Journal of Biotechnology, 7 (1988) 283-292 Elsevier JBT 00302 283 The formation of daptomycin by supplying decanoic acid to Streptomyces roseosporus cultures producing the antibiotic complex A21978C F.M. Huber, R.L. Pieper and A.J. Tietz Antibiotic Development Division, Eli Lilly and Company, Indianapolis, lndiana, U.S.A. (Received 23 November 1987; accepted 9 January 1988) Summary Antibiotic substance A21978C is a complex of compounds having a common cyclic polypeptide nucleus and different fatty acid sidechains. Daptomycin is a semi-synthetic antimicrobial substance derived from the A21978C complex by a very elaborate chemical process. To obtain the daptomycin, the A21978C complex was first isolated from culture filtrates of Streptomyces roseosporus by several resin procedures. The resulting material was then 'blocked' and added to an Actinoplanes culture for deacylation. The protected A21978C nucleus was subsequently isolated by the same procedure as the parent complex and reacylated with the desired fatty acid (decanoic acid). The acylated compound was deblocked to yield daptomycin. This report describes the experimentation undertaken to establish a strategy to supply a very toxic precursor to cultures of S. roseosporus and, thereby, produce daptomycin biosynthetically. Daptomycin; A21978C; Precursing Introduction The production of the antibiotic complex A21978C, by Streptomyces roseosporus NRRLl1379 was originally described by Hamill and Hoehn (1980) and shown to be Correspondence to: F.M. Huber, Antibiotic Development Division, Eli Lilly and Company, Indianapolis, Indiana, U.S.A.
`
`
`1 of 10
`
`FRESENIUS-KABI, Exh. 1011
`
`

`
`284 O ,NH2 .o20 PT- I o / H O , ..CONH2 7--N T T _A.. 1' O ::::~ CH 20 H "k NH I .I.I II ~J. O N--H GO2H ,,._,g° H' -"N'",~ Fig. 1. Basic structure of the A21978C complex. highly active against Gram-positive bacteria. This complex of compounds was demonstrated to have a common cyclic peptide nucleus (Fig. 1) with different fatty acid acyl groups attached by an N-acyl bond (Debono et al., 1980, 1984). Subsequent to the discovery of the natural A21978C antibiotic complex, it was found that substitution of the naturally occurring fatty acid sidechains with de- canoic acid resulted in a compound of superior biological activity (Fukuda et al., 1984; Counter et al., 1984). The latter substance was designated compound LY146032 and then later renamed daptomycin. In order to obtain daptomycin, two biological and three different recovery procedures were required. The processes involved culturing S. roseosporus in primary, secondary and tertiary inoculum development stages for 96 h and then in the producing stage for 140 h. The microorganism was separated from the soluble portion of the culture fluid by filtration. The antibiotic complex was adsorbed and eluted from a resin and concentrated. The free amino function on the complex was 'blocked' with di-tert-butyl dicarbonate and the mixture again concentrated. Actinoplanes utahensis was cultured for 120 h in primary and secondary inoculum development stages. The culture was further grown for 72 h in a stirred reactor and then the concentrated 'blocked' complex was added to the medium for deacylation. After 24 h the spent medium was filtered and the 'blocked' nucleus was adsorbed and eluted from a resin column. The eluate was concentrated, the 'blocked' nucleus was acylated with either the anhydride or halide of decanoic acid and then the protecting group was removed by hydrolysis. The final product was adsorbed and eluted from a resin column and subjected to final purification. In addition to low product concentration, the recovery yields were extremely low. If daptomycin could be made biosynthetically, most of the recovery procedures would be eliminated and overall process yields increased significantly. This report describes a study that resulted in both increased antibiotic concentra- tions and the biosynthetic production of daptomycin.
`
`
`2 of 10
`
`

`
`285 Materials and Methods The organism used in this study was a mutant derived from S. roseosporus NRRL 11379 and designated A21978.65. The inoculum development stages prior to the final reactor have been previously described (Huber et al., 1987). The medium used for the production of the A21978C antibiotic complex contained soybean flour (2.0%, w/v), FeSO 4. 7HzO (0.06%), dextrose (0.75%), potato dextrin (3.0%, w/v), cane molasses (0.25%, w/v), SAG471 silicone antifoam (0.02%, w/v) and water. The medium was adjusted to pH 7.0 with NaOH and sterilized for 45 min at 121°C in a stirred reactor. The reactor was made of 304 stainless steel with an operating volume of 120 1. The reactor had four equally spaced vertical baffles, two impellers with six open turbine blades, and a single open-tube sparger. The reactor was inoculated with 400 ml of the secondary inoculum development stage. The culture was then agitated at 200-400 rpm, aerated a 1 v/v/m and maintained at 30 o C. Biological mass was estimated by mycelial volume and dry weight. Mycelial volume (MV) was determined by centrifuging a known amount of medium at 1200 x g for 10 min and noting the volume of the insohibles in relationship to the total volume of the sample. The dry weight of the mass was estimated by transfer- ring the pellet in distilled water to a tared aluminum weight boat and drying to constant weight at 29 ° C. Respiratory activity was estimated by analyzing the exit gas stream from the reactors by a Perkin Elmer mass spectrometer. Glucose was determined by the Trinder method using glucose oxidase/peri- oxidase (Biodynamics/bmc, Indianapolis, IN). Total carbohydrate was estimated by the anthrone method (Morris, 1948). Inorganic phosphate and ammonia nitrogen were automatically determined by the 'Industrial Method No. 93-70W' (Technicon Industrial Systems, Tarrytown, New York). The concentration of the various components of the A21978C complex were estimated by HPLC. The analytical HPLC column was from ES Industries (C18, 10 cm x 4.6 mm). The mobile phase consisted of 1% ammonium dihydrogen phosphate and acetonitrile (64 : 36, v/v). Results and Discussion The data presented in Fig. 2 and Table 1 are characteristic of the initial experiments with S. roseosporus. The A21978C complex increased with time, but no daptomycin was formed. Oxygen demand reached a maximum at 30 h and then decreased. It was possible that although there was a relatively high concentration of polymeric carbohydrate in the broth, some of that carbon might not be readily available to the microorganism. Thus, more dextrin 700 (1%) was added at 73 h and a slight increase in oxygen demand was observed. After 75 h the oxygen demand decayed further to a steady level at approximately 100 h. Much like the oxygen demand pattern, the mycelial volume also decreased significantly after a maximum had occurred early in the process. It was possible that only the cellular system for catabolizing carbohydrates was decaying. If that was true, the culture could still
`
`
`3 of 10
`
`

`
`286 A 1.0C i o.9C r- E o.8C ~ o.7£ E 0.6C E ~" 0.50 0.40 E a3C ~" 0.2C o.10 x 0 0.000 15 30 45 60 75 90 105 120 135 h Fig. 2. Oxygen demand in original complex medium. readily catabolize other substances (i.e. lipids, amino acids). If such a substance was toxic when in excess, it could be fed at a rate not to exceed the maximum rate for its oxidation. Under those conditions the toxic substance would never accumulate in the culture. Decanoic acid, a very toxic fatty acid, is the desired sidechain for the production of daptomyein. If decanoic acid was supplied to S. roseosporus so as not to accumulate, it might be possible to 'precurse' the A21978C complex and produce daptomycin biosynthetically. The latter achievement would save much processing chemistry. Some antibiotics have been successfully 'precursed' (Higuchi et al., 1946; Moyer and Coghill, 1947; Wolf and Arnstein, 1960). Decanoic acid is a solid at the temperature of cultivation (30 ° C). In order to conveniently add decanoic acid to the stirred reactor, a metabolizable solvent was sought. The original solution was a 5% solution of decanoic acid in an equal mixture of ethanol and water. The resulting solution was fed to the culture at 50 ml h 1 beginning at 40 h post inoculation. The information gained in these experiments is presented in Fig. 3 and Tables 2 and 3. The data clearly indicate that daptomycin could be produced biosynthetically. The concentration of daptomycin in the TABLE 1 CHARACTERISTICS OF CULTURES OF S. roseosporus WITH 1% DEXTRIN 700 ADDED AT 73 h h pH MV Total Glucose PO4-P NH3-N A21978C (%) carbohydrate (mg ml- 1 ) (/~ g ml- 1 ) (/~ g ml- ~ ) complex (mg m1-1) (fig m1-1 ) 0 6.5 6.0 48 6 14.0 63 18 5.5 8.7 44 5 2.0 10 42 7.1 20.0 29 0 0.4 203 66 6.6 18.7 24 0 11.0 146 90 6.0 13.3 29 0 11.0 104 114 5.9 10.6 25 0 5.0 100 138 5.9 11.3 13 0 3.0 100 293 347 387 458
`
`
`4 of 10
`
`

`
`287 1.00 i 090 e" E Q80 ~" 0.70 E 0.60 E ~" 0.50 t-" ~ 0.40 I~ 0.30 ~ Q20 0.10 0.00 140 126 112 h Fig. 3. Changes in oxygen consumption during the initial decanoic acid supplemented culture. A21978C complex approximated 34%. The time the feed was initiated was clearly defined by increased oxygen demand. That the oxygen demand continued to decrease during the remainder of the process suggested that further increases in feed rates were necessary. The decrease in oxygen demand prior to the feed indicated TABLE 2 MEDIUM CHARACTERISTICS DURING INITIAL DECANOIC ACID FED A21978C PROCESS h pH MV Total Glucose PO4-P NH3-N Total (%) CH20 (nag m1-1) (#g m1-1) (#g m1-1) A21978C-like (mg ml- 1) substance (/Lg ml -] ) 0 6.3 7 44 7 19 71 16 5.9 10 40 5 3 32 40 7.35 19 34 0 1 210 64 6.9 12 29 1 11 204 88 6.3 11 27 3 10 156 112 6.2 10 24 6 4 137 136 6.0 8 30 9 4 142 200 306 399 364 TABLE 3 FACTOR COMPOSITION AND CYCLE TIME DURING INITIAL DECANOIC ACID FED PRO- CESS h Factors ( # g ml - 1 ) C 1 C 2 C 5 C 3 LY146032 64 60 71 0 33 36 88 73 105 16 40 72 112 82 133 22 54 108 136 72 109 19 42 122
`
`
`5 of 10
`
`

`
`288 1.00 0.90 )- I o.8o}- g "; o.4o I.- VO.20 ~- 0.10)- 0.00(~ 50 45 o 40 ~ 0 35 E," 30 3_~ 0 15 30 45 60 75 90 105 120 135 h Fig. 4. Oxygen demand and decanoic acid methyl oleate feed rates in a refined process for the production of daptomycin. that the decanoic acid feed could have been initiated earlier. The process was further refined by controlling the pH at 6.5 with ammonium hydroxide and initiating the decanoic acid feed at 28 h. The lipid feed was manipulated in an attempt to maintain constant oxygen demand. In addition to the aforementioned changes, methyl oleate was substituted for the ethanol/water mixture as the solvent for the decanoic acid. The concentration of decanoic acid in methyl oleate was 50% (v/v). The results obtained with the refined system are shown in Fig. 4 and Tables 4 and 5. The data indicate significant changes in the metabohsm of S. roseosporus were caused by the refinement in the process. Oxygen demand, antibiotic and total carbon consumed were markedly increased. The concentration of daptomycin in the A21978C complex was also greatly increased. This type of process was found to be very reproducible, provided all delivery systems containing the decanoic acid were calibrated correctly. Complete lysis of the culture occurred in all cases where the TABLE 4 COMPOSITION OF S. roseosporus CULTURE FLUIDS DURING REFINED DECANOIC ACID FED PROCESS h pH MV Total Glucose PO4-P NH3-N Titer (%) carbohydrate (mg ml 1) (/~g m1-1) (~tg m1-1 ) (/~g m1-1 ) (rag ml- 1 ) 0 6.2 5 46 7 23 71 20 6.4 14 43 6 3 176 44 6.5 21 24 0 1 170 68 6.5 21 23 0 3 221 92 6.5 21 19 0 2 257 116 6.4 19 16 0 2 298 140 6.5 19 14 0 1 347 707 977 1 136 1 282
`
`
`6 of 10
`
`

`
`289 TABLE 5 RELATIONSHIP OF DAPTOMYCIN TO OTHER A21978C COMPONENTS IN THE REFINED DECANOIC ACID FED PROCESS A21978C component (/~ g ml- a) LY146032 C 1 C 2 C s C 3 68 471 79 69 37 51 92 657 98 94 57 71 116 733 96 146 72 89 140 918 105 109 72 78 decanoic acid/methyl oleate solution was fed at rates that caused the fatty acid to accumulate in the culture. To better understand S. roseoporus's inability to readily utilize carbohydrate late in the culture cycle, other substances were tested as carbon sources. When no additional carbon was added to the process, a rapid decrease in oxygen demand occurred at approximately 78 h (Fig. 5). Coincident with that decrease in oxygen demand was a marked decrease in mass as estimated by dry weight (Table 6). The data indicate that both of the decreases in activity occurred when there was sufficient total carbohydrate remaining in the medium. To further confirm the phenomenon of not being able to readily utilize carbohydrate, the same system was continuously fed a total of 2% glucose at a constant rate between 72 and 100 h. The results from that experiment are presented in Fig. 6 and Table 7. Under these conditions a slight increase in oxygen demand occurred at the onset of the glucose feed. The oxygen uptake then decreased as before and the concentration of glucose in culture filtrates increased. To ascertain if the phenomenon was solely related to carbohydrate metabolism, 1% methyl oleate was added to the culture in a batch mode at 72 h. Immediately upon the addition of the ester the oxygen demand rose markedly (Fig. 7 and Table 8). Of great significance was the fact that the dry weight 1.00 I" O.9O ~- °-8°r~, / ............ I o.,oI_,, .--_. ... .... . ................. --. .... E~ .~ 0160|0.50/ t'- ~, - 0.40 030 Ov 0.20 o.10 0.00 0 13 26 39 52 65 78 91 104 117 h Fig. 5. Respiration of S. roseosporus without lipid or carbohydrate additions. 1.200 1.100 1.000 .900 .800 .700 .b .600 ] .500 i .400 .300 .200
`
`
`7 of 10
`
`

`
`290 1.00[ 0.90 I 0-8O f [o7o~ f'o o C g' ~= 050 t'- ~"; o.4o >,,,-- (~ E 0.30 Eo.20 i 0.10 0.00~ 1.200 13 26 39 52 65 78 91 104 h - 1.100 1.000 .900 .800 -.700 .600 ~ i -.500 1 .400 -.300 I .200 117 Fig. 6. Oxygen consumption and respiratory quotient of S. roseosporus to which 2% glucose had been fed between 72 and 120 h, TABLE 6 CHARACTERIZATION OF THE REFINED PROCESS IN WHICH NO CARBON SOURCES WERE FED h pH MV Total Glucose PO4-P NH3-N Dry weight (%) carbohydrate (mgm1-1) (ggml 1) (ggml l) (ggml-1) (mgml 1) 0 6.1 7 42 8 25 85 9.5 19 6.5 20 38 6 3 878 11.3 43 6.5 25 24 2 1 709 12.0 67 6.5 23 16 0 8 890 11.6 91 6.7 20 11 0 9 2205 8.6 115 6.5 9 10 0 27 2716 6.2 TABLE 7 THE EFFECT OF FEEDING 2% GLUCOSE BETWEEN 72 AND 120 h TO THE REFINED PROCESS AS THE SOLE CARBON SOURCE h pH MV Total Glucose PO4-P NH 3-N Dry weight (%) carbohydrate (mg m1-1) (/zg m1-1) (btg ml-1) (Fg ml-1) (rag m1-1 ) 0 6.2 9 42 8 25 91 8.9 19 6.6 20 38 6 2 185 11.1 43 6.6 25 25 2 1 191 11.8 67 6.8 21 18 0 6 371 11.2 91 6.7 20 16 5 4 338 10.4 115 6.5 22 22 11 8 977 9.1
`
`
`8 of 10
`
`

`
`291 1.00 L 0.90~""\ o.801- \ ~. l a70 -' ....... " ........ f. 060 I_ ..... ,... "--. t,"- &T o.4o~ 0.30 I_ I 1 1 1 I 52 65 78 91 104 h 1.200 1.100 1.000 .900 .800 .700~D .600 1 = .500 .400 .300 i .200 117 Fig. 7. The effect of adding methyl oleate (1%) at 72 h on the respiratory activity of S. roseosporus, continued to increase throughout the remainder of the process. When glycerol (2%) was fed to the culture, the response was identical to that observed with methyl oleate. The foregoing data confirm that carbon sources other than glucose are the preferred catabolic substrates for S. roseosporus. Another unique characteristic of S. roseosporus is that it will preferentially use lipid as a carbon source in the presence of high concentration of glucose (Unpubl.). The decay in carbohydrate metabolism and the preferential utilization of lipid in the presence of glucose will be the subject of another investigation. In summary, a biological process was developed to selectively produce daptomy- cin in culture fluids instead of the usual A21978C complex. The process involved feeding a highly toxic substance, decanoic acid, to a S. roseosporus culture that was apparently carbon limited. TABLE 8 CHARACTERIZATION OF THE REFINED PROCESS TO WHICH METHYL OLEATE (1%) WAS FED AT 72 h AS THE SOLE CARBON SOURCE h pH MV Total Glucose PO4-P NH3-N Dry weight (%) carbohydrate (rag m1-1) (/~g m1-1) (/~g m1-1) (#g m] -1) (mg ml- 1) 0 6.2 8 41 7 22 88 9.7 19 6.6 19 38 7 1 155 11.2 43 6.6 27 25 3 1 170 11.7 67 6.7 19 18 1 6 255 10.7 91 6.6 27 11 0 4 278 12.7 115 6.6 15 10 0 2 332 13.3
`
`
`9 of 10
`
`

`
`292 References Counter, F.T., Baker, P.J., Boeck, L.D., Debono, M., Ensminger, P.W., Hamill, R.L., Krupinski, V.M., Molloy, R.M. and Ott, J.L. (1984) LY146032 [N-(n-decanoyl) A21978C nucleus], a new acidic lipopeptide antibiotic: synthesis and biological evaluation. Abstr. 1078, Interscience Conference on Antimicrobial Agents and Chemotherapy, 1984, Washington, DC. Debono, M., Barnhart, M., Carrell, C.B., Hoffman, J.A. and Hamill, R.L. (1980) A21978C, a complex of new acidic peptide antibiotics: Factor definition and preliminary chemical characterization. Abstr, 410 Interscience Conference on Antimicrobial Agents and Chemotheraphy, 1980, New Orleans, LA. Debono, M., Abbott, B.J., Krupinski, V.M., Molloy, R.M., Berry, D.R., Counter, F.E., Howard, L.C., Ott, J.L. and Hamill, R.L. (1984) Synthesis and structure activity relationships of new analogs of the new Gram positive lipopeptide antibiotic A21978C. Abstr. 1077, Interscience Conference on Anti- microbial Agents and Chemotherapy, 1984, Washington, DC. Debono, M., Barnhart, M., Carrell, C.B., Hoffman, J.A., Occolowitz, J.L., Abbott, B.J., Fukuda, D.S. and HamiU, R.L. (1987) A21978C, a complex of new acidic peptide antibiotics: Isolation, chemistry and mass spectral structure elucidation. J. Antibiot. 40, 761-777. Fukuda, D.S., Abbott, B.J., Berry, D.R., Boeck, L.D., Debono, M., Hamill, R,L., Krupinski, V,M. and Molloy, R.M. (1984) Deacylation and reacylation of A21978C, acidic lipopeptide antibiotic: Prepara- tion of new active analogs. Abstr. 1076, Interscience Conference and Antimicrobial Agents and Chemotheraphy, 1984, Washington, DC. Hamill, R.L. and Hoehn, M. (1980) U.S. Patent 4399067. Higuchi, K., Jarvis, F.G., Peterson, W.H. and Johnson, M.J. (1946) Effect of phenylacetic acid derivatives on the types of penicillin produced by Penicilliurn ehrysogenum Q176. J. Am. Chem. Soc. 68, 1669-1670. Huber, F.M., Pieper, R.L. and Tietz, A.J. (1987) Dispersal of insoluble fatty acid and precusors in stirred reactors as a mechanism to control antibiotic factor composition. In: Ho, C.S. and Oldshue, J.E. (Eds.), Scale-up and Mixing of Biotechnological Processes, AIChE Publications, New York, pp. 249-253. Morris, D.L, (1948) Quantitative determination of carbohydrate with Dreywood's anthrone reagent. Science 207, 254-255. Moyer, A.J. and Coghill, R.D. (1947) Penicillin X. The effect of phenylacetic acid on penicillin production. J. Bacteriol. 53, 329-341. Wolf, E.C. and Arnstein, H.R.V. (1960) The metabolism of 6-amino penicillanic acid and related compounds by Penicillium chrysogenum and its possible significance for penicillin biosynthesis. Biochem. J. 76, 375-381.
`
`
`10 of 10