CYANOBACTERIAL SPECIALIZED METABOLITES:
BIOSYNTHESIS, BIOACTIVITY AND STRUCTURE
Lassi Matti Petteri Heinilä
Department of Microbiology Faculty of Agriculture and Forestry
University of Helsinki, Finland
DOCTORAL DISSERTATION
To be presented for public discussion with the permission of the Faculty of Agriculture and Forestry of the University of Helsinki, in Auditorium 2041 at Biocenter 2, Viikinkaari 5, Helsinki on the 7th of
December, 2021 at 10 o’clock.
Helsinki 2021
2 Supervisors Principal supervisor
Professor Kaarina Sivonen Department of Microbiology University of Helsinki, Finland Docent David P. Fewer Department of Microbiology University of Helsinki, Finland Docent Jouni Jokela
Department of Microbiology University of Helsinki, Finland Pre-examiners Professor Nadine Ziemert
Interfaculty Institute of Microbiology and Infection Medicine University of Tübingen, Germany
Professor Shmuel Carmeli
Raymond and Beverly Sackler School of Chemistry and Faculty of Exact Sciences
Tel Aviv University, Israel Thesis committee Professor Jari Yli-Kauhaluoma
Division of Pharmaceutical Chemistry and Technology University of Helsinki, Finland
Professor Päivi Tammela
Division of Pharmaceutical Biosciences University of Helsinki, Finland
Opponent Professor Nicolas Inguimbert
Centre of Island Research and Environmental Observatory University of Perpignan, France
Grading committee Professor Per Saris Faculty representative Department of Microbiology
University of Helsinki, Finland Custos Professor Kaarina Sivonen
Department of Microbiology University of Helsinki, Finland
Dissertationes Schola Doctoralis Scientiae Circumiectalis, Alimentariae, Biologicae ISBN 978-951-51-7747-6 (paperback) and ISBN 978-951-51-7748-3 (PDF) ISSN 2342-5423 (print) and ISSN 2342-5431 (online)
http://ethesis.helsinki.fi Unigrafia
Helsinki 2021
3 Table of Contents
List of publications ... 4
Related publications ... 4
Abbreviations ... 5
Abstract ... 6
Tiivistelmä ... 7
1. Introduction ... 8
1.1. Cyanobacteria ... 8
1.2. Specialized metabolites in cyanobacteria ... 8
1.3. Dolastatin and drug development... 9
1.4. Antifungal compounds ... 10
1.4.1. Laxaphycins ... 13
1.5. Protease inhibitors ... 17
1.6. Nodularin and microcystin ... 19
1.7. Biosynthesis of cyanobacterial compounds ... 20
2. Study aims ... 24
3. Summary of methods ... 25
4. Summary of results and discussion ... 26
4.1. Laxaphycin structures and bioactivity (I, II) ... 26
4.2. Laxaphycin biosynthesis (I, II) ... 28
4.3. Pseudospumigins (III) ... 34
4.4. Nodularin (III) ... 36
5. Conclusions ... 38
6. Acknowledgements ... 40
7. References ... 41
4 List of publications
This thesis is based on the following publications
I Heinilä LMP, Fewer DP, Jokela JK, Wahlsten M, Jortikka A and Sivonen K. 2020. Shared PKS module in biosynthesis of synergistic laxaphycins. Frontiers in Microbiology 11:2173.
DOI: 10.3389/fmicb.2020.578878
II Heinilä LMP, Jokela J, Wahlsten M, Fewer DP, Ouyang X, Permi P, Jortikka A and Sivonen K. 2021. The structure and biosynthesis of heinamides A1-A3 and B1-B5, antifungal members of lipopeptide family laxaphycins. Organic & Biomolecular Chemistry 19:5577-5588. DOI: 10.1039/D1OB00772F
III Jokela J,Heinilä LMP, Shishido TK, Wahlsten M, Fewer DP, Fiore MF, Wang H, Haapaniemi E, Permi P, Sivonen K. 2017. Production of high amounts of hepatotoxin nodularin and new protease inhibitors pseudospumigins by the Brazilian benthicNostoc sp.
CENA 543. Frontiers in Microbiology 8:1963. DOI: 10.3389/fmicb.2017.01963
The author’s contribution
I Lassi Matti Petteri Heinilä participated in the design of the study and experimental work (except MS and NMR), performed the bioinformatics analysis and wrote the article.
II Lassi Matti Petteri Heinilä participated in the design of the study and performed most of the experimental work (except MS and NMR), performed the bioinformatics analysis and wrote the article.
III Lassi Matti Petteri Heinilä executed most of the experimental work (except MS and NMR) and participated in writing the article.
Related publications
IV Fewer DP, Jokela J,Heinilä LMP, Aesoy R, Sivonen K, Galica T, Hrouzek P and Herfindal L. 2021. Chemical diversity and cellular effects of antifungal cyclic lipopeptides from cyanobacteria. Physiologica Plantarum 173:639–650. DOI: 10.1111/ppl.13484
5 Abbreviations
AA Amino acid
BGC Biosynthetic gene cluster
IC50 Half maximal inhibitory concentration
MS Mass spectrometry
NMR Nuclear magnetic resonance NRP Non-ribosomal peptide
NRPS Non-ribosomal peptide synthetase
PK Polyketide
PKS Polyketide synthase SM Specialized metabolite Catalytic domains
A Adenylation
ACP Acyl carrier protein AMT Aminotransferase
AT Acyltransferase
C Condensation
DH Dehydratase
E Epimerization
ER Enoyl reductase
FAAL Fatty acyl-AMP ligase
KR Ketoreductase
KS Ketosynthase
MT Methyl transferases PCP Peptidyl carrier protein
TE Thioesterase
Amino acids
Ada 3-aminodecanoic acid Aoa 3-aminooctanoic acid Argal Argininal
Choi 2-carboxy-6-hydroxyoctahydroindole cHse O-carbamoyl homoserine
Hph Homophenylalanine
Hse Homoserine
Hty Homotyrosine
Leu Leucine
MePro (2S,4S)-4-methylproline NMe-Ile N-Methyl Isoleucine
OHAoa 5-hydroxy-3-aminooctanoic acid OHAsn 3-hydroxy-D-asparagine OHHse 3-hydroxy-D-homoserine OHLeu 3-hydroxy-D-leucine OHMePro 3-hydroxy-4-methyl proline OHPro (2S,4R)-4-hydroxy-L-proline
6 Abstract
Cyanobacteria produce a variety of toxins and a diversity of other specialized metabolites.
Specialized metabolites are compounds produced by an organism to interact with the environment and provide protection against competitors, predators, or abiotic factors. The biosynthetic
pathways for generating specialized metabolites are typically encoded in compact gene clusters that encode multiple biosynthetic enzymes. The structure and biosynthesis of all major
cyanobacterial toxins have been resolved, but new compounds with variable functions are continuously discovered and their biosynthetic origins elucidated. Cyanobacterial specialized metabolites are widely held to have great potential in the pharmaceutical industry given the increasing need for new drugs that target infectious disease and cancer. A better understanding of the chemical structure of the compounds facilitates discovery of their biological targets and their ecological role. The aim of this study was to discover new potential drug leads from cyanobacteria, focusing on antifungal compounds, and describe their structure, activity, and biosynthetic origins.
Laxaphycins are unusual specialized metabolites that consist of two distinct macrocyclic
lipopeptides with either 11 or 12 amino acids. They are known to have synergistic antiproliferative and antifungal activities but unknown biosynthetic origins. Here, new chemical variants of laxaphycin family specialized metabolites were discovered fromNostoc sp. UHCC 0702 and Scytonema hofmanniiPCC 7110. The laxaphycin biosynthetic gene cluster was discovered, organized as a branching pathway, with initiating enzymes participating in the biosynthesis of both different lipopeptide groups. The biosynthetic gene cluster was described in bothScytonema hofmanniiPCC 7110 andNostocsp. UHCC 0702. New laxaphycin variants heinamides were discovered with unforeseen structural moieties and I present predictions for their origins. I confirmed that heinamides also displayed synergistic antifungal activity. The connection of the 11- and 12- amino acid residue compounds is also evident at the genetic level with the common biosynthetic enzymes of the synergistic compounds.
Aeruginosins are common cyanobacterial tetrapeptides with inhibitory activity against serine proteases. Trypsin isoforms have recently been studied as a target in cancer treatment. Here pseudospumigins, new aeruginosin variants, were discovered fromNostocsp. CENA 543.
Pseudospumigins are produced through a PKS/NRPS pathway similar to known aeruginosin biosynthetic gene clusters. Pseudospumigin A acts as weak trypsin inhibitor, with time dependent IC50 value of 4.5 ȝM. Nodularin-R, a cyanotoxin, was also found from the same strain. The nodularin concentration was much higher than what has been seen in aNostoc strain before, comparable to nodularin concentrations in the most common nodularin producersNodularia spumigena.
This study describes new cyanobacterial specialized metabolites and biosynthetic enzymes for their biosynthesis, broadening the knowledge in areas of novel structural elements, biosynthetic pathways and biological activity. The structural and activity information can help in function prediction and rational design of drug candidates or guide the screening for specific targets. The genetic information can be used in mining genomes for discovering new compounds and predicting products for cryptic biosynthetic gene clusters. Probable applications lie also in the emerging fields of combinatory biosynthesis and synthetic biology to produce engineered compounds in biological systems.
7 Tiivistelmä
Syanobakteerit, eli sinilevät, tuottavat monenlaisia myrkkyjä ja muita biologisesti aktiivisia
luonnonyhdisteitä. Luonnonyhdisteitä tuottavat organismit pyrkivät niiden avulla vuorovaikutukseen ympäristönsä kanssa esimerkiksi suojautuakseen saalistajilta, kilpailijoilta tai elottomilta tekijöiltä kuten säteilyltä. Näiden yhdisteiden tuottamiseen tarvittava geneettinen informaatio on tavallisesti koodattu biosynteettisiin geeniryppäisiin, jotka tuottavat useita entsyymejä yhdisteen kokoamiseksi.
Yleisimpien syanobakteerien tuottamien myrkkyjen ja yhdisteiden rakenteet ja biosynteesireitit on selvitetty, mutta uusia yhdisteitä ja niiden geneettistä alkuperää kuvataan jatkuvasti.
Syanobakteerien tuottamissa luonnonyhdisteissä nähdään potentiaalia lääketeollisuudessa.
Erilaisten tartuntatautien ja syöpien hoitotarpeen kasvaessa ja mikrobilääkeresistenssin yleistyessä etsitään uusia lääkkeitä. Syanobakteerien tuottamien yhdisteiden monimuotoisuudessa nähdään potentiaalia lääkekehityksessä tarvittavien johtolankamolekyylien löytämiseen. Tämän tutkimuksen päämääränä oli löytää syanobakteereista uusia johtolankamolekyylejä ja kuvata näiden rakenne, biologinen aktiivisuus ja biosynteettinen alkuperä.
Löysin uusia laksafysiini-perheen yhdisteitä bakteereista Nostoc sp. UHCC 0702 ja Scytonema hofmannii PCC 7110. Laksafysiinit ovat joukko syanobakteerien tuottamia luonnonyhdisteitä, joilla on sienten ja syöpäsolulinjojen kasvua estäviä ominaisuuksia. Laksafysiinit ovat makrosyklisiä lipopeptidejä, jotka jaetaan kahteen rakenteellisesti eroavaan ryhmään, 11- ja 12-aminohapon ryhmiin. Näiden ryhmien välillä on vahva synergistinen muiden organismien kasvua estävä vaikutus. Tässä tutkimuksessa laksafysiinien biosynteesireitti kuvataan ensimmäistä kertaa.
Laksafysiinien tuotanto tapahtuu haarautuvalla biosynteesireitillä, jossa biosynteesin käynnistävät entsyymit osallistuvat molempien yhdisteryhmien tuotantoon. Biosynteettinen geenirypäs kuvattiin molemmista tutkituista laksafysiinejä tuottaneista bakteerikannoista. Uusi ryhmä laksafysiinien rakenteellisia variantteja löydettiin: heinamidit. Heinamideilla on ennennäkemättömiä rakenteellisia ominaisuuksia, joiden alkuperälle esitän ennusteet. Uudet variantit estävät sienten kasvua synergisesti, kuten aiemmin kuvatut laksafysiinit. 11-ja 12-aminohappotähteen yhdisteiden yhteisvaikutus linkittyy mielenkiintoisesti perimätasolla, kun yhdisteet jakavat osan
biosynteesireiteistään. Aeruginosiinit ovat yleisiä syanobakteerien tuottamia tetrapeptidejä, joilla on trypsiinin kaltaisia entsyymejä inhiboiva vaikutus. Trypsiiniä on viime aikoina tutkittu mahdollisena lääkeaineiden kohteena syöpähoidoissa. Tässä työssä löysimme uusia aeruginosiinien
rakenteellisia variantteja, jotka nimesimme pseudospumigiineiksi. Pseudospumigiinit löydettiin bakteerikannasta Nostoc sp. CENA 543. Pseudospumigiinit tuotetaan bakteerisolussa PKS/NRPS biosynteesireittiä pitkin, joka muistuttaa aeruginosiinien biosynteesiä. Pseudospumigiini A toimii trypsiini-inhibiittorina, jonka aikariippuvainen IC50 arvo on 4,5 μM. Myös Nodulariini-R
maksamyrkkyä löydettiin samasta bakteerikannasta. Kannan nodulariinipitoisuus oli paljon suurempi, kuin Nostoc suvun bakteereista on aiemmin löydetty ja verrattavissa yleisimpien nodulariinituottajien, Nodularia spumigena -lajin bakteerien, tuottamiin pitoisuuksiin.
Tämä väitöskirja kuvaa uusia syanobakteerien tuottamia luonnonyhdisteitä, niiden biologista aktiivisuutta ja niiden biosynteesireitit, laajentaen ymmärtämystämme luonnonyhdisteiden mahdollisuuksista. Tieto yhdisteiden uusista kemiallisista rakenteista ja niiden bioaktiivisuudesta voi auttaa lääkeyhdisteiden rationaalista suunnittelua tai ohjata johtolankamolekyylien seulontaa tarkempiin kohteisiin. Tiedon avulla voimme myös tehdä parempia tulkintoja siitä, mitä hyötyä yhdisteistä on niitä tuottaville bakteereille. Geneettistä tietoa voidaan käyttää geenilouhintaan uusia yhdisteitä etsittäessä ja tuntemattomien biosynteettisten geeniryppäiden tuotteiden
ennustamiseen. Todennäköisesti tietoa tullaan myös käyttämään synteettisen biologian ja kombinatoriaalisen biosynteesin kasvavilla aloilla suunniteltujen yhdisteiden tuottamiseen muunnelluissa organismeissa.
8 1. Introduction
1.1. Cyanobacteria
Cyanobacteria are a bacterial phylum capable of oxygenic photosynthesis (Castenholz, 2015).
They emerged around 2.4 billion years ago (Demoulin et al., 2019). Plants have acquired the capability of oxygenic photosynthesis through endosymbiosis of cyanobacteria, which evolved to chloroplasts in the eukaryotic host organisms (Gould et al., 2008; de Vries and Archibald, 2017).
The emergence of cyanobacteria had a huge impact on the ecology of the planet, as the oxygen production altered the chemical composition of the seas and the atmosphere, causing the great oxygenation event and allowing the evolution of aerobic lifestyles (Knoll, 2008; Och and Shields- Zhou, 2012). Through the billions of years with the whole biosphere evolving around them, cyanobacteria have remained a numerous primary producer in many ecosystems, surviving harsh conditions, competition, predation and parasitism (Whitton and Potts, 2012). Cyanobacteria have great tools for survival as sunlight carbon dioxide, and water are widely available energy and carbon sources (Cohen and Gurevitz, 2006). Also some genera of cyanobacteria are able to fix gaseous nitrogen from the atmosphere, making them an important factor in nitrogen cycle (Cohen and Gurevitz, 2006). Because of this special access to nutrients, cyanobacteria make great partners in mutualistic symbioses with plants and fungi in lichen (Kaasalainen et al., 2012; Adams et al., 2013). One success factor for cyanobacteria is their ability to produce a wide variety of specialized metabolites (Dittmann et al., 2015). Through eons of evolution, cyanobacteria have acquired genetic building blocks to produce biologically active compounds with wide variation in structures and functions. The ecological function is unclear and under scientific interest for many of these compounds (Dittmann et al., 2015). New families of compounds produced by cyanobacteria are constantly discovered (Janssen, 2019; Jones et al., 2021).
1.2. Specialized metabolites in cyanobacteria
Bioactive compounds discovered from living organisms are traditionally called natural products (Newman and Cragg, 2020). However, by definition any compound produced by an organism is a natural product. In pharmaceutical literature, talking about natural products differentiates
compounds of natural origin from synthetic molecules (Newman and Cragg, 2020). Another typical alternative term for the same concept is secondary metabolites. Traditionally cell metabolites are categorized to primary and secondary metabolism, where primary metabolism covers compounds necessary for growth and secondary metabolism compounds present in late growth stages with ecological functions (Price-Whelan et al., 2006). However, the secondary metabolites themselves are crucial for the success and growth of an organism in a competitive natural environment and the functions of primary and secondary metabolites are sometimes overlapping which makes the division to primary and secondary metabolism problematic (Price-Whelan et al., 2006). In microbiology, a more suitable functional term distinguishing compounds with biological activity is specialized metabolites (SMs)(Davies, 2013). SMs are relatively small compared to proteins, which typically have molecular weight of hundreds of kilodaltons, whereas these bioactive compounds have molecular mass in the range of 100-2000 (Davies, 2013). SMs typically have a function in killing a threat, digesting a nutrition source, protecting from radiation, or in cell signaling (Davies, 2013).
Some cyanobacterial SMs are known to be very toxic to humans and mammals such as pets and cattle, which has led to the detailed description of their structure, mechanism of action and
9
biosynthetic pathways (Codd et al., 2005; Sivonen, 2009; Elersek et al., 2017; Chorus and Welker, 2021). The study of harmful algal blooms also generated scientific interest in other cyanobacterial SMs and has broadened to search for compounds with value in pharmaceutical, food or chemical industry as antimicrobial and anticancer agents (Patterson et al., 1994; Kehr et al., 2011; Shishido et al., 2015). Cyanobacteria have been shown to produce a great number of diverse SMs with antibacterial, antifungal, antiproliferative and protease inhibitory activities. An open database of cyanobacterial metabolites CyanoMetDB contains structures of 2010 cyanobacterial products (Jones et al., 2021). The function of these compounds for the cyanobacterial cell is mostly
unknown, but they are generally believed to deter other organisms grazing on cyanobacteria (Cruz- Rivera and Paul, 2007). Regardless of the original function of the compounds, research can find new targets in microbial infections and cancer treatments to use them as drug leads in
pharmaceutical research (Atanasov et al., 2021). Outside clinical use, cyanobacterial SMs could be used in other areas, for example as antifouling agents (Demay et al., 2019) or sunscreens (Gao et al., 2021). As cyanobacteria often live as symbionts with other organisms or have a role as prey for higher lifeforms, many cyanobacterial compounds have been first characterized from the tissues of the symbionts such as sponges (Konstantinou et al., 2018) or the predators such as sea hares (Luesch et al., 2002; Huang and Zimba, 2019). After the characterization of these compounds they have been found in isolated cyanobacterial cultures ecologically related to these organisms (Luesch et al., 2002).
Some cyanobacterial SMs and their semisynthetic or synthetic derivatives have reached clinical trials for pharmaceutical use (Vijayakumar and Menakha, 2015). So far one drug derived from a cyanobacterial lead compound has reached the market, dolastatin.
1.3. Dolastatin and drug development
Probably the best known cyanobacterial compound in drug research is dolastatin 10 (Simmons et al., 2005). Dolastatin 10 was originally isolated from a sea hare and later it was defined that the compound is produced by cyanobacteriumSymplocasp. that was a part of the sea hares diet.
Dolastatin 10 was at the time of its discovery the most potent antiproliferative agent known (Pettit et al., 1987). Dolastatin variants have been found in the nature and synthetic analogues have been developed and tested clinically, but none of them has reached the market as such due to their toxicity to the human body. Dolastatins act here as an example for the drug development process from SM drug leads.
When developing a new drug, there are a few things that need to be taken into account (Atanasov et al., 2021). A set of characteristics for a drug lead are the so-called ADMET properties of a compound, standing for absorption, distribution, metabolism, excretion and toxicity. These properties are assessed to see if a compound reaches the target in the body and is compatible with human physiology. One guideline for choosing drug leads for oral use of drugs is the Lipinski’s rule of five (Lipinski et al., 1997; Doak et al., 2014). If one of these characteristics is not met, a lead had poor chances in developmental process: more than 5 hydrogen bond donors, more than 10 hydrogen bond acceptors and molecular weight over 500 or partition coefficient LogP is under 5.
Only few SMs fulfill these criteria as they are discovered. Typically, a SM may have potent bioactivityin vitro, and clinical potential, but the compound cannot be used as a drug due to poor ADMET properties. Either the compounds are not soluble or otherwise do not reach the target in the body, they may accumulate in the tissues or the compounds are simply too toxic to people and cause more harm than good (Atanasov et al., 2021). Drugs can be however, engineered through modelling or combinatorial synthesis and biosynthesis to make them more suitable for the patient and production (Kim et al., 2015). The production of a suitable drug can be economically not
10
plausible if the biological production level is low and chemical synthesis too complicated (Atanasov et al., 2021).
Brentuximab vedotin is a dolastatin monoclonal antibody conjugate that is used to treat blood cancers Hodgkin lymphoma and systemic anaplastic large cell lymphoma (Figure 1) (Senter and Sievers, 2012). It was designed based on the structure of dolastatin 10 through a multistep process to enhance its physiological properties and delivery to the cancer cells (Doronina et al., 2003). A library of dolastatin analogues was produced through combinatorial chemistry. Among these was synthetic dolastatin analog monomethyl auristatin E (MMAE) that was selected for further development. To bypass the toxicity of auristatin a delivery system with a monoclonal antibody (mAB) was designed to target the cancer cells. The final drug compound comprises four parts.
First the mAB that is designed to identify the target cancer cells. The mAB in connected to a cleavage linker consisting of valine and citrulline, which are cleaved through proteolysis. Third part is ap-aminobenzyloxycarbonyl (PABC) spacer placing the auristatin away from the cleavage site to allow effective proteolysis, and fragmenting in the cleavage process. Finally, the MMAE is released to the target cells. This system allows the delivery of the potent drug in high concentration to the target cell while keeping the concentration in the body at tolerable levels.
Figure 1. Structures of dolastatin 10 and Brentuximab vedotin. Monomethyl auristatin E (MMAE) is enhanced analog of dolastatin 10, with potent antiproliferative activity. To deliver the compound to cancer cells without causing toxic concentration of MMAE in other tissues it has been modified with additional parts.
A monoclonal antibody (mAB) recognizes the cancer cell. Between mAB and MMAE are linker parts to cleave mAB off the product without altering MMAE structure.
1.4. Antifungal compounds
Millions of people worldwide contract invasive fungal infections every year, with often over 50%
mortality rates, even when treated (Brown et al., 2012). Invasive fungal infections are rare in healthy individuals and are typically contracted to immunocompromised patients such as receivers of organ transplants, people with HIV/AIDS or suffering major trauma. Better diagnostics and therapies are needed to fight these infections. The deadly infectious fungi belong to the genera Cryptococcus,Candida,Aspergillus andPneumocystis(Hendrickson et al., 2019; Buda De Cesare
11
et al., 2020). Fungi are a part of normal human microbiome, usually living on the skin, but can act as opportunistic pathogens when the immune system is compromised.
Fungi being eukaryotic organisms are physiologically closer to humans than bacteria, which makes it more difficult to develop agents that would target the pathogen and not the human cells. Clinical treatment of fungal infections currently relies on compounds belonging to the echinocandins, triazoles, polyenes and allyloamines structural families (Perlin et al., 2017). The major targets of antifungal drugs are sterol synthesis in the fungal cell (triazoles and allylamines) and fungal cell wall (echinocandins and polyenes). Other treatments are based on the inhibition of DNA synthesis by flucotysine, a pyrimidine analog. Undesirable interactions with other drugs limit the use of triazoles. Unfortunately, fungi are now gaining resistance to these compounds (WHO, 2014;
Hendrickson et al., 2019; Buda De Cesare et al., 2020).
Cyanobacteria are known producers of antifungal compounds such as laxaphycins, hassallidins, nostofungicidine and cryptophycins (Frankmölle et al., 1992b; Trimurtulu et al., 1994; Kajiyama et al., 1998; Neuhof et al., 2005; Shishido et al., 2015). Cyanobacteria face the threat of fungal infection such as parasitic chytrids, in aquatic environments (Agha et al., 2018). Chytrids are host specific fungi, of which some infect cyanobacteria with their zoospores. Oligopeptides of a Planktothrix strain have been shown to deter chytrids and they are assumed to act as a defense mechanism against the fungi (Agha et al., 2018). On the other hand, some cyanobacteria live in lichen symbioses with fungi and even lichen-associated cyanobacteria have been shown to have antifungal activity (Shishido et al., 2015). The mechanism of action for most of these antifungal activities is unknown. Hassallidin D has been shown to disrupt sterol-containing eukaryotic cell membranes on acute myeloid leukemia cells (MOLM-13), normal rat kidney cells (NRK) andC.
albicanscells (Humisto et al., 2019). None of the cyanobacterial compounds with antifungal activity has reached the market as pharmaceuticals.
Cyanobacterial antifungal compounds are typically macrocyclic polypeptides that feature a fatty acid chain as seen in laxaphycins, schizotrin-like analogs, puwainaphycins, hassallidins, nostofungicidine and anabaenolysins (Jokela et al., 2012; Niedermeyer, 2015; Shishido et al., 2015) (Figure 2, Figure 3). The cyclization of the lipopeptides may happen in several ways through an amino group in the fatty acid ȕ-position to cyclizing the peptide with an amino bond or formation of ester or ether bonds. Other types of structures are not exceptional, like depsipeptide
majusculamides (Carter et al., 1984), peptolide cryptophycin (Trimurtulu et al., 1994), polyketide fischerellins (Hagmann and Jüttner, 1996) and alkaloid hapalindole (Moore et al., 1987) (Figure 3).
12
Figure 2. Antifungal compounds from cyanobacteria including laxaphycins A and B, schizotrin A, puwainaphycin A, hassallidins D variant 2 and nostofungicidine.
13
Figure 3. Antifungal compounds from cyanobacteria including fischerellin A, cryptophycin A, majusculamide C, anabaenolysin A and hapalindole A.
1.4.1. Laxaphycins
Laxaphycins are a family of SMs produced by several genera of cyanobacteria (Cai et al., 2018;
Bornancin et al., 2019). They are the subject of articles I and II of this thesis. Laxaphycins are divided to two structurally distinct groups, 11-residue (A-type) laxaphycins and 12-residue (B-type) laxaphycins (Figure 2, Table 1). These two groups enhance each other’s bioactivity notably in a synergistic manner. Laxaphycin producers typically contain compounds of both groups. The bioactivities described for laxaphycins include growth inhibition of fungal, bacterial and cancer cells.
Laxaphycin family compounds were first mentioned in the literature when the biological activities of hormothamnin A were described in 1989 (Gerwick et al., 1989). In this report hormothamnin A was described antimicrobial to both gram-positiveBacillus subtilis and gram-negativePseudomonas aeruginosa, cytotoxic to cancer cell lines and ichtyotoxic to goldfishCarassius. The name laxaphycin was coined when the bioactivity and structures of laxaphycin A and laxaphycin B isolated fromAnabaena laxawere described (Figure 2), with the discovery of strong synergism between 11- and 12-residue laxaphycins (Frankmölle et al., 1992a; Frankmölle et al., 1992b).
Although A and B type laxaphycin structures have great differences, they are generally referred to as a single family or group due to common producers, synergistic effect and the original naming of laxaphycins A and B (Cai et al., 2018; Bornancin et al., 2019). In the study of Frankmölleet al.
(1992a) all tested fungal strains and cancer cell lines were found sensitive to the laxaphycins. After these experiments, new variants with similar bioactivities have been found in multiple genera of cyanobacteria (Table 1, Table 2). The effects have since been reported for antifungal,
antiproliferative, antibacterial and crustacean toxicity toThamnocephalus platyurus (Bonnard et al., 1997; MacMillan et al., 2002; Grewe, 2005; Bonnard et al., 2007; Maru et al., 2010; Luo et al., 2014; Luo et al., 2015; Dussault et al., 2016; Cai et al., 2018; Bornancin et al., 2019; Alvarino et al., 2020). In a recent study laxaphycin B-type peptides were shown to be cleaved by D-peptidase enzyme produced by sea hareStylocheilus striatus(Darcel et al., 2021). The sea hare feeds on
14
laxaphycin producing cyanobacteria and cleaving the laxaphycin to acyclic form renders it less toxic to the mollusk.
All laxaphycin structures include a ȕ-amino fatty acid with 8-10 carbons, which is the only common structure between the two laxaphycin groups (Figure 2, Table 2). 11-residue laxaphycins are characterized by regions of hydrophobic and hydrophilic regions, while 12-residue laxaphycins have also positions with alternating hydrophobic and hydrophilic residues (Table 1, Table 2).
Typical non-proteinogenic amino acids in 11-residue laxaphycins are Hse2, Dhb3 and OHPro4, Hse5 and in 12-residue laxaphycins OHLeu3, OHleu5, N-Me-Leu7, OHAsn8 and OHPro10 (Figure 2, Table 2). Some acyclic variants have been found, but these are expected to be degradation products or biosynthetic intermediates (Bornancin et al., 2015; Bornancin et al., 2019). Total synthesis of laxaphycin B, lygbyacyclamide A and trichormamide A has been carried out to confirm the chemical structure and to allow future research and drug development through synthetic amino acid mutation studies (Boyaud et al., 2013; Gaillard et al., 2018). Hormothamnin A and laxaphycin A have identical structure with the exception of stereochemistry of Dhb, in laxaphycin A it hasE- Dhb configuration and in hormothamnin AZ-Dhb configuration (MacMillan et al., 2002).
Laxaphycins have been predicted to be products of PKS/NRPS biosynthetic pathway (Bornancin et al., 2015). This was confirmed in the articles I and II of this thesis, where the biosynthetic origins of laxaphycins were described for the first time.
15
Table 1. Chemical diversity of 11-residue laxaphycin structures and producing organisms. 11-residue laxaphycinsAmino acid residue ProducerRef. FA12345678910 Laxaphycin A(3R)-Aoa L-HSeE-Dhb(2S,4R)-4-OHPro L-Hse D-PheD-LeuL-IleD-allo-IleL-LeuGly Anabaena laxa UH FK-1-2,Lyngbya majuscula, Hormothamnion enteromorphoides, Anabaena torulosa
1, 2, 3, 4 Laxaphycin A2(3R)-AoaL-HSeE-Dhb(2S,4R)-4-OHProL-HseD-PheD-LeuL-ValD-allo-IleL-LeuGlyHormothamnion enteromorphoides5 Laxaphycin EAdaHSe E-Dhb4-OHProHSePheLeuIleIleLeuGlyAnabaena laxa UH FK-1-21 Hormothamnin A(3R)-AoaL-HSeZ-Dhb(2S,4R)-4-OHProL-HSeD-PheD-LeuL-IleD-allo-IleL-LeuGlyHormothamnion enteromorphoides Grunow6 Lobocyclamide A(3R)-Aoa L-Ser E-Dhb(2S,4R)-4-OHPro L-Hse D-Tyr D-LeuL-IleD-allo-IleL-LeuGlyLyngbya confervoides7 Trichormamide A(3R)-AoaL-SerL-SerL-ProL-SerD-TyrD-LeuL-IleL-IleL-ProGlyTrichormus sp. UIC 103398, 9 Trichormamide D(3R)-Ada L-GlnE-DhbL-ProL-Ser D-Tyr D-Leu L-ValD-PheL-LeuGlyOscillatoria sp. UIC 1004510 Scytocyclamide AAoaL-GlnE-Dhb(2S,4R)-4-OHProL-HSeD-PheD-LeuL-IleL-allo-IleL-LeuGlyScytonema hofmannii PCC 711011 [l-Val8]laxaphycin A(3R)-Aoa L-HSeE-Dhb(2S,4R)-4-OHPro L-Hse D-PheD-Leu L-ValD-allo-IleL-LeuGlyAnabaena torulosa12 [d-Val9]laxaphycin A(3R)-AoaL-HSeE-Dhb(2S,4R)-4-OHProL-HseD-PheD-LeuL-IleL-ValL-LeuGlyAnabaena torulosa12 Acyclolaxaphycin A(3R)-Aoa L-HSeE-Dhb(2S,4R)-4-OHPro L-Hse D-PheD-LeuL-IleD-allo-IleL-LeuGly-OHAnabaena torulosa12 [des-Gly11] Acyclolaxaphycin A(3R)-AoaL-HSeE-Dhb(2S,4R)-4-OHProL-HseD-PheD-LeuL-IleD-allo-IleL-Leu-OHAnabaena torulosa12 [des-(Leu10-Gly11)] acyclolaxaphycin A(3R)-Aoa L-HSeE-Dhb(2S,4R)-4-OHPro L-Hse D-PheD-LeuL-IleD-allo-Ile-OHAnabaena torulosa12 1 = Frankmölleet al. 1992b, 2 = Bonnardet al. 1997, 3 = Bonnardet al. 2007, 4 = Sullivanet al. 2020, 5 = Caiet al. 2018, 6 = Gerwicket al. 1992, 7 = MacMillanet al. 2002, 8 = Luoet al. 2014, 9 = Gaillardet al. 2018, 10 = Luoet al. 2015, 11 = Grewe 2005, 12 = Bornancinet al. 2019 Aoa – ȕ-aminooctanoic acid, Ada –ȕ-aminodecanoic acid, Hse – Homoserine, Dhb – Dehydrobutyrine, OHPro – hydroxyproline Hydrophobic amino acid Hydrophilic amino acid
16
Table 2. Chemical diversity of 12-residue laxaphycin structures and producing organisms. Amino acid residue 12-residue laxaphycinsFA1234567891011 ProducerRef. Laxaphycin B(3R)-AdaL-Val(2R,3S)-3-OHLeuL-Ala (2R,3S)-3-OHLeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsn L-ThrL-ProD-Leu L-ThrAnabaena laxa UH FK-1-2, Lyngbya majuscula, Anabaena torulosa
1, 2, 3, 4, 5 Laxaphycin B2(3R)-AdaL-Val(2R,3S)-3-OHLeuL-AlaD-LeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsnL-ThrL-ProD-LeuL-ThrAnabaena torulosa3 Laxaphycin B3(3R)-AdaL-Val(2R,3S)-3-OHLeuL-Ala (2R,3S)-3-OHLeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsn L-Thr(2S,4R)-4- OHProD-Leu L-ThrAnabaena torulosa3, 5 Laxaphycin B4(3R)-AdaL-Val(2R,3S)-3-OHLeuL-Hse(2R,3S)-3-OHLeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsnL-Thr(2S,4R)-4- OHProD-LeuL-ThrHormothamnion enteromorphoides6 Laxaphycin B5(3R)-AdaL-Ile(2R,3S)-3-OHLeuL-Val (2R,3S)-3-OHLeuL-GlnL-NMe-IleD-AsnL-ThrL-ProD-Tyr L-ThrPhormidium sp. UIC 104847 Laxaphycin B6(3R)-AdaL-Ile(2R,3S)-3-OHLeuL-ValD-LeuL-GlnL-NMe-IleD-AsnL-ThrL-ProD-TyrL-ThrPhormidium sp. UIC 104847 Laxaphycin DAoaVal3-OHLeuAla3-OHLeuGlnNMe-Ile3-OHAsnThrProLeuThrAnabaena laxaUH FK-1-21 Lobocyclamide B(3R)-AdaL-Val(2R,3S)-3-OHLeuL-Ala(2R,3S)-3-OHLeuD-GlnL-NMe-Ile(2R,3R)-4-OHThrL-Thr(2S,4R)-4- OHProD-LeuL-ThrLyngbya confervoides8 Lobocyclamide C(3R)-AoaL-Val(2R,3S)-3-OHLeuL-Ala (2R,3S)-3-OHLeuD-GlnL-NMe-Ile(2R,3R)-4-OHThrL-Thr(2S,4R)-4- OHProD-Leu L-ThrLyngbya confervoides8 Lyngbyacyclamide A(3R)-AdaL-Val(2R,3S)-3-OHLeuL-HseD-LeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsnL-ThrL-ProD-PheL-ThrLyngbya sp.9, 4 Lyngbyacyclamide B(3R)-AdaL-Val(2R,3S)-3-OHLeuL-HseD-LeuL-Gln L-NMe-Ile(2R,3R)-3-OHAsn L-Thr(2S,4R)-4- OHProD-Phe L-ThrLyngbya sp.9 Trichormamide B(3R)-AdaL-Ile(2R,3S)-3-OHLeuL-Hse(2R,3S)-3-OHLeuL-GlnL-NMe-IleD-SerL-ThrL-ProD-TyrL-ThrTrichormus sp. UIC 1033910 Trichormamide C(3R)-AdaL-Val(2R,3S)-3-OHLeuL-Ala (2R,3S)-3-OHLeuL-GlnL-NMe-IleD-AsnL-ThrL-ProD-Leu L-ThrOscillatoria sp. UIC 1004511 Scytocyclamide BAoaL-Val3-OH-D-LeuL-Ala3-OH-D-LeuL-GlnL-NMe-Ile3-OH-D-AsnL-ThrL-ProD-LeuL-ThrScytonema hofmanniiPCC 711012 Scytocyclamide CAoaL-Val3-OH-D-LeuL-AlaD-LeuL-Gln L-NMe-Ile3-OH-D-AsnL-ThrL-ProD-Leu L-ThrScytonema hofmanniiPCC 711012 Acyclolaxaphycin B(3R)-AdaL-Val(2R,3S)-3-OHLeu-OHH-L-Ala(2R,3S)-3-OHLeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsnL-ThrL-ProD-LeuL-ThrStylocheilus striatus13, 5 Acyclolaxaphycin B3(3R)-AdaL-Val (2R,3S)-3-OHLeu-OH H-L-Ala (2R,3S)-3-OHLeuL-GlnL-NMe-Ile(2R,3R)-3-OHAsn L-Thr(2S,4R)-4- OHProD-Leu L-ThrStylocheilus striatus13, 5 1 = Frankmölleet al. 1992b, 2 = Bonnardet al. 1997, 3 = Bonnardet al. 2007, 4 = Boyaudet al. 2013, 5 = Alvarinoet al. 2020, 6 = Caiet al. 2018, 7 = Sullivanet al. 2020, 8 = MacMillanet al. 2002, 9 = Maruet al. 2010, 10 = Luoet al. 2014, 11 = Luoet al. 2015, 12 = Grewe 2005, 13 = Bornancinet al. 2015 Aoa – ȕ-aminooctanoic acid, Ada –ȕ-aminodecanoic acid, NMe-Ile – N-Methyl Isoleucine, OHAsn – hydroxyasparagine, OHLeu – hydroxyleucine, OHThr – hydroxythreonine. Laxaphycin D and Scytocyclamide B share the same mass and amino acid sequence. The stereochemistry for laxaphycin D is not known and for scytocylamide it is not chemically proven. Hydrophobic amino acid Hydrophilic amino acid Hydrophobic/philic amino acid
17 1.5. Protease inhibitors
Proteases are enzymes whose function is to cleave proteins (Leung et al., 2000). Some of them act as digestive enzymes, but at the cell level they have more sophisticated regulatory functions in gene expression, differentiation and cell death. Protease inhibitors are compounds that block the proteolytic action of proteases. In cancer development, tumor cells stimulate protease expression in healthy cells to favoring tumor expansion (Chlipala et al., 2009; Eatemadi et al., 2017).
Obstructing their activity with protease inhibition can block the metastasis, spread of cancer. Thus, inhibition of proteases is a key target in development of cancer drugs. Protease inhibitors are also widely used to treat viral infections HIV/AIDS and hepatitis C, and have been studied for the antiviral potential on other viruses like coronaviruses SARS (Po-Huang, 2006) and COVID-19 (Sagawa et al., 2020). Proteases are essential for the life cycle of these viruses (Chlipala et al., 2009). The genetic material of the viruses translate to precursor polyproteins, which need to be cleaved by proteases to produce functional mature proteins. Inhibiting the proteases prevents the maturation of the virus (Lv et al., 2015; de Leuw and Stephan, 2017).
Proteases are categorized in six classes based on their mechanism of action: serine, threonine, cysteine, aspartic, glutamic, and metalloproteases (López-Otín and Bond, 2008). The protease inhibitors discussed here affect serine proteases. Over one third of proteases are classified as serine proteases (Di Cera, 2009). Serine proteases cleave a polypeptide with a Ser residue as the nucleophilic amino acid in their active site. Catalytic activity is typically dependent on an Asp-His- Ser catalytic triad, which is characteristic for serine proteases. Serine proteases are highly specific to substrate proteins and peptides. Serine proteases are further classified to 13 clans with most prominent being trypsin-like and subtilisin-like serine proteases (Hedstrom, 2002; Di Cera, 2009;
Ovaere et al., 2009). Members of the trypsin-like family of enzymes are common proteases in many organisms. Trypsin itself is produced in the pancreas of mammals and acts as a digestive enzyme (Figarella et al., 1969).They are also present in other tissues such as epithelial cells (Koshikawa et al., 1998) and they are overexpressed in cancer development (Schilling et al., 2018). The class includes typical metabolic proteases trypsins, chymotrypsin and thrombin, that share common structure and catalytic mechanism, but different targets (Hedstrom, 2002; Di Cera, 2009). Trypsin inhibitors are compounds preventing the action of trypsin and have potential roles in biomedical and agricultural applications (Ramachandran et al., 2016; Shamsi et al., 2016). Trypsin inhibitors have biomedical interest, since trypsin overexpression has been seen as a marker and probable mediator in cancer propagation (Kato et al., 1998). Trypsin inhibitors are also used in agriculture in plant protection against microbes, insects and pests blocking digestive proteases (Shamsi et al., 2016) and also to improve digestibility of livestock feed (Hwang et al., 1977).
Many cyanobacterial SMs have been characterized as protease inhibitors (Huang and Zimba, 2019). Among these aeruginosins, cyanopeptolins, carmaphycins, microviridins (do Amaral et al., 2021) and anabaenopeptins (Zafrir-Ilan and Carmeli, 2010) have been described to have trypsin inhibitory activity (Demay et al., 2019) (Figure 4). The aeruginosins, also known as dysinosins, spumigins, microcins and pseudoaeruginosins are well studied with trypsin inhibitory activity (Fujii et al., 1997; Ersmark et al., 2008; Fewer et al., 2009; Liu et al., 2015; Hasan-Amer and Carmeli, 2017). This activity has gained the attention of synthetic chemists creating synthetic aeruginosin variants to develop more potent trypsin inhibitors (Fukuta et al., 2004). Aeruginosins are tetrapeptides, constructed of a hydroxy acid and three amino acids, with more than hundred structural variants (Dittmann et al., 2015; Jones et al., 2021). Protease inhibitory activity of aeruginosins has been shown to affect also thrombin, plasmin and cathepsin B (Demay et al., 2019).Cathepsins are a group of proteases also linked to cancer propagation (Bian et al., 2016).
These compounds have common features, typically including a guanidine group in end of the peptide, a fatty acid in the other end, and aromatic amino acids in their structures.
18
The cyanobacterial protease inhibitors have been shown to inhibit digestive enzymes of
zooplanktonDaphnia grazing on cyanobacteria and cause their lethal disruption (Rohrlack et al., 2003; Rohrlack et al., 2004; Schwarzenberger et al., 2013). This activity has been shown to protect the cyanobacteria from grazers and to reduce the growth ofDaphnia(Schwarzenbergeret al., 2013). Aeruginosin variants pseudospumigins are studied in article III of this thesis.
Figure 4. Chemical structures of cyanobacterial trypsin inhibitors including spumigin E, aeruginosin NAL2, carmaphycin B, cyanopeptolin 1020, oscillarin, anabaenopeptin MM850 and microviridin 1777.
19 1.6. Nodularin and microcystin
The most notorious cyanobacterial toxins are microcystins and nodularins (Janssen, 2019) (Figure 5). These are hepatotoxins that affect potable water reservoirs and recreational water use (Codd et al., 2005). Deaths of dialysis patients (Carmichael et al., 2001), livestock, pets, animals (Sivonen, 2009) have been reported due to microcystins and nodularins. Microcystins are produced by fresh water cyanobacterial genusMicrocystis, especiallyMicrocystis aeruginosa, and other genera Anabaena, Nostoc, Planktothrix, Anabaenopsisand Hapalosiphon(Codd et al., 2005; Catherine et al., 2017). Nodularins are produced by brackish water planktonic cyanobacterial genusNodularia, especially Nodularia spumigena (Sivonen et al., 1989; Codd et al., 2005).Nodularin has also been found in certainNostoc samples in minor amounts (Gehringer et al., 2012) and in lichen symbionts (Kaasalainen et al., 2012). Microcystin is more relevant to public health since the microcystin producers grow in fresh water contaminating potable water reservoirs and the bulk of research on cyanobacterial SMs focuses in microcystins (Janssen, 2019). Nodularin poisonings through ingestion are rare since brackish water is not suitable for drinking, but cases of animals toxicosis involving by nodularin have been reported (Sivonen et al., 1990; Simola et al., 2012). Nodularin and microcystin share structural similarity and are produced through similar biosynthetic pathways, they are proposed to have common evolutionary history (Moffitt and Neilan, 2004; Rantala et al., 2004). Nodularin and microcystin are hepatotoxic, meaning that the toxins act in the liver. When ingested, they are transported to the liver hepatocytes (Runnegar et al., 1991). In the liver nodularin and microcystin, inhibit protein phosphatases, which leads to phosphorylation of
cytoskeletal proteins, their redistribution and collapse of cytoskeletal actin microfilaments leading to hemorrhage and liver failure (Honkanen et al., 1991; Runnegar et al., 1995; Dawson, 1998). The article III of this thesis describes the nodularin production of aNostocstrain in high concentrations.
Figure 5. Chemical structures of the cyanobacterial hepatotoxins nodularin-R and microcystin-LR.
20
1.7. Biosynthesis of cyanobacterial compounds
Cyanobacterial specialized metabolites are typically produced bypolyketide synthases (PKS), non-ribosomal peptide synthetases (NRPS) or PKS/NRPS hybrids. Hassallidins,
puwainaphycins, hapalosins, cryptophycins, aeruginosins, cyanopeptolins, nodularins and microcystins mentioned in previous chapters are produced through PKS/NRPS hybrid gene clusters (Rouhiainen et al., 2000; Moffitt and Neilan, 2004; Magarvey et al., 2006; Rounge et al., 2007; Ishida et al., 2009; Mareš et al., 2014; Vestola et al., 2014; Micallef et al., 2015). Polyketide synthase (PKS) enzymes involved in SM biosynthesis are modular type I PKS enzymes (Hertweck, 2009). Each module of the enzyme adds a new building block to the polyketide chain using typical substrates acetyl-CoA or malonyl-CoA (Smith and Tsai, 2007). Modules are composed of domains with defined functions, with minimal domains in a PKS module beingacyltransferase (AT), acyl carrier protein (ACP) and ketosynthase (KS) domains (Figure 6). AT domain selects a precursor molecule and activates ACP, which then moves the molecule to KS, which catalyzes the
lengthening condensation reaction (Donadio et al., 1991). Common optional domains modifying the incorporated substrates areketoreductase (KR), dehydratase (DH) and enoyl reductase (ER), which can reduce the carbonyl group to a hydroxyl group and further totrans-enoyl unsaturated bonds and to saturated bonds, respectively (Kehr et al., 2011). Non-ribosomal peptide synthetase (NRPS) enzymes are also structured in modules, with each module incorporating an amino acid to the peptide chain (Guenzi et al., 1998; Marahiel, 2016). Each module includes a set of catalytic domains with different activities (Figure 6). Anadenylation (A) domain identifies and binds a matching amino acid (Conti et al., 1997). The activated amino acid is transferred and covalently bound to apeptidyl carrier protein (PCP) domain (Stachelhaus et al., 1996). Next, the bound amino acid is delivered to acondensation (C) domain, where peptide bond is formed with the peptide carried from the previous module (Stachelhaus et al., 1998). Both ACP and PCP domains bind to the growing product intermediate with a phosphopantetheine (Ppant) arm that moves the product to the catalytic sites. Other relevant domains areepimerization (E) and thioesterase (TE) domains (Stachelhaus and Walsh, 2000; Bruner et al., 2002). Epimerization domains transform L- form amino acids to D-form and TE domains release the product in the end of synthesis cyclizing the product or as a linear peptide.
PKS and NRPS biosynthetic pathways commonly encode enzymes that are responsible for supplying non-proteinogenic amino acids or tailoring enzymes that modify the product (Süssmuth and Mainz, 2017). Such asmethyl transferases (MT), which can add a methyl group to the nitrogen or oxygen of the incorporated amino acid, to a carbonyl group or to the carbon backbone (Velkov and Lawen, 2003; Ansari et al., 2008) (Figure 6). Halogenases may introduce halogens such as chlorine or bromine to the structure (von Elert et al., 2005). Some of the non-proteinogenic amino acids are produced prior to their recognition and activation by A domains, such as MePro present in spumigins, 2-carboxy-6-hydroxyoctahydroindole (Choi) present in aeruginosins, homotyrosine (Hty) and homophenylalanine (Hph) present in anabaenopeptins (Ishida et al., 2007;
Fewer et al., 2009; Lima et al., 2017). These unconventional amino acids in the structure change the affinities of the products and allows them to interact with target compounds in exceptional ways. More uncommon domains in PKS/NRPS hybrids relevant to this study arefatty acyl-AMP ligase (FAAL) that activates a fatty acid to serve as substrate for PKS and initiate a PKS synthesis andaminotransferase (AMT) domain, which adds an amino group to the fatty acid (Aron et al., 2005) (Figure 7). Aminotransferases (AMT) have been described for example in mycosubtilin (Duitman et al., 1999; Aron et al., 2005) microcystin (Tillett et al., 2000), nodularin (Moffitt and Neilan, 2004) puwainaphycin (Mareš et al., 2014) and hassallidin (Vestola et al., 2014)
biosynthesis. The AMTs generally act on a ketone or aldehyde carbonyl group as transaminases replacing the oxygen with an amino group in the ȕ-carbon of the fatty acid. The AMT containing PKS module acts as the last PKS module before transferring the product intermediate to NRPS
