2 Isoquinolines and their significance...3

(1)

Jenni Heikkinen

Biosynthesis of tetrahydroisoquinolines and their intermediates with Bacillus subtilis

Master´s Programme in Chemical, Biochemical and Materials Engineering Major in Biotechnology

Master’s thesis for the degree of Master of Science in Technology submitted for inspection, Espoo, 16 November 2022.

Supervisor Professor Markus Linder Instructor PhD Heli Viskari

(2)

Aalto University, P.O. BOX 11000, 00076 AALTO www.aalto.fi Abstract of master's thesis

Author Jenni Heikkinen

Title of thesis Biosynthesis of tetrahydroisoquinolines and their intermediates with Bacillus subtilis

Degree Programme Master’s programme in Chemical, Biochemical and Materials Engineering Major Biotechnolgy

Thesis supervisor Professor Markus Linder

Thesis advisor(s) / Thesis examiner(s) PhD Heli Viskari

Date 16.11.2022 Number of pages 196 Language English

Abstract

The objective of this Master’s thesis was to produce tetrahydroisoquinolines and their derivatives from bulk and inexpensive starting materials by utilizing a chemoenzymatic route. The aim of the work was to design and produce individually inducible expression system for acetolactate synthase, ω-transaminase, (R)-salsolinol synthase, (S)-norcoclaurine synthase and phenylethanoyl-N-methyltransferase (PNMT) in Bacillus subtilis. In addition, the aim of this work was to model the tertiary structure of the enzymes and proteins used in the work using in-silico methods for future use.

A purchase order was placed for the designed de-novo plasmid which the custom gene synthesis company failed to produce. However, a partially constructed plasmid was acquired which was able to replicate in the host and contained intact sequences for the expression of (S)-norcoclaurine synthase and PNMT under eukaryotic promoters. The enzyme production for these enzymes were analyzed with SDS-PAGE. The results indicated successful enzyme expression for (S)- norcoclaurine synthase under Cu²⁺ inducible eukaryotic promoter, but the expression of PNMT under eukaryotic Fe³⁺ promoter was inconclusive.

The expression system was re-designed and purchased which yielded a generic cloning shuttle vector for Escherichia coli and B. subtilis, capable to integrate desired payload into B. subtilis’

alpha-amylase (AmyE) locus. Genes for expression operons, the first consisting of ω-transaminase and the second consisting (R)-salsolinol synthase and (S)-norcoclaurine synthase were purchased and eventually received and cloned successfully into the shuttle vector successfully after modification of the defective promoter driving the fluorescent protein reporter producing gene.

Gene synthesis companies were not able to produce operon for acetolactate synthase. Also the expression of PNMT was omitted to keep the project in reference time.

Electroporation and thus other planned expression experiments for these two plasmids in B.

subtilis could not be performed due the availability problems of laboratory space.

Realistic tertiary structure models of the enzymes and proteins used in the work were successfully created, which can be used in future work.

Keywords expression system Bacillus subtilis, isoquinoline, tetrahydroisoquinoline, acetolactate synthase, transaminase, norcoclaurine synthase, salsolinol synthase, phenylethanoyl N-methyl transferase

(3)

Tekijä Jenni Heikkinen

Työn nimi Biosynthesis of tetrahydroisoquinolines and their intermediates with Bacillus subtilis Koulutusohjelma Master’s programme in Chemical, Biochemical and Materials Engineering Pääaine Biotechnolgy

Työn valvoja Professori Markus Linder

Työn ohjaaja(t)/Työn tarkastaja(t) Heli Viskari

Päivämäärä 16.11.2022 Sivumäärä 196 Kieli englanti

Tiivistelmä

Tämän diplomityön päämääränä oli tuottaa tetrahydroisokinoliineja ja niiden johdannaisia yleisistä ja edullisista lähtöaineista kemoentsymaattista reittiä hyödyntäen. Työn tavoitteena oli tuottaa asetolaktaattisyntaasia, ω-transaminaasia, (R)-salsolinolisyntaasia, (S)- norkoklauriinisyntaasia ja fenyylietanoyyli-N-metyylitransferaasia ekspressiosysteemillä, jossa työtä varten suunniteltu shuttle-ekspressiovektori Bacillus subtilikselle ja Escherichia colille elektroporatoitiin B. subtilis WB600 bakteerikantaan. Lisäksi työssä oli tavoitteena mallintaa in- silico menetelmiä käyttäen työssä käytettävien entsyymien ja proteiinien tertiääristä rakennetta.

Suunniteltu plasmidi tilattiin ostotyönä, se kyettiin toimittamaan ainoastaan osittan. Osittaisessa plasmidissa oli operonit ainoastaan (S)-norkoklauriinisyntaasin ja PNMT:n ekspressoimiseksi.

Täten kokeet entsyymiekspressiolle voitiin suorittaa ainoastaan edellä mainittujen entsyymien osalta. Entsyymituotantoa analysoitiin SDS-PAGE:lla. Tulokset indikoivat onnistunutta enstyymiekspressiota (S)-norkoklauriinisyntaasille Cu²⁺ promoottorin ajamana, PNMT:n ekspression onnistuminen Fe²⁺ promoottorin ajamana jäi tulosten perusteella epäselväksi.

Ekspressiosysteemi suunniteltiin uudelleen B. subtilikselle ja E. colille geneerisestä kloonaussukkalavektorista ja kahdesta ekspressiokasetista joista ensimmäinen sisälsi geenit ω- transaminaasin ja toinen (R)-salsolinolisyntaasista ja (S)-norkoklauriinisyntaasin ekspressoimiseksi. Ekspressiovektoreiden avulla oli tarkoitus kyetä integroimaan halutut geenioperonit B. subtiliksen alfa-amylaasi (AmyE) -lokukseen. Suunnitellut plasmidit tilattiin ostotyönä.

Ekspressiokasettien kloonaus uudelleensuunniteltuun kloonausvektoriin tehtiin onnistuneesti sen jälkeen, kun kloonaussukkalvektorin fluoresoivaa proteiina tuottavan operonin viallinen promoottori saatiin korjattua. Geenisynteesiyritykset eivät pystyneet tuottamaan operonia asetolaktaattisyntaasiin. PNMT:tä vastaava ekspressio-operoni jätettiin pois, ajankäyttösyistä.

Ekspressiosysteemille suunniteltuja kokeita ei voitu suorittaa loppuun tilankäytöllisten ongelmien vuoksi. Työssä käytetyille entsyymeille luotiin realistiset tertiääriset rakennemallit, joita voidaan käyttää tulevaisuudessa.

Avainsanat ekspressiosysteemi Bacillus subtilis, isokinoliini, tetrahydroisokinolini,

asetolaktaattisyntaasi, transaminaasi, salsolinolsyntaasi, norkoklauriinisyntaasi, fenyylietanoyyli- N-metyylitransferaasi

(4)

Preface and acknowledgments

After getting acquainted with Ville Takio and his previous work related to fluorinated isoquinolines, an interesting proposal was made: let's establish biochemical route for producing these valuable pharmaceuticals from bulk and cheap starting materials. This Master's thesis is about the fruits of this proposal.

I would like to express my gratitude to Heli Viskari, the instructor of the thesis work and Markus Linder, the supervisor of the thesis work. I want to thank BioGarage for providing laboratory facilities for this work, professor Merja Penttilä for establishing BioGarage and enabling this project take place in BioGarage. This would not have been possible without Jenny and Antti Wihuri foundation financing BioGarage.

Furthermore, I want to thank James Evans, the laboratory manager of BioGarage, for all the techniques you taught, for the discussions related to the work and endless patience and help with all the practical problems that occurred throughout the way.

Without the funding from Equinorm Ltd this interesting and practical project would not have been possible. Special thanks goes to Mikko Mantere and especially to Ville Takio for inspiration and encouragement, sharing golden ideas and wisdom, and generous amounts of help when I was distressed with the practicalities, fruitful discussions and persistent believing in this work and me: You are a wonderful instructor. Thank you for trusting me.

I want to thank BioFilia laboratory and generally Aalto University’s school of Chemical Engineering research labs for providing additional facilities for this work – especially Juha Linnekoski and Sesilja Aranko for helping with practical issues.

Thanks goes also to Liam McGuffin, the developer of IntFOLD and ReFOLD3 server and James “Jimmy” Stewart, the developer of MOPAC, for personal communication

(5)

and help with issues in protein modeling. Yvonne Tollander de Balch, thank you for the python code that enabled processing the docking data to CVS file and further to Excel format for the convenience of interpreting the results. I also wish to acknowledge CSC - IT Center for Science, Finland, for generous computational resources. Without the online services made possible by Alexandra Elbakyan, this work would not have been completed on time.

Lastly, I want to express my deepest gratitude to my family and close ones for all their support during these years of study.

(6)

Tetrahydroisoquinolines (THIQ), such as 1-Methyl-1,2,3,4-tetrahydroisoquinoline (1MeTIQ), are substances with eminent pharmacological potential and a wide range of action on the brain (Antkiewicz-Michaluk et al. 2013). Isoquinoline compounds have been investigated extensively since their extraction in the nineteenth century (Finley, 2005). 1MeTiQ shows notable neuroprotective activity (Antkiewicz-Michaluk et al. 2013; Mozdzen et al. 2019), due to which it has gained a huge potential as medication for various neurodegenerative illnesses (Antkiewicz-Michaluk et al. 2013). 1MeTIQ also has considerable potential for combating addictions due its antiaddictive properties and effects on dopamine metabolism (Antkiewicz-Michaluk et al. 2013).

1MeTiQ and its derivatives have been studied as a candidate for the treatment of e.g. Parkinson’s disease (Kurnik-Łucka et al. 2017; Wąsik et al. 2016), substance abuse (Antkiewicz-Michaluk et al. 2013), schizophrenia (Białoń et al. 2020), depression (Mozdzen et al. 2019) just to name a few. Fluorinated 1MeTiQ derivatives have been patented and their potential uses include general neuroprotectants, antiparkinsonians, anti-addictives, analgesics, anticonvulsants, antidepressants and mood stabilizers (WO2011FI50707, 2012). Patents on THIQ have been made with e.g. anticancer activity and antimalarial activity (Singh &

Shah, 2016). 6,7-Dimethoxy-2-phenethyl-1,2,3,4-tetrahydroisoquinoline amides and isoesters have been studied as potential reversers of multidrug resistance for chemotherapy (Braconi et al. 2020).

The manufacturing process of tetrahydroisoquinolines, especially those containing a deactivating group at the 6th carbon require harsh toxic chemicals and reaction conditions e.g. boron trifluoride diethyl etherate (BF3•OEt2). Chiral catalysts or chemical resolution of the end-product are required in the process if an enantiomerically pure product is desired (Gogoi et al. 2019).

(14)

To the best of our knowledge, (R)-salsolinol synthase has never been assessed whether it can catalyze sterically unfavorable ring formation, specifically fluorinated isoquinolines.

The scope of this thesis work is to design and build an expression system, hosted by a genetically modified Bacillus subtilis (B. subtilis), applicable for production of isoquinolines and their intermediates. The feasibility of this expression system is evaluated by characterizing the produced enzymes and measuring the formation of the desired end product. An additional aim of this thesis work is to perform a proof of concept experiment, where a step-by-step chemoenzymatic cascade reaction is performed using the constructed expression system.

In molecular biology, inducible promoters are used to drive protein expression in the presence of a chemical inducer, or to stop the expression of the target protein in the absence of the chemical inducer. By selectively choosing which enzymes to express, and when, a multistep reaction pathway can be constructed.

Such has been accomplished for the organic pigment violacein (Jones et al., 2015).

Inducible promoters are used to induce expression, thus the reaction can be terminated at any stage, enabling the production of any intermediate if desired.

This expression system can also be utilized to produce enzymes that are used as catalysts in this work. This approach could enable a cost-efficient enantioselective route for laboratory to industrial scale production of tetrahydroisoquinolines (and their derivatives) instead of conventional organic chemical synthesis with harsh reaction conditions and reagents.

The results of this thesis work could potentially be utilized to produce pharmaceutical compounds related to this pathway more cost-efficiently and environmentally friendly than with conventional chemical synthesis. Additionally, the constructed expression system can potentially be utilized for the production

(15)

2 Isoquinolines and their significance

2.1 Chemistry of isoquinolines

Isoquinoline (Figure 1) is the trivial name for 2-azanaphthalene, a heterocyclic aromatic organic compound which consists of a benzene ring fused to a pyridine ring (Finely, 2005). 1,2,3,4-tetrahydroisoquinoline (Figure 1) is fully hydrogen saturated isoquinoline, the substitutions on the 4th, 6th and 7th carbons modify the chemical character of the molecule. Substitution on the 6th carbon of tetrahydroisoquinoline – or 3rd carbon of the starting material such as benzaldehyde or phenethylamine – can act as an electron withdrawing or donating group (Bringmann et al. 1991). Fluorine substitution causes deactivating steric effect to the para-carbon and thus hinders the formation of the desired isoquinoline derivatives (Bringmann et al. 1991).

Figure 1. Molecular structures of isoquinoline (left) and 1,2,3,4-tetrahydroisoquinoline (right)

From pharmacological viewpoint, fluorine substitution at the 6th and 7th carbons mimic the dopamine scaffold, and affect the binding affinities of the compounds to target receptors. Fluorine substitution at the metabolic active position is utilized to change chemical properties and biological distribution and

(16)

toxicology and interaction of the pharmacological target of a compound. In general only minimal steric effect is caused by substituting hydrogen with fluorine. The fluorine acts similar to a hydroxyl group as hydrogen-bond acceptor but is inert as hydrogen donor for eg. catechol-O-methyltransferase (COMT) dependent methylation. (Park et al. 2001; WO2011FI50707, 2012; Dalvit et al.

2014).

The 4th carbon substitution of tetrahydroisoquinolines affects mainly adrenergic receptors when the compound is used as a pharmaceutical (WO2011FI50707, 2012). Isoquinolines play an important role in the pharmaceutical chemistry of natural products, including papaverine (Li et al. 2016), salsolinol (Kurnik-Łucka et al. 2017) and 1-methyl-1,2,3,4 tetrahydroisoquinoline (Antkiewicz-Michaluk et al.

2013; Mozdzen et al. 2019).

2.2 Formation pathways of tetrahydroisoquinolines in nature

In nature, enzymes play a key role in tetrahydroisoquinoline synthesis, as in nature tetrahydroisoquinolines are formed by enzymatically catalyzed pathways (Matin et al. 2001; Engel et al. 2003; Li et al. 2016). Several types of Pictet- Spenglerases are known e.g. (R)-salsolinol synthase, (S)-norcoclaurine synthase and strictosidine synthase (O’Connor & McCoy 2006; Roddan et al. 2020) as shown in Figure 2.

(17)

Figure 2. Known Pictet-Spenglerases and their natural substrate scope. Each segment in the depicts an enzyme, cyclicizes condensed amine and an aldehyde or ketone to give appropriate product. (Roddan et al. 2020)

2.3 In vitro enzymatic tetrahydroisoquinoline synthesis

Using biotechnologically-produced enzymes, (S)-norcoclaurine (higenamine) has been successfully produced in a two-step one-pot reaction from tyrosine and dopamine catalyzed by recombinant (S)-norcoclaurine synthase (EC 4.2.1.78) produced in E. coli (Bonamore et al. 2010). The biotransformation starts with the oxidative decarboxylation of tyrosine to generate 4-hydroxyphenylacetaldehyde, followed by the addition of enzyme and dopamine substrate with an antioxidant such as ascorbate being present (Bonamore et al. 2010). Similarly, the synthesis of norcoclaurine is based on the Pictet-Spengler reaction (Bonamore et al. 2010;

Calcaterra et al. 2020). The yield of the synthesis was about 80% and

(18)

successfully utilized in production of trolline in one pot synthesis (Zhao et al.

2018).

In 2018 a biosynthetic enzyme called (R)-salsolinol synthase was discovered and characterized, which extends the possibility to exploit it for bioproduction of these valuable pharmaceuticals and nutritional supplements (Chen et al. 2018).

Especially Pictet-Spengler cyclization of phenylethylamines containing deactivating (electron withdrawing) groups such as fluorine at the para and/or meta carbons on the phenyl ring. To make the ring formation more energetically favorable, an electrophilic group such as tosyl or mesyl group is added to the amine to render it more electrophilic and strong enough lewis acid such as boron trifluoride diethyl etherate (BF3· OEt2) is required for the intramolecular cyclization (CN103694170, 2014).

It is believed that in the mammalian brain salsolinol (1-methyl-6,7-dihydroxy- 1,2,3,4-tetrahydroisoquinoline) is formed in condensation of dopamine and acetaldehyde enzymatically catalyzed by (R)-salsolinol synthase (Figure 3) (Chen et al. 2018). (R)-salsolinol synthase, a type of Pictet-Spenglerase, has been successfully isolated from common rat (Rattus norvegicus) brain (Chen et al.

2018). Unlike Pictet-Spengler reaction, which produces racemic (R/S), (R)- salsolinol synthase catalyzed enzymatic reaction produces only R-salsolinol (Chen et al. 2018).

(19)

Figure 3. Pathway of salsolinol synthesis. Reaction A. is Pictet-Spengler reaction yielding rasemic (R/S)-Salsolinol from dopamine and acetaldehyde. In reaction B. (R)-salsolinol synthase catalyzes the formation of (R)-salsoline from dopamine and acetaldehyde. In reaction C. dopamine and pyruvic acid are converted to (R)-salsolinol via formation of salsolinol-1-carboxylic acid. An unknown enzyme marked with a question mark (?), putatively a decarboxylase, is involved in the process. (Chen et al. 2018)

Biocatalytic pathways provide several advantages in synthetic applications compared to traditional chemical methods as they are more environmentally friendly, do not require harsh reaction conditions, and produce excellent stereoselectivity of the reaction product (Zhao et al. 2018).

2.4 Common chemical production methods of tetrahydroisoquinolines

Several synthetic routes for production of isoquinolines exist (Calcaterra, et al.

2020). Bischler-Napieralski reaction is one of the most significant strategies for synthesizing isoquinolines by the Friedel-Crafts based routes. In Bischler- Napieralski synthesis (Figure 4), the presence of POCl3 or P2O5 leads to

(20)

dihydroisoquinoline, which can be oxidized to tetrahydroisoquinoline (Ji Ram et al. 2019).

Figure 4. Bischler-Napieralski synthesis resulting in 1-R-substituted isoquinoline (Ji Ram et al. 2019).

Reaction of POCl3 and amide yields nitrilium salt via imidoyl chloride upon heating and the desired 3,4-dihydroisoquinoline is then formed via intramolecular electrophilic aromatic substitution reaction as depicted on Figure 5 (Ji Ram et al. 2019).

Figure 5. The reaction mechanism of Bischler-Napieralski synthesis employs POCl3 to form 1-R-substituted 3,4-dihydroisoquinolines (Ji Ram et al. 2019).

The Pictet-Spengler isoquinoline synthesis is another important route for producing tetrahydroisoquinolines. Pictet-Spengler cyclization is essential for the biosynthesis of a variety of alkaloids (O’Connor & McCoy, 2006). Also, Pomeranz- Fritsch reaction is a common method for synthesizing isoquinolines (Smith,

(21)

2011). The cyclization is a Mannich-type reaction of an iminium ion formed by condensation of phenethylamine and carbonyl aldehyde (Maryanoff et al. 2004).

The cyclization is affected by electron withdrawing or donating groups present on the phenylic starting material. In the case of an electron withdrawing group, which deactivates the aromatic ring by decreasing the electron density at ortho and para positions through both inductive withdrawing effect and lone pair resonance donating effect, an enhanced electrophilic nature of the N-protonated imine is required to enhance the superacid catalyzed cyclization reaction. As an example, a tosyl group serves for this purpose. (Yokoyama et al. 1999; Silveira et al. 2006; Saha et al. 2008; Zheng et al. 2018)

3 Applicable enzymes for in vitro enzymatic synthesis of isoquinolines

By utilizing enzymatic pathways it is possible to produce eg. N-methylated isoquinolines from cheap and commercially available starting materials such as optionally substituted benzaldehydes or phenethylamines (Figure 6).

Figure 6. An example of enzymatic pathway for synthesizing tetrahydroisoquinolines.

(22)

3.1 Acetolactate synthase

Acetohydroxyacid Synthase I (AHAS, pyruvate:pyruvate decarboxylating acetaldehydetransferase EC 2.2.1.6) catalyzes the first step in branched-chain amino acid biosynthesis, a reaction involving decarboxylation of pyruvate bounded to hydroxyethyl group followed by condensation with another pyruvate molecule or with 2-oxobutyrate forming 2-acetolactate or with 2-ketobutyrate to form 2-aceto-2-hydroxybutyrate (Figure 7) (Bar-Ilan et al. 2001; Chien et al.

2010). The latter is a precursor to isoleucine, while the 2-acetolactate is precursor for leucine and valine biosynthesis (Chien et al. 2010). This route has been exploited for enantiospecific synthesis of (R)-phenylacetylcarbinol (R-PAC), a valuable intermediate for, inter alia, L-ephedrine with high conversion rate (over 99%) (Engel et al. 2003; Engel et al. 2005). In nature AHAS is found in bacteria, fungi and plants (Chien et al. 2010).

(23)

Figure 7. Reaction map of AHAS catalyzed reactions. In the first step lactylThDP is formed from the bound ThDP anion. In step two the intermediate reacts with either ketoacids and forms bounded acetohydroxy acid. In the third step the acetohydroxy acid is released. In the fourth step aryl aldehyde reacts with HEThDP intermediate, which produces bound aryl acetyl carbinol. In the fifth step the carbinol compound is released.

(Engel, 2004)

AHAS I can utilize benzaldehyde as a substrate and requires three cofactors to function, thiamine diphosphate (ThDP), divalent ion (Mg²⁺), and flavin adenine dinucleotide (FAD) (Bar-Ilan et al. 2001; Chien et a. 2010). According to Chien et al. 2010 the optimum temperature and buffer for AHAS I to function has been found to be 37° C in potassium phosphate buffer at pH 7.5.

(24)

3.2 Putrescine aminotransferase (ω-transaminase)

In industry a great interest has arisen towards transaminases as their usage enables producing chiral amines via biocatalytic pathways (Malik et al. 2012;

Cerioli et al., 2015; Planchestainer et al. 2019). Asymmetric transfer of an amino group between ketones and amines is possible when using ω-transaminase catalyzing the reaction (Coscolín et al. 2018; Galman et al. 2018; Han & Shin, 2019). Usage of transaminase in production of chiral amines has several advantages over conventional chemical synthesis including excellent stereoselectivity under mild reaction conditions in aqueous phase enabling greener chemistry (Han & Shin, 2019; Kelly et al. 2020).

Putrescine aminotransferase (EC 2.1.2.2) belongs to omega (ω)-transaminases, a group of enzymes catalyzing an amino group transfer from a non-α position amino acid, or a non-carboxylic amine compound by using pyridoxal 5’- phosphate (PLP) as cofactor (Łyskowski et al. 2014; Shin et al. 2018; Han & Shin, 2019; Kelly et al. 2020). Transaminases are crucial enzymes for amino acid metabolism of many microorganisms and eukaryotic cells (Yun & Kim, 2008).

Putrescine aminotransferase is an amine:pyruvate transaminase capable of (S)- enantioselective transamination of arylic chiral amine (Malik et al. 2012; Cerioli et al. 2015). (R)-selective ω-aminotransferases exist as well (Kim et al. 2018).

Transaminase (ω-TA) an important enzyme in production of several chiral amines (Malik et al. 2012; Han & Shin, 2019), which are widely used as essential precursors for the synthesis of pharmaceutical compounds (Łyskowski et al.

2014; Han & Shin, 2019; Kelly et al. 2020) and are important building blocks in production of biologically active compounds (Cerioli et al. 2015).

The catalytic cycle of transaminases (Figure 8) starts when the amine donor binds to the enzyme bound with PLP as cofactor. PLP takes up the amine and forms pyridoxamine-5'-phosphate (PMP), and the amine donor is released as a carbonyl

(25)

to amine acceptor and PLP is regenerated thus closing the catalytic cycle (reductive amination). (Coscolín et al. 2018)

Figure 8. Asymmetric reductive amination catalyzed by Putrescine transaminase (Cerioli et al. 2015).

(ω)-transaminase from halophilic proteobacteria Halomonas elongata (Uniprot - E1V913) shows S-enantiospecifity, tolerance to organic solvents, excellent conversion of product and wide substrate range (Cerioli et al. 2015).

Green et al. 2014 found that commercially available ortho-xylylenediamine dihydrochloride as the amine donor was superior to other amine donors, such as L-alanine, resulting in almost equimolar conversions and > 99% enantiomeric excess. The resulting isoindole co-product spontaneously polymerizes into colored derivatives which can be removed easily from the reaction mixture by filtration and serve as a method to observe the activity of the transaminase (Green et al. 2014.)

Cerioli et al in 2015 found the maximum activity for Halomonas elongata ω- transaminase (HEωT) to be at pH 10, but the lower pHs were more favorable for the catalyst stability and the optimal balance between activity and stability was found around pH 8-9. Maximum activity of HEωT was observed at 50° C, but the stability of the enzyme decreases when the temperature exceeds 35° C (Cerioli et al. 2015).

(26)

3.3 (R)-Salsolinol synthase

(R)-salsolinol synthase (EC 2.1.2.3) belongs to Pictet-Spenglerases (Chen et al.

2018). By now, three different types of Pictet-Spenglerases have been found in plants; strictosidine synthase, norcoclaurine synthase and deacetylisoipecoside synthase. (R)-salsolinol synthase putatively catalyze the synthesis of salsolinol, (1S)-1-methyl-1,2,3,4-tetrahydroisoquinoline-6,7-diol), in the mammalian brain and (R)-salsolinol synthase has been isolated from rat brain in 2018. After isolation (R)-salsolinol synthase was over expressed in mammalian cells and its catalytic ability in salsolinol production was verified. (Chen et al. 2018)

3.4 (S)-norcoclaurine synthase

(S)-norcoclaurine synthase (EC 4.2.1.78) catalyzes enantioselectively Pictet- Spengler condensation of dopamine and 4-hydroxyphenylacetaldehyde (4-HPAA) yielding the trihydroxylated alkaloid (S)-norcoclaurine (Figure 9) (Samanani &

Facchini, 2002; Li et al. 2016,). This is the first step in benzylisoquinoline alkaloid biosynthesis where formed (S)-norcoclaurine is a common precursor to all benzylisoquinoline alkaloids produced in plants. (Samanani & Facchini, 2002; Li et al. 2016).

(27)

Figure 9. Formation of (S)-norcoclaurine in condensation of dopamine and 4- hydroxyphenylacetaldehyde. Reaction is catalyzed by (S)-Norcoclaurine synthase.

(Samanani & Facchini, 2002)

A mechanism for (S)-norcoclaurine synthase catalyzed (S)-norcoclaurine synthesis (Figure 10) has been proposed by Ilari et al. 2009. The 4- hydroxyphenylacetaldehyde binds with the aspartic acid residue (K141) and dopamine binds with tyrosine (Y108) and glutamic acid (E110) residues. The alanine residue (K122) acts as proton donor and the catalyzes the iminium ion formation. While the dopamine is still bound to the tyrosine residue, the residue E110 catalyses the ring formation reaction leading to (S)-norcoclaurine. (Ilari et al. 2009)

(28)

Figure 10. A proposed mechanism for (S)-norcoclaurine synthase catalyzing an (S)- selective synthesis of norcoclaurine (Ilari et al. 2009).

3.5 Phenylethanolamine N-methyl transferase

Phenylethanolamine N-methyl transferase (PNMT, EC 2.1.1.28) is a 30-kDa enzyme that catalyzes the biosynthesis of catecholamine adrenaline from norepinephrine by methylating the amine of noradrenaline utilizing S-adenosyl-L- methionine as the methyl donor (Caine et al. 1996; Martin et al. 2001). PNMT natively catalyzes the conversion of norepinephrine to epinephrine by using S- Adenosyl-methionine as a co-factor, depicted in Figure 11.

(29)

Figure 11. PNMT catalyzes the conversion of norepinephrine to epinephrine by using S- Adenosyl-methionine as a cofactor (Isobe et al. 2004)

In mammals, including humans, PNMT is found in epinephrinergic cells of adrenal medulla, where epinephrine is synthesized (Isobe et al. 2004).

4 Theoretical background of designing and constructing a de-novo expression system for Bacillus subtilis

4.1 Application potential of B. subtilis as an expression host

Bacillus subtilis is widely used for extracellular production of industrially relevant enzymes (Kakeshita et al. 2012; Liu et al, 2013; Chen et al, 2015; Quesada- Ganuza et al. 2019; Xiao et al. 2020). B. subtilis not only has the ability to produce a wide range of relevant enzymes, but also is capable of secreting them in culture medium in high concentrations (Harwood, & Cranenburgh, 2008; Chen et al, 2015). Although Escherichia coli (E. coli) is the most widely used host for industrial scale expression systems producing pharmaceutical proteins, usage of

(30)

2018). The downstream process in B. subtilis is considerably simpler and cheaper than in E. coli as the proteins can be purified from the culture medium rather than from the cytoplasm (Harwood & Cranenburgh, 2008). Secondly, like many Gram-negative bacteria, E. coli produces harmful endotoxins, which need to be removed from the end-product resulting in even more complicated and expensive purification processes (Westers et al. 2004; Kakeshita et al. 2012). On the other hand, it has been suggested by Wang et al. in 2003 that peptidoglycan (PepG) and lipoteichoic acid (LTA), the major cell wall components of Gram- positive B. subtilis should be regarded as an endotoxins. However, B. subtils has notably advantages over Gram-negative bacterial hosts due its capability to non- specifically secrete recombinant proteins using various signalling peptides (Cui et al, 2018).

In comparison to E. coli as an expression host, the advantage of B. subtilis is its ability to anchor recombinant proteins on the surface of the cell wall by using sortases (Nguyen et al. 2011). The doubling time of individual B. subtilis cells is 120 minutes, when calculated by using modification of the Collins-Richmond principle (Burdett et al. 1986). In rich medium B. subtilis reaches a doubling time of 30 minutes at 37° C (Piggot, 2009) which is only slightly slower than that of E.

coli, 20 minutes in rich medium at 37° C (Schaechter, 2009). However, the difference is marginal for any practical use.

The problematic character of B. subtilis for production of heterologous proteins is its tendency to produce large amounts of extracellular proteases, which degrade secreted foreign proteins (Wu; 1991; Kakeshita et al. 2012). To overcome the problem with extracellular proteases, genetically modified strains such as B. subtilis WB600, which is deficient in six extracellular proteases, have been developed (Wu et al. 1991). In the B. subtilis WB600 strain the inactivation of chromosomal genes encoding neutral protease A, subtilisin, extracellular protease, metalloprotease, bacillopeptidase F, and neutral protease B, showed to decrease the extracellular protease activity to only 0.32% of the wild-type

(31)

strain protein activity (Wu et al. 1991). It has been shown that using B subtilis WB600 strain as a host of expression system significantly increases the stability of secreted enzymes (Wu et al. 1991).

Additionally, in B. subtilis three types of folding factors have been identified which can be exploited: propeptides, isomerases and metal ions (Harwood, &

Cranenburgh, 2008). Lastly, being Generally Recognized as Safe (GRAS), Bacillus subtilis is more desirable for food production and pharmaceutical industry than for instance Escherichia coli. (Westers et al. 2004; Kakeshita et al. 2012; Chen et al, 2014; Chen et al, 2015; Cui et al, 2018). In summarization, B. subtilis has potential to be an excellent expression host for secreted or membrane-bound enzymes in a biotechnological process.

4.1.1 Heterologous protein secretion and anchoring in B.

subtilis

4.1.1.1 Protein secretion pathways

In B. subtilis most proteins destined to leave the cytoplasm are translocated via the SecA-YEG (Sec) pathway across the cytoplasmic membrane (Kakeshita et al.

2012; Tjalsma et al. 2004). Also, additional export pathways exist eg. Twin- arginine translocation (Tat) pathway, ABC pathway and Com pathway (Tjalsma et al. 2004; Natale et al. 2008). The translocation of secretory proteins requires a signal peptide which is fused to the N-terminus of the mature expressed protein (Tjalsma et al. 2004; Kakeshita et al. 2012; Tsirigotaki et al. 2016). In the general secretory pathway (Sec) of B. subtilis, the proteins are secreted in the culture medium (Tsirigotaki et al. 2016). Some of the pathways also enable C-terminal signaling peptide processing while some, such as Tat pathway do not (Kakeshita et al. 2012). The choice of the translocation pathway has also been found to

(32)

As an example of signaling peptide mediation in the general secretory pathway (Sec), the signaling peptide for α-amylase (AmyE, EC 3.2.1.1), a starch degradation enzyme, has been employed to express otherwise hard-to-express heterologous proteins such as human interferon α-2β or pullunase with excellent yields (Kakeshita et al. 2012; Liu et al. 2018). The AmyE signaling peptide has been demonstrated to work in concert with a synthetic propeptide, and addition of a folding chaperone has enhanced the yield even more (Kakeshita et al. 2012).

AmyE is regulated under B. subtilis' ycgB-amyE -operon, where the α-amylase gene is flanked by ycgB, a gene of yet unknown function and lactate dehydrogenase (ldh) (Muñoz et al. 2011).

Sec-pathway -dependent proteins are sorted and targeted to the translocase.

The Sec pathway (Figure 12) results in proteins that are directly lipid layer membrane bound or translocated across the translocase. Mature proteins are released after acquiring their final form through the cleavage of their signal peptides. (Tsirigotaki et al. 2016)

(33)

Figure 12. Sec-pathway-dependent B. subtilis secretion machinery, R represents

ribosome, M represents mature protein, SP represents signal peptide, N and C represent amino and carboxyl termini of the protein respectively (Tjalsma et al. 2000).

4.1.1.2 N-terminus signaling peptide AmyE

A signal peptide directs proteins very effectively to the translocase and is thus necessary for the excretion of recombinant proteins to the culture medium of B.

subtilis (Kakeshita et al. 2012). AmyE signal peptide has been found to be one of the best signal peptides in B. subtilis (Kakeshita et al. 2012). The extracellular secretion mediated in B. subtilis via AmyE signaling peptide has been improved by saturation mutagenesis (Caspers et al. 2010). Discrepancies have been found

(34)

unmodified signaling peptide for modified AmyE signaling peptide. (Caspers et al.

2010). It has been postulated by Caspers et al. 2010, that signaling peptides can affect the speed and efficiency of routing proteins, but also the availability of extra cytosolic folding catalysts is involved. As such, the use of signal peptide is necessary to direct the protein efficiently to the translocase (Kakeshita et al, 2012).

After translocation across the cell membrane, the signaling peptide can be cleaved by a peptidase to mature protein (Kakeshita et al. 2012). For example, AmyE signaling peptide is cleaved by a type I signal peptidase at the AXA cleavage site located in the polar C-region of the signaling peptide (Auclair et al.

2011, Kakeshita et al. 2012). An example of signal peptidase is signal peptidase I (SipT) (Cai et al. 2017). Studies have shown that SipT is sufficient for secreting proteins (Cai et al. 2017). It has also been found that overexpression of sppA (putative signal peptide peptidase) improves cell growth (Cai et al. 2017).

4.1.1.3 Propeptides

Propeptides are not directly involved in translocation but assist in post- translocational folding to reach a stable and active secretory protein (Harwood,

& Cranenburgh, 2008). LEISSTCDA, a nine-residue synthetic propeptide, where the letters represent the single letter code of respective amino acid, fused immediately after the C-terminus of the signal peptide cleavage motif, has been observed to improve heterologous protein secretion in B. subtilis over 4-fold compared to secretion without the propeptide. (Kakeshita et al. 2012).

Improvement of protein secretion is believed to be caused by negative net charge of the N-terminus part of the mature protein (Kakeshita et al. 2012).

The LEISSTCDA encoding sequence is utilized in this work accompanied with signal peptide AmyE. A signal peptide directs proteins very effectively to the

(35)

translocase and is thus necessary for the production of recombinant proteins to the culture medium of B. subtilis (Kakeshita et al. 2012).

Signaling peptidase I (for example, SipT) recognizes the AXA cleavage site of amyE signalling peptide releasing the desired protein fused with propeptide (Figure 13) (Tjalsma et al. 2000; Kakeshita et al. 2012).

Figure 13. Theoretical cell-wall anchored heterologous protein expression in B. subtilis.

Propeptide is negatively charged while the tail of cell-wall anchor is positively charged.

4.1.1.4 Sec pathway protein folding

After translocation, Sec-dependent secretory proteins are in an unfolded state.

Although proteins can fold spontaneously to their native conformation, folding is frequently assisted by folding catalysts (Tjalsma et al. 2004). Lipoprotein PrsA (EC 5.2.1.8) is an essential extracytoplasmic folding factor in B. subtilis and overproduction of PrsA enhances the secretion of recombinant proteins (Viitikainen et al, 2000). Decreased amounts of PrsA dramatically affect secretion in post-translational step and drastic depletion of PrsA has been found to be detrimental for viability of B. subtilis (Viitikainen et al. 2000).

(36)

4.1.1.5 Signal peptidase

Another factor that limits the overexpression of recombinant proteins in B.

subtilis is type I signaling peptidase SipT. B. subtilis encodes leB -like gene producing signal peptidase SipT that processes preproteins and is essential for growth (Schneewind & Missiakas, 2014). The rate of processing of the signal peptide is limited by SipT (Viitikainen, 2001). SipT removes (hydrophobic) amino- terminal signal peptides from the secretory preproteins and thus limits the rate of the signal peptide cleavage (Tjalsma et al. 1998; Viitikainen et al. 2000, Tjalsma et al. 2004). The overexpression of SipT enhances the processing of recombinant proteins, but does not increase the accumulation of proteins in growth medium (Viitikainen et al. 2000).

4.1.1.6 Enzyme immobilization and GPI Anchoring in B. subtilis Typically in industrial applications, enzymes are immobilized by flocculating them (Rehn et al. 2013), cross-linking (Weetall, 1974) them to polymer beads or in calcium alginate because free-bound enzymatic reaction is slow and diffusion limited. Immobilizing the produced enzymes on the cell wall enhances the stability of the enzymes and increases their robustness and decreases their vulnerability towards enviromental changes (Homaei et al, 2013; Bielen et al, 2014). Anchoring on the cell wall allows the prolonged storing of the whole cell bio-catalyst up to 6 months with no significant loss of enzyme activity (Bielen et al, 2014). Cells displaying the active enzymes can be reused multiple times unlike purified enzymes in most cases (Bielen et al, 2014). Additionally, immobilized enzymes are generally easier to handle compared to their free forms (Homaei et al, 2013)

In Gram-positive bacteria, enzymes can be also immobilized on the cell surface by utilizing, for example, glycosylphosphatidylinositol (GPI) anchoring. GPI anchoring is based on covalent attachment of GPIs to the protein C-terminus,

(37)

which is then anchored onto cell membranes through the insertion of GPI lipid into the cell membrane bilayer (Guo, 2013; Yu et al. 2013). The GPI anchor itself consists of a phosphoethanolamine linker, glycan core, and phospholipid tail (Paulick & Bertozzi, 2008). As the Sec-pathway doesn’t require a C-terminal signaling peptide, it opens up a possibility to anchor the gene of interest to the cell surface by GPI anchoring (Kakeshita et al. 2012).

Anchoring takes place via transpeptidation reaction (Figure 14), in which a catalytic cysteine acts as a nucleophile and attacks a peptide bond of secreted protein that serves as its first substrate. Collapsing of the intermediate then releases the C-terminal domain of the substrate and the free amine of the second substrate binds the active site of and attacks the enzyme-substrate intermediate. Newly formed peptide bond is released in collapse of the intermediate and enzyme is regenerated. (McCafferty & Melvin, 2013)

Figure 14. A. Sortase substrates are recognized by their N-terminal signal peptide and a C-terminal LPXTG sorting signal. B. Covalent linkage between sortase substrate and cell wall is formed in following way: (1) Sec machinery exports the protein of interest which is recognized by a signal peptide. (2) LPXTG-motif is recognized which yields to an transpeptidase reaction. (3) Lipid II disassociates the sortase complex via nucleophilic attachment. (4) Lipid II -intermediate is formed. (5) The protein of interest is incorporated into the cell wall. (Call & Klaenhammer, 2013)

(38)

Additionally, the extracellular anchored proteins can be lysed out from the cell surface (Lee et al. 2002; Davies et al. 2009) or be released by utilizing phospholipase C (Verghese et al, 2006; Müller et al, 2012).

4.1.1.7 Significance of sortases in GPI-anchoring

Sortases are transpeptidases found in Gram-positive bacteria and they catalyze the covalent anchoring of proteins to the cell surface, specifically to peptidoglycan cell wall (Liew et al, 2011). Generally, in Gram-positive bacteria, sortases located at the cytoplasmic membrane are capable of anchoring a variety of surface proteins on the cell wall (Yang, 2021). The amount of the proteins anchored on the cell wall is proportional to the amount of sortase present (Liew et al. 2011). For example, when YhcS sortase was overexpressed under optimal conditions, 47300 fusions were displayed per cell (Liew et al. 2011).

Generally, sortases recognize a LPXTG-motif, where the capitalized letters represent the one letter abbreviation of amino acid and X represents any amino acid. In Gram-positive bacteria, after the translocation across the plasma membrane, sortases cleave the peptide between threonine and glycine residues of the LPXTG-motif and anchors the protein via transpeptidation to the peptidoglycan by covalent bond (Liew et al. 2011).

B. subtilis’ subfamily-4 sortase recognizes the sortase motif LPDTSA and the above mentioned cleavage occurs between threonine and serine residues (Oussenko et al. 2004) by sortase D (UniprotKB P54603) which also catalyzes the formation of an amide link between carboxyl-group of the serine and the cell wall precursor lipid II (Comfort & Clubb, 2004). B. subtilis encodes two putative sortases, YhcS and YwpE (Nguyen, 2011). YhcS recognizes the C-terminus motif of the surface protein YhcR (Nguyen, 2011; Yang 2021).

(39)

4.1.2 Folding chaperone PrsA

PrsA (EC 5.2.1.8) is a 266 amino acids long lipoprotein functioning as a peptidylprolyl cis-trans isomerase (PPIase) that assists post-translocational folding of secretory proteins in B. subtilis in the (pseudo) periplasmic space and is required in post-translocational phase for protein stability. The isomerase does not affect the rate of translocation (Harwood, & Cranenburgh, 2008; Wahlström et al. 2003). Evidence has been found that PrsA prevents detrimental interactions with the membrane cell wall and enhances the signal peptide AmyE processing rate (Wahlström et al. 2003; Caspers et al. 2010; Kakeshita et al.

2012).

The majority of secretory proteins in B. subtilis are not PrsA dependent (Harwood, & Cranenburgh, 2008) but PrsA is necessary for growth and depletion of PrsA ultimately causes cell death (Harwood, & Cranenburgh, 2008). It is known that PrsA functions as an extracellular cell-associated folding chaperone or foldase and reduces the susceptibility of the proteolysis of secreted proteins (Harwood, & Cranenburgh, 2008). Overexpression of native PrsA or heterologous co-expression of engineered PrsA is known to increase alpha-amylase production in B. subtilis by up to 2.5 fold (Chen et al. 2014; Quesada-Ganuza et al. 2019). It has also been shown that engineered PrsA chaperone domain has improved amylase secretion with almost inverse correlation to secretion stress level (Quesada-Ganuza et al. 2019).

4.1.3 Potential plasmid transformation methods for B. subtilis and E. coli

Natural bacterial transformation is a process which involves the internalization and chromosomal integration of DNA (Johnston et al. 2014). In molecular biology transformation is utilized to introduce DNA to the bacterial cell by altering the

(40)

(Chen & Dubnau, 2004). For transformation it is necessary to make the bacterial cells competent to enable them to absorb and internalize extracellular DNA (Domingues & Nielsen, 2016).

Escherichia coli cells are commonly prepared using the Hanahan method, where the cells are rendered competent by using CaCl2 and plasmid is transformed into the competent cells by heat shock (Green & Sambrook, 2018). In the Hanahan method, E. coli culture must be in the early exponential stage of the growth (OD600 under 0,26), (Green & Sambrook, 2018). Another common procedure for producing competent cell for E. coli is the Inoue method, which is based on the salt containing Inoue transformation buffer and the cell culture is harvested in the middle of the exponential growth phase (OD600 0,55) (Green & Sambrook, 2020). The advantage of the Inoue method is that the competent cells can be stored in a -80° C freezer for further use for a long time and the frozen cells can be used without any further preparations immediately after thawing them, thus greatly decreasing time required (Green & Sambrook, 2020).

B. subtilis is naturally competent in the stationary phase of the growth (Rahmer et al. 2015). However, the transformation of B. subtilis is more laborious and time consuming than that of E. coli as preparing competent B. subtilis cells typically requires applying two different minimal media and careful growth monitoring of the bacteria culture (Zhang et al. 2014). As an example, a protocol by Bennallack et al. 2014 (Published in supplemental methods section of PMID:

25313391) based on the method described by Anagnostopoulos and Spizizen (1961) requires the optical density (OD600) to reach to 2.0-2.8, corresponding early stationary-phase.

Electrotransformation is a less time consuming method for transforming B.

subtilis than a transformation method exploiting the natural competence of B.

subtilis at high cell densities, and required an OD600 of 0.85-0.95, corresponding to the exponential growth phase (Eppendorf, 2003). In addition, electroporation

(41)

found to be a well functioning transformation method for larger constructs into B. subtilis (Ohse et al. 1995). However, the operation parameters like the growth and electroporation media composition, field strength and competent cell concentration have significant effect on the transformation efficiency (Lu et al.

2012). Storing competent B. subtilis cells in -80℃ freezer decreases the transformation efficiency 2-fold and cells are not viable after long term storage (Bennallack et al. 2014).

4.2 Significance of the codon optimization for the stability and function of the expression system

The secretion of heterologous proteins originating from another organism can be severely hampered due to a poor translation efficiency of heterologous genes (Westers et al. 2004; Lanza et al. 2014). A stable expression system is essential for cost-effective production of heterologous proteins (Westerns et al.

2004).DNA is formed with a combination of four different bases: Adenosine (A), Thymine (T), Cytosine (C) and Guanine (G). In the translation of proteins, triplets of them (a codon) encode a single amino acid within an open reading frame.

There are thus 61 different possibilities to form a codon, excluding three stop codons. In the genetic code, only 22 direct proteinogenic amino acids (excluding- selenocysteine or pyrrolysine) are known, therefore a single amino acid can be encoded by different synonymous codons. The preferences of diverse organisms to encode a specific amino acid vary, this is called codon bias. The reasons for codon preferences of different organisms are not fully understood. (Gustafsson et al. 2004)

Transfer RNAs (tRNAs) have a crucial role in translation and protein synthesis as tRNAs bind to amino acids and transfer them to ribosomes, where they are attached into a forming polypeptide chain. Cellular resources are consumed to produce tRNAs that recognize different codons. If an organism utilizes only a

(42)

small subset of codons, lower amounts of tRNAs are adequate and thus fewer resources are needed in the translation process (Emilsson & Kurland, 1990).

Lower production rate of proteins encoded by codons with low abundance or poorly charged tRNAs may occur compared to production rate of proteins encoded by highly abundant charged tRNAs (Tuller et al. 2010). Misfolded proteins may result if some regions of protein are translated more rapidly than others when codons encoding the protein are a combination of high and low abundance (Pechmann & Frydman, 2013). Varied codon usage allows organisms to regulate protein expression and thus improves its ability to adapt to changing conditions (Dittmar et al. 2005).

Codon optimization is a tool of gene engineering, used most commonly to enhance heterologous gene expression by optimizing the codons of genes according to the host organism’s codon bias (Lanza et al. 2014). The cloned genes, desired to be expressed, may contain codon triplets rarely used in host organisms, and can originate from organisms using non-canonical code or coding sequence containing expression-limiting regulatory elements (Gustafsson et al.2004). This problem can be ameliorated by de-novo synthesizing the gene of interest with codon optimization suitable for the desired host organism (Lanza et al. 2014). Codon optimization is frequently used when expressing functional proteins in hosts that do not naturally express the gene of interest, most commonly, when genes from eukaryotic organisms are expressed in prokaryotic expression systems (Lanza et al. 2014).

It is not fully understood what kind of impact synonymous codon usage has on protein expression, and it’s difficult to predict how the optimal gene sequence should be designed (Webster et al. 2016). There are a multitude of factors to consider, when optimizing a gene sequence (Webster et al. 2016). Codon optimization increases protein production in certain types of expression systems, but can also cause unpredictable harmful changes to protein conformation,

(43)

folding and stability, alter post-translational modification sites and effect on protein function (Mauro & Chappell, 2014).

4.3 Theoretical background of the de-novo expression vector design for B. subtilis

Plasmids are usually circular molecules of double-stranded DNA (Clark et al.

2019; Burks, 2020). A cloning vector, or backbone, is usually a plasmid lacking expression operons such as a promoter for the desired DNA insert to be translated by propagating in bacteria (Carte & Shieh, 2015). Most frequently plasmid-cloning vectors are designed to replicate in E. coli (Burks, 2020).

Enzymes that are necessary for the replication of the plasmid DNA are produced by the host organism (Burks, 2020).

There are three important features in plasmid vectors, origin of replication (ORI), selection marker and a restriction site or sites, called multiple cloning site (MCS) for cloning (Bruks, 2020).

Selection markers enable the selection of transformed cells and they are typically delivered alongside the gene of interest (Jones, 2003; Burks, 2020). A Selection marker can be based on resistance genes conferring the transformed cells the ability to grown in the presence of toxic compounds, such as antibiotics, that would kill the untransformed cells (Jones, 2003; Burks, 2020). However, other types of methods for selection exist as well, for instance, a selection marker can be based on inserting a fluorescent protein expressing gene into host along with the gene of interest (Moore, et al. 2004). Fluorescent protein expression changes the appearance of the transformed cells and the selection can be performed by visually selecting the transformed cells (Moore, et al. 2004).

Another non-antibiotic selection method leans on the usage of auxotrophic bacterial strain, not capable of synthesizing a certain metabolite necessary for its

(44)

survival (Timms et al. 2012). Alongside with the gene of interest, the genes enabling the bacteria to express the protein producing the vital metabolite, are transformed into bacteria cells (Timms et al. 2012). Thus, the positive clones gain the ability grow in the growth media not supplemented with the vital metabolite (Timms, et al. 2012).

To enable the insertion of the foreign DNA, a cloning vector must contain at least one restriction site which can be recognized and cleaved by a restriction enzyme (Burks, 2020). Restriction enzymes are classified according to their structure, recognition site, cleavage site, co-factors and activators (Williams, 2003).

A cloning vector itself, by definition, is incapable of transcription and translation of a gene of interest into a functional protein whereas an expression vector contains the necessary elements for a host cell to transcribe and translate the gene (Carter & Shieh, 2015). A typical expression vector for a bacterial cell consists of a plasmid backbone including an origin of replication, a selection marker, a promoter, a ribosome binding site (typically the Shine-Dalgarno consensus sequence), cloning site(s) and a transcriptional terminator (Cantoia et al. 2021). Other genetic elements may be included as well (Cantoia et al. 2021).

The function of a promoter is to drive the expression of a gene in a cell (Carter &

Shieh, 2015). Both natural and hybrid promoter exist and their purpose is to enable regulated and efficient transcription (Goldstein, 1995). In the selection of a promoter both gene expression and the consequences of the gene expression to the cell need to be taken account (Goldstein, 1995).

Although B. subtilis has significant advantages as a host for heterologous protein production and is widely used for applications for several fields of industry (Wsters et al., 2004; Liu et al, 2013; Cui et al, 2018), unsuitable cloning vectors containing weak promoters or other unsuitable plasmid parts can cause the yields of heterologous proteins to be low (Song et al, 2016). Additionally, the lack

2 Isoquinolines and their significance...3

Preface and acknowledgments

Table of contents

1 Introduction...1

2 Isoquinolines and their significance...3

2.1 Chemistry of isoquinolines...3

2.2 Formation pathways of tetrahydroisoquinolines in nature...4

2.3 In vitro enzymatic tetrahydroisoquinoline synthesis...5

2.4 Common chemical production methods of tetrahydroisoquinolines ...7

3 Applicable enzymes for in vitro enzymatic synthesis of isoquinolines...9

3.1 Acetolactate synthase...10

3.2 Putrescine aminotransferase (ω-transaminase)...12

3.3 (R)-Salsolinol synthase...14

3.4 (S)-norcoclaurine synthase...14

3.5 Phenylethanolamine N-methyl transferase...16

4 Theoretical background of designing and constructing a de-novo expression system for Bacillus subtilis...17

4.1 Application potential of B. subtilis as an expression host...17

4.1.1 Heterologous protein secretion and anchoring in B. subtilis..19

4.1.2 Folding chaperone PrsA...27

4.1.3 Potential plasmid transformation methods for B. subtilis and E. coli...27

4.2 Significance of the codon optimization for the stability and function of the expression system...29

4.3 Theoretical background of the de-novo expression vector design

for B. subtilis...31

4.3.1 Plasmid replication...34

4.3.2 Examples of common molecular cloning strategies...36

5 Aims...40

6 Materials and methods...42

6.1 Chemicals...42

6.2 Laboratory equipment and software...43

6.3 Acquisition of the bacterial strains...43

6.3.1 Antibiotic resistance susceptibility test...44

6.3.2 Catalase test...44

6.3.3 Gram stain of B. subtilis WB600...44

6.4 General protocols...45

6.4.1 Codon optimization...45

6.4.2 Preparation of competent cells and transformation protocols45 6.4.3 Isolation of plasmid DNA...46

6.4.4 Cloning protocols...47

6.4.5 Polymerase chain reaction (PCR)...48

6.4.6 Electrophoresis...51

6.4.7 DNA purification from an agarose gel...52

6.4.8 Phenol-Chloroform extraction and ethanol precipitation...52

6.4.9 Starch assay for plasmid integration to the B. subtilis WB600 AmyE locus...53

6.5 The plasmid pVimea...53

6.5.1 Designing the plasmid pVimea...53

6.5.2 Acquiring the plasmid pVimea...58

6.5.3 Constructing the plasmid pVimea out...58

6.5.4 Electroporation of the plasmid pVimea-fragment B into B. subtilis WB600...58

6.5.5 Expression of (S)-norcoclaurine synthase and PNMT with B.

subtilis WB600 pVimea- fragment B...58

6.5.6 Lysis of GPI-anchor...59

6.5.7 SDS-PAGE...60

6.6 The plasmid pEquinorm-amp...61

6.6.1 Designing the plasmid pEquinorm-amp...61

6.6.2 Acquiring the plasmid pEquinorm-amp...62

6.6.3 Transformation of pEquinorm-amp into E. coli HB101 and DNA isolation...62

6.6.4 Transformation of pEquinorm-amp into B. subtilis...62

6.6.5 Fluorescence spectroscopy methods...63

6.7 The plasmid pEquinorm2-amp...63

6.7.1 Designing of the plasmid pEquinorm2-amp...63

6.7.2 Constructing pEquinorm2-amp...65

6.8 The plasmid pEquinorm2-kanR...70

6.8.1 Designing the plasmid pEquinorm2-kanR...70

6.8.2 Constructing pEquinorm2-kanR...71

6.9 The plasmid pEquinorm2-kanR-intermediates...76

6.9.1 Design of expression cassette for AHAS and transaminase (“intermediates”)...76

6.9.2 Aquiring the expression cassette containing operons for AHAS and ω-transaminase (“intermediates”)...78

6.9.3 Constructing pEquinorm2-kanR-transaminase...78

6.10 The plasmid pEquinorm2-kanR-isoquinolines...82

6.10.1 Designing the expression cassette containing operons for (R)-salsolinol synthase and (S)-norcoclaurine synthase (“isoquinolines”)...82

6.10.2 Aquiring the expression cassette “isoquinolines”...83

6.10.3 Constructing of pEquinorm2-kanR-isoquinolines...83

6.11 In-silico modelling of protein structure...87

6.11.1 Generation of the homolog protein models...87

6.11.2 Optimization and structure analysis of protein models...87

7 Results...89

7.1 Basic testing of B. subtilis WB600 and E. coli HB101 strains...89

7.1.1 Antibiotic resistance susceptibility test for B. subtilis WB600 ...89

7.1.2 Catalase test for B. subtilis WB600 and E. coli HB101...89

7.1.3 B. subtilis WB600 Gram staining...90

7.2 The plasmid pVimea...91