INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely afreet reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Hgher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Infonnation Company 300 North Zed) Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600
ANALYSIS OF THE POLYKETIDE SYNTHASE GENES OF THE DAUNORUBICIN PRODUCER, Streptomyces sp. strain C5: GENERATION OF PKS MUTANTS, AND ANALYSIS OF THE UNUSUAL ANTHRACYCLINE PRODUCTS MADE BY THESE PKS MUTANTS
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Vineet B. Rajgarhia, B.S.
*****
The Ohio State University
1998
Dissertation Committee:
Dr. Tyrrell Conway, Adviser Approved by
Dr. William Strohl
Dr. Tina Henkin Adviser y Dr. Joe Krvzcki Department of Microbiology UMI Number: 9834051
UMI Microform 9834051 Copyright 1998, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized copying under Title 17, United States Code.
UMI 300 North Zeeb Road Ann Arbor, MI 48103 ABSTRACT
Streptomyces sp. strain C5 and Streptomyces peucetius produce the anticancer antibiotics daunorubicin and doxorubicin, which are effectively used for the treatment of several types of cancer. The production of daunorubicin and doxorubicin begins with the sequential decarboxylative condensation of a propionyi-CoA starter moiety with nine malonyl-CoA extender moieties by the anthracycline specific Type II polyketide synthase
(PKS) gene products.
The PKS gene cluster of Streptomyces sp. strain C5 has two unique open reading frames. dpsC and dpsD, encoding a fatty acid KASHI homologue lacking an active site and an acyltransferase of unproven function, respectively. These two genes are directly downstream of the genes dpsA (ketoacylsynthase a), and dpsB (ketoacylsynthase
P), a position normally occupied by the gene encoding the acyl carrier protein. This unique arrangement is speculated to be responsible for Streptomyces sp. strain C5 and
Streptomyces peucetius utilizing propionate as the starter unit instead of an acetate, which is commonly used by most other Type II polyketide synthases. Several deletion mutants of Streptomyces sp. strain C5 were generated in which either dpsD alone, dpsCD, or dpsABCD were replaced with a neomycin resistance gene cassette. The dpsCD mutant strain produced daunorubicin, but significantly enough, also produced unique anthracyclines known as feudomycins. These unique anthracyclines are derived from the
PKS catalyzed decarboxylative condensation of an acetyl-CoA starter moiety with nine malonyl-CoA extender units.
Complementing the mutant strain with dpsC restored normal (propionate initiated) antfiracycline production, while complementation with dpsD failed to restore normal starter unit selection. Similarly, a combination of the PKS genes from
Streptomyces sp. strain C5, excluding the dpsC gene, expressed in S. lividans TK24. produced a mixture of both propionate-derived polyketide intermediate, aklanonic acid, as well as an acetate-derived polyketide intermediate, desmethylaklanonic acid. The production of these early intermediates of the daunorubicin and feudomycin pathways, respectively, confirmed our earlier findings.
These results suggest that the dpsC gene product is responsible for maintaining natural starter unit selection in Streptomyces sp. strain C5. In the absence of which, the PKS in the mutant strain can select both acetyl-CoA as well as propionyl-CoA to prime anthracycline production.
Ill Dedicated to my family for eternal love, support, and understanding.
IV ACKNOWLEDGMENTS
I am grateful to my adviser. Dr. William Strohl, for encouragement, and perpetual intellectual support, which made this dissertation possible.
1 also thank Drs. Tyrrell Conway, Tina Henkin and Joe Kryzcki for their intuitive ideas, and for their time and efforts spent on guiding my research.
1 am indebted to Nigel Priestley for his unceasing support in my research efforts. 1 wish to thank Don Ordaz for his help with fermentations and with the slides 1 used for several of my seminar presentations.
1 also wish to thank the many friends I made in Baltimore, Maryland and Columbus, Ohio for the wonderful experiences and everlasting moments 1 shared with them.
I am grateful to my family members whose constant support and love helped achieve all.
1 thank the past members of the Strohl Lab, Drs. Richard Plater, Yun Li, Gary Kleman, and Mike Dickens who passed on their experiences as well as words of wisdom.
1 am beholden to the current members of the laboratory Anton Woo and Robbie Walczak for their kinship, their friendship, and their support specially during the final year of my graduate school career.
1 am eternally indebted to Balakrishna Shetty and Jignesh Patel for their kinship and their support, which made this process so much smoother. VITA
January 4. 1968 ...... Bom - Calcutta, India
1990...... B.S. Pharmacy, Bombay University.
1990 - 1991...... Graduate Teaching Associate. The University of Maryland at Baltimore
1991 - present...... Graduate Teaching and Research Associate. The Ohio State University
PUBLICATIONS
Research Publications
1. Rajgarhia, V., Priestley, N. D.. and W. R. Strohl. 1995. Efficient synthesis of radiolabeled propionyl-CoA. Anal. Biochem. 224: 159-162.
2. Rajgarhia, V., and W. R. Strohl. 1997. Minimal Polyketide synthase genes of Streptomyces sp. strain C5. J. Bacteriol. 179:2690-2696.
3. Strohl. W. R.. Dickens, M. L.. Rajgarhia, V.. Woo. A., and N. Priestley. Anthracyclines. In: Strohl WR, ed. Biotechnology of Industrial Antibiotics. 2nd ed. New York: Marcel Dekker, Inc, 1997:577 - 657.
4. Strohl, W. R., Dickens, M. L., Rajgarhia. V., Walczak, R., Woo, A., and N. D. Priestley. 1998. Biochemistry, molecular biology and protein - protein interactions in doxorubicin biosynthesis: Proceedings of the Biotechnology of microbial products (BMP 97). C. R. Hutchinson and J. McAlpine. eds. Dev. Indust. Microbiol. 35:15-22.
FIELDS OF STUDY
Major Field: Microbiology
VI TABLE OF CONTENTS
Page Abstract...... ii
Dedication ...... iv
Acknowledgments ...... v
Vita...... vi
Table of Contents ...... vii
List of Tables ...... xiv
List of Figures ...... xvi
Chapters:
I. Introduction ...... 1
A. Anthracyclines and the discovery of daunorubicin
and doxorubicin ...... 4
1. Daunorubicin and doxorubicin producing Streptomyces spp 6
2. Mechanism of action of daunorubicin and doxorubicin ...... 9
3. Quantitative structure/activity relationships ...... 10
4. Toxicity associated with doxorubicin and daunorubicin ...... 12
5. Uses of daunorubicin and doxorubicin ...... 14
VII B. Biosynthetic pathway for daunorubicin and doxorubicin production 15
1. Condensation of acetates and propionates to produce aklanonic
acid...... 16
a. Architecture of the Streptomyces sp. strain C5 polyketide
synthase genes ...... 22
b. Minimal PKS/role of the dpsC and dpsD gene products.. 26
2. Early pathway gene products that convert aklanonic acid to e-
rhodomycinone ...... 28
3. TDP-daunosamine biosynthesis and glycosylation of s-
rhodomycinone ...... 31
4. Conversion of rhodomycin D to doxorubicin ...... 35
5. Predicted gene requirements for daunorubicin biosynthesis 38
C. Hybrid polyketide synthases ...... 43
1. Design rules for combining PKSs functions to make hybrid
polyketides ...... 45
a. Carbon chain length ...... 46
b. Ketoreduction ...... 47
c. First ring cyclization ...... 47
d. First ring aromatization ...... 48
e. Second ring cyclization ...... 49
Vlll f. Choice of starter unit ...... 49
2. Hybrid PKS: actinorhodin pathway and daunorubicin ...... 50
D. In vitro studies determine the programming rules for the PKS- catalyzed
polyketide biosynthesis ...... 53
E. Goals of this work ...... 55
2. Minimal Streptomyces sp. strain C5 daunorubicin polyketide biosynthesis genes
required for aklanonic acid biosynthesis ...... 57
Introduction ...... 57
Materials and Methods ...... 67
Bacterial strain and growth conditions ...... 67
General genetic manipulations ...... 68
Gene replacement methods ...... 71
Detection of anthracyclines and aklanonic acid...... 72
Radiolabeling of aklanonic acid ...... 74
Enzymatic conversion of aklanonic acid ...... 74
MS analysis ...... 75
Results and discussion ...... 75
PKS gene structure ...... 75
IX Gene replacement...... 76
Heterologous biosynthesis of aklanonic acid by minimal PKS 82
Identity of the product ...... 90
Are DauZ and DpsG or their homologues required for efficient
aklanonic acid biosynthesis? ...... 90
Mechanistic and evolutionary implications ...... 92
3. Streptomyces sp. strain C5 polyketide synthase-specific dpsC gene product helps
maintain natural starter unit selection in polyketide synthesis ...... 96
Introduction ...... 96
Materials and Methods ...... 103
Bacterial strains and growth conditions ...... 103
General genetic manipulations ...... 104
Analysis of products isolated using LC/MS. HRFABMS. and NMR.. 104
Isolation and detection of aklanonic acid ...... 105
Isolation and detection of desmethylaklanonic acid ...... 106
Production of anthracyclines and anthracyclinones ...... 106
Isolation of anthracyclines and anthracyclinones ...... 110
Results and discussion ...... 112
Heterologous biosynthesis of aklanonic acid and desmethyl
aklanonic acid by PKS expression in S. lividans TK24 and s. coelicolor CW999 ...... 112
Anthracycline production in the dpsD null mutant...... 123
Anthracycline production in null mutants of dpsCD ...... 124
Significance of the results ...... 134
Comparison of the Streptomyces sp. strain C5 PKS gene cluster with
aclacinomycin and nogalamycin PKS gene clusters ...... 142
Involvement of other cyclases ...... 143
Model for synthesis of aklanonic acid and desmethylaklanonic acid.. 143
Model for feudomycinone C and feudomycin D biosynthesis ...... 146
4. (A) Synthesis of radiolabeled acetyl-and propionyl-CoA; (B) In vitro aklanonic
acid biosynthesis from recombinant S. lividans TK24 transformed with PKS
genes from Streptomyces sp. strain C5; (C) Heterologous combinations of
PKSs from Streptomyces sp. strain C5 and Streptomyces coelicolor A3(2)
that produce aloesaponarin II ...... 150
A. Synthesis of radiolabeled acetyl- and propionyl-CoA
Introduction ...... 150
Materials and Methods ...... 152
Preparation of the 1,1 -carbonyldiimidazole ...... 152
Synthesis of [l-'^C]propionyl-CoA and [3,3.3-dJpropionyl-CoA 153
Synthesis of [l-''*C]acetyl-CoA and [2,2,2-d^]acetyl-CoA ...... 154
XI Results and discussion ...... 154
B. In vitro synthesis of aklanonic acid and SEK 43 by Streptomyces sp.
strain C5 PKS...... 163
Introduction ...... 163
Materials and Methods ...... 167
Culture growth ...... 167
Preparation of the cell free PKSs preparation ...... 168
In vitro aklanonic acid biosynthesis ...... 169
Analysis of the aklanonic acid biosynthesized ...... 172
In vitro biosynthesis of polyketides derived from acetates ...... 172
Analysis of the products ...... 173
Results and discussion ...... 176
In vitro aklanonic acid biosynthesis from PKSs obtained from
S. lividans TK24(pANT782). S. lividans TK24(pANT788).
and S'.TK24(pANT785) ...... 176
In vitro SEK 43 biosynthesis from PKSs obtained from
S. lividans TK24(pANT782), S. lividans TK24(pANT788).
and S. lividans TK24(pANT785) ...... 178
C. Hybrid polyketide production by expressing S. coelicolor actinorhodin producing genes in Streptomyces sp. strain C5 PKS mutants ...... 180
XU Introduction ...... 180
Materials and Methods ...... 186
Bacterial strains and growth conditions ...... 186
General genetic manipulations ...... 187
Gene disruption methods ...... 189
Detection of aloesaponarin 11 biosynthesis ...... 189
LC/MS analysis ...... 192
Results and discussion ...... 193
PKS gene structure ...... 193
Gene replacement...... 196
Heterologous interactions of PKSs: biosynthesis of aloesaponarin 11.. 192
5. Summary ...... 203
List of References ...... 208
Appendices ...... 228
Appendix A. Table of bacterial strains used ...... 228
Appendix B. Table of plasmids and vectors used ...... 230
Appendix C. Maps of plasmids constructed and used ...... 236
Appendix D. Mass spectral data for compounds analyzed ...... 281
xm LIST OF TABLES
Table Page
1.1 Predicted genes required for doxorubicin biosynthesis ...... 41
1.2 Characteristics of products of daunorubicin PKS genes in heterologous hosts ...... 86
3.1 The development of media optimized for feudomycin production in stirred tank fermentors ...... 107
3.2 Characteristics of metabolites produced when Streptomyces sp. strain C5 minimal PKS genes are expressed in heterologous host S. lividans TK24 114
3.3 Production of polyketide intermediates by Streptomyces sp. strain C5 PKS genes expressed in S’, lividans TK24 and S. coelicolor CH999...... 121
3.4 Quantative estimation of aklanonic acid and desmethylaklanonic acid produced by recombinant S. lividans TK24 (pANT785) ...... 123
3.5 Characteristics of the major anthracycline product of Streptomyces sp. strain C5VR5 {dpsCD mutant)...... 125
3.6 '"C NMR data for feudomycinone C ...... 127
3.7 ‘H NMR chemical shift assignments (ppm) for feudomycin D ...... 130
3.8 Major products of Streptomyces sp. strain C5 and Streptomyces sp. strain C5 mutants complemented with PKS genes ...... 132
3.9 Comparison of the anthracycline production by Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR5 ...... 135
4.1. Synthesis of propionyl-CoA and acetyl-CoA ...... 155
XIV 4.2 Analysis of the products from in vitro reactions with Streptomyces sp. strain C5 PKS preparation incubated with [SJ.S-djJpropionyi-CoA and malonyl-CoA ...... 170
4.3 Analysis of the products from in vitro reactions with Streptomyces sp. strain C5 PKS preparation incubated with [2,2.2-d3]acetyl-CoA and malonyl-CoA ...... 174
4.4 Analysis of the aloesaponarin II produced by introducing p ANT817 (S. coelicolor actl-orfsl and IF) into Streptomyces sp. strain C5VR5 {dpsCD mutant) and Streptomyces sp. strain C5VR10 (dpsABCD mutant)...... 198
B.l Table of bacterial strains used ...... 228
C.l Table of plasmid vectors used ...... 236
XV LIST OF FIGURES
Figures Page
1.1 Relationship between growth, differentiation and antibiotic production for streptomycetes ...... 2
1.2 Structures of daunorubicin and related anthracyclines ...... 7
1.3 Hypothetical steps involved in aklanonic acid biosynthesis in Streptomyces sp. strain C5 and Streptomyces peucetius ...... 17
1.4 A comparative pathway of fatty acid synthesis and polyketide biosynthesis... 20
1.5 Comparison of the polyketide synthase genes from several antibiotic producing Streptomyces spp ...... 23
1.6 Pathway for the conversion of aklanonic acid to s-rhodomycinone ...... 29
1.7 Hypothetical pathway for the biosynthesis of TDP-daunosamine ...... 32
1.8 Hypothetical pathway for the biosynthesis of doxorubicin from rhodomycin D ...... 36
1.9 Daunorubicin and doxorubicin biosynthetic gene cluster of Streptomyces sp. strain C5 and Streptomyces peucetius ATCC 29050...... 39
1.10 Actinorhodin and aloesaponarin II biosynthetic pathway ...... 51
2.1 Type II polyketide synthase (PKS) genes from actinomycetes ...... 59
2.2 Southern blot of genomic DNA of strains Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR2 and the restriction maps of the plasmid, parental chromosome and mutant chromosome, showing the expected results obtained by the double crossover ...... 78
XVI 2.3 Southern blot of genomic DNA of strains Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR5. C5VR6, C5VR7. Restriction maps of the plasmid, parental chromosome and mutant chromosome, showing the expected results obtained by the double crossover ...... 80
2.4 Restriction map generated by complete nucleotide sequence data of part of the daunorubicin biosynthetic gene cluster of Streptomyces sp. strain C5 and the genes depicting the minimal biosynthetic genes required for aklanonic acid biosynthesis in heterologous strains ...... 84
2.5 Biosynthesis of tetracenomycin F2 compared to biosynthesis of aklanonic acid...... 88
3.1 LINEUP analysis of the PILEUP comparison (Devereux et al.. 1984) of the deduced amino acid sequences of the dpsC gene product with E. coli (Ec) and spinach (Sp) FabH (P-ketoacyl synthase III) enzymes ...... 98
3.2 Alignment of the acyltransferase active site motifs from several Streptomyces spp. PKS acyltransferases ...... 100
3.3 Structure of the unusual acetate-initiated intermediates, desmethylaklanonic acid and SEK 43. made by the minimal PKS gene products from Streptomyces sp. strain C5 expressed in S. lividans TK24 and S. coe//co/or CH999 ...... 116
3.4 Restriction map, generated by complete nucleotide sequence data, of part of the daunorubicin polyketide synthase gene cluster from Streptomyces sp. strain C5 ...... 119
3.5 Model showing the PKS gene organization, starter moiety, and an example intermediate for daunorubicin, produced by several Streptomyces spp ...... 138
3.6 Hypothetical pathway for biosynthesis of aklanonic acid and desmethylaklanonic acid starting from acetyl and propionyl-CoA and malonyl-CoA, respectively ...... 144
3.7 Proposed pathway for feudomycinone C and feudomycin D biosynthesis 147
4.1 Reverse phase HPLC chromatograph of synthetic [1 -''*C]propionyl-CoA and [l-''*C]acetyl-CoA ...... 158
XVll 4.2 ‘H NMR analysis of synthesized acetyl-CoA (A. 1.2 mg.ml'') and propionyl-CoA (B, 0.9 mg.mL‘‘) in ’HiO at 298 K ...... 160
4.3 Simplified pathway for tetracenomycin biosynthesis (A), and Structure of SEK43 (B) produced from 10 acetates along with the molecular formula and expected molecular mass ...... 165
4.4 Actinorhodin biosynthetic pathway ...... 181
4.5 Model depicting the heterologous combination between the PKS components (depicted by the genes) from S. galilaeus and S. coelicolor ...... 184
4.6 Southern blot of genomic DNA of strains Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR10. Restriction maps of the plasmid, parental chromosome and mutant chromosome ...... 190
4.7 Comparison of the Streptomyces sp. strain C5 and Streptomyces galilaeus (ATCC 31133) PKS gene clusters ...... 194
4.8 Model depicting the interactions among PKS components from Streptomyces sp. strain C5VR5 and C5VR10 and Streptomyces coelicolor 200
C.l MapofpANT716 ...... 237
C.2 Map of pANT726 ...... 238
C.3 MapofpANT740 ...... 239
C.4 Map of pANT749 ...... 240
C.5 Map of pANT750 ...... 241
C.6 MapofpANT751 ...... 242
C.7 Map of pANT752 ...... 243
C.8 MapofpANT753 ...... 244
C.9 MapofpANT754 ...... 245
C.IO MapofpANT755 ...... 246
XVIII c .l I.....MapofpANT756 ...... 247
C.I2.... MapofpANT765 ...... 248
C.13.....MapofpANT767 ...... 249
C.14.... MapofpANT770 ...... 250
C.15.... MapofpANT771 ...... 251
C.16 Map of pANT776 ...... 252
C .l7 .....MapofpANT777 ...... 253
C.18 .....MapofpANT778 ...... 254
C.19 Map of pANT779 ...... 255
C.20 Map of pANT780 ...... 256
C.21 MapofpANT781 ...... 257
C.22 MapofpANT782 ...... 258
C.23 MapofpANT783 ...... 259
C.24 Map of pANT784 ...... 260
C.25 Map of pANT785 ...... 261
C.26 MapofpANT786 ...... 262
C.27 MapofpANT787 ...... 263
C.28 MapofpANT788 ...... 264
C.29 Map of pANT790 ...... 265
C.30 Map of pANT791 ...... 266
C.31 MapofpANT795 ...... 267
XIX C.32 MapofpANT796 ...... 268
C.33 Map of PANT797 ...... 269
C.34 MapofpANT735 ...... 270
C.35 MapofpANT1112 ...... 271
C.36 M apofpA N TlIH ...... 272
C.37 M apofpA N TlllS ...... 273
C.38 MapofpANT1116 ...... 274
C.39 M apofpA N T lin ...... 275
C.40 M apofpA N TllI8 ...... 276
C.41 MapofpANTlI20 ...... 277
C.42 MapofpANT1121 ...... 278
C.43 MapofpANT8I7 ...... 279
C.44 Map of pANT849 ...... 280
D.l Mass spectral data for feudomycinone C ...... 282
D.2 Mass spectral data for feudomycin D (LC/MS) ...... 283
D.3 Mass spectral data for feudomycin D (MS/MS of fragment with molecular mass 516)...... 284
D.4 Mass spectral data for feudomycin D (MS/MS of fragment with molecular mass 369)...... 285
D.5 Mass spectral data for aklanonic acid ...... 286
D.6 Mass spectral data for aklanonic acid from S. lividans TK24(pANT782) [all PKS genes] ...... 287
XX D.7 Mass spectral data for desmethylaklanonic acid from 5. lividans TK24(pANT785) [dpsCD deleted]...... 288
D.8 Mass spectral data for aklanonic acid from S. lividans TK24(pANT788) deleted]...... 289
D.9 Mass spectral data for aklanonic acid from S. lividans TK24(pANT785) complemented with dpsC and dpsD ...... 290
D.IO Mass spectral data for aklanonic acid from S. lividans TK24(pANT785) complemented with ...... 291
D.l 1 Mass spectral data for authentic aklanonic acid made by Streptomyces sp. strain C5A54 in vivo using propionic-[3,3,3-dJacid ...... 292
D.l2 Mass spectral data for aklanonic acid from in vitro biosynthesis using PKSs from S. lividans TK24(pANT782) and [3.3,3-dJpropionyl-CoA 293
D.l3 Mass spectral data for SEK 43 obtained from C. R. Hutchinson ...... 294
D.l4 Mass spectral data for SEK 43 from in vitro biosynthesis using PKS from S. lividans TK24(pANT782) and [2,2.2-d3]acetyl-CoA...... 295
D.l5 Mass spectral data for authentic aloesaponarin 11 ...... 296
D. 16 Mass spectral data for aloesaponarin 11 produced by Streptomyces sp strain C5VR5 and Streptomyces sp strain C5VR10 ...... 297
XXI CHAPTER 1
INTRODUCTION
Members of the genus Streptomyces produce approximately 55% of the 11.900 published antibiotics that have been discovered through 1994 (Berdy. 1995). Antibiotic production by Streptomyces occurs when the organism is in the late logarithmic or early stationary phase of growth. Streptomyces spp. are Gram-positive bacteria which grow as filamentous cellular structures known as mycelia. When grown on solid surface, the mycelia burrow imder the surface to form substrate mycelia. With further growth and nutrient limitation the organism produces aerial mycelia which rise above the solid surface, coil and septate to form chains of spores that can be dispersed to propagate new colonies (Bartel. 1989). Under suitable conditions these spores can germinate and produce new substrate mycelia.
Antibiotic production occurs concurrently with the development of aerial mycelia and spore formation and continues well after the organism ceases to grow.
Concurrent with this antibiotic production is the differentiation of the organism to form spores. The relationship between growth, development, and antibiotic production is shown in Figtire 1.1. Fig. 1.1. Relationship between growth, differentiation and antibiotic production for streptomycetes. Panel A depicts the increase in cell mass and antibiotic production in liquid culture. Panel B shows the differentiation undergone by the organism on solid medium. Antibiotic production and differentiation occurs as mycelial growth slows down (from W. R. Strohl). Dry Wt.
Substrate Anthracyclines
0 24 4 8 72 9 6 120 Hours Free Spore
Aerjal Coiling Sporulation Mycelium Substrate Mycelium
Fig. 1.1. Relationship between growth, differentiation and antibiotic production for streptomycetes. An important class of antibiotics produced by Streptomyces spp. is the polyketides. which includes the macrolides (e.g., tylosin, erythromycin), isochromane quinones (e.g., actinorhodin), anthracyclines (e.g., daunorubicin, doxorubicin), polyethers
(e.g., monensin), and tetracyclines (Bartel, 1989).
A. Anthracyclines and the discovery of daunorubicin and doxorubicin.
“Anthracycline” is the term coined by Brockmann and Brockmann Jr. to describe a group of chemical compounds that are glycosidic derivatives of 7. 8. 9, 10 - tetrahydronaphthacene quinones (Brockmann and Brockmann Jr., 1963). Hans
Brockmann and Klaus Bauer first isolated the anthracyclines P-rhodomycin 1. isorhodomycin, and some of the latters derivatives from Streptomyces purpurascens
(ATCC 25489) in the early 1950s (Brockmann and Bauer. 1950). p-Rhodomycin 1. the first anthracycline to be characterized, displayed strong antibacterial activity against
Staphylococcus aureus, but due to its high toxicity was not clinically useful (Brockmann and Bauer, 1950).
In the mid-1950s, researchers at Farmitalia Research Laboratories (now
Pharmacia Upjohn) in Italy isolated a strain of Streptomyces spp. that produced a red pigment. This strain was later named Streptomyces peucetius (ATCC 29050) (Grein et al., 1963). The red pigment possessed strong antibacterial, anti fungal and cytotoxic properties (Arcamone et al., 1961; Cassinelli et al., 1963; DiMarco et al., 1963; DiMarco et al., 1964b), and was chemically categorized to belong to the group described earlier as anthracyclines (Brockmann and Brockmann, 1963). This red anthracycline compound was named daunomycin by the Farmitalia researchers. Concurrently, a group of researchers (Dubost et al., 1963) at Rh on e-Poulenc in France isolated a red anti tumor drug from cultures of Streptomyces coenileorubidus that they named rubidomycin.
Rubidomycin (Dubost et al., 1963; Despois et al.. 1967; Maral et al.. 1967) and daunomycin (Arcamone et al.. 1961; Cassinelli et al.. 1963; DiMarco et al.. 1963; Grein et al.. 1963; DiMarco et al.. 1964b) were later found to be identical and were renamed daunorubicin, acknowledging the efforts of both research groups (Weiss. 1992).
Daunorubicin was found to possess potent cytotoxic activity in cell cultures and despite its organ toxicity, was considered useful, because of its activity against both solid and ascites tumor (DiMarco et al.. 1963; DiMarco et al.. 1964a; DiMarco et al.. 1964b; Maral et ai. 1967).
In 1969, as a result of a mutation and screening program undertaken at Farmitalia, an N-nitroso-N-methyl-urethane-induced mutant of S. peucetius (ATCC 29050) was generated (Arcamone et al.. 1969a; Arcamone et al, 1969b; Arcamone, 1981a). This isolate was named S. peucetius subsp. caesius (ATCC 27952) and was found to produce a
14-hydroxy analog of daunorubicin with significantly improved antitumor activity over daunorubicin (Arcamone et al., 1969a; Arcamone et al., 1969b; Bonadonna et al.. 1969;
Blum and Carter, 1974). This 14-hydroxy analog of daunorubicin was named adriamycin
(doxorubicin). In the years since the discovery of daunorubicin and doxorubicin, more than 2000 anthracycline analogs have been isolated from natural sources or synthesized in laboratories around the globe (Weiss, 1992; Strohl et al., 1997). Despite the concerted effort of researchers around the world, only five anthracyclines have found clinical acceptance worldwide since the discovery and acceptance of doxorubicin. This exclusive list of drugs includes idarubicin (4-demethoxydaunorubicin), epirubicin (4'- epidoxorubicin), pirarubicin (tetrahydropyranyldoxorubicin), zorubicin (rubidazone) and aclarubicin (aclacinomycin A) (Fig. 1.2; Weiss et al., 1986; Suarato, 1990; Weiss, 1992;
Strohl et al., 1997).
A. 1. Daunorubicin and doxorubicin iprodncmg Streptomyces spp.
A wide range of Streptomyces spp. produce daunorubicin. doxorubicin and structurally related analogs. These strains include: S. peucetius (Arcamone et a/.. 1961;
Cassinelli and Orezzi, 1963; DiMarco et al., 1963; Grein at al.. 1963; DiMarco et ai,
1964a; DiMarco et al., 1964b), Streptomyces bifurcus strain 23219 (Mancy and Florent.
1975), S. coeruleorubidus strains ME 130-A4 (Komiyama et al., 1977; Takahashi et al..
1977), 8899 (Dubost et al., 1963; Despois et al., 1967; Pinnert and Ninet, 1976), 31723
(Dubost et al., 1963; Despois et al., 1967; Pinnert and Ninet, 1976), and JA10092
(Blumauerova et al., 1977), Streptomyces griseoruber (Higashide et al., 1972), S. griseus
IMET JA5142 (Strauss and Fleck, 1975), a patented Streptomyces griseus strain (Mancy and Ninet, 1976), Streptomyces insignis (ATCC 31913) (Kern et al., 1977; Tunac et al.. Fig. 1.2. Structures of daunorubicin and related anthracyclines (Strohl et ai., 1997), Rs
Ri R2 R3 R4 Rs
Doxorubicin =0 -OH -O CH 3 H -OH Daunorubicin =0 -H -O CH 3 H -OH Carminomycin =0 -H -OH H -OH Idarubicin =0 -H -H H -OH Epirubicin =0 -OH -O CH 3 -OH -H
Pirarubicin =0 -OH -O CH 3 H
HO Baumycin Ai =o -H -OCH3 H
Zorubicin —nnhc -H -OCH3 H -OH
Fig. 1.2. Structures of daunorubicin and related anthracyclines 1985), Streptomyces viridochromogenes (Liu and Rao, 1974), Streptomyces sp. strain C5
(ATCC 49111) (McGuire et al.. 1980; McGuire et al.. 1991), and Streptomyces sp. D788
(Fujii et al.. 1986).
A. 2. Mechanism of action of daunorubicin and doxorubicin.
Anthracyclines, in general, have great affinity for DNA, and intercalate in the
DNA double helix noncovalently. Doxorubicin and daunorubicin. predictably, have a very high affinity for cellular DNA. These DNA-anthracycline adducts have a wide range of physiological effects within the cell including inhibition of DNA and RNA polymerase activity (Strohl et al.. 1997). In addition to the above action daunorubicin and doxorubicin also inhibit periribosomal DNA and RNA synthesis, alter cell membranes and generate free radicals due to their quinone action (Doroshow. 1995; Cutts and Phillips. 1995).
However, despite the effects of anthracyclines on DNA. the primary activity of anthracyclines on cellular DNA is believed to be through topoisomerase Il-induced DNA strand breakage (Drlica and Franco, 1988; Cullinane et al.. 1994; Chen and Liu. 1994;
Pommier, 1995). Topoisomerase II, a key enzyme in eucaryotic cells, catalyzes double strand breakage and religation (Liu, 1989). This enzyme is actively involved in eucaryotic cell processes such as chromosomal condensation, chromatid separation, segregation of mitotic and meiotic spindles and potentially also gene expression (Strohl et al., 1997). Doxorubicin and daunorubicin trap topoisomerase II in a drug-enzyme-DNA complex within the cell referred to as a “cleavable complex”. The formation of this cleavable complex interferes with the normal activity of topoisomerase II which leads to inhibition of DNA and RNA synthesis resulting in cessation of cell division (Drlica and
Franco, 1988; Cummings et al., 1991; Chen and Liu, 1994; Osheroff et al.. 1994;
Pommier, 1995; Strohl et al., 1997). The formation of this drug-enzyme-DNA complex also ensures a high concentration of doxorubicin and daunorubicin in the DNA. which results in chromosomal abnormalities and inhibition of DNA replication eventually leading to cell death via apoptosis (Hickman, 1992; Osheroff et al., 1994; Strohl et al..
1997).
Moreover, recent findings suggest that doxorubicin and daunorubicin are exceptional inhibitors of DNA helicases (Bachur et al., 1995). Since DNA helicases are responsible for dissociation of double stranded DNA into single stranded DNA. any alteration in its activity can produce effects such as increased helix rigidity, helix unwinding, and deformation and lengthening of DNA. These effects on DNA are commonly seen with doxorubicin and daimorubicin (Bachtir et al.. 1995. Strohl et al..
1997).
A. 3. Quantitative Structure -Activity Relationships (QSARs) of daunorubicin and analogs.
Structure - activity relationships of daunorubicin and its analogs are based either on modifications made on the aglycone moiety or modifications made of the
10 daunosamine sugar. Some of these analogs of daunorubicin and doxorubicin have gained clinical acceptance or are currently in clinical trials (Strohl et al., 1997). The important modifications and thier effects on the activity of the analogs are described as follows:
• Substitution at C4 with hydrophobic groups increased activity; e.g., idarubicin, with
no substituent group at C4 position, has 50 - 200 fold greater activity than
daunorubicin (Formelli et ai., 1978; Oki, 1982).
• 0-Methyiation of C9 hydroxy group leads to analogs with 100-fold lower activity in
vitro than daunorubicin and complete cessation of antitumor activity (Zunino et al..
1981).
• Side chains at C9 do not seem to have an impact on in vitro or in vivo activity; e.g..
feudomycins (C9 methyl) and auramycins (C9 acetonyl) have similar antitumor
activity as daunorubicin and doxorubicin (Oki et al., 1981; Fujiwara et al., 1981a;
Fujiwara e/a/.. 1981b; Matsuzawa e/a/.. 1981).
• Modification of the C14 substituent of daunombicin by desmethylation and
concurrent oxidation of C13 substituent leads to analogs with carboxy side chain and
significantly increased in vitro activity (DuVemay. 1981). Méthylation of the
carboxy substituent to a methylester produced analogs with activity similar to
doxorubicin and daunorubicin (DuVemay, 1981).
• Bromine, sulphur or nitrogen substitution at C14 results in analogs with lower
activity when compared to doxorubicin, which possesses a 14-hydroxyl moiety
(Horton and Priebe, 1981).
II • CIO(R) substitution, especially with a carbomethoxy moiety, produces analogs that
inhibit topoisomerase II binding of DNA, while the absence of this substitution
makes analogs that trap topoisomerase II -DNA (Sehested and Jensen. 1996). The
stereochemistry of the substituent at CIO position is critical; compounds with a 10(S)
substituent lack biological activity (DuVemay et a/., 1980; Arcamone 1981b;
DuVemay et al., 1982).
• Anthracyclinones (compounds lacking daunosamine sugar moiety) are not
biologically active (Strohl et al.. 1997).
• 4'-Epidoxombicin, a derivative of doxorubicin which has the L-arabino form of the
daunosamine sugar, has excellent pharmacological properties (Arcamone et al..
1975; Arcamone 1981b;).
• The 4'-deoxy analog of doxombicin, esombicin, having the L-threo form of the
daunosamine sugar, has demonstrated lower cardiotoxicity that doxorubicin (Weiss
etal., 1986).
A. 4. Toxicity associated with doxorubicin and daunorubicin therapy.
Doxorubicin and daunorubicin cause a range of side effects including nausea and vomiting, gastrointestinal disturbances, myelosuppression, hepatotoxicity. and cumulative cardiotoxicity (Arcamone 1981a; Arcamone 1981b; Strohl et al., 1997).
There are three forms of cardiotoxicity: (i) acute or subacute cardiotoxicity, which usually occurs after a single dose or course of anthracycline therapy; (ii) chronic cardiotoxicity.
12 which is defined as cardiotoxicity that occurs within one year of initial treatment; and (iii) late-onset cardiotoxicity that occurs after one year of the original treatment, and often after a prolonged asymptomatic period (Shan and Lincoff, 1996). Cardiotoxicity. thus, is not only a significant short-term problem, but a long-term problem as well and limits lifetime dosages of daunorubicin to 500-to-600 mg/m’ and doxorubicin to 550 mg/m’.
Incidence of significant cardiomyopathy (heart disease) caused by doxorubicin treatment is 7%, 15%. and 30-to-40% at 550, 600. and 700 mg/m’, respectively (Fischer et al.,
1993).
The single most studied and well described cause for cardiotoxicity is the generation of free radicals by the quinone redox cycling of anthracyclines (Keiser et ai.
1990; Hale and Lewis. 1994; Doroshow. 1995). especially when complexed with iron
(Keiser et at., 1990; Hale and Lewis, 1994). Most anthracyclines undergo bio transformation by quinone action to a semiquinone free radical that can redox cycle with molecular oxygen to generate a cascade of reactive oxygen species, including superoxide, hydrogen peroxide, and hydroxyl radical. Additionally, free radicals can form by intramolecular reduction of the iron that is chelated to the anthracycline (Keiser et al.,
1990; Doroshow, 1995; Strohl et al, 1997). Free radicals generated by anthracyclines cause lipid peroxidation and effects various cellular membranes leading to cell death
(Keiser et al., 1990; Doroshow. 1995; Strohl et al, 1997). Heart muscle, in particular, has a large number of mitochondria in which these potentially cardiotoxic free radicals can be generated. Moreover, the heart has low levels of antioxidant enzymes, which normally
degrade the destructive free radicals, thus allowing accumulation of these potentially
toxic factors (Hale and Lewis. 1994; Strohl et al. 1997).
A. 5. Uses of doxorubicin and daunorubicin.
Daunorubicin is used in the U.S. for treatment of acute nonlymphocytic leukemia
(adults) and acute lymphatic leukemia (adults and pediatric). Daunorubicin is also effective against chronic myelogenous leukemia, Ewing's sarcoma, neuroblastoma, non-
Hodgkin's lymphoma, and Wilms' tumor (Fischer et al.. 1993, Strohl et al.. 1997).
Doxorubicin is indicated in the U.S. for treatment of acute nonlymphocytic leukemia and
acute lymphatic leukemia. Wilms' tumor, neuroblastoma, soft-tissue and bone sarcomas,
breast sarcoma, ovarian carcinoma, transitional cell bladder carcinoma, thyroid carcinoma, Hodgkin's disease, non-Hodgkin's lymphoma, gastric carcinoma, small-cell
lung cancer. Doxorubicin also may be useful for treatment for chronic myelogenous
leukemia, multiple myeloma, rhabdomyosarcoma, Ewing's sarcoma. Kaposi's sarcoma, and trophoblastic neoplasms, as well as for cancers of the esophagus, endometrium, liver, cervix, islet cell, pancreas, prostate, testes, and head and neck area (Fischer et al., 1993.
Strohl et al.. 1997).
Although useful in treating many forms of cancer, daunorubicin and doxorubicin are yet ineffective in the treatment of colorectal cancers or non-small-cell lung cancer
(Strohl et al.. 1997). The cumulative cardiotoxicity of daunorubicin and doxorubicin, in
14 addition to the certain limitations in their use. portends a need for newer anthracyclines with possibly lower toxicity and greater efficacy. In an effort to generate novel anthracyclines, it is essential to understand the enzymatic reactions of the biosynthetic pathway leading to the production of these drugs within the microorganism. A greater understanding of the functions of the gene products and control points would enable researchers to genetically engineer the pathway to make analogs with the desired pharmacological features. Additionally, knowledge of the productive interactions among gene products of the pathway might spur future research wherein rational interspecies cloning experiments, i.e., gene products from other Streptomyces spp. substituting for the daunorubicin and doxorubicin pathway gene products, will enable researchers to produce
"hybrid drugs’" with enhanced potencies (Neimi et al., 1994; Strohl et al., 1997).
B. Biosynthetic pathway for daunonibicin/doxorubicin production.
'^C-Acetate and propionate labeling experiments (Paulick et al., 1976; Casey et al., 1978; Shaw et al., 1979; ICitamura et al., 1981) have revealed that daunorubicin and doxorubicin are produced by the condensation of a single propionyl starter unit with nine
C2 units derived from malonyl-SCoA. The malonyl-SCoA is obtained from acetates via the activity of the acetyl-SCoA carboxylase (Dekleva and Strohl, 1988a; Dekleva and
Strohl, 1988b). The source of propionyl-SCoA is unknown in Streptomyces sp. strain C5
(Strohl et al., 1997). In some other organisms that utilize propionyl-SCoA as a precursor
15 in antibiotic biosynthesis, it is derived via the catabolism of valine through the action of
L-valine dehydrogenase (Strohl et al., 1997). The source of the methoxy group at the C4 position in the final molecule is methionine and this was determined using méthylation inhibitors (Blumauerova et al.. 1979). The amino sugar daunosamine. essential for antitumor activity (Strohl et al.. 1997), is derived from glucose (Pavanosenkova and
Karpov. 1976).
The entire biosynthetic pathway for daunorubicin and doxorubicin can be divided into four steps: (a) condensation of acetates and propionate to produce an initial hypothetical polyketide which is modified to form the first characterized intermediate aklanonic acid; (b) early enzymatic reactions which convert aklanonic acid to s- rhodomycinone; (c) daunosamine sugar biosynthesis and subsequent glycosylation of s- rhodomycinone; (d) conversion of the glycosylated product to daunorubicin and doxorubicin.
B.l. Condensation of acetates and propionate to produce aklanonic acid.
The initial biosynthetic steps for daunorubicin and doxorubicin (Strohl et al..
1989; Strohl and Connors, 1992; Ye et al., 1994; Grimm et al.. 1994; Strohl et al., 1995;
Rajgarhia and Strohl, 1997; Strohl et al.. 1997; Hutchinson, 1997) result in a hypothetical
C21-polyketide intermediate (Fig. 1.3). This hypothetical intermediate is then cyclized and modified to aklanonic acid, the first characterized chromophore in the daunorubicin biosynthetic pathway (Eckardt et al., 1985a). The series of reactions
16 Fig. 1.3. Hypothetical steps involved in aklanonic acid biosynthesis in Streptomyces sp. strain C5 and Streptomyces peucetius. Polyketide synthase proteins catalyzing each of the steps are also depicted in the figure.
17 Propionyl-SCoA + 9 X maionvl-SCoA
DpsA, OpsB, DpsG
9C02
SE
o o o o
DpsE, DpsF, OauG
4 H2O
COOH
OH O OH O
Aklanonic acid
Fig. 1.3. Hypothetical steps involved in aklanonic acid biosynthesis in Streptomyces sp. strain C5 and Streptomyces peucetius (ATCC 29050).
18 leading to the formation of aklanonic acid from propionyl-SCoA and malonyl-SCoA is
catalyzed by proteins known as the polyketide synthases (PKSs). encoded by the
polyketide synthase genes (Chapter 2; Rajgarhia and Strohl, 1997). The basic pathway of
fatty acid and polyketide biosynthesis is shown in Figure 1.4. Both of these processes
begin with similar steps, i.e., condensation of small acyl-SCoA to give p-keto
intermediates and then differ as P-keto intermediates are reduced and subsequently
modified to give the final products. The PKSs have been classified into different types,
analogous to a system of classification used for differentiating fatty acid synthases.
Type I PKSs are large multi-functional proteins that have several domains with
multiple catalytic sites. The acyl carrier protein functionality, integral to the biosynthesis
of polyketides. is housed as a domain within the large protein. These proteins either work
processively. e.g.. Type I PKS of Saccharopolyspora erythraea which makes
erythromycin, or work iteratively, e.g., fungal Type I PKSs that make aromatic products
such as 6-methyl salicylic acid (Strohl e( al.. 1997).
Type 11 PKSs are comprised of multiple individual proteins, each of which
contributes single or multiple functions that catalyze the synthesis of the polyketide via
an iterative mechanism; e.g., the PKSs of Streptomyces sp. strain C5 and Streptomyces peucetius demonstrate a Type II organization. The acyl carrier protein (AGP) is a
separate, small protein which is encoded by a gene clustered closely with the rest of the
genes encoding the PKS.
A Type III PKS has been found in specialized plants, e.g., the chalcone synthase
19 Fig. 1.4. A comparative pathway of fatty acid synthesis and polyketide biosynthesis.
The reactions undergone by acyl ACP in fatty acid biosynthesis involve reduction, dehydration, enoylreduction followed by an additional acyltransferase reaction to give the fatty acid product (A). In polyketide synthesis, each of the reactions can terminate to give either a polyketide (B), or a reduced polyketide (C). (Figure adapted from Hopwood, 1997) SC oA Acyltransferase ,SACP HOOC HOOC' ir' CO] O Acy transferase SACP SCoA SKS ‘r v o o
R cdnctioa (Ketoreductase)
o o o
O H O POLYKETIDE (C)
O«h>dr»(iott
SACP O O OH
REDUCED POLYKETIDE (B) Rcduciioo (enoyl reductase
SACP
FATTY ACID (A I
Fig. 1.4. A comparative pathway of fatty acid synthesis and polyketide biosynthesis.
21 of parsley {Petroselinum hortense). which catalyze the linking of acyl-SCoA subunits by repetitive decarboxylative condensation (Tropf et al.. 1995). This PKS differs from other known bacterial and fungal PKSs or, for that matter, fatty acid synthases (FASs), and hence have been grouped into a separate class by itself (Hopwood, 1997).
B. 1. a. Architecture of the Streptomyces sp. strain C5 polyketide synthase gene cluster.
The architecture of the anthracycline PKS gene clusters (genes that encode the
PKSs) from Streptomyces sp. strain C5 (Ye et al.. 1994), S. peucetius (ATCC 29050)
(Grimm et ai, 1994), S. galilaeus (ATCC 31133) (Tsukamato et al.. 1994; Hutchinson and Fujii, 1995) and S. nogalater (ATCC 27451) (Ylihonko et ai, 1996) are compared in
Figure 1.5. The actinorhodin PKS gene cluster of S’, coelicolor A3(2) (Femandez-
Moreno et al.. 1992) is included for comparative purposes. The structures of the
Streptomyces sp. strain C5 (Ye et al.. 1994) and S. peucetius (ATCC 29050) (Grimm et al.. 1994) daunorubicin Type II PKS gene regions share 93.1% sequence identity with each other, but differ significantly in organization from that of other known Type II PKS gene clusters (Fig. 1.5).
Reading left from a divergently transcribed promoter region are dpsE and dpsF. encoding polyketide reductase and polyketide cyclase, respectively (Fig. 1.5). Reading right from this sequence are five genes in an apparent operon (Ye et al.. 1994; Grimm et al., 1994): (i) dauG (encoding 12-deoxyaklanonic acid oxygenase); (ii) dpsA Fig. 1.5. Comparison of the polyketide synthase genes from the daunorubicin- producing strain, Streptomyces sp. strain C5 (DAU; the daunorubicin PKS gene cluster from S. peucetius ATCC 29050 is identical), the aclarubicin producer, S. galilaeus ATCC 31133 (ACL), the nogalamycin producer, S. nogalater ATCC 27451
(SNO), the tetracenomycin-producer, S. glaucescens GAO (TCM), and the actinorhodin producer, S. coelicolor (ACT). KAS„ genes (example,dpsA), right slanted arrow; KAS, genes (example,dpsB), left slanted arrow; ACP genes (examples, dpsG, actI-3), short solid arrows; cyclases (example, dpsF), cross pattern arrow; polyketide reductases (example, dpsE), horizontal lined arrow; oxygenases/monooxidases, small short lined arrow. The daunorubicin biosynthesis genesdpsC and dpsD are involved in starter unit selection as described in the text. All other unfilled arrows are genes hypothesized not to be involved with polyketide assembly.
23 * d a u dpsG dpsF dpsE G dpsA dpsBdpsC dpsD
ACL
E F X B C D
^ a ^ lz X l(= ] C||E5) #=)(=] SNO E D C B A 1 2 ACP X Y
ACT
III 1-1 1-2 1-3 VII IV
Fig. 1.5. Comparison of the polyketide synthase genes from several Streptomyces spp. (ketoacylsynthase; KASJ, (iii) dpsB (the ketoacylsynthase homolog that makes up the putative heterodimeric partner; BCASJ. This is described elsewhere as chain length factor
(CLF; McDaniel et al., 1993), but there is evidence now that this protein alone does not confer chain length (Shen at al.. 1996; Rajgarhia and Strohl, 1997; Hopwood, 1997); (iv) dpsC (a homo log of E. coli KASIII that lacks the active site cysteine residue). This gene product, thought to play perhaps a structural role, contains a putative coenzyme A- binding site (Ye et al., 1994); and (iv) dpsD (an acyltransferase speculated, but not proven, to function as propionyl-SCoA;ACP-SH acyltransferase; Ye et al., 1994; Grimm et al.. 1994) (Fig. 1.5). In both S. peucetius and Streptomyces sp. strain C5. the gene encoding the ACP, dpsG, is located about 6.8 kbp upstream of the genes encoding the daunorubicin KAS„ and KASg (Fig. 1.5; Ye et al., 1994; Grimm et al., 1994; Strohl et al..
1995). While comparing the gene clusters from several Streptomyces spp. the most obvious deviations of the daunorubicin and doxorubicin PKS gene clusters from other
Type 11 PKS gene clusters are the presence of the two genes. dpsC and dpsD, directly downstream of the KAS^, and KASg, not found in other Type II PKS gene clusters, and the obvious absence of a gene encoding ACP directly downstream of the genes encoding
KASg (Ye et al., 1994; Grimm et al., 1994; Strohl et al., 1995).
B. 1. b. Minimal PKS / Role of the dpsC and dpsD gene products.
A unique feature of the Type II PKSs from Streptomyces sp. strain C5 and S. peucetius is that they use propionyl-SCoA as the starter molecule, to condense with nine
25 C2 units from malonyl-SCoA, in the biosynthesis of the final product (Ye et al., 1994;
Strohl et al., 1997; Hutchinson, 1997). Most other known Type II PKSs, except the PKS
of S. galilaeus which utilizes propionate and makes aclacinomycin via an aklanonic acid
intermediate (Tsukamato et al., 1994; Hutchinson and Fujii, 1995), use acetate as a starter
unit in priming the biosynthesis of their respective polyketides (Fig. 1.5; Hutchinson et al., 1995; Strohl et al.. 1997; Hutchinson, 1997). Thus, the properties, functions and organization of the polyketide gene region of Streptomyces sp. strain C5 and S. peucetius that confer the propionyl starter moiety might be unique when compared to the organization of the PKS gene region from other Streptomyces spp. Interestingly, the PKS gene cluster encoding aclacinomycin A, an anthracycline also primed with a propionyl moiety, is more analogous to the PKS gene clusters for actinorhodin biosynthesis and other aromatic polyketides. In this case, however, the genes encoding KAS„, KASg, and
ACP are clustered together Just as they are in actinorhodin biosynthesis (Fig. 1.5).
We have recently shown that a minimal PKS gene cluster containing only dpsA
(KASJ. dpsB (KASJ. dauG ( 12-deoxyaklanonic acid oxygenase). dpsF (cyclase), dpsG
(ACP), dpsE (PKR). and daul (transcriptional activator), conferred on both S. lividans
TK24 and S. coelicolor CH999 the ability to produce ample quantities of aklanonic acid
(Chapter 2; Rajgarhia and Strohl, 1997), the first stable chromophore of the daunorubicin and doxorubicin biosynthesis pathways (Fig. 1.3). Additionally, we also observed in the mass spectroscopy data substantial levels of fragments that would have been derived from a methyl-side-chain derivative of aklanonic acid (Chapter 3; Rajgarhia and Strohl, 1997),
26 indicating that both acetyl and propionyl moieties were being incorporated as starter units into products by this recombinant strain.
Moreover, we have also shown that disruption of dpsD or dpsC and dpsD still resulted in the biosynthesis of daunorubicin and 13-dihydrodaunorubicin in mutants of
Streptomyces sp. strain C5 (Chapters 2. 3; Rajgarhia and Strohl. 1997). In addition to these "normal", expected products, we also found that the dpsCD double mutant strains
(SC5-VR5) accumulated substantial quantities of feudomycinone C and feudomycin D
(Chapter 3; Rajgarhia VB. Priestley ND. Strohl WR. unpublished data), compounds that are initiated with an acetyl starter unit (Oki et al., 1981). As far as we can tell from detailed chromatographic analysis, neither feudomycinone C. feudomycin D nor its aglycone are produced by the parental strain. Streptomyces sp. strain C5 (Strohl et al..
1997). A detailed analysis of the experimental data and their significance is presented in this dissertation (Chapters 2. 3). The factors affecting the choice of starter unit (acetate vs propionate) as well as possible involvement of an additional cyclase for the first ring cyclization is also addressed in this dissertation (Chapter 2).
B. 2. Early pathway gene products that convert aklanonic acid to e-rhodomyclnone.
The pathway for the conversion of aklanonic acid to its derivative anthracyclinones was first proposed by Eckardt and co-workers (Eckardt et al., 1988).
27 The pathway was proposed by analyzing accumulated early intermediates, bioconverting
those intermediates to anthracyclinones and analyzing mutants that produced those
intermediates (Wagner et al„ 1981; Wagner et al„ 1984; Eckardt et al., 1985a; Eckardt et
al., 1985b; Schumann et al.. 1986; Eckardt et al., 1988; Eckardt and Wagner. 1988).
The pathway for the conversion of aklanonic acid to e-rhodomycinone in Streptomyces
sp. strain C5 was elucidated by using single and double mutants blocked in daunorubicin
biosynthesis and by in vitro analysis using both parental and blocked mutant strains
(Bartel et al.. 1990; Dickens et al.. 1995; Madduri and Hutchinson. 1995a; Dickens et al..
1996). Using such methods of analysis, aklanonic acid was shown to be converted first to
its methyl ester, then to aklaviketone. followed by aklavinone and finally to s-
rhodomycinone (Fig. 1.6; Connors et al.. 1990; Dickens et al.. 1995; Dickens et al..
1996).
The S-adenosylmethionine dependent conversion of aklanonic acid to its methyl
ester is catalyzed by the product of the dauC gene (Dickens et al.. 1995; Madduii and
Hutchinson. 1995). The subsequent cyclization of aklanonic acid methyl ester to
aklaviketone via an aldol condensation (Kendrew et al.. 1996) is catalyzed by the product of the dauD gene (Dickens et al., 1995; Madduri and Hutchinson, 1995). The ensuing conversion of aklaviketone to aklavinone is via a reduction reaction catalyzed by the product of the dauE gene (Dickens et al., 1996).
Aklavinone is a major precursor common in many anthracycline biosynthetic pathways including those of aclarubicin, daunorubicin, and rhodomycin (Strohl et al..
28 Fig. 1.6. Pathway for the conversion of aklanonic acid to E-rhodomycinone. Proteins catalyzing these steps are: DauC, aklanonic acid methyltransferase; DauD, aklanonic acid methylester cyclase; DauE, aklaviketone reductase; and DauF, aklavinone 11-hydroxylase (Dickens et al., 1995; Madduri and Hutchinson, 1995).
29 COOH
Aklanonic acid
OH 0 OH OH O
DauC
C OOCHj
Aklanonic acid methy ester
OH O OH OH O
DauD
Aklaviketone
OH O OH O
DauE
Aklavinone
OH O OH OH
DauF
O OH COOCH
e-rhodomycinone
OH O OH OH
Fig. 1.6. Pathway for conversion of aklanonic acid to e-rhodomycinone in Streptomyces sp. strain C5 and Streptomyces peucetius (ATCC 29050).
30 1989; Hutchinson and Fujii, 1995; Strohl et al., 1997). The S. peucetius dnrF gene
encodes a flavoprotein aklavinone 11 - hydroxylase (Hong et al., 1994; Flippini et ai,
1995; Swang et al., 1995; Kim et al., 1996) that catalyzes the conversion of aklavinone to
E-rhodomycinone (Fig. 1.6). which is a major product in many daunorubicin
fermentations (Kern et al., 1977; McGuire et al., 1980; Arcamone, 1981a; Arcamone.
1981b; Strohl et al., 1989). This intermediate has been shown to be converted to
daimorubicin and related glycosides by both Streptomyces sp. strain C5 (McGuire et ai.
1980) and S. coeruleorubidus ME 130-A4 (Yoshimoto et al., 1980a; Yoshimoto et al.,
1980b), demonstrating that E-rhodomycinone clearly is an intermediate in daunorubicin
biosynthesis.
B. 3. TDP-daunosamine biosynthesis and glycosylation of E-rhodomycinone.
The precise pathway for the TDP-daunosamine biosynthesis from TDP-glucose is
still unclear. A hypothetical pathway for the bio-synthesis is shown in Figure 1.7
(Hutchinson, 1997). The conversion of glucose-1 phosphate to TDP-glucose is believed
to be catalyzed by S. peucetius DnmL protein (encoded by the dnmL gene) that is similar
in structure to known glucose-l-phosphate:TTP thymidylyltransferases (Gallo et al..
1996). Streptomyces sp. strain C5 also has a dnml homolog (Strohl et al., 1997).
TDP-glucose is a suitable substrate for TDP-glucose 4,6-dehydratase (Thompson et al., 1992) which has been purified to homogeniety from Streptomyces sp. strain C5 and
S. peucetius (Thompson et al., 1992). TDP-4-keto-6-deoxyglucose is a substrate for 3,5-
31 Fig. 1.7. Hypothetical pathway for the biosynthesis of TDP-daunosamine. Genes encoding products, putatively involved in converting glucose-I-phosphate to TDP- daunosamine are indicated for each step. dnmS encodes a glycosyltrasferase postulated to catalyze addition of TDP-daunosamine to s-rhodomycinone (Adapted from Hutchinson, 1997).
32 CH20H CH2OH CHj dnmL j — 0 4,6 -dehydratase J — 0 dnmU OH - 0 = < OH OH""' f OPO3 OH OTDP -f OTDP OTDP OH OH OH OH OH TDP-4-keto-6-deoxy TDP-4-keto-rhamnose D-glucose-1 -OPO3 TDP-D-glucose D-glucose
dnmQ, dnml, dnmZ e-rhodomycinone w
dnmV CO dnmJ E CH3 > ^ 0 = < CH3 5 O H > p ~" ^ O T D P OTDP NHz NHz
TDP-daunosamine
rhodomycinD
Fig. 1.7. Hypothetical pathway for biosynthesis of TDP-daunosamine epimerase (Piepersberg, 1997; Strohl et al., 1997). The dnmU of Streptomyces sp. strain C5 and S. peucetius are closely related to the S. griseus orf4, which apparently encodes a nucleotide sugar 3,5-epimerase. Thus the DnmU protein, encoded by dnmU gene, has been implicated to have a role in this reaction producing TDP-L-4-keto- rhamnose (Fig. 1.7). A hypothetical scheme has been proposed (Otten et at., 1995b;
Hutchinson. 1997; Strohl et al.. 1997) to explain the conversion of TDP-L-4-keto- rhamnose to TDP - daunosamine. The deduced products of dnmJ. dnmU and dnmVhtivt been compared with the products of homologous genes from S. griseus (Krugel et al.,
1993, Otten et al., 1995b; Hutchinson 1997; Strohl et al., 1997) to assign reasonable functions in daunosamine sugar biosynthesis in S. peucetius and Streptomyces sp. strain
C5. These assignments are as yet unproven since none of the substrates and products have been verified (Hutchinson, 1997). Additionally, the exact function of the products of dnmO, dnmT and dnmZ genes putatively involved in C-2 deoxygenation remain unknown (Fig. 1.7; Strohl etal., 1997; Hutchinson, 1997).
Yoshimoto et al. (1980a. 1980b, 1986) and Connors and Strohl (1993) provided strong circumstantial evidence that s-rhodomycinone is the primary aglycone substrate for TDP-daunosamine:anthracycline glycosyltransferase in Streptomyces sp. strain C5 and S. peucetius). This glycosylation step confers biological activity to the s- rhodomycinone (Strohl et al., 1989; Strohl et al., 1997). Two open reading frames dauH
(Dickens et al., 1996) and dnmS (Otten et al., 1995b) from the daunorubicin biosynthetic gene cluster encode deduced proteins with similarity to known glycosyltransferase
34 enzymes (Strohl et ai, 1997). Otten et a i (1995b) complemented S. peucetius mutants that accumulated e-rhodomycinone with dnmS in high copy number and restored daunorubicin production, implicating the DnmS protein with possible glycosylation of e- rhodomyinone. In the absence of substrate specificity data, it is conjectured that dnmS encodes the glycosyltransferase that produces the first glycone intermediate, rhodomycin
D. while dauH is hypothesized to encode a glycosyltransferase that yields higher glycosides of daunorubicin such as baumycins or 4’-daunosamine daunorubicin (Otten et ai, 1995b; Scotti and Hutchinson, 1996; Dickens et ai, 1996; Strohl et ai, 1997).
B. 4. Conversion of rhodomycin D to doxorubicin.
The enzymatic conversion of rhodomycin D to doxorubicin proceeds through the following set of reactions (not particularly in any order): (i) desmethylation of the 16- methoxy residue by a presumed methylesterase; (ii) removal of the resultant free carboxyl moiety at C-10; (iii) hydroxylation at C-13; (iv) oxidation of the newly formed 13- hydroxy moiety to a keto group; (v) 0-methylation of the 4-hydroxy group; and (iv) hydroxylation at C-14 (Strohl et a i, 1997). Figure 1.8 shows the proposed pathway for conversion of rhodomycin D to daunorubicin, doxorubicin, and baumycin Al/,\2. The reactions and genes governing the biosynthesis of daunorubicin and doxorubicin from 8- rhodomycinone have been widely postulated (Oki, 1982; Strohl et ai, 1989; Wagner et ai, 1991; Strohl et ai, 1995; Hutchinson, 1995; Hutchinson, 1997; Dickens et al., 1997), but have not been proven in vitro until very recently (Dickens et al, 1997). Recent
35 Fig. 1.8. Hypothetical pathway for the biosynthesis of doxorubicin from rhodomycin
D. Intermediates are rhodomycin D (RHOD), 10-carboxy-13-dcoxycarminomycin
(CDOC), 13-deoxy daunorubicin (DOD), 13-dihydrodaunorubicin (DHD), daunorubicin (DAU) and doxorubicin (DOX) (Strohl et al., 1989; Walczak et al., unpublished data).
36 o OH COOCHi
RHOD
O OH COOH
OH O OH O OH O OH O CDOC DHD
OH
O OH O
OCH; O OH O DOC DAU
O OH Ü
OCH; O OH O OCH: O OH O DOX DOD
Fig. 1.8. Hypothetical pathway for biosynthesis of doxorubicin from rhodomycin D.
37 evidence, however, indicates that the products of three genes, dauP, dauK and doxA. from the daunorubicin biosynthetic gene cluster are sufficient and necessary to convert rhodomycin D to doxorubicin in a heterologous strain (Dickens ef al„ 1997).
B. 5. Predicted gene required for daunorubicin biosynthesis.
On the basis of the knowledge of the functions of the enzymes and gene products from Streptomyces sp. strain C5, we can now estimate the minimal gene products that would be required to synthesize doxorubicin from propionyl-SCoA, malonyl-SCoA and
TDP-daunosamine in a heterologous host. Table 1.1 is a list of the estimated genes required. Of the 37 genes found in the daunorubicin and doxorubicin biosynthetic gene cluster (Fig. 1.9), we have shown that six gene products from the cluster are required to convert propionyl-SCoA. and malonyl-SCoA to aklanonic acid (Chapter 2; Rajgarhia and
Strohl. 1997; Strohl et al., 1997: Hutchinson. 1997). Dickens et al. (1996) and Madduri and Hutchinson (1995) have shown that four gene products can convert the aklanonic acid to E-rhodomycinone. Moreover, nine additional gene products are required for dTDP-daunosamine biosynthesis as well as the glycosyltransferase reaction (Otten.
1995b: Scotti and Hutchinson. 1996; Gallo et al.. 1996; Hutchinson, 1997; Otten et al..
1997). Finally, the rhodomycin D thus obtained can be converted to doxorubicin in presence of three gene products in a heterologous strain such as S. lividans TK24
(Dickens and Strohl, 1996; Dickens et al., 1997).
38 Fig. 1.9. Daunorubicin and doxorubicin biosynthetic gene cluster ofStreptomyces sp. strain C5 and Streptomyces peucetius ATCC 29050. dps denotes genes encoding proteins involved in the poiyketide synthase reactions; dnm denotes genes encoding proteins involved in the daunosamine sugar biosynthesis as well as the glycosylation reactions; drr denotes genes encoding proteins conferring resistance. The bar represents genes sequenced by Krugelet al., 1993 from S. griseus.
39 s. griseus orfs 1-6
2 1 3 4 5
* nmL t N 0 F drr A drrB drrD X Y dnmZ |dnmV d n m M dnraU
dnmJ I doxA V U ^ dnmT H E dpsFdpsEG dpsA dpsG dauZ
dpsB dpsC dpsD C D K P dnmQ dnmS drrC
2 .0 kbp
Fig. 1.9. Daunorubicin and doxorubicin biosynthetic gene cluster o{Streptomyces sp. strain C5 and Streptomyces peucetius (ATCC 29050) Table 1.1. Predicted genes required for doxorubicin biosynthesis (Strohl et al.,
1997).
41 Substrates Genes Products O RFs
Genes eneoding biosynthetic reactions: Propionyl-SCoA + 9 malonyl-SCoA dpsABEFG, dauG Aklanonic acid 6 Aklanonic acid dauCDEF e-rhodomycinone 4 D-glucose-I-PO 4 dnmJLMQWVZ dTDP-daunosamine 8 dTDP-daunosamine + e-rhodomycinone dnmS rhodomycin D 1 Rhodomycin D daiiKP.doxA doxorubicin 3
Non-structural genes:
Resistance genes drr A BCD - 4
Regulatory genes dnrNOI - 3
Others:
Genes required to ensure propionyl starter unit dpsCD - 2 Poiyketide synthesis (cyclase) dnrY, dauZ - 2 Baumycin biosynthetic proteins? dauH daunorubicin? 1
Unknown (nnction dauU - 1
13-dihydro derivatives dauV - 1
Genes involved in last step of daunorubicin/doxorubicin dnrX - 1 biosynthesis
Total: 37
Table 1.1. Predicted genes required for daunorubicin and doxorubicin biosynthesis. Of the products of the other 15 genes within the cluster, we have shown that dpsC and possibly dpsD help to dictate starter unit in poiyketide biosynthesis; the drrABCD gene products have been shown to be involved in resistance to doxorubicin (Guilfoile and
Hutchinson, 1991; Lomovskaya et al., 1996; Hutchinson, 1997); and dnrNOI have been shown to be regulatory genes (Stutzman-Engwall et al., 1992; Otten et al., 1995a; Tang et al., 1996; Furuya and Hutchinson, 1996). The dauY, dauZ gene have recently been demonstrated to encode a protein that might be involved in cyclization of the initial poiyketide (Lomovskaya et al., 1998). However, recent gene disruptions of dauZ, however, have had no affect on poiyketide biosynthesis (Hutchinson, personal communication). The dnmU and dnmV genes have been shown to encode a putative thymidine diphospho-4-keto-6-deoxygenase-3-(5)-epimerase and thymidone diphospho-
4-ketodeoxyhexulose reductase, respectively, while the dnmZ gene encodes an unknown protein which is also required for daunosamine biosynthesis (Otten et al., 1997). The following open reading frames dauH (probable glycosyltransferase; Strohl et al., 1997;
Scotti and Hutchinson, 1996), daiiX, and dauU encode proteins with no currently known function.
C. Hybrid Poiyketide synthases.
The idea of generating recombinants with poiyketide synthase (PKS) genes from several Streptomyces spp. combined to produce novel polyketides seemed a possibility
43 after Hopwood et al. (1985), developed a new isochromane quinone antibiotic by introducing the PKS genes from S. coelicolor, encoding the actinorhodin biosynthetic pathway proteins, into Streptomyces violaceoruber. The new metabolite produced was granatirhodin, a compound which differs from granaticin, normally produced by S. violaceoruber, at the stereochemistry of a proton at C-3 (Hopwood et al., 1985).
In the second significant “hybrid antibiotics" experiment, Epp et al. (1989) produced a novel molecule 4"-isovaleryl-spiramycin from Streptomyces ambofaciens by transforming the strain with carE, which encodes the isovaleryl-CoA transferase from
Streptomyces thermotolerans (Epp et al., 1989).
Bartel et al., (1990a) transformed ketoreductase mutants of Streptomyces galilaeus with actlll (ketoreductase) from S. coelicolor PKS gene cluster and restored functionality. Bartel et al., (1990a) went further and produced a novel poiyketide, aloesaponarin II (discussed below) in S. galilaeus by transforming wild-type strains with the acti and actlll loci (encoding ketoacylsynthase a and p, acyl carrier protein and reductase from the S. coelicolor PKS cluster, respectively). Wild-type 5". galilaeus did not produce aloesaponarin II (Bartel et al., 1990a).
These early successes were followed by other experiments in which several mutations of the S. coelicolor PKS genes were complemented with genes from other
Streptomyces spp. that were expected to encode products that performed similar functions. In one such experiment, an 5. coelicolor strain with a mutation in actlll
(ketoreductase) gene was complemented by the orf 5 gene from S. violaceoruber, that
44 encodes a ketoreductase (Sherman et al., 1992). The results of these early "mix and match'’ experiments set the stage for a systematic approach wherein novel polyketides generated by recombinants containing hybrid PKS gene clusters could began revealing rules for the interaction between individual functions from different type II PKSs
(Hopwood, 1997).
C.l. Design rules for combining PKSs functions to make hybrid polyketides.
The comparison of the architecture of the poiyketide synthase gene clusters from several Streptomyces spp. revealed that most clusters consisted of structural genes that encoded for a ketoacylsynthase followed immediately by another gene downstream of the ketoacylsynthase that resembled strongly the ketoacylsynthase but lacked the putative catalytic site for condensation of small acyl-SCoAs. A gene encoding the ACP was found directly downstream of these two genes in most cases. In addition, most clusters also contained genes encoding for one or more aromatase/cyclase.
The proteins encoded by the PKS are believed to determine the carbon chain length, and also the specific positions on the poiyketide chain where reduction and cyclization or aromatization may occur. These ftmctions are specific for PKSs and differ considerably among various Streptomyces spp. Hence, by combining the genes encoding
PKSs from different Streptomyces spp. and expressing in suitable hosts, unique molecules are expected to be produced. From a number of such “mixing” experiments, design rules
45 affecting the type of poiyketide structure made by such recombinants have been determined. These rules are described as follows:
C. 1. a. Carbon chain length.
Carbon chain length is determined by the minimal PKS, i. e., the ketoacylsynthase a (KASJ, ketoacylsynthase P, (KASp; previously known as the chain length factor) and the acyl carrier protein (ACP). The ACP is interchangable among several Streptomyces spp.. although with daunorubicin and doxorubicin, as well as with the tetracenomycin biosynthetic PKSs, the specificity for ACP is very high (Strohl et al„ 1997). The ketoacylsynthase P is essential but not sufficient for determining the chain length; however, some heterologous combinations of ketoacylsynthase a and ketoacylsynthase p are non-functional; e.g., the major products produced by the KAS„ and KASp from S. roseofulviis were an octaketide and a nonaketide. while the major product of the KAS„ and the KASp from S. glaucescens was a decaketide. A combination of the KAS^ from S. coelicolor with a KASp from either S. roseofulvus or S. glaucescens failed to produce a product (McDaniel et al., 1993: Hopwood, 1997), indicating probably nonproductive interactions.
Although the KASp gene product was initially suggested to be the chain length determinant (McDaniel et al., 1993), it is currently believed that this gene product might be associated with the KAS„ gene product, to form a heterodimeric ketoacylsynthase similar to the homodimeric E. coll enzymes (Garwin et al., 1995). Recent experiments
46 by Carreras and Khosla (1998) have demonstrated that the KAS^and KASp gene products
indeed associate to form a heterodimeric complex of a^Pi. Phylogenetic analysis has also
implied that the KASp may be a result of KAS„ gene duplication in an ancestral PKS that
diverged (Hopwood, 1997).
C. 1. b. Keto reduction.
Although a specific ketoreductase (KR) is required for specific poiyketide
biosynthesis, the actinorhodin pathway KR works on poiyketide chains ranging from
C l6-24. Most KRs encode products that reduce the C9 carbon from the carboxy terminus of the poiyketide chain as originally hypothesized by bartel et al. (1990a), although actinorhodin ketoreductase {actlll) can also reduce at C7 under specified conditions
(Hopwood et al., 1997).
C. 1. c. First ring cyclization.
Cyclization of the first ring is controlled by the minimal PKS, but regioselectivity can be influenced by the presence of genes encoding the other PKS subunits. For example, the tetracenomycin minimal PKS (KAS„, KASp and ACP) alone make products which are mixture of C7/C12 and C9/C14 cyclized compounds, but in the presence of actinorhodin genes encoding the ketoreductase, aromatase and cyclase, the same tetracenomycin minimal PKS made only C9/C14 cyclized products (Hopwood, 1997).
47 Considerable evidence now indicates that the ketoreductase plays an important role in dictating proper and efficient first ring closure (Hopwood, 1997).
C. 1. d. First ring aromatization.
If the poiyketide chain is unreduced, the process of aromatization is spontaneous and does not need protein catalysis; however, if the poiyketide is reduced, a specific aromatase is required (Hopwood et a/., 1997). Aromatases such as the protein encoded by the actinorhodin act VII gene are didomain proteins that work by extracting two water molecules to effect aromatization of the poiyketide chain. They show specificity for chain lengths; e.g., S. griseus aromatase works only on C20, C l 8 , C16 chains. The S. coelicolor actVII gene product was determined to be the aromatase for the first ring closure in actinorhodin biosynthetic pathway (McDaniel et al., 1994)
Several actVII homologs have been identified from other PKS gene clusters. A combination of the aromatase encoded by tcmN. from the tetracenomycin producer S. glaucescens. with the actl-orfs I, 2. 3 from the actinorhodin producer S. coelicolor. when expressed in S. coelicolor CH999. produced unusual products, indicating that the tcmN- encoded aromatase had some effect on the regiospecificity of first ring closure in unreduced carbon poiyketide chains (McDaniel et al., 1995; Hopwood, 1997). Moreover, this aromatase did not effect first ring closure in reduced chains (Hopwood, 1997).
48 c. 1. e. Second ring cyclization and aromatization.
Actinorhodin biosynthesis requires a second ring cyclase such as the product of the actIV gene. Such cyclases show chain length specificity; e.g., the actU'gene product, responsible for the second ring cyclization leading to a bicyclic precursor of actinorhodin, works on C16 and CIS chains, but does not function on C20 chains. Cyclases for C20 are yet being investigated (Hopwood, 1997). In tetracenomycin biosynthesis, the TcmN protein is responsible for aromatization of the second ring although the first unreduced ring is spontaneously aromatized (McDaniel et al., 1995).
C. 1. f. Choice of start unit.
The choice of starter unit in the majority of aromatic polyketides is acetate. However, anthracyclines such as daunorubicin and doxorubicin biosynthesized by Streptomyces sp. strain C5 and S. peucetius are made using propionate as the starter unit (Ye et al., 1994; Grimm et al., 1994; Strohl et al., 1995; Rajgarhia and Strohl, 1997;
Strohl et al., 1997; Hutchinson, 1997). We have recently shown that in the absence of certain gene products, the PKSs of Streptomyces sp. strain C5 can choose acetate as a starter unit, thus producing imusual compounds (Chapter 3). The choice of the propionyl starter unit seems to be dictated by the poiyketide synthase genes in Streptomyces sp. strain C5 and S. peucetius (ATCC 29050).
In other systems, though, the choice of starter unit is not definitively controlled by the PKS. An example of this phenomenon is the accumulation of decaketides (from ten
49 acetates) when minimal PKS genes from S. rimosus are expressed in S. coelicolor CH999
in the presence of the act ketoreductase (Fu et al., 1994). The PKS of S. rimosus. a producer of oxytetracycline, makes the antibiotic using maJonamate or malonate as a
starter under native conditions (Thomas and Williams, 1983). Hence, it seems that in S. rimosus the PKS can use acetate as the starter unit under unnatural conditions, while
under native conditions seems to be controlled by as yet undiscovered factors (Hopwood,
1997).
The design rules for poiyketide synthesis are obviously incomplete. As more
“mixing" experiments are conducted and more hybrid products are identified, a clear empirical determination of these rules will be possible.
C. 2. Hybrid PKS: actinorhodin pathway and daunorubicin.
The proposed pathway for actinorhodin biosynthesis is shown in Figure 1.10. The genes encoding proteins essential for actinorhodin synthesis are also shown in Figure
1.10. In an early experiment. Bartel et al.. (1990a) showed that actI and actlll encoding the ketoacylsynthase a and ketoacylsynthase p, acyl carrier protein and reductase from S. coelicolor PKS cluster respectively, when introduced into S. galilaeus (ATCC 31133), conferred on the recombinant strain the ability to make substantial amounts of aloesaponarin II. Aloesaponarin II is made utilizing an acetate as starter condensed to seven C2 units derived from malonyl-CoA, as shown in the pathway in Figure 1.10.
Wild-type S. galilaeus (ATCC 31133) makes aclacinomycin via an aklanonic acid
50 Fig. 1.10. Actinorhodin and aloesaponarin II biosynthetic pathways. The catalytic proteins essential for activity are shown along side the arrow. In the absence of the actVI gene product aloesaponarin II is produced by the pathway (Bartel, 1989;
Bartel et al., 1990b).
51 7 X HOOCCHiCOSCoA „cri
COOH CH2COSC0A
A C T I , III
COOH
A C T V II
OH O
COOH
A C T I \
COOH
A C T VI. Va. Vb
OH O OH O
OH O COOH
ALOESAPONARIN II ACTINORHODIN
Fig. 1.10. Actinorhodin and aloesaponarin II biosynthetic pathway.
52 intermediate, using propionate as a starter unit and condensing it with nine C2 units
derived from malonate (Hutchinson and Fujii, 1995). In later experiments, it was
demonstrated that S. galilaeus (ATCC 31133) transformed with act! locus
(ketoacyisynthase a and P along with ACP) alone could make aloesaponarin 11 (Floss and
Strohl. 1991). The S. galilaeus (ATCC 31133) provided the required reductase
functionality for aloesaponarin 11 formation indicating a cooperative interaction between
PKS from either species.
Streptomyces sp. strain C5 also makes aklanonic acid and would be expected to encode the required reductase for combining with S. coelicolor acti locus products to give aloesaponarin 11. Our earlier attempts to show such an interaction had failed (Strohl, personal communication). In this work, however, when actI genes were used to transform Streptomyces sp. strain C5 PKS double mutants (dpsC/dpsD), detectable amounts of aloesaponarin 11 were produced. The significance of the production of aloesaponarin 11 made by generating a possible hybrid PKS in Streptomyces sp. strain C5 is discussed in this dissertation (Chapter 4)
D. In vitro studies determine the programming rules for the PKSs catalyzed polyketide biosynthesis.
Genetic studies using heterologous expressions of the “hybrid polyketide synthase genes" have shown that the results of such experiments, albeit exciting, need to be
53 supplemented with biochemical evidence (Hopwood. 1997). In vitro studies using
recombinant PKSs have been used to supplement genetically derived information
(Gramajo et al., 1991; Shen et al., 1992; Crosby et al., 1995; Crump et al.. 1996; Crump
et al., 1997). In one such experiment, Shen and Hutchinson (1993b) recently obtained the
synthesis of tetracenomycin F2, a tetracenomycin C precursor, from acetyl-SCoA and
malonyl-SCoA in cell free extracts from a recombinant S. glaucescens strain in which the
tetracenomycin PKS genes encoding the KAS„, KASp, ACP, and aromatase/cyclase were
over-expressed. Moreover, when ACP was biochemically depleted from the system, no
synthesis of the precursor was observed (Shen and Hutchinson, 1993).
More recently, Shen and Hutchinson (1996) have shown that when recombinant
TcmN (aromatase/cyclase essential for tetracenomycin F2 biosynthesis) was added to an
in vitro system which lacked it, it caused tetracenomycin production instead of the usual
SEK15. Carreras et al. (1996) have also used similar cell free systems from S. coelicolor
to show the synthesis of SEK4 and SEK4b from acetyl- and maionyl-CoA with minimal
PKS, and carboxylated derivative of aloesaponarin 11, 3,8-dihydroxymethylanthraquinone
carboxylic acid (DMAC) when aromatase, ketoreductase, and cyclase were present along
with the minimal PKS.
We have similarly shown that under in vitro conditions, the Streptomyces sp. strain C5 minimal polyketide synthase genes, over-expressed in S. lividans TK24, can together produce aklanonic acid using malonyl-SCoA and propionyl-SCoA (Chapter 4;
Rajgarhia and Strohl unpublished data). For the purpose of detecting the aklanonic acid
54 produced, we used [ 3 ,3 ,3 -d3]propionyl-SCoA which was synthetically prepared by methods we developed (Chapter 4; Rajgarhia et al.. 1995).
E. Goals of this work.
This dissertation describes the results of my studies on the polyketide synthase genes of the daunorubicin producing strain, Streptomyces sp. strain C5. The primary goal of this research was to determine the minimal polyketide synthase gene products that are necessary to make polyketide intermediates of the daunorubicin pathway.
Characterization of the role, if any, of the unusual polyketide synthase genes dpsC and dpsD of Streptomyces sp. strain C5 was an additional goal of this research.
Realization of these goals would enable others in this area of research to determine how the choice of starter unit in polyketide biosynthesis is determined by gene products that are part of the polyketide synthase gene cluster, and to design experiments whereby additional information can be generated regarding starter unit specificity.
Moreover, this work will enable others in similar areas of research to identify possible starter unit specifying gene products in their pathway. The information generated might be used to further the synthesis of analogues of existing drugs which might have better efficacy and lower toxicity.
This dissertation begins with a description of the experiments performed to determine the required polyketide synthase gene products from Streptomyces sp. strain
C5 that could make detectable quantities of polyketide intermediates of the daunorubicin
55 biosynthetic pathway when expressed in a suitable host. Analysis of the data generated during that work enabled me to identify and chemically characterize unique anthracycline molecules, and that is discussed in the latter part of this dissertation. With the information that was generated in the earlier part of this work, I was able to identify a system where we could show interactions among polyketide gene products, thus far not seen, from Streptomyces sp. strain C5 and other Streptomyces spp. The description of the system and products generated from such interactions is reported in the final part of the dissertation. The generation of the data as described in this dissertation has enabled me to suggest a possible role for certain required, at the same time nonessential, polyketide synthase gene products, dpsC and dpsD. in the assembly of polyketides from
Streptomyces sp. strain C5.
56 CHAPTER 2
MINIMAL Streptomyces sp. strain C5 DAUNORUBICIN POLYKETIDE
BIOSYNTHESIS GENES REQUIRED FOR AKLANONIC ACID BIOSYNTHESIS
INTRODUCTION:
Daunorubicin (daunomycin) and doxorubicin (adriamycin). clinically important anthracycline chemotherapeutic agents produced by Streptomyces sp. strain C5 and
Streptomyces peucetius (ATCC 29050). are synthesized by condensation of nine extender units derived from malonyl-coenzyme A (CoA) onto a propionyl moiety to make a hypothetical poly-^-keto intermediate (Hutchinson, 1995; Hutchinson, 1997; Strohl et al.,
1997). This hypothetical intermediate is reduced at C-9 (Strohl and Connors, 1992). cyclized, and aromatized to form aklanonic acid, which is the first characterized chromophore of the pathway (Eckardt, 1985; Hutchinson, 1997; Strohl et al., 1997).
Aklanonic acid is converted in four enzymatic steps to E-rhodomycinone (Bartel et al.,
1990; Connors et al., 1990; Strohl et al., 1997), then glycosylated (Hutchinson, 1997) and subsequently modified by a series of reactions to form daunorubicin and doxorubicin
(Dickens and Strohl, 1996; Dickens et al., 1997). The initial condensation and modifying
57 reactions leading to aklanonic acid production, is catalyzed by the Streptomyces sp. strain
C5 and S. peucetius Type II polyketide synthases (PKSs) encoded by the polyketide
synthase genes (PKS).
The organizaton of the Type II PKS genes of Streptomyces sp. strain C5 (Ye et
al„ 1994) and S. peucetius (Grimm et al., 1994) is compared with that of several other
known Type II PKS gene clusters in Figure 2.1 (Hutchinson and Fujii, 1995; Hopwood,
1997). This comparison reveals that the arrangement of the PKS genes in Streptomyces
sp. strain C5 and S. peucetius is significantly different fi-om the PKS gene arrangement in
other related species. The main difference is the location of the gene encoding the ACP,
which in Streptomyces sp, strain C5 and S. peucetius is located 6 .8 kbp upstream of the
ketoacyisynthase genes. In most other PKS gene clusters, the gene encoding ACP is
located immediately downstream of the genes encoding the ketoacyisynthase (KAS„ and
KASp). Streptomyces sp. strain C5 and S. peucetius, instead, have two unusual open
reading frames, dpsC and dpsD, immediately after the gene encoding the
ketoacyisynthase. The possible significance of this arrangement and the roles of the gene
products of dpsC and dpsD are discussed in this and subsequent chapters in this
dissertation.
The PKS gene cluster of Streptomyces sp. strain C5 (Ye et a!.. 1994) is virtually
identical with that of Streptomyces peucetius (ATCC 29050) (Grimm et al., 1994) (DNA
identity over this region is 93%; Strohl et al., 1997), and hence the PKS gene names proposed by Grimm et al. (1994) have been adopted for the Streptomyces sp. strain C5
58 Fig. 2.1. Type II polyketide synthase genes (PKS) from actinomycetes. Genes involved in carbon chain assembly, reduction, aromatization or cyclization are shown. Abbreviations: act: actinorhodin PKS from 51 coelicolor, gra: granaticin
PKS from 5. violaceoruber; fren: frenolicin PKS from 5. roseofulvus; gris: griseusin
PKS from 5. griseus; tcm: tetracenomycin PKS from 5. glaucescens', whiE: 5. coelicolor spore pigmentation PKS; cur: 5. curacoi spore pigment PKS; sno: nogalamycin PKS from 5. nogalator, mtm: mitbramycin PKS from 5. argillaceus; urd: urdamycin PKS from 5. fradiae', jad: jadomycin PKS from 5. venezuelae; mon:
5. cinnamonensis PKS making a yet unknown polyketide; bir: Sa. hirsuta PKS making a yet unknown polyketide; ard: Kibdella aridum PKS making a yet unknown polyketide. Arrow legend: White right slanted arrow, e.g., actl-or/l, ketoacyisynthase a; grey right slanted arrow, e.g., actl-orJII, ketoacyisynthase P; black short arrow, e.g., actl-orflll, acyl carrier protein encoding gene; vertical striped arrow, e.g., actVII, aromatase/cyclase; right half slanted, e.g., actlll, reductase; horizontal striped arrow, e.g., actIV, aromatase/cyclase; horizontal filled short arrows, e.g., tcmJ, cyclases/aromatase; shaded arrows, e.g., dpsC/dpsD and tcmN are genes that remain to be fully characterized.
59 * I |irmn. rm zzzzy dau/dpS G F E A B C D - 7 7 7 1 r z Z Z y m S ^ ^ TÏÏTlTjl 7 7 7 : - a C t actm acti-orfl. H, m act VII actIV . 7 7 7 ^ . 7 7 7 : 7 = 7 /h™^^ rmnrij gra 6 5 I 2 3 4 rZTTyBHH^7777 TnTTljl freil 1 2 3 5 4 7777/4* Z777 LHHOjl g h s 1 2 3 5 4
I J K L M N ^c7>=# ♦ ^ whiE n m IV V VI v n ^ 7 7 7 7 / * i \ > sch E l 2 3 4 ^ .7177/ * Il ‘ ^ cur C A B E F G (Irrnm 7777 .7777/Bhh^ §► sno E D 1 2 3 l|miin 7 7 7 /B 0 ^ l^ 7777 Ultlll Q P K S T 7777/^^*- 7777 Urd F A B C D 77777/ 9^ ^ 7777 nTTTTjl j a d 1 2 3 5 4 7777 ^ 7777 Hnn(l 131011 I 2 3 5 4 — —7 ma# hir 5 1 2 3 . 7 7 7 / = # - ^ ard 1 2 3
60 (Fig. 2.1; Strohl et al., 1989; Strohl et al., 1997). The PKS genes are denoted by the name dps proposed by Grimm et al., (1994). The daul gene (Fig 2.1) is not a PKS gene but has been shown to cause overproduction of actinorhodin in S. lividans (Strohl et ai.
1991; Li. 1995). dnri, a homolog of daul in S. peucetius, located in the same relative position as daul, encodes a transcriptional activator of several daunorubicin biosynthesis genes including the polyketide synthase genes (Hutchinson. 1995), and has been shown to have moderate similarity to known activators such as RedD. Actll-0rf4. and AfsR found in other Streptomyces spp. (Stutzman-Engwall et ai, 1992).
Deduced products and functions of the Streptomyces sp. strain C5 daunorubicin polyketide synthase Genes.
In the daunorubicin biosynthetic gene cluster, the dpsG gene is located 2 kb downstream of the daul gene, and encodes a negatively charged protein with a predicted
Mf of 9.437 and with a highly conserved 4-phosphopantetheine binding site (GLDSLAV) that is typical of most ACPs. DpsG (encoded by dpsG) shares a 43% amino acid sequence identity with Saccharopolyspora hirsuta Orf6 . that encodes an ACP putatively involved in the biosynthesis of a polyketide (Le Gouill et ai, 1993). dpsG is the only gene encoding an apparent ACP within the 40+ kbp of DNA representing the
Streptomyces sp. strain C5 daunorubicin biosynthesis gene cluster (Ye et ai, 1994;
Dickens et ai, 1995; Dickens and Strohl, 1996). The source of ACP is not critical for polyketide synthase synthesis by most Type II PKSs (McDaniel et ai, 1994; McDaniel et
61 ai, 1995a; McDaniel et ai, 1995b; KJiosla and Zawada, 1996). Although, when actl-
orflll, the gene encoding actinorhodin ACP was substituted for dpsG, polyketide
production, indicated by the ability to make aklanonic acid, was not observed (Rajgarhia
and Strohl, 1997; Strohl et al., 1997). In addition, Hutchinson (1997) reported that
substitution of the tetracenomycin ACP for S’, peucetius DpsG also resulted in no aklanonic acid formation. While these experiments are not conclusive, they may indicate a higher specificity of the daunorubicin PKS complex for ACP than observed with other
Type II PKSs.
The first gene downstream of dpsG, reading lefl-to-right, dpsH, encodes a small protein (deduced Mj- 16,758) of yet unproven function that has 45%, 35%, and 34% amino acid sequence identities with the orfK gene product from Streptomyces roseofulvus frenolicin polyketide synthase gene cluster (Bibb et ai, 1994), the actVl-orfA gene product from the S. coelicolor actinorhodin biosynthesis gene cluster (Femandez-Moreno et ai, 1994), and the orfX gene product from the Saccharopolyspora hirsuta polyketide biosynthesis gene cluster (Le Gouill et ai, 1993), respectively. These genes, except actVl-orfA. are clustered with polyketide biosynthesis genes. The functions of these genes and their products are as yet unknown, although it has been speculated that dpsH might encode either the second or the third ring cyclase or encode a helper protein that aids in ring cyclization (Gerlitz et al„ 1997; Strohl et ai, 1997).
About 5 kbp downstream of the dpsH gene are two genes, dpsF and dpsE, that are involved in the early stages of daunorubicin polyketide synthesis (Ye et ai, 1994). The
62 dpsF gene product has 44% amino acid sequence identity with ActVII, a bifunctional
cyclase/dehydratase of S. coelicolor involved in actinorhodin biosynthesis (Sherman et
ai, 1991; Femandez-Moreno et ai, 1992). Ye et ai (1994) have shown that dpsF gene
could heterologously complement an act F/7 mutation, which leads to lack of actinorhodin
biosynthesis in S. coelicolor strain B40, and restore production of actinorhodin. In
addition we have recently demonstrated (Rajgarhia and Strohl, 1997), that dpsF gene
product does function as the polyketide cyclase and is involved in aklanonic acid
formation.
The gene product of dpsE, polyketide reductase, has high amino acid sequence
identities with several known or predicted polyketide reductases (Hallam et ai, 1988;
Sherman et ai, 1989; Arrowsmith et ai. 1992; Le Gouill et ai, 1993). dpsE has been
shown to complement polyketide reductase deficient strains of S. galilaeus and S.
coelicolor (Ye et al., 1994).
The deduced dauG gene product (proposed M f 14,288) has a 28-29% sequence
identity with S. glaucescens TcmH (Shen and Hutchinson. 1993) and S. coelicolor
ActVa-OrfiS (Caballero et ai, 1991). S. glaucescens TcmH, has been shown to insert an
oxygen into tetracenomycin FI, a naphthecenone, to produce TcmD3 which is a 5. 12- naphthacenequinone (Shen and Hutchinson, 1993). Similarly, ActVa-0rf6 has been proposed to carry out a hydroxylase-like function in the biosynthesis of actinorhodin
(Caballero et ai, 1991). A 12-deoxyaklanonic acid oxygenase function has been implicated for the dauG product in catalyzing an analogous reaction where a 12-
63 deoxyaklanonic acid is converted to the anthraquinone aklanonic acid in Streptomyces sp.
strain C5 (Strohl and Connors, 1992).
The deduced gene product of Streptomyces sp. strain C5. dpsA, has 60 to 65%
amino acid sequence identities with the gene products of several PKS-orf's (ketoacyl
synthases) including tcmK of S. glaucescens (Bibb et ai, 1989), S. violaceoruber gra-orfi
(Sherman et ai, 1989), Saccharopolyspora hirsuta orfi (Le Gouill et ai, 1993), actl-orfl
of S. coelicolor (Femandez-Moreno et ai. 1992), S. cinnamonensis mon-orfi
(Arrowsmith et al., 1992), S. curacoi cur A (Bergh and Uhlen, 1992), and S. halstedii
spore pigment PKS orfl (Blanco et ai, 1993). DpsA contains the highly conserved
putative condensing enzyme active site GCTSGID (Femandez-Moreno et ai, 1992; Katz
and Donadio, 1993; Strohl et al., 1997), as well as a putative acyltransferase site GHSLG
in the C-terminal portion of the deduced amino acid sequence (Femandez-Moreno et ai.
1992; Strohl et al., 1997). Meurer and Hutchinson (1995) have recently shown that mutation of the C-terminal putative acyltransferase site of TcmK results in no phenotypic defect suggesting that this "conserved active site" is apparently not exclusively responsible for the acyltransferase function. Altematively, other yet unknown proteins from these Streptomyces spp. may substitute for this function.
The deduced gene product of Streptomyces sp. strain C5 dpsB has 52 to 58% amino acid sequence identities with Gra-Orfll of S. violaceoruber (Sherman et ai, 1989).
TcmL of S. glaucescens (Bibb et ai, 1989), Actl-Orfll of S. coelicolor (Femandez-
Moreno et ai, 1992), Sa. hirsuta PKS-Orfll (Le Gouill et ai, 1993). and S.
64 cinnamonensis Mon-Orfll (Arrowsmith et ai, 1992). This protein has been termed the
ketoacyisynthase p (previously known as the polyketide chain length factor (McDaniel et ai, 1993).
The Streptomyces sp. strain C5 dpsC gene is unique, both in the structure of its
deduced product and in its presence within the daunorubicin biosynthesis gene cluster
(Ye et ai, 1994). The dpsC gene product has 20 to 25% amino acid identity and almost a
50% similarity with FabH (P-ketoacyl acyl synthase III [KASIII]) of E. coli (Tsay et ai,
1992) and the module 3-p-ketoacyl synthase domain of erythromycin Type 1 PKS
(Donadio et ai, 1991). KASIII and related proteins all share a highly conserved cysteine active site (e.g., AAACAGF [E. coli FabH sequence]) at which the condensations apparently take place (Tsay et ai, 1992), a domain which is conspicuously absent in the analogous sequences (residues 115-121, RQVSNGG) of DpsC (Ye et al., 1994). The two regions of highest amino acid similarities between DpsC and the KASIII proteins are sequences also similarly conserved in chalcone synthases (Tsay et ai, 1992). i.e.. amino acids VYADGGTALVL (residues 163-173 of DauA-OrfC) and approximately 80 residues at the C-terminus end (Ye et ai, 1994). The putative role of dpsC gene product in polyketide synthase is discussed in Chapter 3 of this dissertation.
The deduced amino acid sequence of dpsD has 28 to 33% amino acid identity with each of the seven acyltransferase domains of the Sa. erythraea Fry A Type 1 PKS
(Bevitt et ai, 1991; Donadio et ai, 1991; Ye et ai, 1994), the two acyltransferase domains of modules 3 and 4 of the Streptomyces antibioticus oleandomycin Type 1 PKS
65 (Swan et ai, 1994), and the putative acyltransferase product of the Rhizobium meliloti fix23-2 gene (Petrovics et ai. 1993). The deduced product of dpsD also has 24% amino acid sequence identity with E. coli FabD (malonyl-CoAracyl carrier protein transacylase;
Verwoert et ai. 1992), with a much higher identity in the region around the active site serine residue. DpsD contains the highly conserved acyltransferase domain containing a serine active site motif (LGHSVGEM) in residues 90-97 (Ye et ai, 1994); no other apparent active sites are present in the deduced amino acid sequence, indicating that dpsD probably encodes an acyltransferase. Most known streptomycete polyketides are synthesized using acetate as a starter unit and need ketoacylsynthases encoded by KAS„.
KASp and a gene encoding the ACP. Since the daunorubicin polyketide synthesis is primed with a propionyl moiety (Bartel et ai, 1990; Strohl et ai, 1990; Strohl and
Connors. 1992), dpsD has been hypothesized to encode a propionyl-CoA:acyl carrier protein acyltransferase required for this polyketide synthesis step (Strohl et a/.. 1997).
This chapter describes the minimal Streptomyces sp. strain C5 genes that are required to produce aklanonic acid in heterologous hosts. In addition, this chapter discusses the effect on daunorubicin biosynthesis, produced by generating strains of
Streptomyces sp. strain C5, with deletions in the chromosomal copy of dpsC and dpsD genes.
66 MATERIALS AND METHODS:
Bacterial strains and growth conditions.
Streptomyces sp. strain C5, originally obtained from the Frederick Cancer
Research Center (McGuire et ai, 1980), has been described in detail elsewhere (Bartel et ai. 1990). Streptomyces lividans TK24 (Hopwood et al., 1985) was obtained from D. A.
Hopwood, and S. coelicolor CH999 (McDaniel et ai. 1993), in which much of the actinorhodin biosynthesis gene cluster has been deleted, was obtained from C. Khosla.
The strains used in this study are further described in Appendix A. Recombinant S. lividans TK24 strains were grown in YEME medium (Hopwood et ai. 1985) containing
10 pg of neomycin (Sigma; St. Louis, MO) and/or 50 pg of thiostrepton (Sigma; St.
Louis, MO) per ml as required for selection. Recombinant S. coelicolor CH999 strains were grown on plates of solid R2YE medium (Hopwood et ai, 1985) containing 10 pg of neomycin and/or 50 pg of thiostrepton per ml as required. Streptomyces sp. strain C5 and mutants derived from it were grown in nitrate-defined plus yeast extract (NDYE) medium as described previously (Bartel et ai. 1990; Connors et al.. 1990). If required for selection of plasmids in recombinant strains, neomycin was added at a concentration of 1 pg/ml. All strains were routinely maintained on R2YE solid medium. When required for selective pressure on plasmids, neomycin and/or thiostrepton was added to R2YE medium at 10 or 50 pg/ml, respectively.
Escherichia coli JM83 (£. coli Genetic Stock Center; Yale University. New
Haven, CT), was used to propagate plasmids and was grown in Luria-Bertani (LB)
67 medium (Sambrook et al., 1989). Plasmids were introduced into E. coli by
transformation, using standard procedures (Sambrook et al„ 1989). Ampicillin. sodium
salt (Sigma; St. Louis, MO), was added at a concentration of 100 pg/ml to cultures of E.
coli harboring pUC19 (Yanisch-Perron et al.. 1985), pWHM3 (Vara et al., 1989). or
derivatives made from them (Appendix B).
General genetic manipulations.
Procedures used herein for protoplast formation, transformation, and regeneration
ÎOT Streptomyces sp. strain C5 and mutants derived from it have been described elsewhere
in detail (Lampel and Strohl, 1986). S. lividans TK24 was transformed with plasmid
DNA as described by Hopwood et al. (1985). Plasmids from E. coli were routinely prepared according to methods described by Carter and Milton (1993). E. coli was grown overnight in 3 ml of Luria Bertani (LB) broth containing 100 pg/ml of ampicillin. The culture was centrifuged and the cell pellet was reconstituted in a buffer A (25 mM Tris-
HCL 10 mM EDTA, pH 7.4) with 50 pg/ml of RNaselA (Gibco BRL Inc.; Gaithersberg.
MD). Cell suspensions were combined with 200 pi of cell lysis solution (0.2 M NaOH).
The cleared mixtures were combined with 200 pi of neutralization solution (3 M potassium acetate, pH 4.8) and mixed by inversion until a white precipitate formed.
Precipitated mixtures were centrifuged at 14,000 x g for 1 min in a microfuge, and the supernatant solution was combined with 1.0 ml of a resin solution (10% [w/v] diatomaceous earth [Sigma; St. Louis, MO] in 4 M guanidine thiocyanate [Sigma; St.
68 Louis, MO], 50 mM NaCI, 20 mM Tris-HCI, 5 mM EDTA, pH 7.5). The
resin/supematant mixture was passed through a “Wizard” minicolumn (Promega,
Madison, WI), and the filtrate was discarded. The Wizard column, containing the DNA
bound to the resin, was washed with 1.0 ml solution of buffer B (200 mM NaCI, 20 mM
Tris-HCI, 5 mM EDTA, pH 7.5; 50% ethanol v/v). The washed column was centrifuged
for 1 min to remove any residual wash buffer and then transferred to a fresh tube, 100 pi
of sterile water was used to elute the DNA from the resin by centrifugation.
Plasmids from S. lividans TK24 were prepared according to methods described by
Hopwood et al. (1985). Large scale plasmid preparation was done according to the
protocols described in Sambrook et al. (1989). Plasmids used for transforming
Streptomyces sp. strain C5 was obtained from E. coli strain ET12567 {dam', dcm')
(Appendix A) using a large scale plasmid preparation method outlined in Sambrook et al.
(1989). The plasmids prepared by this large scale method were purified using a cesium chloride density gradient (Sambrook et al.. 1989). Digestion of DNA with restriction endonucleases was carried out according to the manufacturers' directions. Restriction mapping and other routine molecular methods used in this work were as described by
Sambrook et al. (1989). Plasmids used and constructed in this work are described in
Appendices B and C.
For probes derived from plasmids, the DNA inserts were obtained from plasmid
DNA by digesting with the appropriate restriction endonucleases, followed by size fractionation using gel electrophoresis (Sambrook et al., 1989), and purified from gel
69 slices by electroelution of the DNA using the procedure detailed below. Size fractionated
DNA fragments to be labeled were cut out of agarose gel. The slices were then minced and placed in washed dialysis tubing (Gibco, BRL; Gaithersberg, MD) in 1 X TAE buffer, clamped with plastic dialysis clamps and immersed in a electrophoresis apparatus containing 1 X TAE buffer (Sambrook et al., 1989). The dialysis tubing containing the gel slices was electrophoresed, upon which the DNA migrated out of the gel slices and was retained on the inner sides of the tubing. The I X TAE buffer containing the DNA was removed from the dialysis tubing and placed in sterile 1.5 ml test tube containing 1.0 ml of a resin solution (10% w/v diatomaceous earth in 4 M guanidine thiocyanate, 50 mM Tris-HCI, 20 mM EDTA. pH 7.0). The resin/supematant mixture was passed through a wizard minicolumn (Promega, Madison, WI), and the filtrate was discarded.
The column containing the DNA bound to the resin was washed with 1.0 ml solution of buffer B (200 mM NaCl, 20 mM Tris-HCI, 5 mM EDTA, pH 7.5; 50% ethanol v/v). The washed column was centrifuged for 1 min to remove any residual wash buffer and then transferred to a fresh tube, after which 100 pi of sterile water was used to elute the DNA from the resin. The inserts were then labeled by the ^‘P-random primer procedure
(Feinberg and Fogelstein, 1983), using 50 pCi of [a-^'P]dCTP per pg of DNA and random primer labeling kit (Stratagene; La Jolla, California) according to manufacturers specifications.
Streptomyces sp. strain C5 chromosomal DNA was extracted and purified using
CsCl, density gradients as described by Hopwood et al. (1985). After digesting the
70 chromosomal DNA using restriction enzymes according to manufacturers specifications, the DNA was size fractionated on I.O % agarose gels (Sambrook et al., 1989). The size fractionated DNA was transferred to Nytran nylon membranes (Schleicher & Schuell
Inc.; Keene, New Hampshire) using standard procedures detailed in Sambrook et al.
(1989). Hybridization of the previously labeled probes to the immobilized DNA was performed using the southern blot method (Sambrook et al., 1989).
Gene replacement methods.
For gene replacement experiments, the 3.75-kbp 5c/I-£coRI fragment containing the 3' end of dpsE as well as the entire lengths of genes dpsC and dpsD, was subcloned from pANTl2l (Appendices B and C; Ye et al., 1994) into 5a/wHI-£coRl digested pUCl9 (destroying the BamWl and Bcli sites). The resultant plasmid, pANT7l6
(Appendices B and C), was the starting point for construction of all subsequent replacement vectors used in this work.
To construct the plasmid in which dpsD was replaced with aphi, encoding neomycin phosphotransferase from plJ61 (Thompson et al., 1982), the 616-bp BamWl-
Kpnl fragment of pANT716 was replaced with a 1.27-kbp EcoBl-Kpnl fragment containing aphI from pKK840 (Appendices B and C; Kwak, 1995). For this experiment, pANT716 and pKK840 were digested with BamHl and £coRl, respectively, and those sites were then end-filled using Klenow fragment (Sambrook et al., 1989). The plasmids were then digested with Kpnl, and the fragments, each containing a blunt end and a Kpnl
71 site, were ligated using T4 DNA ligase (Sambrook et al., 1989) to generate pANT726.
This plasmid, in which the aphi gene was flanked to the left and right with 2.05- and
1.08-kbp fragments of homologous Streptomyces sp. strain C5 DNA, respectively, was used to transform protoplasts o f Streptomyces sp. strain C5 as previously described
(Lampel and Strohl. 1986). After 18 h. the regenerated protoplasts were challenged with
10 pg/ml neomycin, and neomycin-resistant transformants were picked over to fresh plates after 7 days.
For replacement of dpsC and dpsD the 1385-bp Sstl-Kpn\ fragment of pANT716. encompassing the 3’ half of dpsC and the 5' half of dpsD, was replaced with the 1.27-kbp
Eco^-Kpnl fragment containing aphi from pKK840 (Kwak, 1995). In this case. pANT716 and pKK840 were digested with 5s/I and £coRI respectively, and those sites were then end-filled using Klenow fragment (Sambrook et al., 1989). The plasmids were then digested with Kpnl. and the fragments, each containing a blunt end and a Kpnl site, as described above, were ligated with T4 DNA ligase (Sambrook et al., 1989) to generate pANT740. This plasmid, in which the aphi gene was flanked to the left and right with
1.28- and 1.08-kbp fragments of homologous Streptomyces sp. strain C5 DNA. respectively, was used to transform protoplasts of Streptomyces sp. strain C5 as described above.
Detection of anth racy dines and aklanonic acid.
Streptomyces sp. strain C5 and mutant strains C5-VR2 (dpsD null mutant) and
72 C5-VR5, C5-VR6, and C5-VR7 (dpsCD null mutants) were grown for 6 days at 30 °C in
50 ml of NDYE medium (1 |ig of neomycin per ml was added to the cultures of the mutant strains) in 250-ml Erlenmeyer flasks with springs for dispersal of the mycelia. At
6 days, the whole broth from each culture was adjusted to pH 8.5 with 5 N NaOH and then extracted with an equal volume of chloroform:methanol (9:1). The anthracyclines in the extracts were reduced to dryness, reconstituted in 100 pi of methanol, and separated on silica gel thin-layer chromatography (TLC) plates as previously described (Dickens et al.. 1996), with chloroformrheptanermethanol (10:10:3) as the liquid phase, or by reverse- phase high-pressure liquid chromatography (HPLC), with a solvent system of methanol:water:acetic acid (65:30:5), as previously described (Dickens et al.. 1996).
For detection of aklanonic acid from recombinant cultures of S. lividans TK24. the strains were grown for 4 days at 30 °C in 50 ml of YEME medium as described above. For extraction, the pH of each culture broth was adjusted to 1.5 with 6 N HCI. and the acidified broths were extracted with an equal volume of chloroform, reduced to dryness, and then reconstituted in 100 pi of chloroform. Aklanonic acid produced by the recombinant strains was compared with authentic aklanonic acid obtained from K.
Eckardt (1985) in parallel and by co-chromatography, using TLC as described previously
(Connors et al., 1990). Aklanonic acid also was analyzed by HPLC as previously described (Connors et al., 1990 ; Dickens et al.. 1996), using a C,g reverse-phase column and a solvent system of methanol:water:glacial acetic acid (65:30:5) (Connors et ai.
1990).
73 For detection of aklanonic acid produced by recombinant S. coelicolor CH999
strains, the strains were grown for 7 days on R2YE solid medium containing 10 and/or 50
pg of neomycin and thiostrepton per ml, respectively, as required for plasmid selection.
The agar containing the recombinant S. coelicolor CH999 lawns was removed from 25
plates, mashed, and extracted with 10 ml of methanolic HCI (12 N HCTmethanoI: 1:1)
per plate. The acidic extracts were filtered through cheesecloth and then extracted with an
equal volume of chloroform. The organic layer was removed, dried and reconstituted in
50 pi of chloroform, and analyzed as described above.
Radiolabeiing of aklanonic acid.
Fifty-milliliter cultures of recombinant S. lividans TK24(pWHM3) (negative
control) and S. lividans TK24(pANT785) were grown for 36 h at 30 °C in YEME
medium plus 50 pg/ml of thiostrepton. Then 50 pCi of filter-sterilized [l-'^C] propionic
acid-sodium salt (84 pCi/mmol) was added to each culture, and the cultures were
incubated for an additional 48 h with the added label. At the end of incubation, the
cultures were extracted, and the radiolabeled aklanonic acid was separated by TLC as described above and analyzed by autoradiography as previously described (Dickens et al..
1996).
Enzymatic conversion of aklanonic acid.
Authenticity of the aklanonic acid produced by recombinant strains was also
74 confirmed by conversion of the product in the presence of 5-adenosyl-L-methionine to aklanonic acid-methyl ester by extracts of S. lividans TK24(pANT154). using the methods described by Dickens et al. (1996). The products were analyzed by TLC and
HPLC (Dickens et al.. 1996).
MS analysis.
For mass spectrometry (MS) analysis. S. lividans TK24 (pANT785) was grown at
30°C for 48 h in 50 ml of YEME medium in 250-ml Erlenmeyer flasks with coiled springs, after which the pH of the whole culture broth was adjusted to 1.5, and the acidified solution was extracted with an equal volume of chloroform. The extracted aklanonic acid was separated from contaminants on TLC plates as described above, scraped from the plates as described previously (Dickens and Strohl. 1996). and extracted again with acidified chloroform. The purified aklanonic acid in chloroform was washed five times with double-distilled water and then dried. The mass spectra were acquired on a Micromass Quattro II (Altrincham) equipped with an atmospheric pressure chemical ionization (APCl) source operating in the positive-ion mode.
RESULTS AND DISCUSSION:
PKS gene structure
The gene organization for the daunorubicin PKS gene cluster of Streptomyces sp. strain C5 is not typical of other streptomycetes Type II PKS gene clusters, in which the
75 ACP gene usually is located immediately downstream of the genes for KAS„ and its homolog (KASp) (Fig. 2.1; Hopwood and Sherman, 1990; Khosla et al.. 1993;
Hutchinson and Fujii, 1995). It has been hypothesized that this novel PKS gene arrangement, unique in the daunorubicin biosynthesis gene cluster, is correlated with the requirement by this strain of a propionyl moiety as the chain initiator unit (Ye et ai.
1994; Grimm et al.. 1994; Meurer and Hutchinson. 1995). In addition, it is speculated that the two unusual genes, dpsC and dpsD (Ye et al.. 1994; Grimm et al., 1994), encode products that are required for initiation of the €21 polyketide chain.
The structure of the Streptomyces galilaeus PKS gene cluster (Fig 2.1; acl.). encoding assembly of aclacinomycin A (aclarubicin). an anthracycline also initiated with a propionyl moiety, has been reported to have an ACP gene in the location analogous to that occupied by dpsCD (Strohl et al.. 1997). This brought to light the question of whether Streptomyces sp. strain C5 or S. peucetius required dpsC. which encodes a homolog of E. coli fatty acid ketoacyisynthase III, but lacks an obvious active site (Ye ei al.. 1994; Grimm et al.. 1994), and dpsD, which encodes an acyltransferase homolog of as yet unproven function (Ye et ai, 1994; Grimm et ai. 1994), for polyketide biosynthesis.
Gene replacement.
A dpsD null mutant was constructed by transforming protoplasts of Streptomyces sp. strain C5 with pANT726 (Appendices B and C), a suicide vector in which dpsD was
76 replaced with aphl. Southern hybridization of several neomycin-resistant transformants, using the integrated DNA as a probe, revealed the presence of recombinants containing either single and double crossovers; a restriction map of a double crossover in which aphl replaced dpsD in the genome is shown in Figure 2.2. This mutant strain, Streptomyces sp. strain C5-VR2, was found to produce in NDYE medium copious quantities of various anthracyclines and anthracyclinones similar to those produced by the parental strain. This result was not totally unexpected, however, since Grimm et al. (Grimm et al.. 1994) had previously found S. peiicetius dpsD to be dispensable in aklanonic acid formation in a heterologous strain.
Null mutants of dpsCD in Streptomyces sp. strain C5 were then constructed by replacement of a 1385-bp Sst\-Kpnl DNA fragment, spanning most of both dpsC and dpsD, with aphl. Transformation of Streptomyces sp. strain C5 protoplasts with this plasmid resulted in recovery of several stable neomycin-resistant transformants. Southern hybridization, using the integrated DNA as a probe, confirmed several strains (C5-VR5.
C5-VR6, C5-VR7) in which double-crossover events had occurred; Figure 2.3 shows the results for three strains in which expected double crossovers occurred, resulting in replacement of dpsC and dpsD in the genome with aphl. These dpsCD mutants also were confirmed by Southern hybridization with the homologously integrating suicide plasmid ampicillin resistance marker, which was present in all single crossovers and absent in all double crossovers (data not shown). All three dpsCD null mutant strains of Streptomyces sp. strain C5 produced copious quantities of anthracyclines as determined by TLC and
77 Fig. 2.2. (A) Southern blot of genomic DNA of strainsStreptomyces sp. strain C5
(lanes 1-3) and Streptomyces sp. strain CSVR2 (lanes 4-6) digested with EcoRI, Pst\ and Sstl, respectively, showing replacement of dpsD in the chromosome of
Streptomyces sp. strain C5 with aphl. The Pstl-EcoRI fragment shown in panel B was used as the probe. (B) Restriction maps of the plasmid, parental chromosome and mutant chromosome, showing the expected results obtained by the double crossover.
78 4.1
B pANT726 Probe
aphl
es Chromosome dpsD E P S B«a>K E S E P I I___ u _ ■ 8.1 kbp 10.4 kbp 4.1 kbp C5-VR2 Chromosome E P S E SIS dpsD E S E P
M - ■ -2.3 kbp - 4.5 kbp 3.0 kbp
Fig. 2.2. Southern blet of genomic DNA of strain Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR2 (dpsD mutant).
79 Fig. 2.3. (A) Southern blot of genomic DNA of strainsStreptomyces sp. strain C5
(lanes 1 and 2) and Streptomyces sp. strain C5VR5 (lanes 3 and 4), C5VR6 (lanes 5 and 6), C5VR7 (lanes 7 and 8) digested with Pst\ and Sstl^ respectively, showing replacement of dpsC and dpsD in the chromosome ofStreptomyces sp. strain C5 with aphl. The Ps/I-EcoRI fragment shown in panel B was used as the probe. (B)
Restriction maps of the plasmid, parental chromosome and mutant chromosome, showing the expected results obtained by the double crossover.
80 B pANT740 Probe H dpiC' s |S dp«DE
aphl X C5 Chromosome dpsC dpsDsD E P E E s E p L_J______J___ LL 10.4 kbp 4.1 kbp C5-VR5.VR6.VR7 Chromosomes E P dpsC'«C s s |S |S dpsD^ c S E P LJ______I I I 4.5 kbp ■ 3.0 kbp
Fig. 2.2. Southern blot of genomic DNA of strain Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR25 {dpsCD mutant).
81 HPLC. Particularly interesting, however, was the fact that e-rhodomycinone, the major product found in fermentations of wild-type Streptomyces sp. strain C5, was virtually absent in fermentations of C5-VR5. C5-VR6. and C5-VR7; instead, the major products were several glycones. including daunorubicin. 13-dihydrodaunorubicin, and some anthracyclines not normally produced by Streptomyces sp. strain C5. These will be described in detail in Chapter 3.
Heterologous biosynthesis of aklanonic acid by minimal PKS.
In parallel with the gene disruption experiments described above, we tested daunorubicin biosynthesis PKS genes for their ability to encode gene products that would synthesize aklanonic acid in heterologous strains, using an approach similar to that used by Grimm et al. (1994). Our first experiments showed that pANT767 (Appendices B and
C), containing the Streptomyces sp. strain daunorubicin PKS gene cluster but not including dpsG (encoding AGP), did not confer aklanonic acid biosynthesis on S. lividans
TK24. This result confirms a similar experiment by Ye et al. (1994). in which S. lividans
TK24 (pANT122) (Appendices B and C; pANT122 contains the same insert as pANT767) did not produce aklanonic acid as possibly expected. Co-transformation of either S. lividans TK24 (Hopwood et al.. 1985) or S. coelicolor CH999 (McDaniel et al..
1993) with pANT767 (conferring neomycin resistance) and pANT751 (thiostrepton resistance) (Appendices B and C; the latter plasmid contains eight intact genes, of which one is dpsG and another is daul [Dickens and Strohl, 1996], a gene highly homologous to
8 2 dnri, a transcriptional activator gene for daunorubicin biosynthesis in S. peucetius
[Stutzman-Engwall et al., 1992]). resulted in the biosynthesis of a compound with the
TLC and HPLC characteristics of aklanonic acid (Connors et al., 1990). This indicated that certain daunorubicin biosynthesis genes in pANT751 were apparently required in addition to the PKS genes in pANT767 to confer the capacity for biosynthesis of the putative aklanonic acid. Similarly. S. lividans TK24(pANT767:pANT771) produced a compound chromatographically identical to authentic aklanonic acid (Rajgarhia and
Strohl.. 1997).
Using the preliminary results shown by Grimm et al. (1994). we then constructed a set of single plasmids (pANT782-pANT788; Fig. 2.4: Appendice B and C) in which
DNA fragments containing intact dpsG (encoding AGP) and daul (transcriptional activator) were subcloned from pANT751 and incorporated into pANT767 and derivatives were made from it. A bright yellow compound chromatographically identical to authentic aklanonic acid (Table 2.1) was produced in large quantities by S. lividans
TK24(pANT785). which contains only dpsG (AGP), daul (regulatory activator). dpsA
(KASJ. dpsB (KASp). dpsE (polyketide reductase). dpsF (aromatase). and dauG
(deoxyaklanonic acid oxygenase) (Fig. 2.5). This experiment confirmed that the two unusual open reading frames, dpsC and dpsD, were not required to synthesize the yellow product. As mentioned previously, Grimm et al. (1994) found that S. peucetius dpsD was not required for aklanonic acid biosynthesis in S. lividans.
83 Fig. 2.4. Restriction map generated by complete nucleotide sequence data (Ye e/ a/.,
1994; Dickens et al., 1995; Dickens and Strohl, 1996) of part of the daunorubicin biosynthetic gene cluster ofStreptomyces sp. strain C5. The genes depicted by black arrows are the minimal biosynthetic genes required for aklanonic acid biosynthesis in heterologous strains as described in the text. The transcriptional activator gene, daul, is indicated by slanted lines. The daunorubicin PKS biosynthetic genes dpsC and dpsD, proven in this work not to be required for anthracycline biosynthesis, are indicated by the shaded arrows. The inserts in plasmids conferring or unable to confer aklanonic acid biosynthesis on S. lividans TK24 and S. coelicolor CH999 are shown below. These inserts are present in the plasmids in the same orientation that is shown in this figure, and the genes are transcribed from their natural promoters.
Complete details for construction of the plasmids are given in Table 1. A bar indicating 2.0 kbp is given. Abbreviations: B, Ba/nHI; Be, BcH', E, EcoRI; S, 5^/1;
Sa, SalV', +, aklanonic acid produced in substantial quantities; -, aklanonic acid not produced; w, only traces of aklanonic acid produced; nd, not done.
84 dps dps dps n dpsH I------1 I------1 L . 2E ______PANT767 ------pANT767/751 + + ------pANT767/771 + nd — ______pANT782 + +
00 LA — ______pANT783 ------pANT784 w — ■ pANT785 + + ------pANT786 w — ______pANT787 — ______pANT788 + +
Fig. 2.4. Restriction map generated by complete nucleotide sequence data of part of the daunorubicin biosynthetic gene cluster 0Ï Streptomyces sp. strain C5 and the inserts in plasmids that confer or is unable to confer aklanonic acid biosynthesis on S. lividans TK24 and S. coelicolor CH999. Table 2.1. Characteristics of products of daunorubicin PKS genes in heterologous hosts
86 O H G O H O
Results of: S am ple TLC (/?/ HPLC APCI-MS (RT in min)'* (M +1)"
S. lividans TK24(pANT785) product 0.44 6.55 335.16 Authentic aklanonic acid 0.44 6.55 335.16
® Solvent system was heptanerCHCljcCHjOH (50:50:25). Compounds were viewed under UV light with a wavelength of 260 nm. RT, retention time. Solvent system was methano I : water : acetic acid (65:30:5). Peaks resulting from mass spectral analysis using APCI-MS as described in Materials and Methods. M+1 indicates a mass of 334 for the decarboxy-anhhydroaklanonic acid (shown above).
Table 2.1. Characteristics of products of daunorubicin PKS genes in heterologous hosts
87 Fig. 2.5. Biosynthesis of tetracenomycin F2 by dpsABtcmMNJ (Meurer and
Hutchinson, 1995), in which the starter unit is acetyl-CoA and the first linkage is a
C-9-C-I2 linkage, compared to biosynthesis of aklanonic acid by dpsABFGdauG
(Rajgarhia and Strohl, 1997), in which propionyl-CoA is the starter unit and the first linkage made is C-7-C-12.
88 dps A dpsB tcmM tcmN tcmJ TCM F2
OH OH OH OH Acetyl-CoA + 9 acetyl units COACP GOOH OH OH OH
dpsA dpsB dpsG dpsF dpsE dauG Aklanonic acid
00 \o COACP o COOH o Propionyl-CoA + — 9 acetyl units
OH O OH O
Fig. 2.5. Biosynthesis of tetracenomycin F2 by dpsABtcmMNJ compared to biosynthesis of aklanonic acid by dpsABFGdauG. Identity of the product.
When the bright yellow compound produced by S. lividans TK24(pANT785) was isolated from TLC plates and incubated in vitro with mycelial extracts containing aklanonic acid methyltransferase (encoded by dauC) by procedures we have described previously (Dickens et al., 1996). it was converted to aklanonic acid methyl ester, indicating that it was biologically-active aklanonic acid. Similarly, when incubated with
[l-''*C]propionate. cultures of S. lividans TK24(pANT785) produced a radioactive compound that co-chromatographed on TLC with authentic aklanonic acid, indicating that the putative aklanonic acid product incorporated propionate as expected (Fig. 2.5).
Finally, purified compound from culture broths of S. lividans TK24(pANT785) that was subjected to APCI-MS analysis gave an M+1 of 335.16 (Table 2.1). which is identical to that obtained with authentic aklanonic acid obtained from K. Eckardt. This corresponds to the expected M + 1 for decarboxy-anhydroaklanonic acid {mz'^ 334). which is similar to the chemical breakdown products observed in mass spectral analyses of aklanonic acid by
Grimm et al. (1994). Eckardt et al. (1985) also had stated that they were unable to obtain a true M + 1 for aklanonic acid, probably because of its instability upon ionization under the conditions of the APCI-MS.
Are DauZ and DpsG or their homologs required for efficient aklanonic acid biosynthesis?
Homologs of dauZ, a gene that is divergently transcribed from dpsG (Fig. 2.1;
90 Dickens et al., 1995), has been found in the PKS gene clusters of Streptomyces roseofulvus (frenolicin biosynthesis; Bibb et al.. 1994), Streptomyces argillaceiis
(mithramycin biosynthesis; Lombo et al.. 1996), and Saccharopolyspora hirsuta (spore pigment biosynthesis; Le Gouille et al.. 1993), suggesting that dauZ might play a role in polyketide synthesis and assembly (Hutchinson. 1997). A dauZ homolog, actVl-orfA (11), also has been found in the actinorhodin biosynthesis pathway and is presumed to be present in both S. lividans TK24 and S. coelicolor CH999. Thus, we tested the effect of overproduction of dauZ on the formation of aklanonic acid in S. lividans TK24 and S. coelicolor CH999. Cultures of S. lividans TK24(pANT782/pANT791) and S. coelicolor
CH999(pANT782/pANT791) contained the minimal genes required for aklanonic acid biosynthesis on one plasmid (Fig. 2.4) and dauZ. expressed from the SnpR-activated snpA promoter (DeSanti, unpublished data; Dickens and Strohl, 1996), on the other plasmid. Both of these cultures produced aklanonic acid but in quantities not significantly different from those produced in these strains by pANT782 alone. While this does not prove the requirement or lack thereof for dauZ or its homologs in PKS function, it does indicate that overproduction of DauZ has no significant impact on product formation.
S. lividans TK24 transformed with either pANT784 or pANT786 also made minute quantities of aklanonic acid, but S. coelicolor CH999 transformed with the same plasmids did not. These results suggest that the actinorhodin biosynthesis AGP in S. lividans TK24 may substitute, albeit weakly, for the daunorubicin AGP. AGPs from several different polyketide biosynthesis gene clusters appear to be functionally
91 interchangeable (Khosla et al.. 1993; McDaniel et al, 1993). While the small amounts of aklanonic acid produced by S. lividans TK24(pANT784) may reflect the copy number difference between the recombinant daunombicin PKS genes and the host actinorhodin gene encoding ACP, we have also observed that recombinant actinorhodin ACP at a high copy number also did not substitute efficiently for DpsG in aklanonic acid biosynthesis.
Similarly, Hutchinson and his colleagues recently found that tcmM, encoding the ACP for tetracenomycin biosynthesis, did not substitute for dpsG with constructs containing S. peucetius dpsABEFdnrl (dpsG or tcmM) to produce aklanonic acid in a heterologous system (Hutchinson. 1997). These results suggest that ACPs utlized in anthracycline biosynthesis possess greater substrate specificity.
Mechanistic and evolutionary implications.
Meurer and Hutchinson (1995) recently carried out experiments in which S. peucetius dpsA and dpsB were expressed together with tcmMNJ (encoding, respectively.
ACP. aromatase. and a protein of unknown function from tetracenomycin C biosynthesis;
Meurer and Hutchinson. 1995) to synthesize tetracenomycin F2. a compound that is initiated with an acetyl moiety (Fig. 2.5). This result led them to speculate that dpsA and dpsB did not contain the information necessary to dictate starter unit specificity (Meurer and Hutchinson, 1995). They suggested that dpsC and dpsD together were the primary contributing factors for starter unit specificity (Meurer and Hutchinson, 1995).
Our results show that Streptomyces sp. strain C5 dpsC and dpsD are not required
92 to specify a propionyl moiety as the starter unit in vivo in either S. lividans TK24 or S. coelicolor CH999 (Fig. 2.5). In light of the results shown by Meurer and Hutchinson
(1995), which indicate that DpsA and DpsB alone do not specify the propionyl starter unit, we propose that the productive protein-protein interactions of multiple proteins within the homologous PKS. including KAS„, KASp (described elsewhere as chain length factor; McDaniel et ai. 1993). "cyclase/aromatase," and polyketide reductase together specify the propionyl starter unit.
We have shown by both gene replacement approaches and heterologous expression that the products of dpsC and dpsD are not required for anthracycline biosynthesis in Streptomyces sp. strain C5. Thus, the information required to confer priming of the daunorubicin polyketide with a propionyl moiety does not lie within these two proteins, as originally hypothesized (Grimm et al.. 1994: Ye et al.. 1994; Meurer and
Hutchinson, 1995). In more recent experiments, however, we determined that the
Streptomyces sp. strain C5 dpsCD mutants C5-VR5. C5-VR6, and C5-VR7 also produce methyl side chain anthracyclines initiated with an acetyl moiety as well as propionyl side chain anthracyclines (see Chapter 3). Similarly, recent preliminary data indicate that S. lividans TK24(pANT785), containing dpsABEFGdauGI (lacking dpsCD), produces not only aklanonic acid but also a methyl side chain analog of aklanonic acid (Chapter 3;
Strohl et al.. 1997). Thus, while dpsCD are clearly not required for incorporation of propionyl-CoA as the starter unit, their presence may ensure that the propionyl moiety will be used to initiate polyketide biosynthesis one hundred percent.
93 Although dpsD was previously hypothesized to encode a propionyl-CoA:ACP acyltransferase. the deduced proteins of dpsD from both Streptomyces sp. strain C5 (Ye er a/., 1994) and S. peucetius (ATCC 29050) (Grimm et al., 1994) have sequences that are slightly more related to sequences of enzymes that carry out acetyI-CoA:ACP acyltransferase reactions than to those that are proposed to carry out propionyl-CoA:ACP acyltransferase reactions (Haydock et al., 1995). Thus, until the product of dpsD is studied biochemically, its actual function in daunorubicin biosynthesis remains unclear.
As mentioned earlier, the PKS gene cluster in S. galilaeus (Fig. 2.1) that encodes aclacinomycin A, an anthracycline also primed with a propionyl moiety, is more analogous to the PKS gene clusters for actinorhodin biosynthesis and other aromatic polyketides initiated with acetyl moieties (Hutchinson and Fujii. 1995) (Fig. 2.1.). In all of these Type II PKS biosynthesis gene clusters, the genes encoding KAS„, KASp, and
ACP are linked in apparent opérons (Hopwood and Sherman, 1990; ; Katz and Donadio.
1993; Hutchinson and Fujii. 1995). It is not known if homologs of dpsC or dpsD are present in the aclacinomycin A PKS gene cluster of S. galilaeus (Hutchinson, 1997).
Alternatively, it is conceivable that the functions of these proteins are utilized when they are present, but alternative enzymes, present in Streptomyces sp. strain CS. S. lividans TK24, and S. coelicolor CH999, may contribute the required activities when these proteins are absent. For example, S. coelicolor was recently shown to possess a
KASHI homolog, encoded byyâô//(Revill et al., 1995; Revill et al., 1996), which might provide the function of dpsC. Considering that S. coelicolor fabH and Streptomyces sp.
94 strain C5 dpsC contain only 23% sequence identity, however, this is probably not the case. Similarly, maionyl-CoA acyltransferase required for fatty acid biosynthesis, and probably also used in polyketide biosynthesis in S. coelicolor and S. glaiicescens
(Summers et al., 1995; Revill et al., 1996), may provide the function of the dpsD gene product. Finally, as mentioned previously, the product of resident actVl-orfA may provide the function for the dauZ product, if this gene were required. Just as with other heterologous PKS expression results (McDaniel et al., 1993; Meurer and Hutchinson.
1995), these types of questions will be best answered by reconstitution of minimal PKS systems in vitro with purified PKS enzymes.
95 CHAPTERS
Streptomyces sp. strain C5 POLYKETIDE SYNTHASE SPECIFIC dpsC GENE
PRODUCT HELPS MAINTAIN NATURAL STARTER UNIT SELECTION IN
POLYKETIDE SYNTHESIS
INTRODUCTION:
The sequences of the daunorubicin Type II PKS regions from Streptomyces sp. strain C5 and S', peucetius ATCC 29050 have been well characterized (Ye et al., 1994;
Grimm et al., 1994; Strohl et al., 1995; Strohl et al., 1997; Hutchinson. 1997). In the region that would usually be occupied by the gene encoding the ACP, there are two unusual ORFs (open reading frames). dpsC and dpsD (Ye et al.. 1994; Grimm et al..
1994).
The dpsC gene product, a homologue of Escherichia coli KASIII that lacks the active site cysteine residue but contains a putative coenzyme A binding site, is thought to play a structural role in forming the PKS complex (Ye et al., 1994; Strohl et al., 1997).
The dpsC gene product, as discussed in Chapter 2, has 20-25% amino acid identity and about 50% similarity with the FabH proteins (KAS III) of E. coli and spinach chloroplast
96 (Fig. 3.1). The two regions of highest amino acid similarities between dpsC gene product
and KASIII proteins are the amino acids VYADGGTALVL (residues 163-173 of dpsC
gene product), and the C-terminal eighty amino acid residues (Fig. 3.1 ; Ye et al., 1994).
The dpsC gene product was unique to Streptomyces sp. strain C5 and S. peucetius
PKS gene clusters until recently, when an orfi gene (tnmcated) from Streptomyces
viridochromogenes TQ57, was found to have about 55% sequence identity to the dpsC
gene (Gaisser et al., 1997). The orfl gene is foimd in the orsellinic acid synthase cluster
of S. viridochromogenes Tü57 and is speculated to encode a protein similar to dpsC gene
product that ensures starter unit specificity (Gaisser et al., 1997).
The dpsD gene product shares 28-33% amino acid identity with several PKS acyltransferases including, each of the seven acyltransferase domains of the Sa. erythreae
Fry A Type I PKS. and two acyltransferase domains of modules 3 and 4 of Streptomyces antibioticus oleandomycin Type 1 PKS product (Fig. 3.2; Ye et al.. 1994). The dpsD gene product contains a highly conserved acyltransferase domain containing a "serine" active site motif (LGHSVGEM) in residues 90 - 97 (Fig. 3.2). The dpsD gene product is speculated to function as a propionyl-CoA: acyl carrier protein (ACP-SH) acyltransferase
(Ye et al., 1994; Strohl et al., 1997). Figure 3.2 shows an alignment of the acyltranferases from several Streptomyces spp. Amino acid sequence of dpsD gene product shows the active site motif and aligns well with the known acyltranferase consensus sequence. Yet it is currently unknown whether dpsD gene product functions as acetyl-CoA:ACP acyltransferase or a propionyl-CoA:ACP transferase, or has no function
97 Fig. 3.1. LINEUP analysis of the PILEUP comparison (Devereux et a/., 1984) of the deduced amino acid sequences of dpsC gene product withE. coli (Ec) and spinach
(Sp) FabH (P-ketoacyl synthase III) enzymes. Only the regions of extensive amino acid similarities are shown, which includes the putative active site cysteine residues present in the FabH proteisn (designated with bullets) but missin in dpsC gene product a region which is also conserved with chalcone synthase (Tsay et a/., 1992), and the conserved C-terminal regions of the deduced proteins. Dashes indicate gaps generated by PILEUP and the consensus sequence shown was determined by
LINEUP.
98 FabH (EC) DEWIVTRTGI RERHIAAPNE TVSTMGFEAA THAIEMAGIE KDQIGLIWA TT— SATHAF PSAACQIQSM LGIKGCP-AF 106 FabH (S p ) DEWIATRTGI RQRHVLSGKD SLVDLAAEAA RNALQMANVN PDDIDLILMC TS— TPEDLF GSAP-QVQRA LGCSRTPLSV 175 D auA -O rfC DRRLASSTRM LSVAV-ADKE TPAEMAASAA RTAVDRSGVP PARIVLVLHA SLYFQGHHLW APASYVQRVA LG-NRCP-AM 112 Consensus DewiatrTgi r.rhv.a.ke t ...maaeAA r.A..magv. pd.I.Lll.a t hlf ,sA..q.q.a LG..rcP.a.
FabH (EC) DVAAACAGFT YALSVADQYV KSGAVK-YAL WGSDVLART C-DPTDRGTI IIFGDGAGAA VLA------A- -SEEPGIIST 177 FabH (S p ) DITAACSGFM LGLVSAACHV RGGGFK-NVL VIGADALSRF V-DWTDRGTC ILFGDAAGAV WQ------AC DSEEDGMFAF 248 D auA -O rfC EVRQVSNGGM AALELARAYL LAAPDRTAAL VTTGDRMSPP GFDRWNRPRH V-ŸADGGTAL I'LSRQGGFAR LRSLVTVSEP 191 Consensus dv.aac.Gfm .aL..A..yv ..g ..K..aL V.q.D .Isr. ..D.tdRgt. 1 . fqD qaqA . Vl*«*(**Aa *Se@ag, , a a
FabH (EC) HLHADGSYGE LLTLPNADRVNP E N S IH L T - MA GNEVFKVAVT ELAHIVDETL AANNLDRSQL 238 FabH (Sp) DLHSDGGGGR ML NASLL NDETDAAIGN NGAVTGFPPK RPSYSCINMN GKEVFRFAVR CITQSIEAAL QKAGLTSSNI 325 D auA -O rfC VLEGMHRGGH PFGPPSAE------EQRTVDLDA HSGRTWPRRE ARSASPRVSA GQE------EALAGAL KAAGVGLDDI 258 C o n s e n s u s a Lb a d g a gG a a 1 . . pnA a a a n.Ba a a a 1 ...... t S a S a a a 1118 G a EVf a a 3V 8L aaagla aSa i \D FabH (EC) DWLVPHQANL RIISATA-KK LGMSMDNVW TLDRHGNTSA ASVPCA-LDE AVRDGRIKPG QLVLLEAFGG GFTWGSALVR 315 FabH (Sp) DWLLLHQANQ RIIDAVA-TR LEVPSERVLS NLANYGNTSA ASIPLA-LDE AVRSGKVKPG NIIATSGFGA GLTWGSSIIR 403 D auA -O rfC SRWLPHHGW RRLSASYFGK WPVPPERTTW EFGRRTGHLG GGDPIAGFOH LVGSGRLAPG ELCLLVSVGA GFSWSCAWE 3 37 Consensus dwlvlhqaoa RilsAaaa ak l.vpaervaa alar.gntsa asaPaAalDe a V rsG r a kPG al llaafr.a r.ft-unaa
Figa 3a 1 a LINEUP analysis of the PILEUP comparison (Devereux et al., 1984) of the deduced amino acid sequences of dpsC gene product with E. coli (Ec) and spinach (Sp) FabH (P-ketoacyl synthase III) enzymeSa Fig. 3.2. Alignment of the acyltransferase active site motifs from several
Streptomyces spp. PKS acyltransferases. The acyltransferase motif of DpsD aligns with those of both acetyl-CoA:ACP acyltransferases as well as the propionyl-
CoA:ACP acyltransferase, although the conserved acyltransferase site, “GHSVG” appears more like that of other known acetyl-CoA:ACP acyltransferases.
100 ALIGNMENT OF ACYLTRANSFERASE ACTIVE SITE MOTIFS:
Active Bite L93 R 1 1 7 0 2 5 0 * * * ACETYL : Rap2,5,8,9,ll ETGYAQPALFALQVALFGLL - llaa - GHSVG Q Averm2 QTRYAQPALFAFQVALHRLL - llaa - GHSLG Q MSAS SSDRVQILTYVMQIGLSALL - llaa - GHSVG Q MCAT KTWQTQPALLTASVALYRVW - 12aa - GHSLG R Q Consensus ETGYAQPALFAMQVALFGLL - llaa - GHSLG R Q DpsD HVTRSQPLLFAVDYALGRMV - llaa - GHSVG R H Consensus RVDWQPALFAVMVSLAALW - llaa - GHSQG W N Rapl,3,4 RVDWQPASWAVMVSLAAVW - llaa - GHSQG N Ery6 RVDWQPVLFSVMVSLARLW - llaa - GHSQG N Cleans,6 RVDWQPALWAVMVSLARTW - llaa - GHSQG N MAS GIDKVQPAVFAVQVALAATM - llaa - GHSMG N STARTER : eryO RVEWQPALFAVQTSLAALW - llaa - GHSIG W S avermO RVDWQPTLFAVMISLAALW - llaa - GHSLG N
Fig. 3.2. Alignment of the acyltransferase active site motifs from several PKS acyltransferases from Streptomyces spp. at all. We have shown that the Streptomyces sp. strain C5 daimorubicin biosynthetic genes ketoacylsynthase a (dpsA), ketoacylsynthase p (dpsB), acyl carrier protein (dpsG). aromatase (dps F), polyketide reductase (dpsE) and putative deoxyaklanonic acid oxygenase dauG together produced aklanonic acid, an intermediate in the daimorubicin biosynthetic pathway, when expressed in a heterologous host (Chapter 2; Rajgarhia and
Strohl, 1997). Additionally we have shown that Streptomyces sp. strain C5VR5. a dpsCD double mutant of Streptomyces sp. strain C5, previously reported (Chapter 2; Rajgarhia and Strohl. 1997), produced detectable amounts of daimorubicin.
This Chapter describes the unusual anthracycline and anthracyclinone. feudomycin D (Oki et al., 1981) and feudomycinone C (Oki et al., 1981; Hoshino and
Fujiwara, 1983), respectively, that are produced along with daimorubicin, in Streptomyces sp. strain C5VR5 (dpsCD mutant). These unusual anthracycline and anthracyclinone products are made when the PKSs in the mutant strain use acetate as a starter unit, instead of the usual propionate. This Chapter also discusses the significance of the results obtained from complementation of the dpsCD double mutant strain with dpsC, which restores the natural starter unit selection by the PKSs in these complemented strain.
In addition, this chapter also describes a new intermediate desmethylaklanonic acid, that is produced along with aklanonic acid, when a minimal PKS gene combinations
(Chapter 2) from Streptomyces sp. strain C5, are expressed in S. lividans TK24 and S. coelicolor CH999.
102 MATERIALS AND METHODS:
Bacterial strains and growth conditions
Streptomyces sp. strain C5, originally obtained from the Frederick Cancer
Research Center, has been described elsewhere (McGuire et al.. 1980). S. lividans TK24
(Hopwood et al.. 1985) was obtained from D. A. Hopwood, and S. coelicolor CH999
(McDaniel et al.. 1993) in which much of the actinorhodin biosynthetic gene cluster has been deleted was obtained from C. Khosla. Recombinant S. lividans TK24 was grown in
YE ME medium (Hopwood et al.. 1985) containing 50 ug/m l thiostrepton as required.
Recombinant S. coelicolor CH999 strains were grown on plates of solid R2YE medium
(Rajgarhia and Strohl, 1997) containing 50 pg/ml thiostrepton. Streptomyces sp. strain
C5 and mutants derived from it were grown in Streptomyces - Connor (SC) medium containing, per L, glucose (22.5 g). malt extract (20 g), yeast extract (8 g). MOPS (20 g). anhydrous MgSO^ (0.1 g), zinc sulphate (10 mg), FeS0^7H;0 (10 mg), and supplemented with sodium nitrate (4.8 g; see below), at a final pH of 7.2. If required for selection, neomycin was added at concentrations of 1 pg/ml. All strains were routinely maintained on R2YE solid medium where 10 pg/m l neomycin was added as required.
Escherichia coli JM83 was used to propagate plasmids and cultures and was routinely grown on LB medium (Sambrook et al., 1989). Plasmids were introduced into
E. coli by transformation using standard procedures (Sambrook et al., 1989). Ampicillin was added at a concentration of 50 fig/ml to cultures of E. coli harboring pUC19
(Sambrook et al., 1989), pWHM3 (Vara et ai, 1989), or derivatives made from them.
103 General genetic manipulations
Procedures for protoplast formation, transformation, and regeneration for
Streptomyces sp. strain C5 and mutants derived from it have been described elsewhere
(Lampel and Strohl. 1986). S. lividans was transformed with plasmid DNA as described by Hopwood et al. (1985). Plasmids were routinely prepared according to the methods described by Carter and Milton (1993). Digestion of DNA with restriction endonucleases was carried out according to manufacturers' directions. Restriction mapping and other routine molecular methods used in this work were carried out as described by Sambrook et al. (1989). Plasmids used in this work and their construction are described in
Appendix B, and the plasmid maps are depicted in Appendix C.
Analysis of products isolated using LC/MS, HRFABMS and NMR analysis.
LC/MS (Liquid Chromatography/Mass Spectrometry) was performed on an API
300 (Perkin Elmer; Norwalk. Connecticut) interfaced with a HPLC system and a UV detector (Shimadzu Scientific Instruments; Columbia. Maryland) at the Ohio State
Campus Chemical Instrumentation Center (Columbus, Ohio). High resolution FABMS
(Fast-Ion Bombardment Mass Spectrometry) was performed at the Mass Spectrometry
Laboratory for Biotechnology at the North Carolina State University where the samples were analyzed using a JEOL (Tokyo. Japan) FIXIIOHF mass spectrometer. High resolution electrospray ionization mass spectrometry was performed on samples which failed to give a HRFABMS signal at the Mass Spectrometry Laboratory for
104 Biotechnology at the North Carolina State University. NMR analysis was performed by
Dr. Nigel Priestley using a Bruker AF270MHz, Broker AF250MHZ. and Bruker DRX400 spectrometers.
Isolation and detection of aklanonic acid.
For isolation of aklanonic acid from recombinant cultures of S. lividans TK24, the strains were grown for 96 h at 30“C in 50 ml YE ME medium. At 96 h. the cultures were centrifuged and the superaatent was decanted away from the cell pellet. The pH of the supematent was adjusted to 1.5 with 6N HCl and the acidified broth was extracted with an equal volume of chloroform, reduced to dryness, and reconstituted with ICO pi of chloroform (Rajgarhia and Strohl, 1997). Aklanonic acid produced by recombinant strains was compared with authentic standard aklanonic acid obtained from K. Eckardt
(1985a) and in parallel by co-chromatography using TLC as described previously
(Eckardt et al.. 1985b; Connors ei al. 1990; Dickens et al.. 1995). Aklanonic acid was analyzed by HPLC as previously described (Rajgarhia and Strohl, 1997) using a C,g reverse phase column and a solvent system of methanol: water: acetic acid (65:30:5)
(Connors et a i. 1990). Isolation of aklanonic acid from recombinant S. coelicolor
CH999 strains has been previously described (Rajgarhia and Strohl, 1997).
105 Isolation and detection of desmethyl aklanonic acid.
Recombinant S. lividans TK24 was grown for 96 h at 30°C in 50 ml YE ME medium. The procedure for isolation of the material was same as the procedure used for aklanonic acid and the final product was similarly reconstituted in 100 pi chloroform. In the absence of available authentic samples of this intermediate (desmethylaklanonic acid) for comparison, we performed LC/MS and high resolution FAB(E1/CI)MS of the material isolated by reverse phase HPLC. The sample eluting at 5.80 minutes, using a C,g reverse phase HPLC column with a solvent system as described above, was concentrated and
LC/MS and HRFABMS was performed to confirm the molecular composition and proposed structure of the product. A similar procedure was followed for isolation of desmethylaklanonic acid from the recombinant S. coelicolor CH999. The final product from S. coelicolor was confirmed by LC/MS analysis.
Production of anthracyciines and anthracyclinones.
The ideal medium for production of feudomycins were determined by carrying out several stirred-tank fermentations (10 L working volume) in different media.
Streptomycete-Connors (SC) medium and NDYE (Nitrate-defined plus Yeast Extract medium) were each supplemented with additional nutrients as described in Table 3.1.
NDYE medium contains, per liter: glucose (22.5 g), yeast extract (5 g), NaNOj (4.2 g).
K2HPO4 (0.23 g), trace salts solution (10 mL) as described in Hopwood et al. (1985). anhydrous MgSO^ (0.12 g), at a final pH of 7.2. The nutrients that were added as
1 0 6 Table 3.1. The development of optimal media for feudomycin production in stirred tank fermentors.
107 NDYE NDYE* NDYE + malt sc* sc SC SC Anthracycline + malt extract + + + sodium product extract + yeast sodium potassium nitrate 4-NaCI extract nitrate phosphate + + NaCI potassium phosphate
O00 feudomycin C (pg/ml) 2.5 0.85 1.4 2.83 3.8 b 1.7
daunorubicin (pg/ml) 0.5 1.06 0.85 0.5 0.25 - 0.45
feudomycin/daunorubicin 5 0.8 1.6 5.66 15.2 - 3.7
The composition of NDYE and SC media is described in materials and methods. No production was observed in this medium.
Table 3.1. The development of an optimal media for feudomycin production by Streptomyces sp. strain C5VR5 in stirred tank fermentors supplements to 1 L NDYE medium were: malt extract (20 g). or malt extract (20 g) +
NaCl (2 g), or malt extract (20 g) + NaCl (2 g) + Yeast Extract (3 g).
Similarly, nutrients that were added as supplements to 1 L SC medium (see above) were: NaNOj (4.2 g), or NaNO, (4.2 g) + K^HPO^ (0.23 g). The ratio of feudomycins to daunorubicin produced by Streptomyces sp. strain C5VR5 in each of the modified media was determined (Table 3.1). SC medium supplemented with 4.8 g of sodium nitrate was best for maximal feudomycin production and was used for all subsequent fermentations.
For the seed inoculum, spores of Streptomyces sp. strain C5-VR2 {dpsC mutant) and strain C5VR5 (dpsCD mutant) were grown in 50 mL of SC medium for 48 h at 30“C.
After 48 h, the inoculum was asceptically transferred to 500 mL of sterile SC medium in a 2 L flask with spring coiled. The flask was agitated at 250 rpm for 36 h at 30°C. This
36 h culture was subsequently used as a seed to inoculate a 14 L immersed bath fermentor containing 9.5 L of SC medium (supplemented with 4.8 g of sodium nitrate for
Streptomyces sp. strain C5VR5). Sterile Mazu OF 60P (5 mL) (PPG chemicals; Gurnee,
Illinois) was used as an antifoam agent. The fermentor was agitated at 250 rpm at 30"C for 132 h. At the end of the fermentation the supematent was separated from the cellular debris by continuous centrifugation using a Heraeus Contiftige 17RS (Heraeus
Instruments; South Plainfield, New Jersey). The clear supematent was then filtered through a 0.45 pm Millipore Pellicon ultrafiltration device and stored at 4“C until further treatment.
109 Isolation of anthracyclinones and anthracyciines.
A 100 cm X 10 cm glass column was packed with methanol pre-treated Amberlite
XAD2 resin (Fluka chemicals; Dorset. UK), after which the column was washed with 5 volumes of distilled water. The clarified extract from the fermentation broth was then passed through the column at flow rates of 2-to-3 ml/min. The anthracyciines and anthracyclinones were retained by the amberlite XAD2 resin. The material bound to the
XAD2 column was eluted by passing 5 volumes of methanol through the column. After the runs were complete, the Amberlite XAD2 resin was regenerated by treating it with 2 volumes of acidic acetone (pH to 2.5 with HCl), followed by treatment with 3 volumes of acetone.
The methanol extract of anthracyciines was evaporated, after which the pH of the aqueous extract was adjusted to 8.5 and extracted 3 times with chloroform; methanol (9:
1 ). The organic layer was pooled, concentrated and stored at 4“C until further treatment.
One-half of the extract was ffactioned by chromatography on silica gel to isolate the anthracyclinones, by eluting with a gradient of 2-to-20% methanol in chloroform. The eluted fractions that had the largest concentration of anthracyclinones (as determined by
TLC) were pooled and further ffactioned by chromatography on silica gel eluting with
30% hexanes in ethylacetate. The final fractions were analyzed by LC/MS, HRFABMS. and NMR.
The anthracyciines were purified by fractionating the other half of the concentrated extract by chromatography using silica gel chromatography, and a solvent
110 system consisting of hexanes; chloroform: methanol (50:50:15). This system separated the anthracyclinones from the anthracyciines. which were retained on the column.
Following this initial separation, a solvent system of chloroform: methanol: acetic acid: water (80:20:16:6) was used to elute the individual anthracyciines. The anthracyciines that were at the largest concentration as determined by TLC were pooled, concentrated and subsequently applied to reverse phase TLC preparatory plates (20 x 20, 1000mm;
Whatman; Fairfield, New Jersey). The plates were developed with a mobile phase of methanohwater (65:35; pH 2.5 with phosphoric acid). The anthracycline band was recovered and reconstituted in methanol. The sample was then further purified using reverse phase chromatography on a semi-prep HPLC column (Waters; Milford,
Massachusettes) using a mobile phase of methanohwater (65:35; 0.2% trifluoroacetic acid). The sample that eluted at 8.5 min was collected. The pure anthracycline was then separated from water by adjusting the pH of the eluent to 8.5 and extracting with chloroform: methanol (9:1). The organic layer was concentrated and the sample was applied to a LH20 (Pharmacia Biotech; Piscataway. New Jersey) column (50 cm x 2 cm) and eluted with 0.5% aqueous NH, in methanol. The isolated anthracycline band was concentrated in vacuo and analyzed using LC/MS and FABHRMS.
The NMR analyses were performed by Dr. Nigel Priestley (The Ohio State
University) on deglycosylated derivatives of feudomycin D, since the feudomycin D obtained by the above purification procedure contained significant levels of fatty acid impurities. The deglycosylation was performed by treating the pure sample with 6N HCl
III and heating at 90“C for 1 h. The deglycosylated sample was extracted in chloroform and the extract was fractioned by chromatography on silica gel to isolate the aglycone of
feudomycin D by eluting with a gradient of 2-to-20% methanol in chloroform. The eluted fractions that had the largest concentration of feudomycinone D (as determined by
TLC and confirmed by LC/MS) were pooled and further fractioned by chromatography on silica gel eluting with 30% hexanes in ethylacetate. This final sample was used for
NMR analyses.
The total anthracyciines produced by Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR5 was estimated by spectrophotometric analysis. Serial dilution of the anthracycline extracts obtained from 10 L fermentations of Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR5 was performed in chloroformimethanol (9:1).
Absorbance of the sample was taken at 430 nm wavelenth in 1 cm plastic cuvettes in a spectrophotometer. The molar extinction coefficient used for the calculations was 13,000
(W. R. Strohl, personal communication).
RESULTS AND DISCUSSION:
Heterologous biosynthesis of aklanonic acid and desmethyl aklanonic acid by PKS expression in S. lividans TK24 and S. coelicolor CH999.
As reported earlier (Chapter 2; Rajgarhia and Strohl, 1997), the minimal PKS genes from Streptomyces sp. strain C5, when expressed in the heterologous hosts S. lividans TK24 and S. coelicolor CH999, produced aklanonic acid. In addition to
112 aklanonic acid, the minimal genes expressed in heterologous hosts produced another yellow chromophoric compound which on thin layer chromatography appeared to have an Rj-\evy similar to that of aklanonic acid. This second metabolite, from recombinant S. lividans TK24 and S. coelicolor CH999. was isolated by HPLC, and analyzed by LC/MS.
Using a combination of HPLC, and HRFABMS (Table 3.2) as described in materials and methods this compound was found to have a molecular mass of 382.0688 (382.0698 by
HRFABMS) and had one methylene group less than aklanonic acid. This intermediatewas named desmethylaklanonic acid. (Fig. 3.3). It was speculated that this novel intermediate is formed when the Streptomyces sp. strain C5 PKS used acetate, instead of the usual propionate, as the starter unit in polyketide biosynthesis. The production of desmethylaklanonic acid along with aklanonic acid in S. lividans TK24 and
S. coelicolor CH999 transformed with the 'minimal PKS’ genes (Rajgarhia and Strohl.
1997) from Streptomyces sp. strain C5 indicates that both acetyl and propionyl moieties are being used concurrently as starter units by the "aberrant” PKSs in the recombinant strain to prime polyketide synthesis, thus indicating that starter unit selection by the
Streptomyces sp. strain C5 PKS in this heterologous recombinant strain is relaxed. In addition. Gerlitz et al. (1997) have also shown that expression of Streptomyces peticetius minimal PKS in S. lividans TK24 produced SEK43 (Fig. 3.3). an intermediate derived from 10 acetates, along with aklanonic acid, hence confirming our finding that starter unit selection is relaxed with such systems.
113 Table 3.2. Characteristics of products produced when Streptomyces sp. strain C5 minimal PKS genes are expressed in heterologous hostS. lividans TK24.
114 Product TLC HPLC (RT in Calculated HR(EI)MS' (/?,)“ min)** mol. wt.
aklanonic acid 0.44 6.55 396.08 desmethylaklanonic acid 0.38 5.80 382.0688 382.0698
“Solvent system was heptane: CHCI 3: MeOH (10:10:3). Compounds were viewed under UV light with a wavelength of
LA 260nm. ** RT, retention time. Solvent system was MeOH: water: acetic acid (65:30:5) ' HR(EI)MS is described in Materials and Methods.
Table 3.2. Characteristics of products produced by Streptomyces sp. strain C5 minimal PKS genes in: S. lividans TK24(pANT785) Fig. 3.3. Structures of the unusual acetate-derived intermediates, desmethylaklanonic acid and SEK 43, made by the minimal PKS gene products from Streptomyces sp. strain C5 expressed in S. lividans TK24 and 5". coelicolor
CH999.
116 COOH
20
OH O OH OH O
Desmethylaklanonic acid OH-n
C2oHi40g 382.32 382.068867 C 62.83% H 3.69% O 33.48% SEK43
368.34 368.089603 C 65.22% H 4.38% O 30.41%
Fig. 3.3. Structures of the unusual acetate-derived intermediates, desmethylaklanonic acid and SEK43.
117 The heterologous hosts S. lividans TK24 and S. coelicolor CH999 were co
transformed with minimal PKS genes from Streptomyces sp. strain C5. and Streptomyces
sp. strain C5 genes dpsC and dpsD (pANTlllS) (Fig. 3.4; Appendices B and C),
Streptomyces sp. strain C5 genes dpsD (pANT 1115) (Fig. 3.4; Appendices B and C).
The secondary metabolites from these recombinant cultures were similarly analyzed by
HPLC and LC/MS. Desmethylaklanonic acid was not produced in cultures containing minimal PKS genes plus dpsC and dpsD (Fig. 3.4), although this recombinant strain produced aklanonic acid as expected. Additionally, recombinant cultures with minimal
PKS genes and dpsD alone continued to make yellow chromophoric products, which gave an LC/MS and HPLC profile similar to desmethylaklanonic acid and aklanonic acid
(Table 3.2). In addition to the above experiments, secondary metabolites from the recombinant S. lividans TK24(pANT788). containing the Streptomyces sp. strain C5 genes dpsA, dpsB, dpsC along with dauG, dpsE, dpsF. were analyzed by HPLC and
LC/MS. This strain only produced aklanonic acid; desmethylaklanonic acid was not detected among the secondary metabolite products produced by this recombinant strain.
Table 3.3 depicts the products (aklanonic acid or desmethylaklanonic acid) made when several combinations of the Streptomyces sp. strain C5 PKS genes are expressed in S. lividans TK24 and S. coelicolor CH999.
A quantitative analysis of the amounts of aklanonic acid produced by S. lividans
TK24 (pANT782) vs S. lividans TK24 (pANT785) indicates that in the absence of dpsCD, the amount of aklanonic acid produced is only 20% that of the amount of
118 Fig. 3.4. Restriction map, generated by complete nucleotide sequence data (Ye e/ a/.,
1994, Grimm et al., 1994) of part of the daunorubicin polyketide synthase gene cluster from Streptomyces sp. strain C5. The genes depicted by black arrows are the minimal PKS genes discussed in Rajgarhia and Strohl (1997). The inserts in plasmids conferring the ability on S. coelicolor CH999 (McDaniel et al., 1993) and S. lividans TK24 (Hopwood et al., 1985) to make aklanonic acid and desmethylaklanonic acid respectively is shown by a + or a - sign. Details of the plasmids are given in Appendix B. Abbreviations: B, BamHl; Be, Bcli', Bg, Bglll;
E, EcoRI; S, Atl; Sa, SaB.
119 dps dps ndpsH I I r desmelhyl- Bg Sa Be BciSa B E Aklanonic aklanonic Acid in: acid in; Plasmidls) TK24 CH999 TK24 C.H999 dnmji doxA V U A dnmT HEFEGA B CD daiil dpsG
pANT7H2
PANT785 + +
PANT788
ë PANT78S/11I8
PANT785/111S
Fig. 3.4. Restriction map, generated by complete nucleotide sequence data of part of the daunorubicin polyketide synthase gene cluster from Streptomyces sp. strain CS and the inserts in plasmids that confer the ability on S. coelicolor CH999 and S. lividans TK24, to make aklanonic acid and desmethylaklanonic acid. Table 3.3. Production of polyketide intermediates by Streptomyces sp. strain C5 PKS genes expressed in S. lividans TK24 and S. coelicolor CH999.
121 Desmethyl Strain(Plasmid) Genes Aklanoni -aklanonic c acid acid
S. lividans TK24(pANT782) dps A, dpsB, dpsC, dps D,dps E, dpsF, dpsG, dauG, - S. coelicolor CH999 (pANT782)
S. lividans TK24(pANT788) dps A, dpsB, dpsC, dpsE, dpsF, dpsG, dauG + - S. coelicolor CH999 (pANT788)
S. lividans TK24(pANT785) dpsA, dpsB, dpsE, dpsF, dpsG, dauG + + S. coelicolor CH999 (pANT785)
« S. lividans TK24(pANt785 + pANTl 115) dpsA, dpsB, dpsE, dpsF, dpsG, dauG + -h S. coelicolor CH999 (pANT785 + pANTl 115) + dpsD^
S. lividans TK24(pANT785 + pANTl 118) dpsA, dpsB dpsE, dpsF, dpsG, dauG + - S. coelicolor CH999 (pANT785 + pANTl 118) + dpsC, dpsD*^
+ / - indicates production and lack of production of aklanonic acid or desmethylaklanonic acid, respectively. 'Genes added to recombinant strains on a separate plasmid.
Table 3.3. Production of polyketide intermediates by Streptomyces sp. strain C5 PKS genes expressed in S. lividans TK24 and S. coelicolor CH999 aklanonic acid produced when dpsC and dpsD gene products was present (Table 3.4).
The ratio of the desmethylaklanonic acid to aklanonic acid produced by the recombinant
strain S. lividans TK24 (pANT785) was 60:40. A quantitative estimation of the amounts
of desmethylaklanonic acid produced was determined indirectly by comparison with the
actual amounts of aklanonic acid produced.
The lower quantities of metabolites produced by S. lividans TK24 (pANT785)
indicates that there might be some other intermediates (e.g.. SEK43) that might also be
accumulating in these strains which we have not evaluated.
Strain Amount of aklanonic Amount of acid desmethylaklanonic acid
S. lividans TK24 (pANT782) 14 mg/L none
S. lividans TK24 (pANT785) 3 mg/L 4.5 mgƓ
* Estimated from the amount of aklanonic acid produced and the mass spectral analysis of the products of the strain.
Table 3.4: Quantitative estimation of aklanonic acid and desmethylaklanonic acid.
Anthracycline production in the dpsD null mutant.
The products from a 10 L fermentation of the dpsD mutant strain Streptomyces sp. strain C5VR2 was isolated and characterized. The major anthracyciines (molecules with daunosamine sugar, e.g., daunomycin) and anthracyclinones (molecules lacking the
12] daunosamine sugar, e.g.. e-rhodomycinone) were analyzed by TLC, HPLC and LC/MS. e-Rhodomycinone was the major anthracyclinone accumulated by this dpsD null mutant, a result which was similar to that observed in fermentation broth of the wild-type
Streptomyces sp. strain C5. Similarly, analysis of the anthracycline compounds produced by the dpsD null mutant indicatd that daunorubicin was the major glycone product accumulated. Wild-type Streptomyces sp. strain C5 also accumulates e-rhodomycinone and daunorubicin in fermentation conditions. Thus, Streptomyces sp. strain C5VR2
(dpsD mutant) produces anthracyciines similar to the wild-type Streptomyces sp. strain
C5 strain. The dpsD mutation does not have any apparant change in anthracycline production. This result was not unexpected, as our group (Rajgarhia and Strohl. 1997). and Grimm et al. (1994), had earlier shown that the dpsD gene product was not essential for the Streptomyces sp. strain C5 polyketide synthase genes to assemble aklanonic acid, a pathway intermediate, in a heterologous host, or for that matter to assemble daunorubicin in mutants of Streptomyces sp. strain C5.
Anthracycline production in null mutants of dpsCD.
Analysis of the fermentation products of Streptomyces sp. strain C5VR5 (dpsCD mutant) indicate that the major anthracyclinone accumulated by the dpsCD mutant was an unusual molecule previously named feudomycinone C (Oki et al., 1981; Table 3.5). The structure of feudomycinone C was determined by HRFABMS (Table 3.5), and NMR analyses (Table 3.6). The structure of feudomycinone indicates that the molecule was
124 Table 3.5. Characteristics of the major anthracycline product of Streptomyces sp. strain C5VR5 {dpsCD mutant).
125 Results of :
Major product TLC (R/ HPLC Calculated HRFABMS (RT in mins)'’ Mol. wt. (M+1)'
Feudomycinone C 0.52 5.0 370.1052 371.1062 10-hydroxydaunomycin (Feudomycin D) 0.63 3.7 515.1791 516.1871
"Solvent system used for the anthracyclinone (feudomycin C) detection was heptane:CHClj:MeOH (10:10:3) while the solvent system for the anthracycline ( 10-hydroxydaunomycin) detection was CHCl]:MeOH:acetic acid:water (80:20:16:6). K Compounds were viewed under UV light with a wavelength of 260 nm. ** RT, retention time. Solvent system was MeOH.water: phosphoric acid (65:30:5) ' HRFABMS is described in Materials and Methods.
Table 3.5. Characteristics of the major anthracycline products oï Streptomyces sp. strain C5VR5 {dpsCD mutant) Table 3.6. ‘^C NMR data for feudomycinone C.
127 o OH 10 OH 12a,
OHOH
Feudomycinone C
Assigned Aglycone of feudomycin C (CDCI3) ppm
1 118.2 2 136.8 3 118.8 4 161.3 5 187.2 12 187.9 6 156.4 11 155.8 4a 119.9 12a 136.1 5a 113.7 11a 113.0 6a 135.9 10a 135.3 7 63.0 8 35.6 9 70.6 10 68.6 13 25.9 OMe 56.8
Table 3.6. ‘^C NMR data for feudomycinone C.
128 derived from polyketides made by the condensation of an acetate molecule with nine C2 moieties derived from malonyl-CoA. In addition to the feudomycinone C. Streptomyces sp. strain C5VR5 also produces daunorubicin, which is made when the PKS in the strain uses propionate as a starter. This indicates that in the absence of the dpsC and dpsD gene products, the PKS in this strain utilizes both acetate and propionate as a starter to prime polyketide biosynthesis.
Further analysis of the anthracycline produced by the dpsCD null mutant revealed that an unusual compound named 10-hydroxydaunomycin (Oki et al., 1981; Hoshino and
Fujiwara, 1983; feudomycin D; Table 3.5) accumulated in the fermentation broth of these strains. The structure of feudomycin D was also determined by HRFABMS (Table 3.5). the aglycone of this unusual anthracycline was determined by NMR analyses (Table 3.7).
The structure of feudomycin D suggests that this compound is also derived from polyketides made by the condensation of one acetate with nine C2 moieties derived from malonyl-CoA. Significantly, feudomycins were not detected in the fermentation broths of either wild-type Streptomyces sp. strain C5 or the dpsD null mutant that was analyzed earlier.
Complementation of Streptomyces sp. strain C5VR5 {dpsCD mutant) with dpsC and dpsD restored the original range of anthracyclines accumulated (Table 3.8). No feudomycins were detected in fermentation broths of these complemented strains. More importantly, transformation of the mutant strain with dpsD alone did not restore wild-type anthracycline production (Table 3.8), as the strain continued to make feudomycin D and
129 Table 3.7. 'H NMR chemical shift assigments (ppm) for (deglycosylated) feudomycin D.
130 OH OH 10 OH 2 12a
3
OH OH
Feudomycin D (deglycosylated)
Shift“/ppm Integral Multiciplicity J/H z Assigned
1.34 3 s 13 -H 1.92 1 d 14.80 H -2 2.09 1 dd 14.80. 4.20 H -2 3.93 s - 4 - OMe 4.62 1 d 1.15 H - 10 5.05 1 d 3.90 H -7 7.47 1 d 8.00 H - 1 7.74 1 d 6.70 H -2 7.91 1 d 6.70 H -3
Relative to CD3OH at 4.78 ppm.
Table 3.7. 'H NMR chemical shift assigments (ppm) for (deglycosylated) feudomycin D.
131 Table 3.8. Major fermentation products ofStreptomyces sp. strain C5 and
Streptomyces sp. strain C5 mutants complemented with PKS genes.
132 Fermentation products :
Strain + plasmid (genes encoded in the plasmids) £-Rhodomyeinonc daunorubicin feudomycinone C and feudomycin D
Streptomyces sp. strain C5 (wild type) + + -
Streptomyces sp. strain C5VR2 {dpsD mutant) + none + + -
Streptomyces sp. strain C5VR5 {dpsC/dpsD mutant) + none - + +
Streptomyces sp. strain C5VR2 + pANT 1116 (dpsD) + + - w Streptomyces sp. strain C5VR5 + pANT 1120 {dpsC and dpsD) + -
Streptomyces sp. strain C5VR5 + pANT 1121 {dpsC) + + -
Streptomyces sp. strain C5VR5 + pANT 1116 (dpsD) - + +
+ / - indicates production or lack of production of daunorubicin or feudomycins, respectively.
Table 3.8. Major fermentation products oïStreptomyces sp. strain C5 mutants and Streptomyces sp. strain C5 mutants complemented with PKS genes. feudomycinone C, while complementation of Streptomyces sp. strain C5VR5 {dpsCD mutant) with dpsC alone restored the strains wild-type product profile and no feudomycin-like compounds were detected (Table 3.8). In the presence of the dpsC gene product, the selection of starter unit seems to be stringent, allowing only propionyl-CoA to be used to prime polyketide biosynthesis. Table 3.8 is a simplified table that depicts the products (daunorubicin or feudomycins) produced by the Streptomyces sp. strain C5 mutants and Streptomyces sp. strain C5 transformed with PKS genes.
Comparison of Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR5 with respect to amounts of anthracyclines produced.
Table 3.9 is a comparison between the Streptomyces sp. strain C5 and
Streptomyces sp. strain C5VR5 with respect to the amounts of anthracyclines produced using SC + nitrate medium (described in materials and methods).
The total anthracycline produced by Streptomyces sp. strain C5 was two-fold higher than the total anthracycline produced by Streptomyces sp. strain C5VR5. The ratio of the amounts of daunorubicin and 13-dihydrodaunorubicin to feudomycin, produced by the Streptomyces sp. strain C5VR5 {dpsCD mutant) was 40:60.
SIGNIFICANCE OF THE RESULTS:
We had originally proposed that the unusual arrangement of the PKS gene cluster in Streptomyces sp. strain C5 and S. peucetius, namely the presence of the two genes
134 Table 3.9. Comparison of the anthracycline production by Streptomyces sp. strain
C5 and Streptomyces sp. strain C5VR5.
135 daunorubicin Total & Feudomycin D** Strain anthracyclines* 13-dihydro daunorubicin'' Streptomyces sp. strain C5 430 mg/L 5.5 mg/L none
Streptomyces sp. strain C5VR5 150 mg/L 1.8 mg/L 2.83 mg/L
Total anthracycline produced was determined by spectrophotometric analysis of the extracted samples as described in materials and methods. ** The concentration of daunorubicin, 13-dihydrodaunorubicin, and feudomycin D was determined by HPLC as described in materials and methods.
Table 3.9. Comparison of the anthracycline production by Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR5.
136 dpsC and dpsD downstream of the ketoacylsynthase KASp (Fig. 3.5), has a role in ensuring the selection of propionyl- CoA as the starter unit in anthracycline biosynthesis
(Rajgarhia and Strohl. 1997). From our earlier results (Rajgarhia and Strohl. 1997), as well as those from Meurer and Hutchinson (1995), it became more evident that starter unit selection in anthracycline biosynthesis was genetically directed rather than a reflection of the concentration of precursors in the intracellular pool (Strohl et al., 1997:
Hutchinson, 1997).
The idea of genetic selection of starter unit is further strengthened by the generation of mutants of S. coeruleorubidus (Oki et al., 1981), which produced feudomycins from acetate and butyrate starter units instead of the usual propionate.
Concurrently, mutants of S. galilaeus (aclacinomycin producer) were isolated that used isobutyrate as starter instead of the usual propionate (Oki et al.. 1981: Soga et al., 1981 ).
These experiments backed the idea that starter unit selection was genetically determined and not a reflection of available precursor pools. Recently, a dpsC homolog, orfl, was reported in the S. vihdochromogenes TÜ57 avilamycin biosynthetic gene cluster (Gaisser et al., 1997). A truncated orfl shares a 55% sequence identity with the corresponding N- terminal region of dpsC gene from S. peucetius 29050 (Gaisser et al., 1997). A role similar to that speculated for dpsC gene product, i.e., determining the starter unit selectivity, has been suggested for the avilamycin orfl gene product.
We conclude from our results that the products of the downstream gene, dpsC, although not essential for the formation of the polyketide precursor initiated with
137 Fig. 3.5. Model showing the PKS gene organization, starter moiety, and an example intermediate for daunorubicin, produced by Streptomyces sp. strain C5
(intermediate: aklanonic acid; Eckardt, 1985, Strohl et al., 1989), aclarubicin,
produced by S. galilaeus ATCC31133 (intermediate: aklanonic acid, Tsukamato et al., 1994), and actinorbodin, produced by S. coelicolor (intermediate: dibydrokalafungin; Femandez-Moreno et al., 1994) biosynthesis. The genes encoding the following functions are indicated: KAS„, solid arrows; KASp, shaded arrows; Acyl carrier protein (AGP), open arrow; polyketide reductase, arrows with left slanted lines, cyclases/aromatases, arrows with right slanted lines; deoxyaklanonic acid oxygenase, dotted arrows; fatty acid KASHI bomolog lacking an apparant active site, arrow with horizontal lines; acyl transferase, arrow with vertical lines. The position of the starter unit within the intermediates are indicated by thicker lines.
138 PKS Gene Cluster Starter Unit Intermediate Final Molecule
4.S kb Propionyl Daunorubicin (daii) (ipsG (IpsF dpsE G dpsA dpsB dpsC dpsD
Propionyl Aclarubicin (akn) E A X B C D
nO Acetyl Actinorbodin
(act) III l-l 1-2 1-3 VI! IV OOH
Fig. 3.5. Model showing the PKS gene organization, starter moiety, and an example intermediate for daunorubicin, produced by Streptomyces sp. strain C5, aclarubicin, produced by S. galilaeus (ATCC 31133), and actinorbodin, produced by S. coelicolor. propionyl-CoA, is required to maintain the integrity of the PKS. In the absence of the dpsC gene product, the complex is "loose" and accepts both propionyl as well as acetyl as starter units in polyketide synthesis. Our hypothesis is based partially on results
(Rajgarhia and Strohl. unpublished data) that we obtained by applying the Two-Hybrid system for protein-protein interactions among the PKS proteins. The Two-Hybrid system
(Iwabuchi et al., 1993; Fields. 1993) is a genetic tool used routinely to analyze interactions among proteins in vivo. We used the system to detect possible interaction(s) between dpsA gene product and gene products of the rest of the PKS genes.
The results we obtained from this analysis indicated that dpsA gene product interacted specifically with dpsC gene product from among the PKS gene products. This interaction was characterized by (3-galactosidase assays as per the protocols for the use of the Two-Hybrid system (Iwabuchi et al., 1993). The assay suggested that the interaction among the dpsA gene product and the dpsC gene product was a weak interaction. This result was not unexpected, since interactions among the PKS gene products if any. could be influenced by the presence of the nascent polyketide product. Such polyketide products would not be present under the conditions of the assay used for the Two-Hybrid system. The detection of interaction between gene products of dpsA and dpsD supports the idea that this interactions holds the possible PKS complex together and allows stict starter unit selection. In the absence of such an interaction as is probably the situation in
Streptomyces sp. strain C5VR5, the complex formed may be irregular or “loose" as we have previously indicated.
140 Our results indicate that Streptomyces sp. strain C5 dpsCD are both not necessary, but dpsC gene product assists in specifying a propionyl moiety as the starter unit in vivo in either S. lividans TK24 or S. coelicolor CH999 (Rajgarhia and Strohl, 1997). Meurer and Hutchinson (1997) earlier showed that DpsA and DpsB alone did not specify the propionyl starter unit or C7-C12 linkage. On the basis of our results we can propose that
DpsC along with DpsA and DpsB together are responsible for starter unit and possibly
C7-C12 linkage.
Our hypothesis is that the daunorubicin PKS is normally primed with propionyl-
ACP possibly as a result of the functions of the dpsD gene products, which catalyze the reaction: propionyl-CoA + ACP-SH propionyl-SACP + CoASH. The dpsC gene product might be playing a structural role in forming a tight PKS complex. The propionyl moiety is transferred to the putative heterodimeric KAS„p to form a propionyl-KAS intermediate, freeing up the ACP-SH to accept a malonyi moiety (Strohl et al.. 1997).
In the absence of the dpsD gene product, we speculate that the daunorubicin and doxorubicin PKSs {dpsA, dpsB, dpsC. dpsG, gene products) can accept propionyl-CoA as the priming substrate that would be transferred directly to the enzyme. Our data suggest that in the absence of dpsC gene product, the aberrant PKSs formed can accept acetyl-
CoA, or perhaps acetyl-SACP, which might be substituting freely for propionyl-CoA.
141 Comparison ofStreptomyces sp. strain C5 PKS gene cluster with aclacinomycin and nogalamycin PKS gene cluster.
The PKS gene cluster encoding aclacinomycin A, an anthracycline also primed with a propionyl moiety, is homologous to the PKS gene clusters for actinorbodin biosynthesis and other aromatic polyketides (Tsukamato et al., 1994; Fig. 3.5). In this case, the genes encoding KAS„, KASg, and AGP are linked, just as they are in actinorbodin biosynthesis (Fig. 3.5). The nogalamycin PKS gene cluster of S. nogalater was recently sequenced (Ylihonko et ai., 1996) and also was shown to have a structure highly similar to that of the S. galilaeus aclacinomycin A PKS gene cluster (Tsukamato et ai, 1994; Fig. 3.5). This is important since nogalamycin biosynthesis is primed with an acetyl moiety, much the same as actinorbodin biosynthesis. More importantly, several strains known to produce anthracyclines using propionyl-CoA as starter also produce, or have been mutated to produce, anthracyclines with acetates used as starter molecules
(Fujiwara et al., 1981a; Fujiwara et al., 1981b; Hegyi and Gerber, 1968; Oki et al., 1981;
Soga et al., 1981). Included in these strains are S. galilaeus ATCC 31133 (Soga et al..
1981), for which part of the PKS gene cluster is known (Fig. 3.5), and a mutant of the baumycin and daimorubicin-producing S. coeruleorubidus ME 130-A4 (Oki et al.. 1981).
If, as predicted earlier, these strains should have a PKS gene arrangement similar to that in Streptomyces sp. strain C5 and S. peucetius (including dpsC and dpsD homologs), it is likely that a dpsC-Wke gene is nonfunctional in the mutant strain, resulting in the accumulation of methyl-side-chain anthracyclines (feudomycins). This
142 would provide a logical and scientifically supported explanation for how a mutation in these strains could generate the different side-chains. It is currently unknown, however, whether S. galilaeus ATCC 31133 or 5. coeruleorubidus ME 130-A4 PKS cluster have a dpsC homolog (Strohl et al., 1997).
Possible Involvement of other cyclases.
The dpsH gene product has recently been implicated as the second cyclase in polyketide biosynthesis (Hutchinson. 1997). Gerlitz et al. (1997) have shown that in the presence of dpsH gene product (Fig. 3.5), the minimal PKS genes expressed in heterologous hosts increased aklanonic acid production by 10-fold. Based on these results, Gerlitz et al. (1997) suggested that DpsH either catalyzes the cyclization of the second or the third ring in the growing polyketide chain, or has a positive effect on the
PKS cyclase DpsF. However, our data do not show any positive effects of DpsH on polyketide biosynthesis (Rajgarhia and Strohl, 1997). Moreover, disruption of dpsH in S. peucetius has demonstrated no effect on polyketide or for that matter anthracycline biosynthesis (C. R. Hutchinson, personal communication).
Model for synthesis of aklanonic acid and desmethylaklanonic acid.
Based on the currently available information, we propose the hypothetical biosynthesis of aklanonic acid from propionyl-CoA and malonyl-CoA in Figure 3.6. This pathway modified from the model proposed by Strohl and Connors (1992). is based on
143 Fig. 3.6. Hypothetical pathway for biosynthesis of aklanonic acid and desmethylaklanonic acid starting from acetyl/propionyl CoA and malonyl-CoA.
[Modified from the model proposed by Strohl and Connors (1992), based on data obtained by Rajgarhia and Strohl (1997), and influenced by results obtained using heterologous PKS functions as shown in other laboratories (Bibbet al., 1989; Shen and Hutchinson, 1993; Hutchinson and Fujii, 1995; Hutchinson, 1997), and aloesaponarin II biosynthesis by actinorbodin biosynthesis genes (Bartel et al., 1990;
McDaniel et al., 1993)].
144 Propionyl-SCoA
(Acetyl-SCoA) a b e f g
9x Malonyl-SCoA co^ | ab(e?) ^ oysAcm Q^S A % O^sAcP
4^ 2H„0 LA
0^S4ACP COOH COOH
HO + ACP-Bh
Fig. 3.6. Hypothetical pathway for biosynthesis of aklanonic acid and desmethylaklanonic acid starling from acetyl/propionyl CoA and malonyl-CoA. the data obtained by Bartel et al. (1990), Strohl et al. (1989), and Rajgarhia and Strohl
(1997), using as models both tetracenomycin F2 biosynthesis by minimal tetracenomycin
PKS genes (Hutchinson. 1995; Hutchinson, 1997; Bibb et al., 1989; Shen and
Hutchinson, 1993; Hutchinson and Fujii, 1995), and aloesaponarin 11 biosynthesis by
actinorbodin biosynthesis genes (Bartel et al., 1990; McDaniel et al., 1993).
Model for feudomycinone C and feudomycin D biosynthesis.
Feudomycinone C and feudomycin D detected in the fermentation broths of
Streptomyces sp. strain C5VR5 (dpsCD mutant) originates from desmethylaklanonic acid
(polyketide intermediate). The desmethylaklanonic acid would have to undergo reactions to form 11-hydroxyauramycinone (analogous to e-rhodomycinone), which would subsequently be glycosylated to auramycin D (analogous to rhodomycin D).
From our earlier work on conversion of rhodomycin D to doxorubicin (Dickens et al., 1997), we can safely assume that rhodomycin D would have to undergo hydrolysis of the CIO carbomethoxy group, followed by decarboxylation and subsequently 4-0- methylation to give feudomycin C (Fig. 3.7). The feudomycinone C (Fig. 3.7) we detected in the fermentation broths of Streptomyces sp. strain C5VR5 (dpsCD mutant) possibly is a deglycosylated product owing its origin to feudomycin C. The feudomycin
D (Fig. 3.7) we found in the fermentation broth is possibly produced as a result of the P-
450 monooxygenase (DoxA) which we predict is the catalyst responsible for adding a hydroxy group to position CIO of feudomycin C. Figure 3.7 is a schematic representation
146 Fig. 3.7. Proposed pathway for feudomycinone C and feudomycin D biosynthesis.
(Based on data obtained from this work; Rajgarhia and Strohl, 1997; Hoshino and
Fujiwara, 1983, and influenced by results obtained for the doxorubicin biosynthetic pathway by Dickens et al., 1997).
147 Acetyl-CoA -i- Malonyl-CoA
dpsA, dpsB. dpsG 9CO2 . CCACP
dpsE. dpsF, dauG
O ODH
Desmethylaklanonic Acid
OH O OH o o 1 1 O OH CGDCH»
'OH 11 - hydroxyauram ycinone
OH O OH OH
(iTtr
OH ^'*= TOP daunosamine
O OH CODCHï CH OH POOP Auramycinone D OH oJ 6 OHh 1^' •‘P x /
a. Decarboxymethylation b. 4-O-methvlation
O OH
Feudomycin C Degiycosyiation w
OŒiO OH Ig 1 0 (H i0 OH OH
"sit \ . ,r., Feudomycinone C DoxA (?)
O OH OH 1 .CH •OH
POOO'CKH, o OH j Feudomycin D sifc
Fig. 3.7. Proposed pathway for feudomycinone C and feudomycin D biosynthesis.
148 of the reactions leading to feudomycinone D and feudomycin C adapted from knowledge of the intermediates and pathway for doxorubicin biosynthesis and the work of Hoshino and Fujiwara (1981), who produced feudomycin D by feeding auramycin (molecules lacking the 11-hydroxy functionality) to mutants of 5". coeruleorubidus (ATCC 31276).
149 CHAPTER 4
PART A; SYNTHESIS OF RADIOLABELED ACETYL- AND PROPIONYL-COA;
PART B: IN VITRO AKLANONIC ACID SYNTHESIS FROM
RECOMBINANT S. lividans TK24 TRANSFORMED WITH PKS GENES FROM
Streptomyces sp. strain C5;
PART C: HETEROLOGOUS COMBINATION OF PKSs FROM
Streptomyces sp. strain C5 AND Streptomyces coelicolor A3(2) TO GIVE
ALOESAPONARIN II
PART A: SYNTHESIS OF RADIOLABELED ACETYL- AND PROPIONYL-COA
INTRODUCTION:
Propionyl coenzyme-A (propionyl-CoA) is a substrate for a variety of biological reactions. For example. propionyl-CoA is reductively carboxylated by Chromatium spp. to form a-ketobutyrate (Buchanan, 1969), organisms synthesizing fatty acids with an odd number of carbon atoms use propionyl-CoA as a starter unit (Voet and Voet, 1990), and certain antitumor antibiotics such as daimomycin, doxorubicin, and aclacinomycin A are assembled by the sequential decarboxylative condensation of nine malonyl-CoA units onto
150 an initial propionyl-CoA starter unit (Strohl et al„ 1989). Daunorubicin and doxorubicin biosynthesis similarly require propionyl-CoA and malonyl-CoA for the synthesis of the initial polyketide (Ye et ai, 1994; Grimm et al., 1994). Despite the importance of labeled propionyl-CoA there previously was no high-yield, cost-effective synthetic procedure for its production in labeled form on a micro scale. Moreover, radiolabeled propionyl-CoA is not available commercially unless custom synthesis of the substrate is ordered (Sigma; St.
Louis, MO). Unlabeled propionyl-CoA can be prepared from propionic anhydride and coenzyme-A by a method originally used for the production of succinyl-CoA (Marsh.
1988). Radiolabeled propionic anhydride, however, is not available, and this procedure would waste half of the label originally present on propionic anhydride. Enzymatic synthesis of long-chain fatty acyl-CoA derivatives has been reported (Komberg and Pricer.
1953). Similarly, an enzymatic exchange reaction has been used for the generation of radiolabeled propionyl-CoA from radiolabeled bicarbonate (Sokatch et al., 1968). [1-
'■*C]Acetyl-CoA has been synthesized from sodium [l-''*C]acetate and CoASH using 1.1'- carbonyldiimidazole to activate the acetate as a mixed anhydride; however, the reported yields for this process were only around 40% (Clugh et al., 1989). Attempting to improve the yield of the reaction by increasing the amount of coupling agent present leads to the production of iso-CoA derivatives which can be difficult to remove (Priestley, 1991).
A simple synthetic method for the production of acetyl-CoA from sodium acetate in near quantitative yield is described in Part A of this chapter. The same procedure has been adapted to synthesize propionyl-CoA from sodium propionate on a micromole scale in up to
151 75% yield and is also described in Part A of this chapter. This synthetic procedure has been adapted to synthesize [3.3,3-dJpropiony 1-CoA and d^-acetyl-CoA to be used later (Part B) in the analysis of polyketide biosynthesis.
MATERIALS AND METHODS:
Sodium [l-‘'*C]propionate and sodium [l-''*C]acetate were purchased from Dupont-
New England Nuclear (Boston, MA), while propionic-3,3,3-d^ acid and acetic-dj acid, were purchased from Isotech, Inc. (Miamisburg, OH). Coenzyme-A (reduced form) was obtained from the Sigma Chemical Company (St. Louis, MO). 1,1'-CarbonyIdiimidazole was obtained from Aldrich (Milwaukee, WT). Propionyl-CoA (unlabeled) was obtained either from Sigma or from the United States Biochemical Corporation (Cleveland, OH).
All other chemicals were of the highest grade available. Anhydrous, oxygen-free, tetrahydrofuran (THF) was distilled as needed from benzophenone and Na-K amalgam.
'Reactivials', for micromolar scale synthesis of these substrates, were purchased from Pierce
Scientific (Rockford, IL).
Preparation of the 1,1-carbonyldiimidazole.
1,1 -Carbonyldiimidazole was purified prior to use by recrystallization from anhydrous THF under argon, using Schlenk techniques (Perrin and Armarego, 1988), dried at 0.05 torr, and then stored under Argon at -20° C in small glass vials, which had been previously dried using acetone and baked in an oven at 70 °C for 90 minutes. The vials
152 were then sealed under an atmosphere of argon to prevent any moisture from effecting the
1,1 -carbonyldiimidazole.
Synthesis of [l-'^CjPropionyl-CoA and [3,3,3-dj]propionyl-CoA.
Sodium [I-''*C]propionate (1.85 MBq; 0.84 pmol) and propionic-3,3 J-d, acid (0.84
|imol) was dissolved in dry ethanol and transferred to a dry reacti vial containing sodium propionate (0.1 mg, 1.2 pmol) which was then sealed further with parafrlm. The ethanol was evaporated in a stream of dry argon, the residue further dried in vacuo, and the atmosphere in the flask replaced with dry argon. Dichlorobenzoic acid (0.38 mg, 2 jimol) in dry THF (0.05 mL) was added to the residue via a dry syringe. Additional dry THF (0.15 mL) was added and the mixture stirred at 65°C for 30 minutes. A white precipitate formed within 5 minutes. Purified 1.1 '-carbonyldiimidazole (0.65 mg, 4 pmol) in dry THF (0.05 mL) was added via syringe and the suspension stirred at 65°C for 90 minutes. The suspension was cooled to room temperature. Reduced coen 2wme-A (1.9 mg, 2.5 pmol) in aqueous imidazole buffer (0.2 mL; 0.5 M. pH 7.0) was added and the mixture was stirred at room temperature for 15 minutes. The crude reaction mixture was stored at - 20°C until purification and analysis. The reaction scheme for this synthesis is similar to that previously reported by Kawaguchi et al. (1981).
Propionyl-CoA prepared by the above method was purified by a modification of the procedure of Ingebretsen and Farstad (1980) in which the propionyl-CoA sample was loaded onto a C l8 Val-U-Pak analytical column (250 x 4.6 mm) and purified by isocratic
153 elution from a CIS Val-U-Pak analytical column (250 x 4.6 mm) using 17.5% v/v methanol
in potassium phosphate buffer (50 mM, pH 4.0) at a flow rate of 1.0 mL/min. The product was detected both by its radioactivity and by its UV absorption at 254 nm and compared against that of the standard propionyl-CoA. The peaks representing propionyl-CoA were collected and lyophilized. The lyophilized product was reconstituted in a minimal volume of water and applied to a PrepSep C8 column (Fisher Scientific; Pittsburgh. PA) previously equilibrated with 5% v/v aqueous methanol. The column was eluted with water (0.5 mL) to remove traces of buffer salts and then with 95% v/v aqueous methanol. The eluate was lyophilized and stored at -20°C.
Synthesis of [l-'^CJAcetyl-CoA and dj-acetyl-CoA.
[l-''*C]Acetyl-CoA and d^-acetyl-CoA were synthesized from sodium [l-''*C]acetate and acetic-dj acid, respectively, and CoASH by the method described for the synthesis of
[l-''*C]propionyl-CoA.
RESULTS AND DISCUSSION:
Propionyl-CoA was synthesized in 68.5% yield (± 5.1 SD; n = 6; Table 4.1 A). The adequate purification of the 1.1 '-carbonyldiimidazole was found to be absolutely necessary for the reproducibility of the procedure. Similarly, acetyl-CoA was synthesized in near
90% yield (± 4.5 SD; n=6; Table 4.IB). Yields of acyl-CoA samples were between 2 and
45% when unpurified, commercial samples of 1,1 '-carbonyldiimidazole
154 Table 4.1(A) Synthesis of propionyl-CoA. The table shows the yield of propionyl-CoA obtained in six separate attempts. (B) Synthesis of acetyl-CoA. The table shows the yield of acetyl-CoA obtained in six separate attempts.
155 4.1(A). Synthesis of propionyl-CoA* Yield (%)
1 63.4 2 65.0 3 73.0 4 72.5 5 64.0 6 74.5
' Results of the several attempts at synthesis with cold propionate prior to the actual synthesis with radiolabeled starting product.
4.1(B). Synthesis of acetyl-CoA* Yield (%)
1 87.0 2 94.5 3 93.0 4 92.5 5 82.5 6 90.2
“ Results of the several attempts at synthesis with cold acetate prior to the actual synthesis with radiolabeled starting product.
Table 4.1.(A) Synthesis of propionyl-CoA. (B) Synthesis of acetyl-CoA.
156 was used. Indeed, HPLC analysis of these latter synthesized samples showed a range of
contaminant species to be present, some of which were also present in commercial samples
of propionyl-CoA.
The identity and purity of the acyl-CoA samples were established by several
techniques. The UV spectra, in particular the ratio of absorbances at 232 and 260 nm
(Appendix C), for the synthesized compounds were identical to that of the authentic
materials. HPLC analysis confirmed that the radiolabeled acyl-CoA derivatives were the
only radiolabeled compounds present in the samples: the synthesized compounds co-eluted
with authentic standards (Fig. 4.1 ). Conclusive evidence for the purity and identity of the
synthesized compounds comes from their ‘H NMR spectra (Fig. 4.2). The chemical purity of the samples can be estimated as being greater than 95%. In particular, no resonances associated with either propionyl-wo-CoA. contaminants which can be inhibitors in biological reactions, can be seen in these 'H NMR spectra (Fig. 4.2).
Synthesized [l-‘‘‘C]propionyl-CoA was incubated with cell extracts of Rhodospirillwn rubrum known to contain propionyl-CoA carboxylase (Wurtele and Nikolau. 1990) using standard assay conditions (Olsen and Merrick, 1968). MethyImalony 1-CoA was shown to be the sole radioactive product of the reaction by HPLC analysis of the incubation mixture.
Synthesized samples of propionyl-CoA were turned over by the cell-free system with efficiencies comparable to commercial samples of propionyl-CoA. Similarly, synthesized acetyl-CoA was analyzed enzymatically by standard procedures (Knight, 1962; Olsen and
Merrick, 1968) using cell extracts of R. rubrum. In the latter case.
157 Fig. 4.1. Reverse phase HPLC chromatograph of synthetic [l-"C]prop:onyl-CoA and
[l-"C]acetyl-CoA. The solid line represents absorbance at 254 nm. The dashed line represents radioactivity (cpm), indicating elution of radioactive compounds. CoASH
(a) elutes at 3.7 min. (A) Propionyl-CoA (b) elutes at 17.5 min. (B) Acetyl-CoA (c) elutes at 6.5 min.
158 0.8 10 0.6 12
0.5 0.5 10 if 0.4 0.4 8 O 6 O X X 0.3 |{? 0.3
6 0.2 Pi 0.2 4 Ü Ü
0.0 0.0
5 10 15 200 0 5 10 15 20 Time (min) Time (min)
Fig. 4.1. Reverse phase HPLC chromatograph of synthetic [l-'^CJpropionyl-CoA and [l-'^C]acetyl-CoA. Fig. 4.2. ‘H NMR analysis of synthesized acetyl-CoA (A, 1.2 mg.mi ') and propionyl-
CoA (B, 0.9 mg mL ') in H%0 at 298 K. ‘H Spectra were run on an IBM AF-270 spectrometer at 270.133 MHz. The spectra are referenced to the residual 'HO H signal at 4.78 p.p.m. The majority of the HO H signal was removed by acquiring the spectra using a presaturation sequence. FID's were transformed using a slight resolution enhancement (LB -0.2, GB 0.05). (NMR was run and interpreted by Dr.
Nigel D. Priestley)
160 il
I
• 5 9 0 IS to 75 7 0 6 5 6.0 1 5 1 0 PM#4 5 4 0 15 3 0 2.5 2.0 t 5 1 0 0.5 0
9 5 9 0IS 75 70 IS 1 0 5 5 4 0 3530 2 5 2 0 0.5 10
Fig. 4.2. 'H NMR analysis of synthesized acetyi-CoA (A, 1.2 mg.ml ') and propionyl-CoA (B. 0.9 mg.mL ') in 'HiO at 298 K.
161 malonyl-CoA was shown to be the sole radioactive product of the reaction. The synthesized [3, 3, 3-dJpropionyl-CoA and [2, 2, 2-dJacetyi-CoA was used subsequently in detecting in vitro akianonic acid biosynthesis described in Part B of this chapter.
162 PART B: IN VITRO SYNTHESIS OF AKLANONIC ACID AND SEK 43 BY THE
Streptomyces sp. strain C5 PKS
INTRODUCTION:
The Streptomyces sp. strain C5 polyketide synthase (PKS; Chapter 2; Rajgarhia and Strohl. 1997) is comprised of a group of enzymes that together catalyze the biosynthesis of polyketide products from activated carboxylic acids (propionyl-CoA and malonyl-CoA). The resulting polyketide chain is modified by reduction, cyclization, and hydroxylation to yield akianonic acid (Strohl et al., 1997). These reactions are catalyzed by several monofunctional (some possibly bi-functional) proteins that make up PKS. and are expressed by the PKS genes (Ye et al., 1994; Grimm et al., 1994). The possible reactions and the gene products catalyzing reactions leading to akianonic acid have been previously discussed (Chapter 3). In the absence of any available intermediates in the reactions leading up to akianonic acid, it is assumed that the growing polyketide chain is attached to the PKS. and is only released upon being modified to a stable intermediate such as akianonic acid in Streptomyces sp. strain C5. The reactions leading up to akianonic acid, therefore, are at best hypothetical since none of the intermediates in the pathway has ever been isolated. Unraveling the role played by each of the enzymes of the
PKS involved in the modifications leading to akianonic acid is essential to a better understanding of the biosynthetic pathway for daunorubicin and doxorubicin.
Additionally, such studies will allow hybrid PKS experiments where individual
163 components of the PKS from several Streptomyces spp., could be combined productively, to give unique hybrid products.
In an effort to better understand the reactions involved in forming polyketide intermediates, in vitro biosynthetic studies involving individual PKS proteins have been attempted by several groups (Shen and Hutchinson. 1993; Carreras et al., 1996; Carreras and Khosla, 1998). Recent works of Shen and Hutchinson (1993. 1996) have described the first cell free system for aromatic polyketide biosynthesis. Using tetracenomycin
(tcm) biosynthesis as a model system, Shen and Hutchinson have characterized the roles of some of the polyketide synthase gene products of wild type Streptomyces glaucescens which produce the antibiotic tetracenomycin (Shen and Hutchinson, 1996). In their studies, the tcm PKS genes were over-expressed in S. glaucescens, and cell free extracts from the recombinant culture was incubated with radiolabeled acetyl-CoA and malonyl-
CoA. to produce intermediates of the biosynthetic pathway. By deleting certain gene products from this in vitro experiment, and by analyzing the products formed or lack of product thereof, the possible roles of some of the tcm PKS components were deciphered.
Figure 4.3.(A) is a representation of the genes encoding the proteins shown required to produce tetracenomycin F2 (Hutchinson, 1997).
Subsequently, in a similar study, Carreras et al. (1996) used a similar cell free
PKS system to analyze actinorhodin biosynthesis by S. coelicolor, and confirmed the functions of some of the PKS gene products. In a recent experiment, Carreras and Khosla
(1998) have shown that purified ketoacylsynthase a (encoded by S. coelicolor actl-orfi).
164 Fig. 4 J. (A). Simplified pathway for tetracenomycin biosynthesis. Genes encoding the
PKS proteins required to produce tetracenomycin F2. (B). Structure of SEK43 produced from 10 acetates along with the molecular formula and expected molecular mass.
165 tcmK tcmL tcmM tcmN tcmJ TCM F2 TCM FI
Acatyt-CoA
9 acetyl units COOH COOH OH OH OH OH O OH CH)
OH
SEK43 CzoHiaOy 368.34 368.089603 CH, C 65.22% H 4.38% O 30.41%
OH
Fig. 4.3. (A). Simplified pathway for tetracenomycin biosynthesis. Genes encoding the PKS proteins required to produce tetracenomycin F2. (B). Structure of SEK43 produced fi-om 10 acetates.
166 ketoacylsynthase (3 (encoded by S. coelicolor actl-orfll), the S. coelicolor PKS AC?, and
fatty acid synthase malonyl-CoA:ACP acyl-transferase (MAT), together, were able to produce polyketide intermediates. This result is significant in that it shows that ketoacylsynthase a and ketoacylsynthase P form a complex while catalyzing polyketide biosynthesis. This experiment also identified the involvement of enzymes from primary metabolism (MAT) in polyketide biosynthesis.
A cell-free system for the synthesis of akianonic acid derived from recombinant S. lividans TK24, transformed with Streptomyces sp. strain C5 PKS genes, is described in this (Part B) chapter. Radiolabeled propionyl-CoA, synthesized as described in Part A of this chapter, has been used to produce akianonic acid in vitro by the PKSs from this recombinant S. lividans TK24. The detection of akianonic acid produced, and an analysis of the results of the in vitro experiment is also reported in this (Part B) chapter.
MATERIALS AND METHODS:
Culture growth.
Seed cultures of wild type and recombinant S. lividans TK24, containing pANT782 or pANT785, were grown in 50 mL of SC {Streptomyces-Coxmexs) medium
(Chapter 3), supplemented with 3 x MOPS in 500 mL flasks with spring coils at 30“C for
48 h. Thiostrepton (25 mg/mL) was added for plasmid selection. The extra MOPS allowed the medium to remain buffered at pH 7.2, even when significant amounts of akianonic acid was made. After 48 h, 10 mL of this culture was used to inoculate 200 mL
167 of s c medium supplemented with 3 x MOPS contained in 1 L flasks with spring coils.
Thiostrepton (25 mg/mL) was added for maintaining the plasmid selection. This culture
was incubated at 30“C for 36 h, harvested by centrifugation for 10 min at 10,000 x g in a
Sorval centrifuge (Newtown, Connecticut), and the cells were washed with 150 mM
phosphate buffer, pH 7.1. The cells were harvested after the washing by centrifugation
for 10 min at 10,000 x g and the pellet was stored at 4°C until further treatment.
Preparation of the ceil free PKSs preparation.
Two grams (wet weight) of the recombinant S. lividans TK24 cells were resuspended in 10 mL of buffer A (150 mM phosphate buffer, pH 7.1; 15% glycerol; 2 mM DTT; 2 mM EDTA) after which the cells were lysed by passaging through a French pressure cell (American Instrument Co.; Urbana, IL) at 15,000 p.s.i.. The lysed cellular material was then clarified by centrifugation at 10,000 x g for 30 min at 4“C. The
Supematent obtained was decanted into a chilled beaker and brought to 80% saturation with ammonium sulphate [(NH^)iSO^]. The ammonium-sulphate saturated supematent was centrifuged at 10,000 x g for 30 min at 4“C. The cell pellet obtained was separated from the supematent and resuspended in 5 mL of buffer A, and desalted on a Sephade.x
G-25 column at 4°C to yield the cell-free PKS preparation. This preparation was maintained at 4°C until use. The total protein content of the PKS preparation was determined by the Bradford assay (Bradford, 1976), and contained 15 mg/mL of protein.
168 In vitro akianonic acid biosynthesis.
Several 2 mL micro-centrifuge tubes were prepared for the in vitro synthesis of akianonic acid. The main reaction tube contained 2 mg of cell free PKS preparation
(quantified by Bradford Assay) obtained from S. lividans TK24 transformed with pANT782 (all PKS genes including dpsC and dpsD), 250 p,M malonyl-CoA, 10 pM
[3,3,3-d]]propiony 1-CoA (made as described in Part A of this chapter), 500 pM NADPH and buffer A to make a total volume of 500 pL. Control tubes had either unlabeled propionyl-CoA instead of the labeled propionyl-CoA, or one of the other components of the reaction mixture omitted as described in Table 4.2. In each reaction, the cell free PKS preparation was added only after the rest of the mixture had been pre-incubated for 30 min at 30°C. After the addition of the cell-free PKS extract, the tubes were again incubated for 1 h at 30“C and the reaction was stopped by addition of 50 pL of concentrated HCl to each tube. Chloroform (500 pL) was added to the tubes and after shaking the mixture, the tubes were centrifuged at 5,000 x g for 1 min. The chloroform layer was removed into fresh tubes. This chloroform extraction was repeated twice and the pooled chloroform layers were evaporated to dryness in vacuo. The material was resuspended in 50 pL methanol and analyzed by LC/MS.
Similar reactions were performed with cell-free PKS preparation obtained from S. lividans TK24 transformed with pANT788 (all PKS genes except dpsD) and pANT785
(all PKS genes except dpsC and dpsD), as listed in Table 4.2.
169 Table 4.2. Analysis of the products from in vitro reactions with Streptomyces sp. strain C5 PKS preparation incubated with [S^^-dajpropionyl-CoA and malonyl-
CoA.
170 In vitro reaction with |3,3,3-dj|propionyl-CoA Product determined by T L C , H PL C
1 • PKSs from S. lividans TK24 (pANT782) incubated with *: [3,3,3-dj]propionyl-CoA + malonyl-CoA + NADPH akianonic acid [3,3,3-dj]propionyl-CoA + malonyl-CoA - [3,3,3-d3]propionyl-CoA + NADPH - malonyl-CoA + NADPH -
2. PKS from S. lividans TK24 (pANT788) incubated with *: [3,3>3-d3]propionyl-CoA + malonyl-CoA + NADPH -
3. PKS from S. lividans TK24 (pANT785) incubated with *: [3,3,3-d3]propionyl-CoA + malonyl-CoA + NADPH -
" pANT782 includes all Streptomyces sp. strain C5 PKS genes (dpsA, dpsB, dpsC, dpsD, dauG, dpsE, dpsF)\ pANT788 includes all Streptomyces sp. strain C5 PKS genes except dpsD-, pANT785 includes all Streptomyces sp. strain C5 PKS genes except dpsC and dpsD. Akianonic acid was analyzed and confirmed by LC/MS as described in materials and methods.
Table 4.2. Analysis of the products from the in vitro reactions with Streptomyces sp. strain C5 PKS preparations incubated with [3,3,3-d]]propionyl-CoA and malonyl-CoA. Analysis of the akianonic acid biosynthesized:
Authentic akianonic acid obtained from K. Ekhardt was used as standard for TLC and HPLC analysis of the products of the in vitro akianonic acid biosynthesis. For
LC/MS analysis, the dj-aklanonic acid standard was generated by feeding cultures of
Streptomyces sp. strain C5-69, a dauC mutant that accumulates akianonic acid (Bartel ei al., 1989), with propionic- 3,3,3-d3-acid. Akianonic acid was isolated from the culture by adjusting the pH to 1.5 using concentrated HCl, and extracting with equal volumes of chloroform. The chloroform extract was then dried in vacuo, reconstituted in methanol and chromatographed on pre-coated silica gel TLC plates (Merck KGaA. Darmstadt,
Germany) to isolate akianonic acid.
The material obtained by the in vitro reaction was analysed by LC/MS (Liquid
Chromatography/Mass Spectrometry) on a API 300 (Perkin Elmer; Norwalk.
Connecticut) MS instrument, interfaced with a HPLC system and a UV detector
(Shimadzu Scientific Instruments; Columbia, Maryland), at the Ohio State Campus
Chemical Instrumentation Center (Columbus, Ohio).
In vitro biosynthesis of polyketides derived from acetates (e.g., SEK 43 or desmethylaklanonic acid).
Several 2 mL centrifuge tubes were prepared for the in vitro synthesis of polyketides derived from 10 acetates [Chapter 3; Fig. 4.3.(B)j. The main reaction tube contained 2 mg of cell free PKS preparation (quantified by Bradford Assay) obtained
172 from 5. lividans TK24 transformed with pANT785, 250 pM malonyl-CoA, 10 pM d,- acetyl-CoA (made as described in Part A of this chapter), 500 pM NADPH and buffer A to make a total volume of 500 pL. Control tubes had either unlabeled acetyl-CoA or one of the other components of the reaction missing and is described in Table 4.3. In each reaction, the cell free PKS preparation was added only after the rest of the mixture had been pre-incubated for 30 min at 30“C. After the addition of the cell-free PKSs extract, the tubes were again incubated for 1 h at 30°C and the reaction was stopped by addition of 50 pL of concentrated HCl to each tube. Chloroform (500 pi) was added to the tubes and after shaking the mixture, the tubes were centrifuged at 5,000 ;c g for 1 min. The chloroform layer was removed into fresh tubes. The chloroform extraction was repeated and the pooled organic layer was dried. The material was resuspended in 50 pL methanol and analyzed by LC/MS and MS/MS analysis.
Similar reactions were performed with cell-free PKS preparations obtained from
5. lividans TK24 transformed with pANT782 (all PKS genes) and pANT788 (all PKS genes, except dpsD), as listed in Table 4.3.
Analysis of the products.
SEK 43 was isolated from S. lividans transformed with a plasmid encoding tetracenomycin PKS genes known to produce SEK43 was obtained from Dr. Richard
Hutchinson, University of Wisconsin. SEK43 obtained from this strain was confirmed by
LC/MS and was used as a standard to detect the products of in vitro biosynthesis using
173 Table 4.3. Analysis of the products from in vitro reactions with Streptomyces sp. strain C5 PKS preparation incubated with [Z^^l-d^jacetyl-CoA and malonyi-CoA.
174 In vitro reaction with dj-acetyl-CoA Product determined by TLC, HPLC and LC/MS ^
1. PKSs from S. lividans TK24 (pANT782) incubated with [2,2,2-dj]acetyl-CoA + malonyl-CoA + NADPH -
2. PKSs from S. lividans TK24 (pANT788) incubated with [2,2,2-d,]acetyl-CoA + malonyl-CoA + NADPH -
3. PKSs from S. lividans TK24 (pANT785) incubated with *: [2,2,2-dj]acetyl-CoA + malonyl-CoA + NADPH SEK43 [2,2,2-d,]acetyl-CoA + malonyl-CoA - Ui [2,2,2-d)]acetyl-CoA + NADPH -
malonyl-CoA + NADPH -
* pANT782 includes all Streptomyces sp. strain C5 PKS genes (dpsA, dpsB, dpsC, dpsD, dauG, dpsE, dpsF)\ pANT788 includes all Streptomyces sp. strain C5 PKS genes except dpsD; pANT785 includes all Streptomyces sp. strain C5 PKS genes except dpsC and dpsD. SEK43 was analyzed and confirmed by LC/MS and MS/MS as described in materials and methods.
Table 4.3. Analysis of the products from in vitro reactions with Streptomyces sp. strain C5 PKS preparation incubated with [2,2,2-d,jacetyl-CoA and malonyl-CoA. PKS from 5. lividans transformed with pANT785. SEK 43 produced in the in vitro
reaction was expected to co-migrate on silica gel TLC plates with the authentic material,
and was expected to have a M+H of 369 (SEK 43 produced using dj-acetyl-CoA would
have a M+H of 372). Desmethylaklanonic acid, isolated earlier (Chapter 3), also was used as a standard in determining the products of the reaction.
The material obtained from the in vitro reactions was analyzed by LC/MS and
MS/MS (Mass Spectrometry) on a API-300 MS instrument (Perkin Elmer; Norwalk.
Connecticut), interfaced with a HPLC system and a UV detector (Shimadzu Scientific
Instruments; Columbia, Maryland), at the Ohio State Campus Chemical Instrumentation
Center (Columbus. Ohio).
RESULTS AND DISCUSSION:
In vitro akianonic acid biosynthesis from PKSs obtained from S. lividans TK24
(PANT782), S. lividans TK24 (pANT788), and S. lividans TK24 (pANT785).
TLC, HPLC and LC/MS analyses (Appendix C) of the product obtained from the in vitro reactions involving PKS from S. lividans TK24 (pANT782) indicate that akianonic acid is made when propionyl-CoA and malonyl-CoA are present. The akianonic acid co-spotted with authentic akianonic acid (Eckhardt et al., 1985) on TLC had an identical Kf. When either malonyl-CoA, propionyl-CoA, or NADPH was excluded (Table 4.2), no akianonic acid product was detected. Using [3,3,3-dJpropionyl-
CoA, a LC/MS peak eluting at 6.55 min gave a M+H of 400 (Appendix C). That
176 corresponded to the M+H obtained for authentic d^-aklanonic acid that was made using
Streptomyces sp. strain C5-A64, as described in materials and methods.
These results agreed with the results obtained from genetic analysis, wherein the
PKS genes from Streptomyces sp. strain C5, when expressed in heterologous hosts, produced aklanonic acid (Rajgarhia and Strohl. 1997). What was significant was that
PKS preparations from recombinant S. lividans TK24 functioned after isolation from the cellular material. (NHJiSO^ fractionation, and desalting on a sephadex G-25 column. In a similar experiment, but with a buffer that did not contain any EDTA, identical results were obtained, indicating that EDTA did not have any effect on product formation in vitro.
Similarly, analysis of the product from reactions involving PKS from S. lividans
TK24 (pANT788) [all PKS genes except dpsD], propionyl-CoA and malonyl-CoA produced no yellow chromophoric-aklanonic acid like products. This result, although in disagreement with the genetic evidence obtained earlier (Chapter 2. 3). is not entirely unexpected. It is probable that in the absence of dpsD gene product, the PKS might be unstable and hence might not be able to prime polyketide biosynthesis. Under the milder in vivo environment, the absence of dpsD might be compensated by other cellular components leading to a stable and functional PKS. It is also possible that other acyltransferases borrowed from either primary or secondary metabolism might function in vivo in the absence of dpsD in polyketide biosynthesis. In support of this possibility is
177 the recent discovery by Carreras and Khosla (1998) that malonyl-CoA:ACP acyl transferase is borrowed by polyketide synthases from fatty acid synthases.
In vitro reactions with PKS obtained from S. lividans TK24 (pANT785) [all PKS genes except dpsC and dpsD] did not produce aklanonic acid. This result suggests that in the absence of the dpsC and dpsD gene products, the aberrant PKSs does not utilize propionate primed polyketide biosynthesis. Under a favorable in vivo condition, either an alternative propionyl-CoAiACP acyltransferase functions in the absence of the putative dpsD gene product, or the PKS formed is more stable and able to utilize propionate and acetate in priming polyketide biosynthesis.
In vitro SEK43 biosynthesis from PKS obtained from S. lividans TK24 (pANT782),
S. lividans TK24 (pANT788), and S. lividans TK24 (pANT785).
TLC and HPLC analyses of the product obtained from the in vitro reactions involving PKSs from S. lividans TK24 (pANT782) [all PKS genes], and S. lividans TK24
(pANT788) [all PKS genes, except dpsD], indicate no product is made when d-acetyl-
CoA, malonyl-CoA and NADPH is present in the reaction mixture. The lack of product formation is an expected result and agrees with the results from the genetic analysis
(Chapter 3), that when the dpsC gene product is present, the PKS from Streptomyces sp. strain C5 failed to produce polyketide intermediates initiated with acetate (Chapter 3).
Significantly, the PKS from S. lividans TK24 (pANT785), incubated with acetyl-CoA. malonyl-CoA, and NADPH produced SEK43 (made from 10 acetates). SEK43 is an
178 aberrantly cyclized polyketide intermediate derived from 10 acetates [Fig. 4.3.(B)].
SEK43 was compared to the authentic SEK43 product obtained from Dr. Richard
Hutchinson (University of Wisconsin). This result is significant since Gerlitz et al. (1997) have recently shown that the Streptomyces peucetius minimal PKS (PKS genes except dpsC and dpsD), expressed in S. lividans TK24, produced SEK43.
Table 4.2 and Table 4.3 show the summary of the results of the in vitro reactions involving the PKS. It is evident that under in vitro conditions, the PKS preparations obtained from S. lividans TK24 (pANT785) are probably unstable as a result of ±e exclusion of the dpsC and dpsD gene products. This apparant instability yields SEK43 from 10 acetates, an aberrantly cyclized polyketide intermediate. These data not only confirm the genetic evidence, that in the absence of dpsC gene product, acetates is used as a starter to prime polyketide biosynthesis, but also support the theory that the deletion of dpsD and dpsC leads to instability in the PKS formed under in vitro conditions.
179 PART C: HYBRID POLYKETIDE PRODUCTION BY EXPRESSING 5.
COELICOLOR ACTINORHODIN PRODUCING GENES IN Streptomyces sp. strain C5
PKS MUTANTS.
INTRODUCTION:
Streptomyces coelicolor produces the benzoisochromanequinone (polyketide) antibiotic, actinorhodin. The biosynthetic pathway for the production of these antibiotics has been well characterized using a series of genetic and chemical tools (Fig. 4.4:
Malpartida and Hopwood, 1986; Bartel et aL, 1990). The study of the actinorhodin biosynthetic pathway has furthered the study of other Streptomyces spp. antibiotic biosynthetic routes (Malpartida et al., 1987). Additionally. S’, coelicolor actinorhodin biosynthetic genes have been used in demonstrating hybrid antibiotic production.
Streptomyces violaceoruber Tii22, transformed with the actinorhodin biosynthetic genes from S. coelicolor produced dihydrogranatirhodin, a novel antibiotic, not produced in either the donor strain (S’, coelicolor) or the recipient strain (S’, violaceoruber Tii22)
(Hopwood et al.. 1985). This example is the earliest demonstration of hybrid molecules produced as a result of interactions between the biosynthetic genes from two different
Streptomyces spp.
Bartel et al. (1990) were the first to characterize the role of the polyketide synthase genes in generating hybrid molecules. S. galilaeus strains (ATCC 31133) and
(ATCC 31671) were transformed by Bartel et al (1990). with S. coelicolor actinorhodin
1 8 0 Fig. 4.4. Actinorhodin biosynthetic pathway (Bartel et a/., 1990b). The gene products are designated to depict their catalytic role in the pathway.
181 Acetyi-CoA Starter Unit -t- 7 Malonyl-CoA
Polyketide synthase :OOH
Polyketide reductase (aa/II) COOH
(Polyketide synthase) | COOH OK
Polyketide cyclase/dehydratase COOH
OH
OH o Polyketide cyclase/dehydratase I {actIV) COOH
OH o act VI r- (acr V7-minus I >2 conditions) COOH
OH o OH O CH, actVa/actVb COOH
OH o 2 OH O
ALOESAPONARIN D ACTINORHODIN
Fig. 4.4. Actinorhodin biosynthetic pathway (Bartel et al., 1990b).
182 PKS genes, resulting in the production of a novel antibiotic aloesaponarin II and its derivative, desoxyerythrolaccin, in the two strains, respectively. Subsequently, a series of different gene constructs o f S. coelicolor PKS region was introduced by transforming S. galilaeus strains (Bartel et al.. 1990). The results suggested that the novel metabolites were produced as a result of interactions among the gene products of actl-orfl
(ketoacylsynthase a) and actl-orfll (ketoacylsynthase P) from S. coelicolor, and the antibiotic specific acyl carrier protein (ACP). polyketide cyclase and reductase from S. galilaeus strains in the absence of an actVI gene product. Figure 4.5 is a representation of the possible interacting proteins (gene products) that are required to make aloesaponarin
II in S. galilaeus strains. The early work of Bartel et al. (1990) has received a significant boost in recent years by C. Khosla and his collègues, who have added a wealth of knowledge by producing additional hybrid molecules by generating similar interactions among the PKS gene products from several other Streptomyces spp. (McDaniel et al..
1995).
Streptomyces sp. strain C5 (daunorubicin producer) and S. galilaeus (aclarubicin producer) share similarity in that, both these strains, produce the final antibiotic products via a common intermediates aklanonic acid and aklavinone. These intermediates are derived from the propionyl-CoA and C2 units from malonyl-CoA (discussed earlier), by reactions catalyzed by the PKS gene products. Hence, it thought to be quite likely that the PKS genes in these strains would be rather similar and share a great deal of homology. Since S. galilaeus PKS gene products were able to productively interact with
183 Fig. 4.5. Model depicting the heterologous combination between the PKS components
(depicted by the genes) from S. galilaeus and S. coelicolor. Aklanonic acid and aloesaponarin II were the products made indicating a productive combination between the PKS components as shown.
184 aknX aknD /I______I___ ( aknE \ aknA aknB aknC \\ \| actl-1) actl-2
AknD COOCH,
OH O OH OH Aklanonic acid
AknD OH 0 9^3 AknA Act-2
Aloesaponarin II
Fig. 4.5. Model depicting the heterologous combination between the PKS components (depicted by the genes) from S. galilaeus and S. coelicolor.
185 s. coelicolor PKS gene products, in making aloesaponarin II, the Streptomyces sp. strain
C5 PKS genes also were expected to show similar results. But when Streptomyces sp.
strain C5 was transformed with actl-orfl and actl-orfll genes from S. coelicolor. no
aloesaponarin II was detected (P. Bartel. Y. Li., and W. R. Strohl, unpublished data).
Before the structure of Streptomyces sp. strain C5 PKS was known this result was confounding, and was laid aside. Later, after the PKS structure was determined, this
failure was attributed to a possible hinderance in productive interactions among PKS
proteins, due to the presence of dpsC and dpsD gene products that are unique to the
Streptomyces sp. strain C5 PKS cluster. Although a detailed analysis of the S. galilaeus
PKS cluster is pending, dpsC and dpsD thus far have not been reported in that strains.
Attempts at showing hybrid interaction with mutants of Streptomyces sp. strain
C5. in which the dpsC and dpsD genes have been deleted, are described in this (part C) chapter. In addition, similar experiments to generate hybrid molecules using a C5 strain in which dpsA. dpsB. dpsC. and dpsD (entire PKS backbone) have all been deleted are also described in this part of the chapter.
MATERIAL AND METHODS:
Bacterial strains and growth conditions.
Streptomyces sp. strain C5. originally obtained from the Frederick Cancer
Research Center (McGuire et al., 1980), has been described in detail elsewhere (Bartel et ai, 1990). Streptomyces sp. strain C5VR5 {dpsCD mutant) has been described earlier
186 (Chapter 2; Rajgarhia and Strohl, 1997). Streptomyces lividans TK24 (Hopwood et al..
1985) was obtained from D. A. Hopwood. Streptomyces galilaeus (ATCC 31133) was obtained from the American Type Culture Collection (Manassus. VA). The strains used in this study are further described in Appendix A. For plasmid isolation, recombinant S. lividans TK24 strains were grown in YEME medium (Hopwood et ai. 1985) containing
50 pg of thiostrepton (Sigma; St. Louis, MO) per ml as required. Recombinant
Streptomyces sp. strain C5 and mutants derived from it, were grown GPS complex medium. GPS medium contains, per IL: glucose (30 g), Proflo (10 g; Buckeye Cellulose
Corp.; Memphis, TN), NaCl (3 g), CaCOj (3 g) as described previously (Dekleva and
Strohl, 1987). If required for selection, thiostrepton was added at a concentration of 10 pg/ml. All strains were routinely maintained on R2YE solid medium. When required for selective pressure, thiostrepton was added to R2YE medium 50 pg/ml. respectively.
General genetic manipulations.
Procedures for protoplast formation, transformation, and regeneration of protoplasts for Streptomyces sp. strain C5 and mutants derived from it have been described elsewhere (Lampel and Strohl, 1986). Streptomyces galilaeus (ATCC 31133) and S. lividans TK24 were transformed by methods described by Hopwood et al. (1985).
Plasmids pANT849 (SnpR-activated snpA promoter controls gene expression;
Appendices B and C), and pANT817 (pANT849 with actl-orfl and actl-orfll expression controlled by the SnpA snpR; Appendices B and C) were obtained from Charles DeSanti
187 (The Ohio State University). These plasmids were amplified by introducing into S. lividans TK24 and subsequently were introduced into Streptomyces sp. strain C5 and mutants using procedures described in Hopwood et al. (1985).
Plasmids from 5". lividans TK24 were routinely prepared according to Hopwood ei al. (1985). Plasmids used in this work are given in Appendices B and C. For probes derived from plasmids, the DNA inserts were obtained from plasmid DNA by digesting with the appropriate restriction endonucleases, followed by size fractionation using gel electrophoresis (Sambrook et al.. 1989), and purified from gel slices by electroelution of the DNA using the procedure detailed in Chapter 2. The DNA probes were then labeled by the "‘P-random primer procedure (Feinberg and Fogelstein, 1983), using 50 mCi of [a-
'■P]dCTP per mg of DNA and random primer labeling kit (Stratagene; La Jolla.
California) using manufacturers specifications.
Streptomyces sp. strain C5 and Streptomyces sp. strain C5VR10 (described later) chromosomal DNA was extracted and purified using CsCU density gradients as described by Hopwood et al. (1985). After digesting the chromosomal DNA using restriction enzymes according to manufacturers specifications, the DNA was size ffactioned on 1.0
% agarose gels (Sambrook et al.. 1989). The size ffactioned DNA was transferred to
Nytran nylon membranes (Schleicher & Schuell Inc.; Keene, New Hampshire) using standard procedures detailed in Sambrook et al. (1989). Hybridization of the previously labeled probes to the immobilized DNA was performed using the Southern blot method
(Sambrook er a/., 1989).
188 Gene disruption methods.
For dpsABCD gene replacement, the 5.905-kbp Xhol-EcoBl fragment containing entire dpsE and its putative promoter region, all the way to the 3’ end of dauC gene from pANTI21 (Ye et al., 1994) was cloned into Xho\-Eco^ digested pANT841 (Appendix
A). The resultant plasmid, pANT735. was used to make the subsequent replacement vector used in this work.
To construct the plasmid in which dps ABCD were replaced with aphl. encoding neomycin phosphotransferase from plJ61 (Thompson et al., 1982), the 2.553-kbp BamHl fragment of pANT735 was replaced with a 1.40-kbp BgBl fragment containing aphl from pKK840 (Kwak. 1995) to generate pANTl 106 (Fig. 4.6; Appendices B and C). This plasmid, in which the aphl gene was flanked to the left and right with 1.75- and 1.67-kbp fragments of homologous Streptomyces sp. strain C5 DNA. respectively, was used to transform protoplasts o f Streptomyces sp. strain C5 as previously described (Lampel and
Strohl. 1986). After 18 h, the regenerated protoplasts were challenged with 10 mg/ml neomycin, and neomycin-resistant transformants were picked over to fresh plates after 7 days.
Detection of aloesaponarin II biosynthesis.
Recombinant Streptomyces sp. strain C5 and recombinant mutant strains C5VR5
(dpsCD null mutant), and C5-VR10 (dpsABCD null mutant), were each transformed with pANT849 and pANT817, and grown for 6 days at 30°C in 50 ml of GPS medium in 250-
189 Fig. 4.6. (A) Southern blot of genomic DNA of strainsStreptomyces sp. strain C5
(lanes 1-3) and Streptomyces sp. strain CSVRIO (lanes 4-6) digested with Bcli, Pstl^ and BamW., respectively, showing replacement of dpsA through dpsD in the chromosome of Streptomyces sp. strain C5 with aphl. The Mlul-BamHl fragment shown in panel B was used as the probe. (B) Restriction maps of the plasmid, parental chromosome and mutant chromosome, showing the expected results obtained by the double crossover.
190 A
B
Prohc Psi
üph X X Bel C5 Chromosome IJcl I*'' l> B ! ^s.
Jp\A JpsB JpsC JpsD
------8 .6 k b p 1 0 3 k b p
2.9 k b p
C5-VRI0 IWI IM g,.; Chromosome
jp \ I uphi Jpsï)
- i.b kpp 10.5 k b p 4.11 k b p
Fig. 4.6. (A) Southern blet of genomic DNA of strains Streptomyces sp. strain C5 (lanes 1-3) and Streptomyces sp. strain CSVRIO (lanes 4-6). (B) Restriction maps of the plasmid, parental chromosome and mutant chromosome, showing the expected results obtained by the double crossover.
191 ml Erlenmeyer flasks with springs for dispersal of the mycelia.. Twenty-five micrograms of thiostrepton per ml were added to the cultures of the mutant strains. At 6 days, the whole broth from each culture was adjusted to pH 8.5 with 5 N NaOH and then extracted with an equal volume of chloroform:methanol (9:1). The anthracyclines and anthracyclinones in the extracts were reduced to dryness, reconstituted in 100 pi of methanol, and separated on silica gel thin-layer chromatography (TLC) plates as previously described (Dickens et al., 1996), with chloroform;heptane:methanol (10:10:3) as the mobile phase. Aloesaponarin II standards obtained earlier (Bartel et al.. 1990) were used to detect the band of interest. The aloesaponarin was quantified by measuring the absorbance of the isolated sample at 409 nm, in a 1 cm cuvette, using a UV- spectrophotometer. [(e ) for 1 gm/L = 0.0884]
LC/MS analysis.
For mass spectrometry (MS) analysis, extracts from Streptomyces sp. strain
C5VR5 {dpsCD double mutant) transformed with pANT817, and from Streptomyces sp. strain C5 VRIO {dpsABCD mutant) transformed with pANT8l7 were resolved on TLC plates as described by Bartel el al. (1990a). The bands with the same Ryas authentic aloesaponarin II was scratched out from the plates and the aloesaponarin II was extracted in chloroform, evaporated in vacuo and finally resuspended in methanol. LC/MS (Liquid
Chromatography/Mass Spectrometry) was performed on the samples using a API 300
(Perkin Elmer; Norwalk, Connecticut) mass spectrometer, that was interfaced to a HPLC
192 system and a UV detector (Shimadzu Scientific Instruments; Columbia, Maryland), at the
Ohio State Campus Chemical Instrumentation Center (Columbus, Ohio). Authentic aloesaponarin 11 standard (Bartel et al., 1990) was used to compare the results.
RESULTS AND DISCUSSION:
PKS gene structure.
Streptomyces galilaeus PKS, encoded by the polyketide synthase genes, produce aklanonic acid, which is a common intermediate in the aclacinomycin and daunorubicin biosynthetic pathway. Since aklanonic acid is made starting with propionyl-CoA. analogous to the synthesis in Streptomyces sp. strain C5, it is expected that the PKS gene cluster arrangement in these two Streptomyces spp. would be similar. Instead, the gene organization for the daunorubicin PKS gene cluster of Streptomyces sp. strain C5 was found to differ in structure from the organization of the aclacinomycin PKS cluster of
Streptomyces galilaeus (Fig. 4.7; Strohl et al.. 1997). In Streptomyces sp. strain C5, the dpsC and dpsD genes are present downstream of and presumably co-transcribed with the
KASg and KASp. This position is occupied by the ACP-encoding gene in S. galilaeus. It is yet unknown whether S. galilaeus has dpsC or dpsD homo logs elsewhere in the biosynthetic gene cluster. It has been shown earlier that dpsC has a role in ensuring correct starter unit selection, hence this unusual gene cluster organization in Streptomyces sp. strain C5, and the resultant PKS stmcture, might be important in allowing or disallowing heterologous interactions among the individual PKS units.
193 Fig. 4.7. Comparison of the Streptomyces sp. strain C5 and Streptomyces galilaeus
(ATCC 31133) PKS gene clusters (Tsukamato et al., 1994; Hutchinson and Fujii,
1994), The genes encoding the following functions are indicated: KAS„, and KASp, e.g., dpsA and dpsB; aknB, aknC; Acyl carrier protein (ACP), e.g., dpsG, aknD', polyketide reductase, dpsE, aknA, cyclases/aromatases, e.g., dpsF, aknE-, deoxyaklanonic acid oxygenase, e g.,dauG, aknX.
194 akn>C aknD '1______< aknA
dpsG dauG
^ 1
Fig. 4.7. Comparison of the Streptomyces sp. strain C5 and Streptomyces galilaeus (ATCC 31133) PKS gene clusters.
195 Thus dpsC and dpsD were speculated to be involved in forming a possible PKS
structure with the rest of the gene products, and could be interfering with interactions of
the individual components of the PKSs with other heterologously expressed PKS. This
might explain the lack of aloesaponarin II formation when actl-orfl and actl-orfll (KAS^
and KASp, respectively), from S. coelicolor, were expressed in wild-type Streptomyces
sp. strain C5. A similar expression in S. galilaeus (with no dpsC and dpsD transcribed
with KAS„ and KASp), produced the hybrid aloesaponarin II molecule. This led us to
attempt similar experiments using the S. coelicolor KAS„ and KASp, in C5 strains which
had the dpsC and dpsD genes deleted. Additionally, we also attempted this experiment
using another mutant C5 strain in which dpsA-dpsD (in effect the backbone genes) were
deleted.
Gene replacement.
A dpsABCD null mutant was constructed by transforming protoplasts of
Streptomyces sp. strain C5 with pANTI 106. a suicide vector in which dpsABCD was
replaced with aphl. Southern hybridization of several neomycin-resistant transformants,
using the integrated DNA as a probe, revealed a double crossover in which aphl replaced dpsA through dpsD in the genome is shown in Figure 4.6. This mutant strain.
Streptomyces sp. strain C5-VR10. was found to produce in NDYE medium no
anthracycline or anthracyclinone like compoimds. This result was expected since the
196 polyketide synthase genes were replaced, generating a null mutant in which no anthracycline would be produced.
Fermentation extracts from Streptomyces sp. strain CSVRIO transformed with pANT817 {actl-orfl and actl-orfll from S. coelicolor) was analyzed for aloesaponarin II formation as described in materials and methods. These extracts produced relatively low amounts of aloesaponarin II. The products were confirmed by TLC and subsequently by
LC/MS. Control strains transformed with pANT849 (vector used for constructing pANT817) did not show any aloesaponarin Il-like products. Similarly, extracts from
Streptomyces sp. strain C5VR5 (Chapter 2; dpsCD null mutant) transformed with pANT817 and pANT849 were also analyzed. LC/MS and TLC data confirmed the presence of aloesaponarin II (Appendix C). An M+1 fragment of 255 corresponding to that obtained for authentic aloesaponarin II was obtained for the aloesaponarin II isolated from the mutant strains (Table 4.4). The amounts of aloesaponarin II, produced by these
Streptomyces sp. strain C5 mutant strains (0.02 mg/ml), were considerably lower than those produced by the 5. galilaeus strains (2.4 mg/ml) when transformed with the same plasmid. In the control, no aloesaponarin II was detected in the fermentation broth of
Streptomyces sp. strain C5 (wild type strain) harboring pANT817.
Heterologous interaction of PKSs: biosynthesis of aloesaponarin II.
Figure 4.8 shows a cartoon representation of my interpretation for the results of the experiment. It is evident from the experimental data that in the absence of the dpsC
197 Table 4.4. Analysis of the aloesaponarin II produced by introducing pANT817 (S. coelicolor actl-orfsl and II) into Streptomyces sp. strain C5VR5 (dpsCD mutant) and
Streptomyces sp. strain CSVRIO (dpsABCD mutant).
198 Results of:
Strain (plasmid) TLC (R^)** LC/MS '
S. galilaeus (ATCC 31133) (pANT817)“ 0.45 255
Streptomyces sp. strain C5 (pANT817) --
Streptomyces sp. strain C5VR5 (pANT817) 0.45 255
Streptomyces sp. strain C5VR10 (pANT817) 0.45 255
“ Source of authentic aloesaponarin II ** SolventSolvent system system used used for for TLC TLC analysis analys was CHCl^iheptaneiCH^OH (10:10:3). LC/MS was performed as described in materials and methods.
Table 4.4. Analysis of the aloesaponarin II produced by introducin pANT817 into strains Streptomyces sp. strain C5VR5 and Streptomyces sp. strain C5VR10. Fig. 4.8. Model depicting the interactions among PKS components fromStreptomyces sp. strain C5VR5 and CSVRIO and Streptomyces coelicolor. In wild type C5 strain aklanonic acid is the only product produced, while in the absence of dpsC and dpsD or dpsA throughdpsD aloesaponarin II was detected.
200 dpsG dauG I I ^§F\( dpsE\ dpsA dpsB dpsC dpsD
actl-1 ' actl-2
COOCH, O OH I j Actl-1IActl-2
OH O OH OH
dpsG dauG
I I m F % V
actl-1/ actl-2
dpsG dauG
actl-1 / actl-2 V : OH O
DpsG DpsG
COOCH, O OH I ■ i
OH O OH OH
Fig. 4.8. Model depicting the interactions among PKS components from Streptomyces sp. strain C5VR5 and CSVRIO and Streptomyces coelicolor.
201 and dpsD gene products, and in the absence of dpsA through dpsD. the remaining enzymes of the PKS from Streptomyces sp. strain C5 mutant are able to produce aioesaponarin II with the help of actl-orfi and orfll gene products from S. coelicolor. A similar production is not seen in wild type Streptomyces sp. strain C5 in several experiments. It is possible that the dpsC and dpsD by some yet unknown mechanism prevents the productive interaction between the actl-orfsl and II gene products and the native PKS from Streptomyces sp. strain C5. A model depicting such activity is diagrammed in Fig. 4.8. Further investigation of this role will be possible by isolating and purifying the individual proteins of the PKS and carefully reconstituting them to obtain the polyketide products. The results described in this chapter provides one more step towards that final objective.
202 CHAPTER 5
SUMMARY
Daunorubicin and doxorubicin are clinically important antibiotics used to treat a wide variety of neoplasias. These complex drugs are made from the decarboxylative condensation of nine malonates onto a simple molecule such as propionate followed by several modifying reactions. These reactions are catalyzed by several enzymes of a complex that have been extensively studied and characterized. An understanding of the enzymatic reactions involved in producing the drugs is a key factor in manipulating the pathway proteins to engineer anticancer drugs that would be more efficacious than doxorubicin. Although much is known about the role of the gene products that catalyze the later reactions in the pathway, the early reactions in which the simple propionate is condensed with malonates and modified to a chromophoric intermediate aklanonic acid is as yet unknown. These reactions are catalyzed by the polyketide synthase (PKS) gene products that are encoded by the PKS genes.
The research presented in this dissertation describes the gene products that are involved in the early reactions leading up to aklanonic acid. The study described in this
203 dissertation provides a better understanding of the roles played by certain core gene products in selecting propionate to produce these anticancer drugs. Additionally, the work, presented here has shown a way to allow productive interactions among heterologous polyketide synthase genes from other Streptomyces spp. with the PKS genes of Streptomyces sp. strain C5. This interaction was earlier not observed in this strain, possibly as a result of the unusual PKS structure.
When this study was undertaken, the PKS gene architecture in Streptomyces sp. strain C5 was known. What had been established was that certain genes were unique in this cluster (e.g., dpsC and dpsD), and that certain genes were placed in an irregular position e.g., dpsG, when compared with other known streptomycete PKS gene clusters.
The work by Grimm et al. (1994) had shown that dpsD might not be significant under heterologous situations for polyketide production via a propionate. dpsD is a putative acyl-CoA:ACP acyltransferase and its function could be substituted by other acyltransferases or it could be a gene product whose function maybe unnecessary for polyketide formation. What had not been studied was whether dpsD could be eliminated under native conditions, i.e.. in Streptomyces sp. strain C5 without effecting polyketide biosynthesis. Moreover, there was no research that was done to study the role of dpsC. if any, in polyketide biosynthesis. The research presented in this dissertation describes three specific findings that contribute to a better understanding of anthracycline biosynthesis.
204 The most essential discovery presented here is the minimal gene products from
Streptomyces sp. strain C5. that can in heterologous hosts catalyze polyketide biosynthesis. This analysis was performed by expressing a series of gene constructs, representing several permutations of the PKS genes, from Streptomyces sp. strain C5. in heterologous hosts. The results from the analysis suggest that dpsC and dpsD gene products might not be as essential in priming polyketide biosynthesis as it was hypothesized earlier. Moreover, the deletion of these unique genes. dpsC and dpsD, in the native strain {Streptomyces sp. strain C5) resulted in production of daunorubicin as well as acetate-derived anthracycline products. This result is key in understanding the role of dpsC and dpsD. Transforming this deleted strain with dpsC and then dpsD further demonstrated that dpsC alone seems responsible for propionate selection in polyketide and anthracycline biosynthesis. The function of dpsD seems to be compensated or redundant to the system. These findings provide a clue to the possible role of the unique gene products dpsC and dpsD in selecting propionate as the only starter in polyketide biosynthesis.
The research presented in this dissertation also describes methods that were developed to test the genetic results using biochemical techniques. A cell-free system for polyketide biosynthesis in Streptomyces sp. strain C5 was developed. This cell free system is significant as it is the only known system for Streptomyces sp. strain C5 and only the third such system used for analyzing antibiotic pathways (Shen and Hutchinson.
1993; Carreras et al., 1996). More importantly, the system developed for studying
205 Streptomyces sp. strain C5 PKS is innovative in the use of d^-propionyl-CoA as well as in the use of PKS systems generated after deleting certain essential gene products {dpsC and dpsD). The analysis of the results obtained by the biochemical methods have added significant knowledge about the PKSs and have opened the doors for future research in this area.
Finally, this study has paved the way for producing heterologous PKS combinations, between PKS components firom Streptomyces sp. strain C5 and PKS components from Streptomyces coelicolor. Such combinations were earlier found not to be productive between the respective PKS of these strains. By generating specific mutant
C5 strains such combinations have been shown to be productive.
These specific findings, along with the techniques and tools developed during this research, have shown the way for an exciting phase of research in this field where a combination of biochemistry and genetic tools can now be applied to dissect the reactions leading to aklanonic acid. Since the intermediates prior to aklanonic acid are unavailable in solution, it is popularly believed that these intermediates might be channeled among the components of the PKS and released only when aklanonic acid is fully formed. Using the information provided by this research, biochemical tools can be applied to study and test this possibility. Purifying individual proteins of the PKSs and reconstituting the putative complex should yield interesting results.
This research has also shown that dpsC encodes an enzyme that possibly selects propionate in priming polyketide biosynthetic enzymes. Streptomyces sp. strain C5
206 mutants lacking dpsC gene might be tested to see if tliey can utilize any other starter
moieties such as the CoA-esters of butyrate or isopropionate. This could lead to
additional unique anthracyclines that might be more efficacious anticancer activity over
the existing drugs. Similarly, searching for dpsC-\\ke gene products in other pathways
such as the aclacinomycin pathway in S. galilaeus strains, or other propionate utilizing strains is another line of research that can be worth pursuing. Protein-protein interactions seem to be involved in PKS formation and activity. Preliminary data implicating DpsC
interactions with DpsA (KASJ have been obtained for Streptomyces sp. strain C5 PKS
(Rajgarhia and Strohl. unpublished data). This could mean that additional protein-protein interactions among other PKS proteins might exist and is an interesting area worth investigation.
207 LIST OF REFERENCES
Arcamone, F. 1981a. Doxorubicin: Anticancer antibiotics. New York: Academic Press, 1981.
Arcamone, F. 1981b. Structure-activity relationships, p. 79-84. In G. Mathé, M. Maral. and R. De Jager R (ed.), Anthracyclines. Current status and future developments. Masson Publishing USA, Inc., New York.
Arcamone, F., A. DiMarco, M. Gaetani, and T. Scotti. 1961. Isolamento ed attività antitumorale di un antibiotico da Streptomyces sp. Giom Microbiol. 9:83-90.
Arcamone, F., G. Cassinelli, G. Fan tin:, A. G rein, P. Orezzi, C. Pol, and C. Spaila. 1969a. Adriamycin, 14-hydroxydaunomycin, a new antitumor antibiotic from S. peucetius var. caesius. Biotechnol. Bioeng. 11:1101-1110.
Arcamone, F., G. Franceschi, and S. Penco. 1969b. Adriamycin (14- hydroxydaunomycin), a novel antitumor antibiotic. Tetrahedron Lett. 13:1007-1010.
Arcamone, F., S. Penco, A. Vigevani, S. Redaelli, G. Franchi, A. DiMarco, A. M. Casazza, T. Dasdia, F. Formelli, A. Necco, and C. Soranzo. 1975. Synthesis and anti tumor properties of new glycosides of daunomycinone and adriamycinone J. Med. Chem. 18:703-707.
Arrowsmith, T. J., F. Malpartida, D. H. Sherman, A. Birch, D. A. Hopwood, and J. A. Robinson. 1992. DNA encoding polyketide synthase genes from the monensin ^xodMCQx Streptomyces cinnamonensis. Mol. Gen. Genet. 234:254-264.
Bachur, N. R., R. Johnson, F. Yu, R. Hickey, and L. Maikas. 1995. Anthracycline antihelicase action. New mechanism with implications for guanosine-cytidine intercalation specificity, p. 204-221. In W. Priebe (ed.), Anthracycline antibiotics. New analogues, methods of delivery, and mechanisms of action. American Chemical Society Press, Washington, D. C.
Bartel, P. 1989. Analysis of natural and hybrid antibiotic formation in anthracycline- producing streptomycetes. Ph. D. dissertation. The Ohio State University, Columbus.
208 Bartel, P. L., C. -B. Zhu, J. S. Lampel, D. C. Dosch, N. C. Connors, W. R. Strohl, J. M. Beale, Jr., and H. G. Floss. 1990. Biosynthesis of anthraquinones by interspecies cloning of actinorhodin biosynthesis genes in streptomycetes: Clarification of actinorhodin gene functions. J. Bacteriol. 172:4816-4826.
Bartel, P. L., N. C. Connors, and W. R. Strohl. 1990b. Biosynthesis of anthracyclines: Analysis of mutants of Streptomyces sp. strain C5 blocked in daunomycin biosynthesis. J. Gen. Microbiol. 136:1877-1886.
Berdy, J. 1995. Are actinomycetes exhausted as a source of secondary metabolites?, p. 13-34. In V. G. Debabov, Y. V. Dudnik, and V. N. Danilenko (ed.). Proceedings of the Ninth International Symposium on the Biology of the Actinomycetes. All-Russia Scientific Research Institute for Genetics and Selection of Industrial Microorganisms. Moscow.
Bergh, S., and M. Uhlen. 1992. Analysis of a polyketide synthesis-encoding gene cinslQv o f Streptomyces curacoi. Gene 117: 131-136.
Bevitt, D. J., J. Cortes, S. F. Haydock, and P. F. Leadlay. 1991. 6- DeoxyerythronoIide-P-synthase from Saccharopolyspora erythraea: Cloning of the structural gene, sequence analysis and inferred domain structure of the multifunctional enzyme. Eur. J. Biochem. 204:39-49.
Bibb, M. J., D. H. Shennan, S. Omura, and D. A, Hopwood. 1994. Cloning, sequencing and deduced functions of a cluster of Streptomyces genes probably encoding biosynthesis of the polyketide antibiotic ffenolicin. Gene 142:31-39.
Bibb, M. J., S. Biro, H. Motamedi, J. F. Collins, and C. R. Hutchinson. 1989. Analysis of the nucleotide sequence of the Streptomyces glaucescens tcml genes provides key information about the enzymology of polyketide tetracenomycin C antibiotic biosynthesis. EMBO J. 8:2727-2736.
Blanco, G., P. Brian, A. Pereda, C. Mendez, J. A. Salas, and K. F. Chater. 1993. Hybridization and DNA sequence analyses suggest an early revolutionary divergence of related biosynthetic gene sets encoding polyketide antibiotics and spore pigments in Streptomyces spp. Gene 130:107-1 16.
Blum, R., and S. Carter. 1974. Adriamycin, a new anticancer drug with significant clinical activity. Ann. Intern. Med. 80:249-259.
209 Blumauerovâ, M., E. Kralovcovà, J. Mateju, E. Mateju, J. Hostalek, and Z. Vanek. 1979. Biotransformations of anthracyclinones in Streptomyces coeruleorubidus and Streptomyces galilaeus. Folia Microbiol. 24:117-127.
Blumauerovâ, M., J. Mateju, K. Stajner, and Z. Vanek. 1977. Studies on the production of daunomycinone-derived glycosides and related metabolites in Streptomyces coeruleorubidus and Streptomyces peucetius. Folia Microbiol. 22:275-285.
Bonadonna, G., S. Monfardini, M. de Lena, and F. Fossati-Bellani. 1969. Clinical evaluation of adriamycin, a new antitumour antibiotic. Brit. Med. J. 3:503-506.
Bradford, M. M. 1976. A rapid and sensitive method for quantitation of micrograpm quantitites of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.)
Brockmann, H., and K. Bauer. 1950. Rhodomycin, ein rotes Antibioticum aus Actinomyceten. Naturwissenschaften 37:492-493.
Brockmann, H., and H. Brockmann Jr. 1963. Rhodomycine. VIII; Antibiotica aus Actinomyceten, L. Chem. Ber. 96:1171-1178.
Buchanan, B. B. 1969. Role of ferredoxin in the synthesis of a-ketobutyrate from propionyl-CoA and carbon dioxide by enzymes from photosynthetic and non photosynthetic bacteria. J. Biol. Chem. 244:4218-4223.
Caballero, J. L., E. Martinez, F. Malpartida, and D. A. Hopwood. 1991. Organization and functions of the act VA region of the actinorhodin biosynthetic gene cluster of Streptomyces coelicolor. Mol. Gen. Genet. 230:401-412.
Carreras, C. and C. Khosla. 1998. Purification and in vitro reconstitution of the essential protein components of an aromatic polyketide synthase. Biochemistry 37:2084-2088.
Carreras, C. W., R. Pieper and C. Khosla. 1996. Efficient synthesis of aromatic polyketides in vitro by the actinorhodin polyketide synthase. J. Am. Chem. Soc. 118:5158- 5159.
Carter, M. J., and 1. D. Milton. 1993. An inexpensive and simple method for DNA purification on silica particles. Nucleic Acids Res. 21:1044.
Casey, M. L., R. C. Paulick, and H. W. Whitlock. 1978. Carbon- 13 nuclear magnetic resonance study of the biosynthesis of daunomycin and islandicin. J. Org. Chem. 43:1627-1634.
210 Cassinelli, G., and p. Orezzi. 1963. La daunomicina: Un nuovo antibiotico ad attività citostatica isolamento e proprietà. Giom Microbiol. 11:167-174.
Chen, A. Y., and L. F. Liu. 1994. DNA topoisomerases - essential enzymes and lethal targets. Annu. Rev. Pharmacol. Toxicol. 34:191-218.
Clough, R. C., S. R. Bamum, and J. G. Jaworski. 1989. Synthesis of radiolabeled acetyl-CoA from sodium acetate. Anal. Biochem. 176:82-84.
Connors, N. C., and W. R. Strohl. 1993. Partial purification and properties of carminomycin 4-O-methyltransferase from Streptomyces sp. strain C5. J. Gen. Microbiol. 139:1353-1362.
Connors, N. C., P. L. Bartel, and W. R. Strohl. 1990. Biosynthesis of anthracyclines: carminomycin 4-O-methyltransferase, the terminal enzymatic step in the formation of daunomycin. J. Gen. Microbiol. 136:1895-1898.
Connors, N. C., P. L. Bartel, and W. R. Strohl. 1990. Biosynthesis of anthracyclines: Enzymatic conversion of aklanonic acid to aklavinone and s-rhodomycinone by anthracycline producing S'rrep/owyce^. J. Gen. Microbiol. 136:1895-1898.
Crosby, J., D. H. Sherman, M. J. Bibb, W. P. Revill, D. A. Hopwood, and T. J. Simpson. 1995. Polyketide synthase acyl carrier proteins from Streptomyces: Expression in Escherichia coli, purification and partial characterisation. Biochim. Biophys. Acta 1251:32-42.
Crump, M. P., J. Crosby, C. E. Dempsey, J. A. Parkinson, M. Murray, D. A. Hopwood, and T. J. Simpson. 1997. Solution structure of the actinorhodin polyketide synthase acyl carrier protein from Streptomyces coelicolor A3(2). Biochemistry 36:6000-6008.
Crump, M. P., J. Crosby, C. E. Dempsey, M. Murray, D. A. Hopwood, and T. J. Simpson. 1996. Conserved secondary structure in the actinorhodin polyketide synthase acyl carrier protein from Streptomyces coelicolor A3 (2) and the fatty acid synthase acyl cectncx TÇtxoiém from Escherichia coli. FEBS Lett. 391:302-306.
Cullinane, C., A. van Rosmalen, and D. R. Phillips. 1994. Does adriamycin induce interstrand cross-links in DNA? Biochemistry 33:4632-4638.
Cummings, J., L. Anderson, N. Willmott, and J. F. Smyth. 1991. The molecular pharmacology of doxorubicin in vitro. Eur. J. Cancer. 27:532-535.
211 Cutts, S. M., and D. R. Phillips. 1995. Use of oligonucleotides to define the site of interstrand cross-links induced by adriamycin. Nucl. Acids Res. 23:2450-2456.
Dekleva, M. L., and W. R. Strohl. 1988. Activity of phosphoewo/pymvate carboxylase of an anthracycline-producing streptomycete. Can. J. Microbiol. 34:1241-1246.
Dekleva, M. L., and W. R. Strohl. 1988. Biosynthesis of s-rhodomycinone from glucose by Streptomyces sp. strain C5. Can. J. Microbiol. 34:1235-1240.
DeSanti, C., J. S. Lampel, K. A. Lampel, and W. R. Strohl. Unpublished data.
Despois, R., M. Dubost, D. Mancy, R. Maral, L. Ninet, S. Finnert, J. Preud’homme, Y. Charpentie, A. Belloc, N. de Chezelles, J. Lunel, and J. Renaut. 1967. Un nouvel antibiotique doué d’activité antitumorale: la rubidomycine (13.057 R.P.). I. Préparation et propriétés. Arzneim Forsch. 17:934-939.
Devereux, J., P. Haeberli, and O Simithies. 1984. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387-395.
Dickens, M. L., Priestley, N. D., and W. R. Strohl. 1997. In vivo and in vitro bioconversion of s-Rhodomycinone glycoside to doxorubicin: Functions of DauP. DauK. and DoxA. J. Bacteriol. 179:2641-2650.
Dickens, M. L., and W. R. Strohl. 1996. Isolation and characterization of a gene from Streptomyces sp. strain C5 that confers on Streptomyces lividans TK24 the ability to convert daunomycin to doxorubicin. J. Bacteriol. 178:3389-3395.
Dickens, M. L., J. Ye, and W. R. Strohl. 1995. Analysis of clustered genes encoding both early and late steps in daunomycin biosynthesis by Streptomyces sp. strain C5. J. Bacteriol. 177:536-543.
Dickens, M. L., J. Ye, and W. R. Strohl. 1996. Cloning, sequencing and analysis of akiaviketone reductase from Streptomyces sp. strain C5. J. Bacteriol. 178:3384-3388.
DiMarco, A., M. Gaetani, L. Dorigotti, M. Soldati, and O. Bellini. 1963. Studi sperimentali sull ‘attivita* antineoplastica del nuovo antibiotico daunomicina. Tumori 49:203-217.
DiMarco, A., M. Gaetani, L. Dorigotti, M. Soldati, and O. Bellini. 1964a. Daunomycin: a new antibiotic with antitumor activity. Cancer Chemother. Rep. 38:31- 38.
212 DiMarco, A., M. Gaetani, P. Orezzi, B. M. Scarpinato, R. Slivestrini, M. Soldati, T. Dasdia, and L. Vaientini. 1964b. ‘Daunomycin’, a new antibiotic of the rhodomycin group. Nature (London) 201:706-707.
Donadio, S., M. J. Staver, J. B. McAlpine, S. J. Swanson, and L. Katz. 1991. Modular organization of genes required for complex polyketide biosynthesis. Science 252:675-679.
Doroshow, J. H. 1995. Role of reactive-oxygen metabolism in cardiac toxicity of anthracycline antibiotics, p. 259-267. In W. Priebe (ed.), Anthracycline antibiotics. New analogues, methods of delivery, and mechanisms of action. American Chemical Society Press, Washington, D. C.
Drlica, K., and R. J. Franco. 1988. Inhibitors of DNA topoisomerases. Biochemistry 27:2253-2259.
Dubost, M., P. Ganter, R. Maral, L. Ninet, S. Pinnert, J. Prud’homme, and G. H. Wemer. 1963. Un novel antibiotique a properties cytostatique: la rubidomycine. CR Acad. Sci. 257:1813-1815.
DuVemay, V. H. 1981. Molecular pharmacology of anthracycline antitumor antibiotics, p. 233-271. In S. T. Crooke, and A. W. Prestayko (ed.). Cancer and chemotherapy, Vol 111. Antineoplastic agents. Academic Press, New York.
DuVemay, V. H. Jr., D. Eubanks, R. Perales, A. W. Prestayko, and S. T. Crooke. 1982. The antitumor effects of anthracyclines. 11. The stereospecificity of the carbomethoxy group at position 10 of the class II anthracycline molecule. Mol. Pharmacol. 21:196-203.
DuVemay, V. H., S. Mong, and S. T. Crooke. 1980. Molecular pharmacology of anthracyclines: Demonstration of multiple mechanistic classes of anthracyclines, p. 61- 123. In S. T. Crooke, and S. D. Reich (ed.), Anthracyclines. Current status and new developments. Academic Press, New York.
Eckardt, K., and C. Wagner. 1988. Biosynthesis of anthracyclinones. J. Basic Microbiol. 28:137-144.
Eckardt, K., D. Tresselt, G. Schumann, W. Ihn, and C. Wagner. 1985. Isolation and chemical structure of aklanonic acid, an early intermediate in the biosynthesis of anthracyclines. J. Antibiot. 38:1034-1039.
213 Eckardt, K., D. Tresselt, G. Schumann, W. Ibn, and C. Wagner. 1985. Isolation and chemical stnicture of aklanonic acid, an early intermediate in the biosynthesis of anthracyclines. J. Antibiot. 38:1034-1039.
Eckardt, K., G. Schumann, D. Tresselt, and W. Ihn. 1988. Biosynthesis of anthracyclinones: Isolation of a new early cyclization product akiaviketone. J. Antibiot. 41:788-793.
Eckardt, K., G. Schumann, U. Grafe, W. Ihn, C. Wagner, W. F. Fleck, and H. Thrum. 1985. Preparation of labeled aklanonic acid and its bioconversion to anthracyclinones by mnVanXs oiStreptomyces griseus. J. Antibiot. 38:1096-1097.
Feinberg, A. P., and B. Fogelstein. 1983.A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-13. (Addendum, Anal. Biochem. 137:266-267, 1984.)
Femandez-Moreno, M. A., E. Martinez, L. Boto, D. A. Hopwood, and F. Malpartida. 1992. Nucleotide sequence and deduced functions of a set of cotranscribed genes of Streptomyces coelicolor A3(2) including the polyketide synthase for the antibiotic actinorhodin. J. Biol. Chem. 267:19278-19290.
Femandez-Moreno, M. A., E, Martinez, J. L. Caballero, K. Ichinose, D, A. Hopwood, and F. Malpartida. 1994. DNA sequence and functions of the actVl region of the actinorhodin biosynthetic gene cluster of Streptomyces coelicolor A3(2). J. Biol. Chem. 269:24854-24863.
Fields, S. 1993. The two-hybrid system to detect protein-protein interactions. Methods (Orlando) 5:116-124.
Filippini, S., M. M. Solinas, U. Breme, M. B. SchlQter, D. Gabellini, G. Biamonti, A. L. Colombo, and L. Garofano. 1995. Streptomyces peucetius daunorubicin biosynthesis gene, dnrF: sequence and heterologous expression. Microbiol. 141:1007- 1016.
Fischer, D. S., M. T. Knohf, and H. J. Durivage. 1993. Cancer themotherapy and pharmacology, p. 527. In The cancer chemotherapy handbook, fourth edition. Mosby, St. Louis.
Formelli, F., A. DiMarco, A. M. Casazza, G. Pratesi, R. Supino, and A. Mariani. 1978. Biological properties of 4-demethoxydaunorubicin: In vitro and in vivo studies. Curr. Chemother. 2:1240-1242.
214 Fujii, S., K. Kubo, O. Johdo, A. Yoshimoto, T. Ishikura, H. Naganawa, T. Sawa, T. Takeuchi, and H. Umezawa. 1986. A new anthracycline metabolite D788-I (10- carboxy-13-deoxocarminomycin) in daunorubicin beer. J. Antibiot. 39:473-475.
Fujiwara, A., T. Hoshino, M. Tazoe, and M. Fujiwara. 1981. Auramycins and sulfiirmycins, new anthracycline antibiotics: characterization of aglycones, auramycinone and sulfurmycinone. J. Antibiot. 34:608-610.
Fujiwara, A., M. Tazoe, T. Hoshino, V. Sekine, S. Masuda, and S. Nomura. 1981. New anthracycline antibiotics, 1-hydroxyauramycins and l-hydroxysulfiirmycins. J. Antibiot. 34:912-915.
Furuya, K., and C. R. Hutchinson. 1996. The DnrN protein of Streptomyces peucetius, a pseudo-response regulator, is a DNA-binding protein involved in the regulation of daunorubicin biosynthesis. J. Bacteriol. 178:6310-6318.
Gaisser, S., A. Trefeer, S. Stockert, A. Kirschning, and A. Bechthoid.1997. Cloning of an avilamycin biosynthetic gene cluster from Streptomyces viridochromogenes Tü57. J. Bacteriol. 179:6271-6278.
Gallo, M. A., J. Ward, and C. R. Hutchinson. 1996. The dnrM gene in Streptomyces peucetius contains a naturally occurring frameshift mutation that is suppressed by another locus outside of the daunorubicin-production gene cluster. Microbiol. 142:269-275.
Gerlitz, M., G. Meurer, E. Wendt-Pienkowski, K. Madduri, and C. Richard Hutchinson. 1997. Effect of the daunorubicin dpsH gene on the choice of starter unit and cyclization pattern reveals that Type II polyketide synthases can be unfaithful yet intriguing. J. Am. Chem. Soc. 119:7392 -7393.
Gramajo, H. C., J. White, C. R. Hutchinson, and M. J. Bihh. 1991. Overproduction and localization of components of the polyketide synthase of Streptomyces glaucescens involved in the production of the antibiotic tetracenomycin C. J. Bacteriol. 173:6475- 6483.
G rein. A., C. Spalla, A. Di Marco, and G. Canevazzi. 1963. Descrizione e classificazione di un attinomicette (Streptomyces peucetius sp. nova) produttore di Una Sostanza ad attivita anti tumorale: la daunocinina. Giom Microbiol. 11:109-118.
Grimm, A., K. Madduri, A. Ali, and C. R. Hutchinson. 1994. Characterization of the Streptomyces peucetius ATCC 29050 genes encoding doxorubicin polyketide synthase. Gene 151:1-10.
215 Guilfoile, P. G., and C. R. Hutchinson. 1991. A bacterial analog of the mdr gene of mammalian tumor cells is present in Streptomyces peucetius, the producer of daunorubicin and doxorubicin. Proc. Nat'l. Acad. Sci. USA 88:8553-8557.
Hale, J. P., and I. J. Lewis. 1994. Anthracyclines: Cardiotoxicity and its prevention. Arch. Dis. Childhood. 71:457-462.
Hailam, S. E., F. malpartida, and D. A. Hopwood. 1988. Nucleotide sequence, transcription and deduced function of a gene involved in polyketide antibiotic synthesis in Streptomyces coelicolor. Gene 74:305-320.
Haydock, S. F., J. F, Aparicio, 1. Molnar, T. Schweeke, L. E. Khaw, A. Konig, A. F. A. Marsden, i. S. Galloway, J. Staunton, and P. F. Leadlay. 1995. Divergent sequence motifs correlated with the substrate specificity of (methyl)malonyl-CoA:acyl carrier protein transacylase domains in modular polyketide synthases. FEMS Lett. 374:246-248.
Hegyi, J. R., and N. N. Gerber. 1968. N,-Pyrromycinone. a new tetracenequinone (anthracyclinone) from Streptomyces sp. 1-8. Tetrahedron Lett. 13:1587-1589.
Hickman, J. A. 1992. Apoptosis induced by anticancer drugs. Cancer Metastasis Rev. 11:121-139.
Higashide, E., T. Hasegawa, M. Shibata, and K. Mizumo. 1972. U.S. Patent 3655878.
Hong, Y-S., C. K. Hwang, S-K. Hong, V. H. Kim, and J. J. Lee.1994. Molecular cloning and characterization of the aklavinone 11-hydroxylase gene of Streptomyces peucetius subsp. caesius (ATCC 27952). J. Bacteriol. 176:7096-7101.
Hopwood, D. A. 1997. Genetic contributions to understanding polyketide synthases. Chem. Rev. 97:2465-2497.
Hopwood, D. A., M. J. Bibb, K. F. Chater, T. Kieser, C. J. Bruton, H. M. Kieser, D. J. Lydiate, C. . Smith, J. M. Ward, and H. Schrempf. 1985a. Genetic manipulation of Streptomyces-. a laboratory manual. The John limes Foundation, Norwich, England.
Hopwood, D. A., F. Malpartida, H. M. Keiser, H. Ikeda, J. Duncan, I. Fujii, B. A. M. Rudd, H. G. Floss, and S. Omura. 1985b. Production of hybrid antibiotics by genetic engineering. Nature 314:642-644.
Hopwood, D. A., and D. H. Sherman. 1990. Molecular genetics of polyketides and its comparison to fatty acid biosynthesis. Annu. Rev. Genet. 24: 37-66
216 Horton, D., and w. Priebe. 1981. New adriamycin analogs. Synthesis and antitumor activity of 14-substituted 7-0(3,4-di-0-acetyl-2,6-dideoxy-a-L-/yro-hexopyranosyl) daunomycinones. J. Antibiot. 34:1019-1025.
Hoshino, T., and A. Fujiwara. 1983. Characterization of the microbial conversion products of auramycinone by Streptomyces coeruleorubidus ATCC 31276. J. Antibiot. 36:1463-1467.
Hutchinson, C. R. 1995. Anthracyclines, p. 331-357. In: L. C. Vining, C. Stuttard (eds.). Genetics and biochemistry of antibiotic production. Butterworth-Heinemann, Boston, Massachusettes.
Hutchinson, C. R. 1997. Biosynthetic studies of Daunorubicin and Tetracenomycin C. Chem. Review. 97:2525 - 2535.
Hutchinson, C. R., and I. Fujii. 1995. Polyketide synthase gene manipulation: A structure-function approach in engineering novel antibiotics. Annu. Rev. Microbiol. 49:201-238.
Hutchinson, C. R., H. Decker, K. Madduri, S. L. Otten, and L. Tang. 1994. Genetic control of polyketide biosynthesis in the genus Streptomyces. Antonie van Leeuwenhoek 64:165-176.
Hwang, C. K., H. S. Kim, Y-S. Hong, V. H. Kim, S-K. Hong, S-J. Kim, and J. J. Lee. 1995. Expression of Streptomyces peucetius genes for doxorubicin resistance and aklavinone 11-hydroxylase in Streptomyces galilaeus ATCC 31133 and production of a hybrid aclacinomycin. Antimicrob. Agents Chemother. 39:1616-1620.
Ingehretsen, O. C., and M. Farstad. 1980. Direct measurement of free coenzyme A in biological extracts by reverse phase high performance liquid chromatography. J. Chromatogr. 202:439-445.
Iwahuchi, K., B. Li, P. Bartel, and S. Fields. 1993. Use of the Two-Hybrid system to identify the domain of p53 involved in oligomerization. Oncogene 8:1693-1696.
Janssen, G. R., and M. J. Bihh. 1993. Derivatives of pUC18 that have Bglll sites flanking a modified multiple cloning site and that retain the ability to identify recombinant clones by visual screening of Escherichia coli. Gene 124:133-142
Katz, L., and S. Donadio. 1993. Polyketide biosynthesis: prospects for hybrid antibiotics. Annu. Rev. Microbiol. 47:875-912.
217 Kawaguchi, A., T. Yoshimura, and S. Okuda. 1981. A new method for the preparation of acyl-CoA thioesters. J. Biochem. 89:337-339.
Keizer, H. G., H. M. Pinedo, G J. Schuurhuis, and H. Hoenje. 1990. Doxorubicin (adriamycin): A critical review of free radical-dependent mechanisms of cytotoxicity. Pharm. Ther. 47:219-231.
Kendrew, S. G., E. Deutsch, K. Madduri, and C. R. Hutchinson. 1996. Aklanonic acid methyl ester cyclase: An intramolecular aldol condensation in the biosynthesis of daunorubicin. Bloomington: Abstr Sixth Conf Gen Mol Biol Industr Microorg, P4.
Kem, D. L., R. H. Bunge, J. C. French, and H. W. Dixon. 1977. The identification of E-rhodomycinone and 7-deoxy-daunorubicinol aglycone in daunorubicin beers. J. Antibiot. 30:432-434.
Khosla, C., R. McDaniel, S. Ebert-Khosla, R. Torres, D. H. Sherman, M. J. Bibb, and D. A. Hopwood. 1993. Genetic construction and functional analysis of hybrid polyketide synthases containing heterologous acyl carrier proteins. J. Bacteriol. 175:2197-2204.
Kim, H-S, Y-S. Hong, Y-H. Kim, O-J. Yoo, and J. J. Lee.1996. New anthracycline metabolites produced by the aklavinone 11-hydroxylase gene in Streptomyces galilaeus ATCC 31133. J. Antibiot. 49:355-360.
Kitamura, I., H. Tobe, A. Yoshimoto, T. Old, H. Nagahawa, T. Takeuchi, and H. Umezawa. 1981. Biosynthesis of aklavinone and aclacinomycins. J. Antibiot. 34:1498- 1500.
Knight, M. 1962. The photometabolism of propionate by Rhodospirillum rubrum. Biochem. J. 84:170-173.
Komiyama, T., Y. Matsuzawa, T. Old, T. Inui, Y. Takahashi, H. Naganawa, T. Takeuchi, and H. Umezawa. 1977. Baumycins, new antitumor antibiotics related to daunomycin. J. Antibiot. 30:619-621.
Komberg, A., and W. E. Pricer, Jr. 1953. Enzymatic synthesis of the coenzyme A derivatives of long chain fatty acids. J. Biol. Chem. 204:329-343.
Kriigel, H., G. Schumann, F. Hanel, and G. Fiedler. 1993. Nucleotide sequence analysis of five putative Streptomyces griseus genes, one of which complements an early
218 function in daunorubicin biosynthesis that is linked to a putative gene cluster involved in TDP-daunosamine formation. Mol. Gen. Genet. 241:193-202.
Kwak, J. 1995. Characterization of nonsporulating mutants of Streptomyces griseus. Ph.D. dissertation. The Ohio State University, Columbus. Lampel, J. S., and W. R. Strohl. 1986. Transformation and transfection of anthracycline-producing streptomycetes. Appl. Environ. Microbiol.; 51:126-131.
Le Gouili, C., D. Desmarais, and C. V, Dary. 1993. Saccharopolyspora hirsuta 367 encodes clustered genes similar to ketoacyl synthase, ketoacyl reductase, acyl carrier protein, and biotin carboxyl carrier protein. Mol. Gen. Genet. 240:146-150.
Li, Y. 1995. Structure and function analysis of the asaA, ptpA, and cysA loci of Streptomyces coelicolor. Ph. D. dissertation. The Ohio State University, Columbus.
Liu, W. €., and K. U. Rao. 1974. U.S. Patent 3852425.
Lombo, F., G. Blanco, E. Fernandez, C. Mendez, and J. A. Salas. 1996. Characterization of Streptomyces argillaceus genes encoding a polyketide synthase involved in the biosynthesis of the antitumor mithramycin. Gene 172:87-91.
Lomovskaya, N., S-K. Hong, S-U. Kim, L. Fonstein, K. Furuya, and C. R. Hutchinson. 1996. The Streptomyces peucetius drrC gene encodes a UvrA-like protein involved in daunorubicin resistance and production. J. Bacteriol. 178:3238-3245.
Lomovskaya, N., V. Doi-Katayama, S. Filippini, C. Nastro, L. Fonstein, M. Gallo, A. L. Colombo, and C. R. Hutchinson. 1997. The Streptomyces peucetius dpsY and dnrX genes govern early and late steps of daunorubicin and doxorubicin biosynthesis. J. Bacteriol. 180:2379 - 2386.
Madduri, K., and C. R. Hutchinson. 1995. Functional characterization and transcriptional analysis of a gene cluster governing early and late steps in daunorubicin biosynthesis in Streptomyces peucetius. J. Bacteriol. 177:3879-3884.
Malpartida, F. and D. A. Hopwood. 1986. Physical and genetic characterization of gene clusters of the antibiotic antinorhodin in S. coelicolor A3 (2). Mol. Gen. Genet. 205:66-73.
Malpartida, F., S. E. Hallam, H. M. Keiser, H. Motamedi, C. R. Hutchinson, M. J. Butler, D. A. Sugden, M. Warren, C. McKillop, C. R. Bailey, G. O. Humphreys, and D. A. Hopwood. 1987. Homology between Streptomyces genes coding for synthesis of different polyketides used to clone antibiotic biosynthetic genes. Nature 325:818-822.
219 Mancy, D., and L. Ninet. 1976. U.S. Patent 3997661.
Mancy, D., J. Florent, and J. Preud’homme. 1975. U.S. Patent 3855010
Maral, R., G. Bourat, J. Foumel, P. Ganter, Y. de Ratuld, and G. H. Wemer. 1967. Un nouvel antibiotique doué d’activité antitumorale: la rubidomycine (13.057 R.P.). II. Activité antitumorale expérimentale. Arzneim Forsch 17:939-948.
Marsh, N. (1988). Molecular aspects of succinyl-thiokinase. Ph.D. Thesis. University of Cambridge, Cambridge, U.K.
Matsuzawa, Y., T. Oki, T. Takeuchi, and H. Umezawa. 1981. Structure-activity relationships of anthracyclines relative to cytotoxicity and effects on macromolecular synthesis in L1210 leukemia cells. J. Antibiot. 34:1596-1607.
McDaniel, R., S. Ebert-Khosla, D. A. Hopwood, and C. Khosla. 1993. Engineered biosynthesis of novel polyketides. Science 262:1546-1550.
McDaniel, R., S. Ebert-Khosla, H. Fu, D. A. Hopwood, and C. Khosla. 1994. Engineered biosynthesis of novel polyketides: influence of a downstream enzyme on the catalytic specificity of a minimal aromatic polyketide synthase. Proc. Natl. Acad. Sci. USA 91:11542-11546.
McDaniel, R., C. R. Hutchinson, and C. Khosla. 1995. Engineered biosynthesis of novel polyketides: Analysis of TcmN function in tetracenomycin biosynthesis. J. Am. Chem. Soc. 117:6805-6810.
McGuire, J. C., G. Glotfelty, and R. J. White. 1980a. Use of cerulenin in strain improvement of the daunorubicin fermentation. FEMS Microbiol. Lett. 9:141-143.
McGuire, J. C., M. C. Thomas, R. C. Pandey, M. Toussaint, and R. J. White. 1991. Biosynthesis of daunorubicin glycosides: analysis with blocked mutants, p. 117-122. In M. Moo-Young (ed.). Advances in Biotechnology, Vol 3. Permagon Press, New York.
McGuire, J. C , M. C. Thomas, R. M. Stroshane, B. K. Hamilton, and R. J. White. 1980b. Biosynthesis of daunomycin glycosides: role of e-rhodomycinone. Antimicrob. Agents Chemother. 18:454-464.
Meurer, G., and C. R Hutchinson. 1995. Daunorubicin Type II polyketide synthase enzymes DpsA and DpsB determine neither the choice of starter unit nor the cyclization pattern of aromatic polyketides. J. Am. Chem. Soc. 117:5899-5900.
220 Meurer, G., M. Gerlitz, E. Wendt Pienkowski, L. C. Vining, J. Rohr, and C. R. Hutchinson. 1997. Interactive type II polyketide synthases, cyclases and ketoreductases exhibit context-dependent behavior in the biosynthesis of linear and angular decapolyketides. Chem. Biol. 4:433-443.
Niemi, J., K. YUhonko, J. Hakala, R. Parssinen, A. Kopio, and P. Mântsâlâ. 1994. Hybrid anthracycline antibiotics: production of new anthracyclines by cloned genes from Streptomycespurpurascens \n Streptomyces galilaeus. Microbiol. 140:1351-1358. Oki T. 1982. Microbial transformations of anthracycline antibiotics and development of new anthracyclines. p. 75-96. In H. S. El Khadem (ed.), Anthracycline antibiotics. Academic Press, New York.
Oki, T., Y. Matsuzawa, K. Kiyoshima, and A. Yoshimoto. 1981. New anthracyclines. feudomycins, produced by the mutant from Streptomyces coeruleorubidus ME 130-A4. J. Antibiot. 34:783-790.
Olsen, I., and J. M. Merrick. 1968. Identification of propionate as an endogenous CO, acceptor in Rhodospirillum rubrum and propionate of purified propionyl-CoA carboxylase. J. Bacteriol. 95:1774-1777.
OsherofT, N., A. H. Corbett, S. H. Elsea, and M. Westergaard. 1994. Defining functional drug-interaction domains on topoisomerase II by exploiting mechanistic differences between drug classes. Cancer Chemother. Pharmacol. 34(Suppl):S19-S25.
Otten, S. L., J. Ferguson, andC. R. Hutchinson. 1995a. Regulation of daunorubicin production in Streptomyces peucetius by the dnrR2 locus. J. Bacteriol. 177:1216-224.
Otten, S. L., K. J. Stutzman Engwall and C. R. Hutchinson. 1990. Cloning and expression of daunorubicin biosynthesis genes from Streptomyces peucetius and Streptomyces peucetius subsp. caesius. J. Bacteriol. 172:3427-3434.
Otten, S. L., M. A. Gallo, K. Madduri, X. Liu, and C. R. Hutchinson. 1997. Cloning and characterization of the Streptomyces peucetius dnmZUV genes encoding three enzymes required for biosynthesis of the daimorubicin precursor thymidine diphospho-L- daunosamine. J. Bacteriol. 179:4446-4450.
Otten, S. L., X. Liu, J. Ferguson, and C. R Hutchinson. 1995b. Cloning and characterization of the Streptomyces peucetius dnrQS gene encoding a daunosamine biosynthesis enzyme and a glycosyl transferase involved in daunorubicin biosynthesis. J. Bacteriol. 177:6688-6692.
221 Paulick, R. C., M. L. Casey, and H. W. Whitlock. 1976. A '^C nuclear magnetic resonance study of the biosynthesis of daunomycin from '^CH^'^COiNa. J. Am. Chem. Soc. 98:3370-3371.
Pavanosenkova, V. I., and V. L. Karpov. 1976. Studies on biosynthesis of carbohydrate fragment of rubomycin. Antibiotiki 21:299-301
Perrin, D. D., and W. L. F. Armarego. 1988. Purification of l,r-carbonyldiimidazole. p.208. In Purification of Laboratory Chemicals. 3rd Ed., Pergamon Press, Oxford, England.
Petrovics, G. P., P. Putnovsky, B. Reuhs, J. Kim, T. A. Thorp, K. D. Noel, R. W. Carlson, and A. Kondorosi. 1993. The presence of a novel type of surface polysaccharide in Rhizobium meliloti requires a new fatty acid synthase-like gene cluster involved in symbiotic nodule development. Mol. Microbiol. 8:1083-1094.
Piepersberg, W. 1997. Molecular biology, biochemistry, and fermentation of aminoglycoside antibiotics, p. 81-163. /n W. R. Strohl (ed.). Biotechnology of industrial antibiotics. Second Edition. Marcel-Dekker, Inc., New York.
Pinnert, S., and L. Ninet. 1976. U.S. Patent 3989598.
Pommier, V. 1995. DNA topoisomerases and their inhibition by anthracyclines, p. 183- 203. In W. Priebe (ed.), Anthracycline antibiotics. New analogues, methods of delivery, and mechanisms of action. American Chemical Society Press, Washington, D. C.
Priestley, N. D. 1991. Studies of enzymes involved in the primary and secondary metabolism of antibiotic producing streptomyces. Ph.D. Thesis, University of Southampton, Southampton, U.K.
Rajgarhia.V. B. and W. R. Strohl. 1997. Minimal Streptomyces sp. strain C5 daunorubicin polyketide biosynthesis genes required for aklanonic acid biosynthesis. J. Bacteriol. 179:2690-2696.
Revill, W. P., M. J. Bihh, and D. A. Hopwood. 1995. Purification of a malonyItransferase from Streptomyces coelicolor A3(2) and analysis of its genetic determinant. J. Bacteriol. 177:3947-3952.
Revill, W. P., M. J. Bibb, and D. A. Hopwood. 1996. Relationships between fatty acid and polyketide synthases from Streptomyces coelicolor A3(2): characterization of the fatty acylsynthase acyl carrier protein. J. Bacteriol. 178: 5660-5667.
222 Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Schumann, G., C. Stengel, K. Eckardt, and W. Ihn. 1986. Biotransformation of aklanonic acid by blocked mutants of anthracycline-producing strains of Streptomyces galilaeus and Streptomyces peucetius. J Basic Microbiol. 26:249-255.
Scotti, C., and C. R. Hutchinson. 1996. Enhanced antibiotic production by manipulation of the Streptomyces peucetius dnrH and dnmT genes involved in doxorubicin (adriamycin) biosynthesis. J. Bacteriol. 178:7316-7321.
Sehested, M., P. B. Jensen. 1996. Mapping of DNA topoisomerase II poisons (etoposide, clerocidin) and catalytic inhibitors (aclarubicin. ICRP-187) to four distinct steps in the topoisomerase II catalytic cycle. Biochem. Pharmacol. 51:879-886.
Shan, K., A. M. LincofT, and J. B. Young. 1996. Anthracycline-induced cardiotoxicity. Ann. Intern. Med. 125:47-58.
Shaw, G. J., G. W. A. Milne, and A. Minghetti. 1979. Propionate precursors in the biosynthesis of daimomycin and adriamycin: A ‘^C nuclear magnetic resonance study. Phytochemistry 18:178-179.
Shen, B., and C. R. Hutchinson. 1993a. Enzymatic synthesis of a bacterial polyketide from acetyl and malonyl coenzyme A. Science (Washington. D.C) 262:1535-1540.
Shen, B., and C. R. Hutchinson. 1993b. Tetracenomycin FI monooxygenase: Oxidation of a naphthacenone to a naphthacenequinone in the biosynthesis of tetracenomycin C m Streptomyces glaucescens. Biochemistry 32:6656-6663.
Shen, B., and C. R. Hutchinson. 1996. Deciphering the mechanism for the assembly of aromatic polyketides by a bacterial polyketide synthase. Proc. Natl. Acad. Sci. (USA) 93:6600-6604.
Shen, B., R. G. Summers, E. Wendt-Pienkowski, and C. R. Hutchinson. 1995. The Streptomyces glaucescens tcmKL polyketide synthase and tcmN polyketide cyclase genes govern the size and shape of aromatic polyketides. J. Am. Chem. Soc. 117:6811-6821.
Shen, B., R. G. Summers, H. Gramajo, M. J. Bihh, and C. R. Hutchinson. 1992. Purification and characterization of the acyl carrier protein of the Streptomyces glaucescens tetracenomycin C polyketide synthase. J. Bacteriol. 174:3818-3821.
223 Sherman, D. H., E. S. Kim, M. J. Bibb, and D. A. Hopwood. 1992. Functional replacement of genes for individual polyketide synthase components in Streptomyces coelicolor A(3)2 by heterologous genes from a different pathway. J. Bacteriol. 174:6184-6190.
Sherman, D. H., F. Malpartida, M. J. Bibb, H. M. Keiser, M. J. Bibb, and D. A. Hopwood. 1989. Structure and deduced ftmction of the granaticin-producing polyketide synthase gene cluster of Streptomyces violaceoruber Tü22. EMBO J. 9:2717-2725.
Soga, K., H. Furusho, S. Mori, and T. Oki. 1981. New antitumor antibiotics: 13- methylaclacinomycin A and its derivatives. J. Antibiot. 34:770-773.
Sokatch, J. R., L. E. Sanders, V. P. Marshall. 1968. Oxidation of methylmalonate semialdehyde to propionyl-CoA in Pseudomonas aeruginosa grown on valine. J. Biol. Chem. 243:2500-2506.
Strauss, D., and W. F. Fleck. 1975. Luekaemomycin, an antibiotic with antitumor activity. II. Isolation and identification. Zeitschr. Allg. Mikrobiol. 15:615-623.
Strohl, W. R., P. L. Bartel, N. C. Connors, C-b. Zhu, D. C. Dosch, J. M. Beale Jr., H. G. Floss, K. Stutzman-Engwall, S. L. Otten, and C. R. Hutchinson. 1989. Biosynthesis of natural and hybrid polyketides by anthracycline-producing streptomycetes, p. 68-84. In C.L. Hershberger, S. W. Queener. and G. Hegeman (ed.). Genetics and Molecular Biology of Industrial Microorganisms. American Society for Microbiology, Washington, D. C.
Strohl, W. R., P. L. Bartel, Y. Li., N. C. Connors, and R. H. Woodman. 1991. Expression of polyketide biosynthesis and regulatory genes in heterologous Stretpomycetes. J. Indust. Microbiol. 7:163-174.
Strohl, W. R. and N. C. Connors. 1992. Significance of anthraquinone formation resulting from the cloning of actinorhodin genes in heterologous streptomycetes. Mol. Microbiol. 6:147-152.
Strohl, W. R., M. L. Dickens, V. B. Rajgarhia, A. Woo, and N. D. Priestley. 1997. Anthracyclines, p. 577-658. In W. R. Strohl (ed.). Biotechnology of antibiotics, 2nd ed. Marccl-Dekker, Inc., New York.
Strohl, W. R., J. Ye, and M. L. Dickens. 1995. Daunomycin biosynthesis genes of Streptomyces sp. strain C5, p. 45-54. In V. G. Debabov, Y. V. Dudnik, and V. N. Danilenko (ed.), Proceedings of the Ninth International Symposium on the Biology of the
224 Actinomycetes. All-Russia Scientific Research Institute for Genetics and Selection of Industrial Microorganisms, Moscow, Russia.
Stutzman-Engwall, K. J., S. L. Otten, and C. R. Hutchinson. 1992. Regulation of secondary metabolism in Streptomyces spp. and overproduction of daunorubicin in Streptomyces peucetius. J. Bacteriol. 174:144-154.
Suarato, A. 1990. Anti tumor anthracyclines, p. 30-62. Im D. E. V. Wilman (ed.). The chemistry of antitumour agents. Blackie and Son, Ltd., Glasgow.
Summers, R. G., A. Ali, B. Shen, W. A. Wessel, and C. R. Hutchinson. 1995. Malonyl-coenzyme Aracyl carrier protein acyltransferase of Streptomyces glaucescens: a possible link between fatty acid and polyketide biosynthesis. Biochemistry 34:9389- 9402.
Swan, D. G., A. M., Rodriguez, C. Vilches, C. Mendez, and J. A. Salas. 1994. Characterization of a Streptomyces antibioticus gene encoding a type 1 polyketide synthase which has an unusual coding sequence. Mol. Gen. Genet. 242:358-362.
Takahashi, Y., H. Naganawa, T. Takeuchi, H. Umezawa, T. Komiyama, and T. Oki and T. Inui. 1977. The structure of baumycins Al, A2, Bl, B2, Cl, and C2. J. Antibiot. 30:622-624.
Tang, L., A. Grimm, Y-X. Zhang, and C. R. Hutchinson. 1996. Purification and characterization of the DNA-binding protein Dnrl. a transcriptional factor of daunorubicin hxosynthesxs m Streptomyces peucetius. Mol. Microbiol. 22:801-813.
Thomas, R., and D. J. Williams. 1983. Oxytetracycline: biosynthetic origin of the caroboxamide substituent. J. Chem. Soc., Chem. Commun. 677-679.
Thompson, C. J , T. Kieser, J. M. Ward, and D, A. Hopwood. 1982.Physical analysis of antibiotic-resistance genes from Streptomyces and their use in vector construction. Gene 20:51-62.
Thompson, M. W., W. R. Strohl, and H. G. Floss. 1992. Purification and characterization of TDP-D-glucose 4,6-dehydratase from anthracycline-producing streptomycetes. J. Gen. Microbiol. 138:779-786.
Tsay, J.-T., W. Oh, T. J. Larson, S. Jackowski, and C. O. Rock. 1992. Isolation and characterization of the p-ketoacyl-acyl-carrier protein synthase III gene ifabH) from Escherichia coli K-\2. J. Biol. Chem. 267:6807-6814.
225 Tsukamato, N., L Fujii, Y. Ebizuka, and U. Sankawa. 1994. Nucleotide sequence of the aknA region of the aklavinone biosynthetic gene cluster of Streptomyces galilaeus. J. Bacteriol. 176:2473-2475. Tunac, J. B., B. D. Graham, W. E. Dobson, and M. D. Lenzini. 1985. Fermentation by a new daunomycin-producing organism, Streptomyces insignis ATCC 31913. Appl. Environ. Microbiol. 49:265-268.
Vara, J., M. Lewandowska-Skarbek, V. -G. Wang, S. Donadio, and C. R. Hutchinson. 1989. Cloning of genes governing the deoxysugar portion of the erythromycin biosynthesis pathway 'm Saccharopolyspora erythraea. J. Bacteriol. 171:5872-5881.
Verwoert, 1.1. G. S., E. C. Verbree, K. H. Van der Linden, H. J. J. Nijkamp, and A. R. Stuitje. 1992. Cloning, nucleotide sequence, and expression of the Escherichia coli fabD gene, encoding malonyl coenzyme A-acylcarrier protein transacylase. J. Bacteriol. 174:2851-2857.
Voet, D., and J. G. Voet. 1990. Biosynthesis of lipids, p. 268-288. In D. Voet (ed.). Biochemistry, John Wiley & Sons Inc., New York.
Wagner, C., K. Eckardt, W. Ihn, G. Schumann, C. Stengel, W. F. Fleck, and D. Tresselt. 1991. Biosynthese der Anthracycline: eine Neuinterpretation der Ergebnisse zur Daunomycin-Biosynthese. J. Basic Microbiol. 3:233-240.
Wagner, C., K. Eckardt, G. Schumann, W. Ihn, and D. Tresselt. 1984. Microbial transformation of aklanonic acid, a potential early intermediate in the biosynthesis of anthracyclines. J. Antibiot. 37:691-692.
Wagner, C., K. Eckardt, D. Tresselt, W. Ihn, G. Schumann, and W. F. Fleck. 1981. Leukaemomycin-geblockte Mutanten des Streptomyces griseus und ihre Pigmente. Zeitschr. Allg. Mikrobiol. 21:751-760.
Weiss, R. B. 1992. The anthracyclines: Will we ever find a better doxorubicin? Semin. Oncol. 19:670-686.
Weiss, R. B., G. Sarosy, K. Clagett-Carr, M. Russo, and B. Leyland-Jones. 1986. Anthracycline analogs: The past, present, and future. Cancer Chemother. Pharmacol. 18:185-197.
Wurtele, E. S., and B. J. Nikolau. 1990. Plants contain multiple biotin enzymeszdiscovery of 3-methylcrotonyl-CoA caborxylase, propionyl-CoA carboxylase and pyruvate carboxylase in the plant kingdom. Arch. Biochem. Biophys. 278:179-186
226 Yanisch-Perron, C., J. Vieira, and J. Messing.1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 33:103-119.
Ye, J., M. L. Dickens, R. Plater, Y. Li, J. Lawrence, and W. R. Strohi, 1994.Isolation and sequence analysis of polyketide synthase genes from the daunomycin-producing Streptomyces sp. strain C5. J. Bacteriol. 176:6270-6280.
Ylihonko, K., J. Hakala, J. Niemi, J. Lundell, and P. Mântsâlâ. 1994. Isolation and characterization of aclacinomycin A-non-producing Streptomyces galilaeus (ATCC 31615) mutants. Microbiol. 140:1359-1365.
Ylihonko, K., J. S. Tuikkanen, S. Jussila, L. Cong, and P. Mântsâlâ. 1996. A gene cluster involved in nogalamycin biosynthesis from Streptomyces nogalater: sequence analysis and complementation of early-block mutations in the anthracycline pathway. Mol. Gen. Genet. 251:113-120.
Yoshimoto, A., S. Fujii, O. Johdo, K. Kubo, T. Ishikura, H. Naganawa, T. Sawa, T. Takeuchi, and H. Umezawa. 1986. Intensely potent anthracycline antibiotic oxaunomycin produced by blocked mutant of daunorubicin-producing organism. J. Antibiot. 39:902-909.
Yoshimoto, A., T. Oki, T. Takeuchi, and H. Umezawa. 1980a. Microbial conversion of anthracyclinones to daunomycin by blocked mutants of Streptomyces coeruleorubidus. J. Antibiot. 33:1158-1166.
Yoshimoto, A., T. Oki, and H. Umezawa. 1980b. Biosynthesis of daunomycinone from aklavinone and s-rhodomycinone. J. Antibiot. 33:1199-1201.
Zunino, F., A. M. Casazza, G. Pratesi, F. Formeili, and A. DiMarco. 1981. Effect of méthylation of aglycone hydroxyl groups on the biological and biochemical properties of daunorubicin. Biochem. Pharmacol. 30:1856-1858.
227 Appendix A
Table of bacterial strains used
228 Strain Description Reference/source
Escherichia coli JM83 E. coli Genetic Stock Center Escherichia coli ET12567 dam', dcm' MacNeil et ai, 1992 Streptomyces lividans TK24 Hopwood et ai, 1985 Streptomyces peucetius - Arcamone et al., 1981 (ATCC 29050) Greinetfl/.. 1963 Streptomyces sp. strain C5 - McGuire et al.. 1980 (ATCC 49111)
Table A.I. Table of bacterial strains used
229 Appendix B
Table of plasmid vectors used
230 Plasmid vector Description Reference
pUC19 2.686 kbp; Amp'; E. coli cloning vector Yanisch-Perron et ai, 1985 pWHM3 7.20 kbp; E. coli - Streptomyces shuttle vector constructed from Vara et ai, 1989 plJ486 and pUC19; HC; Thio' pANT735 pUC19 8.62 kbp; 5.90 khp Xhol - EcoRI fragment from pANT 121 into This work pUC19 (contains the daunorubicin PKS genes dauE, dpsA, dpsB, dpsC, dpsD) pANT755 pUC19 10.15 kbp; pANT 121 derivative that contains all PKS genes Rajgarhiaand Strohi, 1997 except dpsD which is truncated pANT841 pUC19 2.746 kpb; pUC19 derivative DeSanti unpublished data pANT782 pWHM3 17.96 kbp; 8.08 kbp EcoRl fragment from pANT121 into pANT Rajgarhia and Strohi, 1997 779 containing daid the activator gene along with all the daunorubicin PKS genes) pANT785 pWHM3 16.58 kbp; 6.69 kbp £coRl fragment from pANT755 into pANT Rajgarhia and Strohi, 1997 Wro 779 containing daul the activator gene along with all the daunorubicin PKS genes except dpsC and dpsD) pANT 788 pWHM3 17.35 kbp; 7.46 kbp EcoRl fragment from pANT754 (37) into Rajgarhia and Strohi, 1997 pANT 779 containing daul the activator gene along with all the daunorubicin PKS genes except dpsD) pANT 797 pANT849 8.43 kbp; E. coli - Streptomyces shuttle vector where expression Rajgarhiaand Strohi, 1997 of genes cloned into polylinker is driven by SnpR - activated snpA promoter; HC; Neo’^ pANT 857 pANT849 7.97 kbp; E. coli - Streptomyces shuttle vector where expression DeSanti unpublished data of genes cloned into polylinker is driven by SnpR - activated snpA promoter; HC; Thio"^ (Continued)
Table B. 1. List of plasmid vectors used. Plasmid Vector Description Reference
pKK840 plJ2925 4.1 kbp; containing 1.4-kbp PCR-amplified aphI gene from Kwak, 1992 plJ6122 pANT152 pUC19 10.36 kbp'; 7.67-kbp A/?/?! fragment from phage P7 in Ye e/ al., 1994 pUC19 pANT235 pUC19 9.2 kb pb; pUC 19 containing a 6.48-kbp BamHl-BglU DNA Ye et al., 1994 fragment from Streptomyces sp. strain C5 carrying the dnrl- hybridizing region pANT716 pUC19 6.41 kbp; pUC19 with 3.74-kbp Bcl\ fragment from pANT121 Rajgarhia and Strohi, 1997 ligated into the BamH\- EcoRl sites pANT726 pUC19 7.09 kbp; pUC 19 with dpsBC\ dpsD replaced by aphi cassette Rajgarhia and Strohi, 1997 pANT740 pUC19 6.26 kbp; pUC 19 with c//7.s-5, partial dpsCD with aphI cassette Rajgarhiaand Strohi, 1997 pANT749 pUC19 7.38 kbp; 4.65-kbp Ay;/7l-£coR1 fragment from pANT152in Rajgarhia and Strohi, 1997 pUC19 pANT750 - 12.77 kbp; 4.89-kbp Kpnl -Hindlll fragment from pANT235 into Rajgarhia and Strohi, 1997 pANT749 pANT751 16.74 kbp; 9.54-kbp EcoRl -HindlU fragment from pANT750 Rajgarhia and Strohi, 1997 into pWHM3 (has daul, doxA, dauv, dauU, dpsG, dauZ, dnmT, and dauH) pANT752 pANT841 6.45 kbp; 3.74-kbp Bell - fcoRI fragment cloned into BamHI- Rajgarhia and Strohi, 1997 &'oRl -digested pANT841 pANT753 5.83 kbp; 600-bp BamWl-Kpnl fragment deleted from pANT752 Rajgarhia and Strohi, 1997 pANT754 5.06 kbp; 1 36-kbp Sstl-Kpnl fragment deleted from pANT752 Rajgarhia and Strohi, 1997 pANT755 pUC19 10.15 kbp; 3.13-kbp Bell- £coRl fragment from pANT753 Rajgarhia and Strohi, 1997 ligated to 4.33-kbp EeoRl-Bell fragment from pANTI21 and EeoKl -digested pUC19 (reconstructs all PKS genes except dpsD) (Continued) Plasmid Vector Description Reference
pANT756 pUC19 9.38 kbp; 2.3-kbp 5c7I - EcoRI fragment from pANT754 ligated Rajgarhia and Strohi, 1997 to 4.33-kbp £coRI - Bcl\ fragment from pANT121 and Æ'coRl - digested pUC19 pANT767 - 15.23 kbp; 8.08-kbp EcoRI fragment from pANT121 into Rajgarhia and Strohi, 1997 pANT765 (contains all daunorubicin PKS genes of pANTl2l) pANT770 - 10.78 kbp; 1.6-kbp Bam\\\ fragment from pANTl52 into Rajgarhia and Strohi, 1997 pANT235 (reconstitutes dauH gene) pANT771 15.29 kbp; 8.09-kbp /scoRI-Z/mdlll fragment from pANT770 Rajgarhia and Strohi, 1997 into pWHM3 pANT776 - 3.17 kbp; 442-bp A7/ol-5c/l fragment from pANT235 into Rajgarhia and Strohi, 1997 pANT84l (gives intact dauG gene) pANT777 - 5.37 kbp; 2.62-kbp fragment from pANT776 into pANT84l Rajgarhia and Strohi, 1997 (yields intact dauG and daul genes) U>NJ pANT778 - 5.33 kbp; 1.36-kbp BamW\-Bcl\ fragment deleted from Rajgarhia and Strohi, 1997 pANT777 (leaves intact daul activator gene) pANT779 pWHM3 9.78 kbp; 2.73-kbp EcoK\-Hind\\\ fragment from pANT777 into Rajgarhia and Strohi, 1997 pWHM3 pANT780 pWHM3 9.85 kbp; l.37-kbp EcoR\-Hindlll fragment from pANT778 into Rajgarhia and Strohi, 1997 pWHM3 pANT781 pWHM3 7.50 kbp; 540-bp EcoRl-Hindlll fragment from pANl'776 into Rajgarhia and Strohi, 1997 pWHM3 pANT783 - 15.77 kbp; 8.08-kbp EcoR\ fragment from pANTl2l into Rajgarhia and Strohi, 1997 pANT78l pANT784 - 17.93 kbp; 8.08-kbp EcoR\ fragment from pANTl2l into Rajgarhia and Strohi, 1997 pANT780 (Continued) Plasmid vector Description Reference pANT 1112 pANT841 5.21 kbp;2.5 kbp S.v/I fragment from pANT121 with the dpxD This work gene cloned into S.s7l digested pANT841 pANTlllS pANT797 10.18 kbp; 1.87 kbp EcoRi fragment from pANTl 114 cloned This work into £coRl digested pANT797 correct orientation gives dpsD expression driven by SnpR - activated snpA promoter in a E. coli - Streptomyces shuttle vector; neo*^) pANT 1116 pANT857 9.84 kbp; 1.87 kbp £coRI fragment from pANTl 114 cloned into This work £coRl digested pANT857 (correct orientation gives dpsD expression driven by SnpR - activated snpA promoter in a E. coli - Streptomyces shuttle vector; Thio*^) pANT 1117 pANT841 5.59 kbp; 2.90 kbp EcoAl III - £coRI frament from pANT735 This work into smal - £coRI digested pANT 841 (contains dpsC and dpsD) pANT 1118 pANT797 11.17 kbp; 2.74 kbp Cla\ - EcoRl fragment from pANT 1117 This work into CVal - EcoRl digested pANT797 {dpsC and dpsD driven by the SnpR - activated snpA promoter in a E. coli - Streptomyces shuttle vector; Neo") pANT 1120 pANT857 10.85 kbp; 2.88 kbp Bglll fragment from pANT 1118 into Bglll This work digested pANT 857 (correct orientation allows dpsC and dpsD to be driven by the SnpR - activated snpA promoter in a E. coli - Streptomyces shuttle vector;Thio"^) pANT 1121 pANT857 10.85 kbp; pANT 1120 digested with Kpnl site introduced in This work dpsD T4 DNA polymerase blunted and religated to give disrupted dpsD (Contains only an active dpsC driven by the SnpR - activated snpA promoter in a E. coli - Streptomyces shuttle vector with Thio’') (Continued) Plasmid Vector Description Reference
pANT786 - 16.54 kbp; 6.69-kbp fcoRl fragment from pANT755 into Rajgarhia and Strohi, 1997 pANT780 pANT787 - 14.39 kbp; 6.69-kbp EcoR\ fragment from pANT755 into Rajgarhia and Strohi, 1997 pANT781 pANT790 pANT841 3.53 kbp; 817-bp Nco\-BamH\ fragment from pANT235 into Rajgarhia and Strohi, 1997 pANT841 pANT791 pANT797 9.10 kbp; 642-bp HincnU-Ec(A7\\l fragment from pANT790 Rajgarhia and Strohi, 1997 into HindlW- and Hpa\- digested pANT797 pANT795 - 7.10 kbp; EcoRl-digested pANT765 end-filled with Klenow Rajgarhia and Strohi, 1997 fragment and religated; HC; Neo' pANT796 pANT841 5.34 kbp; 1,331-bp Kpnl-Mlul fragment from pANT849 into Rajgarhia and Strohi, 1997 pANT841 u>ro LA Appendix C
Maps of plasmid vectors constructed.
236 P stI SphI H indlll
dpsB
ori(ColEl)
SstI
dpsC ampR
dpsD
Kpnl EcoRI SphI SphI
Fig.C.l. MapofpANT716
237 EcoRI WamHI SstI BamHI SstI PstI dpsC aphi
pANT726 dpsB
7080 bps
PstI Hindlll EcoRI ori(GolEl)
ampR
Fig.C.2. MapofpANT726
238 PstI [SphI iH in d lll
ori(ColEl)
pANT740 6263 bps
SstI
PstI /
EcoRI SphI SphI
Fig. C.3. M apofpANT740
239 H indlll SphI PstI BamHI LKpnl I .Bell .SphI
ori(CoIEl)
SphI
pANT749
7328 bps BamHI
dnmT
EcoRI
dauH
BamHI
Fig. C.4. MapofpANT749
240 Hindlll SphI iP stI SstI
ori(ColEl) dnmj
EcoRI
pANT750 ...Bell 12227 bps doxA dauH
BamHI
dnmT SphI SphI
BamHI
Fig. C.5. MapofpANT750
241 Kpnl, Bell
ori(ColEl) orfSo ampR * D A Bell tsrR <« / Hindlll /SphI /PstI dauE pANT751
dauH 16548 bps SstI
BamHI
dnmT
doxA
BamHI dpsGu V Bell
SstI SphI SphI
Fig.C6. MapofpANT751
242 SphI P stI .Bell i Bglll
ori(ColEl)
dpsB
ampR
.....SstI dpsC
SphI
EcoRI dpsD BamHI SphI
SphI Kpnl
Fig.C.7. MapofpANT752
243 H indlll SphI PstI .Bell l Bglll
ori(ColEl)
dpsB
ampR
dpsC SstI
SphI
dpsD’
EcoRI SphI SphI
Fig.C8. MapofpANT753
244 Hindlll SphI PstI Bell /.Bglll
ori(ColEl)
pANT754
5082 bps
EcoRI SphI
Fig.C9. MapofpANT754
245 Hindlll SphI PstI BamHI Kpnl SstI lEcoR I I .SphI .SphI
ori(ColEl)
dpsD ampR SphI \ ...SstI
dpsB dpsF PstI ^ BamHI SstI Bglll dpsE Bell SstI dpsA SphI SphI BamHI
Kpnl
Fig.C.10. MapofpANT755
246 .EcoRI ISstl IKpnl IBamHI iP s tI iS p h I SphL iHindlll SphI.
ori(ColEl)
pANT756
Bell 9402 bps
EcoRI SphI
PstI BamHI SstI SstI SphI
Fig.Cll. MapofpANT756
247 K pnl,
rep ori(ColEl) Bell
ampR orf56
Bell aphi
EcoRI SstI PstI BamHI SstI
Fig. C.12. MapofpANT765
248 .Kpnl
ori(coIEl) .Bell EcoRI ampR rep SphI.
SphI orf56 Bell
pANT767 SstI dpsD PstI 15040 bps aphi SstI dpsC SphI .... EeoRI SstI dpsF dpsB PstI dpsE dpsA G \ SstI Bell SstI SphI SphI Kpnl
Fig. C.13. Map of pANT767
249 H indlll SphI PstI
ori(CoIEl)
EcoRI /.SstI £K pnI C BamHI pANT770 10785 bps Bell doxA
dauH
dnmT
BamHI
Bell SphI
Fig. C 14. MapofpANT770
250 H indlll SphI P stI
ori(colEl) dnmJ
ampR
pANT771
10609 bps Bell dauH
BamHI dpsC U
Fig. C.15. Map of pANT771
251 Bell PstI SphI SphI SstI H indlll Bglll BamHI; Kpnl dpsG Sstl. EcoRI.
ori(CoIEl
ampR
Fig.C.16. MapofpANT776
252 SphI Sstl BglII, BamHI.i Kpnlli S s tl# EcoRI 1
pANT777 am pR 5710 bps
S stl
SphI ori(ColEl) dnmj
I PstI iSphl H indlll
Fig.C.17. MapofpANT777
253 Kpnl Sstl, EcoRI.
Sstl ampR pANT778 4472 bps daul
SphI
PstI
ori(ColEl) PstI SphI Hindlll
Fig. C.18. MapofpANT778
254 Kpnl. Bell
rep
ori(ColEl)
orf56 Bell
ampR
tsrR
EeoR Ij Bell Sstl I dpsG Kpnl I BamHI I H indlll Bglll] SphI Bell PstI SphI PstI Sstl SphI SphI Sstl Bell
Fig. C. 19. lVIapofpANT779
255 K p n l. Bell rep
ori(ColEl)
orf56 Bell
ampR pANT780 9658 bps tsrR
EcoRI Bell Sstl Kpnl dnmJ Hindlll SphI Sstl PstI SphI Sstl PstI
Fig. C.20. Map of pANT780
256 K pnl.
Bell rep
ri(ColEl)
orf56 ampR B ell
tsrR
dpsC
EcoRI i B ell Sstl j Hindlll Kpnl; SphI BamHI PstI Bglll Sstl SphI Fig. C.5îî^Map of PANT781
257 KpnL B ell
ori(CoIEl) or£56
EcoRI
/ PstI
Sstl.... pANT782 ....SphI
= Bc I BamHI I Bglll SphI iBam H I IKpnl EcoRI
SphI Fig.C.22. MapofpANT782
258 K p n l, B ell
ori56
pANT783 ? Bell LSphI PstI 15583 S stl I B glll gB am H I iK p n l p s t I \ EcoRI \ SphI SphI Kpnl BamHI SphI BamHI Sstl SphI B ell
Fig. C.23. MapofpANT783
259 K pnl, B ell
ori(ColEl) orf56 ampR
/ P s t I P s tI ... pANT784 S s tl.... 17741 bps
:\K pnI i Sstl
Fig. C.24. M apofpANT784
260 K p n l, B ell
ori(ColEl) orfSb
C Pstl SphI.... pANT785 ...PstI
\ Bell SphI SphI / S stl BamHI I SphI tB c II I B glll IB am H I iK pnl iSstl EcoRI PstI BamHI
Fig. C.25. Map of pANT785
261 EcoRI
BamHI Saul
Sstl X bal /P stI /.H in d lll
pANT786 ....Clal dauE 16356 bps crf56
Sstl
BamHI AsuII A vril EcoRI Sspl Xmnl
Fig.C.26. MapofpANT786
262 K pnl. B ell
orf56
Bell pANT787 .SphI ,-LPstI 14989 bps Bell ... SphI Sstl t B g l l l S K pnl p s t I \ EeoRI SphI SphI
SphI Sstl
BamHI Bglll SphI Bell
Fig.C.27. MapofpANT787
263 K p n l, B ell
ori56
6 PstI pANT788 ...PstI
Sstl 17181 bps Sstl
SphI
|\ SphI dpsBdpsC A Bell SphI I B glll Bell {Kpnl Bglll I S stl EcoRI SphI SphI
Fig. C.28. Map of pANT788
264 Hindlll SphI P stI ori(ColEl)
.SphI dauZ
pANT790 3526 bps dnmT
à\ BamHI % Kpnl \SstI EcoRI
Fig. C.29. MapofpANT790
265 .Kpnl
orfSo pANT791
8918 bps
BaPiHI dnmT Sstl dauZ P-snpA
SphI i P s t I iS p h I iH in d lll iSstI Bglll EcoRI
Fig. C.30. Map of pANT791
266 K pnl,
rep ..Bell ori(ColEl)
pANT795
6949 bps ampR orf56
Bell
aphi
S stl P stI BamHI Sstl
Fig.C.31. MapofpANT795
267 ,H in d lll [SphI iP s tI S .Hindlll I iS stI I §. Bglll I ftEcoR I .SphI oriCColEl) P-snpA
pANT796 4014 bps
Kpnl Sstl EcoRI
Fig. C.32. Map of pANT796
268 .Kpnl Bell. rep
ori(GolEl
orf56
Bell ampR
aphI Sstl / PstI ^ P-snpA BamHI Sstl snpR Hindlll Sstl Bglll EcoRI SphI
Fig.C.33. MapofpANT797
269 Hindlll Xhol I Notl
ori(ColEl)
MluI pANT735
8620 bps
Kpnl / Bsml BamHI Notl
Fig.C34. MapofpANT735
270 H in d lll PstI Bell iB g lII / Bam HI
ori(ColEl)
dauC
pANT1112 5216 bps
BamHI
Fig. C.35. MapofpANT1112
271 Hindlll PstI Bell B glll .BamHI LSmal iK p n l iS s tI I.EcoRI
ori(ColEl)
..Smal
ampR
Kpnl dpsD
Smal
Sm al dpsC
BamHI Il Smal iB clI I Bglll BamHI Smal
Fig. C.36. Map of pANT1114
272 ori(ColEl)
orfSo
pANTlllS ampR L... Seal
10160 bps Sstl .-
M lul ^ H p a l f=SpeI I Sstl P-snpA dpsD iB glll EcoRI
Clalf BamHI S p h I: EcoRI I S stl I Kpnl I BamHI | Fig. of pANTlllS
273 .BamHI
ori(ColEl)
oriDO pANTine Clal 9841 bps
P-snpA \M luI H pal iS p e l Clal I IH in d lll SphI I fSacI EcoRI I I B glll Sstl I EcoRI Kpnl Bell BamHI Fig. C.38. Map of pANT 1116
274 .H indlll [SphI iP s tl I .Clal I iBclI I i B g l l l I |/. BamHI
oriC olE l)
pANTlin
5595 bps
EcoRI
Fig.C.39. MapofpANTlll?
275 Bell
rep
ori(CoIEl)
B ell... orfS6
ampR Sstl pANTlllS PstI aphi 11172 bps
Sstl Mlul Sstl Bglll EeoRI snpR StuI dpsD P-snpA dpsC
Clal Bell S tuI B glll BamHI
Fig. C.40. Map of pANTlllS
276 BamHI Smal Smal Smal
ori(GolEl)
orfSo Bell
pANT1120 10855 bps
Bell
H indlll P-snpA SstI Bglll EcoRI EcoRI ;f Smal B g lll/ Kpnl BamHI Sm al Sm al Smal BamHI Smal
Fig.C.41. MapofpANT1120
277 BamHI Smal Smal Sm al
ori(ColEl)
ori56 Bell
pANTim 10851 bps
Bell
i\H indIII P-snpA I SstI \B glII EeoRI EcoRI I B g lll/ BamHI Smal Smal BamHI Smal
Fig. C.42. Map of pANTI 121
278 BamHI Xhol
-snpA pANT817 orfSo
8 2 0 3 bps
Clal
PvuII
Bglll SphI BamHI
Fig. C.43. Map of pANT817
279 BamHI Xhol
N otl pANT849
5343 bps
TsnpA
S p h l^ / orf56 EcoRI // B g lll/ M lul
Fig. C.44. Map of pANT849
2 8 0 Appendix D
Mass spectral data for compounds analyzed
281 353 0
39 0 0 0
371 0 Feudo C 36 0 0 0
33 0 0 0
30000*
27 0 0 0
24000*
5 21 0 0 0 *
r o s 15000
12000 *
9 0 0 0 '
6000*
3000*
160 240 3 00 360 4 80420 540 m/z. amu
Fig. D. 1. Mass spectral data for feudomycinone C (Data generated by API-LC/MS) 369 45000-
42000-
39000-
36000-
33000-
30000-
27000- Feudomycin D
K 24000-
c 2 1 0 0 0 -
Kl 00Ul 18000-
15000-
12000 - 264 908 9000- 309
6000- 848 750
3000- 177 946 445
200 300 400 500 600 700 800 900 m/z. amu Fig. D.2. Mass spectral data for feudomycin D (LC/MS). (Data generated by APl-LC/MSJ 309
2 2 0 0 0 *
2 0 0 0 0
16000
16000
351 14000*
g 12000
10000 t o 3 2 3 s 8 0 0 0 3 3 6
6 0 0 0
4 0 0 0 Agl>cone uf feudomycin U
113 1 3 0 2 9 5 2 0 0 0 72 » 3 69
J- —r — I— 6 0 120 180 2 4 0 300 360 540 m/z. amu
Fig. D.3. Mass spectral data for feudomycin C (MS/MS of fragment with molecular mass 516). [Data generated by API-LC/MS] 2! 4
3 6000-
3 3000-
300 0 0
27000
336
& 21000 202
S 16000
2 17 K) 15000- 00LA
12000 - 309
9000 290 Aglycone of feudomycin D 189 6000- 174 256 135 271 3000- 146 161 234 121 351 107 59 370
80 120 160 200 240 2 8 0 320 3 6 0 m/2 , amu
Fig. D.4. Mass spectral data for feudomycin C (MS/MS of fragment with molecular mass 369). [Data generated by API-LC/MS] AA
5 2eS
4 BflS* /
4 4e5
4 OoS-
379 0 3 6 eS
^ 3 , . S f I 2 Be5
N) 2 4e5- S 2 OeS
1 6 0 S* 3 2 3 .0 I 2eS
8 084* 360 8
4 0*4
-Y l. />■ T u 180 240 300 360 420 4 8 0 540 mil, amu
Fig. D.5. Mass spectral data for Akianonic acid. [Data generated by API-LC/MS] 3 .OeS* 37
2 .6eS
2.4eS -
2 .2eS
2 .OeS
I.BeS
3 6 0 J oor o *o I.OeS AA 306.8
200 300 400 SCO 600 0 8 0 0 9 0 070 m/z, emu
Fig. D.6. Mass spectral data for akianonic acid from S. lividans TK24(pANT782) [all PKS genes]. [Data generated by API-LC/MS] 2.105 3 3 0 2 OeS- I.SeS I)csinc(h\)Ak)aiioiuc acid 1 8e5-
1.7g5 320 8 y 1 6e&
1 5o5
1.405-
1.3o5
1 2o5
1 le S
V 0o5
K) 9 0e4 00 00 8 0o4
7 Oo4-
6 Oo4
5.004-
4 Oo4
3 .004-
2 084
1.084
-r 2 4 0 3 0 0 360 420 540 m/z. amu
Fig. D.7. Mass spectral data for desmethyiaklanonic acid from S. lividans TK24(pANT785) [dpsCD deleted]. [Data generated by API-LC/MS] 5 6o5 AA
5 2ûS
4 8e5
4 OeS
379 0 3 6o5*
a " 3 2o5- s
2 4e5 K) NO00 2 OeS-
\ 6o5- 323 0 1 2eS
360
4 064
100 240 300 3 60 420 480 540 m/2 , emu
Fig. D.8. Mass spectral data for akianonic acid from S. lividans TK24(pANT788) [dpsD deleted]. [Data generated by API-LC/MS] AA
7 Û0e4
6.3 0 e4
§ 5 6064
360 6 £
170.
44 3 .2 s
260 0
7 0003-
ISO 240 3 0 0 3 60 rn/z. amu
Fig. D.9. Mass spectral data for akianonic acid from S. lividans TK24(pANT785) complemented with dpsC and dpsD. [Data generated by API-LC/MS] 3 4e5 3 2 ) 0
3 ?o5-
3 OeS- 366 8
2 BeS
2 6eS
2 465 3 4 0 .6
2 2eS
2 OeS & 4 0 0 0 ^ 1 BeS 1 I 1 OeS
1 4eS lO VO 1 2oS
1 DoS
6 Oo4 Ocsmcihylaklanonlcecid 6 Oo4 3 8 3 .r 4 Oo4 379 0 2 De4 8 2 66 8 \ 428,2 WVA(WvwWi \jiW /i v JV j 2 6 0 320 340 360 380 400 420 440 mil, amu
Fig. D.IO. Mass spectral data for akianonic acid from S. lividans TK24(pANT785) complemented with dpsD. [Data generated by API-LC/MS] 1 20E-»06
400 Î2
1.00E*06 AA
8 OQE+05
aOOEiOS 381 S8
4COE+05 §
2 00E*05
0.00E«00 X Lju U so too ISO 200 2S0 300 3S0 400 4S0 SOO
-2 00E«05
Fig. D. 11. Mass spectral data for authentic akianonic acid made using propionic-[3,3,3-d]]acid. [Data generated by API-LC/MS] 9 00E«06 400 32
8 00E*O6 AA
7 O0E*06 /
6 00E*06
5 00E«06
4 00E«06 w wVO 3 00E*06
2 00E«06 381 98
1 00E«06
0 00E«00 50 100 ISO 200 250 300 350 400 450 500
-1 ooE«oe
Fig. D.12. Mass spectral data for akianonic acid from in vitro biosynthesis using PKSs from S. lividans TK24(pANT782) and dj-propionyl-CoA. [Data generated by API-LC/MS] 6 00E«O5
369 2ü
SEK 43 5 OOE+05
4 00E«05
3 00E*05
325 0 w s 200E*OS
1 OOE'05
00OE«O0 I, . 1^ I iluJuil.L i-eJL -ll JuA 50 100 150 200 250 300 350 400 450 500
-1 00E«05
Fig. D.13. Mass spectral data for SEK 43 obtained from Dr. C. R. Hutchinson. [Data generated by API-LC/MS] 1 60E*05 Î12 : SEK43
1 40E^05
1 20E«05
1 O0E*O5
8 00E«04
6 00E«04 u>
4 00E«04
200E *04
0O0E4O0 300
Fig. D.14. Mass spectral data for SEK 43 from in vitro biosynthesis using PKSs from S. lividans TK24(pANT782) and [2,2,2- djJacetyl-CoA. [Data generated by API-LC/MS] 25 5.265
4 .BeS*
4 4eS"
4 .OeS
3 .BeS
3 2eS
. 2 BeS
B 2 4e5
2 Oo5
1 6o5
1 2e5
180 0 300 3 6 0 4 20 4B0 54024 m/z. amu
Fig. D.15. Mass spectral data for authentic aioesaponarin II. [Data generated by API-LC/MS] 2 S I a
7.5e6- AlocMpon«rm II
7.OeS / G.SeS
G.OeS
5.5e5
5 OeS-
M 4 565- & i? 4 .OeS I — 3 SeS
3 OeS
2 SeS
2 .OeS
I SeS
1 OeS
S.0e4
I JL.aL. ■><> — I— T— ‘ V — (— I------— I— ISO 240 3 0 0 360 420 46 0 540 m/z. amu
Fig. D.16. Mass spectral data for aioesaponarin II produced by Streptomyces sp. strain C5VR5 and Streptomyces sp. strain C5VRI0. [Data generated by API-LC/MS] IMAGE EVALUATION TEST TARGET (Q A -3 ) /
%
1.0 I j f l a M mil 2 .0 l.l 1.8
1.25 1.4 1.6
150m
V 6> / /A PPL IE D ^ 1IVWC3E . Inc 1653 East Main Street ^ Rochester, NY 14609 USA Phone: 716/482-0300 -^=~- = Fax: 716/288-5989
O 1993. Applied Image. Inc.. Ail Rights Reserved
7