~ sec? Uneammeks Heer af 3 f oy tn t7.3 me eget 1856 Seseerch Report « Bacterial Genetica - Wisconsin - J. mapa ae PE CIMOL meoterial heredity. For the most part, this preceeding aleng the Lim of ont past work, with various collaborators and students concerning then selyeg with different aspects. The detailed work is better swenarized in the enfentifie publications, which are listed separately. Abseched 2 = a + 7 a aecount of tivo main livres of work, having te do, respectively, wit pseterial preteplaste and a prophage~Llinked system of transduction, ihe protoplast work is a new line, and is being followed primarily in the hope that it may lead te e DNAemediated transduction, ar te nev mechanisna of 1 fesien, Im addition, 1+ ie a basis of = blochemical~gens 4. i . wyon, 4S “2; ep mgpg ode 2 bacterial celieuall syntheis. nelentific voerkers connected with this project are: Professor Jd, Lederberg, Principal Investigator. ' $ E Assistant Professor N. 5, Morten; Dr, ©, M, Lederberg, Agsociates, Dr. F, Grakev, De. T. Geskov {Copenhagen} Dr, W. Heumann (Praunschweiz. Germany) Visiting Fellows. De. M, le Morsa. Aesectate to dune 30, 1956. Present address, Dept, Hie~ owls a ot - a 7 “ “ ~ , University of Coloratio Medical. School, Penver. De, 5, G, Beacley, Fellow to dune 30, 1956, Present address, Dept. Uscteriolesy, University of Minnesota Medical School, Minneapolis. ope i, Tino, A. Richter, R, Weight, J, 5s, Clair, 4. Cook, Graduate students and assisteute. Q35 1956 Research Report + Bacterial Genetics - Wiscongin ~ J, Lederbery wo]. Protoplasts, In botanical usage a protoplast is the living content of a cell, lying within the vigid cell wall, Attention was focussed on bacterial protoplasts in 1982 by Neibull's demonstretion that the protoplasts of Fecilius magstertin could be released when the outer wall was dissolved by lysozyme. The proto- plasts appeared as spherical bodies which could be maintained only ina hyper- tonic protective medium, and promptly lysed when transferred to water. These two criteria, spherical shape and osmotic fragility, are the easiest to dis- tinguish bacterial protoplasts from intact cells, In gram-positive (bat not gramensgative) bacteria, moreover, the cell wall is readily demonstrated by simple staining procedures. The protoplasts of B, megaterium have already been the subject of several important studies--directed on one hand to the missing functions of the cell wall, on the other to the use of protoplasts as easily disrupted and extracted biosynthetic units, These results (e.g. Spiegelman and Landman 1956) stimu- lated the hope of comparable experiments with E, coli a) for the use of already well-studied enzymatic mutants, phage, etc., in physiological studies, and b) to test for the uptake of DNA (as genetic fragments) by the wall-less protoplasts, Early in 1956, several investigators became interested in this problem (Repaske and collaborators at Indians; Zinder at Rockefeller Institute) and succeeded in finding special conditions under which lysozyme would dissolve the existing walls of E, coli cells, Simultaneously, in this laboratory, penicillin was found to inhibit wall formation in growing celis, It head long been known that penicillin lysed growing cells; by analogy with Weibull’s results» it was found that in a growth medium containing M/3 suecrese and M/50 Me**, this lysis was averted, and the cells transformed instead into spherical, osmotically fragile 'protoplasts' (Lederberg, 1956), This ob- servation also furnished clear evidence that the antibiotic action of penicillin depended cn walleinhibition: the protoplasts would remain viable only so long as they were suspended in the protective medium, and rapidly lysed (and of course died) when diluted in ordinary medium. On the other hand, if the peni- cillin was removed, the protoplasts gradually reverted to viable bacillary forms, presumably by resynthesis of the normal wall, This hypothesis on the mode of action of penicillin is in excellent accord with Park's finding of the accumulation of celi-wall precursors in penicillin-bloecked staphylococci (cf. Lederberg 19573; Park and Strominger 1957). The synthatic abilities of E, coli protoplasts were then studied: they were found to continue to grow (increase in mags and volume) many fold when S4ncubated in penicillin-broth, but without increase in numbers, or micro~ scopically detectable proliferation. More specifically, the protoplasts were also shown to be capable of synthesizing large amounts of an inducible enzyme, Begalactosidase (Lederberg 1956). Provocative biosynthetic studies on this system are being actively pursued by Professor Spiegelman at the University of Tllinois, Our interests have been concentrated on a) a search for genetic inter~ actions between protoplasts of different genotypes, or between protoplasts and extracts, and b) the biclogy of protoplasts and L-forms, As to the former, while compatible genotypes have been found to mate with either parent a prote- plast, no new forms of genetic interaction (such as recombination betucen Fe x Fe, or protoplasts and extracts) have been found so far. 1956 Research Report - Bacterial Genetics - Wisconsin = J, Lederberg on Poss Tt is difficult to be certain precisely what is new in these findings, except their application to E, coli. Many authors have described the formation of spherical ‘large bodies! under the influc {see a vocaent review by Lisbermeister and Kelloents - 1956 end the one tensive reports by Dicnes and Klieneberger quoted there) and there are suny hints (at least) of the osmotic fragility of the L-forms to which the large bodies are related, The several studies on L-forms have, however, so far lacked a unifying principle that might transfer them from obscure morpholo= gical curiosities to generally interesting objects for quantitative, : physiological study, The current studies on E, coli protoplasts suggest that the absence of a normal wall, whether by external inhibition or internal genetic defect, accounts for the L-form cycle. As already mentioned, in broth the protoplasts E., coli fail to divide, and after extensive increase in mass, eventually burst. After some empirical experimentation, the conditions for proliferation ints beforms have been found. In addition to the protective medium already mentioned, these condi- tions are 1) a suitable strain, 2) at least 1,000 units of penicillin per ml. 3) One percent agar (Difco) and ) One percent meat extract (Lemeo). Seme of these conditions are quite rigorous, and have no certain explanation at present, For exemple, while 100 units of penicillin is quite sufficient for the production of protoplasts in broth or agar, 10 times this concentrae tion is required for L-forms, Perhaps the lower concentration dcez nos completely abolish wall-synthosis, and this is necessary for Leferm develop ment. The transformations of individual cells have been followed by direct phase microscopy of agar cultures, (In broth, a single blister appears whie’: balloons out to form a spherical protoplast). One or several blisters apseer, to give either a spherical or a lobulate 'protoplast', Then, further blisters may appear, te generate new lobes, or thin processes develop, the tips of which then balleon out at some distance from the original body, The connection may later disappear, As a result of these processes, the cell develops into a cluster of granvles and spheres of various sizes, the oftedescribved L-colony. In 2 or three days, the Lecolonies may reach a diameter of 2=3 mm. The Lecolonies may be transferred in series by grinding blocks of agar in broth and plating the minces, When plated with penicillin, they regenerate new Lecolonies; without penicillin, only normal bacillary colonies develop (even after 10 serial passages as L-forms). However, the viability of the L-coleny 1s very low, averaging only about 10 plating unite per colony which may contain 103 te 10? visible granules and spheres, The requirement for agar is a physical one: L-coloniea will develop in small blecks of agar, while protoplasts at the surface or immersed with the agar blocks in the same fluid medium will enlarge ani explode, L-~colonies which penetrated to an agareglass interface formed a conglomerate of large and small droplets which adhered imperfectly to the glass, These observations can be rationalized in the following working hypo- thesis: in the absence of the normal wall, the proteplast assumes a spherical shape in an isobaric medium by virtue of interfacial forces. There is no mechanism by which it can divide its mass, It has a fluid or senifluid consistency and, in contact with glass which it wets to some extent, the balance of interfacial forces breaks it up into droplets of varying size. In agar, however, the expanding protoplast is subject to local mechanical stresses, from the fibrils in which it is enmeshed, and can therefore flow into local crevices in the agar, to give blisters or processes, At these points, there 1956 Research Report ~ Bacterial Genetics - Wisconsin - J, Lederberg =3e is a fursher weakening of the tesidual (lipid?) membrane and secondary or daughter protoplasts are ultimately budded off, Ima sense, the agar mesh hag taken over the functions of the missing wall, The lov viability may be a reflection of the cxtent to shich a daughter gliobulc receives un in. tact genetic complement (nucleus), or simply te the mechanical fragility of these elencnts. This interpretation of L-forms is at present a workingehypothesis, The basic observations are not preatly different from those of Dienss and others working on different species, It happens that in E. coli, the osmotic fragility of the protoplasts is especially pronounced, ard therefore more readily analyzed, In other bacteria, notably Streptobacillus moniliformis whose L-forms can be cultivated in liquid medi, we may suppose a stronger residual wall or another layer is left, The liquid-culture L-forms of Proteus appear to be instances where the structural ghosts of exploded pretso~ plasts remain as an outer shell to provide the framework in which still viable units can proliferate at the center. Reports on the filtrability of viable elements of Leforms require some circumspection: on the one hand, most of the elements are not dramatically smaller than ordinary bacteria; on the other, the fluidity or plasticity of the L-forms raises some questions as to the dimensions of the filtrable units, ‘Insofar as these protoplasts do behave 2s naked globules of bacterial 'proteplasm' they may be expected to lead to new avenues of experimentation. However, the existing data cive no basis to doubt that an essential organizational unit cerresponding to the content of an intact cell is still required for the persistent viability of the protoplastic elements. An interesting corroboration of these proposals comes from the behavior of a mitant of E, coli (B, Davis) which requires diaminopimelic acid (DAP), This substance had been found to be a unique constituent of the cell walls of a number of bacteria, The mutant cells, when deprived of DAP in a pro- tective mediun, undergo the same transformations (protoplasts and 1-colonies) as described for the effect of penicillin. When DAP is restored to these forms, they then revert to the bactlilary type. However, as DAP does not ree verse the antibiotic effect of penicillin, compounds relate to different as- pects of cell wall synthesis, A systematic search is umierway for other biosynthetic mutants defective in wall formation, It is still uncertain exactly what reaction is inhibited by penicillin, The penicillin~protoplasts appear to be Limited only by a thin residual membrane; however, india-Ink preparations show a prominent but transparent capsule: immunological tests are contemplated to determine whether this capsule represents the remains of the old wall, or the accentuated develop- ment of a cartohydrate (capsular, K) antigen, When a protoplast is lysed in water, it does not freely digsolve; instead, india-ink preparations show @ large clear area, probabiy the cell-contents in a gelatinous lump, which may help account for the persistence of protein- synthetic preparations in partial lysates (Spiegeiman), (The foregoing observations are being coordinated into a manuscript for publication in the near future). * Ja 2t ex/lt2w Plt, L+te, 12%, 1-2 and other more comflex types. Further research has been devoted (1) to the physiologicel relations of differert mutants (2) to the concomitant behavior of propkaze in transduction and (3) the effects of irradiation of the phage. a. Almost all galactose-nsgative mutants are closely lir’-ed to ore another, Their linkage sequence has however not been determined as yet. No definite recurrences of mutations at the same locus have bsen tdentified. A series of at least 12 distinct loci has been established already, b. Kalekar and Kurahashi (at N.I,H., Bethesda) have worked out the engy= matic steps in galactose fermentation, according to the following scheme: 1, Galactose + ATP ¢)Galactose ~1-P + ADP (galactokinase) 2, Gal -1-P +UDP Glu<«?UDP Gal + Gln -1-P. (transferase) 3, UDP GalGlu «1-P + ADP by the Then Glu “1-P-900,, lactic acid/general glycolytic patiney. They have also analyzed the enzymatic defects of a number of mitants, Group 4 (Gal 2, Gal 8) lack galactokinase, Group B (Gal 1, Gal hh, Gai 6, Gal 7) lacke the transferase, and tims corresponds to the genetic disease, congenital galactosemia, in man, No epimerase mutants have been found so far. (See Kurahashi, K, 1957 Enzyme formation in galactose-negetive mutants of Escherichia coli Science 125, 114-116). e, Certain beterogenotes, c.g., 1+2=/le2+ give a normal wild type, galactose positive reaction. However, others, ltl-/l-lj+ are galactose~ negative, in contrast to 1+)+/l-lj-, galactose positive, That is, Cal 1/Cal ), forms 5 cisetrans positiomeffect group. In sum, Ual 1, Gal , Gal 6, Gat 7 form one position effect group (or "eistron"-Denzer), any pair from the group giving a cis-trans position-effect. Gal 2 and Gal & form a ssconrd cistron, Combinations of a mutant from one cistron with one from she other are galactose positive, Thus, as far as these experiments go, each cistron corresponis to one enzyme. However, one other mutant, Gal 3° shows position-effects with both cistrons, Its enzymology is not yet known, «25 r0 1956 Research Report ~ Bacterial Genetics - Wisconsin « J, Lederberg Currens theories of gene~enzyme relationships are converging on 2 linear correzpomence between the informational content of a protein pro duct, and of a polynucleotide string in the chromosome, Many experiments are in accord with this simple correspomience, ©.f., the relationship between the group A and B cistrons and the enzymes, For this reason, it is especially important to analyze apparent exceptions, like Gal.~ to sce whether they support, or contradict what may be an oversimplified schene, Prophage relations in transduction, The only markers known to be transduced by lambda are the Gal complex, A concerted search for additional transducible markors is underway, so far without success, The phage obteined by lytic srowth is ineffective 9 ard it must be obtained by the ultra-violet light induction of lysogenic donors. Since Gal is the only marker closely Linked to lambda (Lp) in sexual crosses, its specificity in transduction is connected with this linkage to the pro- phage. In fact, when lambda infects a sensitive cell, and lysogenic survivors are obtained, this may be considered a transduction of the lambda propharce . When this is coupled with Gal, we may write: wa e + Gal” Lp* x Gal Lr —> Cal” Lp . As the phace from heterogenotes has a very high efficiency of transduction, we have been able to study individual transformed clones, These are ane variably mixed with unchanged Gal” Lp® cells, which probably arg the unin fected segrezants from the multimucleate initial cejl. The Gal” components of these transductional clones have usually been Lp, a minority (about 1/3} give the Lp” (immune) reaction, The Lp, or prophage ta therefore directly associated with Gal” transduction, We will now deal with experiments de- signed to Leck for the segregation of prophage from heterogenotic cells $ where Gal is segregating, 1. Gal* Lp -ox Gat” Lp. When both parents are lysogenic, the heterogenotie progeny are pure lysogenic, as expected on almos any hypothesis, 2, Gal* Lp* «=x Gal" Lp® , Two types are found, a) Gal {Gak” which are pure lysegenic, that is either homogenotic (Lp*/Lp") or hemogenotic (is*) for prophage, This would seem to contradict the expectation that prophase is coupled with Gal, and therefore should segregate with it, ») Gel? /Gai" which give the immune (Lp") reaction, However, these types age invariably unstable: ali haploid segregants (with respect to Gal) are Ly ; and vice versa, while all Lp’ derivatives are still homo= er hotero-genotie for Gal, and sub- ject to further segregation. Onr first hypothealg was that this type was segregating for a defective prophage, Lp’ /ip*, However as no tF homogenstes have bgen found, it seems more Likely that the immune phenotype is either an Lp®/Lp’, homogenotie for the Lp®.mutant-prophage, or an Lp® on which the exs« genote confers immunity, Since this would ieply that the exofenote carrias at least part of the prophage, the two hypotheses are nore or less equivalents, For further discussion, we will postulate that this phenotype is Gel” fp? /Cai> Lp®, and abbreviate the prophage condition as Lo? “the immundzine effeck af the exogenotic Lp® would then be either a dosage effect of Lp® (wi: : compensation, cf. the sensitive phenotype of Lp®/up” homozygotes) or = : effect of Lp? in the exogenotic location, On this forsmiation, th: werwks a) is also homogenotic, Lp*/Lp*, sae {Spoae 1956 Research Report - Bacterial Genetics ~- Wisconsin « J, Lederberg This hypothesis does not explain why no sersreegstias prevhace heterogonoies, Lp /Lp” ave found, and we simply have to regort to 2d hoe proseripiions against this type. This type of segregation would be the only tangible evidence for our hypothesis that the prophage is coupled with Gal in the exogenote, The deeper significance of this hypothesis is the homology between prophage and other segments of the bacterial chromosome, 3, Gal Lp’ wx Gal” Lp / Gal” Lp° ( that is, homogenotic recipients). This transduction has not been sufficiently studied. However, they have been found to give lysogenic heterogenotes, which then segregate to give Lp’, Lp™S, and Lp® progeny, as if they wore Gal* Lp*/Gal” Lp*/Gal™ ip*. No effort has been made to characterize the simple Gal* Lp* / Gal” Lp® which mieht issue, The result does, however, give the most direct support so fsr for the representation of Lp’ on the exogenote, he Gal” Lp ox Gal” Lp*, To circumvent the possible pecularities of Lp*/ip® (i.e. prophage/sensitive) confrontctions, we have used the h (hos terange) mutant of lambda furnished by Appleyard, here designated simply as Lp", The usual re- sults are, however, pure Lp* or, less cften, pure Lp“, Again, we would have to find an ad hoc rule against prophage segregation, Perhaps the first event in establishment, end a necessary one for the survival of a heterogenote, is 2 cross- ‘over which ensures homogenosis for the Lp locus. In favor of this is the coinci- dence of Lp* ami LpT® hsterogenctes in the same primary traneductional clone, 2s has been observed once or twice, i.e., these would be the complementary Lpt/ip* and Lp®/Lp® homogenotes fron the initial, forbidden Lp*/Lp*. 5, Gal* Lp’ --x Gal” Lp’. The recipient here is a haploid, immune (defective prophage) type, not to be confused with Lp™®, The results are mostly Lp", with rare Lp* heterogenctes, This is, therefore, analogous to experiment 4, One or two types which behave as if Lp*/tp™ were also isclated, but the segregatioral coupling of Lp to Cel was not maintained, This is not understcod, 6. Thyy- Benzer (1955) has reported that Lp’ is killed by certain r mutants of phage Ty, but dees not support their growth, while Lp® supports normal growth of Thr, This stock is thercfore a further test of prophage action, All lambda lysogenic and immune stocks, including Lp’, ard Lp” as well as ip /Lp® diploids were found to be ‘immune’ to They . Te Gal” Lp’ PL3h” wx Gal” Lp” Pah. Ph3h is a new temperate phage, partly homologous with lambda, described by Jacob. It was used here as an additional means of marking the exogenote in transduction., A wide variety of lyeogeny- types has issued from this transduction, including Lp , pt and Lp® in verions combina tions with P3)* and P38, The most pertinent one, behaves as a Gal Lp PL3he/ Gal” Lp® Ph3h*, i.e., it is segregating both prophages, in coupling with Gal, This result together with experiment 3, gives strong, but tentative, support to the exogenotic prophage hypothesis, Yhe special immemity relationships between lenods and P)3)) may have something to do with the contrast between this result and the simpler experiment 2, The complexity of these results awaits a unifying generalizetion, Experiments 3 and 6 support the working hypothesis of prophage seeregation; the others are equivocal, At issue is the conclusion that the prophage is sn integral part of the bacterial chromosome. Its behavior wnat be analysed further te see whether its peculiarities ere incidental to a fundamental homology of prophage with other genres, or whether entirely new principles must be postulated, Q35 1956 1956 1956 1956 1957 1957 Publications of work supported by NSF Lederberg, J, Bacterial protoplasts induced by penicillin, Proc, Nat, Acad, Sei, 2:57h-577. Morse, M.L., E.M. Lederberg and J. Lederberg Transductional heterogenotes in Escherichia coli. Genetics 11:758+779. Lederberg, J. ard Tino, T. Phase variation in Salmonella, Genetics 41:7h3=757. Lederberg, J. Linear inheritance in transductionsl clones, Genetics 11:8),5-871. Morse, M.L. Transduction and transformation, Ann, N. Y. Acad, Set. ‘Lederberg, J. Viruses, geros and cells, Bacteriological Reviews,