Florida International University FIU Digital Commons

Department of Chemistry and Biochemistry College of Arts, Sciences & Education

4-6-2016 Indole Alkaloids of the Stigonematales (Cyanophyta): Chemical Diversity, Biosynthesis and Biological Activity Katherine E. Walton Department of Chemistry and Biochemistry, Florida International University, [email protected]

John P. Berry Department of Chemistry and Biochemistry, Florida International University, [email protected]

Follow this and additional works at: http://digitalcommons.fiu.edu/chemistry_fac Part of the Chemistry Commons

Recommended Citation Walton, K.; Berry, J.P. Indole Alkaloids of the Stigonematales (Cyanophyta): Chemical Diversity, Biosynthesis and Biological Activity. Mar. Drugs 2016, 14, 73.

This work is brought to you for free and open access by the College of Arts, Sciences & Education at FIU Digital Commons. It has been accepted for inclusion in Department of Chemistry and Biochemistry by an authorized administrator of FIU Digital Commons. For more information, please contact [email protected]. marine drugs

Review Indole Alkaloids of the Stigonematales (Cyanophyta): Chemical Diversity, Biosynthesis and Biological Activity

Katherine Walton and John P. Berry *

Department of Chemistry and Biochemistry, Marine Science Program, Florida International University, 3000 NE 151st Street, North Miami, FL 33181, USA; kwalt001@fiu.edu * Correspondence: berryj@fiu.edu; Tel.: +1-305-919-4569; Fax: +1-305-919-4030

Academic Editor: Orazio Taglialatela-Scafati Received: 2 February 2016; Accepted: 31 March 2016; Published: 6 April 2016

Abstract: The cyanobacteria are well recognized as producers of a wide array of bioactive metabolites including toxins, and potential drug candidates. However, a limited number of taxa are generally considered with respect to both of these aspects. That said, the order Stigonematales, although largely overlooked in this regard, has become increasingly recognized as a source of bioactive metabolites relevant to both human and environmental health. In particular, the hapalindoles and related indole alkaloids (i.e., ambiguines, fischerindoles, welwitindolinones) from the order, represent a diverse, and phylogenetically characteristic, class of secondary metabolites with biological activity suggestive of potential as both environmental toxins, and promising drug discovery leads. The present review gives an overview of the chemical diversity of biologically active metabolites from the Stigonematales—and particularly the so-called hapalindole-type alkaloids—including their biosynthetic origins, and their pharmacologically and toxicologically relevant bioactivities. Taken together, the current evidence suggests that these alkaloids, and the associated cyanobacterial taxa from the order, warrant future consideration as both potentially harmful (i.e., “toxic”) algae, and as promising leads for drug discovery.

Keywords: cyanobacteria; blue-green algae; Stigonematales; indole alkaloids; hapalindole; ambiguine; fischerindole; welwitindolinone; toxins; harmful algal blooms

1. Introduction Cyanobacteria are ubiquitous in the environment, but arguably best recognized as the conspicuous “blue-green algae” in freshwater and coastal habitats, and particularly in association with eutrophication and episodic “harmful algal blooms” (HABs). This is reinforced by the recognized diversity of toxic, and more generally bioactive, metabolites by this phylum. Studies of the cyanobacteria have, in particular, focused on the role of several metabolites—the so-called cyanotoxins—with documented toxicity (e.g., human and livestock poisonings, chronic health effects, etc.). In parallel, pharmacological investigations have sought to leverage this chemical diversity in regards to possible drug discovery leads. The chemical diversity of cyanobacteria with respect to both areas is more generally reviewed elsewhere [1–5]. Among studies of bioactive metabolites from cyanobacteria, the order Stigonematales remains relatively underrepresented with regards to both environmental health (i.e., toxin) and biomedical (i.e., drug discovery) aspects. Lack of investigation, with respect to the former, arguably stems from the focus on a limited number of recognized toxic metabolites with established links to HAB events, or human health concerns (e.g., contamination of drinking water), and the thus taxa (i.e., Chroococcales, Nostocales, Oscillatoriales) which produce these compounds. Specifically,

Mar. Drugs 2016, 14, 73; doi:10.3390/md14040073 www.mdpi.com/journal/marinedrugs Mar. Drugs 2016, 14, 73 2 of 28 investigations of the toxic potential of cyanobacteria have consequently focused on well-described neurotoxins (e.g., saxitoxin, anatoxin-a, BMAA) and “hepatotoxic” metabolites (e.g., microcystins, nodularin, cylindrospermopsin), and therefore, the taxa most frequently associated with these compounds (e.g., Microcystis, Nodularia, Cylindrospermopsis, Anabaena, Aphanizomenon, Lyngbya) [3,5].

Similarly,Mar. the Drugs vast 2016 majority, 14, x of investigations of cyanobacterial metabolites explored2 of with 28 respect to potential drug leads have been, likewise, largely biased toward—and presumptively due to the tremendousneurotoxins chemical (e.g., diversity saxitoxin, of—a anatoxin rather‐a, BMAA) small taxonomic and “hepatotoxic” subset metabolites of cyanobacteria, (e.g., microcystins, and most notably nodularin, cylindrospermopsin), and therefore, the taxa most frequently associated with these the polyphyletic genus, Lyngbya (Order Oscillatoriales), and limited number of other genera (e.g., compounds (e.g., Microcystis, Nodularia, Cylindrospermopsis, Anabaena, Aphanizomenon, Lyngbya) [3,5]. Nostoc, SymplocaSimilarly,)[ 1the,2 ,vast4]. majority of investigations of cyanobacterial metabolites explored with respect to That said,potential an drug accumulated leads have knowledgebeen, likewise, of largely the chemistry biased toward—and of the Stigonematales, presumptively due from to the more than three decadestremendous of investigation, chemical diversity suggests of—a rather that the small order taxonomic represents subset of a uniquecyanobacteria, repertoire and most of bioactive notably the polyphyletic genus, Lyngbya (Order Oscillatoriales), and limited number of other genera metabolites with relevance to both drug discovery and environmental health. In particular, a chemically (e.g., Nostoc, Symploca) [1,2,4]. diverse familyThat of bioactivesaid, an accumulated indole alkaloids knowledge is of seemingly the chemistry characteristic of the Stigonematales, and, in from fact, more phylogenetically than restrictedthree to the decades order. of Herein, investigation, we specifically suggests that reviewthe order this represents family a unique of unique repertoire indole of bioactive alkaloids, and to a lesser extentmetabolites other with bioactive relevance metabolites, to both drug from discovery the Stigonematalesand environmental with health. respect In particular, to their a chemical chemically diverse family of bioactive indole alkaloids is seemingly characteristic and, in fact, diversity—including biosynthesis and pharmacology/toxicology—specifically as it relates to their phylogenetically restricted to the order. Herein, we specifically review this family of unique indole potential asalkaloids, drug leads, and to possiblea lesser extent ecological other bioactive function metabolites, and relevance from the Stigonematales to environmental with respect health. to their chemical diversity—including biosynthesis and pharmacology/toxicology—specifically as it relates 2. Biologyto oftheir the potential Stigonematales as drug leads, possible ecological function and relevance to environmental health.

Members2. Biology of the of Stigonematalesthe Stigonematales are classified as heterocystous (i.e., nitrogen-fixing) cyanobacteria, and within thisMembers classification, of the specificallyStigonematales distinguished are classified as as “true heterocystous branching” (i.e. filamentous, nitrogen‐fixing) cyanobacteria (Figure1).cyanobacteria, Heterocysts formand within within this filaments, classification, particularly specifically during distinguished nitrogen as starvation, “true branching” and are able to enzymaticallyfilamentous (via nitrogenases) cyanobacteria (Figure utilize 1). dinitrogen Heterocysts (formi.e.,N within2). Taxa filaments, which particularly produce during heterocysts nitrogen have been shown [6]starvation, to represent and a are monophyletic able to enzymatically group (via (based nitrogenases) on 16s rDNA) utilize dinitrogen within the (i.e. cyanobacteria, N2). Taxa which previously produce heterocysts have been shown [6] to represent a monophyletic group (based on 16s rDNA) referred towithin as the the orders cyanobacteria Nostocales previously and referred Stigonematales, to as the orders but Nostocales classified and by Stigonematales, more recent but taxonomic nomenclatureclassified as Section by more IVrecent and taxonomic V, respectively. nomenclature That as Section said, the IV and Stigonematales V, respectively. That have said, been the shown, in fact, to representStigonematales a polyphyletic have been groupshown, sub-classifiedin fact, to represent as “T-branching” a polyphyletic group and “Y-branching”sub‐classified as “T groups‐ [7,8]. In the currentbranching” review, and however, “Y‐branching” we will groups employ [7,8]. In the the classic current nomenclature review, however, of we Order will employ Stigonematales. the classic nomenclature of Order Stigonematales.

Figure 1. Photomicrograph of Fischerella 52‐1 as a representative of the branched filamentous habit of Figure 1. Photomicrographthe Stigonematales. of Fischerella 52-1 as a representative of the branched filamentous habit of the Stigonematales. In addition to being heterocystous, the Stigonematalean cyanobacteria are also capable of producing hormogonia (motile trichomes) and akinetes (thick‐walled dormant cells) adding, thereby, In additionto their ecological to being success. heterocystous, In fact, the Stigonematales the Stigonematalean represent the most cyanobacteria complex prokaryotes, are also and capable of producingwere hormogonia originally thought (motile to trichomes)be directly ancestral and akinetes to eukaryotes (thick-walled [9]. Moreover, dormant the order cells) represents adding, an thereby, to their ecologicalecologically success. diverse Ingroup fact, with the representatives Stigonematales isolated represent from a thewide most range complex of terrestrial prokaryotes, and and freshwater habitats, including numerous thermophiles, as well as notable epiphytic species. were originally thought to be directly ancestral to eukaryotes [9]. Moreover, the order represents an ecologically diverse group with representatives isolated from a wide range of terrestrial and freshwater habitats, including numerous thermophiles, as well as notable epiphytic species. Mar. Drugs 2016, 14, 73 3 of 28

With respect to bioactive metabolites, a relatively limited number of taxa within the order, however, have been investigated. And, to-date, only six genera representing only three (of the eight described) families of the order, namely Fischerellaceae (Fischerella, Westelliopsis), Mastigocladaceae (Hapalosiphon, Westiella, Mastigocladus) and Stigonemataceae (Stigonema), have been specifically studied with respect to bioactive metabolites. Arguably, this pursuit has been inherently hindered by inabilities to culture representative species of other taxa, as well as lack of clear links to either toxicity or drug potential which might, otherwise, fuel research in this regard. As one noteworthy example, in this regard, a novel species from the order was only recently isolated [10] as an epiphyte of the plant, Hydrophilla verticillata, and linked to intoxication of waterfowl (see Section8, Relevance as Harmful Algae, below). However, the relevant toxic metabolites, in this case, remain to be identified.

3. Stigonematales as a Source of Bioactive Metabolites Although metabolites typically classified as “cyanotoxins” are not generally linked to this family, there have notably been documented instances of members of the order producing known toxins. For example, a freshwater isolate of Fischerella was found [11] to produce the widely distributed hepatotoxin, microcystin-LR (MC-LR). Similarly, MC-LA was identified from a collection of a terrestrial species of Hapalosiphon [12]. And more recently, Fischerella isolated from a freshwater system in Australia was found [13] to produce several microcystin variants (i.e., MC-LR, MC-LA, MC-LF, MC-FR and demethyl-MC-LR). In addition, the neurotoxic amino acid, β-methylamino-L-alanine (BMAA), was recently identified from an epiphytic stigonematalean species [14]. It was also previously found that a freshwater isolate of the species, Umezakia natans, apparently produced the hepatotoxic alkaloid, cylindrospermopsin; however, subsequent taxonomic analysis confirmed that this species belongs, in fact, to the Nostocales rather than the Stigonematales [15]. That said, a wide range of bioactive compounds with both potential environmental (i.e., ecological, environmental toxicology) and biomedical significance have been isolated and characterized from the Stigonematales (Figure2). Interestingly, multiple polychlorinated aromatic compounds, specifically including the polyphenols ambigol A, B and C, 2,4-dichlorobenzoic acid and a carbazole alkaloid, tjipanazole D (Figure2), have been isolated from the order; and, in turn, associated with diverse biological activity including antibacterial, anti-parasitic (i.e., Plasmodium, Trypanosoma), molluscicidal and cytotoxic activity, as well as inhibition of cyclooxygenases (an enzyme responsible for inflammatory responses) and HIV-1 reverse transcriptase [16,17]. The diversity of chlorinated compounds, in particular, is perhaps notable in light of the very recent identification of a halogenase from the order (see Section5, Biosynthesis of Indole Alkaloids from the Stigonematales) which is uniquely capable of chlorinating “freestanding” substrates [18]. More recently, studies of the capsular polysaccharides from the species, Mastigocladus laminosus, identified apparent cytotoxicity, as well as inhibition of cell migration and invasion, in a human carcinoma cell line, and specifically pointed to a role of secreted metalloproteinases in this regard; however, chemical characterization of the polysaccharides remains to be accomplished [19]. Not surprisingly, given the recognized diversity of non-ribosomal peptides from cyanobacteria, several bioactive peptides have been, likewise, identified from the Stigonematales, including several depsipeptides and bistratamides [20–23]. With respect to depsipeptides, hapalosin (Figure2) was isolated from Hapalosiphon welwitschii, and specifically shown to inhibit P-glycoprotein-mediated multidrug resistance (MDR) activity more effectively, in fact, than clinically utilized verapamil [21]. And, more recently, stigonemapeptin (Figure2), specifically containing the unusual amino acid, Ahp, i.e., (3-amino-6-hydroxy-2-piperidone), was isolated from freshwater field samples of the genus, Stigonema, and found to act as a protease (i.e., elastase, chymotrypsin) inhibitor [22]. “Bistratatamide-type” peptides (Figure2) have also been isolated from order. Members of this unusual class of macrocyclic peptides containing thiazole and oxazole rings, and named for the bistratamides originally isolated from a marine worm Lissoclinum bistratum [24], have been isolated, in fact, from other cyanobacterial taxa (e.g., nostocyclamide from Nostoc spp. [25,26]). Representatives of the class from Stigonematales specifically include: (1) the dendroamides, isolated Mar. Drugs 2016, 14, 73 4 of 28

fromMar.Stigonema Drugs 2016 dendroideum, 14, x , and shown—similar to hapalosin—to reverse MDR via P-glycoprotein4 of 28 inhibition [23]; and (2) westiellamide, isolated from Westiellopsis prolifica, and shown to be cytotoxic towardWestiellopsis human epithelial prolifica, and (KB) shown and colon to be (LoVo)cytotoxic cancer toward cell human lines [ 20epithelial]. Finally, (KB) a considerableand colon (LoVo) diversity of alkaloidscancer cell has lines been [20]. isolated Finally, from a considerable the Stigonematales. diversity of Amongalkaloids these, has been the isolated enediyne-containing from the fischerellinsStigonematales. (Figure Among2) were these, isolated the enediyne from Fischerella‐containing muscicola fischerellins, and (Figure characterized 2) were isolated by a wide from range Fischerella muscicola, and characterized by a wide range of bioactivities including antialgal, herbicidal of bioactivities including antialgal, herbicidal and antifungal activity, as well as toxicity towards and antifungal activity, as well as toxicity towards rotifers and crustaceans [27,28]. Of the alkaloids rotifers and crustaceans [27,28]. Of the alkaloids isolated from the order, however, a family of indole isolated from the order, however, a family of indole alkaloids (Figure 3) is clearly the most chemically alkaloidsdiverse, (Figure important3) is and clearly widespread the most group chemically of compounds diverse, [29–45], important and is andthe primary widespread focus of group the of compoundsremainder [29 of–45 this], andreview. is the primary focus of the remainder of this review.

FigureFigure 2. Examples2. Examples of of bioactive bioactive metabolites from from the the Stigonematales. Stigonematales.

4. Indole Alkaloids from the Stigonematales: Chemical Diversity A chemically diverse family of indole alkaloids (Figure3) produced by the Stigonematalean cyanobacteria, with more than 80 variants reported so far [29–45], arguably represent the most characteristic bioactive secondary metabolites from the order. And, in fact, these metabolites have not been found outside of the Stigonematales. Although diverse, these molecules all bear certain structural similarities, likely owing to shared biosynthetic origins (see Section5, Biosynthesis of Indole Alkaloids from the Stigonematales) including: a polycyclic, i.e., tricyclic, tetracyclic or pentacyclic, ring-system, specifically based on an indole core; a pendant vinylic group; and, in almost all cases, a potentially reactive functional group comprised of either isonitrile or isothiocyanate—or, in a relatively few cases, nitrile and thiocarbamate. In addition, site-specific chlorination is frequently observed among the variants. Based on carbon skeletons observed so far, indole alkaloids from the Stigonematales can be effectively classified into nine structural groups (Figure3). Thus far, however, these alkaloids have been only identified from four genera of the Stigonematales (i.e., Fischerella, Hapalosiphon, Westiella, Westiellopsis), but as they span multiple families within the order, it is likely that related metabolites remain to be identified from others.

Mar. Drugs 2016, 14, 73 5 of 28

Mar.Mar. DrugsDrugs 20162016,, 1414,, xx 55 ofof 2828

FigureFigure 3.3. ClassificationClassification schemescheme ofof indoleindole alkaloidsalkaloids fromfrom OrderOrder Stigonematales.Stigonematales. TheThe previouslypreviously Figure 3. Classification scheme of indole alkaloids from Order Stigonematales. The previously describeddescribed indoleindole alkaloidsalkaloids areare classifiedclassified herehere asas ninenine proposedproposed groups;groups; givengiven areare groupgroup numbernumber (as(as described indole alkaloids are classified here as nine proposed groups; given are group number defineddefined here),here), andand correspondingcorresponding chemicalchemical namename ofof eacheach groupgroup (based(based onon previouslypreviously namednamed carboncarbon (as definedskeletons).skeletons). here), NoteNote and thatthat corresponding stereochemistrystereochemistry chemical isis illustratedillustrated name onlyonly of each forfor thosethose group subsub (based‐‐classesclasses on forfor previously whichwhich eithereither named aa veryvery carbon skeletons).limitedlimited Note numbernumber that ofof stereochemistry variantsvariants havehave beenbeen is illustratedidentifiedidentified ((i.e.i.e. only,, GroupsGroups for those 3,3, 66 sub-classesandand 8),8), and/orand/or for nono which variabilityvariability either inin a very limitedstereodispositionstereodisposition number of variants (e.g.,(e.g., 1212‐‐epimers, haveepimers, been CC‐‐10/11/15)10/11/15) identified hashas ( beeni.e.been, Groups yetyet reportedreported 3, 6 ( and(i.e.i.e.,, GroupGroup 8), and/or 55 andand no9).9). ForFor variability moremore in stereodispositiondetaileddetailed explanationexplanation (e.g., 12-epimers, ofof variationvariation withinwithin C-10/11/15) thethe subclasses,subclasses, has been seesee yet FiguresFigures reported 4–11.4–11. (i.e., Group 5 and 9). For more detailed explanation of variation within the subclasses, see Figures4–11. 4.1.4.1. HapalindolesHapalindoles 4.1. HapalindolesOfOf thethe indoleindole alkaloidsalkaloids producedproduced byby thethe Stigonematales,Stigonematales, thethe hapalindoleshapalindoles areare thethe largestlargest andand mostmost widespreadwidespread group,group, andand ostensiblyostensibly representrepresent thethe archetypalarchetypal subsub‐‐classclass ofof thethe indoleindole alkaloidsalkaloids Of the indole alkaloids produced by the Stigonematales, the hapalindoles are the largest and fromfrom thethe orderorder [29,30].[29,30]. InIn fact,fact, biosynthesisbiosynthesis studiesstudies suggestsuggest thatthat hapalindoleshapalindoles areare thethe precursorsprecursors toto mostotherother widespread subsub‐‐classesclasses group, ofof thethe and soso‐ ostensibly‐calledcalled hapalindolehapalindole represent‐‐typetype the alkaloidsalkaloids archetypal fromfrom thethe sub-class Stigonematales.Stigonematales. of the indole FirstFirst alkaloidsidentifiedidentified from the ordermoremore [thanthan29,30 3030]. years Inyears fact, agoago biosynthesis byby MooreMoore etet studies al.al. [29][29] from suggestfrom H.H. intricatusintricatus that hapalindoles,, specificallyspecifically are basedbased the on precursorson aa suggestedsuggested to other sub-classesallelopathicallelopathic of the rolerole so-called (see(see SectionSectionhapalindole-type 7,7, PossiblePossible EcologicalEcological alkaloids from Function),Function), the Stigonematales. aa diversitydiversity ofof hapalindoleshapalindoles First identified havehave more beenbeen than 30 yearssincesince ago isolatedisolated by Moore fromfrom numerousnumerouset al. [29] speciesspecies from H. ofof intricatus Hapalosiphon,Hapalosiphon,, specifically asas wellwell asas based FischerellaFischerella on aandand suggested WestelliopsisWestelliopsis allelopathic.. role (see SectionStructurally,Structurally,7, Possible thethe Ecological hapalindoleshapalindoles Function), areare furtherfurther a diversity subsub‐‐classifiedclassified of hapalindoles intointo tetracyclictetracyclic have (Group(Group been since1)1) andand isolated tricyclictricyclic from numerous(Group(Group species 2)2) variants—withvariants—with of Hapalosiphon, thethe latterlatteras beingbeing well moremore as Fischerella common—and,common—and,and Westelliopsis withinwithin bothboth. ofof thesethese subsub‐‐classes,classes, therethere is considerable variation including both chlorinated and non‐chlorinated (at C‐13; see Figures 4 and 5), Structurally,is considerable the variation hapalindoles including are both further chlorinated sub-classified and non‐chlorinated into tetracyclic (at C‐13; (Groupsee Figures 1) 4 and and tricyclic 5), andand occasionallyoccasionally hydroxylatedhydroxylated (e.g.,(e.g., hapalindolehapalindole O),O), representatives.representatives. InIn somesome rarerare casescases (e.g.,(e.g., (Group 2) variants—with the latter being more common—and, within both of these sub-classes, there is hapalindolehapalindole NN andand P),P), thethe vinylicvinylic groupgroup isis replacedreplaced byby anan epoxide.epoxide. Moreover,Moreover, allall membersmembers ofof thethe subsub‐‐ considerable variation including both chlorinated and non-chlorinated (at C-13; see Figures4 and5), and classclass and,and, moremore generally,generally, nearlynearly allall membersmembers ofof thethe largerlarger classclass identifiedidentified soso far,far, containcontain reactivereactive occasionally hydroxylated (e.g., hapalindole O), representatives. In some rare cases (e.g., hapalindole

N and P), the vinylic group is replaced by an epoxide. Moreover, all members of the sub-class and, more generally, nearly all members of the larger class identified so far, contain reactive functional Mar. Drugs 2016, 14, 73 6 of 28

groupsMar. Drugs at the 2016 C-11, 14, x (Figures4 and5) position; in this regard, isonitrile- and isothiocyanate-bearing6 of 28 hapalindolesMar. Drugs are,2016, 14 by, x far, the most common, but thiocarbamates (e.g., hapalindole T) [31] have6 of 28 also beenmost identified. notably including Finally, within epimers the at group, the C‐12 there position, is variability and the withdiastereomers respect to at stereochemistry the C‐10 and C‐ most15 notablypositionsmost including notably (Figure including epimers 4). Chemical atepimers the C-12variants at the position, Cof‐12 the position, and hapalindoles the and diastereomers the (and,diastereomers indeed, at the C-10atother the and Csub‐10‐ C-15 classes)and positionsC‐15 are generally named alphabetically (i.e., hapalindole A, B, C, etc.) in order of their identification with (Figurepositions4). Chemical (Figure variants 4). Chemical of the hapalindolesvariants of the (and, hapalindoles indeed, other(and, sub-classes)indeed, other are sub generally‐classes) are named most of the core structures (hapalindoles A, B, C–Q and T–V) notably identified by Moore et al. [29,30]; alphabeticallygenerally named (i.e., hapalindole alphabetically A, (i.e. B,, C, hapalindoleetc.) in order A, B, ofC, theiretc.) in identification order of their withidentification most of with the core however, additional nomenclature (e.g., 12‐epi‐hapalindoles, deschloro hapalindoles) are also structuresmost of (hapalindoles the core structures A, B, (hapalindoles C–Q and T–V)A, B, C–Q notably and T–V) identified notably byidentified Moore byet Moore al. [29 et, 30al. ];[29,30]; however, frequentlyhowever, found additional in the literaturenomenclature to further (e.g., elaborate 12‐epi‐hapalindoles, structural variants. deschloro As hapalindoles)an example, hapalindole are also additional nomenclature (e.g., 12-epi-hapalindoles, deschloro hapalindoles) are also frequently found A frequentlyis a tetracyclic, found in chlorinated the literature isonitrileto further‐ containingelaborate structural congener, variants. whereas As an hapalindoleexample, hapalindole B is the in the literature to further elaborate structural variants. As an example, hapalindole A is a tetracyclic, correspondingA is a tetracyclic, isothiocyanate. chlorinated Hapalindole isonitrile ‐Jcontaining and M, in congener,turn, are thewhereas non‐chlorinated hapalindole isonitrile B is the and chlorinatedisothiocyanatecorresponding isonitrile-containing congeners, isothiocyanate. respectively congener, Hapalindole (Figure whereas J and 4), M,whereas hapalindole in turn, 12 are‐epi B the‐ ishapalindole the non corresponding‐chlorinated J has beenisonitrile isothiocyanate. identified and Hapalindole[32]isothiocyanate as an epimer J and congeners,of M, the in former. turn, respectively are the (Figure non-chlorinated 4), whereas isonitrile12‐epi‐hapalindole and isothiocyanate J has been identified congeners, respectively[32] as an (Figure epimer4), of whereas the former. 12- epi-hapalindole J has been identified [32] as an epimer of the former.

FigureFigureFigure 4. 4.Previously Previously4. Previously described described described tetracyclic tetracyclic tetracyclic hapalindoles hapalindoles (Group(Group (Group 1) 1) 1) from from from Stigonematales. Stigonematales. Stigonematales.

Figure 5. Previously described tricyclic hapalindoles (Group 2) from Stigonematales. FigureFigure 5. 5.Previously Previously described described tricyclictricyclic hapalindoles (Group (Group 2) 2) from from Stigonematales. Stigonematales. Notably alongside hapalindoles from H. fontinalis, Moore et al. [30,33] isolated several more Notablypolar,Notably minor alongside alongside constituents, hapalindoles hapalindoles namely fromfontonamides, fromH. H. fontinalis fontinalis hapalonamides, Moore, Mooreet al.et and [al.30 ,hapaloxindoles[30,33]33] isolated isolated several (Figure several more 6). more It polar, minorpolar,was constituents, minor experimentally constituents, namely shown namely fontonamides, [30] that fontonamides, these congeners hapalonamides hapalonamides were likely and oxidation hapaloxindolesand hapaloxindoles products of (Figure hapalindoles, (Figure6). It6). was It experimentallywasand experimentally specifically shown hapalindole shown [30] that [30] A, these that formed these congeners by congeners singlet were oxygen were likely withlikely oxidation hapaloxindole oxidation products products serving of hapalindoles, ofas hapalindoles,a precursor and specificallyandfor specifically fontonamide hapalindole hapalindole and A,hapalonamide, formed A, formed by singletwhereby by singlet oxygen subsequent oxygen with with oxidation hapaloxindole hapaloxindole leads to serving opening serving as of as athe a precursor precursor indole‐ for ring in the latter (Figure 6). These oxidation products are seemingly widespread as they have also fontonamidefor fontonamide and hapalonamide, and hapalonamide, whereby whereby subsequent subsequent oxidation oxidation leads leads to opening to opening of the of indole-ring the indole‐ in ringrecently in the latterbeen found(Figure [34], 6). alongsideThese oxidation hapalindoles, products from are Fischerella seemingly. Moreover, widespread an oxindoleas they have ring alsois the latter (Figure6). These oxidation products are seemingly widespread as they have also recently recentlysimilarly been found found in other[34], alongsidesub‐classes hapalindoles, including hapalindolinones from Fischerella and. Moreover, welwitindolinones an oxindole (Figure ring 3) is been found [34], alongside hapalindoles, from Fischerella. Moreover, an oxindole ring is similarly found similarlysuggesting found a similar in other role sub of ‐oxidationclasses including in the observed hapalindolinones chemical diversity and welwitindolinones of these groups. (Figure 3) in other sub-classes including hapalindolinones and welwitindolinones (Figure3) suggesting a similar suggesting a similar role of oxidation in the observed chemical diversity of these groups. role of oxidation in the observed chemical diversity of these groups.

Mar.Mar. Drugs Drugs2016 2016, 14, 7314, x 7 of7 28 of 28 Mar. Drugs 2016, 14, x 7 of 28

Figure 6. Oxidized hapalindoles isolated from Stigonematales. Figure 6. Oxidized hapalindoles isolated from Stigonematales. 4.2. Hapalindolinones 4.2. Hapalindolinones The hapalindolinones (Group 3) represent a unique structural sub‐class containing a spiro‐fused cyclopropaneThe hapalindolinones ring, and oxindole (Group 3) ring. represent Hapalindolinones a unique structural A and sub sub-classB (chlorinated‐class containing and non a spiro spiro-fused‐chlorinated,‐fused cyclopropanerespectively; ring, see Figure and oxindole 3) were, ring. in fact, Hapalindolinones originally identified A and from B (chlorinated a species of and Fischerella non non-chlorinated,‐chlorinated, isolated from respectively;a Florida Everglades see see Figure Figure soil3)3) were, were, sample in in fact, [35]. fact, originally It originally is tempting identified identified to speculate from from a thatspecies a species hapalindolinones of Fischerella of Fischerella isolated mayisolated represent from froma Floridaderivatives a Florida Everglades of Everglades the soil tricyclic sample soil hapalindoles, sample [35]. It is [35 tempting]. Itspecifically is tempting to speculate formed to speculate that via hapalindolinones cyclization that hapalindolinones of C may‐11 torepresent yield may the representderivativescyclopropane derivatives of the ring, tricyclic of and the tricyclicsinglethapalindoles, oxygen hapalindoles, specifically oxidation specifically (asformed suggested formed via cyclization for via cyclizationhapaloxindole of C‐11 of C-11to derived yield to yield thefrom thecyclopropanetetracyclic cyclopropane hapalindolering, ring, and and singlet A, singlet see oxygenabove) oxygen tooxidation oxidationform the (as oxindole (as suggested suggested ring. for forHowever, hapaloxindole hapaloxindole the biosynthesis derived derived from of this tetracyclicsubclass hapalindole has yet to beA, investigated,see above) to and form the thethe hapalindoles oxindoleoxindole ring.ring. are However, classified the separately biosynthesis here, of in this part, subclassbecause hashas of yet yettheir to to be uniquebe investigated, investigated, biological and andactivity the hapalindolesthe (Sectionhapalindoles 6, are Biological classified are classified Activity separately separately of here, the Hapalindole in here, part, becausein part,‐Type ofbecause theirAlkaloids). unique of their biological unique biological activity (Section activity6, Biological(Section 6, Activity Biological of theActivity Hapalindole-Type of the Hapalindole Alkaloids).‐Type Alkaloids). 4.3.4.3. Ambiguines Ambiguines 4.3. Ambiguines TheThe second second largest largest group group of indoleof indole alkaloids alkaloids of Stigonematalesof Stigonematales is theis the ambiguines. ambiguines. Members Members of of thisthisThe class class second are are specificallyspecifically largest group distinguished distinguished of indole alkaloids by by an an isoprene of isoprene Stigonematales unit unit attached attached is atthe the ambiguines. at C‐ the2 position C-2 Members position of the indole of thethisring indoleclass [36–38] are ring specifically including [36–38] includingdistinguishedtetracyclic tetracyclicand by pentacyclic an isoprene and pentacyclicvariants unit attached with variants the at latter the with Cformed‐2 position the by latter cyclization of formed the indole of by the cyclizationringisoprene [36–38] ofvia including the an isopreneepoxide tetracyclic bridge via an and (Figures epoxide pentacyclic 7 bridge and 8, variants (Figures respectively), with7 and the8 although respectively), latter formed hydroxyl, although by cyclization carbonyl hydroxyl, and of the even carbonylisoprenenitrile viagroups and an even epoxide may nitrile replace bridge groups the (Figures epoxide. may 7 and replace Also 8, respectively),similar the epoxide. to the althoughhapalindoles, Also similarhydroxyl, both to chlorinatedcarbonyl the hapalindoles, and and even non ‐ bothnitrilechlorinated chlorinated groups maycongeners and replace non-chlorinated are the common. epoxide. In congenersAlso addition, similar arehydroxyl to common. the hapalindoles,‐containing In addition, variants—spe both chlorinated hydroxyl-containingcifically and at non the‐ C‐ variants—specificallychlorinated10 (Figure congeners 8)—of at theare the common.ambiguines C-10 (Figure In addition,8are)—of quite the hydroxyl ambiguinescommon.‐containing Unlike are quite hapalindoles, variants—spe common. Unlike cificallyno hapalindoles,isothiocyanate at the C‐ ‐ no10 containing isothiocyanate-containing(Figure 8)—of variants the areambiguines known, variants however, are are quite known, interestingly common. however, Unlikenitrile interestingly‐ bearinghapalindoles, nitrile-bearing congeners no ofisothiocyanate both congeners tetracyclic of‐ bothcontaining[39] tetracyclic and variants pentacyclic [39] andare pentacyclic known,(i.e., ambiguine however, (i.e., ambiguine Ginterestingly [40]) Gambiguines [40 ])nitrile ambiguines‐bearing have havebeen congeners been identified. identified. of both Moreover, tetracyclic Moreover, like like[39]hapalindoles, hapalindoles,and pentacyclic 12 12-epimers‐epimers (i.e., ambiguine exist. exist. Specifically, Specifically, G [40]) WaltonWalton ambiguineset et al. al.[ 39 [39]have] recently recently been identified identified.identified a 12- aMoreover, 12epi‐epi-ambiguine‐ambiguine like Bhapalindoles, nitrileB nitrile from from 12 an ‐anepimers isolate isolate of exist. Fischerellaof Fischerella Specifically,, and, and subsequent Walton subsequent et al. studies [39]studies recently have, have, likewise, identified likewise, identified identifieda 12‐epi‐ theambiguine the related, related, non-chlorinatedB nitrilenon‐chlorinated from an congeners isolate congeners of (e.g.,Fischerella (e.g., 12- 12epi, ‐-ambiguineandepi‐ambiguine subsequent C nitrile C studiesnitrile [41 [41])]) have, from from likewise, the the same same identified strain. strain. the related, non‐chlorinated congeners (e.g., 12‐epi‐ambiguine C nitrile [41]) from the same strain.

FigureFigure 7. Previously7. Previously described described tetracyclic tetracyclic ambiguines ambiguines (Group (Group 4) from 4) from Stigonematales. Stigonematales. Figure 7. Previously described tetracyclic ambiguines (Group 4) from Stigonematales.

Mar. Drugs 2016, 14, 73 8 of 28 Mar. Drugs 2016, 14, x 8 of 28

Mar. Drugs 2016, 14, x 8 of 28

FigureFigure 8. 8.Previously Previously described described pentacyclic pentacyclic ambiguines (Group (Group 5) 5) from from Stigonematales. Stigonematales. Figure 8. Previously described pentacyclic ambiguines (Group 5) from Stigonematales. TheThe ambiguines ambiguines were were first first isolated isolated byby SmitkaSmitka et al. [36][36] who who identified identified six six variants, variants, ambiguines ambiguines A–FA–F isonitriles, Theisonitriles, ambiguines including including were both firstboth tetra- isolated tetra and‐ and pentacyclicby Smitkapentacyclic et structures, al. structures,[36] who from identified from three three separate six variants,separate genera ambiguinesgenera and speciesand species (A–F from F. ambigua, A and E from H. hibernicus and D and E from W. prolifica). Subsequently, (A–FA–F from isonitriles,F. ambigua, includingA and both E fromtetra‐ H.and hibernicus pentacyclicand structures, D and E from from threeW. prolifica separate). Subsequently,genera and Huber et al. [40] isolated ambiguine G (from H. delicatulus), and specifically identified it as a novel Huberspecieset al.(A–F[40 from] isolated F. ambigua, ambiguine A and E G from (from H.H. hibernicus delicatulus and), D and and specificallyE from W. prolifica identified). Subsequently, it as a novel nitrile‐bearing variant. Identification of additional variants, i.e., ambiguines H–Q, followed [37,38,42]. nitrile-bearingHuber et al. [40] variant. isolated Identification ambiguine ofG additional(from H. delicatulus variants,),i.e. and, ambiguines specifically H–Q,identified followed it as a [37 novel,38,42 ]. Notable among these, ambiguine P isolated from F. ambigua represents an unusual variants lacking Notablenitrile‐ amongbearing these, variant. ambiguine Identification P isolated of additional from F. variants, ambigua i.e.represents, ambiguines an unusual H–Q, followed variants [37,38,42]. lacking the the typical isonitrile, or less common nitrile, found among ambiguines (or alternative functional typicalNotable isonitrile, among orthese, less ambiguine common nitrile, P isolated found from among F. ambigua ambiguines represents (or an alternative unusual variants functional lacking group, thegroup, typical e.g., isonitrile,isothiocyanate, or less thiocarbamate, common nitrile, present found in otheramong indole ambiguines alkaloids (or from alternative the order). functional e.g., isothiocyanate, thiocarbamate, present in other indole alkaloids from the order). group, e.g., isothiocyanate, thiocarbamate, present in other indole alkaloids from the order). 4.4. Fischambiguines 4.4. Fischambiguines 4.4. FischambiguinesClearly related to, but classified separately here from, the ambiguines are fishambiguines. This Clearly related to, but classified separately here from, the ambiguines are fishambiguines. pentacyclicClearly sub related‐class to, is butdistinguished classified separately by the presence here from, of a sixthe‐ memberedambiguines (rather are fishambiguines. than 7‐membered) This This pentacyclic sub-class is distinguished by the presence of a six-membered (rather than 7-membered) pentacycliccyclization ofsub the‐class C‐2 isisoprene distinguished group (Figureby the presence9). Fischambiguines of a six‐membered A and B (rather were first than isolated, 7‐membered) rather cyclization of the C-2 isoprene group (Figure9). Fischambiguines A and B were first isolated, rather cyclizationrecently, from of the F. ambiguaC‐2 isoprene by Mo group et al. (Figure[42], specifically 9). Fischambiguines based on weak A and antibacterial B were first activity. isolated, And, rather to recently, from F. ambigua by Mo et al. [42], specifically based on weak antibacterial activity. And, recently,date, only from two F. variants ambigua (Aby andMo etB) al. have [42], beenspecifically characterized, based on and weak specifically antibacterial differ activity. based And, on the to to date, only two variants (A and B) have2 been characterized, and specifically differ based on the date,presence only of two either variants an exomethylene (A and B) (=CH have ) beencarbon characterized, or epoxide, respectively and specifically (Figure differ 9). Similar based toon other the presencesub‐classes, of either the an fischambiguines exomethylene (=CHinclude) carbon both orchlorinated epoxide, respectively(e.g., fischambiguine (Figure9). SimilarB) and tonon other‐ presence of either an exomethylene (=CH22) carbon or epoxide, respectively (Figure 9). Similar to other sub-classes,subchlorinated‐classes, the (e.g.,the fischambiguines fischambiguinefischambiguines include A) include congeners. both both chlorinated chlorinated (e.g., fischambiguine(e.g., fischambiguine B) and non-chlorinatedB) and non‐ (e.g.,chlorinated fischambiguine (e.g., fischambiguine A) congeners. A) congeners.

Figure 9. Previously described fischambiguines (Group 6) from Stigonematales. FigureFigure 9. 9.Previously Previously described described fischambiguinesfischambiguines (Group (Group 6) 6) from from Stigonematales. Stigonematales.

Mar. Drugs 2016, 14, 73 9 of 28 Mar. Drugs 2016, 14, x 9 of 28

4.5. Fischerindoles Fischerindoles, firstfirst isolated isolated from fromF. muscicola F. muscicolaby Park byet Park al. [43 et], areal. tetracyclic[43], are tetracyclic alkaloids considered alkaloids toconsidered be chemically to be relatedchemically to the related hapalindoles, to the hapalindoles, but specifically but specifically distinguished distinguished by ring-fusion by ring at the‐fusion C-2 (ratherat the C than‐2 (rather C-6) ofthan the C indole‐6) of the ring indole forming ring a forming typically a five-membered typically five‐membered C-ring (Figure C‐ring 10 (Figure), although 10), exceptionsalthough exceptions to this have to beenthis reportedhave been (see reported below). (see Similar, below). to the Similar, hapalindoles, to the thehapalindoles, fischerindoles the includefischerindoles both isonitrile- include both or isothiocyanate-containing isonitrile‐ or isothiocyanate (at C-11,‐containing see Figure (at C10‐11,), as see well Figure as chlorinated 10), as well and as non-chlorinatedchlorinated and congenersnon‐chlorinated (at C-13, congeners see Figure (at 10C‐),13, and see all Figure of the 10), same and stereocenters all of the same (i.e., stereocenters C-10/11 and 12-epimers)(i.e., C‐10/11 found and 12 among‐epimers) the found hapalindoles among are, the likewise,hapalindoles found are, in likewise, the fischerindoles. found in the Most fischerindoles. interestingly perhaps,Most interestingly this includes perhaps, 12-epimers this includes as the 12 apparently‐epimers as most the commonapparently stereoisomer. most common As stereoisomer. an example, fischerindoleAs an example, L, fischerindole the first representative L, the first isolated, representative was named isolated, (similar was tonamed all subsequent (similar to variants) all subsequent on the basisvariants) of similarity on the basis to the of similarity stereochemistry to the stereochemistry of hapalindole L of (a hapalindole 12-epimeric L variant) (a 12‐epimeric that was variant) isolated that in thewas same isolated study in [the43]. same And notably,study [43]. in fact,And hapalindole notably, in Lfact, and hapalindole fischerindole L Land have fischerindole similar antifungal L have propertiessimilar antifungal (see Section properties6, Biological (see Section Activity 6, Biological of the Hapalindole-Type Activity of the Hapalindole Alkaloids).‐Type Interestingly, Alkaloids in). additionInterestingly, to the in typical addition isonitrile to the and typical isothiocyanate isonitrile variants, and isothiocyanate nitrile-bearing fischerindolesvariants, nitrile have‐bearing been identifiedfischerindoles [44]. have Moreover, been identified variants (i.e. [44]., fischerindole Moreover, variants W) characterized (i.e., fischerindole by a six-membered W) characterized (rather thanby a 5-membered)six‐membered aromatic (rather than C-ring 5‐membered) have been aromatic recently C described‐ring have [44 been]; however, recently unlikedescribed other [44]; sub-classes however, andunlike variants, other sub no‐classes biological and activityvariants, has no beenbiological reported, activity and has little been is, reported, otherwise, and known little is, about otherwise, these unusualknown about congeners. these unusual congeners.

Figure 10. Previously described fischerindoles fischerindoles from Stigonematales.

4.6. Welwitindolinones Finally, oneone ofof thethe mostmost chemicallychemically uniqueunique sub-classessub‐classes are the tetracyclictetracyclic indolinone-containingindolinone‐containing welwitindolinones (Figures 33 andand 11 11).). FirstFirst isolatedisolated fromfrom Westiella intricata and H. welwitschii by Stratmann et al. [45],[45], the group is further sub sub-categorized‐categorized into Type A and B subclasses with the latter being thethe most most common. common. In fact,In fact, the formerthe former currently currently includes includes only the only unusual the cyclobutane-containingunusual cyclobutane‐ welwitindolinonecontaining welwitindolinone A. The remaining A. The welwitindolinones remaining welwitindolinones (B–D) are classified (B–D) asare Type classified B (Figure as 11Type) that B represent(Figure 11) considerable that represent structural considerable variability; structural however, variability; it has been however, suggested it has that been all welwitindolinones suggested that all arewelwitindolinones biosynthesized are from biosynthesized the cyclobutane-containing from the cyclobutane welwitindolinone‐containing A welwitindolinone as a common precursor. A as a Interestingly,common precursor. all of theInterestingly, welwitindolinones, all of the welwitindolinones, with the exception with of the the unusual exception and of highlythe unusual oxidized and welwitindolinonehighly oxidized welwitindolinone D (see below), are D chlorinated,(see below), andare chlorinated, similar to other and sub-classes,similar to other all known sub‐classes, variants all haveknown either variants an isonitrile have either or an an isothiocyanate isonitrile or an (at isothiocyanate C-11; see Figure (at 11 C).‐11; More see notably,Figure 11). multiple More variants,notably, specificallymultiple variants, of welwitindolinone specifically of C, welwitindolinone are characterized byC, N-methylationare characterized of theby indoleN‐methylation ring (Figure of 11the). Owingindole ring to the (Figure remarkable 11). Owing scaffold to the of theremarkable welwitindolinones, scaffold of the as wellwelwitindolinones, as their emerging as well potential as their as anticanceremerging potential drugs (see as Section anticancer6, Biological drugs (see Activity Section of the6, Biological Hapalindole-Type Activity of Alkaloids), the Hapalindole considerable‐Type effortAlkaloids), has been considerable placed on theeffort enantiospecific has been placed total synthesis on the enantiospecific of these compounds total [46synthesis,47]. of these compoundsSimilar [46,47]. to the hapalindoles, several oxidized products of welwitindolinones have been isolatedSimilar [48]. to These the hapalindoles, include the hydroxylated several oxidized 3-hydroxy- productsN-methylwelwitindolinone of welwitindolinones have C isothiocyanate been isolated and[48]. isonitrile,These include and thethe unusual,hydroxylated non-chlorinated 3‐hydroxy‐NN‐methylwelwitindolinone-methylwelwitindolinone C D isothiocyanate isonitrile. It wasand isonitrile, and the unusual, non‐chlorinated N‐methylwelwitindolinone D isonitrile. It was specifically proposed that these variants were derived from photooxidation of the previously isolated N‐methylwelwitindoline C isonitrile [48].

Mar. Drugs 2016, 14, 73 10 of 28 specifically proposed that these variants were derived from photooxidation of the previously isolated

N-methylwelwitindolineMar. Drugs 2016, 14, x C isonitrile [48]. 10 of 28

Figure 11. Previously described Type B welwitindolinones (Group 9) from Stigonematales. Figure 11. Previously described Type B welwitindolinones (Group 9) from Stigonematales. 5. Biosynthesis of Indole Alkaloids from the Stigonematales 5. Biosynthesis of Indole Alkaloids from the Stigonematales In support of the impressive chemical diversity observed, a continuing body of knowledge has Inevolved support with of respect the impressive to the biosynthesis chemical of the diversity hapalindole observed,‐type alkaloids. a continuing Species (and body strains of knowledge within has evolvedspecies) of with the respectStigonematales to the biosynthesisfrequently produce of the several hapalindole-type of the indole alkaloid alkaloids. sub Species‐classes, (andand the strains withinobserved species) isolation of the Stigonematalesof diverse congeners frequently from a single produce source several leads ofto thethe conclusion indole alkaloid that a sub-classes,similar and thebiosynthetic observed pathway isolation occurs of diversefor these congenerscompounds. from Indeed, a single a unified source biosynthesis leads toof the hapalindole conclusion‐ that type alkaloids was accordingly proposed soon after their discovery [45]. However, very recent a similar biosynthetic pathway occurs for these compounds. Indeed, a unified biosynthesis of the genomic data have led, in many cases, to considerable revision of this proposed biogenetic pathway. hapalindole-type alkaloids was accordingly proposed soon after their discovery [45]. However, very In the first assessments of the biosynthesis of the hapalindole alkaloids, Stratmann et al. [45] recentspecifically genomic data suggested have led, that ininitial many steps cases, occurred to considerable (Figure 12) revision by way of of a this chloronium proposed or biogenetic hydrogen ion pathway.‐ Ininduced the first condensation assessments of of3‐ the(Z‐2′‐ biosynthesisisocyanoethenyl)indole of the hapalindole (i.e., “indole alkaloids,‐isonitrile”) Stratmann and (Z)et‐3,7 al.‐ [45] specificallydimethyl suggested‐1,3,6‐octatriene that (i.e. initial, β‐ocimene), steps whereby occurred these (Figure precursors 12) are by formed, way ofrespectively, a chloronium from or hydrogentryptophan ion-induced and geranyl condensation pyrophosphate of 3-( (GPP),Z-21-isocyanoethenyl)indole leading to the formation of (i.e. a ,presumptive “indole-isonitrile”) tricyclic and (Z)-3,7-dimethyl-1,3,6-octatrienehapalindole. Based on this biosynthetic (i.e., β-ocimene), scheme, 12 whereby‐epimers of these the hapalindoles precursors are were formed, proposed respectively, to be fromalternatively tryptophan andderived geranyl from pyrophosphate either E‐ or Z‐ (GPP),isomers leading of β‐ocimene to the formation (see [49]). ofVery a presumptive recent studies, tricyclic hapalindole.incorporating Based genomic on this and biosynthetic other molecular scheme, approaches, 12-epimers however, of the clearly hapalindoles suggest werean alternative proposed to biosynthetic route, and specifically direct prenylation of the indole‐isonitrile by GPP, rather than be alternatively derived from either E- or Z-isomers of β-ocimene (see [49]). Very recent studies, involvement of β‐ocimene (as discussed below). Similarly, the presence of a cis and trans isomers (i.e., incorporating genomic and other molecular approaches, however, clearly suggest an alternative Z‐ and E‐isomers, respectively) of the indole‐isonitrile were proposed to underlie the stereochemistry biosyntheticat C‐10/11 route, in hapalindoles; and specifically however, direct very recent prenylation genomic of studies the indole-isonitrile suggest that only bythe GPP,Z‐isomer rather is a than involvementsubstrate of forβ the-ocimene condensation (as discussed step (also below). discussed Similarly, below). the presence of a cis and trans isomers (i.e., Z- and E-isomers,With respect respectively) to the biosynthesis of the indole-isonitrile of the indole‐isonitrile, were it proposed was proposed to underlie in early the studies stereochemistry that the at C-10/11isonitrile in functional hapalindoles; group however, found in this very precursor, recent genomic and consequently, studies suggest the majority that of only the sub the‐classes,Z-isomer is a substrateis either for derived the condensation from inorganic step cyanide, (also or discussed cyanide produced below). from C‐2/N‐2 of amino acids [50]—as Withreported, respect in fact, to the for biosynthesiscyanobacteria of [51]—by the indole-isonitrile, way of the tetrahydrofolate it was proposed (THF) in C1 early pathway, studies and that the isonitrilespecifically functional a 5‐formimino group found‐THF in intermediate. this precursor, Contribution and consequently, of amino the acid majority‐derived ofC theand sub-classes, N to the is isonitrile is supported by studies of hapalindole A, and specifically feeding studies of H. fontinalis either derived from inorganic cyanide, or cyanide produced from C-2/N-2 of amino acids [50]—as utilizing isotopically (i.e., 13C, 15N) labeled glycine [52]. The incorporation of inorganic cyanide is reported, in fact, for cyanobacteria [51]—by way of the tetrahydrofolate (THF) C1 pathway, and indirectly supported by feeding studies of sponge‐derived, isonitrile‐containing compounds [53], as specificallywell as aradiolabeling 5-formimino-THF (i.e., 14C‐labeled intermediate. cyanide) Contribution studies in the ofcyanobacteria. amino acid-derived However, attempts C and N to to the isonitrilereplicate is supported the latter feeding by studies studies of with hapalindole inorganic A,cyanide and specificallylabeled with feedingstable isotope studies (i.e. of, 13C)H. were fontinalis utilizingconfounded, isotopically as the (i.e. requisite, 13C, 15 N)concentrations labeled glycine proved [52 ].toxic The to incorporation the cyanobacterial of inorganic cultures cyanide [52]. is indirectlyMoreover, supported recent genomic by feeding studies studies [54] suggest of sponge-derived, an alternative isonitrile-containing biogenic route for the indole compounds‐isonitrile. [53 ], as well asIdentification radiolabeling of a gene (i.e., cluster14C-labeled (amb) associated cyanide) with studies ambiguine in the biosynthesis cyanobacteria. revealed However, three homologs attempts to replicate(i.e., theAmbI1 latter‐3) feedingof previously studies characterized, with inorganic bacterial cyanide isonitrile labeled synthase with stable genes, isotope and (followingi.e., 13C) were confounded,purification as the of the requisite gene products—and concentrations in provedvitro enzyme toxic studies—demonstrated to the cyanobacterial cultures specific production [52]. Moreover,

Mar. Drugs 2016, 14, 73 11 of 28 recent genomic studies [54] suggest an alternative biogenic route for the indole-isonitrile. Identification of a gene cluster (amb) associated with ambiguine biosynthesis revealed three homologs (i.e., AmbI1-3) of previously characterized, bacterial isonitrile synthase genes, and following purification of the gene products—and in vitro enzyme studies—demonstrated specific production of the cis-isomer (and, Mar. Drugs 2016, 14, x 11 of 28 notably,Mar. not Drugs the 2016trans, 14, xform) of the indole-isonitrile from tryptophan and ribulose-5-phosphate11 of 28 (as a carbonof the source) cis‐isomer as shown (and, notably, in Figure not 13the[ 54trans]. Similarform) of resultsthe indole were‐isonitrile found from for homologstryptophan fromand the wel gene clusterribuloseof the associated‐ 5cis‐phosphate‐isomer (and, with (as anotably, biosynthesiscarbon source)not the of as trans welwitindolinonesshown form) in Figure of the 13 indole [54]. [49 ‐Similarisonitrile]. These results from studies, were tryptophan therefore,found for and support an alternativehomologsribulose ‐from route5‐phosphate the which wel (asgene does a carboncluster not source)associated utilized as shown cyanide.with biosynthesis in Figure Specifically, 13 of[54]. welwitindolinones Similar in this results proposed were [49]. found These pathway, for the studies,homologs therefore, from thesupport wel gene an alternative cluster associated route which with does biosynthesis not utilized of welwitindolinonescyanide. Specifically, [49]. in Thesethis nitrogenproposedstudies, of the pathway,therefore, isonitrile the support nitrogen is derived, an alternativeof the instead, isonitrile route fromis which derived, tryptophan, does instead, not utilized from and tryptophan, cyanide. the carbon Specifically, and fromthe carbon in a donorthis (i.e., ribulose-5-phosphatefromproposed a donor pathway, (i.e. in ,the feedingribulose nitrogen studies);‐5 ‐ofphosphate the isonitrile and in subsequent isfeeding derived, studies);instead, oxidative fromand decarboxylation, tryptophan, subsequent and oxidative the in carbon turn, yields the indole-isonitriledecarboxylation,from a donor intermediatein turn,(i.e., yieldsribulose the (Figure‐ 5indole‐phosphate 13‐isonitrile). in intermediatefeeding studies); (Figure 13).and subsequent oxidative decarboxylation, in turn, yields the indole‐isonitrile intermediate (Figure 13).

FigureFigure 12. 12.Originally Originally proposed biosynthesis biosynthesis of a tricyclic of a tricyclic hapalindole, hapalindole, as an intermediate as an for intermediate other for hapalindoleFigure 12.‐ typeOriginally alkaloids proposed (see Figure biosynthesis 14), via chloroniumof a tricyclic or hapalindole, H+ induced ascondensation an intermediate [45]. Recentfor other other hapalindole-type alkaloids (see Figure 14), via chloronium or H+ induced condensation [45]. molecularhapalindole studies,‐type however, alkaloids suggest (see Figure an alternative 14), via chloronium biosynthetic or route H+ induced [55]. condensation [45]. Recent Recent molecular studies, studies, however, however, suggest suggest an alternative an alternative biosynthetic biosynthetic route [55]. route [55].

Figure 13. Biosynthesis of the indole‐isonitrile precursor as inferred based on the identification of the FigureWelI1 13.Figure‐Biosynthesis3 and 13. BiosynthesisAmbI1‐3 of genes the of the indole-isonitrileas homologsindole‐isonitrile of previously precursorprecursor characterized as as inferred inferred basedisonitrile based on thesynthase on identification the identification genes, andof the of the subsequent in vitro studies of gene products [49,54]. WelI1-3WelI1 and‐ AmbI1-33 and AmbI1 genes‐3 genes as homologs as homologs of of previously previously characterized characterized isonitrile isonitrile synthase synthase genes, genes,and and subsequent in vitro studies of gene products [49,54]. subsequentThe originin vitro of thestudies isothiocyanate of gene products found in [ many49,54]. of the variants, on the other hand, has yet to be investigated.The origin Several of the feeding isothiocyanate studies found of isonitrile in many‐ ofand the variants,isothiocyanate on the‐ containingother hand, terpenoidshas yet to be Theproducedinvestigated. origin by ofmarine Several the sponges isothiocyanate feeding generally studies suggest found of isonitrile inter in many‐conversion‐ and of isothiocyanate the between variants, cyanide‐containing on and the thiocyanate, otherterpenoids hand, has yet toandproduced be subsequent investigated. by marine incorporation sponges Several generallyto feedingproduce suggest studiesthe intercorresponding‐conversion of isonitrile- isonitrilesbetween and cyanide and isothiocyanate-containing isothiocyanates, and thiocyanate, respectivelyand subsequent [53]. However,incorporation the toaforementioned produce the correspondinggenomic studies isonitriles [54] which and argueisothiocyanates, against terpenoids produced by marine sponges generally suggest inter-conversion between cyanide incorporationrespectively of[53]. cyanide, However, and ratherthe aforementioned donation of carbon genomic to the studies amine of[54] tryptophan, which argue would, against in and thiocyanate,general,incorporation conflict and withof cyanide, this subsequent proposal. and rather An incorporation alternative donation hypothesisof carbon to produce proposesto the amine the the correspondingsubsequent of tryptophan, incorporation would, isonitriles in and isothiocyanates,of general,sulfur into conflict respectively the withisonitrile this proposal. [[49,53];53]. However, and, An alternativemost therecently, aforementioned hypothesis it was proposessuggested, genomic the based subsequent studieson a the incorporation [lack54] whichof a argue againstdedicatedof incorporation sulfur gene into withinthe of isonitrile cyanide, biosynthetic [49,53]; and gene rather and, clusters, most donation recently, that either of it carbon wasa “pathway suggested, to the independent amine based of on tryptophan, sulfura the lackcarrier of would, a in general,protein,”dedicated conflict or withgene alternative thiswithin proposal. biosyntheticenzymes, An such gene alternative clusters,as a cysteine hypothesis that either desulfurase, a proposes “pathway could the independent subsequentbe responsible sulfur incorporation carrier for of sulfurisothiocyanates intoprotein,” the isonitrile or alternativefound [49 in,53 several ];enzymes, and, hapalindole most such recently, as‐type a cysteine italkaloids was suggested, desulfurase, [49]. Likewise, based could as on tobe a the theresponsible rare lack nitrile of afor dedicated variants,isothiocyanates specifically found as found in several in ambiguines hapalindole and‐type fischerindoles, alkaloids [49]. no studiesLikewise, have as investigatedto the rare nitrilethe gene within biosynthetic gene clusters, that either a “pathway independent sulfur carrier protein,” or biosyntheticvariants, specifically origin of this as foundgroup, in and ambiguines previous andstudies fischerindoles, simply propose no studies an otherwise have investigated unexplained the alternative“rearrangement”biosynthetic enzymes, origin of such the of isonitrile asthis a cysteinegroup, [40]. and desulfurase, previous studies could simply be responsible propose an forotherwise isothiocyanates unexplained found in several hapalindole-type“rearrangement”According to the of originallythe alkaloids isonitrile proposed [49 [40].]. Likewise, scheme (Figure as to the 12), rare therefore, nitrile the variants, first step specifically in the common as found in ambiguinesbiosyntheticAccording and fischerindoles,pathway to the might originally yield no studies proposedone of four have scheme of investigatedthe (Figure tricyclic 12), hapalindoles the therefore, biosynthetic the (i.e. first, C–F), step origin depending in the of thiscommon on group, and biosynthetic pathway might yield one of four of the tricyclic hapalindoles (i.e., C–F), depending on previous studies simply propose an otherwise unexplained “rearrangement” of the isonitrile [40].

Mar. Drugs 2016, 14, 73 12 of 28

According to the originally proposed scheme (Figure 12), therefore, the first step in the common biosynthetic pathway might yield one of four of the tricyclic hapalindoles (i.e., C–F), depending on whether an isonitrile or isothiocyanate indole precursor, was involved (i.e., C/E or D/F, respectively), i.e. and whetherMar. Drugs a 2016 hydrogen, 14, x or chloronium ion was utilized in the condensation ( , C/D12 of 28 or E/F, respectively). That said, recent molecular studies cast doubt on the role of chloronium ion-induced condensation,whether an and isonitrile suggest or isothiocyanate instead an indole enzymatically precursor, was driven involved condensation (i.e., C/E or D/F, reaction respectively), followed by subsequentand enzymaticwhether a hydrogen halogenation or chloronium of the resulting ion was hapalindole utilized in the [18 ,condensation55]. In this case, (i.e., eitherC/D or hapalindole E/F, respectively). That said, recent molecular studies cast doubt on the role of chloronium ion‐induced C or D (i.e., isonitrile and isothiocyante variants, respectively) would be the expected tricyclic precursor, condensation, and suggest instead an enzymatically driven condensation reaction followed by but if subsequentsubsequent introductionenzymatic halogenation of sulfur toof thethe isonitrileresulting (tohapalindole produce [18,55]. the isothiocyanate) In this case, either is assumed, and whichhapalindole is supported C or D by(i.e. recent, isonitrile data, and then isothiocyante hapalindole variants, C is respectively) the most likely would tricyclic be the expected intermediate. However,tricyclic if stereochemistry precursor, but if of subsequent the indole-isonitrile/isothiocyanate introduction of sulfur to the orisonitrile monoterpenoid (to produce precursorsthe does explainisothiocyanate) the existence is assumed, of various and which stereoisomers is supported (i.e. by, 12-epimers,recent data, then C10/11/15 hapalindole isomers), C is the most as has been propose,likely it is tricyclic likely that intermediate. other, yet However, to be identified, if stereochemistry intermediates of the indole might‐isonitrile/isothiocyanate exist. or monoterpenoid precursors does explain the existence of various stereoisomers (i.e., 12‐epimers, After the formation of a tricyclic intermediate, and thus Group 2 alkaloids, the proposed C10/11/15 isomers), as has been propose, it is likely that other, yet to be identified, intermediates biosyntheticmight schemeexist. diverges into three paths as shown in Figure 14. The first path suggests the cyclization betweenAfter the formation the C-4 carbon of a tricyclic in the intermediate, indole ring, and and thus the Group C-16 2 carbon alkaloids, in thethe proposed isopropylidene group [biosynthetic45,56]. The scheme cyclization diverges would, into three thereby, paths as produce shown in the Figure tetracyclic 14. The first hapalindoles path suggests of the Group 1. The formationcyclization of between oxindoles the andC‐4 carbon formamides in the indole (Figure ring,6) areand proposedthe C‐16 carbon to occur in the through isopropylidene oxidation of these tetracyclicgroup [45,56]. hapalindoles. The cyclization Notably, would, no thereby, studies produce have, the to tetracyclic date, elucidated hapalindoles the biosyntheticof Group 1. The origin of formation of oxindoles and formamides (Figure 6) are proposed to occur through oxidation of these the hapalindolinonestetracyclic hapalindoles. (Group Notably, 3), although no studies it is have, assumed to date, to elucidated occur via the oxidation biosynthetic (to formorigin theof the oxindole ring), andhapalindolinones subsequent (Group rearrangement 3), although (to it is form assumed the to spiro-fused occur via oxidation cyclopropane (to form the ring). oxindole Prenylation ring), of C-2 of theand indole subsequent ring rearrangement (of a tetracyclic (to form hapalindole the spiro‐fused precursor) cyclopropane would ring). produce Prenylation tetracyclic of C‐2 of ambiguines the from Groupindole 4. ring And (of if a the tetracyclic isoprene hapalindole group reacts precursor) further, would intermolecular produce tetracyclic cyclization ambiguines would from lead to the pentacyclicGroup ambiguines 4. And if the and isoprene fischambiguines group reacts of further, Group intermolecular 5 and 6. Oxidation, cyclization and would various lead rearrangement to the pentacyclic ambiguines and fischambiguines of Group 5 and 6. Oxidation, and various rearrangement of these compounds, would subsequently afford variants within the ambiguines sub-classes. of these compounds, would subsequently afford variants within the ambiguines sub‐classes.

Figure 14.FigureBiosynthesis 14. Biosynthesis of indole of indole alkaloid alkaloid sub-classes sub‐classes from from a atricyclic tricyclic precursor. precursor. The scheme The scheme shown shownis is specificallyspecifically based based on a hapalindoleon a hapalindole C ( Ci.e. (i.e., isonitrile-bearing), isonitrile‐bearing) precursor. precursor.

Mar. Drugs 2016, 14, 73 13 of 28

The second pathway would support cyclization of the tricyclic hapalindole between C-2 of the indole ring, and C-16 of the isopropylidene group, producing the tetracyclic fischerindoles of Group 7. The condensation of the hapalindole would presumably occur via acid catalysis, and be possibly governed by an enzymatic reaction. Park et al. [43] showed that the cyclization reaction was possible without the presence of an enzyme using hapalindole C formamide as a starting product. A strong acid was used to convert hapalindole C formamide to fischerindole C formamide and fischerindole C amine (Bonjouklian et al., 1988). Using the same conditions, however, this same reaction was not reproducible with hapalindole E formamide that only differs from hapalindole C formamide by presence of a chlorine atom. Recent identification of a halogenase [18] capable of utilizing fischerindole as a substrate, however, suggests chlorination could occur post-cyclization. Moreover, the identification of novel fischerindoles (Figure 10) with 6-membered “B-rings” suggest an alternative, but yet investigated, biosynthetic route. Finally, a third pathway suggests the oxidation of the tricyclic hapalindole to produce an intermediate indolinone with a structure essentially similar to a tricyclic variant of anhydrohapaloxindole A. Anhydrohapaloxindole A has been shown to be produced from singlet oxygen oxidation of hapalindole A[30,33], and similar oxidation of a tricyclic hapalindole could presumably occur. According to the biosynthesis proposed by Stratmann et al. [45] this intermediate would then undergo intermolecular cyclization between C-3 of the indole ring, and C-16 of the isopropylidene group, via acid catalysis to produce welwitindolinone A from Group 8. In order to form the remaining welwitindolinones from Group 9, it was proposed that an isocyano epoxide intermediate would be formed, and further intermolecular rearrangements would occur. Subsequent, oxidation and methylation of the welwitindolinones would, in particular, produce variants such as the oxidized welwitindolinones including 3-hydroxy-N-methylwelwitindolinone C isonitrile. That said, it is worth noting that more recent assessments have proposed an alternative pathway for the welwitindolinones, and specifically oxidative ring contraction of 12-epi-fischerindole I to give welwitindolinone A [56], however, this has yet to be investigated. As previously mentioned, only very recently have relevant gene clusters linked to the biosynthesis of the indole alkaloids been elucidated [49,54,55,57,58]. Work by Hillwig et al. [49,54] identified gene clusters from F. ambigua and H. welwitschii, and established a link to the biosynthesis of ambiguine and welwitindolinone alkaloids, respectively. The 42 kbp amb gene cluster, encoding 32 proteins, was confirmed by both bioinformatics, and in vitro enzyme studies, to be involved in the biosynthesis of ambiguines including steps common to all hapalindole alkaloids, and specifically genes associated with biosynthesis of tryptophan, prenylation precursors (i.e., isopentyl pyrophosphate and dimethylallyl pyrophosphate [DMAPP]) and the indole-isonitrile intermediate, as well as key steps in the prenylation of tetracyclic hapalindoles to ambiguines [54]. Furthermore, the presence of several oxygenases suggests a contribution to the “late stage” biosynthetic steps characteristic of the diversity of ambiguines. Concurrently, characterization of the wel cluster from H. welwitschii [49] identified genes related to the biosynthesis of the welwitindolinones including key precursors, namely of 3-(Z-21-isocyanoethenyl)indole, and the putative monoterpenoid building blocks. Gene products associated with indole-isonitrile biosynthesis (i.e., AmbI1-3 and WelI1-3) in both amb and wel gene clusters were evaluate in enzyme studies, and specifically found to produce only the cis form of the indole-isonitrile intermediate (Figure 13). Finally in addition to these common genes, the wel gene cluster included apparent methyltransferases suggested to have a role in the N-methylation of several of the welwitindolinones, and were demonstrated in in vitro enzyme studies to, in fact, function as such. Most generally, both studies support condensation of geranyl pyrophosphate (GPP) and the indole-isonitrile, rather than conversion of GPP to β-ocimene, as previously proposed [45,56]. A particularly comprehensive, subsequent study [57] coupled whole genome sequencing of multiple stigonematalean species—including Fischerella, Hapalosiphon and Westiella—and previously published, and unpublished, sequence data for gene clusters associated with hapalindole-type alkaloid biosynthesis (e.g., amb, wel [49,54]) to elaborate a unified biosynthetic pathway. Comparative studies, Mar. Drugs 2016, 14, 73 14 of 28 accordingly, identified 19 gene sets, with 92% sequence identity at the protein level, which were common to all gene clusters (and species/strains) investigated. Supporting the proposed early steps of the biosynthesis of the alkaloids, these “core” genes include those involved in tryptophan biosynthesis, and isonitrile biosynthesis genes involved in the formation of the 3-(Z-21-isocyanoethenyl)indole precursor. Notably in this study, the gene products of the latter, and specifically WelI1/3, were found to produce in vitro both cis and trans isomers of the indole-isonitrile, and were thereby suggested to explain the stereodisposition between C10 and C11 observed among the hapalindole-type alkaloids. However, concurrent studies suggested (as discussed above) that only the cis form (i.e., Z-isomer) is produced, and subsequently utilized, in the condensation [49,55]. This discrepancy was attributed to differences in incubation time of feeding experiments in the two studies. However, it should be noted that WelI genes assessed in the conflicting studies [49,57] were, in fact, isolated from two different genera/species (i.e., H. welwitschii and W. intricate), and differ to some extent in their sequence; and it is possible, therefore, that differences in enzymatic activity could also reflect phylogenetic differences in the biosynthesis. Conserved genes additionally included those involved in isoprene biosynthesis, specifically including geranyl phosphate synthases, as well as aromatic prenyltransferases which was proposed to be involved in the condensation of the indole-isonitrile to GPP (Figure 15) although this could not be confirmed in the study. In addition to these common genes, several others seemingly specific to the different sub-types were identified including various putative oxygenases, an aromatic prenylase involved in the C-2 prenylation of the ambiguines (from the amb cluster), and methyltransferases involved in the N-methylation of the welwitindolinones (from the wel cluster). Finally, a phylogenetic analysis of the strains and the gene clusters—based on both 16s rDNA and relevant, conserved biosynthetic genes (i.e., prenyltransferase) —indicated a monophyletic group, and thus vertical inheritance, of these biosynthetic genes. This latter finding, thereby, further supports the unique distribution of the indole alkaloids within the Stigonematales. Two very recent studies, based on genomic data, and in vitro enzymatic studies, have attempted to elucidate a potential mechanism for condensation of GPP and the indole-isonitrile precursor [55,58]. In one study, a homologous series of strictly magnesium-dependent aromatic prenyltransferases (AmbPI, WelP1 and FidP1) was identified across related biosynthetic gene clusters (i.e., wel, amb and unpublished fid), and specifically shown to direct the formation of 3-geranyl 3-isocyanovinyl indolenine—via condensation of GPP and the Z-isomer of the indole isonitrile—as a common intermediate [55]. Following an unusual Cope rearrangement, it was suggested that a tricyclic hapalindole could be formed via aza-Prins-type cyclization, and subsequently provide a starting point for a wide range of hapalindole-type alkaloids (Figure 15). Concurrent studies, however, by Li et al. [58] identified the fam gene cluster from F. ambigua (involved in the biosynthesis of ambiguines), and demonstrated through in vitro studies with cell-free lysates, the role of two prenyltransferases (FamD1 and FamD2) and a cyclase (FamC1) within this cluster. It was proposed that FamD2 and FamC1 “act in concert” to catalyze the condensation of the indole-isonitrile and GPP via similar Cope rearrangement, and a similar quaternary intermediate, but distinct from the studies by Liu et al. [55] suggested the formation of a tetracyclic hapalindole core, and namely 12-epi-hapalindole U. Subsequent to this, FamD1 was shown to prenylate (with DMAPP) at the C-2 position to give rise to the ambiguines. Furthermore, it was suggested that the cyclase step (via FamC1, or perhaps homologs) could be involved in determining stereochemistry (i.e., C-12 epimers, C10/C15), and also directing biosynthesis toward the fischerindoles. These findings, in general, reinforce a biosynthesis that circumvents the previously proposed β-ocimene precursor in this key step. Intriguingly, both studies point to dependence of the condensation step on pH and/or Mg2+, and might, thereby, underscore potential plasticity, relevant to structural diversification, at this step. Mar. Drugs 2016, 14, 73 15 of 28 Mar. Drugs 2016, 14, x 15 of 28

FigureFigure 15. 15.Most Most recently recently proposed proposed pathway pathway for synthesis of of common common tricyclic tricyclic hapalindole hapalindole precursor, precursor, viavia a monogeranylateda monogeranylated indolenine indolenine intermediate,intermediate, as as supported supported by by recent recent molecular molecular studies studies [55]. [55 ]. SubsequentSubsequent steps steps to to yield yield other other alkaloids alkaloids would would occuroccur via Paths A, A, B B or or C C as as shown shown in in Figure Figure 14. 14 .

Finally, recent work from the Liu group [18] has identified a non‐heme, iron‐dependent halogenaseFinally, recent(i.e., welO5 work) from from the thewel gene Liu cluster, group and [18] has has specifically identified demonstrated a non-heme, the iron-dependent ability of this halogenaseenzyme to ( i.e.catalyze, welO5 chlorination) from the ofwel thegene C‐13 cluster, of a tricyclic and has hapalindole specifically (i.e., demonstrated 12‐epi‐hapalindole the abilityC) and of thisfischerindole enzyme to ( catalyzei.e., 12‐epi chlorination‐fischerindole of U). the This C-13 finding of a not tricyclic only explains hapalindole the presence (i.e., 12- ofepi chlorinated-hapalindole C)and and non fischerindole‐chlorinated ( i.e.congeners, 12-epi-fischerindole found among U). many This of finding the hapalindole not only‐type explains alkaloids, the presence but also of chlorinatedsuggests a and role non-chlorinated of this chlorination congeners as a first found committed among many step ofin thethe hapalindole-typesubsequent formation alkaloids, of butwelwitindolinones also suggests a role [18] of (Figure this chlorination 14), and further as a first explains, committed therefore, step in the the previously subsequent unexplained formation of welwitindolinonesabsence of dechloro [18 ]variants (Figure within14), and this further sub‐class. explains, therefore, the previously unexplained absence of dechloro variants within this sub-class. 6. Biological Activity of the Hapalindole‐Type Alkaloids 6. Biological Activity of the Hapalindole-Type Alkaloids The indole alkaloids from the Stigonematales have been generally associated with a wide range of Thebiological indole activity alkaloids (Table from the1) which Stigonematales is the focus have of beenthe remainder generally associatedof this review. with The a wide reported range of biologicalbiological activity activity (Table suggests1) which a role is of the these focus compounds of the remainder in the chemical of this review.ecology Theof the reported order, as biological well as activityboth environmental suggests a role and of thesebiomedical compounds relevance in theas potential chemical toxins ecology and of promising the order, leads as well to asdrug both environmentaldiscovery, respectively. and biomedical As to relevance the potential as potential as drugs, toxins a preponderance and promising of leadssynthesis to drug studies discovery, has, respectively.accordingly, As focused to the on potential these alkaloids, as drugs, and a is preponderance reviewed extensively of synthesis elsewhere studies [47]. has, accordingly, focused on these alkaloids, and is reviewed extensively elsewhere [47].

Mar. Drugs 2016, 14, 73 16 of 28

Table 1. Summary of the reported biological activity of hapalindole-type alkaloids from Stigonematales including inhibition of microbes, cytotoxicity and plant/animal toxicity, as well as biochemical, molecular and cellular targets.

Biochemical, Molecular Plant/Animal Antimicrobial Cytotoxicity or Cellular Toxicity Bacteria Fungi Algae Hapalindoles Insecticidal A, G–H, I–J, [32,60,61] A, J, X A, H–J, U, X Modulate Na+ channels Tetracyclic (Group 1) T, X G[42] Teratogenicity/ [34] [34] in neuroblastoma [59] [29,31,34,42] vertebrate toxicity [39,41] Modulate Na+ channels E–F in neuroblastoma [59] Tricyclic (Group 2)E[62]E[62] C, E [34,62] Insecticidal [32,60] [62,63] RNA polymerase inhibition [62,64] Inhibits arginine Hapalindolinones vassopressin binding, and (Group 3) adenylate cyclase [35] Ambiguines Teratogenicity/ A–C, H A–C, H Tetracyclic (Group 4) A–C [42] vertebrate [38,42] [38,42] toxicity [39,41] Apoptosis, inhibition of E–F, G NF-kb, IKKb and E–F, I, E–F, G nitrile, I, ICAM-1 [65] Pentacyclic (Group 5) K–O, P nitrile, K–O Phytotoxic [66] K–O Inhibits mitosis (plant [37,38,42] [37,42] [37,38,42] cells) [66] Elevates ROS [65,66] Fischambiguines A, B [42]A[42] (Group 6) descholor I Teratogenicity/ Fischerindoles L[34]L[34] nitrile, L vertebrate (Group 7) [34,44] toxicity [41] Welwitindolinones Welwitindolinone A A[45] (Group 8) Inhibits microtubules and N-methyl C Welwitindolinone B mitosis [67] isothiocyanate Insecticidal [45] (Group 9) Inhibit P-glycoprotein [67] MDR [67,68]

6.1. Antimicrobial Activity Arguably, the primary bioactivity of the hapalindole-type has been inhibition of microbes including bacteria, fungi and algae. In fact, the hapalindoles, as the archetypal member of the class, were first isolated from H. fontanalis based on antimicrobial—and specifically antialgal—activity of extracts [29,30]. Subsequent studies, in the course of identifying new variants, have confirmed antimicrobial activity of both tri- and tetracyclic hapalindoles [34,42,69]. Most recently, Kim et al. [34] reported the isolation of several hapalindoles including a novel variant, hapalindole X, and demonstrated relatively potent activity (MIC ě 2.5 µM) of this congener against Mycobacterium tuberculosis, and the yeast, Candida albicans, as well as potent activity of several other hapalindoles (i.e., A, I, J) and presumptive oxidative products (i.e., hapalonamide H). It is noteworthy that both chlorinated and non-chlorinated congeners (e.g., hapalindole A and J) display similar antimicrobial activity [34]. Furthermore, comparable antimicrobial activity has been reported for 12-epi-hapalindoles [62,64]. Relatively limited information exists, on the other hand, as to the role of the variable C-11 functional group of the hapalindoles; in particular, studies have generally focused on the more common isonitrile-bearing variants compared, for example, to the less common isothiocyanate congeners. Moore et al. [29], in the original isolation of the hapalindoles, did indicate that hapalindole A “was responsible for most of the antialgal and antimycotic activity of H. fontinalis,” and reported no biological activity for hapalindole B, the isothiocyanate equivalent, isolated as a minor component in the same study; however, no specific bioactivity data were, in fact, presented in this study. That said, the thiocarbamate-containing variant, hapalindole Mar. Drugs 2016, 14, 73 17 of 28

T, was found to be antibacterial against a range of Gram-positive and -negative representatives, and quantitatively comparable to other hapalindoles in this regard [31]. In addition to inhibition of heterotrophic bacteria and fungi, it has been shown that hapalindoles effectively inhibit the growth of other algae, and cyanobacteria, in particular. Etchagaray et al. [63] specifically used antialgal activity to guide isolation of several metabolites from an Amazonian isolate of Fischerella, and identified two putative hapalindoles, including the 12-epi-hapalindole F (as confirmed by NMR studies), which inhibited a representative strain of the cyanobacterial genus, Synechococcus. This finding is notable in that hapalindole F is an isothiocyanate-bearing variant; however, Doan et al. [62] also demonstrated inhibition of cyanobacteria and eukaryotic green algae by the isonitrile counterpart of 12-epi-hapalindole F (i.e., 12-epi-hapalindole E). Similar to the hapalindoles, Smitka et al. [36] originally isolated ambiguines based on inhibition of several representative fungi including Aspergillus oryzae, Candida albicans, Penicillium notatum, Saccharomyces cerevisiae and Trichophyton mentagrophytes. And subsequent studies [37,38,42] have, likewise, demonstrated diverse antibacterial and antifungal activity of both tetra- and pentacyclic ambiguines encompassing rather diverse structural variability within the sub-class (i.e., ambiguines A–P). Also similar to the hapalindoles, both chlorinated and non-chlorinated (at C-13) variants generally showed comparable antimicrobial activity; however, unlike hapalindoles, there have to-date not been any isothiocyanate variants of the ambiguines reported. That said a multiple nitrile-containing ambiguines have been identified. In particular, ambiguine G nitrile was isolated from F. ambigua, and shown to have antibacterial and antifungal activity comparable to ambiguine isonitriles. More recently,the fischambiguines were isolated, alongside several previously recognized ambiguines, from F. ambigua specifically based on antibacterial activity of extracts toward Mycobacterium tuberculosis and Bacillus anthracis [42]. Subsequently, purified fischambiguines A and B were found to be differentially active toward a range of bacteria at concentrations as low as 2 µM; in particular, fischambiguine B was generally more active (MIC ě 2–28 µM), and specifically inhibitory toward M. tuberculosis and B. anthracis, as well as Staphylococcus aureus and M. smegmatis, whereas fischambiguine A was moderately active against only S. aureus (MIC > 82 µM). It is proposed that presence of the epoxide in fischambiguine B might underlie the observed antibacterial activity of this variant, as the presence and absence of a chlorine atom (at C-13) has generally been found to have no significant effect on the antibacterial activity of the hapalindole-type alkaloids including the related ambiguines [37,38]). Finally, antimicrobial activity has been also demonstrated for fischerindoles and welwitindolinones (Groups 7, 8 and 9). Fischerindole L was, when first isolated [43], found to be antifungal, and more recent studies [44] have, likewise, demonstrated inhibition of a range of bacteria and yeast by other variants. Similarly, the welwitindolinones—and particularly welwitindolinone A—were fungicidal [45]; however, members of the “welwitindolinone B” class (Group 9), on the other hand, have been exclusively linked to a reversal of multidrug resistance in human cancer cells (see Section 6.2, Biochemical, Molecular and Cellular Bioactivity).

6.2. Biochemical, Molecular and Cellular Bioactivity Aside from their antimicrobial activity, subsequent studies have suggested a wide range of biological activity at the biochemical, molecular and cellular level that is, likewise, relevant to both the potential of the hapalindole-type alkaloids as drug leads, and to their possible role as toxins. At the cellular level, several of the indole alkaloids have been shown to be cytotoxic. In addition to antimicrobial activity, several of the hapalindoles (including A, C, E, H–J, U and X) were shown to be cytotoxic to both normal mammalian (i.e., Vero) and various cancer cell lines [34,62]. Similar to antimicrobial activity, tetracyclic hapalindole A and J—and the unusual X—were particularly active in these studies, as was hapalindole H (although it was not comparatively tested for antimicrobial activity in the same study), whereas others (i.e., I and U) were only moderately and/or variably cytotoxic [34]. The difference in cytotoxicity between, for example, hapalindole H and U is perhaps notable as they differ only in relative stereochemistry at C-10/11/15, and suggest a stereoselectivity of Mar. Drugs 2016, 14, 73 18 of 28 this biological activity. Similarly, in a prior study, the tricyclic variant, 12-epi-hapalindole E, was found to be moderately cytotoxic to a mouse myeloma cell line [62]. In addition, some oxidized hapalindoles (e.g., hapalonamide H, anhydrohapaloxindole A) shared cytotoxicity equivalent to the hapalindoles, whereas others (i.e., 13-hydroxydechlorofontonamide) did not [34]. This finding is intriguing as both active and inactive variants (e.g., hapalonamide H and 13-hydroxydechlorofontonamide, respectively) include representatives which lack an intact indole ring. Both tetra- and pentacyclic ambiguine isonitriles including numerous congeners (A–C, E, F, I, K–O), as well as the unique ambiguine G nitrile, were also found to be cytotoxic to a mammalian cell line [37,42], whereas ambiguine D isonitrile was shown to inhibit mitosis in plant cells [66]. In a recent study, it was specifically shown that ambiguine I isonitrile induced apoptosis in the MCF-7 breast cancer cell line, and led to arrest of cells in G1 phase [65]. Notably, the chemically related fischambiguines A and B were effectively not found to be cytoxotoxic (Mo et al., 2010), although further investigation is clearly needed. Interestingly, however, given the considerable deviation in carbon backbone, the fischerindoles were found to be cytotoxic, and fischerindole L, in particular, was shown to be as equally cytotoxic (to both normal and cancer cells) as hapalindole-type alkaloids [44]. Arguably, the sub-class which has received the most attention in terms of cytotoxicity, particularly in relation to potential as anticancer drug, has been the welwitindolinones, and specifically welwitindolinone C isothiocyanate and its N-methyl derivative—or welwistatin—that were found to inhibit cell proliferation in various cancer cell lines, and moreover, to circumvent (as discussed below) multidrug resistance [67,68]. As a result of this compelling biological activity, considerable attention has been devoted, in particular, to synthetic studies of the welwistatin and other welwitindolinones [47]. With respect to biochemical or molecular underpinnings of the pharmacology of the hapalindole-type alkaloids, several possible mechanisms have, in fact, been identified (Figure 16). In a study of ambiguine I, for example, it was shown that observed induction of apoptosis and cell-cycle arrest could be specifically attributed to the inhibition of pathways involving the transcription factor, NF-κB, which has a well-established role in regulating multiple pathways involved in cancer progression including suppression of apoptosis [65]. It was generally shown in these studies that apoptosis occurred by way of a caspase-independent mechanism, and involved mitochondrial dysfunction including elevated levels of reactive oxygen species (ROS). Interestingly, studies of the antimitotic activity of ambiguine D in a plant cell model, likewise, demonstrated elevated ROS, and associated increases in lipid peroxidation, and was suggested to play a role in the observed cell death [66]. Moreover, immunoblotting studies of breast cancer cells treated with ambiguine I showed down-regulation of both p65 and p50 subunits of NF-κB, and the upstream kinase, IKKβ; in turn, ambiguine I treatment led to a down-regulation of intracellular cell adhesion protein, ICAM-1, which is involved in migration and invasion of cancer cells, suggesting a potential novel mechanism of the compound as an anticancer drug [65]. As discussed above, a considerable amount of attention has been placed on the welwitindolinones with respect to antiproliferative—and potential anticancer—activity. Although much of this research has focused on synthetic studies, the original characterization of the molecules and their biological activity identified compelling mechanistic details [67,68]. Specifically, welwitindolinones were isolated as part of studies focused on the identification of compounds which reverse MDR of cancer cells. Two variants, welwitindolinone C isothiocyanate and its N-methyl derivative, were specifically shown to abate acquired drug-resistance of a breast cancer cell line (i.e., MCF-7/ADR) with respect to a diverse range of anticancer drugs, including taxol, vinblastine, colchicine, daunomycin and actinomycin D, at sub-micromolar concentrations [68]. In particular, N-methylwelwitindolinone C isothiocyanate showed pronounced activity (at concentrations below that of the described MDR-reversing drug, verapamil). And notably, the isonitrile variant of welwitindolinone C showed no activity, suggesting role of the isothiocyanate functional group. Similarly, accumulation of radiolabeled [3H]-taxol was observed in an MDR variant of an ovarian carcinoma cell line (i.e., SK-VBL-1), and paralleled the potency of MDR-reversing potential of the variants in breast cancer cells; interestingly, however, the Mar. Drugs 2016, 14, x 19 of 28

Similarly, accumulation of radiolabeled [3H]‐taxol was observed in an MDR variant of an ovarian Mar. Drugs 2016, 14, 73 19 of 28 carcinoma cell line (i.e., SK‐VBL‐1), and paralleled the potency of MDR‐reversing potential of the variants in breast cancer cells; interestingly, however, the N‐methyl derivative was not found to lead N-to methylthe accumulation derivative was[3H]‐ nottaxol found in treated to lead cells, to the whereas accumulation the non [3H]-taxol‐methylated in treated welwitindolinone cells, whereas C theisothiocyante non-methylated and verapamil welwitindolinone did. These C studies, isothiocyante moreover, and elucidate verapamil an did. apparent These interaction studies, moreover, with the elucidateP‐glycoprotein an apparent (i.e., ATP interaction‐binding withefflux the pump) P-glycoprotein pathway of (i.e. MDR., ATP-binding Furthering efflux the interest pump) in pathway these two of MDR.compounds, Furthering a thesubsequent interest instudy these twodemonstrated, compounds, ain subsequent addition study to demonstrated,MDR‐reversing in additionactivity, toantiproliferative MDR-reversing activity activity, of antiproliferative the welwistatins activity themselves, of the welwistatins and specifically, themselves, interference and specifically, with interferencemicrotubule withformation microtubule via inhibition formation of tubulin via inhibition polymerization of tubulin [66]. polymerization [67].

Vasopressin

Hapalindolinones Na+

Hapalindoles AVPR

ATP

+ AC Na cAMP RNA Pol MT (tubulin) NF IKK ICAM-1 P-gp

Ambiguines Welwitindolinones

Figure 16. Overview of biochemical, molecular and cellular targets targets of the the hapalindole hapalindole-type‐type alkaloids. alkaloids. Abbreviations: ACAC = adenylate= adenylate cyclase; cyclase; AVPR AVPR = arginine = arginine vasopressin vasopressin receptor; ICAM-1 receptor; = interceullular ICAM‐1 = adhesioninterceullular molecule adhesion 1; IKK moleculeβ = inhibitor 1; IKK of nuclearβ = inhibitor factor kappa-Bof nuclear kinase factor subunit kappa beta;‐B kinase MT = subunit microtubulin; beta; P-gpMT = =microtubulin; p-glycoprotein; P‐gp NF = pκB‐glycoprotein; = nuclear factor NFκ kappaB = nuclear B; RNA factor pol kappa = RNA B; polymerase.RNA pol = RNA polymerase.

With respect to the hapalindoles, studies by Doan et al. [60,61] have pointed to inhibition of RNA With respect to the hapalindoles, studies by Doan et al. [62,64] have pointed to inhibition of polymerase as a possible mode of action. Investigating allelopathic potential, it was shown that 12‐ RNA polymerase as a possible mode of action. Investigating allelopathic potential, it was shown epi‐hapalindole E inhibited the growth of a wide range of bacteria, fungi and algae, and more that 12-epi-hapalindole E inhibited the growth of a wide range of bacteria, fungi and algae, and specifically, in a Gram‐positive bacterial model (i.e., Bacillus subtilis 168), inhibited both RNA and more specifically, in a Gram-positive bacterial model (i.e., Bacillus subtilis 168), inhibited both RNA protein synthesis [60]. It was further demonstrated in vitro that the hapalindole inhibited RNA and protein synthesis [62]. It was further demonstrated in vitro that the hapalindole inhibited RNA polymerase directly (rather than interacting with the DNA template). And subsequent studies using polymerase directly (rather than interacting with the DNA template). And subsequent studies using Gram‐negative E. coli RNA polymerase both confirmed the direct interaction with the enzyme (and Gram-negative E. coli RNA polymerase both confirmed the direct interaction with the enzyme (and not DNA), and also demonstrated several notable difference from binding activity of other RNA not DNA), and also demonstrated several notable difference from binding activity of other RNA polymerase inhibiting antibiotics, i.e., rifampicin [61]. The study concluded, however, based on the polymerase inhibiting antibiotics, i.e., rifampicin [64]. The study concluded, however, based on the relatively high concentration required for RNA polymerase inhibition, that additional mechanisms relatively high concentration required for RNA polymerase inhibition, that additional mechanisms were likely involved in the antimicrobial activity of the hapalindoles. were likely involved in the antimicrobial activity of the hapalindoles. Interestingly, several hapalindoles—including the tricyclic hapalindoles E and C isonitrile, and Interestingly, several hapalindoles—including the tricyclic hapalindoles E and C isonitrile, and tetracyclic hapalindole L—were recently shown to modulate sodium channels in a neuroblastoma tetracyclic hapalindole L—were recently shown to modulate sodium channels in a neuroblastoma cell cell line [67]. The inhibition of veratridine‐induced depolarization (as measure of sodium channels) line [59]. The inhibition of veratridine-induced depolarization (as measure of sodium channels) occurred without cytotoxicity, and it is unlikely, therefore, to explain antimicrobial or cytotoxic occurred without cytotoxicity, and it is unlikely, therefore, to explain antimicrobial or cytotoxic activity. However, this mechanism was suggested to explain an observed insecticidal activity of these activity. However, this mechanism was suggested to explain an observed insecticidal activity of compounds [32,68] (see below). Moreover, this activity supports the potential of the hapalindoles as these compounds [32,60] (see below). Moreover, this activity supports the potential of the hapalindoles potential environmental toxins similar to other neurotoxic metabolites from cyanobacteria (and other as potential environmental toxins similar to other neurotoxic metabolites from cyanobacteria (and “harmful algae”) including saxitoxins and anatoxin‐a (see Section 8, Relevance as Harmful Algae). other “harmful algae”) including saxitoxins and anatoxin-a (see Section8, Relevance as Harmful Algae). In addition to the hapalindoles, oxidized congeners have also been investigated with respect to In addition to the hapalindoles, oxidized congeners have also been investigated with respect relevant pharmacological activity. Specifically, the hapalindolinones (i.e., Group 3; Figure 3) were to relevant pharmacological activity. Specifically, the hapalindolinones (i.e., Group 3; Figure3) were first identified in fact, as part of screening program to identify vasopressin antagonists [35]. Mar. Drugs 2016, 14, x 20 of 28

Mar. Drugs 2016, 14, 73 20 of 28 first identified in fact, as part of screening program to identify vasopressin antagonists [35]. Following isolation and chemical characterization, hapalindolinone A was shown to inhibit both binding of arginineFollowing vasopression isolation and to chemicalkidney cells, characterization, and also inhibit hapalindolinone vasopressin stimulated A was shown adenylate to inhibit cyclase. both Accordingly,binding of arginine this compound vasopression has toreceived kidney attention cells, and as also a possible inhibit vasopressindrug lead for stimulated various associated adenylate diseases.cyclase. Accordingly, this compound has received attention as a possible drug lead for various associated diseases. 6.3. Toxicity to Plants and Animals 6.3. Toxicity to Plants and Animals Finally, alongside pharmacological studies of the hapalindole‐type alkaloids, several studies have Finally,investigated alongside their pharmacological toxic effects in multicellular studies of the organismal hapalindole-type systems. alkaloids, In some several cases, these studies effects have couldinvestigated be attributed their toxic to biochemical, effects in multicellular molecular organismal or cellular systems. mechanisms. In some For cases, example, these inhibition effects could of mitosisbe attributed by ambiguine to biochemical, D isonitrile molecular was linked or cellular to phytotoxic mechanisms. effects, For example,and specifically inhibition inhibition of mitosis of seedlingby ambiguine germination D isonitrile [63]; the was phytotoxic linked to phytotoxicactivity was, effects, in fact, and successfully specifically employed inhibition for of bioassay seedling‐ guidedgermination fractionation [66]; the and phytotoxic purification activity of the was, compound. in fact, successfully However, employedin most cases, for bioassay-guided the connection betweenfractionation mechanism and purification and toxicity of remains the compound. to be elucidated. However, in most cases, the connection between mechanismThe hapalindole and toxicity‐type remains alkaloids to be have, elucidated. for example, been shown as toxic to invertebrate animals, and particularlyThe hapalindole-type insects. The alkaloids larvae of have, the forfly, example, Chironomus been riparius shown, was as toxic used to as invertebrate a model system animals, to isolateand particularly several hapalindoles insects. The (i.e. larvae, C, E, of J and the fly,L includingChironomus 12‐epimers) riparius, wasas insecticidal used as a modelagents systemfrom a biofilmto isolate‐forming several species hapalindoles of Fischerella (i.e., [32,68]; C, E, J in and the L same including study, 12-epimers)hapalindole asJ (identified insecticidal as agentsa new variant)from a biofilm-formingwas also found to species be lethal of toFischerella the freshwater[32,60 crustacean,]; in the same Thamnocephalus study, hapalindole platyurus J. Moreover, (identified recentas a new studies variant) identified was also apparent found tosodium be lethal channel to the modulating freshwater activity crustacean, of theseThamnocephalus same compounds, platyurus as . discussedMoreover, above, recent and studies it was identified proposed apparent that this sodium activity channel might modulating explain the activity observed of insecticidal these same activitycompounds, [67]. In as fact, discussed other above,studies and have, it waslikewise, proposed reported that thisthe apparent activity might insecticidal explain activity the observed of the hapalindoleinsecticidal‐ activitytype alkaloids. [59]. In fact,It was other observed studies have, that likewise,welwitindolinones, reported the and apparent particularly insecticidal N‐ methylwelwitindolinoneactivity of the hapalindole-type C isothiocyanate, alkaloids. It isolated was observed as the thatmajor welwitindolinones, alkaloid from W. and intricate particularly, were insecticidalN-methylwelwitindolinone against blowfly larvae C isothiocyanate, [45]. More recently isolated an as evaluation the major alkaloid of crude from extracts,W. intricate and purified, were alkaloids,insecticidal from against a freshwater blowfly larvaeFischerella [45]. isolate More was recently shown an to evaluation inhibit development of crude extracts, of mosquito and purified (Aedes aegyptialkaloids,) larvae from [69]. a freshwater Crude extracts,Fischerella in thisisolate case, were was lethal shown to tolarvae inhibit by day development 4 of development, of mosquito and larvae(Aedes aegyptinever )transitioned larvae [61]. Crude beyond extracts, 1st instar in this stages. case, were Both lethal increased to larvae mortality by day 4 ofand development, delays in developmentand larvae never were, transitioned likewise, observed beyond for 1st larvae instar exposed stages. to Both12‐epi increased‐hapalindole mortality H isonitrile and compared delays in todevelopment control (untreated) were, likewise, larvae; observed specifically, for larvaelarvae exposedexposed to to 12- theepi alkaloid-hapalindole generally H isonitrile were arrested compared at 1stto control instar (untreated)stage after larvae;5 days specifically, (whereas control larvae exposed larvae transitioned to the alkaloid to generally 2nd instar were by arrested day 2), at and 1st developmentinstar stage after was 5 observably days (whereas impaired control for larvae larvae transitioned (only one) which to 2nd did instar transition by day to 2), 2nd and instar development (Figure 17).was observably impaired for larvae (only one) which did transition to 2nd instar (Figure 17).

Figure 17. Inhibition of mosquito (Aedes aegypti) larval development by 12‐epi‐hapalindole H Figure 17. Inhibition of mosquito (Aedes aegypti) larval development by 12-epi-hapalindole H isonitrile. isonitrile. Shown are 2nd instar larvae treated with the alkaloid (a) showing development Shown are 2nd instar larvae treated with the alkaloid (a) showing development abnormalities compared abnormalitiesto untreated controlcompared larvae to untreated (b). Most control larvae larvae treated (b). with Most 12-epi-hapalindole larvae treated with H, 12 however,‐epi‐hapalindole do not H,transition however, from do 1stnot to transition 2nd instar. from 1st to 2nd instar.

Mar. Drugs 2016, 14, 73 21 of 28 Mar. Drugs 2016, 14, x 21 of 28

Most recently, the zebrafishzebrafish (Danio rerio) embryo as a vertebrate model of teratogenicity ( (i.e.i.e.,, development toxicity) toxicity) was was used used to to identify, identify, purify purify and and characterize characterize several, several, diverse diverse alkaloids alkaloids from from a Fischerellaa Fischerella isolateisolate [39,41,70]. [39,41,70 ].Teratogenicity Teratogenicity of of the the strain strain was was first first identified identified in in screening screening studies, and initial characterizationcharacterization supported supported their their identity identity as hapalindole-type as hapalindole‐ alkaloidstype alkaloids [70]. Subsequent [70]. Subsequent studies werestudies able were to able utilize to utilize bioassay-guided bioassay‐guided fractionation, fractionation, based based on observed on observed teratogenicity, teratogenicity, to purify to purify and andchemically chemically characterize characterize several several variants variants including including hapalindoles, hapalindoles, ambiguines ambiguines andand fischerindoles, fischerindoles, including new new nitrile nitrile ambiguines ambiguines [39,41]. [39,41]. Toxicity Toxicity to to embryos embryos included—in included—in addition addition to tomortality mortality at theat the highest highest exposure exposure concentrations—impairment concentrations—impairment of of melanophores, melanophores, leading leading to to lack lack of of pigmentation, pigmentation, as well as deformity of the body axisaxis andand severesevere edemasedemas (Figure(Figure 1818).). The observed impairment of melanophore development, in particular, is notable as melanocytes are derived from neural crest cells of embryos, embryos, may may thereby thereby support, support, as asdiscussed discussed below below (Section (Section 8, Relevance8, Relevance as Harmful as Harmful Algae), Algae), a role ofa rolethese of thesecompounds compounds in myelinopathies in myelinopathies observed observed among among waterfowl waterfowl exposed exposed to to epiphytic stigonematalean cyanobacteria [71]. [71]. The tetracyclic 12 12-‐epi‐-hapalindolehapalindole H H isonitrile isonitrile was most closely linked to the developmental deformity observed in these studies, and was generally the most potent teratogen [39]. [39]. However, However, both both ambiguine ambiguine isonitrile isonitrile and and nitrile nitrile variants, variants, as as well well as fischerindoles, fischerindoles, purifiedpurified inin subsequent subsequent studies studies demonstrated demonstrated observable observable teratogenicity teratogenicity albeit at higher albeit concentrations, at higher concentrations,i.e., ě5–20 µM[ 41i.e.]., ≥ Generally5–20 μM [41]. speaking, Generally these speaking, studies perhapsthese studies best supportperhaps abest potential support contribution a potential contributionof the hapalindole-alkaloids of the hapalindole to‐alkaloids the toxicity to the of cyanobacteria toxicity of cyanobacteria from the order from as the a order “harmful as a “harmful algae” in algae”aquatic in systems. aquatic systems.

Figure 18.18. TeratogenicityTeratogenicity of of hapalindole-type hapalindole‐type alkaloids alkaloids in thein zebrafishthe zebrafish (Danio (Danio rerio) rerio embryo) embryo as a model as a modelof vertebrate of vertebrate development. development. Shown Shown are control are control (untreated) (untreated) embryos embryos at 3 days at post-fertilization 3 days post‐fertilization (a); and (embryosa); and embryos exposed forexposed three daysfor three to 12- daysepi-ambiguine to 12‐epi‐ambiguine B nitrile at B 10 nitrileµg/mL at (10b); μ andg/mL to 12-(b);epi and-hapalindole to 12‐epi‐ hapalindoleH isonitrile at H 5 isonitrile and 10 µ atg/mL 5 and (c 10and μg/mLd, respectively) (c and d, respectively) [61]. [69].

7. Possible Possible Ecological Ecological Functions Functions A functional ecological role of the hapalindole-typehapalindole‐type alkaloids has been long suggested. Indeed, Indeed, as discussed above, hapalindoles were firstfirst characterized from Hapalosiphon following observations of apparent allelopathy, and in particular, antialgal activity suggestive of a role in the inhibition of potentially competing aquatic microbes, i.e.i.e.,, algae. More More specifically, specifically, the inhibition of the growth of Anabaena by sympatric H. intricatus and H. fontinalis led to thethe isolation,isolation, and subsequentsubsequent chemical characterization, of hapalindole A and B as the first first archetypal variants of the class [29]. [29]. And And several, several, subsequent studies pointed to the possible allelopathic role of the hapalindoles and ambiguines including antialgal activity [[38,60,62].38,62,63]. Further supporting a role of the hapalindole‐type alkaloids in allelopathy, studies by Gantar et al. [72] recently investigated numerous cyanobacterial and green algal isolates from the Florida

Mar. Drugs 2016, 14, 73 22 of 28

Further supporting a role of the hapalindole-type alkaloids in allelopathy, studies by Gantar et al. [72] recently investigated numerous cyanobacterial and green algal isolates from the Florida Everglades, and associated South Florida waterways, by means of pairwise co-culturing experiments, and observed an isolate of Fischerella (52-1) as the most frequent inhibitor of sympatric algae. Moreover, extracts from Fischerella 52-1 were subsequently shown to inhibit photosynthesis, and disrupt thylakoid membranes, in Chlamydomonas, as a model green alga [72]. This same strain was subsequently shown to produce a diversity of indole alkaloids including hapalindoles, ambiguines and fischerindoles [39,41], and it was proposed that observed allelopathy may, therefore, be associated with these compounds; however, a definitive link between the observed allelopathy and these compounds has yet to be established. That said, hapalindoles and ambiguines were isolated from this same strain as major components of “exudates” produced by cultures [39], suggesting possible transport of these compounds to the extracellular environment as would be consistent with an allelopathic role. In addition to the indole alkaloids, other bioactive metabolites from the order have been proposed to have a similar ecological role. Fischerellins A and B were shown to be only marginally antimicrobial (when tested alongside hapalindole alkaloids) against numerous species of bacteria and fungi [34]. However, the fisherellins were potently antialgal, inhibiting cyanobacteria and green-algae at nanomolar concentrations, despite having little or no effect on eubacteria. Indeed, this antialgal activity was used to isolate fischerellin A from Fischerella sp. [27,73]. Moreover, fischerellin A was subsequently shown to specifically inhibit photosystem II of algae and other photosynthetic organisms. (i.e., plants) [27,74]. Similarly, although not directly investigated in this regard, the dendroamides isolated from Stigonema belong to the same class of cyclic peptides as—and bear, in fact, considerable structural similarity to—nostocyclamide from the cyanobacterial genus, Nostoc, which have been linked to allelopathy, and specifically inhibition of cyanobacteria, green-algae and diatoms [25,26]. Finally, in addition to the antimicrobial activity observed for hapalindole-type alkaloids, and potential relevance to allelopathy, several representatives have demonstrated insecticidal, and particularly insect larvicidal activity, as well as effects on freshwater zooplankton [32,45,60,61]. It is, therefore, tempting to speculate a possible role of these secondary metabolites against potential micrograzers. Such a role of cyanobacterial metabolites has, in fact, been generally proposed [75]. The observation of sodium channel modulation in neurons [59] by the hapalindoles—as an animal-specific mechanism of toxicity—further underscores a potential targeting of potential grazers, and most likely microinvertebrates (e.g., insect larvae, crustaceans). To date, however, no studies have directly investigated the anti-grazer potential of these compounds.

8. Relevance as Harmful Algae As ubiquitous components of freshwater habitats, a link between production of toxic metabolites by the stigonematalean cyanobacteria, and possible environmental human health concerns, is proposed. In terms of potential toxicity of the indole alkaloids from Stigonematales there have, however, been rather limited documented links to negative impacts on human, animal or ecocystem health in this regard. That said, recognized cyanobacterial toxins, and specifically the microcystins [11–13] and BMAA [14] have been identified from members of the order (see Section3, Stigonematales as a Source of Bioactive Metabolites). Moreover, a preponderance of toxicological studies of the indole alkaloids effectively supports the potential of these metabolites as naturally occurring toxins. Perhaps most relevant, in this regard, is the recent identification of teratogenicity of the indole alkaloids from a freshwater Fischerella strain, specifically using the zebrafish (Danio rerio) embryo model of vertebrate development [39,41,70]. In these studies, developmental toxicity was observed in zebrafish embryos ambiently exposed to the alkaloids, suggesting the potential from direct exposure in the environment (i.e., the water column). And, as a freshwater teleost fish model, these findings would directly support a potential effect of these teratogens on fish populations in these habitats. Moreover, as a vertebrate model, it is, of course, possible that these observations could be extrapolated to other vertebrates including humans. Mar. Drugs 2016, 14, 73 23 of 28

Alongside observed vertebrate toxicity, the recently reported effects of hapalindoles on sodium channels of neurons, likewise, indirectly supports potential neurotoxicity of these compounds, and possible concerns, therefore, as toxigenic algae in aquatic systems [59]. Indeed, the authors of these studies compared activity of the hapalindoles, in this regard, to the known sodium-channel blocker, and recognized HAB toxin, neosaxitoxin, and suggested a common mechanism. It was accordingly proposed that hapalindoles should be considered alongside other neurotoxic cyanotoxins (e.g., saxitoxins, anatoxin-a). Although there have been no direct links between characterized bioactive (i.e., potentially toxic) metabolites from the Stigonematales, and either human or animal intoxiciations, there has been indirect evidence in this regard. Most notably, on-going studies have documented a link between apparent metabolites of a stigonematalean cyanobacterial species, and well documented avian ventricular myelinopathy (AVM) among various bird species including waterfowl (e.g., American coots, various duck species), and other birds (e.g., bald eagles). Numerous studies, in this regard, have documented AVM as a neurological disease among birds, and feeding studies have linked the disorder to consumption of aquatic vegetation, and specifically the invasive Hydrilla verticillata [76]. More recently, however, subsequent studies identified, specifically using molecular techniques, an apparent cyanobacterial epiphyte of H. verticillata, and correlated presence to the occurrence of AVM [77]. And most recently, these studies have characterized a new species, Aetokthonos hydrillicola, specifically within the Stigonematales, as the epiphytic cyanobacterium [10]. Feeding studies with the plant material containing this epiphytic cyanobacterium, as well as extracts prepared from this material, generally confirm a link between AVM and a cyanobacterial metabolite [71,78]. To date, however, the actual toxins involved in AVM have yet to be identified, and work is currently in progress to do so (S. Wilde, personal communication). That said, hapalindole-type alkaloids represent clear candidates as toxins associated with AVM. It was, in fact, recently noted by Cagide et al. [59] that neurotoxicity, and apparent sodium channel blocking, of hapalindole could contribute to this neurological disorder. Likewise, the observed teratogenicity of the alkaloids [39,70] might suggest a mechanism. Specifically, the teratogenic alkaloids from Fischerella seemingly affected neural crest-derived cells (i.e., melanophores) in the zebrafish embryo model [39]. Glia, as the producers of myelin in the nervous system, are, likewise, derived from neural crest; and the vacuolization of myelin (i.e., “white matter”) in the brains of AVM-affected birds might be consistent with apparently cell-lineage specific activity of the teratogenic indole alkaloids. In addition, a recent study identified the neurotoxic non-proteinogenic amino acid, β-methylamino-L-alanine (BMAA), from this epiphyte [14], and suggested a possible role in AVM. Clearly, identification of the toxic principles involved in AMV will be key to establishing a link to any of these metabolites, or others.

9. Potential Source of Drug Leads Finally, in terms of the potential of these indole alkaloids as leads to pharmaceutical, the literature is, in fact, rather limited. However, pharmacological data—along with continued synthetic interest in these molecules—do support potential future development in this regard. As detailed above, cellular and molecular targets identified for the hapalindole-type alkaloids include appreciable antibiotic activity, antiproliferative effects against diverse cancer cells, reversal of MDR, selective inhibition of RNA polymerases and antagonism of vasopressin. Underscoring this pharmacological potential it has been regularly noted that antibiotic and antimycotic activity of the hapalindoles and ambiguines are comparable to established drugs in this regard; for example, hapalindole T inhibited M. tuberculosis as potently as the established anti-tuberculosis drug, rifampicin, and ambiguines were comparable with commercially available antibiotics (e.g., streptomycin) and antifungal drugs (e.g., puramycin, amphotericin B) [31,38]. On the other hand, the welwistatins (i.e., welwitindolinones) have been perhaps most actively studied as compelling anticancer drug candidates due to a combined antimitotic activity, and MDR-reversing activity which is comparable with established drugs (e.g., veraprimil) in Mar. Drugs 2016, 14, 73 24 of 28 this regard [67,68]. Accordingly, to-date there have been no less than twenty-three published synthetic studies targeting this latter sub-class of alkaloids [47]. Consistent with pharmacological potential, several patents have been filed for a diverse range of metabolites from the Stigonematales. Numerous patents have been filed in relation to the reported biological activity of the hapalindoles and ambiguines including, in particular, antifungal, antibacterial and antitumor activity, as well as to methods for their production including biosynthesis from cultures. Similarly, patents regarding the anticancer potential of welwistatin, and vasopressin antagonist activity of the hapalindolinones, specifically as a treatment for congestive heart failure, hypertension, edema and hyponatremia, have been filed within the past 20 years. The hapalindole alkaloids are clearly poised with respect to future development as drug leads. To date, however, there have been no clinical studies of the hapalindole-type alkaloids, and future studies in this regard are obviously needed, along with continued pharmacological characterization, and supporting chemical synthesis (and, alternatively, biosynthetic studies), to realize this potential.

10. Conclusions Taken together, the existing data regarding the bioactive secondary metabolites from the Stigonematales paint a burgeoning picture of the role of these compounds—and particularly the hapalindole-type alkaloids—with respect to human and environmental health. Continued chemical and pharmacological/toxicological characterization of these metabolites, and rapidly evolving insight as to their biosynthetic origins, reveals the impressive depth of this intriguing metabolic class of compounds. However, realization of the potential of these compounds as drugs (for an apparently wide range of important biomedical targets), as well as a suggested role of these metabolites with respect to both ecology of the order, and as a novel class of environmental toxins, remains an exciting area for future scientific exploration. In terms of ecological roles of these compounds, much remains to be studied. Likewise, identification (and clarification) of links between bioactive metabolites from the stigonematalean cyanobacteria and intoxications will be key to confirming a role in environmental health (i.e., as HABs). Finally, while a preponderance of pre-clinical data support the potential of these compounds as drugs, clinical studies are, in particular, needed. The current state of the science with respect to these widespread cyanobacteria clearly support future studies with respect to all of these aspects.

Author Contributions: J.B. and K.W. contributed equally to this manuscript. The review was specifically developed as part of the M.Sc. (Chemistry) thesis of K.W., and later modified to the current form. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Singh, R.K.; Tiwari, S.P.; Rai, A.K.; Mohapatra, T.M. Cyanobacteria: An emerging source for drug discovery. J. Antibiot. 2011, 64, 401–412. [CrossRef][PubMed] 2. Dixit, R.B.; Suseela, M.R. Cyanobacteria: Potential candidates for drug discovery. Antonie Van Leeuwenhoek 2013, 103, 947–961. [CrossRef][PubMed] 3. Zanchett, G.; Oliveira-Filho, E.C. Cyanobacteria and cyanotoxins: From impacts on aquatic ecosystems and human health to anticarcinogenic effects. Toxins 2013, 5, 1896–1917. [CrossRef][PubMed] 4. Dittmann, E.; Gugger, M.; Sivonen, K.; Fewer, D.P. Natural product biosynthetic diversity and comparative genomics of the cyanobacteria. Trends Microbiol. 2015, 23, 642–652. [CrossRef][PubMed] 5. Ibelings, B.W.; Backer, L.C.; Kardinaal, W.E.; Chorus, I. Current approaches to cyanotoxin risk assessment and risk management around the globe. Harmful Algae 2015, 49, 63–74. [CrossRef][PubMed] 6. Wilmotte, A.; Herdman, M. Phylogenetic relationships among the cyanobacteria based on 16s rRNA sequences. In Bergey’s Manual of Systematic Bacteriology, 2nd ed.; Boone, D.R., Castenholz, R.W., Eds.; Springer: New York, NY, USA, 2001; Volume 1, pp. 487–493. 7. Anagnostidis, K.; Komárek, J. Modern approach to the classification system of Cyanophytes 5-Stigonematales. Arch. Hydrobiol. Suppl. 1990, 86, 1–73. Mar. Drugs 2016, 14, 73 25 of 28

8. Gugger, M.F.; Hoffmann, L. Polyphyly of true branching cyanobacteria (Stigonematales). Int. J. Sytemat. Evol. Microbiol. 2004, 54, 349–357. [CrossRef][PubMed] 9. Dagan, T.; Roettger, M.; Stucken, K.; Landan, G.; Koch, R.; Major, P.; Gould, S.B.; Goremykin, V.V.; Rippka, R.; Tandeau de Marsac, N.; et al. Genomes of Stigonematalean cyanobacteria (Subsection V) and the evolution of oxygenic photosynthesis from prokaryote to plastids. Genome Biol. Evol. 2013, 5, 31–44. [CrossRef][PubMed] 10. Wilde, S.B.; Johansen, J.R.; Wilde, H.D.; Jiang, P.; Bartelme, B.A.; Haynie, R.S. Aetokthenos hydrillicola gen. et sp. nov.: Epiphytic cyanobacteria on invasive aquatic plants implicated in Avian Vacuolar Myelinopathy. Phytotaxa 2014, 181, 243–260. [CrossRef] 11. Fiore, M.F.; Genuária, D.B.; da Silva, C.S.; Shishido, T.K.; Moraes, L.A.; Cantúsio, N.R.; Silva-Stenico, M.E. Microcystin production by a freshwater spring cyanobacterium of the genus Fischerella. Toxicon 2009, 53, 754–761. [CrossRef][PubMed] 12. Prinsep, M.R.; Caplan, F.R.; Moore, R.E.; Patterson, G.M.L.; Honkanen, R.E.; Boynton, A.L. Microcystin-LA from a blue-green algae belonging to the Stigonematales. Phytochemistry 1992, 31, 1247–1248. [CrossRef] 13. Cirés, R.; Alvarez-Roa, C.; Wood, S.A.; Puddick, J.; Loza, V.; Heiman, K. First report of microcystins producing Fischerella sp. (Stigonematales, Cyanobacteria) in tropical Australia. Toxicon 2014, 88, 62–66. [CrossRef] [PubMed] 14. Bidigare, R.R.; Christensen, S.J.; Wilde, S.B.; Banack, S.A. Cyanobacteria and BMAA: Possible linkage with avian vacuolar myelinopathy (AVM) in the southeastern United States. Amyotroph. Lateral Scler. 2009, 10, 71–73. [CrossRef][PubMed] 15. Niiyama, Y.; Tuji, A.; Tsujimura, S. Umezakia natans does not belong to the Stigonemataceae to Nostocaceae. Fottea 2011, 11, 163–169. [CrossRef] 16. Falch, B.S.; König, G.M.; Wright, A.D.; Sticher, O.; Angerhofer, C.K.; Pezzuto, J.M.; Bachmann, H. Biological activities of cyanobacteria: Evaluation of extracts and pure compounds. Planta Med. 1995, 61, 321–328. [CrossRef][PubMed] 17. Wright, A.D.; Papendorf, O.; König, G.M. Ambigol C and 2,4-dichlorobenzoic acid, natural products produced by the terrestrial cyanobacterium Fischerella ambigua. J. Nat. Prod. 2005, 68, 459–461. [CrossRef] [PubMed] 18. Hilliwig, M.L.; Liu, X. A new family of iron-dependent halogenases acts on freestanding substrates. Nat. Chem. Biol. 2014, 10, 921–923. [CrossRef][PubMed] 19. Gloaguen, V.; Morvan, H.; Hoffmann, L.; Catherine, O.S.; Kraemer, M.; Krausz, P.Bioactive capsular polysaccharide from the thermophilic cyanophyte/cyanobacterium Mastigocladus laminosus—Cytotoxic properties. Planta Med. 2007, 73, 1402–1406. [CrossRef][PubMed] 20. Prinsep, M.R.; Moore, R.E.; Levine, I.A.; Patterson, G.M. Westiellamide, a bistratamide-related cyclic peptide from the blue-green alga Westelliopsis prolifica. J. Nat. Prod. 1992, 55, 140–142. [CrossRef][PubMed] 21. Stratmann, K.; Burgoyne, D.L.; Moore, R.E.; Patterson, G.M.L.; Smith, C.D. Hapalosin, a cyanobacterial cyclic depsipeptide with multidrug-resistance reversing activity. J. Org. Chem. 1994, 59, 7219–7226. [CrossRef] 22. Kang, H.S.; Krunic, A.; Orjala, J. Stigonemapeptin, an Ahp-containing depsipeptide with elastase inhibitory activity from the bloom-forming freshwater cyanobacterium Stigonema sp. J. Nat. Prod. 2012, 75, 807–811. [CrossRef][PubMed] 23. Ogino, J.; Moore, R.E.; Patterson, G.M.; Smith, C.D. Dendroamides, new cyclic hexapeptides from a blue-green alga. Multidrug resistance reversing activity of dendroamide A. J. Nat. Prod. 1995, 59, 581–586. [CrossRef][PubMed] 24. Foster, M.P.; Concepcion, G.P.; Caraan, G.B.; Ireland, C.M. Bistratamides C and D. Two new oxazole-containing cyclic hexapeptides isolated from a Philippine Lissoclinum bistratum ascidian. J. Org. Chem. 1992, 57, 6671–6675. [CrossRef] 25. Todorova, A.K.; Jüttner, F. Nostocyclamide: A new macrocyclic, thiazole-containing allelochemical from Nostoc sp. 31. J. Org. Chem. 1995, 60, 7891–7895. [CrossRef] 26. Jüttner, F.; Todorova, A.K.; Walch, N.; von Philipsborn, W. Nostocyclamide M: A cyanobacterial cyclic peptide with allelopathic activity from Nostoc 31. Phytochemistry 2001, 57, 613–619. [CrossRef] 27. Hagmann, L.; Jüttner, F. Fischerellin A, a novel photosystem-II inhibiting allelochemical of the cyanobacterium, Fischerella muscicola, with antifungal and herbicidal activity. Tetrahedron Lett. 1996, 37, 6539–6542. [CrossRef] Mar. Drugs 2016, 14, 73 26 of 28

28. Papke, U.; Gross, E.M.; Francke, W. Isolation, identification and determination of the absolute configuration of fischerellin B. A new algicide from the freshwater cyanobacterium Fischerella muscicola. Tetrahedron Lett. 1997, 38, 379–382. [CrossRef] 29. Moore, R.E.; Cheuk, C.; Patterson, G.M.L. Hapalindoles: New alkaloids from the blue-green alga Hapalosiphon fontinalis. J. Am. Chem. Soc. 1984, 106, 6456–6457. [CrossRef] 30. Moore, R.E.; Cheuk, C.; Yang, X.-Q.G.; Patterson, G.M.L.; Bonjouklian, R.; Smitka, T.; Mynderse, J.S.; Foster, R.S.; Jones, N.D.; Swartzendruber, J.K.; et al. Hapalindoles, antibacterial and antimycotic alkaloids from the cyanophyte Hapalosiphon fontinalis. J. Org. Chem. 1987, 52, 1036–1043. [CrossRef] 31. Asthana, R.K.; Srivastava, A.; Singh, A.P.; Deepali; Singh, S.P.; Nath, G.; Srivastava, R.; Srivastava, B.S. Identification of an antimicrobial entity from the cyanobacterium Fischerella sp. isolated from the bark of Azadirachta indica (Neem) tree. J. Appl. Phycol. 2006, 18, 33–39. [CrossRef] 32. Becher, P.G.; Keller, S.; Jung, G.; Süssmuth, R.D.; Jüttner, F. Insecticidal activity of 12-epi-hapalindole J isonitrile. Phytochemistry 2007, 68, 2493–2497. [CrossRef][PubMed] 33. Moore, R.E.; Yang, X.-Q.G.; Patterson, G.M.L.; Bonjouklian, R.; Smitka, T.A. Hapalonamides and other oxidized hapalindoles from Hapalosiphon fontinalis. Phytochemistry 1989, 28, 1565–1567. [CrossRef] 34. Kim, H.; Lantvit, D.; Hwang, C.H.; Kroll, D.J.; Swanson, S.M.; Franzblau, S.G.; Orjala, J. Indole alkaloids from two cultured cyanobacteria, Westelliopsis sp. and Fischerella muscicola. Bioorg. Med. Chem. 2012, 20, 5290–5295. [CrossRef][PubMed] 35. Schwartz, R.E.; Hirsch, C.F.; Spring, J.P.; Pettibone, J.; Zink, D.L. Unusual cyclopropane-containing hapalindolinones from a cultured cyanobacterium. J. Org. Chem. 1987, 52, 3704–3706. [CrossRef] 36. Smitka, T.A.; Bonjouklian, R.; Doolin, L.; Jones, N.D.; Deeter, J.B.; Yoshida, W.Y.; Prinsep, M.R.; Moore, R.E.; Patterson, G.M.L. Ambiguine isonitriles, fungicidal hapalindole-type alkaloids from three genera of blue-green algae belonging to the Stigonemataceae. J. Org. Chem. 1992, 57, 857–861. [CrossRef] 37. Mo, S.; Krunic, A.; Chlipala, G.; Orjala, J. Antimicrobial ambiguine isonitriles from the cyanobacterium Fischerella ambigua. J. Nat. Prod. 2009, 72, 894–899. [CrossRef][PubMed] 38. Raveh, A.; Carmeli, S. Antimicrobial ambiguines from the cyanobacterium Fischerella sp. collected in Israel. J. Nat. Prod. 2007, 70, 196–201. [CrossRef][PubMed] 39. Walton, K.; Gantar, M.; Gibbs, P.D.; Schmale, M.C.; Berry, J.P. Indole alkaloids from Fischerella inhibit vertebrate development in the zebrafish (Danio rerio) embryo model. Toxins 2014, 6, 3568–3581. [CrossRef] [PubMed] 40. Huber, U.; Moore, R.E.; Patterson, G.M.L. Isolation of a nitrile-containing indole alkaloid from the terrestrial blue-green alga Hapalosiphon delicatulus. J. Nat. Prod. 1998, 61, 1304–1306. [CrossRef][PubMed] 41. Steele, D.; Berry, J.P. Isolation, chemical characterization and comparative toxicology of structurally diverse indole alkaloids from Fischerella as teratogens. Mar. Drugs forthcoming. 42. Mo, S.; Krunic, A.; Santarsiero, B.D.; Franzblau, S.G.; Orjala, J. Hapalindole-related alkaloids from the cultured cyanobacterium Fischerella ambigua. Phytochemistry 2010, 71, 2116–2123. [CrossRef][PubMed] 43. Park, A.; Moore, R.E.; Patterson, G.M.L. Fischerindole L, a new isonitrile from the terrestrial blue-green alga Fischerella muscicola. Tetrahedron Lett. 1992, 33, 3257–3260. [CrossRef] 44. Kim, H.; Krunic, A.; Lantvit, D.; Shen, Q.; Kroll, D.J.; Swanson, S.M.; Orjala, J. Nitrile-containing fischerindoles from the cultured cyanobacterium Fischerella sp. Tetrahedron 2012, 68, 3205–3209. [CrossRef] [PubMed] 45. Stratmann, K.; Moore, R.E.; Bonjouklian, R.; Deeter, J.B.; Patterson, G.M.L.; Shaffer, S.; Smith, C.D.; Smitka, T.A. Welwitindolinones, unusual alkaloids from the blue-green algae Hapalosiphon welwitschii and Westiella intricate: Relationship to the fischerindoles and hapalindoles. J. Am. Chem. Soc. 1994, 116, 9935–9942. [CrossRef] 46. Wood, J.L. Total synthesis: Welwitindolinone is well worth it. Nat. Chem. 2012, 4, 341–343. [CrossRef][PubMed] 47. Bhat, V.; Dave, A.; MacKay, J.A.; Rawal, V.H. The chemistry of hapalindoles, fischerindoles, ambiguines and welwitindolinones. Alkaloids Chem. Biol. 2014, 73, 65–160. [PubMed] 48. Jimenez, J.I.; Huber, U.; Moore, R.E.; Patterson, G.M. Oxidized welwitindolinones from terrestrial Fischerella sp. J. Nat. Prod. 1999, 62, 569–572. [CrossRef][PubMed] 49. Hilliwig, M.L.; Fuhrman, H.A.; Ittiiamornkul, K.; Sevco, T.J.; Kwak, D.H.; Liu, X. Identification and characterization of a welwitindolinone alkaloid biosynthetic gene cluster in the Stigonematalean cyanobacterium Hapalosiphon welwitschii. ChemBioChem 2014, 15, 665–669. [CrossRef][PubMed] Mar. Drugs 2016, 14, 73 27 of 28

50. Knowles, C.J. Microorganisms and cyanide. Bacteriol. Rev. 1976, 40, 662–680. 51. Pistorius, E.K.; Voss, H. Some properties of a basic L-amino acid oxidase from Anacystis nidulans. Biochim. Biophys. Acta 1980, 611, 227–240. [CrossRef] 52. Bornemann, V.; Patterson, G.M.L.; Moore, R.E. Isonitrile biosynthesis in the cyanophyte Hapalosiphon fontinalis. J. Am. Chem. Soc. 1988, 110, 2339–2340. [CrossRef] 53. Garson, M.J.; Simpson, J.S. Marine isocyanides and related natural products—Structure, biosynthesis and ecology. Nat. Prod. Rep. 2004, 21, 164–179. [CrossRef][PubMed] 54. Hilliwig, M.L.; Zhu, Q.; Liu, X. Biosynthesis of ambiguine indole alkaloids in cyanobacterium Fischerella ambigua. ACS Chem. Biol. 2014, 9, 372–377. [CrossRef][PubMed] 55. Liu, X.; Hilliwig, M.L.; Koharudin, L.M.; Gronenborn, A.M. Unified biogenesis of ambiguine, fischerindole, hapalindole and welwitindolinone: Identification of a monogeranylated indolenine as a cryptic common biosynthetic intermediate by an unusual magnesium-dependent aromatic prenyltransferase. Chem. Commun. 2016, 52, 1737–1740. [CrossRef][PubMed] 56. Richter, J.M.; Ishihara, Y.; Masuda, T.; Whitefield, B.W.; Llamas, T.; Pohjakallio, A.; Baran, P.S. Enantiospecific total synthesis of the hapalindoles, fischerindoles and welwitindolinones via a redox economic approach. J. Am. Chem. Soc. 2008, 130, 17938–17954. [CrossRef][PubMed] 57. Micallef, M.L.; Sharma, D.; Bunn, B.M.; Gerwick, L.; Viswanathan, R.; Moffitt, M.C. Comparative analysis of hapalindole, ambiguine and welwitindolinone gene clusters and reconstitution of indole-isonitrile biosysynthesis from cyanobacteria. BMC Microbiol. 2014, 14, 213. [CrossRef][PubMed] 58. Li, S.; Lowell, A.N.; Yu, F.; Raveh, A.; Newmister, S.A.; Bair, N.; Schaub, J.M.; Williams, R.M.; Sherman, D.H. Hapalindole/ambiguine biogenesis is mediated by a Cope Rearrangement, C-C bond-forming cascade. J. Am. Chem. Soc. 2015, 137, 15366–15369. [CrossRef][PubMed] 59. Cagide, E.; Becher, P.G.; Louzao, M.C.; Espiña, B.; Vieytes, M.R.; Jüttner, F.; Botana, L.M. Hapalindoles from the cyanobacterium Fischerella: Potential sodium channel modulators. Chem. Res. Toxicol. 2014, 27, 1696–1706. [CrossRef][PubMed] 60. Becher, P.G.; Jüttner, F. Insecticidal compounds of the biofilm-forming cyanobacterium Fischerella sp. (ATCC 43239). Environ. Toxicol. 2005, 20, 363–372. [CrossRef][PubMed] 61. Walton, K.E. Identification, Isolation and Characterization of Developmental Toxins from the Cyanobacterium Fischerella 52–1 Using the Zebrafish (Danio rerio) Embryo Model. Master’s Thesis, Florida International University, Miami, FL, USA, 30 March 2012. 62. Doan, N.T.; Rickards, R.W.; Rotschild, J.M.; Smith, G.D. Allelopathic actions of the alkaloids 12-epi-hapalindole E isonitrile and calothrixin A from cyanobacteria of the genera Fischerella and Calothrix. J. Appl. Phycol. 2000, 12, 409–416. [CrossRef] 63. Etchagaray, A.; Rabello, E.; Dieckmann, R.; Moon, D.H.; Fiore, M.F.; von Döhren, H.; Tsai, S.M.; Neilan, B.A. Algicide production by the filamentous cyanobacterium Fischerella sp. CENA 19. J. Appl. Phycol. 2004, 16, 237–243. [CrossRef] 64. Doan, N.T.; Stewart, P.R.; Smith, G.D. Inhibition of bacterial RNA polymerase by the cyanobacterial metabolite 12-epi-hapalindole E isonitrile and calothrixin A. FEMS Microbiol. Lett. 2001, 196, 135–139. [CrossRef][PubMed] 65. Acuña, U.M.; Zi, J.; Orjala, J.; Carcache de Blanco, E.J. Ambiguine I isonitrile from Fischerella ambigua induces caspase-independent cell death in MCF-7 hormone dependent breast cancer cells. Int. J. Cancer Res. 2015, 49, 1655–1662. 66. Koodkaew, I.; Sunohara, Y.; Matsuyama, S.; Matsumoto, H. Isolation of ambiguine D isonitrile from Hapalosiphon sp. and characterization of its phytotoxic activity. Plant Growth Regul. 2012, 68, 141–150. [CrossRef] 67. Zhang, X.; Smith, C.D. Microtubule effects of welwistatin, a cyanobacterial indolinone that circumvents multiple drug resistance. Mol. Pharmacol. 1996, 49, 288–294. [PubMed] 68. Smith, C.D.; Zilfou, J.T.; Stratmann, K.; Patterson, G.M.; Moore, R.E. Welwitindolinone analogues that reverse P-glycoprotein-mediated multiple drug resistance. Mol. Pharmacol. 1995, 47, 241–247. [PubMed] 69. Klein, D.; Daloze, D.; Braekman, J.C.; Hoffmann, L.; Demoulin, V. New hapalindoles from the cyanophyte Hapalosiphon laingii. J. Nat. Prod. 1995, 58, 1781–1785. [CrossRef] 70. Berry, J.P.; Gantar, M.; Gibbs, P.D.L.; Schmale, M.C. The zebrafish (Danio rerio) embryo as a model system for identification and characterization of developmental toxins from marine and freshwater microalgae. Comp. Biochem. Physiol. C Toxicol. Pharmacol. 2007, 145, 61–72. [CrossRef][PubMed] Mar. Drugs 2016, 14, 73 28 of 28

71. Wiley, F.E.; Wilde, S.B.; Birrenkott, A.H.; Williams, S.K.; Murphy, T.M.; Hope, C.P.; Bowerman, W.W.; Fischer, J.R. Investigation of the link between avian vacuolar myelinopathy and a novel species of cyanbacteria through laboratory feeding trails. J. Wildl. Dis. 2007, 43, 337–344. [CrossRef][PubMed] 72. Gantar, M.; Berry, J.P.; Thomas, S.; Wang, M.; Perez, R.; Rein, K.S. Allelopathic activity among cyanobacteria and microalgae isolated from Florida freshwater habitats. FEMS Microbiol. Ecol. 2008, 64, 55–64. [CrossRef] [PubMed] 73. Gross, E.M.; Wolk, C.P.; Jüttner, F. Fischerellin, a new allelochemical from the freshwater cyanobacterium Fischerella muscicola. J. Phycol. 1991, 27, 686–692. [CrossRef] 74. Srivastava, A.; Jüttner, F.; Strasser, R.J. Action of the allelochemical, fischerellin A, on photosystem II. Biochim. Biophys. Acta Bioenerg. 1998, 1364, 326–336. [CrossRef] 75. Berry, J.P.; Gantar, M.; Perez, M.H.; Berry, G.; Noriega, F. Cyanobacterial toxins as allelochemicals with potential applications as algaecides, herbicides and insecticides. Mar. Drugs 2008, 6, 117–146. [CrossRef] [PubMed] 76. Birrenkott, A.H.; Wilde, S.B.; Hains, J.J.; Fischer, J.R.; Murphy, T.M.; Hope, C.P.; Parnell, P.G.; Bowermann, W.W. Establishing a food-chain link between aquatic plant material and avian vacuolar myelinopathy in mallards (Anas platyrhynchos). J. Wildlif. Dis. 2004, 40, 485–492. [CrossRef][PubMed] 77. Wilde, S.B.; Murphy, T.M.; Hope, C.P.; Habrun, S.K.; Kempton, J.; Birrenkott, A.; Wiley, F.; Bowermann, W.W.; Lewitus, A.J. Avian vacuolar myelinopathy linked to exotic aquatic plants and novel cyanobacterial species. Environ. Toxicol. 2005, 20, 348–353. [CrossRef][PubMed] 78. Wiley, F.E.; Twiner, M.J.; Leighfield, T.A.; Wilde, S.B.; Van Dolah, F.M.; Fischer, J.R.; Bowerman, W.W. An extract of Hydrilla verticillata and associated epiphytes induces avian vacuolar myelinopathy in laboratory mallards. Environ. Toxicol. 2009, 24, 362–368. [CrossRef][PubMed]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons by Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).