Biogenesis of an Antitumor Antibiotic Protein, Neocarzonostatin

A study of the biogenesis of the antitumor protein antibiotic neocarzinostatin (NCS) was undertaken. The production of NCS, as well as the growth of Streptomyces carzinostaticus in a production medium, was sensitive to puromycin, chloramphenicol, and actinomycin D. However, when a 12-hr culture in production medium was transferred to a nongrowth medium consisting of a phosphate buffer with Mg2+ and Ca2+, rapid NCS synthesis and liberation occurred. NCS production in this medium was no longer sensitive to actinomycin D, but was sensitive to puromycin and chloramphenicol. The conversion of a precursor NCS to an active form was shown to occur in this medium. Subcellular analysis suggested that NCS synthesis occurred by a mechanism similar to that of protein synthesis by membrane polysomes.

A study of the biogenesis of the antitumor protein antibiotic neocarzinostatin (NCS) was undertaken. The production of NCS, as well as the growth of Streptomyces carzinostaticus in a production medium, was sensitive to puromycin, chloramphenicol, and actinomycin D. However, when a 12-hr culture in production medium was transferred to a nongrowth medium consisting of a phosphate buffer with Mg2+ and Ca2+, rapid NCS synthesis and liberation occurred. NCS production in this medium was no longer sensitive to actinomycin D, but was sensitive to puromycin and chloramphenicol. The conversion of a precursor NCS to an active form was shown to occur in this medium. Subcellular analysis suggested that NCS synthesis occurred by a mechanism similar to that of protein synthesis by membrane polysomes.
The antitumor antibiotic neocarzinostatin (NCS; 7, 10) is a protein with a molecular weight of 11,000 (H. Maeda, personal communication) produced by Streptomyces carzinostaticus. It is highly active against various experimental tumors, including L-1210 (3). Studies on the mode of action of NCS revealed that this antibiotic selectively inhibits the synthesis of deoxyribonucleic acid in Sarcina lutea (11,12) and HeLa cells (6).
This study was undertaken to reveal the mode of biogenesis of NCS in a liquid culture of S. carzinostaticus. In the biogenesis of the peptide antibiotics so far known, nonribosomal peptide synthesis is generally accepted (1). Recently, Ingram (5) reported that nisin (molecular weight, 7,000), or a related compound, is synthesized through a ribosomal system. Since NCS has a considerably larger molecular size than other peptide antibiotics, the mode of biogenesis is of great interest, as well as its chemical structure, which is being clarified (H. Maeda, C. B. Glaser, and J. Meienhofer, in preparation). In the present study, one of the main objectives was to determine whether the biogenesis of NCS is similar to that of polypeptide antibiotics such as bacitracin (13), gramicidin (15), and tyrocidine (9), which do not go through a ribosomal system, or to that of nisin (5), diphtheria toxin (14), and botulinus toxin (2).
NCS production in nongrowth medium. To analyze the process of liberation of NCS into the fluid, a nongrowth medium (NG medium) was used in which no growth of mycelium occurred. The NG medium was a 0.067 M phosphate buffer of pH 7.2 containing 10-2 M Mg'+ and Ca2+. After the organism had been cultured for 12 hr in production medium, mycelia were spun down and production media were discarded. NG medium was added to the mycelia. In the NG medium, no detectable growth of the mycelium, as determined by protein concentration, was observed up to 12 hr of incubation, but the liberation of NCS into the medium (100 ug/ml) could be detected.
Bioassay for antibacterial activity. The activity of NCS was measured by the conventional paper-disc diffusion method with S. lutea PCI 1001 as a test organism. Standard NCS used in this experiment was a gift from Kayaku Antibiotic Research Co. Ltd., Tokyo, Japan. Before the activity of NCS in the production medium and the NG medium was measured, other antibiotics (puromycin, chloramphenicol, and actinomycin) were removed by dialysis against 0.001 M acetic acid solution.
Assay for mycelial growth. Mycelia were disrupted by grinding with quartz sand in 10 mm tris(hydroxymethyl)aminomethane (Tris)-hydrochloride buffer (pH 7.4) containing 10 mM MgCl2 and 60 mr KCl. The supematant fluid obtained after centrifugation at 5,000 rev/min for 20 min was used for protein estimation by the method of Lowry et al. (8).
Radioactive amino acids. The 14C-labeled amino acids used in this experiment were randomly labeled chlorella hydrolysate, obtained from the Institute of Applied Microbiology, University of Tokyo. The specific activity was 8.0 mCi/mmole. 3I-L-alanine (3.03 mCi/mmole) was purchased from Daiichi Pure Chemical Co. Ltd., Tokyo, Japan. Amounts of 2 or 20 ,Ci of labeled amino acids in 0.5 ml of phosphatebuffered saline were added to 100 ml of production medium at the desired time, and incubation was continued for an appropriate period.
Purification of radioactive NCS (10). To determine the specific incorporation of labeled amino acids into the NCS molecule, culture filtrates were first precipitated with saturated ammonium sulfate; the precipitate was dissolved in water, dialyzed against 0.001 M acetic acid solution, and freezed-dried. This crude preparation was dissolved in water and passed through a column of Sephadex G-50; bioactive fractions were combined and lyophilized. This preparation was further chromatographed on a carboxymethyl (CM)cellulose column. Either stepwise or gradient elution of NCS with acetic acid and sodium acetate was made, and 0.5 ml of the highest antibacterial fraction was neutralized with K2CO8. The insoluble salt formed was removed by centrifugation, and a sample was dried on a metal planchet for counting.
Density gradient sedimentation analyses. Streptomyces mycelia in NG medium were collected by centrifugation and disrupted by grinding with quartz sand as described above. They were then extracted with a solution containing 10 mm Tris-hydrochloride buffer (pH 7.4) with 10 mr MgCl2 and 60 mm KCl in 0.25 M sucrose. Quartz sand, intact cells remaining after quartz sand treatment, and cell walls were removed by centrifugation, and the resulting cell-free extract was further fractionated into soluble and particulate fractions by centrifugation at 105,000 X g for 120 min. The yield in the particulate fraction was determined by measuring the total protein contents from the disrupted cells. Samples of 0.3 ml of particulate fraction were layered on 3.6 ml of a 10 to 30% linear sucrose gradient in 10 mm Tris-hydrochloride buffer (pH 7.4) containing 10 mM MgCl2 and 60 mm KCI, which was then overlaid on a bottom layer which contained 0.6 ml of 50 to 70% sucrose in the same buffer. After centrifugation at 23,000 rev/min for 120 min, three-drop fractions were collected from the bottom of the tube. Each sample after dilution was used for the determination of absorbance at 260 and 280 nm and of radioactivity. The radioactivity was determined after precipitation with an.equal volume of 10% trichloroacetic acid solution in the presence of one drop of 1% bovine serum albumin. The precipitates were washed with cold 5% trichloroacetic acid solution and then with a mixture of ethanolether-chloroform (2:1:1, v/v). The precipitates were ANTIMICROB. AG. CHEMOTHER. then dissolved in dilute ammonia solution and dried on a metal planchet for counting.

RESULTS AND DISCUSSION
Time course of NCS production in relation to the growth of S. carzinostaticus. A 20-hr seed culture of S. carzinostaticus was inoculated into the production medium, and the culture flask was incubated at 27 C with shaking. Under these conditions, the organisms formed bundles of aggregated mycelia after overnight cultivation. The results presented in Fig. 1 reveal that NCS was produced and released into the medium by the organisms at the logarithmic phase of growth (judged by protein content). The production of NCS was almost parallel to the mycelial growth until the mycelium reached maximal growth (20 mg of protein/ml) at 24 hr. The pH of the culture medium reached 6.8 at maximal production. This 3to 4-hr delayed production curve is quite similar to that of diphtheria toxin (14) but quite different from that of botulinus (2).
Effect of actinomycin D on mycelial growth and production of NCS. Puromycin, chloramphenicol, and actinomycin D were each added to different production media at zero time to give a concentration of 40 pg/ml. In addition to inhibiting the production of NCS, all of these antibiotics inhibited the growth of the organisms completely, although their concentrations were below the lethal level.
The effect of actinomycin D (40 jig/ml) at  different time periods of cultivation on growth and NCS production (Fig. 2) amined in the production medium. tion at 8 hr resulted in growth inhil about 90% as measured by mycelial at 36 hr. Mycelial protein content at I less than 10% of that at 36 hr. In oth there was almost no apparent differen mycelial protein content at 8 and 36 actinomycin D was added at 8 hr. I 30% production of NCS was found at spite of the sustained presence of actinom When the drug was added at 12 hr, th was almost one-tenth of the control, production of NCS measured at 36 hr wi 60% of the control. The addition oi mycin D at 24 hr did not have any i] effect on either mycelial protein or N duction.
In these experiments, it is evident addition of actinomycin D at the begi exponential growth inhibited NCS pr( but drug addition at the late logarithn did not inhibit NCS production.
This observation was further exarr pulse administration of labeled amino various times during logarithmic gro test the incorporation of labeled ami into the NCS molecule, 20 ,uCi of 3Hper 100 ml was added to the production every 4 hr from zero time to 16 hr. A of pulse-labeling followed by contin mentation for 40 hr, the culture was ce at 3,000 rev/min for 15 min to precipitate the c mycelia. From this supernatant fluid, NCS was purified by precipitation with 60% ammonium sulfate, molecular seiving with Sephadex G-50, and CM-cellulose column chromatography.
The results in Table 1 clearly show that the 2 most effective incorporation occurred from 4 to 8 hr. However, active NCS was never detected at 8 hr in either culture fluid or mycelia. These 8 results may indicate the formation of an NCSrelated protein (precursor protein) through messenger ribonucleic acid (mRNA) at the 4-4 to 8-hr period and eventual conversion into radio-0 active NCS as described below.
Critical period for 14C-amino acid incorporation into NCS-related protein. The incorporation of production "'C-amino acids into NCS-related protein and pry growth conversion into NCS were further studied by ated after short-period pulse-labeling. Except for the 1-hr us, actino-period of labeling, the experimental design was dture. Ar-the same as that illustrated in Table 1. Pulse-4rves show labeling for 1 hr was done at 0 to 1, 4 to 5, 8 to 9, 12 to 13, and 16 to 17 hr after inoculation in the production medium. The results are shown in Lmycelial Fig. 3, in which specific incorporation of 14C was ex-amino acids into the NCS molecule determined [he addi-after 36 hr of incubation is shown by histograms. bition of The results clearly indicate that the greatest in-I protein corporation occurred during the period of 8 8 hr was to 9 hr, followed by that at 4 to 5 hr and 12 to er words, 13    shown at 4 to 5 hr, without an appreciable increase in mycelial protein content. Synthesis of mRNA for NCS production. The data presented in Fig. 2 illustrate the fact that addition of actinomycin D at 12 hr allowed almost 60%' production of NCS at 36 hr, irrespective of the restricted mycelial growth. This may indicate that mRNA for NCS-related protein was already synthesized before 12 hr; however, mRNA for mycelial protein seems to be synthesized at a much later stage. In the experiment shown in Fig. 3, NCS-related protein was formed even as early as 4 hr.
The effect of actinomycin D was further investigated in the following experiment. Actinomycin D was added to the production medium at 4, 6, 8, and 10 hr, and at 12 hr the mycelia were transferred into NG medium without actinomycin D. The latter medium did not allow the increase of mycelium protein content. Results obtained by this procedure (Fig. 4) indicate that the addition of actinomycin D to the production medium at 8 hr (i.e., just at the beginning of the logarithmic growth phase) allows the production of NCS in the NG medium.
These results show (Fig. 4) that liberation of NCS occurred in this NG medium as early as 6 hr without further supply of extracellular amino acids. The activity of NCS in this NG medium [100 jg/ml in the control (Fig. 4)] is almost one-third of that in the production medium. These result may indicate that the mRNA for NCS was made before 8 hr.
Effect of various inhibitors on NCS production at the actinomycin D-insensitive phase. After cultivation of the organisms in the production medium for 12 hr, precipitated mycelia were transferred to NG medium and incubated as described above. The effect of various concentrations of puromycin, chloramphenicol, and actinomycin D on NCS production in NG medium was examined (Fig. 5). An inhibitory effect on NCS production was found with puromycin and chloramphenicol ( Fig. 5A and 5B), but not with actinomycin D (Fig. 5C). Although radioactive amino acids are incorporated into the inactive form of NCS (NCS-related protein) at an early stage in the production medium, the inhibitory effect of puromycin and chloramphenicol in Fig. 5A and SB indicate that further new protein synthesis may be involved in the activation process of NCS-related protein. It may be suggested that the organisms transferred into NG medium have already finished the synthesis of mRNA for NCS or NCSrelated protein (precursor protein; Fig. 3), whereas the conversion of precursor protein into active NCS in NG medium requires further protein synthesis, as evidenced by tk effect of puromycin and chloramphe later stage (actinomycin D-insensiti Intracellular mechanism of NUO From the above experiments, it nr cluded that NCS-related protein (I} sor) is synthesized in the mycelium in the production medium, and the to the active form occurs in the NG i requires protein synthesis de novo the latter process further, the fol experiments were conducted.
The first was conducted to clarify t the protein excreted from myceliui medium. For this purpose, '4C-a (2 ,Ci/100 ml) were added at 0, 4, at the inoculation of Streptomyces in duction medium, and cultivation tinued. All cultures were harvestec The mycelia were transferred to NG incubated for a further 4 hr. The 1 radioactivity into the NG medium w with a sample of culture fluid at 0. The specific biological activity of N4 examined. All three experimental r trated in Table 2 show that 70% o activity liberated into the medium pure NCS, which was proven by iso] active NCS. From this result and th, it is evident that the conversion fro: protein (bioinactive) to NCS (bi( curred in the NG medium without fu of amino acids. In the next experiment, the site o duction in the mycelium was invest amino acids (2 ,uCi/100 ml) were after inoculation, and incubation tinued for 12 hr. The collected nm divided into three portions and ii NG medium: one portion receives chloramphenicol/ml, and the oth Control ceived none. After incubation periods of 0.5 and 4 hr without chloramphenicol, and after -30 4 hr with chloramphenicol, each culture was centrifuged in the cold to separate the mycelia. These mycelia were disrupted by grinding with quartz sand and were extracted with 10 mm Tris-hydrochloride buffer (pH 7.4) containing 10 mM MgCl2 and 60 mi KCl in 0.25 M sucrose.
Quartz sand, undisrupted cells, and cell walls 3 6 were removed by centrifugation at 7,000 rev/min for 60 min, and the resulting cell-free extracts hibitors added were further fractionated into soluble and paricrograms per ticulate fractions by centrifugation at 105,000 x duction in NG g for 120 min. This particulate fraction was used phenicol. (C) for sucrose density gradient analysis to separate membrane fractions from microsomal fractions. When the particulate fraction from the myie inhibitory celium was subjected to linear sucrose gradient nicol at this centrifugation by the procedure of Hallberg and ive stage).
Hauge (4), the results illustrated in Fig. 6 were S synthesis. obtained. The heaviest fraction in Fig. 6 was iay be con-thought to be a membrane-associated fraction, ICS precur-and the average ratio of A280/A260 was 0.8. before 8 hr The light fraction, assumed to be ribosomal, conversion possessed an average A280/A260 ratio of 0.5. medium and After cultivation for 4 hr, almost 45% of radio-. To clarify activity in the membrane fraction was lost ( Fig.  Rlowing two 6B). The radioactivity lost from the membrane fraction was almost equal to that liberated into he nature of the medium (Table 2). However, the radioactivm into NG ity in the ribosomal fraction was almost unalmino acids tered. When a sufficient amount of chloramnd 8 hr after phenicol to inhibit peptide synthesis was mainto the pro-tained for 4 hr, the radioactivity of the memwas con-brane fraction was not lost and NCS production I at 12 hr. was inhibited (Fig. 6A). These results suggest X media and that the synthesis of NCS-related protein (preiberation of cursor protein) was carried out on this membrane as examined fraction which contains the polysome system. 5 and 4 hr. Consequently, from this and the other findings CS was also mentioned above, it appears reasonable to deresults illusduce that the site of NCS protein synthesis in of the radiowas due to  the Streptomyces mycelia must be closely associated with the membrane fraction.
The question of whether or not membranebound ribosomes are directly involved as the actual site in the NCS synthesis is not resolved by these results, although the inhibition by actinomycin D indicates necessary synthesis of mRNA for NCS production.
The above results can be summarized as follows: (i) NCS production by S. carzinostaticus has been divided into two different stages. The first stage is de novo protein and mRNA synthesis, as shown by the effective incorporation of the labeled amino acids into NCS-related protein and by the susceptibility to the action of actinomycin D at this stage (Fig. 2) which indicates the necessity of mRNA for NCS production. The second stage is the activation stage of NCSrelated protein into NCS, and this stage was insensitive to the action of actinomycin D, but was sensitive to the action of puromycin and chloramphenicol (Fig. 4). These results may indicate that early synthesis of mRNA for NCS or NCS-related protein is necessary before a substantial increase of mycelial protein content can be attained.
(ii) NCS production is inhibited in the second stage by the late addition of chloramphenicol and puromycin, although NCS-related protein has been synthesized by this time (0.5 to 12 hr). These results may reveal that the biosynthetic mechanism for NCS requires a further activation stage which is sensitive to chloramphenicol and puromycin.
(iii) When a 12-hr-old culture was incubated in NG medium, activation of NCS-related protein and liberation of NCS into the medium were observed. Using this system, together with subcellular fractionation and isotopic techniques, we attempted to determine which fraction of the mycelium is directly associated with NCS peptide synthesis in S. carzinostaticus. The results showed that the membrane-associated fraction was responsible for the peptide synthesis. Bodanszky and Perlman (1) suggested that peptide antibiotics are made by a different type of mechanism from that of protein. Biosynthesis of peptide antibiotics is generally insensitive to inhibitors of protein synthesis such as puromycin and chloramphenicol. Ingram (5) reported that a polypeptide antibiotic produced by Streptococcus lactis, a compound related to nisin, was synthesized by a ribosomal mechanism, as the incorporation of amino acids into the antibiotic molecule was inhibited by puromycin and chloramphenicol. The present results may also suggest that NCS-related protein and NCS are synthesized similarly at a site closely associated with the membrane, by use of mRNA and a mechanism of biosynthesis different from that of bacitracin (13), gramicidin (15), and tyrocidine (9).