- NOTED .AEMORANDLM “°TE? BRISTOL LABORATORIES A. GOUREVITCH UNIT OF BRISTOL-MYERS COMPANY J. Lederber “arch 26, 19 FROM e _ DATE on 76— 1955 J. Lein Sew &ntiblotica Sersening “rogram - TO i ee eect eee SUBJECT __conguttrntentv ArranPaneat eae CL. vi Toe, Here are gome further, I hope rather better organized, comments on the -uast assignment. {sn elso sending an outline of these resarke but I think some more casual ciscussion way helo fill in the outline. [ am st111 feeling my way aroun: your thinking on thie croblem anc trying to orient my Own eo these will not in ony senne be the last words, ne if there was any chotee that they ever could be. first of all, it seems to me that there are three acrroaches to tne whol® cuestion. “he firat is the sort of muddled trial and error aat-up which has been vratty much the pattern uo until now. A second might be the theoreticn} eoproach, the one which -ekemean has sesoused on the theory, but not very mich in oractice, which ic breed on the cueetion °"hat is an antidiotie?". The third ie so-ething of a fuaion hetveen these two and I might eall an operational aporoach which te based on the retroe-ect on the enccessen tant have been made and te an attemrt on operational zrounte to try and fine whet furtner vsthe there nave been. The ovsrational anpronch ie, 1 think, reasonably well exemelified by your own program on looking for new kings of Actinomycetes by #lective isolation. Sow I must admit that I will be somewhet cresudieed in terms of tne theoreticn] aporosch to it. This is rerhane not the one that ylelda the obvious, immediate cividends but is obviously the only basis on which one knows what one ie actusily soins. Unfortunately I do not think thet there is enough informntion cregently evailsble to nnewer the question "“Ynat is an antiblotic?", that le, fron the coint of the orsaniem thet is making it. You seam to ve pretty well convinced that they do not function as antibiotics under neture) cirecumeteances. I heve seen onty a limited aiscussion of thie cuestion and whet I nave seen noes not arceer to be very convincing one way or snother. Can you refer ne to oublished sources for this concert? It heroly seame to me aufficient to fay thet anslyree heve been made on certrin “ince of soil or at certein seasons and that aetectadle levels of antibiotic have not been found on those cccagions. The potentiality of croducing antibiotic eubet«nces in sufficient cuantities may be shared by the laboretory on the one hend end rother acecial nabitats on the other. Im fact, I wonder if there isn't a etienificent cueation in the notion thet if it does hrve a factual bestia thet antiblotic croduetion ia limited in the soil but I met eay that I am very mich in the dark on this whole euestion. & more curiour yoint would be, if we followed throuch on the notion thet antibiotics function se auch under some conditions, +hiet icinsa of organieams might they in fact cataronize. It seems to me that the most likely competitors to which such antibiotics mieht be cirected are other sveciss of the Actinomycetes themselves, These, after all, ere the dominant soil flors; they «re the ones which wre moat effective comretitors for the common «colomrical niche. I wou’: also euote the anslogies «ith the bacteriaieine of other orgeniens of which, for examole, the colicina, are the moat outetendine tor Jo Leim woos ? Mareh 26, 1955 New Antibiotics Sereening Srogran - Coneultantship Progran there anveares to be a coneiderable degree of, let's say genus epecificity, in the inhibitory action of these compounds one from the other. From thie voint of view antibecterial activity mizht eimply wean that the by-product based on metabolic similarity of other bacteria to the Actinomycetes which were the immediate targets with resvect to the adactive value of antiblotic function. If thie is the case, then the mere fact that a given Actinomycete species apoears to be dominant in » given habitat may be an index that it has some kind of competitive nechaniem and if we are ever to eet oureelvas on a sound baaie it might be worthwhile doing sone exploration on thie point. It seems to me that it could be most readily teeted by mixing soil samples, culturing them together, testing that type which eventually does sredominate from the mixed culture againet the others that hed been originally inoculsisd. There may be some real point in doing this kind of expariment since it may show the way by weans of which the moat active antibiotic-producing organiams can ve autome tically selected for. Now there may be nothing at all in this but if there were it might lead to protective methods not only in fresh isolation of active srecies but also of the weans of selecting for them in the mutation experiments, I would also be curious to learn if efforteare ever made to examine antibiotic treduction enough fron the ltodividual pure cultures of Actinomycetes which are isolated from sotl sanples but from crude inocula of the entire sample without murification. All of thie comment will make sense at all only in the event thet antibiotic activity 1a meenineful in the normal life history of the Actinomycete. A@ a subsidiary question, have you ever concerned yourself with what it is thet deternines the natural biological profile with respect to the Actinomycetes in eny sol) cnmple. I have the iapression that the Actinomycetes on the one hend are a rather broadly confident erour of organiems but on the other thet one is just about like the next so that there would not be very much to choose from among the grouv of them with rescect to specific shyeiological characteristics. Have you any impressions on this? The obvious factors that come to mind in suggesting the determination of the blological composition of a soll sanple would be the chemical nutrition, the chysical environnent, tencerature (ite variability and sdeture in the same) but there leaving -eneral chance factors sdrift for the whole question of antagoniem on the one hand by other Actinomycetes and on the other by other microorganiame as well. I'm sot proposing that we have a complete solution to all the problems but I do suggest thet any rrogranm in such as your own hinges on these in such a way that one should slwaye be thinking about thes. ‘nis has particularly to do with your question of the isolation of new and hithertea perhass overlooked forme by the methode that you hove indicated. Regrettably tne baale information thet one would like to heve ia essentially lacking becaure of the orientation and the research on Actinomycetes that has been effective over the cast several years. At any rate the one suggestion that doses come out of this is that if antibiotics do potentially function «s such in the soil, then the likely targete for their action are perhana other Actinomycetes and that this micht in some way be used as the Basie for selection sroesdures for the imprevement of antibiotic function. Now the alternative croposal that antiblotica are sigcellaneour metabolites is, of course, not really exclusive of the first notion because in any event that's whet they are. One would however substitute metabolitea which function es antibiotics in the soll for tne notion that theee sre setabolites without any known function whatsoever and whose antibiotic activity made them be presumed to be more or less by way of an necident. 1 find this rather difficult to believe, frankly, but im the absence of any rarticuler evidence one way or the other it's an anprogeh that should hardly be neglected. Ae fr na Te. Latin cessese 3 March 26, 1955 Hew Antibioties Screening Frogran - Consultantehip Arrangement IT ean see it makes very little difference «hich way you think about it for the purvose s of designing the mutation program. In any event antibiotics are metabolites and insofar as they are metabolites their production is assumed to be under genstic control. We, fortunately, already heave the precedent of Nelner's work which does show that even in what apveare at firat sight to be a cualitative sense, the abllity to form antibiotica is aubjeact to mutational variation, If then, thin orogram ie essentially a search for miscellaneous and bizarre nsw chemical structures, conceivably it ia a mistake to preselect them first on the baeie of antibiotic activity. Instead there comes to mind what one might call the Hneetrick appraoch, that is, one which recognizes the wide diversity of organic comcounde which may be produced by Actinomycetes. One might then attemnt to screen Actinoaycetes not in terns of their vcroduciag en antibiotie in the first instance but in terse of their producing large quantities of new organic compounds, Theee, thea, might be screened in some rather broad way for their potential antibiotic activity: this crocedure would only be economical if there were some easy method of firat determining whether large amounta of unicue, new metabolites are being produced ctherwine it would, of course, be more economical to look for antibiotic activity in the filtrates in the first instance, keeping in mind however thet ineofer as thie is the standard orerating procedure of the other companies it would be less likely to uncover quite new cevelopmente. I, unfortunately, cannot readily suggest nethode of detecting, and mors narticnlerly characterizing, even moderete or large smounte of new unknowns, The only thing thet comes to mind immediately would be some technicue of chromatograhy of filtrates. ‘ith little experience, I suppose, most of the familisr compounds would be spotted auc discarded. ‘fhe question sight be how to locate a compound of unknown composition. You'll need some more help from the chemiste on a point of this sort but I might think on the one hand if chrometography on glaes or some such compound with any fseneral method of texting organic compounds, cherring or perhace (well this is possibly prohibitively excensive) the use of autoradiogravhy to test the accumulation of verticuler snote. Off hand though, I would have to admit that thie kind of reverse screening would be unlively to be even nearly as efficient as the dirset method of antibiotic detection but it conceivably may engzest wn anele that hoan't occurred ta me yet. March 31, 1955 Continuation Hi again Joe, I think I had left off with the remark that one might have a more straightforward approach ta the whole cuestion of detection of new producte if theee were in the first inetance detectable without regard to their biological activity so that thie could be left for an entirely objective and iscependent evaluation. I only mean to leave you thie as something to think sbout because, frankly, I have no immediate suggestions on how this could actually be sceomplished in # screening program, Yow my next remark might be considered new. Regardlees of whether the antibiotics function as much in the soll there is no question but that they hove to be considered metabolites be they intermediaries in ths eynthesie of other compound or themselves happen to be the end vroduct of a considerable sequence. I don't think it very such matters one way or another and I'm not sure one could really oreclagly define the difference between thece two notions. At any rete if these are J. Lain coccreee & March 25, 1955 Sew Antibiotics Seresening Program ~ Consultentehip Arrangement metabolites it then, of course, follews that their production te under genetic control and we already have the actual exverimente of Kelner which show this voaribility. I em still conelderably purrled ae to what Yelner's experiments ectuzlly sean, thet is to say, what hepvrens when you discover a mtant which now produces an antibiotic which had net previously been detected. I showld say in the first plaee that I am extremely skentical of being able to aenage what I might call true genstic novelties. ‘low, of course, such mutations, we micht call then necmervhic or watations fer new enzymatic apecificities, very likely, in fact inevitably, must have occurred fa evolution. However, the tepreseion that I have gotten from all of the work in thie direction that T have done has been that if these occurred at all they oust be extrenely rare and thet most of the sutations which one now observes can be considered as falling into one of two categories: the firet ie the removal of the effective functioning of on envyre or enzymes or second © reverse mutation, thet is to say, one which puts the origins! systex beck in balance, Now on truely aprieri grounds that ia exactly whet one would exnect. It would be wuch more difficult to synthesize or create an entirely new system than it would be on the ons hand to indues some aneecific dame in the oxiating system or on the other, and this rperhape even sore rarely, to repair danege that head been done creviously. Thie is a very different kettle of fish from mutations whose effect would be to create entire novelties in the formation of enzyres. ow 1 con't think that Actinomycetes are the most likely organtems in which to look for evidence along this peint, but it way turn out that they will be in fact useful in just that connection. I would voint however to the fact that, for exavle, in eucrose negntive strains of Escherichis Soli I at least ~ bave found that essentially it te impossible to obtein emtents which would ferment this sugnr. Om the other hand, the lactose negative varisnt atrain thet had been obtained fron wild tyre in their turn have, asa rule, a well-recognivable cazacity of reverse mutation or of renponding to suprreseive mutations at doer loci. This king of exveriment hac been sone many times and in essentially every cage where an apparently new fuaction hae beea noticed 1t would be at least as reasonable and in the majority cf cases mech wore reasonable to srgue that there had simply been the reparation of vrevious damace either by reveree mutation or by other changes in the baekground gena tyve which had 2 comrorable effect. Thie argument is based on only two voiats however: one the aoriori notion that the formalation of en entirely new process would be exbected to be » such less srobable event than either of the other two that I nad mentioned and second the actual fact thet in sont cases naturally occurring deficiencies tend to be far sere stable then those “hich are obteined by Inboratery mutation, the excertion to this rule being moet readily exclained in terms of previous matetion. For example, in the “elmonella grou it te not at «ll infrequent to find tryptochane-devendent orgenieme which in thelr turn will readily revert, s0 to speak, to eutonuxotrozhs, ty-ea not recuiring tryptochane. ‘Since the ceneral sattern of nutrition im the Salmonella group, however, is autcauxotrochy it is moat convenient to regard these acquisitions of tryotovneane synthesia aa effective reverse sutations rether then the novel exoerience of first learning how to make tryrtochane. Having etated this suspicion that the development of the abllity to orecuce antibistics by mutation is not likely to be a completely novel pathway, let me indiente what I consider eome of the wore likely alternatives to be. The first is, ef course, thet we mast distinguish between syntheele and excretion. Wow we heve ne convenient method of determining the synthetic cavneity of a given strain of Actinomyeete execent in terme of the materia) which is coured out inte the mediue and on the one hand we night have te congicer the fact thet ciffereat media and therefore different cena tynee are going to heve a profound quantitative influence on the extent of synthesis and on the other that the very fact of accumiation and excretion aight be a fuaction which ia quite different from tant for the initial synthesis. The necond voas#ibility ie already tmolicit in vhat I and J. Leim secceee § Mareh 28, 1955 New Antibiotics Sereening Frogram Conaultantaship Arrangement said before and one that is, of course, obvious that these are not new functions in any senee that are being brough, out by cutation but the simcle quantitative improvenent on any grounds of a function which ie already present. Thies senin would be something that would be rather difficult to settle definitely withont a coneiderable amount of experimental work although one can imagine use of tracer recovery exoeriments, isotope dilution, end whet have you, in order to detect minimal amounts of previous synthesis, Finally, while I would consider 1t a most isprobable ewent indeed that s true evolutionsry novelty in terns of the development of a new pathway, usually involving ® number of enzymatic steve, could de controlled in the laboratory, it is, of course, conceivable that there will be mutations which will influence rather slight changes in specificity and might therefore lead to derived products which on one ground or another aight be more desirable than the original. I have in mind hers, for exasple, the diversity of related penicillins. There the balance between one kind of penicillin and another is certainly under genetic control and it iseven conceivable that there might be mutational changes in the srecificity of a single enzyme which vould result in a qualitative sense in the ability to form some new and distinctive antibiotics. I must state, however, that ae far as I know there are no good clear cases of changes in enzymatic specificity as the result of a laboratory mutation in any meterial. Most matations thet have been examined have concerned the conditions or the extent or the mere ability to which the organism can respond by making 2 given enzyme or not and chenges in ececifleity of enzymes have been extremely rare, tf indeed there are any clear-cut cases. The best example is perharve Moss and avis on the pantothenate synthetase in Jecherichia golf where tempersture-sensitive allelos have been picked up ano have been shown in turn thet temperature-sensitive enzymes ;*e involved. It would be a very difficult thing te determine whether this meant that thy gene change had directly resulted in the change in the kind of ensyee produeed or whether the orgenian had originally possessed potentiality of making both the normal and the teaverature- sensitive enzyme where the temperanture-sensitive allelo stock had only the one. Ag a final wey in which mutation sight influence the ability to form an antibiotic, it would seem to me that far and away the most likely event would be thet mutations could interpose setabolic blocks which would result in the accumulation of intermediates in the syntheais of other compounds, these intermediates not ordinarily deing detected, and it is of course quite conceivable that some of these intermediates will have interesting antibiotic activity. On these grounds, however, one would be tempted to eachasize in one's search stocks which had already oroduced a variety of complicated end products in the hove that there will be a sufficient wariety of well-used pathways thet blocks at eny one of a number of points might lead to the accumulation of any of a number of distinctive compounds. That was one of the points that I hed in aind when I suzreated using organisens which were already producing lerge amounts of some distinctive end yreduct. You are, of course, taking the chance thet whatever that end product my be that none of the intermediates leading to its formation will be of any particular intereet. That vas another reason for vanting to have other meane of detecting those stocke which are producing large amounts of complicated molecules aince whether or not the end products are antibiotic intermediates in their formation might or wizht not de. Gn thewnole, theese theoretical preconceptions give ma the iepreseion at least that the entire operation asy be very chancy. The basie reason why we suggest going on with it fe the hove that in the very process of continuing along this aporoach that other Je Loin wercseves & Harch 26, 1955 New Antibiotics Sereening Progran ~ Consultantshio Arrangerent notions of the concepts will crop up which will in their turn be of nore immediate use. If indeed the major effect of autational blocks is going to be to interrupt the aynthesis of compounde whoes pathways are already well established in that strain, then the obvious conclusion is that it would not be worthwhile to concentrate too much effort OR any one strain since the sarticular pathway which is sost vulnerable or most effective inone atrain may not in fact have any intermediate station which would be of any particular interest to you. The kind of thing that one might most profitably expect from mutational programs ie, enfortunately, not the most technically feasible. I would be willing to bet as an example that by mutational operations one could obtain from an aureomyein-producing orgeniem derivatives which would produce either terramycin or tetracycline in view of the close chemical relationship of these compounds one to the other. What I would be most concerned about is how exactly one would readily detect euch changes. Of course here one Ase the advantaze with aureofaciens of the Fellow coler of the aureomycin which wieght lend to the detection of colorless wariants which still retain antibiotic activity. I wonder, in fact, if this kind of experiment hes not really been done. Ferhaps this is something I shouldn't delve into too closely. There would not as a rule however be any convenient method hy which one could pick up euch minor deviatione in the character of the end product and at least for basic reasons of anemotherapy there would be only a very few eituetions in which such deviant end producte would, in fact, be atrikingly useful, though of course there's alwayea the chence of augmentation of activity, reduction of toxicity ete. These however are just the things which would be most inconvenient to sereen for on any lerge seale. This seems a convenient point at which to leave what I have cal!_4 the theoretical approach and take up what I might now call the overational approach which is based essentially on e purely impirics] consideretion of the details of vrocedures vhich in the vast have given useful entibloties and whet modifileations might be expected to give some further diversity. Yow for this purpose I will sisply take at its face your statenent that further random teating is not the procedure that your comonny wants to geo into. If am mot well enough acquainted with the detaile of the prograa to be able to comment on your statement that this would be an uneconomical procedure. The queation its then basically how to vary the standard procedure in auch a way ee to have the highest orobability of picking up new naterials. I was particularly happy to note thet you made no mention of the orimitive crose-atreaking method ae the initial ater in entiblotie screening. My thinking on this matter is somewhat limited by my lack of appreoiation of what the limiting factors are in your screening program. I have no notion, for example, of what vreportion of your effort is devoted to the collection of enztles of the creportion to the prelivinary tests, what fraction of samples are eufficiently crowising in producing any antibiotic to be submitted for further tests and no om and so ferth. I think any operational analyaie would attest to review the componente of the screening crocedure on the one hence giving some notion of the amount of labor which f# axvended in then and the other the efficiency or the screening value to the proportion of the cultures which Pace and so forth that do come through there. For that reason until I can ret that iad of background many of my suggestions are bounc to be sonewhat srecious. Now soat of your first assignaent and discussion to me was an secount of your interest in the use of especial mutritional and similar selections as a aeans of restricting the range of crganiens that you were considering. I am still, as I cerhape already Consultantship Arrangement indicated at sufficient length, somewhat uncertain of exactly how much vork you have to eliminate in order to accompliech something uesful by this approach. If there were indeed a reiatively small sumber of very prevalent tyres whose nutritional behavior wee well known and which it would be desirable to exclude selectively in the first instance then I would any there would be no question whatever in the use of these sveelalizsed proceedures, If the previously known and unwanted tyves are Aowever a considerable scattering of physiological forma of which some are likely to be included in this new, selective precedure and some are not, then I am mot sure that you would be aceomplishing very wach in your specialired sample ae compared to 2 random cample ef the entire Actinonycete vepal ation. In all I should say that your notion of special selection is extremely well worthwhile trying if only for the reaeon that it represents an approach thet, as far as I know, hae not been used before. However, I would be lending to some effort to validating the srocedure that is in seeing what criteria can be set ur to learn if it 1a worthwhile eontinuing with it or not and thia may indeed be rather difficult. One of the things that I am afraid of that you may ran inte ie that you will be developing selective sedia for isolating a very few species and you will very promptly be running through the chance of whether they are or are not croductive, and once a vary small number of isolates obtained in this way had been tested the very snseificity of your selection procedure may exclude not only the bulk cf wawanted, well-known cultures but siso new novelties which might otherwise have been of some interest to you. Instead you woulc be reieoleting essentially the same few organiems again and agnin ang again. uzactly how you would know when you had reached that stage at any given point I de not rnow. , For these reasons I think it would be well worthwhile to do ea preliminary validation of these selective procedures by teking a fairly sized group of cultures which represent the thinge that you don't want to have and seeing how they, in fact, would respond if they are put through the mill through your selective procedure. What I ama little suspicious may harpen ie that you will be picking ur not only specialized tyres but new, Feveres mutations,and what have you, of the existing populations. At any rate I would ve particularly leery of any selective medium which was toe rigorous in excluding too many cifferent types. I don't know what the cost accounting factors are, es I cuestioned above, in your operations but I would suppose that if you are limiting more than 9C-954 of distinctive esecies, shell I call them, that your selection may be too rigorous. I think it would be more meaningful too if the agents that you were asing for your selection were vhysiologically reaningful beenuse I think this would have the largest ehance of picking ur a new kind of ecological wariaty or eet of varieties of your Actinomycetes. The kinds of variables for use in the initial screening that have the moet obvious bearing would be pH, temperature, selenity anc Ridex potentiel, and I wonder how much work has been done based on the original screening of Actinonycete sooulationa on medin which are not optimal ia these elementary terns for the