Use of Constructed Double Mutants for Determining the Temporal Order of Expression of Sporulation Genes in Bacillus subtilis

Transcription

1 JOURNAL OF BACTERIOLOGY, June 1973, p Copyright 1973 American Society for Microbiology Vol. 114, No. 3 Printed in U.S.A. Use of Constructed Double Mutants for Determining the Temporal Order of Expression of Sporulation Genes in Bacillus subtilis J. G. COOTE' AND J. MANDELSTAM Microbiology Unit, Department of Biochemistry, University of Oxford, Oxford, England Received for publication 5 March 1973 Double mutants containing two Spo mutations concerned with different stages of sporulation were constructed. In these, the phenotype that is exhibited is that of the earlier sporulation block. The same procedure was applied to sporulation mutants damaged in the same stage of development. The results provide a basis for placing in a temporal order different mutations concerned in stage II and stage IV of spore development. In general, the order indicated by the phenotypes of the double mutants is in agreement with the order derived on biochemical grounds. Double oligosporogenous mutants have also been constructed. Their sporulation incidence is roughly equal to the product of the incidences of the parent strains, idicating that separate factors are involved in overcoming each oligosporogenous block. The number of dependent sequential steps in sporulation is estimated as not less than about 12. The biochemical and morphological events that accompany sporulation can be subdivided into several categories (6). In summary, these are (i) the primary sequences of dependent events specifically concerned with the process (A, B, C, etc., see diagram below). The term dependent in this context applies to an event which will not occur unless the earlier events have been successfully completed. The dependent nature of the sequence suggests that sporulation mutations are pleiotropic, a fact that was emphasized even in early studies of sporulation mutants (12, 14). For some stages of development, several events (E, F, G) might have to occur simultaneously. Examples of dependent events are formation of the spore septum, separation of the spore protoplast from the membrane of the mother cell, etc. (ii). The second category is side effects (Bl, Cl, etc.). An example is the appearance of dipicolinic acid, a spore-specific compound which shows up at a definite stage in the process. Nevertheless, later events do not seem to be dependent upon its appearance, and in mutants of Bacillus cereus which lack this compound viable spores are still formed (16). (iii) The third category consists of changes in vegetative functions that take place because the cultural conditions that induce 1 Present address: Institut de Microbiologie, Universite de Paris-Sud, 91 Orsay, France. sporulation differ from those that promote growth. An example is the appearance of aconitase, and other examples will be found elsewhere (6). C2 A +B C D F )jh...sp ORE B1 C1, It is often not possible to distinguish between events of type (i) and those of type (ii). For instance the appearance of alkaline phosphatase is correlated with the transition of stage II to stage III of sporulation, but since the function of the enzyme is unknown it is impossible to tell whether later events are dependent upon it or whether it is a side effect. A further complication has been introduced by the observation that the alkali-soluble protein of the spore coat which is incorporated into the structure at a very late stage is, in fact, synthesized several hours earlier (17). Events of this type (C2) are also indicated in the diagram. Because events in the main sequence are dependent, it is easy to distinguish mutants blocked early in the process from those blocked later since the mutations give rise to phenotypes which are morphologically distinct or which differ in the biochemical marker events that accompany sporulation (15, 1). It is, however, 1254

2 VOL. 114, 1973 TEMPORAL ORDER OF EXPRESSION OF SPORE GENES 1255 much more difficult to decide which gene expression precedes another when both are expressed at about the same time. One might have thought that the mapping of sporulation genes would help, but this is not so because mutations that block the process at some particular time, for instance stage II, may be widely separated on the chromosome, and, conversely, genes that are very close to each other may be associated with events that are widely separated in time (4, 2). However, when some of these mutants are induced to sporulate, the process proceeds normally up to a given point and then gives rise to obvious maldevelopment. This is particularly likely to occur with stage II mutants, and curious, often bizarre, forms are likely to be seen (see 10, 18, 15). Examples include the production of multiple septa, abortive disporic forms, overproduction of membrane, etc. It follows, again from the dependent nature of the system, that if one introduced into a single organism two mutant genes, each of which gives rise to a distinct sporulation phenotype, one should be able to distinguish the earlier mutation from the later because, in such a mutant, sporulation would proceed normally until the time came for expression of the first altered gene. This would produce its damaging effect and the process would go wrong, so that the organism would exhibit the characteristic phenotype for that gene. The second gene, being concerned with a later event, would not be expressed at all. The work to be described was carried out with the following objectives: (i) to establish in principle that constructed double mutants could be used to distinguish early and late genes; (To this end they were constructed from single mutants blocked at obviously different stages of development [e.g., II and IV].); (ii) to order mutations affecting events connected with the same stage of sporulation which are morphologically distinguishable; (iii) to determine whether constructed double mutants could afford any further insight into the nature of oligosporogeny which might well be caused by leaky genes (2, 12). Even if this is so, it is still a puzzle to know why, within an oligosporogenous population, a small proportion of the cells are able to sporulate successfully. This success could be due to some general phenotypic factor (e.g., age of the cell or an unusually high content of adenosine 5'-triphosphate) that allowed a few cells to overcome the block in the process. If such a general factor enabled cells to overcome more than one type of oligosporogeny, it follows that if one constructed a double mutant from strains sporulating with a frequency of, for example, 1% and 10%, any cell in the population having enough of the general factor to overcome the more severe block would even more easily overcome the less severe one. The double mutant should accordingly sporulate at 1%. If, however, the factors involved in overcoming the two types of block are independent, then the probability of sporulation in the double mutant would be the product of the separate probabilities or 0.1%. The results to be described show that double mutants exhibit the phenotype of the earlier gene so that the method is, in principle, applicable to mutations affecting the same stage of sporulation. In addition, it appears that whatever factor allows a cell to overcome one oligosporogenous block does not help it overcome another. MATERIALS AND METHODS Organism. Bacillus subtilis 168 (trpc2) was used. It forms spores normally in resuspension medium supplemented with tryptophan (see below) and is referred to as the wild type. The Spo and Osp mutants (see below) derived from the wild type and their morphological and biochemical properties have been described elsewhere (1, 15). We are indebted to S. R. Ayad for a prototrophic strain of B. subtilis 168. All mutations in the strain except DG2, N25, and NG17.29 were transformed into the standard 168 strain to make them isogenic. The exceptions were tested for the absence of multiple spore mutations by using Spo+, trp+ deoxyribonucleic acid (DNA) at a saturating concentration (2,ug/ml) to transform each of them to prototrophy. All yielded between 5 and 10% Spo+ colonies among the trp+ transformants as a result of the integration of a second DNA fragment possessing the unlinked wild-type allele of the spore mutation. If the mutants had carried a second unlinked spore mutation, the expected proportion of Spo+ transformants obtained would have been about 0.5%. A list of the strains carrying auxotrophic markers which were used for transduction analysis has been given elsewhere (2), and the map positions of the sporulation mutations had already been determined earlier by Piggot (8) and Coote (2) by using transduction mapping with phage PBSl and also in some instances by transformation. Media. PAB, antibiotic assay medium no. 3 (Difco, Detroit, Mich.), was used. Lactate-glutamate minimal agar plates were prepared as already described (1) Ṅomenclature. The asporogenous and oligosporogenous phenotypes will be referred to as Spo and Osp, respectively. Growth and sporulation. The organisms were grown with shaking at 37 C in a medium containing hydrolyzed casein, L-tryptophan, and inorganic ions (13). Growth was measured spectrophotometrically by using a calibration curve relating extinction at 600

3 1256 COOTE AND MANDELSTAM J. BACTERIOL. nm to bacterial dry weight. To initiate sporulation, a culture of 0.25 mg (dry wt)/ml was centrifuged, and the cells were transferred at the same density to a resuspension medium containing L-glutamate, L-tryptophan, and inorganic ions (13). With the wild type this procedure produces about 80% refractile spores in 7 to 8 h. Time in hours after transfer of cells to resuspension medium is indicated by to, ti, etc. Stages 0 to VI are the generally recognized cytological stages of spore formation (9). Biochemical events during sporulation. Associated with the morphological stages of development are a number of biochemical marker events which occur at characteristic times during sporulation (15). Those used for the work described here are exo-protease (stage 0), alkaline phosphatase (stage II-III), glucose dehydrogenase (late stage III), and heat resistance (stage VI). They were assayed in samples removed from the resuspension cultures (see reference 1). Electron microscopy. Cells initiated to sporulation in the resuspension medium were sampled at t7 and fixed, and thin sections were prepared (5). Morphological assessment was made only on complete cells in longitudinal sections. Transduction. The preparation of transducing lysates with phage PBS1 and the procedure used for transuction were as already described (2). Transformation. DNA was extracted with minor modifications by the method of Marmur (7), and the transformation procedure has been described previously (2). For the construction of double mutants, DNA was used at a concentration of /ig/ml. Construction of double sporulation mutants. Most of the double mutants were constructed by using transduction. The most convenient procedure involved donor strains carrying spore mutations linked to trpc2. The recipient (also trpc2) carried a different spore gene mutation not linked to trpc2, but to another auxotrophic marker elsewhere on the chromosome map. Each donor strain was first transformed to prototrophy by using DNA prepared from the trp+ strain of B. subtilis. Lysates prepared from each trp+ derivative were then used to infect the recipient strain, selection being made for trp+ transductants. A percentage of these, depending on the degree of linkage, took up the sporulation mutation of the donor and became double-sporulation mutants. In one donor, X8, the sporulation mutation was so closely linked to the auxotrophic marker phe-12 that transformation could be more conveniently used. In this case the recipients were previously constructed phe- derivatives carrying spore mutations unlinked to phe-12. Each was transformed to prototrophy by using DNA prepared from X8, phe+. Selection and screening of double mutants. Transductants and transformants obtained in the crosses just described were plated on lactate-glutamate minimal agar and incubated at 37 C for 2 to 3 days to allow development of the pigment associated with sporulation in this strain (11). Spo+ colonies were dark brown, whereas Spo or Osp colonies were either translucent, white, or light brown. In some instances, in the crosses involving two sporulation mutants, it was possible to distinguish by color two types of spore-defective colony, one exhibiting the phenotype of the recipient and the other the phenotype associated with the double mutant. Whether two colony types were distinguishable by color or not, it was essential to know that the strain taken for investigation contained both sporulation mutations. This was checked by back-crossing as follows: eight single colonies, presumptive double mutants, from each cross were picked at random and subcultured on nutrient agar. A phage lysate was then prepared from each of the eight separate cultures and used to transduce to prototrophy two recipient strains carrying the auxotrophic markers to which each of the two sporulation mutations was known to be linked. The resulting prototrophic transductants in each case were then checked to ensure that the expected percentage of cells had acquired the appropriate sporulation phenotype. RESULTS The phenotypes of the four donor mutants are summarized in Table 1. In resuspension cultures, mutant Y13 produced long elongated cells which were blocked at stage 0. It was included as an obvious control because any recipient taking up its mutation should give a double mutant in which the majority of the cells (>99.9%) remained blocked at stage 0. Mutant P18 was damaged at spore septum formation (stage II). It developed septa at both ends of the cell and also laid down cell wall material between the double membranes of both septa (Fig. 2). Mutant P20 was characterized by deposition of spore coat material in the cytoplasm of the mother cell instead of around the developing prespore (Fig. 7). Although, for biochemical reasons, this mutant would appear to be blocked between stages III and IV, its appearance in electron micrographs suggested that it might be blocked at a much later stage since coat proteins do not normally become visible until stage V. It was included in the series to determine whether the double mutant method would help to place this gene in the temporal sequence. Finally, mutant X8 was used as an example of an organism blocked at a later stage. In these cells, cortex formation was initiated but no coat material was laid down (Fig. 4). The phenotypes of the recipients are summarized in Table 2 and will be referred to below. Phenotypic characteristics of double mutants containing mutation Y13 (stage 0). Double mutants were constructed from Y13 and each of the following: P14 (stage II), P9 (stage II), X8 (stage IV), W10 (stage V). All the double mutants exhibited the stage 0 phenotype characteristic of mutant Y13.

4 VOL. 114, 1973 TEMPORAL ORDER OF EXPRESSION OF SPORE GENES 1257 TABLE 1. Phenotypes of the strains used as donors in constructing double mutantsa Stage Pr ~~~~~~~~Alkaline Glucose Heat Geti Strain T Morphologic al description Pro- -phospha- dehydro- resist- Genetic blocked tease ~~~~~tase genase ance linkage Y13 Osp 0 Long vegetative cell trpc2 (61) P18 Osp II Septa at both poles of cell trpc2 (33) P20 Osp III-IV Incomplete cortex, coat mate trpc2 (12) rial deposited in mother cell X8 Osp IV Incomplete cortex, no coat ma phe-12 (3) terial a Alkaline phosphatase and glucose dehydrogenase were measured in samples taken from resuspension cultures at t5, and heat resistance was measured at t7. Values are expressed as percentages of those obtained with the wild type. A fuller description of the donor phenotypes has been given previously (1). The last column lists the percentage recombination values obtained by two-factor transduction crosses with the auxotrophic marker indicated (2). Phenotypic characteristics of double mutants containing mutation P18 (stage II). The sporulation phenotypes of the double mutants resulting from the introduction of this mutation are summarized in Table 3. It was to be expected that, since mutation P18 affects some process at stage II, expression of the gene containing this mutation would precede the expression of genes concerned in later stages. The table shows, in fact, that all the double mutants involving recipients blocked at later stages exhibited a phenotype which was characteristic of P18. In addition, it was clear that the gene containing mutation P18 also preceded in expression the genes containing mutations NG17.29 and P9. Both of these mutations block sporulation after a single septum at one end of the cell has been formed in the normal manner. Development in mutant NG17.29 goes no further, but in mutant P9 the septum proceeds to bulge into the cytoplasm of the mother cell (Fig. 3). Both produce alkaline phosphatase in amounts comparable to those of the wild type (Table 2). The double mutants (P9/P18 and NG17.29/P18) in keeping with their morphological resemblance to P18 both produced small amounts of phosphatase. However, when mutation P18 was introduced into organisms carrying two other stage II mutations (DG2 and N25), both of which block sporulation before production of alkaline phosphatase, the resulting phenotype was indistinguishable from that of the recipient. It seems clear then that expression of these two mutations precedes that of P18 (Table 3 and see Fig. 9). The two recipients are recognizable because mutation N25 causes production of a single septum at one pole of the cell and in some circumstances an overproduction of septum membrane (15). Mutation DG2 was morphologically recognizable because it caused a single septum to be laid down at one pole of the cell with an exceptionally thick layer of cell wall between the twin membranes (Fig. 1). In addition, in some cells several cross walls were laid down apparently in a random way. The results shown in Table 3 do not help one to establish the temporal order of the gene containing mutation DG2 relative to that containing N25, nor to establish the relative order of NG17.29 and P9. To do this it would be necessary to construct double mutants containing these pairs of mutations. Phenotypic characteristics of double mutants containing mutations P20 (stage III- IV) and X8 (stage IV). The sporulation phenotypes of double mutants containing P20 are also shown in Table 3. It is manifest that the expression of the gene containing P20 precedes those containing the late mutations W10 and W5, as the phenotypes of the double mutants were identical with that of the donor. Both mutants W10 and W5 formed spores almost normally to a late stage, but in W10 the cortex had an unusual, streaked appearance (Fig. 5), and in W5 the cortex was almost entirely missing (Fig. 6). A somewhat curious result was obtained with mutant X8 in which the recipient mutation was presumed to concern a slightly earlier stage, i.e., the initiation of cortex formation (see Fig. 4). The double mutant (X8/P20) exhibited the phenotype of the donor but in a modified way. For example, coat material, instead of being deposited in concentric layers in the cytoplasm of the mother cell, tended to wrap itself around the developing prespore (Fig. 8). The remaining double mutants described in Table 3 all exhibited the morphology ch#racteristic of the recipient, and it would thus appear that the expression of mutation P20 could be placed temporally after that of Y10 (stage III) and before that of X8 (Fig. 9).

5 1258 COOTE AND MANDELSTAM J. BACTrERIOL. I 3 Downloaded from 4 on July 3, 2018 by guest FIG. 1. Mutant DG2 is blocked at stage II and is aberrant in that it lays down a thick layer of cell wall between the two membranes of the spore septum. FIG. 2. Mutant P18 forms spore septa at both poles of the cell and lays down cell wall between the septum membranes. FIG. 3. Mutant P9 forms a single septum in the normal way which bulges into the mother cell, although it remains attached to the original cell wall invaginations. FIG. 4. Mutant X8 develops as far as the first stage of cortex formation, the primordial germ cell wall. This is seen as an electron-dense band between the prespore membranes. The bars represent 0.2 jm. Mutant X8 (stage IV, Fig. 4) was used as donor with four recipient strains, P14 and P9 (both stage LI), and W5 and W10 (stage IV and V) (see Table 2 and Fig. 3, 5, and 6). The phenotype of the double mutant P14/X8 is discussed below. Double mutant P9/X8, as expected, had a phenotype indistinguishable from that of P9. However, the double mutants W1O/X8 and W5/X8 were both blocked at stage III, i.e., at an earlier stage than either of the

6 VOL. 114, TEMPORAL ORDER OF EXPRESSION OF SPORE GENES 1259 Downloaded from 7 on July 3, 2018 by guest FIG. 5. In mutant W10 the cortex has an unusual, striated appearance and development of the coat layers is only complete at the poles of the spore. FIG. 6. Mutant W5 shows well-developed coat layers, but growth of the cortex which normally occurs between the two membranes of the prespore is almost entirely absent. FIG. 7. Mutant P20 develops normally up to the free prespore stage, but the coat material is deposited in concentric layers in the mother cell cytoplasm instead of around the prespore. FIG. 8. Double mutant X8/P20 exhibits a modified P20 phenotype where the coat material is attempting to wrap itself around the prespore in a more normal way instead of forming a mass in the mother cell cytoplasm. The bars represent 0.2,gm.

7 1260 COOTE AND MANDELSTAM J. BACTERIOL. TABLE 2. Phenotypes of strains used as recipients in constructing double mutantsa I Stage Pro- Alkaline Glucose Strain Type Ha eei blocked blocked Morphological description tease phospha- dehydro- genase ~~~resistance rheat Genetic tase genase linkage DG2 Spo II Single septum with thick < cysa 14 (5) cell wall between membranes N25 Osp II Single septum and over cysa14 (7) production of membrane P14 Osp II Septa at both poles of cysa14 (27) cell NG Spo II Single septum < hisa I (85) P9 Osp JI Single septum that hisal (57) bulges into mother cell Y1O Osp III Free prespore ura-1 (84) W5 Osp IV-V Incomplete cortex, nor ura-1 (45) mal coat layers Wio Osp V Striated cortex, normal ura-1 (54) coat layers a For descriptive previously (1, 14). TABLE 3. details see Table 1. A fuller description of the recipient phenotypes has been given Phenotypes of double mutants produced by introduction of mutations P18 and P20 into recipient strains blocked at various stagesa Recipient Donor P18 (II) Donor P20 (III-IV) Alka- Glu- No. and stage line cose Heat Alkaline Glucose Heat blocked Morphology phos- dehy- resistance Morphology phospha- dehydropha- dro- tase resistance genase tase genase DG2 (II) II (recipient) < II (recipient) < N25 (II) II (recipient) < II (recipient) < NG17.29 (II) II (donor) < II (recipient) < P9 (Il) II (donor) < II (recipient) < Y10 (III) II (donor) < III (recipient) < X8 (IV) II (donor)' 2.0 NRe < III-IV (modified < donor, see Fig. 8)b W5 (IV-V) II (donor) 2.0 NR < III-IV (donor) < W1o (V) II (donor) 3.0 NR < III-IV (donor) < a Details are as in Table 1. "The double mutants X8/P18 and X8/P20 were constructed in two ways. Firstly, using a lysate of P18 or P20, trp+ to transduce X8, trpc2 to protrophy, and secondly, transforming P18 or P20, phe-12 to prototrophy using X8, phe+ DNA. The same double mutants, constructed by either procedure, had indistinguishable phenotypes. e Not recorded; the values were presumed to be similar to those of the donor. parent organisms. Both double mutants formed a complete spore protoplast, free within the mother cell, but no cortex of coat formation was initiated (see Discussion). Phenotypic properties of double Osp mutants. A double mutant was constructed by using as recipient mutant P14. In this strain 80% of the cells are blocked at stage II, whereas the remainder go on to form heat-resistant spores (Table 2). The donor was another Osp mutant, X8, which sporulated at an incidence of 1%, the remaining cells being blocked at stage IV (Fig. 4). Electron micrographs of the double mutant prepared from resuspension cultures at t7 showed that the majority (about 80%) of the cells were blocked at stage II, and that a much smaller proportion (about 20%) were blocked at stage IV. The incidence of heat-resistant spores was 0.12%. Similarly, when P14 (spore incidence 20%) was used as recipient with mutant Y13 as donor (spore incidence 0.06%), the re-

8 VOL. 114, 1973 TEMPORAL ORDER OF EXPRESSION OF SPORE GENES sulting double mutant sporulated at 0.003%. When X8 (1%) was used as recipient with Y13 (0.06%) as donor, the double mutant had a spore incidence of about 0.001%. DISCUSSION The oligosporogenous double mutants will be considered first. Previous studies of oligosporogeny have shown that in sister cells the initiation of sporulation is highly correlated; i.e., if a cell has been induced to start sporulating the probability is high that its sister cell will have been initiated at the same time (1). This is also what is found with sibling wild-type cells (3). This synchrony, however, is lost as soon as the cells encounter an oligosporogenous block when the occurrence of a spore in one cell does not help to indicate the probable behavior of its sibling (1). This must be taken to mean that overcoming the block results from the possession of some factor, presumably biochemical, which was either distributed asymetrically at the time of cell division or which has been generated randomly in a small proportion of cells sometime after cell division. The present experiments were undertaken to discover whether possession of this factor, whatever it might be, would help an organism to overcome oligosporogenous blocks in general. If there were such a general factor the double mutants should sporulate at the incidence characteristic of the more severely damaged of the two parents (see Introduction). If, on the other hand, there is no general factor, the probability of sporulation should be the product of the separate probabilities. Reference to the results will show that the latter is in fact what is found, and it seems to show that independent factors are involved in overcoming different oligosporogenous blocks at least in the pairs of mutants we have examined. The results with the double mutants generally support the validity of the contention that if such a strain is constructed, the sporulation phenotype will be that of the earlier mutation. Thus, all the double mutants which carried the stage 0 mutation Y13 exhibited the stage 0 phenotype. Similarly, when the double mutant contained a mutated stage II gene and a muta tion producing a block at any later stage, the prevailing phenotype was stage II (see Table 3). In light of these results, it seems justifiable to use this method to order, at least tentatively, the mutations affecting stage II. The results summarized in Table 3 indicate that the mutation which blocks sporulation after the formation of a single spore septum (e.g., N25) precedes the expression of the mutation P18 which causes septa and cell wall to be laid down at both ends of the cell. From the data a composite diagram (Fig. 9) has been drawn to indicate the temporal order of expression of sporulation genes. For mutations blocking sporulation before stage III the order is unambiguous and can be readily inferred, but the placing of later sporulation genes is more difficult. The results (Table 3) indicate that the gene containing mutation P20 should be placed after that containing Y10 (stage III). This is because the double mutant exhibits the phenotype of Y1O both morphologically and biochemically. However, the temporal position of P20 relative to X8 (stage IV) is not clear. In the double mutant the coat material, instead of forming a mass in the cytoplasm of the mother cell as it does in mutant P20 (Fig. 7) has now apparently settled out around the developing prespore so that the resulting product differs from both parent phenotypes and is more nearly "normal" than either (Fig. 8). The failure of the double mutant to reproduce the phenotype of either one parent or the other might be attributed to the fact that the coat protein is not part of the primary sequence of events. Instead, and this would appear to follow from the work of Wood (17), it is a side product, formed during stage III, which re-enters the main sequence after a time (see Introduction). The result in a double mutant might then be a phenotype in which both aberrations are expressed, but expressed in altered form. For morphological reasons, however, expression of P20 is shown as preceding that of X8 in Fig. 9. An interesting result was obtained with the double mutant W5/X8 in which the phenotype was typically stage III, i.e., an earlier stage than either of the parent strains. This is what would be anticipated if there were simultaneous expressions of two mutated genes (see Introduction). Thus, it is very probable that several gene products are necessary for the successful transition from stage III to stage IV. Failure to produce one of these might lead to the appearance of feebly refractile, stage IV mutants, whereas failure of two products simultaneously would prevent even this amount of development and the phenotypic appearance would be that of a stage III mutant. We have therefore been unable to assign an order of expression for genes containing mutations W5 and X8 and have assumed tentatively that they are expressed at the same time (Fig. 9). The double mutant W1O/X8 behaved similarly, but in this case a clear distinction could be made because W10 produces dipicolinic acid.

9 1262 COOTE AND MANDELSTAM J. BACTERIOL. 'AX3 FIG. 9. Composite schematic diagram indicating the probable order of expression of sporulation genes. It was obtained by considering the phenotypes of constructed double mutants and also the biochemical properties of the parent strains. The parent strains are described in Tables 1 and 2 and the double mutants in Table 3 (see also Fig. 1-8). Stages of spore development are indiated by roman numerals. Numbers in brackets refer to biochemical or morphological properties as follows: (1) protease, (2) phosphatase, (3) glucose dehydrogenase, (4) refractility (note that none of the mutants developed more than feeble refractility and appeared gray rather than bright in the phase contrast microscope), (5) dipicolinate, (0) indicates that the mutant did not exhibit any of the indicators of sporulation. This characterizes it as being blocked later than either W5 or X8. It will be apparent from this discussion that the method of constructed double mutants is probably of limited value in assigning an order to late sporulation genes. However, for mutations producing blocks up to stage III or IV the results are much more clear-cut and the assignment of an order of gene expression is in agreement with the order indicated on biochemical grounds. Finally, if we accept that exclusion of one mutant sporulation phenotype by another indicates the sequential order of gene expression, we can use Fig. 9 to obtain an estimate of the number of dependent sequential steps involved in the process. The order of expression is as follows: (i) Y13, (ii) DG2, (iii) N25, (iv) P18, etc., and the number of sequential steps is about 12. It should be noted that this is a minimum estimate which might be greatly increased by examining the properties of more constructed double mutants. Nevertheless, for reasons given in this paper and also by Piggot (8), the total number of sequential steps is almost certain to be substantially less than the total number of operons concerned in sporulation. ACKNOWLEDGMENTS We thank D. Kay for advice on electron microscopy and P. J. Piggot for many helpful discussions. We are greateful to D. Torgersen for skilled technical assistance. The work was supported by the Science Research Council of England.

10 VOL. 114, 1973 TEMPORAL ORDER OF EXPRESSION OF SPORE GENES 1263 LITERATURE CITED 1. Coote, J. G Sporulation in Bacillus subtilis. Characterization of oligosporogenous mutants and comparison of their phenotypes with those of asporogenous mutants. J. Gen. Microbiol. 71: Coote, J. G Sporulation in Bacillus subtilis. Genetic analysis of oligosporogenous mutants. J. Gen. Microbiol. 71: Dawes, I. W., D. Kay, and J. Mandelstam Determining effect of growth medium on the shape and position of daughter chromosomes and on sporulation in Bacillus subtilis. Nature (London) 230: Ionesco, H., J. Michel, B. Cami, and P. Schaeffer Genetics of sporulation in Bacillus subtilis Marburg. J. Appl. Bacteriol. 33: Kay, D., and S. C. Warren Sporulation in Bacillus subtilis. Morphological changes. Biochem. J. 109: Mandelstam. J Regulation of bacterial spore formation. Symp. Soc. Gen. Microbiol. 19: Marmur, J A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3: Piggot, P. J Mapping of asporogenous mutations of Bacillus subtilis: a minimum estimate of the number of sporulation operons. J. Bacteriol. 114: Ryter, A Etude morphologique de la sporulation de Bacillus subtilis. Ann. Inst. Pasteur 108: Ryter, A., P. Schaeffer, and H. Ionesco Classification cytologique, par leur stade de blocage, des mutants de sporulation de Bacillus subtilis Marburg. Ann. Inst. Pasteur 110: Schaeffer, P., and H. Ionesco Contribution a l1'tude genetique de la sporogenese bacterienne. C. R. Acad. Sci. 251: Schaeffer, P., H. Ionesco, A. Ryter, and G. Balassa La sporulation de Bacillus subtilis: etude genetique et physiologique. Colloq. Int. Centre Nat. Rech. Sci., p Sterlini, J. M., and J. Mandelstam Commitment to sporulation in Bacillus subtilis and its relationship to development of actinomycin resistance. Biochem. J. 113: Spizizen, J Analysis of asporogenic mutants in Bacillus subtilis by genetic transformation, p In L. L. Campbell and H. 0. Halvorson (ed.), Spores III. American Society for Microbiology, Ann Arbor, Michigan. 15. Waites, W. M., D. Kay, I. W. Dawes, D. A. Wood, S. C. Warren, and J. Mandelstam Sporulation in Bacillus subtilis. Correlation of biochemical events with morphological changes in asporogenous mutants. Biochem. J. 118: Wise, J., A. Swanson, and H. 0. Halvorson Dipicolinic acid-less mutants of Bacillus cereus. J. Bacteriol. 94: Wood, D. A Properties and time of synthesis of alkali-soluble protein of the spore coat. Biochem. J. 130: Young, I. E Characteristics of an abortively disporic variant of Bacillus cereus. J. Bacteriol. 88: Downloaded from on July 3, 2018 by guest