Biochimie (1992) 74, 661-667 © Soci6t6 fran~;aise de biochimie et biologie mol6culaire / Elsevier, Pads
661
P r o t e i n f i l a m e n t s m a y initiate the a s s e m b l y o f the Bacillus subtilis s p o r e coat AI Aronson l, L Ekanayake l, PC Fitz-James 2 IDepartment of Biological Sciences, Purdue University, West Lafayette, IN, 47907 USA; 2Department of Microbiology, University of Western Ontario Medical School, London, Canada (Received 29 November 1991; accepted 20 February 1992)
S u m m a r y D The Bacillus subtilis spore coat consists of three morphological layers: a diffuse undercoat, a striated inner coat and a densely staining outer coat. These layers are comprised of at least 15 polypeptides and the absence of one in particular, CotE, had
extensive pleiotropic effects. Only a partial inner coat was present on the spores which were lysozyme-sensitive. The initial rate of germination of these spores was the same as for the wild type but the overall optical density decrease was greater apparently due to the loss of the incomplete spore coat from germinated spores. Suppressors of the lysozyme-sensitive phenotype had some outer coat proteins restored as well as some novel minor polypeptides. These spores still lacked an undercoat and germinated as did those produced by the cotE deletion strain. The CotE protein was synthesized starting at stage II-III of sporulation, long before the appearance of the coat on spores at stage IV-V. Despite its apparent hydrophilic properties, this protein was present in the crude insoluble fraction from sporulating cells. CotE was not solubilized by high or low ionic strength buffers nor by detergents used for the solubilization of membrane proteins. Either 8 M urea or 6 M guanidine HC! was required and dialysis against a low ionic strength buffer resulted in aggregation into long, sticky filaments. Both the CotE and CotT spore coat proteins appeared to be necessary for the formation of these filaments. Each of these proteins contains sequences related to a bovine intermediate filament protein so their interaction could result in an analogous structure. The properties of these filaments are consistent with their deposition on the outer forespore membrane as the initial components of the spore undercoat layer. B subtUis I spore / filament / coat assembly
Introduction The B subtUis spore coat is comprised of 15-20 polypeptides arranged in three morphological layers [ 1-3]. In addition, 20-30% of the coat protein can not be solubilized and probably consists of cross-linked polypeptides [3]. Seven spore coat protein genes (cotA-F, cotT) have been cloned [4-8] and deletions of most of them had no detectable phenotypic effects (usually lysozyme-sensitivity or germination defects). Absence of the CotT polypeptide resulted in some alteration of the inner coat layer as well as a selective germination defect [6]. There were extensive pleiotropic effects clue to the absence of the CotE protein [7]. Both the under and outer spore coat layers were absent although cotA-lacZ and cotC-lacZ fusions were expressed as well as in the wild type [7]. These two proteins are probably components of the outer coat and they must accumulate in the mother cell cytoplasm in the cotE deletion strain. Spores produced by this strain were sensitive to lysozyme but responded
normally to germinants [7]. Transcription of the cote gene began at stage II-III from a a E dependent promoter and then later transcription from a second promoter was dependent upon the SpoIIID protein as well as o K [9]. The early synthesis of this coat protein was surprising since transcription of other coat proteins begins at stages IV-V, at the time when the coat is assembled on the outer forespore membrane. The orderly assembly of the three-layered spore coat from 15-20 polypeptides must be dependent upon transcriptional control, the interaction of the polypeptides and perhaps the processing of some from precursors [5, 8]. The early synthesis of CotE and the pleiotropic effects resulting from a cotE deletion implied that it was a critical protein for initiating coat assembly. Some, but not all, of the phenotypic effects of the cotE deletion could be suppressed supporting the importance of this protein to coat structure. In addition, CotE has unique solubility properties and aggregates with at least one other spore coat protein to form filaments which may be the structural basis for the spore undercoat.
662
Materials a n d m e t h o d s Bacterial strains a n d growth The wild type B subtilis strain JH642 and a derivative (NB200) lacking a functional cotT gene have been described elsewhere [5, 61. The wild type strain PYI7 and a cotE deletion strain were kindly provided by Dr L Zheng [7]. Cells were grown and sporulated at 37°C in a nutrient sporulation medium (NSM l l0l) with or without 7 lag ml -I chloramphenicol. A minimal salts-glucose medium was also used [ ! l ]. Spores were purified by washing once with ! M KCI, once with deionized water and then pelleting suspensions in water through 50% Renografin (66% sodium diatrizoate meglumine plus 10% sodium diatrizoate; Solvay) by centrifuging in a Sorvall HB4 swinging bucket rotor at 8000 rpm for ! h.
Isolation o f cotE suppressors Spores produced by the cotE deletion strain are sensitive to iysozyme [71 so suppressors of this phenotype were isolated. Because of the extensive manipulations involved and the possibility of contamination by strains commonly used in the laboratory, the deletion was transformed into a B subtilis prototroph (Marburg strain) by selection for chloramphenicol resistance. This strain grew on the glucose-minimal salts medium with chloramphenicol but without any supplements in contrast to all the other B subtilis strains. The phenotypic properties of the spores produced by the transformants were the same as those of the original cote deletion mutant and the Southern hybridization of Hindlll-digested DNA to a cotE probe was as expected 171. About !() a° spores were treated with 50 lag ml -I lysozyme at 37°C tbr 2 h in 20 ml 30 mM Tris, pH 7.8 with slow shaking. Following centrifugation at 8000 rpm for 15 rain in a Sorvall RC2 centrifuge, the pellet was suspended in 8 mi of 30 mM Tris, pH 7.8 and portions layered onto 8 ml of 50% Renografin in deionized water. The tubes were centrifuged at 8000 rpm for 1 h in a Sorvall HB4 swinging bucket rotor. The supematant containing the lower density, phase dark spores was discarded and the pellets were washed once with 10 ml of 30 mM Tris, pH 7.8. These pellets were suspended in 200 ml of the glucosesalts medium plus 7 pg ml -I chloramphenicol. Following sporulation at 37°C, the cycle was repeated at least twice. The spore pellets from the final Renografin gradients were then streaked on NSM agar containing 7 lag ml -I chloramphenicol. After incubation at 37°C for 2 days, portions of individual colonies were tested for the presence of lysozyme-resistant spores by incubating suspensions in 50 lag ml -I lysozyme at 37°C for 2 h in microtiter plates. Each was then examined in the phase microscope and two, designated Rev~ and Rev 2, were selected for further analysis. Both suppressor strains still contained the original cotE deletion as determined by Southern hybridization as previously described. In order to do preliminary PBS-I transduction mapping of the suppressor loci, the cotE deletion was transformed into each of the tester strains of Dedonder et al [12]. Transducing lysates from strains Rev I and Rev 2 were used for transduction of the tester strains. 50-100 of each of the prototrophs were then streaked on NSM-chloramphenicol agar and the resulting spores suspended in microtiter plates in order to test for the lysozyme-sensitive phenotype.
Gel electrophoresis Spores were prepared in NSM medium, purified and the spore coats isolated [3]. Spore coat proteins were solubilized in 8 M urea-l% sodium dodecyl sulfate-50 mM dithioerythritol-2 mM phenylmethyl sulfonyl fluoride-2 mM 2[N cyclohexylamine] ethane sulfonic acid, pH 9.6, by incubating the suspensions at 37°C for 20 min and then repeating the extraction. Portions of the soluble fractions were resolved in 12, 15 or 20% urea SDS-PAGE. Gels were either stained with Coomassie blue [3] or transferred to lmmobilin P membranes (Millipore) for reaction with antibodies [13]. Antibody to the CotT protein has been described [5]. The CotE protein is not very abundant [7] so the gene sequence was used to synthesize a peptide embracing the carboxyl-terminal 20 residues. The peptide was purified by HPLC chromatography on a Synchron RP-P reverse phase column employing a gradient from 0.1% trifluoroacetic acid in water to 0.1% trifluoroacetic acid-100% acetonitfile. The peptide fraction was lyophilized and rechromatographed before linking to bovine serum albumin [14] for subcutaneous injection into rabbits. Following an initial inoculation in complete Freund's adjuvant and three subsequent inoculations at weekly intervals in incomplete Freund's adjuvant, the rabbits were bled and the gamma globulin fraction purified by elution from a DEAE column with 7 mM sodium phosphate, pH 7.0, and precipitation with 50% saturation of ammonium sulfate. Detection of antibodies involved a second antirabbit antibody coupled to alkaline phosphatase (Promega).
Cell fractionation Cells were grown at 370C in NSM plus or minus 7 lag ml -I chloramphenicol for 8-18 h. Progress into sporulation was monitored by following the optical density at 660 nm in a Klett
colorimeter and the appearance of phase-white endospores. 100 ml samples were taken at 90-min intervals usually starting 1-2 h after the end of exponential growth and continuing until at least 70% of the cells contained phase-white endospores (stage V). Cells were centrifuged at 8000 rpm for 8 min, washed once with 30 mi ! M KCI-5 mM EDTA, once with deionized water-2 mM phenyl methyl sulfonyl fluoride (PMSF) and finally with 30 mM Tris-2 mM [5-mercaptoethanol (I3ME) 2 mM PMSF, pH 8.0. The cells were suspended in 1-2 ml of the latter buffer and passed twice through a French pressure cell at 9000 psi. The suspension was sonicated for 20 s to reduce viscosity and centrifuged at 4500 rpm for 4 min. The pellet which contained intact spores and cells was discarded. The supernatant was centrifuged at 12 000 rpm for 15 min and again at 30 000 rpm for 30 min in a Spinco 65 rotor. The pellets were pooled and washed successively with 10 ml portions of 10 mM Tris--2 mM [~ME, 1% Triton X-100 plus 2 mM PMSF, pH 7.8; I 0 mM Tris-2 mM PMSF-2 mM [~ME1% Triton X-100--i M KCi, pH 8.0 and deionized water plus 2 mM PMSF. Each suspension was centrifuged at 30 000 rpm for 30 min in the Spinco 65 rotor and the supernatants discarded. Portions of the washed pellet fraction were suspended in one of the following: 50 mM octyl [3-D-glucoside, 1% deoxycholate, 2% Brij 35, 1% sodium dodecyl sulfate, 8 M urea in 20 mM Tris, pH 9.0 or 6 M guanidine HCI in 20 mM 'Iris, pH 9.0. Each suspension was centrifuged in the Spinco 65 rotor as described above and the extraction repeated. Portions of the final pooled supernatants were fi'actionated on urea SDS-PAGE. The 8 M urea-soluble fractions were also dialyzed at 4°C for 12 h versus 2000 vol 20 mM Tris-10 mM
663 [3ME, pH 7.4. The dialyzed suspension was then centrifuged in an Eppendorf microfuge for 10 min. The pellets were dissolved in the solubilization buffer used for spore coat proteins or suspended in the dialysis buffer for examination in the electron microscope.
Electron microscopy Suspensions of the pellet fractions following dialysis of the 8 M urea soluble fractions were examined by negative staining. Portions were mixed with ml equal vol of filtered 2% phosphotungstic acid in l0 mM Tris, pH 7.4 and drops placed on carbon-coated grids. Following drying, the grids were examined in a Phillips 200 microscope.
Results
Properties of cotE-deletion spores The initial germination responses of heat activated (70°C for 30 min) mutant spores to the B subtilis germinants, L-alanine plus inosine, Penassay broth or a mixture of glucose-fructose-asparagine [6] were as for the wild-type. There was often a longer lag for the mutant spores but the biggest difference was in the extent of decrease of optical density (fig 1). The phase-dark mutant spores were much smaller than the swollen wild-type spores but they were viable if
plated on NSM-chloramphenicol agar within 1-2 h after the initiation of germination. Subsequently, they lost viability such that after 4 h in the germination medium, less than 5% survived (versus >90% for the wild-type). The smaller size and instability of the mutant germinated spores was probably due to the loss of the spore coat upon germination (unpublished results). Apparently the germ cell wall was not adequate to protect the emerging cell for a prolonged period. Suppressors of the lysozyme-sensitive phenotype were selected as described in Materials and methods. Despite the lysozyme-resistance of the spores, they germinated as did the mutant (fig l) with an apparent loss of the spore coat from the germinated spores. There were additional proteins present in the spore coat of these suppressor strains which were missing from the cotE deletion spores (fig 2). Several had the same sizes as those found in wild type spore coats
GERMINATION IN L-ALANINE
100
90-
0 U
80-
PY17 8
m L_ 0
cote Rev.2.1
70
Rev.13.2
.4[ .4[
60
50
~00
1 50
minutes
Fig I. Germination in 30 mM Tfis, pH 7.8, plus 10 mM L-alanine and 1 mM inosine of heat activated (70°C for 30 min) spores of the wild type (PYI7), the cotE-deletion mutant (cotE) and of the two lysozyme-resistant suppressors (Rev~ and Rev2). Spores were suspended to an optical density of 0.9-1.0 at 580 nm, incubated at 37°C and the decrease in absorbance followed in a Perkin Elmer Junior Model 35 spectrophotometer. Values are plotted as the percent of maximum absorbance at 580 nm.
Fig 2. A Coomassie blue stained 20% urea SDS-PAGE of spore coat extracts (30 l~g protein in each lane) of strain PYI7 (WT), the cotE-deletion (A cotE) and of the two lysozyme-resistant suppressor strains (Rev~ and Rev2). Unmarked arrowheads on the left indicate bands found in the wild-type but not the AcotE extract. The position of the Cote protein is indicated (E arrowhead). Arrowheads on the fight indicate bands unique to the suppressor strains.
664 including some of 40--60 kDa which are probably constituents of the outer coat [7]. At least two genes encoding presumptive outer coat proteins, cotA and cotC, are transcribed in the cote deletion strain [7] and these proteins are probably aggregated in the cytoplasm. In the suppressor strains, these and other outer coat proteins may somehow be det~osited on the spore surface in sufficient quantity to confer lysozyme resistance to the spores. The germination properties suggest, however, that there has not been compensation for all of the functions of the CotE protein, including synthesis of an undercoat (unpublished results). There are one or two novel, minor bands present in coat extracts from the suppressors (fig 2) and their presence may facilitate the deposition of outer coat proteins. There is no obvious over-production of a well-defined spore coat protein so the nature of the suppressor locus product is unknown. Both suppressors have been mapped by PBS-I transduction and are 20-30% linked to aroD with much less linkage to lys-l.
Properties and synthesis of the CotE protein Antibody to a synthetic peptide comprised of the carboxyl-terminal 20 residues of the Cote protein was used to examine the time of synthesis of this polypeptide (fig 3). Antigen was first detected at about stage II-IIl of sporulation thus confirming results obtained with a lacZ fusion [9]. Addition of 50 ~tg ml-I chloramphenicol to cells at stage Ill resulted in no decrease in antigen content over the subsequent 4-5 h. Virtually all of the CotE antigen was present in the crude insoluble fraction. The antigen was often present as multiple and/or smeared bands in contrast to a cross-reacting soluble antigen (fig 3). The latter was present in soluble extracts of exponentially growing wild type cells or of the cotE deletion mutant, as well as in extracts of a spoOB mutant prepared after the end of exponential growth. The solubility properties of the CotE antigen were unexpected based on the deduced amino acid sequence of this protein [7] which indicated a hydrophilic, acidic polypeptide. There is modest sequence similarity to a bovine intermediate filament protein (BIF) [15]:
COTE:
90 100 110 120 RYRDNNYLDDEHEVIAKVLQQPNCLEVTISPNGNKI I1:::11
BIF:
:11:
:
130 140 150 160 170 WQAEREFLAEVVGETKVVVEVNPDWEEDDEEDWEDELEEELEDI :
II:::::
:
II
:1:::
:::
LKEELAYLKKNHEEEMSVLKGQVGGQVSVEVDSAPG 210 220 230 240
::
:
:I:I:I
:
:III:
ID ...... LAKILSDMRSQYEVIAEKNRPd)AEAWFISQ~EELNRE 250 260 270 280
wt 1 S -Z
Acot E Z S g
wt l V S Z
A cot E ZV S Z
Fig 3. Immunoblot of extracts of sporulating cells of the parental strain, PYI7 (wt), and the cotE-deletion strain (AcotE) prepared at about stage II and late stage IV (at least 60% phase white endospores) of sporulation. The crude soluble (S) and insoluble (1) fractions were prepared as described in Materials and methods and 20 ~g of protein from each was electrophoresed on a 15% urea SDS-PAGE. Transfer and treatment with CotE antibody were as described in Materials and method.~. The broad lower bands (E arrows) are unique to the wild-type and are CotE antigens whereas the sharp, upper bands are cross-reacting antigens found in exponential cells of the wild type and of the CotE deletion strain as well as in post-exponential, soluble extracts of a mutant (spoOB) blocked at the initiation of sporulation.
There is a 23% identity, 45% with conserved differences and 62% (dotted lines) with replacements based on the mutation data matrix of Schwartz and Dayhoff as per the GCG program (Madison, WI), over a 73 amino acid overlap in a region of high a-helical content of the BIF. This region is involved in the coiled coil-polypeptide interactions required for filament formation [15, 16]. The CotE protein could also contain a lot of u-helical structures so such interactions are possible with this polypeptide. In addition, intermediate filament proteins contain repeat motifs of small residues (ser, gly) at their amino terminus and the BIF has the repeat sequence GGYGGG [15]. This sequence is present at the carboxyl terminus of another spore coat protein, CotT which also contains the repeat motif (GGGY)3 [5], so an interaction of CotT and CotE may be involved in filament formation (fig 4). Most intermediate filament proteins are distinguished by their unique solubility properties, ie generally insoluble in low or high ionic strength buffers and require 8 M urea or 6 M guanidine for solubilization [ 16]. They also reaggregate as 10-nm wide filaments upon dialysis against a low ionic strength buffer [16]. The low speed pellet from stage IV-V sporulating cells was extracted with reagents used for the solu-
665
Acidic Cytokeratin Intermediate Filamentprotein (> 40 kDa)
NH~~
0~ [
Head groups ['GGYGGG;'] LSSSNSSSS.]
Rod domain Qt-helix-coiled coils
(24 kDa)
~ racidicq or Facidicq = [.acidicl2 [. basicJ2
N
I
-
Cot T C O O H ~
10 nm filiament
I I
2
~
COOH 9
(7.8 k D a )
COOH
protofilament~
I Cot E (acidic)
~
filament
P,Y rich
~a2"~-
Fig 4. A schematic diagram of the structure of an acidic cytokeratin intermediate filament protein and its assembly into 10 nM filaments. The spore coat proteins, CotE with sequence homology in the a-helical rod domain (bracket) and CotT with homology at its carboxyl end to the cytokeratin head group may interact to form filaments.
l 1
2
3
4
5
2
3
4
6
........ ~
_qD
E-"~ |
L
~-T
Fig 5. Immunoblot with CotE (A) or CotT (B) antibodies of fractions resolved on a 15% urea SDS-PAGE. The crude insoluble fraction from late stage IV-V sporulating cells of the wild type (strain PYI7) or the cotE-deletion strain were washed and extracted with 8 M urea as described in Materials and methods. Following dialysis of this latter fraction, 10 ~tg of protein in the precipitates and soluble fractions were electrophoresed. Lanes A 1 and B 1: insoluble fraction from strain PYI7; lanes A2 and B2: insoluble
bilization of membrane proteins (n-octyl-glucoside, deoxycholate, Brij 35 or Triton X-100), as well as with low and high ionic strength buffers (see Materials and methods). Little Cote antigen was solubilized. The washed pellet was then extracted with 8 M urea and multiple Cote antigen bands were found (fig 5A, lane 1). Dialysis of the 8 M urea soluble fraction against 20 mM Tris, pH. 7.4, resulted in aggregation and most of the Cote antigen was again present in the insoluble fraction (fig 5A, lane 5). Portions of this fraction were mixed with an equal vol of 2% phosphotungstic acid and spread on carbon-
fraction from the cotE-deletion; lane A3: soluble fraction from strain PYI7; lane A4: soluble fraction from the cotE deletion. The insoluble fractions (as in lanes 1 and 2) were dissolved in 8 M urea and redialyzed. Following centrifugation, the precipitates from strain PYI7 (lanes A5 and B3) and the cotE-deletion (lanes A6 and B4) were dissolved in the buffer used for spore coats and electrophoresed. E arrow on left indicates major CotE antigen probably the monomer and T arrow on right indicates CotT monomer.
666 coated grids (fig 6). There was an extensive network of filaments present in preparations from stage IV-V of the wild type (fig 6A). These were 1.5-2 mn at the thinnest but riley tended to stick together to form thicker and larger fibers. These filaments were not found in extracts from the cotE deletion strain but there were less abundant, shorter and wider filaments present (fig 6B). Filaments were less prevalent and had a different appearance in preparations from the wild type at stage HI of sporulation (fig 6C). A less abundant network was present in preparations from stage V cells of strain NB200, a mutant lacking the CotT antigen (fig 6D). Multiple forms of both the CotE and CotT antigens were present in the precipitates following dialysis of the 8 M urea-soluble fraction (fig 5). This precipitate was dissolved in 8 M urea and again dialyzed against 20 mM Tris-10 mM 13ME, pH 7.4. There appeared to be fewer filaments (unpublished results) and while both antigens were still present in the insoluble fraction, they were less abundant (fig 5A, lane 5; 5B lane 3). In addition to containing a glycine-rich carboxyl end, CotT is relatively rich in tyrosine residues [5]. The multiple CotT antigen bands present in these precipitates could be due to cross-linking to each other or to CotE.
Discussion A deletion of the cotE gene resulted in extensive pleiotropic effects on the bacterial spore coat with the absence of both an under and outer coat and a deficiency of the inner coat [7]. At least some outer coat proteins are synthesized in mutant cells but apparently cannot be deposited on the spore due either to the absence of the undercoat and/or the deficiency of the inner coat. This defect can be partially corrected in a suppressor strain which forms lysozymeresistant spores. While these spore coats contained some outer coat proteins, the spores still germinated as did those from the parental strain (fig 1), so while they are protected against lysozyme, the coat was still unstable and presumably came off after germination. Given the relatively early time for the initiation of synthesis of CotE, it could be a major constituent of the undercoat which is deposited on the outer forespore membrane. Perhaps in conjunction with the CotT protein, an intermediate filament-like structure is formed which is anchored to this membrane. Since the CotT protein is synthesized later in sporulation (stage IV) than CotE, the initial deposition of CotE may be in a different conformation than when it subsequently interacts with CotT. This latter protein is present in the crude insoluble fraction containing CotE (fig 5B) and is also probably a component of the
¢
0
Fig 6. Negative stain of precipitates following 8 M urea extraction and dialysis of crude, washed insoluble fractions from sporulating cells (see Materials and methods). A. Strain PY 17 at stage IV-V of sporulation. B. The cotEdeletion strain at stage IV-V of sporulation. C. Strain PYI 7 at about stage III of sporulation. D. Strain NB200 (absence of CotT antigen) at stage V of sporulation. Magnification was X 77 000; 1 mm is 12.98 nm. striated inner coat [6], implying a dual function in coat structure. It may be significant that much of the CotT protein in spore coats from the cote deletion strain was present in its precursor form ([5] unpublished results). Processing may be integrated with the deposition of the outer coat. In addition to the sequence relatedness of Cote to a bovine intermediate filament protein, there is some homology to the active site of peroxidases (T Deits, personal communication). This activity has been implicated in the formation of dityrosine cross-linked proteins of the cortical membrane which is formed
667
following fertilization of sea urchin eggs [ 17]. CotT is a tyrosine-rich protein [5] and there is evidence for dityrosine in the coat insoluble fraction [3]. The presence of high molecular weight bands of CotT especially in the 8 M urea-soluble fraction of strain PYI7 (fig 5B) may be indicative of such crosslinking. The cross-linking could stabilize the interaction of Cote and CotT in filaments. CotT monomers may also be cross-linked to each other as components of the under and/or inner spore coat layers.
7
8 9 10
Acknowledgments
11
Research was supported by Grant DAAL03-87-K-0133 from the US Army. The technical assistance of Joan Clapper and Doryth Loewy is deeply appreciated.
12
13
References 1 Aronson AI, Fitz-James PC (1976) Structure and morphogenesis rff the bacterial spore coat. Bacteriol Rev 40, 360--402 2 Jenkinson HF, Sawyer WE), Mandelstam J (1981) Synthesis and order of assembly of spore coat proteins in Bacillus subtilis. J Gen Microbiol 123, 1-16 3 Pandey NK, Aronson At (1979) Properties of the Bacillus subtilis spore coat. JBacteriol 137, 1208-1218 4 Donovan W, Zheng L, Sandman K, Losick R (1987) Genes encoding spore coat polypeptides from Bacillus subtilis. J Mol Biol 196, l-d0 5 Aronson AI, Song HY, Bourne N (1989) Gene structure and precursor processing of a novel Bacillus subtilis spore coat protein. Mol Microbiol 3, 437 AqA 6 Bourne N, Fitz-James PC, Aronson AI (1991) Structural and germination defects of Bacillus subtilis spores with
14
15
16
17
altered contents of a spore coat protein, d Bacterioi ! 73, 6618-6625 Zheng L, Donovan WP, Fitz-James PC, Losick R (1988) Gene encoding a morphogenetic protein required in the assembly of the outer coat of the Bacillus subtilis endospore. Genes Dev 2, 1047-1054 CuRing S, Zheng L, Losick R (1991) Gene encoding two alkali-soluble components of the spore coat from Bacillus subtilis. J Bacterioi 173, 2915-2919 Zheng L, Losick R (1990) Cascade regulation of spore coat gene expression in Bacillus subtilis. J Mol Biol 212~ 645-660 Schaeffer P, Ionesco H, Ryter A, Ballassa G (1963) La ~,porulation Bacillus subtilis, 6tude g6n6tique et physiologique. Colloq Int CNRS 124, 553-563 Spizizen J (1958) Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc Nati Acad Sci USA 44, 1072-1078 Dedonder RA, Lepesant JA. Lepesant-Kejzlarov J, Billault A, Steinmetz Z, Kunst F (1977) Construction of a kit of reference strains for rapid genetic mapping in Bacillus subtilis 168. Appl Environ Microbio133, 989-993 Towbin H, Staehlin T, Gordon J (1979) Electrophoretic transfer of proteins from polyacrylamide gel to nitrocellulose sheets: procedure and some applications. Proc Nat! Acad Sci USA 76, 4350-4354 Bulinski JC, Kumar S, Titani K, Hauschka SD (1983) Peptide antibody specificity for the amino terminus of skeletal muscle a-actin. Proc Natl Acad Sci USA 80, 1506-15~0 Bader BL, Magin TM, Hatzfeld M, Franke WW (1986) Amino acid sequence and gene organization of cytokeratin no 19, an exceptional tail-less intermediate filament protein. Embo J 5, 1865-1875 Hatzfeld M, Franke WW (1985) Pair formation and promiscuity of cytokeratins: formation in vitro of heterotypic complexes and intermediate-sized filaments by homologous and heterologous recombinations of purified polypeptides. J Cell Bio1101, 1826-1841 Shapiro BM (1991) The control of oxidant stress at ferlilization. Science 252, 533-536
661
P r o t e i n f i l a m e n t s m a y initiate the a s s e m b l y o f the Bacillus subtilis s p o r e coat AI Aronson l, L Ekanayake l, PC Fitz-James 2 IDepartment of Biological Sciences, Purdue University, West Lafayette, IN, 47907 USA; 2Department of Microbiology, University of Western Ontario Medical School, London, Canada (Received 29 November 1991; accepted 20 February 1992)
S u m m a r y D The Bacillus subtilis spore coat consists of three morphological layers: a diffuse undercoat, a striated inner coat and a densely staining outer coat. These layers are comprised of at least 15 polypeptides and the absence of one in particular, CotE, had
extensive pleiotropic effects. Only a partial inner coat was present on the spores which were lysozyme-sensitive. The initial rate of germination of these spores was the same as for the wild type but the overall optical density decrease was greater apparently due to the loss of the incomplete spore coat from germinated spores. Suppressors of the lysozyme-sensitive phenotype had some outer coat proteins restored as well as some novel minor polypeptides. These spores still lacked an undercoat and germinated as did those produced by the cotE deletion strain. The CotE protein was synthesized starting at stage II-III of sporulation, long before the appearance of the coat on spores at stage IV-V. Despite its apparent hydrophilic properties, this protein was present in the crude insoluble fraction from sporulating cells. CotE was not solubilized by high or low ionic strength buffers nor by detergents used for the solubilization of membrane proteins. Either 8 M urea or 6 M guanidine HC! was required and dialysis against a low ionic strength buffer resulted in aggregation into long, sticky filaments. Both the CotE and CotT spore coat proteins appeared to be necessary for the formation of these filaments. Each of these proteins contains sequences related to a bovine intermediate filament protein so their interaction could result in an analogous structure. The properties of these filaments are consistent with their deposition on the outer forespore membrane as the initial components of the spore undercoat layer. B subtUis I spore / filament / coat assembly
Introduction The B subtUis spore coat is comprised of 15-20 polypeptides arranged in three morphological layers [ 1-3]. In addition, 20-30% of the coat protein can not be solubilized and probably consists of cross-linked polypeptides [3]. Seven spore coat protein genes (cotA-F, cotT) have been cloned [4-8] and deletions of most of them had no detectable phenotypic effects (usually lysozyme-sensitivity or germination defects). Absence of the CotT polypeptide resulted in some alteration of the inner coat layer as well as a selective germination defect [6]. There were extensive pleiotropic effects clue to the absence of the CotE protein [7]. Both the under and outer spore coat layers were absent although cotA-lacZ and cotC-lacZ fusions were expressed as well as in the wild type [7]. These two proteins are probably components of the outer coat and they must accumulate in the mother cell cytoplasm in the cotE deletion strain. Spores produced by this strain were sensitive to lysozyme but responded
normally to germinants [7]. Transcription of the cote gene began at stage II-III from a a E dependent promoter and then later transcription from a second promoter was dependent upon the SpoIIID protein as well as o K [9]. The early synthesis of this coat protein was surprising since transcription of other coat proteins begins at stages IV-V, at the time when the coat is assembled on the outer forespore membrane. The orderly assembly of the three-layered spore coat from 15-20 polypeptides must be dependent upon transcriptional control, the interaction of the polypeptides and perhaps the processing of some from precursors [5, 8]. The early synthesis of CotE and the pleiotropic effects resulting from a cotE deletion implied that it was a critical protein for initiating coat assembly. Some, but not all, of the phenotypic effects of the cotE deletion could be suppressed supporting the importance of this protein to coat structure. In addition, CotE has unique solubility properties and aggregates with at least one other spore coat protein to form filaments which may be the structural basis for the spore undercoat.
662
Materials a n d m e t h o d s Bacterial strains a n d growth The wild type B subtilis strain JH642 and a derivative (NB200) lacking a functional cotT gene have been described elsewhere [5, 61. The wild type strain PYI7 and a cotE deletion strain were kindly provided by Dr L Zheng [7]. Cells were grown and sporulated at 37°C in a nutrient sporulation medium (NSM l l0l) with or without 7 lag ml -I chloramphenicol. A minimal salts-glucose medium was also used [ ! l ]. Spores were purified by washing once with ! M KCI, once with deionized water and then pelleting suspensions in water through 50% Renografin (66% sodium diatrizoate meglumine plus 10% sodium diatrizoate; Solvay) by centrifuging in a Sorvall HB4 swinging bucket rotor at 8000 rpm for ! h.
Isolation o f cotE suppressors Spores produced by the cotE deletion strain are sensitive to iysozyme [71 so suppressors of this phenotype were isolated. Because of the extensive manipulations involved and the possibility of contamination by strains commonly used in the laboratory, the deletion was transformed into a B subtilis prototroph (Marburg strain) by selection for chloramphenicol resistance. This strain grew on the glucose-minimal salts medium with chloramphenicol but without any supplements in contrast to all the other B subtilis strains. The phenotypic properties of the spores produced by the transformants were the same as those of the original cote deletion mutant and the Southern hybridization of Hindlll-digested DNA to a cotE probe was as expected 171. About !() a° spores were treated with 50 lag ml -I lysozyme at 37°C tbr 2 h in 20 ml 30 mM Tris, pH 7.8 with slow shaking. Following centrifugation at 8000 rpm for 15 rain in a Sorvall RC2 centrifuge, the pellet was suspended in 8 mi of 30 mM Tris, pH 7.8 and portions layered onto 8 ml of 50% Renografin in deionized water. The tubes were centrifuged at 8000 rpm for 1 h in a Sorvall HB4 swinging bucket rotor. The supematant containing the lower density, phase dark spores was discarded and the pellets were washed once with 10 ml of 30 mM Tris, pH 7.8. These pellets were suspended in 200 ml of the glucosesalts medium plus 7 pg ml -I chloramphenicol. Following sporulation at 37°C, the cycle was repeated at least twice. The spore pellets from the final Renografin gradients were then streaked on NSM agar containing 7 lag ml -I chloramphenicol. After incubation at 37°C for 2 days, portions of individual colonies were tested for the presence of lysozyme-resistant spores by incubating suspensions in 50 lag ml -I lysozyme at 37°C for 2 h in microtiter plates. Each was then examined in the phase microscope and two, designated Rev~ and Rev 2, were selected for further analysis. Both suppressor strains still contained the original cotE deletion as determined by Southern hybridization as previously described. In order to do preliminary PBS-I transduction mapping of the suppressor loci, the cotE deletion was transformed into each of the tester strains of Dedonder et al [12]. Transducing lysates from strains Rev I and Rev 2 were used for transduction of the tester strains. 50-100 of each of the prototrophs were then streaked on NSM-chloramphenicol agar and the resulting spores suspended in microtiter plates in order to test for the lysozyme-sensitive phenotype.
Gel electrophoresis Spores were prepared in NSM medium, purified and the spore coats isolated [3]. Spore coat proteins were solubilized in 8 M urea-l% sodium dodecyl sulfate-50 mM dithioerythritol-2 mM phenylmethyl sulfonyl fluoride-2 mM 2[N cyclohexylamine] ethane sulfonic acid, pH 9.6, by incubating the suspensions at 37°C for 20 min and then repeating the extraction. Portions of the soluble fractions were resolved in 12, 15 or 20% urea SDS-PAGE. Gels were either stained with Coomassie blue [3] or transferred to lmmobilin P membranes (Millipore) for reaction with antibodies [13]. Antibody to the CotT protein has been described [5]. The CotE protein is not very abundant [7] so the gene sequence was used to synthesize a peptide embracing the carboxyl-terminal 20 residues. The peptide was purified by HPLC chromatography on a Synchron RP-P reverse phase column employing a gradient from 0.1% trifluoroacetic acid in water to 0.1% trifluoroacetic acid-100% acetonitfile. The peptide fraction was lyophilized and rechromatographed before linking to bovine serum albumin [14] for subcutaneous injection into rabbits. Following an initial inoculation in complete Freund's adjuvant and three subsequent inoculations at weekly intervals in incomplete Freund's adjuvant, the rabbits were bled and the gamma globulin fraction purified by elution from a DEAE column with 7 mM sodium phosphate, pH 7.0, and precipitation with 50% saturation of ammonium sulfate. Detection of antibodies involved a second antirabbit antibody coupled to alkaline phosphatase (Promega).
Cell fractionation Cells were grown at 370C in NSM plus or minus 7 lag ml -I chloramphenicol for 8-18 h. Progress into sporulation was monitored by following the optical density at 660 nm in a Klett
colorimeter and the appearance of phase-white endospores. 100 ml samples were taken at 90-min intervals usually starting 1-2 h after the end of exponential growth and continuing until at least 70% of the cells contained phase-white endospores (stage V). Cells were centrifuged at 8000 rpm for 8 min, washed once with 30 mi ! M KCI-5 mM EDTA, once with deionized water-2 mM phenyl methyl sulfonyl fluoride (PMSF) and finally with 30 mM Tris-2 mM [5-mercaptoethanol (I3ME) 2 mM PMSF, pH 8.0. The cells were suspended in 1-2 ml of the latter buffer and passed twice through a French pressure cell at 9000 psi. The suspension was sonicated for 20 s to reduce viscosity and centrifuged at 4500 rpm for 4 min. The pellet which contained intact spores and cells was discarded. The supernatant was centrifuged at 12 000 rpm for 15 min and again at 30 000 rpm for 30 min in a Spinco 65 rotor. The pellets were pooled and washed successively with 10 ml portions of 10 mM Tris--2 mM [~ME, 1% Triton X-100 plus 2 mM PMSF, pH 7.8; I 0 mM Tris-2 mM PMSF-2 mM [~ME1% Triton X-100--i M KCi, pH 8.0 and deionized water plus 2 mM PMSF. Each suspension was centrifuged at 30 000 rpm for 30 min in the Spinco 65 rotor and the supernatants discarded. Portions of the washed pellet fraction were suspended in one of the following: 50 mM octyl [3-D-glucoside, 1% deoxycholate, 2% Brij 35, 1% sodium dodecyl sulfate, 8 M urea in 20 mM Tris, pH 9.0 or 6 M guanidine HCI in 20 mM 'Iris, pH 9.0. Each suspension was centrifuged in the Spinco 65 rotor as described above and the extraction repeated. Portions of the final pooled supernatants were fi'actionated on urea SDS-PAGE. The 8 M urea-soluble fractions were also dialyzed at 4°C for 12 h versus 2000 vol 20 mM Tris-10 mM
663 [3ME, pH 7.4. The dialyzed suspension was then centrifuged in an Eppendorf microfuge for 10 min. The pellets were dissolved in the solubilization buffer used for spore coat proteins or suspended in the dialysis buffer for examination in the electron microscope.
Electron microscopy Suspensions of the pellet fractions following dialysis of the 8 M urea soluble fractions were examined by negative staining. Portions were mixed with ml equal vol of filtered 2% phosphotungstic acid in l0 mM Tris, pH 7.4 and drops placed on carbon-coated grids. Following drying, the grids were examined in a Phillips 200 microscope.
Results
Properties of cotE-deletion spores The initial germination responses of heat activated (70°C for 30 min) mutant spores to the B subtilis germinants, L-alanine plus inosine, Penassay broth or a mixture of glucose-fructose-asparagine [6] were as for the wild-type. There was often a longer lag for the mutant spores but the biggest difference was in the extent of decrease of optical density (fig 1). The phase-dark mutant spores were much smaller than the swollen wild-type spores but they were viable if
plated on NSM-chloramphenicol agar within 1-2 h after the initiation of germination. Subsequently, they lost viability such that after 4 h in the germination medium, less than 5% survived (versus >90% for the wild-type). The smaller size and instability of the mutant germinated spores was probably due to the loss of the spore coat upon germination (unpublished results). Apparently the germ cell wall was not adequate to protect the emerging cell for a prolonged period. Suppressors of the lysozyme-sensitive phenotype were selected as described in Materials and methods. Despite the lysozyme-resistance of the spores, they germinated as did the mutant (fig l) with an apparent loss of the spore coat from the germinated spores. There were additional proteins present in the spore coat of these suppressor strains which were missing from the cotE deletion spores (fig 2). Several had the same sizes as those found in wild type spore coats
GERMINATION IN L-ALANINE
100
90-
0 U
80-
PY17 8
m L_ 0
cote Rev.2.1
70
Rev.13.2
.4[ .4[
60
50
~00
1 50
minutes
Fig I. Germination in 30 mM Tfis, pH 7.8, plus 10 mM L-alanine and 1 mM inosine of heat activated (70°C for 30 min) spores of the wild type (PYI7), the cotE-deletion mutant (cotE) and of the two lysozyme-resistant suppressors (Rev~ and Rev2). Spores were suspended to an optical density of 0.9-1.0 at 580 nm, incubated at 37°C and the decrease in absorbance followed in a Perkin Elmer Junior Model 35 spectrophotometer. Values are plotted as the percent of maximum absorbance at 580 nm.
Fig 2. A Coomassie blue stained 20% urea SDS-PAGE of spore coat extracts (30 l~g protein in each lane) of strain PYI7 (WT), the cotE-deletion (A cotE) and of the two lysozyme-resistant suppressor strains (Rev~ and Rev2). Unmarked arrowheads on the left indicate bands found in the wild-type but not the AcotE extract. The position of the Cote protein is indicated (E arrowhead). Arrowheads on the fight indicate bands unique to the suppressor strains.
664 including some of 40--60 kDa which are probably constituents of the outer coat [7]. At least two genes encoding presumptive outer coat proteins, cotA and cotC, are transcribed in the cote deletion strain [7] and these proteins are probably aggregated in the cytoplasm. In the suppressor strains, these and other outer coat proteins may somehow be det~osited on the spore surface in sufficient quantity to confer lysozyme resistance to the spores. The germination properties suggest, however, that there has not been compensation for all of the functions of the CotE protein, including synthesis of an undercoat (unpublished results). There are one or two novel, minor bands present in coat extracts from the suppressors (fig 2) and their presence may facilitate the deposition of outer coat proteins. There is no obvious over-production of a well-defined spore coat protein so the nature of the suppressor locus product is unknown. Both suppressors have been mapped by PBS-I transduction and are 20-30% linked to aroD with much less linkage to lys-l.
Properties and synthesis of the CotE protein Antibody to a synthetic peptide comprised of the carboxyl-terminal 20 residues of the Cote protein was used to examine the time of synthesis of this polypeptide (fig 3). Antigen was first detected at about stage II-IIl of sporulation thus confirming results obtained with a lacZ fusion [9]. Addition of 50 ~tg ml-I chloramphenicol to cells at stage Ill resulted in no decrease in antigen content over the subsequent 4-5 h. Virtually all of the CotE antigen was present in the crude insoluble fraction. The antigen was often present as multiple and/or smeared bands in contrast to a cross-reacting soluble antigen (fig 3). The latter was present in soluble extracts of exponentially growing wild type cells or of the cotE deletion mutant, as well as in extracts of a spoOB mutant prepared after the end of exponential growth. The solubility properties of the CotE antigen were unexpected based on the deduced amino acid sequence of this protein [7] which indicated a hydrophilic, acidic polypeptide. There is modest sequence similarity to a bovine intermediate filament protein (BIF) [15]:
COTE:
90 100 110 120 RYRDNNYLDDEHEVIAKVLQQPNCLEVTISPNGNKI I1:::11
BIF:
:11:
:
130 140 150 160 170 WQAEREFLAEVVGETKVVVEVNPDWEEDDEEDWEDELEEELEDI :
II:::::
:
II
:1:::
:::
LKEELAYLKKNHEEEMSVLKGQVGGQVSVEVDSAPG 210 220 230 240
::
:
:I:I:I
:
:III:
ID ...... LAKILSDMRSQYEVIAEKNRPd)AEAWFISQ~EELNRE 250 260 270 280
wt 1 S -Z
Acot E Z S g
wt l V S Z
A cot E ZV S Z
Fig 3. Immunoblot of extracts of sporulating cells of the parental strain, PYI7 (wt), and the cotE-deletion strain (AcotE) prepared at about stage II and late stage IV (at least 60% phase white endospores) of sporulation. The crude soluble (S) and insoluble (1) fractions were prepared as described in Materials and methods and 20 ~g of protein from each was electrophoresed on a 15% urea SDS-PAGE. Transfer and treatment with CotE antibody were as described in Materials and method.~. The broad lower bands (E arrows) are unique to the wild-type and are CotE antigens whereas the sharp, upper bands are cross-reacting antigens found in exponential cells of the wild type and of the CotE deletion strain as well as in post-exponential, soluble extracts of a mutant (spoOB) blocked at the initiation of sporulation.
There is a 23% identity, 45% with conserved differences and 62% (dotted lines) with replacements based on the mutation data matrix of Schwartz and Dayhoff as per the GCG program (Madison, WI), over a 73 amino acid overlap in a region of high a-helical content of the BIF. This region is involved in the coiled coil-polypeptide interactions required for filament formation [15, 16]. The CotE protein could also contain a lot of u-helical structures so such interactions are possible with this polypeptide. In addition, intermediate filament proteins contain repeat motifs of small residues (ser, gly) at their amino terminus and the BIF has the repeat sequence GGYGGG [15]. This sequence is present at the carboxyl terminus of another spore coat protein, CotT which also contains the repeat motif (GGGY)3 [5], so an interaction of CotT and CotE may be involved in filament formation (fig 4). Most intermediate filament proteins are distinguished by their unique solubility properties, ie generally insoluble in low or high ionic strength buffers and require 8 M urea or 6 M guanidine for solubilization [ 16]. They also reaggregate as 10-nm wide filaments upon dialysis against a low ionic strength buffer [16]. The low speed pellet from stage IV-V sporulating cells was extracted with reagents used for the solu-
665
Acidic Cytokeratin Intermediate Filamentprotein (> 40 kDa)
NH~~
0~ [
Head groups ['GGYGGG;'] LSSSNSSSS.]
Rod domain Qt-helix-coiled coils
(24 kDa)
~ racidicq or Facidicq = [.acidicl2 [. basicJ2
N
I
-
Cot T C O O H ~
10 nm filiament
I I
2
~
COOH 9
(7.8 k D a )
COOH
protofilament~
I Cot E (acidic)
~
filament
P,Y rich
~a2"~-
Fig 4. A schematic diagram of the structure of an acidic cytokeratin intermediate filament protein and its assembly into 10 nM filaments. The spore coat proteins, CotE with sequence homology in the a-helical rod domain (bracket) and CotT with homology at its carboxyl end to the cytokeratin head group may interact to form filaments.
l 1
2
3
4
5
2
3
4
6
........ ~
_qD
E-"~ |
L
~-T
Fig 5. Immunoblot with CotE (A) or CotT (B) antibodies of fractions resolved on a 15% urea SDS-PAGE. The crude insoluble fraction from late stage IV-V sporulating cells of the wild type (strain PYI7) or the cotE-deletion strain were washed and extracted with 8 M urea as described in Materials and methods. Following dialysis of this latter fraction, 10 ~tg of protein in the precipitates and soluble fractions were electrophoresed. Lanes A 1 and B 1: insoluble fraction from strain PYI7; lanes A2 and B2: insoluble
bilization of membrane proteins (n-octyl-glucoside, deoxycholate, Brij 35 or Triton X-100), as well as with low and high ionic strength buffers (see Materials and methods). Little Cote antigen was solubilized. The washed pellet was then extracted with 8 M urea and multiple Cote antigen bands were found (fig 5A, lane 1). Dialysis of the 8 M urea soluble fraction against 20 mM Tris, pH. 7.4, resulted in aggregation and most of the Cote antigen was again present in the insoluble fraction (fig 5A, lane 5). Portions of this fraction were mixed with an equal vol of 2% phosphotungstic acid and spread on carbon-
fraction from the cotE-deletion; lane A3: soluble fraction from strain PYI7; lane A4: soluble fraction from the cotE deletion. The insoluble fractions (as in lanes 1 and 2) were dissolved in 8 M urea and redialyzed. Following centrifugation, the precipitates from strain PYI7 (lanes A5 and B3) and the cotE-deletion (lanes A6 and B4) were dissolved in the buffer used for spore coats and electrophoresed. E arrow on left indicates major CotE antigen probably the monomer and T arrow on right indicates CotT monomer.
666 coated grids (fig 6). There was an extensive network of filaments present in preparations from stage IV-V of the wild type (fig 6A). These were 1.5-2 mn at the thinnest but riley tended to stick together to form thicker and larger fibers. These filaments were not found in extracts from the cotE deletion strain but there were less abundant, shorter and wider filaments present (fig 6B). Filaments were less prevalent and had a different appearance in preparations from the wild type at stage HI of sporulation (fig 6C). A less abundant network was present in preparations from stage V cells of strain NB200, a mutant lacking the CotT antigen (fig 6D). Multiple forms of both the CotE and CotT antigens were present in the precipitates following dialysis of the 8 M urea-soluble fraction (fig 5). This precipitate was dissolved in 8 M urea and again dialyzed against 20 mM Tris-10 mM 13ME, pH 7.4. There appeared to be fewer filaments (unpublished results) and while both antigens were still present in the insoluble fraction, they were less abundant (fig 5A, lane 5; 5B lane 3). In addition to containing a glycine-rich carboxyl end, CotT is relatively rich in tyrosine residues [5]. The multiple CotT antigen bands present in these precipitates could be due to cross-linking to each other or to CotE.
Discussion A deletion of the cotE gene resulted in extensive pleiotropic effects on the bacterial spore coat with the absence of both an under and outer coat and a deficiency of the inner coat [7]. At least some outer coat proteins are synthesized in mutant cells but apparently cannot be deposited on the spore due either to the absence of the undercoat and/or the deficiency of the inner coat. This defect can be partially corrected in a suppressor strain which forms lysozymeresistant spores. While these spore coats contained some outer coat proteins, the spores still germinated as did those from the parental strain (fig 1), so while they are protected against lysozyme, the coat was still unstable and presumably came off after germination. Given the relatively early time for the initiation of synthesis of CotE, it could be a major constituent of the undercoat which is deposited on the outer forespore membrane. Perhaps in conjunction with the CotT protein, an intermediate filament-like structure is formed which is anchored to this membrane. Since the CotT protein is synthesized later in sporulation (stage IV) than CotE, the initial deposition of CotE may be in a different conformation than when it subsequently interacts with CotT. This latter protein is present in the crude insoluble fraction containing CotE (fig 5B) and is also probably a component of the
¢
0
Fig 6. Negative stain of precipitates following 8 M urea extraction and dialysis of crude, washed insoluble fractions from sporulating cells (see Materials and methods). A. Strain PY 17 at stage IV-V of sporulation. B. The cotEdeletion strain at stage IV-V of sporulation. C. Strain PYI 7 at about stage III of sporulation. D. Strain NB200 (absence of CotT antigen) at stage V of sporulation. Magnification was X 77 000; 1 mm is 12.98 nm. striated inner coat [6], implying a dual function in coat structure. It may be significant that much of the CotT protein in spore coats from the cote deletion strain was present in its precursor form ([5] unpublished results). Processing may be integrated with the deposition of the outer coat. In addition to the sequence relatedness of Cote to a bovine intermediate filament protein, there is some homology to the active site of peroxidases (T Deits, personal communication). This activity has been implicated in the formation of dityrosine cross-linked proteins of the cortical membrane which is formed
667
following fertilization of sea urchin eggs [ 17]. CotT is a tyrosine-rich protein [5] and there is evidence for dityrosine in the coat insoluble fraction [3]. The presence of high molecular weight bands of CotT especially in the 8 M urea-soluble fraction of strain PYI7 (fig 5B) may be indicative of such crosslinking. The cross-linking could stabilize the interaction of Cote and CotT in filaments. CotT monomers may also be cross-linked to each other as components of the under and/or inner spore coat layers.
7
8 9 10
Acknowledgments
11
Research was supported by Grant DAAL03-87-K-0133 from the US Army. The technical assistance of Joan Clapper and Doryth Loewy is deeply appreciated.
12
13
References 1 Aronson AI, Fitz-James PC (1976) Structure and morphogenesis rff the bacterial spore coat. Bacteriol Rev 40, 360--402 2 Jenkinson HF, Sawyer WE), Mandelstam J (1981) Synthesis and order of assembly of spore coat proteins in Bacillus subtilis. J Gen Microbiol 123, 1-16 3 Pandey NK, Aronson At (1979) Properties of the Bacillus subtilis spore coat. JBacteriol 137, 1208-1218 4 Donovan W, Zheng L, Sandman K, Losick R (1987) Genes encoding spore coat polypeptides from Bacillus subtilis. J Mol Biol 196, l-d0 5 Aronson AI, Song HY, Bourne N (1989) Gene structure and precursor processing of a novel Bacillus subtilis spore coat protein. Mol Microbiol 3, 437 AqA 6 Bourne N, Fitz-James PC, Aronson AI (1991) Structural and germination defects of Bacillus subtilis spores with
14
15
16
17
altered contents of a spore coat protein, d Bacterioi ! 73, 6618-6625 Zheng L, Donovan WP, Fitz-James PC, Losick R (1988) Gene encoding a morphogenetic protein required in the assembly of the outer coat of the Bacillus subtilis endospore. Genes Dev 2, 1047-1054 CuRing S, Zheng L, Losick R (1991) Gene encoding two alkali-soluble components of the spore coat from Bacillus subtilis. J Bacterioi 173, 2915-2919 Zheng L, Losick R (1990) Cascade regulation of spore coat gene expression in Bacillus subtilis. J Mol Biol 212~ 645-660 Schaeffer P, Ionesco H, Ryter A, Ballassa G (1963) La ~,porulation Bacillus subtilis, 6tude g6n6tique et physiologique. Colloq Int CNRS 124, 553-563 Spizizen J (1958) Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc Nati Acad Sci USA 44, 1072-1078 Dedonder RA, Lepesant JA. Lepesant-Kejzlarov J, Billault A, Steinmetz Z, Kunst F (1977) Construction of a kit of reference strains for rapid genetic mapping in Bacillus subtilis 168. Appl Environ Microbio133, 989-993 Towbin H, Staehlin T, Gordon J (1979) Electrophoretic transfer of proteins from polyacrylamide gel to nitrocellulose sheets: procedure and some applications. Proc Nat! Acad Sci USA 76, 4350-4354 Bulinski JC, Kumar S, Titani K, Hauschka SD (1983) Peptide antibody specificity for the amino terminus of skeletal muscle a-actin. Proc Natl Acad Sci USA 80, 1506-15~0 Bader BL, Magin TM, Hatzfeld M, Franke WW (1986) Amino acid sequence and gene organization of cytokeratin no 19, an exceptional tail-less intermediate filament protein. Embo J 5, 1865-1875 Hatzfeld M, Franke WW (1985) Pair formation and promiscuity of cytokeratins: formation in vitro of heterotypic complexes and intermediate-sized filaments by homologous and heterologous recombinations of purified polypeptides. J Cell Bio1101, 1826-1841 Shapiro BM (1991) The control of oxidant stress at ferlilization. Science 252, 533-536
