The Bacillus subtilis spore coat provides “eat resistance” during phagocytic predation by the protozoan Tetrahymena thermophila

Abstract

Bacillus spores are highly resistant to many environmental stresses, owing in part to the presence of multiple “extracellular” layers. Although the role of some of these extracellular layers in resistance to particular stresses is known, the function of one of the outermost layers, the spore coat, is not completely understood. This study sought to determine whether the spore coat plays a role in resistance to predation by the ciliated protozoan Tetrahymena, which uses phagocytosis to ingest and degrade other microorganisms. Wild-type dormant spores of Bacillus subtilis were efficiently ingested by the protozoan Tetrahymena thermophila but were neither digested nor killed. However, spores with various coat defects were killed and digested, leaving only an outer shell termed a rind, and supporting the growth of Tetrahymena. A similar rind was generated when coat-defective spores were treated with lysozyme alone. The sensitivity of spores with different coat defects to predation by T. thermophila paralleled the spores' sensitivities to lysozyme. Spore killing by T. thermophila was by means of lytic enzymes within the protozoal phagosome, not by initial spore germination followed by killing. These findings suggest that a major function of the coat of spores of Bacillus species is to protect spores against predation. We also found that indigestible rinds were generated even from spores in which cross-linking of coat proteins was greatly reduced, implying the existence of a coat structure that is highly resistant to degradative enzymes.

Spores of Bacillus species are formed in sporulation, a process that is induced by low nutrient levels. The spores are metabolically dormant and very resistant to a variety of environmental stress factors including heat, radiation, desiccation, and freeze–thaw cycles (1). As a consequence of their dormancy and resistance, spores can survive for very long periods, certainly hundreds of years, and there are reports suggesting that spores may even survive for millions of years (2–4).

There are a number of factors that contribute to spore resistance to acute stress treatments (1). Many of these factors are a reflection of the different overall structure of the spore in comparison with that of a growing cell. One of the unique structural features of the spore is the coat, the outermost layer of spores of many Bacillus species, including B. subtilis (5–8). The B. subtilis spore coat is composed of several distinct layers and contains ≥30 proteins, almost all of which are spore-specific gene products (5–12). There is extensive cross-linking of a number of proteins in the coats, at least in part by a transglutaminase and perhaps by the formation of dityrosine cross-links as well (13–18).

Although there is much information on the composition, organization, and assembly of the spore coat, the precise function of this structure is less clear. The coat does protect against some, although not all chemicals, perhaps by serving as a reactive armor that detoxifies harmful chemicals before they damage essential components located further within the spore (1, 7, 19). However, it seems unlikely that the complex spore coat would have evolved solely to allow the dormant spore to resist toxic chemicals. It has also been suggested that the spore coat is important in the resistance of spores to mechanical disruption (7). However, most spore coat proteins are not essential for this resistance property (20). Another well established function of the coat is to provide a barrier against lytic enzymes that can degrade the peptidoglycan cortex lying below the coats (5, 7). Cortex hydrolysis can result either in spore death directly or in spore germination, a process that normally takes place only when nutrients return to the spore's environment (21). Although high levels of lytic enzymes are not typically encountered in natural environments, such enzymes are a part of the offensive enzymatic repertoire of organisms that prey on bacteria, including bacteriophage and unicellular eukaryotes (22–24). Because the latter are often found in environments in which spores can remain dormant (e.g., soil and water), it seems likely (i) that spores would resist predation by such organisms, and (ii) that the spore coat would be an important factor in spore resistance to such predation. Indeed, B. subtilis spores are resistant to passage through the mouse gastrointestinal tract (25), Bacillus thuringiensis spores are readily ingested by the ciliated protozoan Tetrahymena pyriformis but fail to be digested (26, 27), and at least some B. anthracis spores are resistant to killing and digestion within mammalian macrophages (28, 29). However, the specific role of the spore coat in resistance to predation has never been studied.

In this article, we report studies on the role of the spore coat in the resistance of B. subtilis spores to predation by the unicellular eukaryote Tetrahymena thermophila. In its natural environment, Tetrahymena meets its nutritional needs by ingesting and degrading other microorganisms, including bacteria, by phagocytosis. This behavior has allowed us to assess the resistance of wild-type B. subtilis spores, as well as a series of spores with mutations affecting the coat structure, to predation by Tetrahymena, and to obtain some insights into the mechanism of spore destruction.

Methods

B. subtilis Strains and Spore Preparation. All B. subtilis strains used in this work are isogenic derivatives of strain PS832, a prototrophic derivative of strain 168. The strains used and details of their construction are described in Table 2, which is published as supporting information on the PNAS web site. New B. subtilis strains constructed in this work were prepared by transformation (17) of strains with low amounts of chromosomal DNA with selection for resistance to appropriate antibiotics including: kanamycin, 10 μg/ml; chloramphenicol, 5 μg/ml; tetracycline, 10 μg/ml; spectinomycin, 100 μg/ml; or lincomycin (25 μg/ml) plus erythromycin (1 μg/ml).

Spores of all strains were prepared at 37°C on 2× SG medium agar plates without antibiotics and cleaned and stored as described (30). In a few cases, pyridine-2,6-dicarboxylic acid [dipicolinic acid (DPA)] was added to 200 μg/ml to the sporulation medium. All spore preparations used were free (≥98%) of growing cells, germinated spores or cell debris as observed in a phase contrast microscope. Before feeding to Tetrahymena, spores were rewashed with water, collected by centrifugation, and resuspended in 10 mM Tris·HCl (pH 7.4).

Spore resistance to lysozyme or 0.25% sodium hypochlorite (pH ≈ 11) was measured by assessing spore viability after these treatments as described in refs. 17, 19, and 31. The spore's large depot (≈10% of dry weight) of DPA was extracted and analyzed as described in refs. 32 and 33.

Culturing of Tetrahymena and Spore Feeding Conditions. T. thermophila, strain CU428.2, was grown on SPPA medium containing 250 μg/ml penicillin-G, 250 μg/ml streptomycin sulfate, and 0.25 μg/ml amphotericin B (34) at 30°C. For spore ingestion experiments, Tetrahymena cultures were initiated by adding 0.5 ml of a stock culture to 9.5 ml of fresh SPPA medium and culturing overnight with shaking at 100 rpm. The Tetrahymena were then harvested by centrifugation (1 min; 500 × g) and rinsed with an equal volume of 10 mM Tris·HCl (pH 7.4), and the final cell pellet was suspended in an equal volume of 10 mM Tris·HCl (pH 7.4). The cells were starved for 2–4 h at room temperature before use in spore ingestion experiments. T. thermophila cell concentrations were determined by appropriately diluting an aliquot of cells in 5% formaldehyde fixative and counting the number of cells in four 5-μl droplets by light microscopy (×100 magnification). To prepare a Tetrahymena cell-free exudate, cells grown as described above were incubated at room temperature for 24 h in 10 ml of 10 mM Tris·HCl (pH 7.4) at a concentration of 1 × 10⁴ cells per ml with 20 μl of 2-μm-diameter red fluorescent polystyrene microspheres (Duke Scientific, Palo Alto, CA) to stimulate phagocytosis. The Tetrahymena cells were removed initially by centrifugation and then by filtration of the supernatant through an 8-μm Nitex membrane (Tetko, Elmsford, NY).

Spores of various strains (1.5 × 10⁸ ml; final OD_600nm = 7.5) were incubated at 24°C with gentle shaking (50 rpm) in 10 mM Tris·HCl (pH 7.4) with an initial inoculum of 2 × 10³ Tetrahymena cells per ml. Typically, 0.5-ml cultures were set up in 2-ml polypropylene tubes. At various times, the OD_600nm of the culture was measured, and the numbers of T. thermophila cells were counted as described above. Aliquots were also diluted 10-fold in sterile water, heated at 70°C for 30 min to inactivate T. thermophila, sonicated briefly (≈1 min) in a bath sonicator to disperse aggregated material, and serially diluted in sterile water. This treatment may not release all ingested spores from T. thermophila, but at any one time these are only a small minority of total spores in incubations. Aliquots (10 μl) of the dilutions were spotted on LB medium (32) agar plates containing appropriate antibiotics, the plates were incubated for ≈24 h at 30–37°C, and bacterial colonies were counted. Incubation for longer than 24 h gave no increase in colonies. For incubations with spores of strains that lack all nutrient germinant receptors (strains FB72 and PS3329), after the mild sonication step, the spores were treated with 50 mM Ca²⁺-DPA as described in ref. 35 to trigger spore germination, because the spores of these strains germinate extremely poorly with nutrients.

All of the above analyses were carried out in duplicate, with average values reported. In most cases where spores were digested, results were confirmed by repeating all analyses with independent preparations of Tetrahymena and spores.

Samples of cotE spores (strain PS3394) at an OD_600nm of 5 in 10 mM NaPO₄ buffer (pH 7.4) plus 1 mM MgSO₄ were incubated for 30 min at 37°C with egg white lysozyme (Sigma; 250 μg/ml) plus DNase I (Sigma; 1 μg/ml). The DNase was added to reduce the viscosity of the resultant lysate. As expected, this lysozyme treatment inactivated ≈99% of the cotE spores.

Light and Electron Microscopy. Tetrahymena cells were monitored by brightfield microscopy (×200 magnification), and spores were examined by phase-contrast microscopy (×1,000 magnification). Live Tetrahymena cells were photographed by using a Nikon Diaphot-TMD microscope at ×200 magnification. For measurements of phagosome size, cells were fixed in 2.5% formaldehyde before photography.

For electron microscopy, 3-ml cultures of wild-type spores (OD_600nm = 5.0) and cotE spores (OD_600nm = 7.5) were incubated with Tetrahymena (initial concentrations of 1 × 10⁴ cells per ml and 2 × 10³ cells per ml, respectively) for 20 h. cotE spores were also digested with lysozyme as described above. Samples (1 ml from cultures with Tetrahymena, and 2 ml from lysozyme-treated cotE spores) were collected by centrifugation, fixed in 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 2 h, rinsed in 0.1 M cacodylate buffer, and embedded in SeaPlaque low-melting-temperature agarose (BioWhittaker). The agarose-embedded pellets were postfixed in 1% osmium tetroxide/0.8% potassium ferricyanide in 0.1 M cacodylate buffer (pH 7.4), treated with 1% aqueous uranyl acetate, and dehydrated in ascending ethanol solutions. They were then infiltrated with mixtures of PolyBed epoxy resin (Polysciences) and propylene oxide, embedded in pure PolyBed resin, and polymerized at 60°C for 48 h. Thin sections were cut with a diamond knife, collected on 200-mesh uncoated specimen grids, stained with uranyl acetate and lead citrate, and observed and photographed in a Philips CM10 transmission electron microscope. The developed negatives were scanned in a DuoScan T2500 scanner (Agfa). Levels, brightness, and contrast of photographs were adjusted by using photoshop (Version 6.0.1, Adobe Systems, San Jose, CA).

Results

Ingestion and Digestion of Spores of B. subtilis Strains with Various cot Mutations by T. thermophila. We initially examined the fate of wild-type B. subtilis spores (strain PS533) and cotE mutant spores during a 48-h coincubation with T. thermophila. For the wild-type spores, there was no significant decrease in spore titer over the coincubation period nor any increase in T. thermophila numbers (Fig. 1 and Table 1). Moreover, incubation for an additional 3 d also resulted in no changes in cell titer or spore appearance (data not shown). In contrast, spores of a cotE strain, which have a very defective coat (5, 7), displayed a decrease of ≈100-fold in the spore titer over the 48-h incubation period, and there was a 10- to 30-fold increase in T. thermophila cell numbers (approximately three to five doublings; Fig. 1 and Table 1). Control experiments showed that there was no decrease in cotE spore numbers when these spores were incubated for 48 h without T. thermophila (data not shown).

Fig. 1.

Growth of Tetrahymena on B. subtilis spores. Plots of spore survival versus time (A) and Tetrahymena cells per ml versus time (B) are shown for wild-type spores (strain PS533) (open circles) and cotE spores (strain PS3394) (filled circles). Incubations were initiated with 2 × 10³ Tetrahymena cells per ml and spores at an OD_600nm = 7.5.

Table 1. Tetrahymena, hypochlorite, and lysozyme resistance of B. subtilis spores.

	Tetrahymena coculture assay*			Spore sensitivity
Spore strain (genotype)	Tetrahymena (cells per ml × 103)	OD600nm	Spore rinds†	OCI-‡	Lys.§
PS533 (wt)	2	6.8	-	-	-
FB72 (ger3)	1	6.0	-	nd	nd
PS3394 (cotE)	47	0.4	++	+++	+++
PS3738 (safA)	43	0.6	++	++	++
PS3735 (spoVID)	32	0.8	++	++	++
PS3739 (ysxE)	20	2.5	+	+	+
PS3737 (yutH)	24	2.6	+	+	+
PS3736 (cotH)	2	4.2	-	+	-
KB29 (gerQ)	2	7.0	-	-¶	-¶
PS2495 (sodA)	1	8.1	-	-	-
PS3759 (sodA tgl)	1	8.4	-	-	-
KB81 (tgl)	1	7.0	-	-¶	-¶
PS3329 (cotE ger3)	30	0.6	++	nd	nd
PS3330 (cotE sleB)	39	0.5	++	nd	nd
PS3741∥ (cotE sleB spoVF)	52	1.0	++	nd	nd
PS3741** (cotE sleB spoVF)	60	0.6	++	nd	nd
PS3753 (cotE tgl)	17	1.1	++	nd	nd
PS3760 (cotE sodA tgl)	56	0.8	++	nd	nd

nd, not determined.

^*Spores of various strains were incubated with Tetrahymena for 48 h as described in Methods.

^†-, <2% of spores present as rinds; +, 10-40% spores converted to rinds; ++, >95% spores converted to rinds.

^‡Hypochlorite sensitivity. The symbols used reflect the time for 99% killing by hypochlorite. -, >45 min; +, 25-35 min; ++, 5-12 min; +++, <1 min.

^§Lysozyme sensitivity. -, <20% spore killing; +, 20-50% spore killing; ++, 50-90% spore killing; +++, >95% spore-killing.

^¶Data taken from refs. 17 and 32.

^∥Sporulated without added DPA.

^**Sporulated with 200 μg/ml DPA.

The failure of wild-type spores to support Tetrahymena growth was not the result of a failure to ingest the spores. Both light and electron microscopy revealed that the wild-type spores were efficiently ingested by Tetrahymena, forming numerous food vacuoles (Figs. 2 and 3A). However, during the course of coincubation, spherical particles accumulated in the cultures (Fig. 2C), which appeared to be clusters of spores representing the expelled contents of food vacuoles. Electron microscopy revealed that both intra- and extracellular wild-type spores remained intact (Fig. 3 A and C). The cotE spores were also efficiently ingested, and over the course of incubation, flocculent debris accumulated in the culture (Fig. 2 D–F). Electron microscopy of Tetrahymena cells at 20 h after cotE spores addition revealed some phagosomes containing relatively intact spores and others with hollow spherical structures that we term “rinds” (Fig. 3B). Rinds also accumulated outside the Tetrahymena cells (Fig. 3 B and D), presumably after being released from phagosomes.

Fig. 2.

Coculture of Tetrahymena and B. subtilis spores. Tetrahymena cells fed wild-type (wt) (strain PS533) spores for 1 h (A) and 18 h (B) are shown. At both times, multiple dark bodies representing spore-filled phagosomes were seen within the Tetrahymena cells. Debris resembling the contents of phagosomes (C) is also evident at 18 h. Also shown are Tetrahymena fed cotE spores (strain PS3394) (cotE) for 1 h (D) or 18 h (E), and debris remaining from digested spores after 18 h (F). Images were obtained by brightfield microscopy at ×200 magnification. (Scale bar: 20 μm.)

Fig. 3.

Electron micrographs of Tetrahymena cells and spores. (A) Section of a Tetrahymena cell that had been coincubated with wild-type (strain PS533) spores for 20 h. Arrows indicate selected phagosomes or food vacuoles, and “ma” denotes the macronucleus. (B) Section of a Tetrahymena cell coincubated with cotE (strain PS3394) spores for 20 h. An early phagosome with partially digested spores (black arrow) and a late phagosome containing mostly spore rinds (white arrow) are indicated. (C) Extracellular wild-type spores after a 20-h coincubation with Tetrahymena. Note that intact spores were not completely permeable to fixatives but can be differentiated from spore rinds by the presence of internal, darkly staining material. (D) Extracellular rinds from cotE spores after a 20-h coincubation with Tetrahymena. A spore that appears to be germinating (arrowhead) and a vegetative bacterium (arrow) are also indicated. (E) Rinds produced by digestion of cotE spores with lysozyme. (Scale bars: 2 μm.)

The cotE spores have a major defect in coat formation that is likely responsible for their ability to be digested by Tetrahymena (5, 7). This cotE spore coat defect is also indicated by their increased sensitivity to lysozyme and hypochlorite, as compared with that of wild-type spores (Table 1) (5, 7, 17, 19). To obtain further insight into the role of the spore coat in resistance to predation, spores of strains with mutations affecting the spore coat to different degrees were examined. Spores with safA or spoVID mutations were also almost completely digested in 48 h and supported Tetrahymena growth to levels comparable to that with cotE spores (Table 1 and data not shown). SafA and SpoVID play major roles in spore coat morphogenesis (7, 8, 11), and in their absence, spore coats are defective such that the spores are sensitive to both lysozyme and hypochlorite, although not as sensitive as cotE spores (Table 1).

Many genes coding for spore coat proteins, including cotE, safA, and spoVID, are transcribed by RNA polymerase containing the sporulation-specific sigma factor, σ^E (6, 7, 8, 36). In a preliminary study using spores of strains with the PY79 background carrying single mutations in a number of σ^E-controlled genes, ysxE and yutH mutations caused slight spore sensitivity to hypochlorite (R. I. Tennen, K.R., and P.S., unpublished results). In our PS832 background, spores of ysxE and yutH strains were also partially digested by T. thermophila in 48 h and supported Tetrahymena growth, albeit to lower levels than cotE, spoVID, or safA spores (Table 1). Incubation with Tetrahymena for 120 h, or with 5-fold more Tetrahymena for 48 h, did not significantly reduce the number of viable ysxE or yutH spores (data not shown), suggesting the existence of a subpopulation of digestion-resistant spores in these spore preparations. Spores of the ysxE and yutH strains with the PS832 background also exhibited slight sensitivity to hypochlorite and were slightly sensitive to lysozyme (Table 1). To date, specific roles for YsxE and YutH in spore coat assembly have not been established, although both proteins have been shown to be in the spore coat (11, 36, 37). Finally, cotH, gerQ, sodA, sodA tgl, and tgl spores incubated with Tetrahymena displayed no decreases in numbers over 48 h, and there was no growth of Tetrahymena with these mutant spores (Table 1 and data not shown). CotH plays a role in assembly of other coat proteins, albeit a minor one, GerQ is a major cross-linked coat protein, SodA is a superoxide dismutase found in spore coats that is involved in coat protein cross-linking, and Tgl is a transglutaminase that also catalyzes cross-linking of coat proteins, including GerQ (11, 13, 15, 17, 38). In our PS832 background, the gerQ, sodA, sodA tgl, and tgl spores were not sensitive to lysozyme or hypochlorite (Table 1). Other workers have found that a cotH mutation causes a significant defect in spores of a B. subtilis strain with a different genetic background than the strain used in our work (39, 40). However, such a significant defect was not observed when using cotH spores with our PS832 genetic background, beyond a slight sensitivity of these spores to hypochlorite (Table 1). Overall, the sensitivity of the mutant spores to digestion by Tetrahymena correlated quite well with their sensitivity to lysozyme, and to a lesser degree with their sensitivity to hypochlorite.

Mechanism of Spore Eating by T. thermophila. Although spores with significant coat defects were killed and digested by T. thermophila, the mechanism of this process was not immediately clear. Possible explanations include (i) that spores with defective coats are readily germinated within the T. thermophila phagosomes, and it is the germinated spores that are digested; (ii) that phagosomal degradative enzymes lyse the coat-defective spores directly; or (iii) that the spores with defective coats are lysed outside the T. thermophila by degradative enzymes released by the cell (41). The last possibility is unlikely in this case, because a cell-free exudate from 1 × 10⁴ Tetrahymena cells per ml incubated for 24 h without spores, but with polystyrene beads to stimulate phagocytosis, gave no lysis of cotE spores in 48 h (data not shown). However, germination of B. thuringiensis spores within Tetrahymena phagosomes has been reported (27). Some support for the germination of cotE B. subtilis spores within phagosomes was provided by electron microscopy, because spores that appear to be in the process of germinating were sometimes observed within phagosomes, and a small number of germinated spores and vegetatively growing cells were observed in the medium (Fig. 3 B and D). To more critically distinguish between the first and second explanations, we examined the fate of spores of a strain (PS3329) that lacks not only CotE, but also all functional germinant receptors, and thus germinates extremely slowly in response to nutrient germinants (42). Spores of this strain were digested as readily as cotE spores (Table 1). In contrast, spores that lacked all functional nutrient germinant receptors but had a wild-type spore coat (strain FB72) were not digested by T. thermophila (Table 1). Spores of a strain (PS3330) that lacks both CotE and SleB, the latter being one of the spore's two lytic enzymes either of which is required for cortex degradation during spore germination (21), were also digested as readily as cotE spores (Table 1). Although the spores of strain PS3330 retain the gene for the second redundant cortex lytic enzyme, CwlJ, this enzyme is not present in cotE spores because it either is not assembled in the coat of these spores or is rapidly degraded following its assembly (38, 43). Thus, the cotE sleB spores lack both CwlJ and SleB and cannot complete germination on their own, yet are still readily digested by Tetrahymena.

Effects of DPA on Spore Eating. In the early stages of this work, we noted that the phagosomes in T. thermophila that had ingested wild-type spores sometimes appeared somewhat smaller than phagosomes containing cotE spores (Fig. 2 and data not shown). Although there could be many reasons for this difference, one intriguing possibility was that the phagosome enlargement with the cotE spores was due to the release of a large amount of DPA upon spore digestion. To test this possibility directly, we examined the appearance of phagosomes in T. thermophila that were eating spores of strain PS3741. In addition to CotE, this strain also lacks the spoVF operon that encodes DPA synthetase, and spoVF spores normally lack DPA (44, 45). However, strain spoVF strains including PS3741 can take up exogenously added DPA and accumulate ≈75% of normal DPA levels in the dormant spores (44) (data not shown). Strain PS3741 also contains a sleB mutation to prevent the rapid germination of the DPA-less spores, as shown previously (35). DPA-less and -replete spores of strain PS3741 were both readily consumed and supported Tetrahymena growth (Table 1). The average size of phagosomes in Tetrahymena cells eating DPA-less spores (6.9 ± 0.8 μm, n = 20) was not statistically different from that in cells eating DPA-replete spores (7.7 ± 1.1 μm, n = 28). As a result, the data provide no support for the hypothesis that DPA release causes an increase in the size of phagosomes.

Production of Spore Rinds upon Eating of Spores of Different Strains. It was surprising in some respects that even though cotE and other coat-defective spores were digested, a significant “rind” remained (Fig. 3D). The rinds were clearly apparent in the electron microscope but could also be readily observed by phase-contrast microscopy as relatively transparent spherical or hemispherical structures with no internal contents. The rinds have a diameter similar to that of an intact spore (Fig. 3 C and D), and this strongly suggests that they represent at least a portion of a spore coat layer that is resistant to digestion by Tetrahymena's phagosomal enzymes. Because there is significant cross-linking of spore coat proteins, it is possible that this rind is composed largely of highly cross-linked coat protein. To determine whether this was the case, we examined the production of rinds from either cotE tgl or cotE sodA tgl spores, because sodA and tgl encode a superoxide dismutase and transglutaminase, respectively, thought to be involved in generating cross-links in spore coat proteins (13, 17). However, the production and appearance of the rinds was not reduced by the tgl or sodA tgl mutations (Table 1).

Because lysozyme sensitivity correlated well with the ability of Tetrahymena to digest spores (Table 1), we also examined cotE spores that had been treated solely with lysozyme by electron microscopy (Fig. 3E). Structures similar to rinds were also observed in this analysis, as well as by phase-contrast microscopy (data not shown), although much additional debris was also observed. The debris presumably represents spore constituents that are not digestible by lysozyme (e.g., ribosomes, cell membranes, etc.) but which are susceptible to digestion in Tetrahymena phagosomes.

Discussion

The major conclusion of this study is that the coats are a major defense against spore predation by protozoa. This is perhaps not surprising given the importance of the coats in preventing access of lytic enzymes to the spore's cortical peptidoglycan. However, this crucial role for the coat in protecting spores against predation has never been shown. The good parallel between spore resistance to predation by T. thermophila and spore lysozyme resistance strongly suggests that it is through some type of lytic enzyme that this protozoan kills and digests spores. Indeed, Tetrahymena phagosomes are known to contain numerous degradative enzymes (46, 47) (M. E. Jacobs, L. DeSouza, K. W. M. Siu, and L.A.K., unpublished results), and the recently completed T. thermophila macronuclear genome sequence (The Institute for Genome Research, http://tigrblast.tigr.org/er-blast/index.cgi?project=ttg) includes three genes predicted to encode lysozymes (http://db.ciliate.org/cgi-bin/gbrowse/tt-genomic).

An alternative possibility is that spores, in particular coat-defective spores, first germinate in the T. thermophila phagosome and it is these germinated spores that are then digested. However, spores lacking all functional nutrient germinant receptors as well as CotE were consumed as well as cotE spores, and even cotE spores that lack the two cortex lytic enzymes, CwlJ and SleB, either of which is essential for completion of spore germination, were readily digested. Thus, these results further suggest that the loss of coat-defective spores is due to the direct digestion and lysis of these spores by protozoal enzymes.

The specific role of some individual coat proteins in resistance to spore lysis by protozoal enzymes is clear in that the three coat proteins whose loss has the most dramatic effect (CotE, SafA, and SpoVID) are morphogenic proteins responsible for assembly of a number of other coat proteins (6–8, 11). Indeed, cotE mutants lack the spore's outer coat as well as many inner coat proteins. However, cotH mutants are also defective in the assembly of some coat proteins (6–8, 11), yet cotH spores are neither lysozyme sensitive nor subject to digestion by T. thermophila. The partial digestion of yutH and yxsE spores is consistent with their partial lysozyme sensitivity, and this partial digestion and partial lysozyme sensitivity was seen with two independent spore preparations of these strains. The reason that only a fraction of the yutH and ysxE spores are digested by T. thermophila is not clear, but this may reflect some stochastic defect in these spores such that only a fraction has a severe defect with the majority having a wild-type or near-wild-type phenotype. Unfortunately, specific coat defects in ysxE and yutH spores, beyond the absence of YsxE or YutH, have not been identified. The sensitivity of spores with coat defects to predation by T. thermophila and the range in predation sensitivity observed suggest that assessment of susceptibility to protozoal predation may be a potential bioassay to assess spore coat integrity. However, measurement of spore lysozyme sensitivity appears to be a good surrogate for measurement of spore sensitivity to protozoal predation.

Another notable finding in this work is that the release of large amounts of DPA and its associated divalent cations upon spore digestion had no noticeable effect on the appearance of phagosomes in T. thermophila, or on T. thermophila itself. DPA is present in spores at extremely high concentrations (≥1 M if all were soluble), and its release in phagosomes would likely give a DPA concentration in the phagosome of ≥5 mM if all released DPA remained in the phagosome. This concentration might be high enough to affect the intraphagosomal pH, which is normally reduced to a value of pH of 4–5 during phagosome acidification (48, 49). DPA has also been reported to be toxic, at least for insects (50). However, the phagosomes and T. thermophila appear capable of dealing with the large amount of DPA released upon cotE spore lysis.

A final notable observation concerns the residue remaining after protozoal digestion of coat-defective spores. This residue looks like coat material, but its specific origin is not clear, because it was present in cotE spores in which at least some components of the insoluble fraction of the spore coat (cotX, Y, and Z gene products) are not assembled (7, 12). The rinds were also not eliminated by sodA and tgl mutations that eliminate much of the cross-linking of spore coat proteins (13, 17). Perhaps direct analysis of the proteins in these spore rinds may indicate their composition.