Utah State University DigitalCommons@USU
All Graduate Theses and Dissertations Graduate Studies
5-2016
Investigating Biosynthetic Steps of an Angucycline Antifungal
S. Gabrielle Gladstone Utah State University
Follow this and additional works at: https://digitalcommons.usu.edu/etd
Part of the Biological Engineering Commons
Recommended Citation Gladstone, S. Gabrielle, "Investigating Biosynthetic Steps of an Angucycline Antifungal" (2016). All Graduate Theses and Dissertations. 4952. https://digitalcommons.usu.edu/etd/4952
This Thesis is brought to you for free and open access by the Graduate Studies at DigitalCommons@USU. It has been accepted for inclusion in All Graduate Theses and Dissertations by an authorized administrator of DigitalCommons@USU. For more information, please contact [email protected]. INVESTIGATING BIOSYNTHETIC STEPS OF AN ANGUCYCLINE ANTIFUNGAL
by
S. Gabrielle Gladstone
A thesis submitted in partial fulfillment of the requirements for the degree
of
MASTER OF SCIENCE
in
Biological Engineering
Approved:
Jixun Zhan Ron Sims Major Professor Committee Member
Anhong Zhou Mark R. McLellan Committee Member Vice President for Research and Dean of the School of Graduate Studies
UTAH STATE UNIVERSITY Logan, Utah
2016 ii
Copyright © S. Gabrielle Gladstone 2016
All Rights Reserved iii
ABSTRACT
Investigating Biosynthetic Steps of an Angucycline Antifungal
by
S. Gabrielle Gladstone, Master of Science
Utah State University, 2016
Major Professor: Jixun Zhan Department: Biological Engineering
From the bacterium Streptomyces sp. SCC-2136 (ATCC 55186), two angucycline nat- ural products are produced, designated Sch 47554 and Sch 47555. These compounds are produced through a type II polyketide biosynthetic pathway. The early biosynthetic steps to these molecules were confirmed. These include the minimal polyketide synthase (PKS), the C-9 ketoreductase, the first-ring aromatase, the subsequent ring cyclase, and two oxy- genases. Also confirmed were the biosynthetic genes responsible for production of the first amicetose moiety, as well as the glycosyltransferase that creates a C -glycosidic bond be- tween the angucyclic scaffold and the amicetose moiety. In confirming these pathways, two new natural products were produced: GG31, an amitosylated rabelomycin, and GG53, ra- belomycin hydroxylated at C-12b. Future work will be to understand the late biosynthetic steps and generate new angucyclines through combinatorial biosynthesis.
(64 pages) iv
PUBLIC ABSTRACT
Investigating Biosynthetic Steps of an Angucycline Antifungal
S. Gabrielle Gladstone
The species of bacterium Streptomyces sp. SCC-2136 which has the American Type
Culture Collection index 55186 naturally produces two chemical compounds, labeled Sch
47554 and Sch 47555. These compounds were previously reported to posess antifungal activity. This thesis sets out to confirm the mechanism by which the bacterium produces these compounds; specifically which genes are responsible for producing the enzymes that make and shape the molecules.
The genes and enzymes that were characterized are the minimal polyketide synthase, the ketoreductase that specifically acts on the ninth carbon, the first-ring aromatase, the subsequent ring cyclase, and two oxygenases. Also elucidated were the genes responsible for production of the first sugar (called amicetose) in the chain of two sugars attached to carbon 9 as well as the glycosyltransferase enzyme that attaches this sugar to the rest of the molecule.
In confirming these pathways, two new compounds were produced: GG31, which has a base structure of the antibiotic rabelomycin but is attached at carbon 9 to a molecule of the deoxysugar amicetose; and GG53 which has the base structure of rabelomycin but with a hydroxyl group at carbon 12b. Future work will be to understand the later biosynthetic steps and generate new related molecules through combinatorial biosynthesis. v
Maria Montessori once said, ”The preparations for life are indirect”. This is certainly true.
This thesis is dedicated to my parents, Melissa and Russell Gladstone; to Andrew Tori Hamblin; to Kandy Napan-Hsieh; and to Aerin Richardson. Wasn’t this a wild ride. vi
ACKNOWLEDGMENTS
I must thank Dr. Jixun Zhan for his guidance and patience while I learned about molecular biology and combinatorial biosynthesis. I also wish to thank the members of my
Masters committee, Drs. Anhong Zhou and Ron Sims for their attention and help these last two semesters. Paul Veridian, the graduate coordinator in BioEngineering, was invaluable in the management of all the paperwork, deadlines, and requirements. I could not have
finished this without you.
Additionally, I am eternally grateful for all of the people who were friendly and en- couraging over the last four years. These include my labmates: Dr. Zeng Jia, Dr. Kandy
Napan, Dr. Siyuan Wang, Dr. Jiachen Zi, Dr. Lei Shao, Dr. Tong Zhou, Dr. Bingji Ma,
Dr. Shuwei Zhang, Dr. Dayu Yu, Dr. Jianfeng Mei, Dr. Anfeng Xiao, Dr. Riming Yan,
Dr. Huiwen Yan, Lei Sun, and Fuchao Xu.
For their help outside the lab, I wish to thank the Painter family, Kat Combs, Jessica
Bills, Steven Collins, Stetson Barnhouse, Kassie and Rick Cressal, Jesse Walker, the Smurf family, the USU library staff, and anyone who gave me a reassuring smile.
And most importantly, I thank my parents, my grandparents, and my partner Andrew
Hamblin: we made it.
S. G. Gladstone
Logan, UT
2016 vii
CONTENTS
Page ABSTRACT ...... iii PUBLIC ABSTRACT ...... iv ACKNOWLEDGMENTS ...... vi LIST OF TABLES ...... viii LIST OF FIGURES ...... ix 1 Introduction ...... 1 2 Literature Review ...... 4 2.1 Type I and Type III PKS ...... 4 2.2 Type II PKS ...... 6 2.3 Angucycline Compounds ...... 9 2.3.1 Angular Hydroxyls ...... 10 2.3.2 Glycosylation ...... 11 3 Experimental Methods ...... 16 3.1 Extraction of Genomic DNA ...... 16 3.2 Polymerase Chain Reaction and Amplification of Genes ...... 16 3.3 Creating Plasmids for Heterologous Expression ...... 17 3.4 Extraction of Compounds ...... 18 4 Results ...... 19 4.1 Rabelomycin Pathway Elucidation ...... 19 4.1.1 Minimal PKS ...... 19 4.1.2 Ketoreductase ...... 21 4.1.3 Aromatases and Cyclases ...... 22 4.2 Glycosylation ...... 23 4.3 Angular Hydroxyls ...... 24 5 Conclusions ...... 31 6 Future Work ...... 34 6.1 Glycodiversification ...... 34 6.2 Co-expression of Two Plasmids to Assist with Combinatorial Biosynthesis . 35 APPENDICES ...... 43 A Compounds ...... 44 B Nuclear Magnetic Resonance Spectra for Structure Confirmation of GG31 . 47 viii
LIST OF TABLES
Table Page
3.1 Sequences of primers used for plasmid construction ...... 17
3.2 General PCR mixture ...... 18
4.1 Nuclear magnetic resonace spectroscopy (NMR) data for GG31 ...... 25
4.2 Plasmid construction ...... 26
5.1 Gene functions ...... 32
A.1 Numbered compounds ...... 45 ix
LIST OF FIGURES
Figure Page
1.1 Structures of 1, 2, 3, 4 ...... 2
2.1 The proposed organization of the herboxidiene biosynthetic gene cluster . . 5
2.2 Polyketide chains of certain lengths spontaneously form predictable structures 7
2.3 Biosynthesis of oxytetracycline and reported shunt products ...... 8
2.4 Oxidoreductase UrdM catalyzes a 2-step reaction ...... 13
2.5 Catalysis of 12 and 12b oxidation ...... 14
2.6 Examples of glycosylated natural products ...... 14
2.7 Deoxysugars that can be synthesized from common intermediates ...... 15
2.8 The four urdamycin glycosyltransferases ...... 15
4.1 Putative gene cluster of ATCC 55186 ...... 20
4.2 The vector used to build all of the expression plasmids ...... 20
4.3 HPLC traces of expressed biosynthetic products ...... 27
4.4 Structures of urdamycin A and simocyclinone D8 ...... 28
4.5 Structure elucidation of GG31 ...... 28
4.6 Comparison of retention times of 18 to 16 ...... 29
4.7 Proposed pathway to 1 and 2 ...... 30
6.1 Putative route from glucose-6-phosphate to the three sugars ...... 34 CHAPTER 1
Introduction
There is a joke about the history of medicine that appeared in the now-discontinued
Western Journal of Medicine. It is presented as a table:
The History of Medicine 2001 BC Here, eat this root.
1000 AD That root is heathen. Here, say this prayer.
1850 AD That prayer is superstition. Here, drink this potion.
1920 AD That potion is snake oil. Here, swallow this pill.
1945 AD That pill is ineffective. Here, take this penicillin.
1955 AD Oops, bug mutated. Here, take this tetracycline.
1960-1999 AD 39 more “oops”...Here, take this more powerful antibiotic.
2000 AD The bugs have won! Here, eat this root.
Natural products, a term which refers to any compound made by a living organism, are often responsible for why “this root” is still theoretically a viable medicinal treatment.
Natural products have been studied in organic chemistry since at least 1891, when Albrecht
Kossel classified them into two categories: primary and secondary metabolites [21]. Pri- mary metabolites are vital for the survival of the organism, e.g. structural or nutritional.
Conversely, secondary metabolites are not vital to the immediate survival of the living or- ganism, but provide a competitive advantage. Usually this is accomplished by affecting some other organism- for example to deter predators, control parasites, or in the case of antimicrobials, to kill off competition.
Antimicrobial natural products have a variety of structures, and are often classified based on the structural properties they possess. The first natural product with antibacterial activity to be discovered was penicillin, discovered by Alexander Fleming in 1929 [21], and 2 the category of antibacterial agents that has grown from this discovery is called β-lactam drugs, because of the 4-membered lactam ring that is present in each structure. Other categories of include macrolactones, because each has a lactone ring containing more than
8 carbons; tetracyclines, which contain four rings; and angucyclines, which are similar in structure to tetracyclines, but the four rings are not attached linearly, but at an angle.
Naturally occurring macrolactones, tetracyclines, and angucyclines are all secondary natural products from the genus Streptomyces, a genus of soil- and marine-dwelling Gram- positive bacteria. Because of their complex secondary metabolism, various Streptomyces species are responsible for over two-thirds of naturally occurring antibiotics [20], including many members of the families listed above.
Fig. 1.1: Structures of angucycline compounds Sch 47554 (1), Sch 47555 (2), aquayamycin (3), and urdamycin A (4)
This thesis details the pursuit of understanding the biosynthetic pathway of one specific pair of natural products, Sch 47554 (1) and Sch 47555 (2), which belong to the angucycline family of compounds. Angucyclines are part of a larger classification of natural products that are built by type II polyketide synthases. These two molecules, which are identical ex- cept for the second sugar in the disaccharide moiety on ring D, were discovered in 1993 from
Streptomyces sp. SCC-2136. They possess an aglycon very similar to that of aquayamycin 3
(3) and the synthetic compound SS-228Y5, except for the differing sugars. Biologically,
(1) and (2) possess moderate antifungal activity and are also weak inhibitors of tyrosine hydroxylase [6], an integral enzyme in the conversion of tyrosine into neurotransmitters. De- spite their potential as medicines, the biosynthetic pathway of these compounds remained unknown. In 2006, a group from South Korea sequenced the gene cluster of Streptomyces sp. SCC-2136, and used BLAST searching to find homologous DNA sequences. Results showed that many genes were similar to those involved in the synthesis of urdamycin A (4) by Streptomyces fradiae, but some enzymes had homologies to other products’ gene clus- ters [5]. The genes putatively involved in the biosynthesis of (1) and (2) were annotated but not functionally characterized, which has hampered the understanding of the biosynthetic process of these antifungal molecules. 4
CHAPTER 2
Literature Review
The study of natural product biosynthesis gives inspiration toward engineering new
“unnatural” natural products, because a molecule’s novel functionality might suggest new pathways or a way to improve an existing molecule. Of the natural products available, the most intriguing are the polyketides, which have been isolated from bacteria, fungi, and plants [44]. These molecules are assembled by three types of polyketide synthases (PKSs): type I, type II and type III. All three do similar actions: synthesis of natural products via chaining together molecules of acetyl-coenzyme A, and then tailoring this chain into various conformations. The mechanisms, though, are the differences between the three types.
2.1 Type I and Type III PKS
Type I PKSs are large, multimodular enzymes with covalently fused domains that perform individual catalytic activities [45]. The essential domains are the ketosynthase (KS) domain, the acyl transferase (AT) domain, and the acyl carrier protein (ACP). These three domains catalyze the elongation of the polyketide chain. Other domains are ketoreductase
(KR), dehydratase (DH), enoyl reductase (ER) and methyltransferase (MT). Often these are present in the type I PKSs to modify the chain, but there are occurrences of the tailoring steps taking place via enzymes separate from the large PKS.
One example of post-PKS tailoring occurs in the biosynthesis of the anticholesterol natural product herboxidiene, first discovered from Streptomyces chromofuscus. There are seven genes in the gene cluster for herboxidiene; the three that code for the PKS are hypothesized to be herB, herC and herD. It is predicted that HerB contains a loading module and three elongation modules; HerC consists of three elongation modules, and HerD has one elongation module. This allows the electrophilic attack of the vinyl adjacent to the
first carbonyl, causing the separation of the nascent chain. The molecule is then available for 5
Fig. 2.1: The proposed organization of the herboxidiene biosynthetic gene cluster modification by domains HerE, HerF, HerG, which are an epoxidase, a methyl transferase, and hydroxylase, respectively [36]. Recently, HerF was confirmed via knockout experiments and in vitro reactions to catalyze the formation of the methoxy at C-25, depicted in Figure
2.1.
In contrast to noniterative PKSs, some type I PKSs are iterative (iPKS), meaning that their architecture involves fewer domains that get repeatedly used [7]. Among iPKSs are highly reducing PKSs (hrPKS), partially reducing PKSs (prPKSs) and nonreducing
PKSs (nrPKSs). The biosynthesis of macrolactone natural products like radicicol involves collaborative actions of a hrPKS and an nrPKS. The hrPKS (consisting of KS, AT, ACP,
KR, DH, and ER domains) synthesizes a reduced polyketide chain, which is transferred to the partner nrPKS by a starter unit/acyl-carrier protein transacylase (SAT) domain. This nrPKS elongates the chain to a certain length, utilizes the product template (PT) domain to form an aromatic ring by specific aldol condensation, and then the thioesterase domain 6 closes the macrolactone ring [42].
Type III PKSs are different from type I PKSs. The enzyme complex is made up of two small homogeneous enzymes, each with a single active site. The homodimer uses the two active sites iteratively to catalyze the Claisen condensations of one or more extender units onto a starter substrate [16], the species of substrate depending on the type of organism.
Most plant type III PKSs use free acyl-CoA as a starter unit, but bacteria and fungi use an acyl-ACP. As is to be expected, the final product will differ depending on the number and type of extender substrates used, as well as the cyclization pattern [43].
2.2 Type II PKS
Type II PKSs have similar intermediate products as type I and type III (a long polyke- tide chain), but the mechanism by which they create this chain this is different. A type
II PKS cluster is made of a series of discrete enzymes that form multi-enzyme complexes to catalyze the synthesis of polyketide chains. Additionally, depending on the substrates required, some components of type II PKS are interchangeable across species. Thus, the minimal PKS of a type II PKS gene cluster is comprised of an ACP to bring malonyl
CoA extenders to the growing chain; a KS to attach the extender, and a chain length fac- tor (CLF) to control the length of the chain. In a system where these three enzymes are all the only type II PKS components that are expressed, the generated polyketide chain will spontaneously fold into more stable products. In the absence of tailoring enzymes like aromatases (AROs) and cyclases (CYCs), a newly-formed decaketide (made from one acetyl-CoA and nine malonyl-CoAs) will spontaneously fold into one SEK15 (5); and an octaketide chain will likewise fold into SEK4 (7), as seen in Figure 2.2. Hence the minimal
PKSs from different species are comparable. When expressed in Streptomyces, the genes of various minimal PKS will all generate spontaneous products determined by the length of the original polyketide chain.
With the addition of a KR in the expression vector, different sets of spontaneous products are observed, again depending on the length of the chain. This is because the reduction of one specific ketone group, usually the one attached to C-9, is enough to make the 7
Fig. 2.2: The polyketide chain folds spontaneously into different, but predictable, structures depending on how many malonyl CoA molecules they contain, and whether or not Carbon 9 is reduced spontaneous cyclization include a heterocyclic pyran ring in the scaffold. A decaketide chain reduced at C-9 folds into RM20b/c (6); the C-9 reduced octaketide chain becomes mutacin
(8). Both the minimal PKS and the KR functionalities are conserved across different type
II PKS clusters, in that it is possible to combine the minimal PKS from one gene cluster
(e.g. oxytetracycline) and the KR from a different cluster (e.g. tetracenomycin [13, 32]) and arrive at the same products (again, see Figure 2.2).
This conservation is not necessarily the case for CYCs. This type of enzyme catalyzes the formation of the subsequent rings. Aromatases are also included in this family, since
AROs are simply CYCs that are specialized for the first ring. The proposed mechanism is formation of an enol via nucleophilic attack of a carbonyl to the alpha carbon on the 8 other side of the chain. The enol then collapses, resulting in the first-ring cyclization
[23]. Subsequent ring formations follow a similar mechanism. However, in heterologous expression with other PKS enzymes, it was observed that some CYCs seem to have a minimal requirement for the length of the polyketide chain. This is logical: if a CYC is geared to catalyze a fourth ring closure but is presented with a chain from which only three rings can form, that particular CYC will not work. Experiments also suggested that certain
CYCs cannot work on polyketide chains that have not been reduced at specific carbons [29], especially if the enzyme is an ARO. This is because most aromatases are specialized to the
first ring of the scaffold. Typically, the ARO will catalyze the formation of the first ring, and the CYCs that follow will direct the formation of the second ring and so on. But if the chain is reduced at a specific carbon, the steric conformation may be incorrect for the catalytic mechanism of the ARO. The result is that the enzyme has a different level of activity towards the reduced chain, and the end product may be different.
Fig. 2.3: Biosynthesis of oxytetracycline and reported shunt products 9
This is apparent in the biosynthetic pathway of 12, oxytetracycline (Figure 2.3), which contains OxyABCD, OxyJ, and OxyK, as the minimal PKS, C-9 KR, and ARO respectively
[32]. Following the aromatase is the CYC OxyN, which cyclizes the second and third rings.
The final ring forms when the C-1 carbonyl attacks C-18 as the molecule breaks the thioester bond and separates from the enzyme [32].
Unique to tetracycline compounds, however, is the amidotransferase gene from S. rimosus as oxyD and in Streptomyces sp. SF2575 as ssfD, as evidenced by the nitrogen- containing compounds WJ85 (9) and WJ35 (10) that spontaneously form when only oxyABCD and oxyABCDJ or ssfABCD and ssfABCDJ are expressed, respectively [32,33]. Following amination and ketoreduction, several different tailoring enzymes work to decorate the newly formed pretetramid. OxyF introduces the methyl group at C-6; OxyE and OxyL cooper- ate to create the double-hydroxylation between the third and fourth ring. OxyQ catalyzes the reductive amination at C-2, which is then followed by the methylation of the nitrogen via OxyT to make a tertiary amine. The resulting molecule is anhydrotetracycline (11).
A precursor to 12, 11 is the first of the intermediates in the tetracycline family to be a functional antibiotic, due to the groups in position on the fourth ring.
2.3 Angucycline Compounds
Angucycline natural products have been extensively studied since at least 1959, when scientists in the Miyago prefecture of Japan described ayamycin, a compound from Strep- tomyces flaveoulus ATCC 13118, although the structure was never elucidated. Four years later, scientists in the Aomori prefecture cultured a different Streptomyces strain and iso- lated aquayamycin (3). Growth inhibition assays were conducted, and it was found that
3 inhibited Gram-positive bacteria, with the exception of mycobacteria [20]. The Aomori group also hypothesized an approximate molecular formula based on NMR and IR spec- tra, but the structure was unknown until 1970 when it was elucidated as a derivative of benzene[α]-anthracene with one C-linked deoxysugar [30]. Fifteen years later, Dr. J¨urgen
Rohr discovered urdamycin A (4)[11], one of the most important angucycline natural prod- ucts. Extensive studies have been done on 4, and it is a standard to which other angucycline 10 compounds are compared.
The structure and biosynthesis pathways of several angucycline products have been deciphered, but most are yet to be determined. Like other type II PKS pathways, those that produce angucyclines begin with a minimal PKS, and then continue with the aforementioned tailoring enzymes. The angular ring is formed by a specific type of CYC that works to close the fourth ring in the benz-α position instead of as an anthracene as with chromomycin
[35], or an octahydrotetracene, as with oxytetracycline. This differently shaped scaffold allows for new patterns of functional groups, created by glycosynthases, glycosyltransferases, oxidoreductases and methyltransferases.
2.3.1 Angular Hydroxyls
The hydroxyl groups at the angle of the α ring are an example of the unique function-
ality. It has been hypothesized that the enzymes responsible for these oxidations catalyze
a Baeyer-Villager rearrangement [30], and the deletion of such oxygenases result in spon-
taneous formation of products like rabelomycin [11]. It is proposed that the oxidation of
C-4a and C-12b is catalyzed in each gene cluster by the same enzyme that oxidizes the
hydroxyl at C-12 into a carbonyl. Many pathways have confirmed the enzymes responsible
for this type of oxidation in different compounds; namely LanE, PgaE, CabE catalyze both
the oxidation at C-12 and C-12b. Other related enzymes likewise do the oxidation at C-12,
but actually perform a dehydration reaction at C-4a/C-12b, resulting in an unsaturated
angular bond; essentially the opposite of the previously described double-oxidizers.
Other pathways, like that of 4, split this oxygenation up into two enzymes: one for
C-12, and one for C-12b. In the Urd pathway, these are UrdE and UrdM respectively it
was once thought that UrdE was like the dual-functioning oxygenases as described above,
especially when compared with LanE (to which it shares a 73% amino acid identity) and
PgaE (77% identity). However, when UrdM was inactivated, all urdamycin A production
was eliminated, and instead the shunt products rabelomycin and urdamycin L were formed
(see Figure 2.4)[19].
It would seem that if UrdE is present in the reaction system, UrdE oxidizes C-12 to 11 a carbonyl, and UrdM does the above two-step oxidation, showing that the two enzymes do indeed work together. Curiously, though, further experiments showed that when the reduction domain of UrdM is isolated separately, the reaction proceeds in the same way, as described in Figure 2.5[?]. This is also supported by the notion that urdamycin L, the shunt product described in Figure 2.4, is made in very small amounts (less than 1 mg per liter) [35]
2.3.2 Glycosylation
In addition to oxidization, another category of angucyclines, and indeed many natural products, is those natural products that contain sugar moieties, including rebeccamycin, an indolocarbazole antitumor candidate [4]; and puerarin, an isoflavone antioxidant [24]
(Figure 2.6) as well as the aforementioned 4.
For angucyclines, the glycoside moieties tend to be deoxysugars, which are derived from glucose-1-phosphate, which itself is synthesized directly from glucose either by anomeric phosphorylation or derived via phosphoglucomutase [25,40] from glucose-6-phosphate. Once the glucose-1-phosphate has been activated via coupling to a nucleoside phosphate (usually thymidine diphosphate, or TDP), the pathway continues and modifies the sugar. This is accomplished by a specific sequence of reductases, dehydratases, and epimerases that are unique to each sugar, although the pathways to build similar sugars are very homologous across different species of Actinobacteria. Figure 2.7 depicts this phenomenon, as several deoxysugars are formed from glucose-1-phosphate, via dehydratases, epimerases, ketoreduc- tases, and methyltransferases. It is important to note that most of these sugars are made by different species of bacteria; although a few sugars show up frequently in different nat- ural products. Additionally, there have been successful expressions of sugar plasmids with heterologous sugar genes to create both the expected and novel deoxysugars [1].
Once the activated deoxysugars have been synthesized, they are now able to be cou- pled to the polyketide scaffold by one or more glycosyltransferases (GTFs). The reaction catalyzed by GTFs is simply a redox reaction between a sugar donor and a sugar acceptor.
For angucyclines and other type II PKS natural products, the sugar donor is the activated 12 deoxysugar, and the acceptor is the aromatic polyketide itself.
In the case of urdamycin, there are four glycosyltransferases. Two that act upon the scaffold: UrdGT2, which catalyzes the C-C link between C-9 of the scaffold and the aliphatic carbon of a D-olivose moiety [9, 19, 25]; and UrdGT1a, responsible for the catalyst of an
O-linked L-rhodinose at C-12b. UrdGT1c extends the sugar chain at C-9 by attaching an
L-rhodinose via an α(1-3) O-linkage to the D-olivose mentioned above; and finally UrdGT1b
completes the trisaccharide chain with the α(1-4) O-linkage of D-olivose to the L-rhodinose
(2.8).
Yet because individual enzymes are potentially very specific to the aglycan sugar accep- tor, and the individual saccharide sugar donor, it was assumed that enzymes like UrdGT2 and LanGT2 (a similar angucycline with O-linked sugars) were specific to their respec- tive final products and related intermediates. On the contrary, the enzyme ElmGT was discovered as being structurally homologous to each of the glycosyltransferases from the urdamycin and landomycin pathways [41], that was able to catalyze “at least eleven differ- ent sugars” [34] onto its tetracenomycin derivatives. This provides evidence that further novel natural products might be made by simply switching glycosyltransferases between
Streptomyces species, resulting in glycodiversification. 13
Fig. 2.4: Oxidoreductase UrdM catalyzes a 2-step reaction at carbon 12b in the pathway to Urdamycin A 14
Fig. 2.5: Catalysis of 12 and 12b oxidation by UrdE and UrdMred, the reducing domain of UrdM
Fig. 2.6: Rebeccamycin and puerarin, two of many natural products containing sugar moi- eties 15
Fig. 2.7: Deoxysugars that can be synthesized from common intermediates
Fig. 2.8: The four urdamycin glycosyltransferases 16
CHAPTER 3
Experimental Methods
3.1 Extraction of Genomic DNA
Once a suitable culture of Streptomyces sp. SCC 2136 ATCC 55186 was established on solid YM media, a single colony was selected into liquid YM media. After approximately
12 hours of growth at 250 RPM at 28 oC, the genomic DNA was extracted using the phenol-chloroform technique [46], and was purified by washing with 70% ethanol.
3.2 Polymerase Chain Reaction and Amplification of Genes
Primers for each gene of interest were designed to include restriction sites to facilitate
cloning and ligation procedures. The sequences of these primers are presented as Table 3.1 below. The primers were ordered from Sigma-Aldrich and reconstituted with TE buffer to the approximate concentration of 100 ng/mL. The PCR was then performed for the genes using the recipe printed in table 3.2 and Phusion DNA polymerase (ThermoFisher).
The agarose gel purified PCR product was ligated into the pJET 1.2 cloning vector (Ther- moFisher) using T4 DNA ligase (New England Biolabs) and transformed into E. coli XL1
Blue competent cells. After an overnight culture at 37 oC on solid LB media supplemented with carbenicillin (50 µg/mL), several colonies were picked into liquid LB with carbenicillin, grown at 37 oC for an additional 8-10 hours, and the plasmids extracted from each colony using alkaline lysis and column purification (ThermoFisher Gene-JET). A small amount of each plasmid sample was digested using the corresponding restriction endonucleases in order to verify the size of the inserted gene, as compared with the Gene Ruler 1kb Plus
DNA ladder (ThermoFisher). When a sample had been identified as likely correct, high concentrations of the plasmid were sent out for sequencing (Eton Biosciences) in order to ensure that the gene was correct and free of mutations. 17
Table 3.1: Sequences of primers used for plasmid construction
Genes 5’Primer Restriction Enzymes
oxyAoxyB 5’AATTAATTAAGGAGGAGCCCATATGTCCAAGATCCATGACGC3’ PacINheI 5’AAGCTAGC TCAGTCCCGGCCGCTGACCA3’
oxyC 5’AAACTAGTGGAGGAGCCCATATGACCCTGCTCACCCTCTC3’ SpeI/NheI 5’AAGCTAGC TCACTTGTCCCGCGCGGCGC3’
oxyD 5’AAACTAGTGGAGGAGCCCATATGTGCGGAATCGCCGGCTG3’ SpeI/NheI 5’AAGCTAGC TCATAGCTCCAGGCTGACGC3’
schP3 5’AAACTAGTGGAGGAGCCCATATGGAGATCCTCGACGGTCT3’ SpeI/NheI 5’AAGCTAGC TCAGCGCAGCAGCGTGCCGC3’
schP4 5’AAACTAGTGGAGGAGCCCATATGACCGTCCGTGAGGTCGA3’ SpeI/NheI 5’AAGCTAGC TCAGCGCCGCGCTTCGGCGT3’
schP5 5’AAACTAGTGGAGGAGCCCATATGACGACTGCCACGTCCGG3’ SpeI/NheI 5’AAGCTAGC TCAGAAATTGCCGAGGCCGC3’
schP6schP7schP8 5’AATTAATTAAGGAGGAGCCACCACTGGACCGCACTGTGA3’ PacI/NheI 5’AAGCTAGC TCAGCTGGACCGGACGGGCA3’
schP9 5’AAACTAGTGGAGGAGCCCATATGCACAGCACGCTGATCGT3’ SpeI/NheI 5’AAGCTAGC TCACAGTGCGGTCCAGTGGT3’
schP10 5’AAACTAGTGGAGGAGCCCATATGGATGCTTCGGTCATAGT3’ SpeI/NheI 5’AAGCTAGC TCAGGACTTCGGGCCCCGCG3’
3.3 Creating Plasmids for Heterologous Expression
Each gene or block of genes was cloned as described above, then ligated together via cutting with the appropriate restriction endonuclease (New England Biolabs) and inserting the genes. Often, genes were ligated using compatible, but non-identical sticky ends (e.g., the family of NheI, SpeI, XbaI, and AvrII) that cannot be re-cut with the same enzymes.
This allowed us to form large fragments of DNA that could be easily ligated into the expression vector pRM5 between the PacI and NheI restriction sites.
Once the pRM5-derived plasmid had been purified and its concentration measured, the plasmid was introduced into Streptomyces lividans K4 protoplasts via polyethlylene glycol (PEG) aided transformation. After allowing the transformed protoplasts to recover on solid R5 media for 12-16 hours at 28 oC, an antibiotic was overlaid to select for positive
transformants: 50 µg/mL thiostrepton in R5 media with 0.8% agar. Positive transformants 18
Table 3.2: General PCR mixture
Sterilized water 14.2 µL
High fidelity buffer 4 µL
DMSO 0.6 µL
100 mmol dNTP 0.4 µL
3’ primer 0.2 µL
5’ prime 0.2 µL
Genomic DNA template 0.2 µL
Phusion DNA polymerase 0.2 µL
Total Volume 20 µL were identified after 4-5 days at 28 oC, and spread onto new R5-thiostrepton plates. If the culture continued to grow healthily, and the dark brown secondary metabolites were observed to be excreted from the mycelia, the culture was amplified over several weeks to at least 5 liters of culture.
3.4 Extraction of Compounds
Once these plates were confluent with mycelia and brown metabolite, the agar was diced into small pieces and mashed with a solution of 89% ethyl acetate, 10% methanol, and 1% acetic acid (v/v) to partition the metabolites from the agar into the solution. The extract was concentrated by rotary evaporation, purified by HP-20 open column chromatography, and finally purified by liquid chromatography using an Agilent 1100 chromatograph with a
C-18 Eclipse column and a gradient of 45% acetonitirile and 55% water, supplemented with
0.1% formic acid.
ESI-MS were acquired on an Agilent 6130 LC-MS HPLC (Agilent Zorbax SB-C18 column, 5 m, 21.2 mm 150 mm). 1D and 2D NMR spectra were recorded on a JEOL instrument (300 MHz). 19
CHAPTER 4
Results
4.1 Rabelomycin Pathway Elucidation
4.1.1 Minimal PKS
It had been reported that schP8, schP7, schP6 were the KS, CLF, and ACP, re- spectively, due to their homology with urdA, urdB, and urdC [5]. The minimal PKS of urdamycin had already been found to be comparable to jadABC of jadomycin [8] and thus all of the other minimal PKS clusters that yielded an unreduced polyketide chain of 20 carbons in length. These included the minimal PKSs of actinorhodin (act-ORF1, 2, 3 ) tetracenomycin (tcmKLM )[22][29], and oxytetracycline (oxyABC )[32]. In the absence of other tailoring enzymes, these enzymes produce a dodecaketide that spontaneously folds into the molecule known as SEK15 (5)[13,29,32]. We therefore hypothesized that if schP8, schP7, schP6 were indeed the minimal PKS for 1 and 2, they would work together to produce what would be observed as (5).
In order to verify this hypothesis, two comparable plasmids were produced, both using the pRM5 vector as an expression system, depicted in Figure 4.2. One of these was pTZ3, containing oxyA, oxyB and oxyC immediately after the promoter; the other was pME1, which had schP6, schP7 and schP8 after the promoter. A list of plasmids can be found in
Table 4.2 below.
When each of these plasmids was transformed into the heterologous host, Streptomyces lividans K4 via polyethylene glycol-assisted protoplast transformation [20], the resulting transformants yielded a slightly yellow colored compound that could be seen within the 20
schA5 schA11 schA1 schA3 schA7 schA9 schA13 schA15 schA17 schA19 schP2 schS2 schS4
schA2 schA4 schA6 schA8 schA12 schA14 schA16 schA18 schP1 schS1 schS3 schA10
schP7 schA22 schS6 schS8 schS10 schP3 schP5 schP9 schP11 schA24 schA26 schA28
schS5 schS7 schS9 schA20 schP4 schP8 schP10 schA21 schA23 schA25 schA27 schP6 schA29
Fig. 4.1: Gene cluster upon which the functions of each step in the biosynthetic pathway were hypothesized
Fig. 4.2: The vector used to build all of the expression plasmids. Genes were inserted in between the PacI and NheI restriction sites 21 growth medium. Extraction and analysis of both samples resulted in very similar data: a product peak with [M+H]+ = 359 and [M-H]- = 357 (Figure 4.3 A and B). These data
verify that the product produced by both pME1 and pTZ3 is (5). As such, SchP8, SchP7
and SchP6 were identified as the dedicated ketosynthase, chain length factor and acyl carrier
protein, respectively.
4.1.2 Ketoreductase
We then sought to prove that SchP5 was, as reported, the C-9 ketoreductase. As stated
above, (5) always forms in the absence of such a ketoreductase. However, when a minimal
PKS is expressed with certain C-9 KR genes, various other products form [13].
We were curious how this putative ketoreductase would behave. There were two hy-
potheses proposed: one, that SchP5 would either act directly on the nascent polyketide
chain, similar to actKR [29]; or two, that it would show up in synthesis later, as seen in the
pdm gene cluster of pradimicin [44]. Thusly, schP5 was introduced to the minimal PKS in pTZ3 to get pME3 via ligation after oxyC. Transformation of this plasmid into S. lividans
K4 resulted in a slightly darker compound excreted by the cells into the growth medium.
HPLC analysis of pME3 showed a different set of peaks from either of the previous plas- mids’ compounds, notably the primary product had an earlier retention time than 5 (4.3
C. ESI-MS spectra revealed that the mass of this product is 386, which is consistent with the reported RM20 b/c (6)[17].
This result confirms that SchP5 specifically reduces the C-9 keto group of the nascent chain [13], confirming our first hypothesis that it is very similar to ActKR from the acti- norhodin pathway of S. coelicolor A3 [27]. The presence of the ketoreductase allows the polyketide chain to fold spontaneously into different fused-ring structures, and also allows the chain to be folded by certain aromatases, which could not utilize it as a substrate were it not reduced at C-9. 22
4.1.3 Aromatases and Cyclases
The next step in the pathway is the formation of the first ring, a procedure typically catalyzed by an aromatase. In the biosynthetic pathway for 1 and 2, SchP4 is the putative
ARO. This is based on its gene schP4 being 73% homologous to simA5, and 72% homol- ogous to urdL [5], both of which are confirmed first-ring cyclases for simocyclinone in S. antibioticus [37] and 4 in S. fradiae [11], respectively.
For both simocyclinone and urdamycin A as well as numerous other aromatic polyketide molecules [3,10,22,23, 37], the aromatase is closely related to the subsequent cyclaze; referred to in a pair as an ARO/CYC. This is because assembly of all other rings is dependent on the ARO. That is to say, the guided formation of the rest of the molecule is contingent upon the first ring coming together between C-7 and C-12, whether the molecule is of an anthracycline formation, an angucycline formation, or neither. Like the ARO, the putative cyclase SchP9 was identified as such based on its homology with both the urdamycin A cyclase UrdF and the cyclase of simocyclinone SimA4.
Therefore, to verify the identity of first schP4 and then schP9, the putative ARO was cloned and inserted after the ketoreductase schP5 in the plasmid pME3, resulting in a plasmid designated pME4. By expressing oxyABC, schP5 and now schP4 all together, the resultant molecule would be SEK43 (16), with a molecular weight of 368. Expressing pME4 and comparing its resultant metabolites with that of pME1, pTZ3, and pME3, showed a comparatively less polar molecule (Figure 4.3 D). ESI-MS of the extract produced molecular ion peaks of [M+H]+ = 369 and [M-H]- = 367, confirming that the major metabolite of pME4 was SEK43 (16) and therefore that SchP4 is the dedicated aromatase in the pathway.
Having established the ARO, we proceeded to introduce schP9 to the expression plas- mid, which would finish the core structure and result in a recognizable angucycline product.
Ligating schP9 into pME4 after schP4 resulted in the plasmid pME5. When transformed into S. lividans K4, a compound was produced that was significantly darker than the pre- vious samples had shown. Although it was only detected in trace amounts, the molecule 23 was found to be UWM6 (13), from the ESI-MS that indicated the molecular weight was
342 [22].
We believe that rather than remain with 13 as the major product, the extracted metabolite mixture would spontaneously oxidize into the structure known as rabelomycin
(14), which has a molecular weight of 338. This was the major peak in the mass spectra of pME5s products (Figure 4.3 E). Curiously, it was also the major peak of pME6, which is a plasmid created by adding schP10 to pME5. The enzyme SchP10 is a putative monooxy- genase, responsible for catalyzing a carbonyl group forming at C-12. We thus were able to confirm that SchP10 is indeed the monooxygenase that oxidizes the second ring into the quinone form. Even without SchP10, rabelomycin can still be produced by oxidation of 13.
4.2 Glycosylation
Many deoxysugars can be made from glucose-1-phosphate [26], a metabolite of glu- cose. The enzymes responsible for the conversion of glucose-1-phospate to any of several deoxysugar species seem to be conserved over Streptomyces species. Indeed, the putative functions of SchS1, 2, 3, 4, 5, and 6 are closest in homology to UrdS, UrdQ, UrdZ3, UrdH, and UrdG, in that order – all of which are responsible for L-rhodinose and D-olivose syn- thesis in Urdamycin A [5].
Additionally, the disaccharide moiety attached to C-9 was of particular interest to us since biologically active molecules with C-linked carbohydrate moieties are better drug candidates, due to the robustness of the C-glycosidic bond through the harsh pH conditions of the digestive tract. The putative C-glycosyltransferase of the Sch pathway is listed as
SchS7, with 71% homology to UrdGT2, a confirmed C-glycosyltransferase.
Since it is difficult to verify individually transformed monosaccharides from in vivo
culture, we set out to express the entire putative sugar pathway in a plasmid along with
the rabelomycin scaffold. Furthermore, since UrdGT2 has been reported to be unselective
towards dextrorotary or levorotary deoxysugars [14], we hypothesized that an additional
epimerase, SchS8, might be all that is necessary for completion of the disaccharide attached
to C-9. Based on this reasoning, pKN57, which contained schS2, schS1, schS3, schS8, 24 schS7, schS, schS5, schS4 was ligated to pME8 to form pGG31.
GG31 was therefore predicted to be rabelomycin with a D-amicetose moiety attached via a C-glycosidic bond to carbon 9; with the possibility of an L-amicetose moiety attached via a (1,4) O-glycosidic bond to the D-amicetose. When transformed into K4, GG31 (17) was formed, with earlier retention time than 16 (Figure 4.3 F), Ion peaks corresponding to [M-H]- at m/z 451.1, [M+H]+ at m/z 453.2 in the ESI-MS spectra suggested that the molecular weight of 17 is 452. 1H and 13C NMR of 17 showed very similar signals to 16, except one D-amicetose. The structure was suggested as amitosylated rabelomycin, which was confirmed by plentiful HMBC correlations; these are summarized in Table 4.1, with one- dimensional and two-dimensional spectra in Appendix B. The position of the D-amicetose attachment was also at C-9 based on the HMBC correlations from H-1’ to C-8, 9, 10 and from H-10 to C-1’ (Figure 4.5). The stereostructure was proposed the same as 16.
4.3 Angular Hydroxyls
Since both 1 and 2 have two hydroxyl groups at C-4a and C-12b, but neither 13 nor
17 do, it was apparent that the oxidoreductase responsible for these groups was neither
SchP9 nor SchP10. Indeed, the enzymes responsible for catalysis of the hydroxyl formation at the angle are being investigated in many different angucycline biosynthetic pathways
[15, 18, 19, 31]. The two hypotheses for any given angucyclic natural products are that there is one dioxygenase that results in formation of both C-4a and C-12b hydroxyls; or a monooxygenase that yields an epoxide intermediate before being opened by a water attack
[38].
In the pathway of 1 and 2, there are six putative oxygenases: SchA18, SchA19, SchP3,
SchP10, SchA24, and SchA26. Having identified SchP10 as the C-12 oxygenase and SchA26 as very likely responsible for reduction of side-chain deoxysugar moieties [2,5], four are left.
SchP3 is homologous with UrdM of S. fradiae, and has been evaluated as responsible for the catalysis of both C-4a and C-12b hydroxyls (see Figure 2.4 in the Literature Review).
By ligating schP3 into pGG31 after the last sugar synthase gene schS4, we created the 25
Table 4.1: 1 13 H (300 MHz) and C (75 MHz) NMR summary for GG31 in acetone-d 6; (δ in ppm, J in Hz)
No. δ C δ H
1 156.3
2 121.6 7.16 (s, 1H)
3 142.0
4 120.2 7.26 (s, 1H)
4a 139.3
5 121.4 8.15 (d, 1H, 7.6)
6 121.6 8.31 (d, 1H, 7.6)
6a 135.2
7 188.5
7a 115.9
8 159.4
9 139.4
10 133.6 7.89 (d, 1H, 7.2)
11 121.4 7.87 (d, 1H, 7.2)
11a 133.1
12 189.7
12a 132.7
12b 122.0
13 21.4 2.49 (s, 3H)
1’ 73.3 4.85 (d, 1H, 9.4)
2’ 32.0 1.50 (m, 1H); 2.28 (m, 1H)
3’ 33.3 1.73 (m, 1H); 2.25 (m, 1H)
4’ 79.0 3.46 (m, 1H)
5’ 72.3 3.40 (m, 1H)
6’ 14.1 1.40 (d, 1H, 5.8) 26
Table 4.2: Plasmid construction
Plasmid Genes Major Product
pTZ3 oxyA, oxyB, oxyC SEK15
pME1 schP6, schP7, schP8 SEK15
pME3 schP6, schP7, schP8, schP5 RM20 b/c
pME4 schP6, schP7, schP8, schP4, schP5 SEK43
pME5 schP6, schP7, schP8, schP4, schP5, schP6 UWM6
pME6 oxyA, oxyB, oxyC, schP4, schP5, schP9, schP10 Rabelomycin
pME8 schP4, schP5, schP6, schP7, schP8, schP9, Rabelomycin
schP10
pKN57 schS1, schS2, schS3, schS4, schS5, schS6, schS7, Amicetose and transfer
schS8
pGG31 schP6, schP7, schP8, schP4, schP5, schP9, GG31
schP10, schS1, schS2, schS3, schS4, schS5, schS6,
schS7, schS8
pGG53 schP6, schP7, schP8, schP4, schP5, schP9, GG53
schP10, schP3 plasmid pGG53. When analyzed, the major product had a retention time less than that of 17. The ESI-MS spectrum depicted [M+H]+ at m/z equal to 355; [M-H]- at m/z equal to 353, indicating a molecular weight of 354 (see Figure 4.6). This compound was dubbed
GG53, and is hypothesized to have a structure similar to the molecule indicated as 18, although stereochemical correlations have not yet been found to verify this. 27
5
A 300 nm 5
300 nm B
6
300 nm C 16
D 300 nm
14 13 E 420 nm 17 420 nm F
5 10 15 20 25 min
Fig. 4.3: HPLC comparison of products from expression plasmids transformed into S. livi- dans K4. A: pTZ3; B: pME1; C: pME3; D: pME4; E: pME5; F: pGG31 28
Fig. 4.4: Structures of urdamycin A and simocyclinone D8, which have ARO/CYC homol- ogous to those in the Sch pathway
Fig. 4.5: Structure elucidation of GG31 29
16 M-1 = 353
m 300 400 /z
M+1 = 355 18
m 300 400 /z
12.5 15 17.5 20 22.5 min
Fig. 4.6: Comparison of retention times of GG53 (18), which is more polar than rabelomycin (16), and LC-MS spectra 30
Fig. 4.7: Proposed pathway to 1 and 2 31
CHAPTER 5
Conclusions
In the search for novel bioactive compounds, we produced two compounds, 15 and 16
using combinatorial biosynthesis. We believe that these “unnatural” natural products are
intermediates in the pathway to compounds 1 and 2, the natural secondary metabolites
of Streptomyces sp. 2136. These two products signify the promise of engineering novel
compounds via metabolic pathway recombination. The original compounds 1 and 2 have shown promising antifungal activity for many different species, and potential anticancer activity via inhibition of tryptophan 5-monooxygenase and tyrosine hydrolase. Moreover, by expressing these and other type II polyketide synthase genes together, new molecules can be made without the environmental costs of chemical synthesis.
We first confirmed that the minimal PKS of the sch gene cluster was SchP8, SchP7, and SchP6. This minimal PKS is able to produce the same well-known compounds as several other minimal PKSs from various Streptomyces species: SEK15 (5 and RM20b/c
6). Therefore, these genes synthesize a 20-carbon polyketide chain that will be shaped into
the polyaromatic backbone of the natural products 1 and 2. A summary of all the gene
functions is listed below in Table 5.1.
We also verified the functions of the other SchP enzymes, SchP4 through SchP10, as
crucial enzymes in the synthesis of rabelomycin (14), another intermediate in the pathway
to the biosynthesis of 1 and 2. Of these, SchP4 is the aromatase (ARO), SchP5 is the
ketoreductase (KR), SchP9 is the cyclase (CYC), and SchP10 is an oxygenase (OX). Not
only were the functions confirmed, but the sequence specificity of many of the enzymes were
also resolved. (For example, the selectivity of SchP5 to work on the nascent polyketide
chain, instead of working on the structure later in the synthesis.) Additionally, the minimal
PKS genes from the oxytetracycline pathway oxyA, oxyB and oxyC can be exchanged with
schP6, 7, and 8, suggesting a broader conclusion that the Type II PKS pathway of ATCC 32
55186 shares many similarities with other type II PKS pathways.
We were also able to confirm the functions of the sugar synthase enzymes SchS1 through
SchS6, which have very similar functions to other sugar synthases. Together, these enzymes work to first synthesize the deoxysugar D-amicetose, and then the C-glycosyltransferase
SchS7 attaches the sugar onto the rabelomycin backbone at the C-9 position. As noted, the attachment of the sugar to C-9 was expected, but ongoing experiments are being performed to test the substrate selectivity of SchS7 for both deoxysugar and aglycan specificity.
Table 5.1: Gene functions
Gene Function
schP6 Acyl Carrier Protein
schP7 Chain Length Factor
schP8 Ketoacyl synthase
schP5 C-9 ketoreductase
schP4 Aromatase
schP9 Cyclase
schP10 C-12 oxygenase
schS7 C-glycosyltransferase
schS1 Hexose 3-ketoreductase
schS2 Hexose 3-ketoreductase
schS3 Hexose 3-dehydratase
schS4 Hexose 4-ketoreductase
schS5 Hexose 4,6-dehydratase
schS6 Hexose NDP-glucosynthase
A few steps still remain before we have the full pathway to 1 and 2, however. We are still pursuing the synthesis and attachment of the other sugars, L-amicetose and aculose, because other studies have shown that the full complement of deoxysugars is crucial to compounds biological activity [1, 40]. 33
In addition, the gene cluster contains many other proteins with unknown functions that we remain curious about. Perhaps some of them synthesize a different secondary metabolite under other conditions; perhaps they modify 1 and 2 or assist in the emission from the cell.
There is strong functional homology between several organisms biosynthetic machinery, so the answers to these questions may be within reach. 34
CHAPTER 6
Future Work
6.1 Glycodiversification
The second sugar in the disaccharide connected to C-9 is L-amicetose in Sch-47554, and
is aculose in Sch-47555. The pathway for the synthesis of these two different deoxysugars
is proposed in Figure 6.1, and is potentially just a different branch off of the pathway from
glucose-6-phosphate to amicetose.
Fig. 6.1: Putative route from glucose-6-phosphate to the three sugars. We have confirmed the pathway to D-amicetose
With this hypothesis, the only genes that need to be included in an expression plas- mid are schA26, a nucleotide that codes for an oxidoreductase with homology to AknOx from the aclacinomycin biosynthetic pathway of S. galilaeus [2]; and the appropriate O-
glycosyltransferase, either SchS9 or SchS10. Knockout studies are currently being done by
other members of our laboratory to determine in the full pathway which of the glycosyl-
transferase enzymes is responsible for each of the O-glycosidic bonds. 35
Additionally, some exploration has been done towards the coupling of different sugars onto angucyclic or tetracyclic scaffolds. One of these is olivose, which was discussed in the
Literature Review as being attached to Urdamycin A at C-9 by UrdGT2; but other studies have shown that UrdGT2 is able to catalyze the transfer of different sugars [14]. In this way, the library of Sch-like antifungal compounds can grow, lending insight into the usefulness of the deoxysugars to mitigate infections.
6.2 Co-expression of Two Plasmids to Assist with Combinatorial Biosynthesis
All of the heterologously expressed compounds discussed in this thesis were expressed by transforming a single expression plasmid with the required polyketide synthase genes into
S. lividans K4. The vector used was derived from the plasmid pRM5 [27] that itself was constructed from the SCP2* self-replicating plasmid found in S. coelicolor. pRM5 allows for easy ligation of a block of genes between the PacI and NheI, or PacI and EcoRI restriction sites (see Figure 4.2 in the Results section). The smallest of the expression plasmids used was pME1, which expressed 5, of size 18.8 kilobase pairs in length. The largest is pGG31, which is a 31.5 kilobase pairs in length.
Further adding individual genes is tedious at best, due to the increasing size of the
plasmid. Direct insertion of the new gene into a single-cut, e.g., NheI, of such an expression
plasmid is extremely improbable; and verification of correct ligation proves to be difficult
and time-consuming. Thus, instead of growing an existing plasmid from upwards of 32,000
base pairs by adding more nucleotides, the best possible procedure is to add the new genes
to a plasmid with different antibiotic resistance, and co-transform the two plasmids into the
Streptomyces protoplasts.
A few attempts at such co-transformation were made, but yielded the same results
as the control: the second plasmid was never seen to express. This could be due to the
use of pSET152 [28] as the co-expression plasmid. While it has suitable restriction sites
for cloning as well as a suitably different antibiotic resistance gene (apramycin), pSET152
was originally created to be integrative rather than self-replicating, thus the efficiency of
gene transcription and translation is lower. Additionally, because of its original integrative 36 purpose, pSET152 lacks a promoter. This was remedied by others in our lab via the inclusion of the erythromycin promoter from Saccharopolyspora erythraea [39], but their studies only utilized it to express a single gene; the ideal co-expression vector would have the capability to express several nucleotides.
In fact, the ideal co-expression vector would be identical to pRM5, with the exception of the antibiotic resistance gene. A suitable resistance gene would provide selection for an antibiotic other than thiostrepton; the two historic choices seem to be apramycin (aac(3)
IV ) or hygromycin (hyg)[28], both of which are aminoglycosides that disrupt protein synthesis. Some candidates include p6perme or pRLE6 [12]. Conversely, it is possible that the vector might have to be engineered from a commercial vector like pJET or pUC19, and have the desired promoter, replicon, and restriction sites placed in optimal positions. The fact remains that co-expression is very likely the key to novel natural product biosynthesis. 37
REFERENCES
[1] Aguirrezabalaga, I., Allende, N., and Bran, A. F. Engineering Deoxysugar Biosynthetic Pathways from Antibiotic-Producing Microorganisms. Chemistry & Bi-
ology 9, 2 (2002), 721–729.
[2] Alexeev, I., Sultana, A., Mants¨ al¨ a,¨ P., Niemi, J., and Schneider, G. Acla- cinomycin oxidoreductase (aknox) from the biosynthetic pathway of the antibiotic acla-
cinomycin is an unusual flavoenzyme with a dual active site. Proceedings of the National
Academy of Sciences of the United States of America 104, 15 (2007), 6170–6175.
[3] Ames, B. D., Lee, M.-Y., Moody, C., Zhang, W., Tang, Y., and Tsai, S.-C. Structural and biochemical characterization of zhui aromatase/cyclase from the r1128
polyketide pathway. Biochemistry 50, 39 (2011), 8392–406.
[4] Bailly, C., Qu, X., Graves, D. E., Prudhomme, M., and Chaires, J. B. Calo- ries from carbohydrates: energetic contribution of the carbohydrate moiety of rebec-
camycin to dna binding and the effect of its orientation on topoisomerase i inhibition.
Chemistry & Biology 6, 5 (1999), 277–86.
[5] Basnet, D. B., Oh, T.-j. J., Vu, T. T., Sthapit, B., Liou, K., Lee, C. H., Yoo,
C. J., Sohng, J. K., Thu, T., Vu, H., Lee, H. C., and Yoo, J.-c. Angucyclines sch 47554 and sch 47555 from streptomyces sp . scc-2136 : Cloning , sequencing , and
characterization. Molecules and Cells 22 (2006), 154–162.
[6] Chu, M., Yarborough, R., and Schwartz, J. Sch 47554 Sch 47555, Two Novel Antifungal Antibiotics Produced from a Streptomyces sp. Journal of Antibiotics 46, 5
(1993).
[7] Crawford, J. M., Thomas, P. M., Scheerer, J. R., Vagstad, A. L., Kelleher,
L., Townsend, C. A., Jason, M., Scheerer, R., Vagstad, L., Kelleher, N. L., 38
and Town, C. A. Deconstruction Polyketide of Iterative Multidomain Function Synthase. Science (New York, N.Y.) 320, 5873 (2008), 243–246.
[8] Decker, H., and Haag, S. Cloning and characterization of a polyketide synthase gene from Streptomyces fradiae Tu-2717, which carries the genes for biosynthesis of the
angucycline antibiotic urdamycin A and a gene probably involved in its oxygenation.
Journal of Bacteriology 177, 21 (1995), 6126–36.
[9] Durr,¨ C., Hoffmeister, D., Wohlert, S.-E., Ichinose, K., Weber, M.,
Von Mulert, U., Thorson, J. S., and Bechthold, A. The glycosyltransferase UrdGT2 catalyzes both C- and O-glycosidic sugar transfers. Angewandte Chemie (In-
ternational ed. in English) 43, 22 (2004), 2962–5.
[10] Erb, A., Luzhetskyy, A., Hardter, U., and Bechthold, A. Cloning and se- quencing of the biosynthetic gene cluster for saquayamycin z and galtamycin b and the
elucidation of the assembly of their saccharide chains. ChemBioChem : a European
journal of chemical biology 10, 8 (2009), 1392–1401.
[11] Faust, B., Hoffmeister, D., Weitnauer, G., Westrich, L., Haag, S., Schnei-
der, P., Decker, H., Kunzel,¨ E., Rohr, J., and Bechthold, A. Two new tailoring enzymes, a glycosyltransferase and an oxygenase, involved in biosynthesis of
the angucycline antibiotic urdamycin a in streptomyces fradiae tu2717. Microbiology,
1 (2000), 147–154.
[12] Fleige, C., and Steinbuchel, A. Construction of expression vectors for metabolic engineering of the vanillin-producing actinomycete Amycolatopsis sp. ATCC 39116.
Applied Microbiology and Biotechnology 98, 14 (2014), 6387–95.
[13] Fu, H. Engineered Biosynthesis of Novel Polyketides: Dissection of the Catalytic Specificity of the act Ketoreductase. Journal of the American Chemical Society 116, 6
(1994), 4166–4170. 39
[14] Hoffmeister, D., Drager,¨ G., Ichinose, K., Rohr, J., and Bechthold, A. The C-Glycosyltransferase UrdGT2 Is Unselective toward D - and L -Configured Nucleotide-
Bound Rhodinoses. Journal of the American Chemical Society (2003), 4678–4679.
[15] Kallio, P., Patrikainen, P., Belogurov, G., Mants¨ al¨ a,¨ P., Yang, K., Niemi,
J., and Metsa-Ketel¨ a,¨ M. Tracing the evolution of angucyclinone monooxygenases: structural determinants for C-12b hydroxylation and substrate inhibition in PgaE.
Biochemistry 52, 26 (2013), 4507–16.
[16] Katsuyama, Y., and Ohnishi, Y. Type III polyketide synthases in microorganisms., vol. 515. Elsevier Inc., jan 2012.
[17] Kharel, M. K., Pahari, P., Lian, H., and Rohr, J. Enzymatic total synthesis of rabelomycin, an angucycline group antibiotic. Organic Letters 12, 12 (2010), 2814–
2817.
[18] Kharel, M. K., Pahari, P., Shaaban, K., Wang, G., Morris, C., and Rohr, J. Elucidation of post-PKS tailoring steps involved in landomycin biosynthesis. Organic
& Biomolecular Chemistry 10, 21 (2012), 4256–65.
[19] Kharel, M. K., Pahari, P., Shepherd, M. D., Tibrewal, N., Nybo, S. E.,
Shaaban, K., and Rohr, J. Angucyclines: Biosynthesis, mode-of-action, new natu- ral products, and synthesis. Natural Product Reports 29, 2 (2012), 264–325.
[20] Kieser, T., Bibb, M., Buttner, M., Chater, K., and Hopwood, D. Practical Streptomyces Genetics. John Innes Foundation, 2000.
[21] Kossel, A. Ueber die chemische zusammensetzung der zelle (the chemical composition of the cell). Archiv fr Physiologie (1891), 181–186.
[22] Kulowski, K., and Wendt-Pienkowski, E. Functional characterization of the jadI gene as a cyclase forming angucyclinones. Journal of the American Chemical Society,
5 (1999), 1786–1794. 40
[23] Lee, M.-Y., Ames, B. D., and Tsai, S.-C. Insight into the molecular basis of aro- matic polyketide cyclization: crystal structure and in vitro characterization of WhiE-
ORFVI. Biochemistry 51, 14 (2012), 3079–91.
[24] Liu, G., Sun, L., Wang, S., Chen, C., Guo, T., Ji, Y., Li, X., Huang, G., Wei,
H., Dai, Y., and Yuan, S. Hydroxylation modification and free radical scavenging activity of puerarin-7-O-fructoside. Folia Microbiologica 56, 4 (2011), 305–11.
[25] Lombo,´ F., Abdelfattah, M. S., Brana,˜ A. F., Salas, J. a., Rohr, J., and
Mendez,´ C. Elucidation of oxygenation steps during oviedomycin biosynthesis and generation of derivatives with increased antitumor activity. ChemBioChem : a Euro-
pean Journal of Chemical Biology 10, 2 (2009), 296–303.
[26] Lombo,´ F., Olano, C., Salas, J. a., and Mendez,´ C. Chapter 11. Sugar biosyn- thesis and modification. Methods in Enzymology 458, 9 (2009), 277–307.
[27] McDaniel, R., Ebert-Khosla, S., Hopwood, D. A., and Khosla, C. Engi- neered biosynthesis of novel polyketides. Science (New York, N.Y.) 262, 5139 (Dec
1993), 1546–50.
[28] Medema, M. H., Breitling, R., and Takano, E. Synthetic biology in Streptomyces bacteria., vol. 497. Elsevier Inc., jan 2011.
[29] Meurer, G., Gerlitz, M., Wendt-Pienkowski, E., Vining, L. C., Rohr, J.,
and Hutchinson, C. R. Iterative type II polyketide synthases, cyclases and ketore- ductases exhibit context-dependent behavior in the biosynthesis of linear and angular
decapolyketides. Chemistry & Biology 4, 6 (1997), 433–43.
[30] Patrikainen, P., Kallio, P., Fan, K., Klika, K. D., Shaaban, K. A., and Ma,
P. Tailoring Enzymes Involved in the Biosynthesis of Angucyclines Contain Latent Context-Dependent Catalytic Activities. Chemistry & Biology 19 (2012), 647–655. 41
[31] Patrikainen, P., Niiranen, L., Thapa, K., Paananen, P., Tahtinen,¨ P.,
Mants¨ al¨ a,¨ P., Niemi, J., and Metsa-Ketel¨ a,¨ M. Structure-based engineering of angucyclinone 6-ketoreductases. Chemistry & Biology 21, 10 (oct 2014), 1381–91.
[32] Pickens, L., and Tang, Y. Oxytetracycline biosynthesis. The Journal of biological chemistry 285, 36 (2010), 27509–15.
[33] Pickens, L. B., Kim, W., Wang, P., Zhou, H., Watanabe, K., Gomi, S., and
Tang, Y. Biochemical analysis of the biosynthetic pathway of an anticancer tetracy- cline SF2575. Journal of the American Chemical Society 131, 48 (Dec 2009), 17677–89.
[34] Ramos, A., Olano, C., Brana,˜ A. F., Mendez,´ C., and Salas, J. A. Modulation of deoxysugar transfer by the elloramycin glycosyltransferase ElmGT through site-
directed mutagenesis. Journal of Bacteriology 191, 8 (apr 2009), 2871–5.
[35] Rix, U., Remsing, L. L., Hoffmeister, D., Bechthold, A., and Rohr, J. Urdamycin L: a novel metabolic shunt product that provides evidence for the role of
the urdM gene in the urdamycin A biosynthetic pathway of Streptomyces fradiae TU
2717. Chembiochem : a European Journal of Chemical Biology 4, 1 (jan 2003), 109–11.
[36] Shao, L., Zi, J., Zeng, J., and Zhan, J. Identification of the herboxidiene biosyn- thetic gene cluster in Streptomyces chromofuscus ATCC 49982. Applied and Environ-
mental Microbiology 78, 6 (mar 2012), 2034–8.
[37] Trefzer, A., Hoffmeister, D., Kunzel,¨ E., Stockert, S., Weitnauer,
G., Westrich, L., Rix, U., Fuchser, J., Bindseil, K. U., Rohr, J., and
Bechthold, A. Function of glycosyltransferase genes involved in urdamycin A biosyn- thesis. Chemistry & Biology 7, 2 (feb 2000), 133–42.
[38] Udvarnoki, G., and Henkel, T. Biosynthetic origin of the oxygen atoms of aquayamycin: aspects for the biosynthesis of the urdamycin family and for
aquayamycin-containing angucycline antibiotics. The Journal of Organic Chemistry,
21 (1992), 1274–1276. 42
[39] Vara, J., Lewandowska-Skarbek, M., Wang, Y. G., Donadio, S., and
Hutchinson, C. R. Cloning of genes governing the deoxysugar portion of the ery- thromycin biosynthesis pathway in Saccharopolyspora erythraea (Streptomyces ery-
threus). Journal of Bacteriology 171, 11 (nov 1989), 5872–81.
[40] White-Phillip, J., Thibodeaux, C. J., and Liu, H.-w. Chapter 21. Enzymatic synthesis of TDP-deoxysugars. Methods in Enzymology 459, 09 (jan 2009), 521–44.
[41] Wohlert, S., Blanco, G., Lombo, F., and Ferna, E. Novel Hybrid Tetraceno- mycins through Combinatorial Biosynthesis Using a Glycosyltransferase Encoded by
the elm Genes in Cosmid 16F4 and Which Shows a Broad Sugar Substrate Specificity.
Journal of the American Chemical Society, 16 (1998), 10596–10601.
[42] Xu, Y., Zhou, T., Zhang, S., Xuan, L.-J., Zhan, J., and Molnar,´ I. Thioesterase domains of fungal nonreducing polyketide synthases act as decision gates
during combinatorial biosynthesis. Journal of the American Chemical Society 135, 29
(jul 2013), 10783–91.
[43] Yu, D., Xu, F., Zeng, J., and Zhan, J. Type III polyketide synthases in natural product biosynthesis. IUBMB Life 64, 4 (apr 2012), 285–95.
[44] Zhan, J. Biosynthesis of bacterial aromatic polyketides. Current Topics in Medicinal Chemistry 9, 17 (jan 2009), 1958–610.
[45] Zhang, Q., Pang, B., Ding, W., and Liu, W. Aromatic Polyketides Produced by Bacterial Iterative Type I Polyketide Synthases. ACS Catalysis 3, 7 (jul 2013),
1439–1447.
[46] Zhang, W., Watanabe, K., Wang, C. C. C., and Tang, Y. Heterologous biosyn- thesis of amidated polyketides with novel cyclization regioselectivity from oxytetracy-
cline polyketide synthase. Journal of Natural Products 69, 11 (nov 2006), 1633–6. 43
APPENDICES 44
Appendix A
Compounds 45
Table A.1: Numbered Compounds in this Thesis
№ Name Formula Structure
1 Sch 47554 C37H38O13
2 Sch 47555 C37H42O13
3 Aquayamycin C25H26O10
4 Urdamycin A C43H56O17
5 SEK15 C20H16O8
6 RM20 b/c C20H18O8
7 SEK4 C18H18O5
8 Mutacin C16H18O6
9 WJ85 C19H13NO7 46
№ Name Formula Structure
10 WJ35 C19H17NO8
11 Anhydrotetracycline C22H22N2O7
12 Oxytetracyline C22H24N2O9
13 UWM6 C19H18O7
14 Rabelomycin C19H14O6
15 Urdamycin L C19H14O7
16 SEK43 C20H16O7
17 GG31 C26H26O7
18 GG53 C19H14O7 47
Appendix B
Nuclear Magnetic Resonance Spectra for Structure Confirmation of GG31
As stated previously in the Experimental section, the nuclear magnetic resonance spec- trophotometer used for characterization of molecules in this study was a JEOL ECX-300 instrument. I wish to specially acknowledge Dr. Tong Zhou for his assistance. 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64