Interaction between Individual Protein Components of the GerA and GerB Nutrient Receptors That Trigger Germination of Bacillus subtilis Spores

Abstract

Germination of Bacillus subtilis spores via the GerA nutrient receptor was suppressed by GerAC lacking the diacylglycerylated cysteine essential for receptor function. Overexpression of the C protein of the GerB nutrient receptor also suppressed the function of both the GerA receptor and a variant GerB receptor, GerB*. These findings suggest that GerAC and GerBC interact with their respective A and B proteins in GerA or GerB receptors and that GerBC potentially interacts with GerAA-GerAB. However, GerAC did not appear to interact with GerBA-GerBB.

Spores of Bacillus species are metabolically dormant and can remain so for long periods of time. However, nutrient receptors located in the dormant spore's inner membrane can sense the presence of exogenous nutrients and trigger the process of spore germination that leads to a resumption of metabolism in spore outgrowth (7, 13, 14, 15, 20, 21, 26). There are three functional nutrient receptors in Bacillus subtilis spores, the products of the homologous tricistronic gerA, gerB, and gerK operons (termed gerA operon homologs) (13, 14, 15, 21, 25, 32). Based on their amino acid sequences, the A and B proteins encoded by gerA operon homologs are most likely integral membrane proteins (13, 21, 32). The third encoded protein, C, is probably not an integral membrane protein, but the C proteins have a signal peptide followed by a consensus sequence for diacylglycerol addition to a cysteine residue and a signal peptidase II cleavage site (8, 13, 21, 23, 32). Consequently, it is likely that some combination of the signal peptide and diacylglycerol addition ensures that the C proteins are also in the spore's inner membrane.

The addition of diacylglycerol to the C proteins of GerA receptor homologs is important for receptor function, as loss of the only B. subtilis prelipoprotein diacylglycerol transferase, Lgt (also called GerF), reduces significantly the spore germination with nutrients (8, 10, 23). The lack of diacylglycerol addition to C proteins almost completely eliminates the function of the GerA receptor that responds to l-alanine, significantly but not completely abolishes spore germination via the GerB receptor, but has little effect on germination via the GerK receptor (8).

The presence of genes for the A, B, and C proteins of GerA receptor homologs in an operon suggests that each receptor is a complex of the appropriate A, B, and C proteins, and loss of any cistron of a particular gerA operon homolog eliminates the function of the encoded receptor (13, 14, 15). While there is as yet no biochemical evidence for physical interaction between the A, B, and C proteins of any individual receptor, there is genetic evidence that GerAA interacts with GerAB and that GerBA interacts with GerBB (15, 18). However, there is no evidence for the physical interaction of the C protein with the A and B proteins of either of these receptors. It was shown recently that substitution of alanine for the diacylglycerylated cysteine residue in GerAC (residue 18) or GerBC (residue 20) reduces GerA and GerB receptor function 200- and 7-fold, respectively (8). One possible explanation for this result is that without diacylglycerol addition GerAC and GerBC do not localize to the spore's inner membrane, leading to lack of assembly or stability of the appropriate A and B proteins. While this explanation has not been dismissed for the GerA receptor, it is not true for the GerB receptor, as GerBA is at normal levels in the inner membrane of lgt spores (8). In addition, the GerK receptor functions normally in lgt spores, as does GerKC in which alanine has replaced the diacylglycerylated cysteine, although the loss of gerKC eliminates GerK receptor function (8). These findings suggest that the GerB and GerK receptors, and by inference the GerA receptor, are in the inner spore membrane even if the receptor's C proteins are not diacylglycerylated. Indeed, a number of membrane proteins that are normally diacylglycerylated are in the plasma membrane of growing lgt cells, although the functions of these proteins are often compromised (9, 10, 30). An alternative explanation for the reduced function of the GerA and GerB receptors that are not diacylglycerylated, therefore, is simply that this modification is essential for normal GerA and GerB receptor function. If this alternative explanation is correct, then gerAC or gerBC in which an alanine codon has replaced the diacylglycerylated cysteine codon might be dominant negative to the wild-type allele if all three proteins encoded by the gerA or gerB operons physically interact. This communication presents results in support of this prediction.

Plasmid and strain construction.

All work was carried out with B. subtilis strains (Table 1) derived from and isogenic with strain PS832, a prototrophic derivative of strain 168. Only outlines of plasmid construction in Escherichia coli and generation of B. subtilis strains by transformation with linearized plasmid DNA are given; details are available on request. PCR and DNA sequence analyses confirmed the chromosomal structures of B. subtilis transformants.

TABLE 1.

B. subtilis strains used in this study

Strain	Genotype and phenotypea	Source or reference
FB10	gerBB1*	18
FB72	ΔgerA ΔgerB ΔgerK Cmr Spr Tcr	19
PS832	Wild type	Laboratory stock
PS3604	gerBCCys20 Spr	8
PS3605	gerBCAla20 Spr	8
PS3608	ΔgerA Cmr	8
PS3611	ΔgerB ΔgerK Cmr Emr Spr	8
PS3612	gerACAla18 ΔgerB ΔgerK Cmr Emr Spr	8
PS3629	ΔamyE::(PgerA-gerACCys18) gerACAla18 ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3630	ΔamyE::(PgerA-gerACAla18) ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3631	ΔamyE::(PsspB-gerACCys18) gerACAla18 ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3632	ΔamyE::(PsspB-gerACAla18) ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3633	ΔamyE::(PsspB-gerACCys18) ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3636	ΔamyE::(PsspB-gerACAla18) gerBB1* Kmr	This work
PS3637	ΔamyE::(PsspB-gerACCys18) gerBB1* Kmr	This work
PS3656	gerBB1* gerBCCys20 Spr	This work
PS3657	gerBB1* gerBCAla20 Spr	This work
PS3658	ΔamyE::(PsspB-gerBCCys20) gerBB1* Kmr	This work
PS3659	ΔamyE::(PsspB-gerBCAla20) gerBB1* Kmr	This work
PS3660	ΔamyE::(PsspB-gerBCCys20) ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3661	ΔamyE::(PsspB-gerBCAla20) ΔgerB ΔgerK Cmr Emr Kmr Spr	This work
PS3662	ΔamyE::(PsspB-gerBCCys20) ΔgerA Cmr Kmr	This work
PS3663	ΔamyE::(PsspB-gerBCAla20) ΔgerA Cmr Kmr	This work
PS3692	ΔamyE::(PgerA-gerBCCys20) ΔgerA Cmr Kmr	This work
PS3693	ΔamyE::(PgerA-gerBCAla20) ΔgerA Cmr Kmr	This work
PS3694	ΔamyE::(PgerA-gerBCCys20) gerBB1* Kmr	This work
PS3695	ΔamyE::(PgerA-gerBCAla20) gerBB1* Kmr	This work

^aCm^r, resistance to chloramphenicol (5 μg/ml); Em^r, resistance to erythromycin (1 μg/ml) and lincomycin (25 μg/ml); Km^r, resistance to neomycin (7 μg/ml); Sp^r, resistance to spectinomycin (100 μg/ml); Tc^r, resistance to tetracycline (10 μg/ml).

Strains expressing gerAC^Cys18 (wild-type allele with Cys at position 18) or gerAC^Ala18 at the amyE locus from the gerA promoter (P_gerA) were generated as follows. A Km^r cassette from plasmid pDG780 (6) was inserted into plasmid pTEF1/Zeo (Invitrogen, Carlsbad, Calif.), giving plasmid pTI1. A ∼400-bp fragment including P_gerA and the ribosome-binding and translation start sites of the gerAA cistron was PCR amplified from chromosomal DNA of strain PS3611. The sequence around the gerAA translation start site in this fragment was 5′-GGTGACCCATATGAGA-3′ (the initiation codon is shown in boldface type) instead of 5′-GGTGACCTCATTGGAA-3′ as in wild-type gerAA. The spacing between the ribosome-binding site and the translation initiation codon is the same in both cases, but the translation initiation codon and the three nucleotides upstream of the initiation codon differ. This fragment was inserted in plasmid pTI1, giving plasmid pTI2. A 1.3-kb fragment including the gerAC translation start site and coding sequence and extending beyond the likely gerA operon's transcription terminator was amplified from strains PS3611 and PS3612 (5, 32). These fragments were inserted in plasmid pTI2, giving plasmids pTI3a (fragment from PS3611) and pT13b (fragment from PS3612). Fragments (3.2 kb) containing P_gerA upstream of gerAC from pTI3a and pTI3b were inserted in plasmid pDG364 (4), giving plasmids pTI4a (from pTI3a) and pTI4b (from pTI3b). The transformation of B. subtilis strain PS3612 with pTI4a gave strain PS3629, and the transformation of strain PS3611 with pTI4b gave strain PS3630.

The construction of strains overexpressing gerAC at amyE from the strong forespore-specific sspB promoter (P_sspB) (12, 29) was as follows. A ∼400-bp fragment including P_sspB and up to the sspB translation initiation codon was amplified from chromosomal DNA of strain PS832. This fragment was combined with most of plasmids pTI6a and pTI6b (derived from pTI4a and pTI4b, respectively), giving plasmids pTI7a (from pTI6a) and pTI7b (from pTI6b). The transformation of strains PS3611 and PS3612 with pTI7a gave strains PS3633 and PS3631, respectively; the transformation of strain PS3611 with pTI7b gave strain PS3632. Previous work (1, 12, 29) suggests that the level of GerAC overexpressed from P_sspB will be ∼100-fold higher than when gerA is expressed from P_gerA.

To test the effects of GerAC overexpression on spore germination via the GerB receptor, we used spores of strains carrying a missense mutation in gerBB termed gerBB1*; the encoded GerB* receptor is sufficient to trigger spore germination in response to d-alanine, and even better with d-alanine plus d-glucose or with l-asparagine (1, 18). The transformation of strain FB10 with pTI7a and pTI7b gave strains PS3636 and PS3637, respectively. The transformation of strain FB10 with plasmids pBC^C20 and pBC^A20 (8) gave strains PS3656 and PS3657, respectively.

B. subtilis strains overexpressing gerBC^Cys20 or gerBC^Ala20 at amyE from P_sspB were constructed as follows. The gerBC gene was amplified from chromosomal DNA of strains PS3604 and PS3605; the fragments were combined with most of pTI7a, giving plasmids pTI8a (gerBC^Cys20) and pTI8b (gerBC^Ala20) in which gerBC transcription is from P_sspB. Transformation of strains FB10 and PS3611 by pTI8a gave strains PS3658 and PS3660, respectively; transformation of strains FB10 and PS3611 by pTI8b gave strains PS3659 and PS3661, respectively (Table 1). The transformation of strain PS3608 with chromosomal DNA of strains PS3660 and PS3661 gave strains PS3662 and PS3663, respectively. Since GerBA levels in spores with the gerB or gerB* operons transcribed from P_sspB were 200- to 500-fold higher than wild-type GerB levels (1, 20), levels of GerBC overexpressed from P_sspB will also likely be elevated 200- to 500-fold (2, 5, 12, 32).

For construction of B. subtilis strains overexpressing gerBC^Cys20 or gerBC^Ala20 at amyE from P_gerA, P_gerA from plasmid pTI6a replaced P_sspB in pTI8a and pTI8b, giving plasmids pTI9a (from pTI8a) and pTI9b (from pTI8b). The transformation of strains PS3608 and FB10 with these plasmids gave strains PS3692 and PS3694 (from pTI9a), respectively, and PS3693 and PS3695 (from pTI9b), respectively.

Spore preparation and measurement of spore germination.

Spores of various strains were prepared on 2× Schaeffer’s glucose medium agar plates without antibiotics at 37°C and cleaned and stored as described previously (16, 17). All spore preparations used were free (>98%) from growing or sporulating cells or cell debris, as determined by phase contrast microscopy.

Spore germination was measured in two ways. In one, spores at an optical density at 600 nm (OD₆₀₀) of 20 were heat shocked (30 min; 70°C) and cooled on ice. The heated spores were diluted in water, aliquots were spotted on Luria-Bertani medium agar plates (19) without antibiotics, and colonies appearing after 12 h at 37°C were counted. Since the gerA operon homologs are expressed only during sporulation (13, 21), mutations in these operons have no effects on cell growth or spore outgrowth, and this assay measures only differences in rates of spore germination (8, 19).

In the second assay, heat-shocked spores at an OD₆₀₀ of 1.5 were germinated at 37°C in (i) 1 mg of l-alanine/ml-6.95 mg (each) of KH₂PO₄ and NaHPO₄ (pH 7.25)/ml-100 mM NaCl-200 mM KCl, (ii) 10 mM d-alanine-25 mM Tris-HCl (pH 8), (iii) 10 mM d-alanine-10 mM d-glucose-25 mM Tris-HCl (pH 8), (iv) 3 mM l-asparagine-25 mM Tris-HCl (pH 8), or (v) 3 mM l-asparagine-500 μg of d-glucose/ml-500 μg of d-fructose/ml-50 mM KPO₄ (pH 7.4) (AGFK) (1, 8, 18). At various times, aliquots (1 ml) of the suspensions were centrifuged for 2 min in a microcentrifuge and the OD₂₇₀ of the supernatant fluid was determined. This assay measures the release of the spore core's depot of pyridine-2,6-dicarboxylic acid (dipicolinic acid [DPA]), an early event in spore germination (1, 24, 26). Ninety percent or more of the OD₂₇₀ released by germinating B. subtilis spores is due to DPA, as the UV spectrum of the material released is identical to that of pure DPA (data not shown), as expected based on previous work (22). Maximum rates of DPA release were calculated as described previously (1, 24).

Effects of GerAC^Ala18 on spore germination via the GerA receptor.

Analysis of the germination of spores of various strains by determination of colonies formed from spores applied to Luria-Bertani medium plates (Table 2) showed that germination of spores containing only the GerA receptor (strain PS3611) was ∼4-fold lower than that of wild-type spores (strain PS832). Germination of spores lacking all nutrient receptors (strain FB72) or lacking the GerB and GerK receptors and with a GerA receptor containing GerAC^Ala18 (strain PS3612) was greatly decreased in this assay (Table 2), as found previously (8, 18). When spores lacked the GerB and GerK receptors and contained both gerAC^Cys18 and gerAC^Ala18, at either amyE or gerA (strains PS3629 and PS3630), there was significant spore germination. However, spores of strains with the gerAC at amyE expressed from P_gerA exhibited different rates of germination depending on the locations of the wild-type and mutant gerAC cistrons. Spores of the strain with gerAC^Cys18 in the gerA operon and gerAC^Ala18 expressed from P_gerA at amyE (strain PS3630) germinated more rapidly than spores of the strain with the positions of the gerAC alleles reversed (strain PS3629). The same general result was obtained when gerAC alleles at amyE were overexpressed from P_sspB (∼100-fold stronger than P_gerA) (29) (strains PS3631 and PS3632). However, while overexpression of gerAC^Ala18 reduced spore germination significantly (strain PS3632), overexpression of gerAC^Cys18 increased spore germination (strains PS3631 and PS3633) (Table 2).

TABLE 2.

Germination of spores of various strains via the GerA receptor^a

Strain	Genotype	CFU/12 h	Rate of DPA release (% of maximum value)
FB72	ΔgerA ΔgerB ΔgerK	1.5 × 105	<2
PS832		1.5 × 108	104
PS3611	ΔgerB ΔgerK	4.1 × 107	100b
PS3612	gerACAla18 ΔgerB ΔgerK	1.5 × 105	<2
PS3629	ΔamyE::(PgerA-gerACCys18) gerACAla18 ΔgerB ΔgerK	1.4 × 107	25
PS3630	ΔamyE::(PgerA-gerACAla18) ΔgerB ΔgerK	4.7 × 107	104
PS3631	ΔamyE::(PsspB-gerACCys18) gerACAla18 ΔgerB ΔgerK	1.5 × 108	360
PS3632	ΔamyE::(PsspB-gerACAla18) ΔgerB ΔgerK	7 × 106	10
PS3633	ΔamyE::(PsspB-gerACCys18) ΔgerB ΔgerK		470
PS3660	ΔamyE::(PsspB-gerBCCys20) ΔgerB ΔgerK	3.4 × 105	9
PS3661	ΔamyE::(PsspB-gerBCAla20) ΔgerB ΔgerK	4.2 × 107	115

^aSpores were prepared, cleaned, and tested for colony-forming ability and for DPA release upon germination with l-alanine as described in the text. CFU values are the averages of duplicate determinations with two independent spore preparations; individual values differed by ≤25% from the average value. DPA release rates are the averages of duplicate measurements with two independent spore preparations; individual values differed by ≤25% from the average values.

^bThis value was set at 100%.

The results from assays of spore germination by colony formation were confirmed when rates of spore germination with l-alanine were measured by DPA release (Table 2). In particular, the overexpression of gerAC^Ala18 at amyE (strain PS3632) reduced the rate of spore germination with l-alanine 10-fold, while the overexpression of gerAC^Cys18 at amyE (strains PS3631 and PS3633) stimulated the rate of spore germination ∼4-fold (Table 2).

Effects of GerBC^Ala20 on spore germination via the GerB receptor.

Previous work has shown that GerB receptor function is significantly decreased when GerBC cannot be diacylglycerylated, as spore germination with AGFK which triggers germination through cooperative action of the GerB and GerK receptors, is reduced ∼7-fold in spores carrying gerBC^Ala20, although the analogous modification of gerKC has very little effect (8). Diacylglycerylation of GerBC was also required for GerB* receptor function, as germination of spores of strain PS3657 was reduced ∼11-fold compared to that of FB10 or PS3656 spores (Table 3). The overexpression of GerBC^Ala20 from P_sspB (strains PS3663 and PS3659) reduced the germination of gerA spores with AGFK and germination of gerBB1* spores with various germinants, but these reductions were small and not especially significant (Table 3). The overexpression of GerBC^Cys20 also reduced spore germination slightly via the GerB receptor (strain PS3662); via the GerB* receptor (strain PS3658) it gave even more of a reduction than that given by the overexpression of GerBC^Ala20 (strains PS3659 and PS3663) (Table 3). In contrast, the expression of either GerBC^Cys20 or GerBC^Ala20 from P_gerA did not inhibit notably the germination via the GerB or GerB* receptor (strains PS3692, PS3693, PS3694, and PS3695) (Table 3).

TABLE 3.

Germination of spores of various strains via the GerB or GerB* receptor^a

Strain	Genotype	Rate of spore germination (% of maximum) with:b
		AGFK	d-ala	d-ala/d-glu	l-asn
FB10	gerBB1*		100c	100c	100c
PS3608	gerA	100c
PS3636	amyE::(PsspB-gerACAla18) gerBB1*		71	64	68
PS3637	amyE::(PsspB-gerACCys18) gerBB1*		80	91	85
PS3656	gerBB1* gerBCCys20		104	103	97
PS3657	gerBB1* gerBCAla20	2	9	8	9
PS3658	amyE::(PsspB-gerBCCys20) gerBB1*		32	16	27
PS3659	amyE::(PsspB-gerBCAla20) gerBB1*		78	55	78
PS3662	amyE::(PsspB-gerBCCys20) gerA	56
PS3663	amyE::(PsspB-gerBCAla20) gerA	63
PS3692	amyE::(PgerA-gerBCCys20) gerA	106
PS3693	amyE::(PgerA-gerBCAla20) gerA	121
PS3694	amyE::(PgerA-gerBCCys20) gerBB1*		87	100	121
PS3695	amyE::(PgerA-gerBCAla20) gerBB1*		88	85	112

^aSpores of different strains were germinated with various agents, germination was measured by the release of DPA, and the maximum rates of germination were calculated as described in the text.

^bValues are the averages of duplicate measurements on three different spore preparations. Individual values differed by ≤35% from the average values. d-ala, d-alanine; d-glu, d-glucose; and l-asn, l-asparagine.

^cThis value was set at 100%.

Effects of GerAC^Ala18 or GerBC^Ala20 on spore germination via the GerB or GerA receptor, respectively.

The inhibition of spore germination via the GerA receptor by the overexpression of GerAC^Ala18 and of germination via the GerB or GerB* receptor by the overexpression of GerBC^Cys20 or GerBC^Ala20 suggested that the overexpressed C proteins may sequester some spore components that are limiting for germination in an inactive complex. Such limiting components might be the A and B proteins of the GerA or GerB receptors or other proteins altogether. To distinguish between these possibilities, we examined the effects of overexpressed GerAC on spore germination via the GerB* receptor and the effect of overexpressed GerBC on spore germination via the GerA receptor. The effects of overexpression of GerAC^Ala18 from P_sspB on spore germination via the GerB* receptor (strains PS3636 and PS3637) were small and likely not significant (Table 3), even though the GerB* receptor proteins are likely present in spores at ∼10-fold lower levels than the GerA receptor proteins (1, 2, 5). The overexpression of GerAC^Cys18 (strain PS3637) gave a ≤20% decrease in spore germination with the various nutrients to which the GerB* receptor responds, and there was only a slightly larger decrease upon the overexpression of GerAC^Ala18 (strain PS3636); while neither of these decreases is likely to be significant, the overexpression of GerBC^Cys20 decreased germination via the GerA receptor ∼11-fold (strain PS3660), although the overexpression of GerBC^Ala20 had essentially no effect on GerA receptor function (strain PS3661) (Table 2).

Alternative explanations for the results.

We had hoped to interpret the results of the experiments described above as due only to alterations in the levels of the various C proteins. However, there are a number of alternative explanations that should be considered, including the following. (i) Previous work has shown that the overexpression of the gerA operon abolishes sporulation (1), and even a small alteration in sporulation can alter the germination properties of the resultant spores (25). However, we saw no alteration in sporulation upon overexpression of the various C proteins, except for GerAC^Ala18, whose overexpression under the control of P_sspB (but not P_gerA) reduced sporulation efficiency ∼3-fold (data not shown). The reason for the reduced sporulation efficiency is not clear, but the effect was potentially worrisome. However, it seems unlikely that this is the reason for the effects of overexpression of GerAC^Ala18 on spore germination via the GerA receptor, since (a) GerAC^Ala18 expressed from P_gerA had similar, albeit smaller effects, and (b) GerAC^Ala18 overexpressed from P_sspB had only a slight effect on spore germination via the GerB* receptor. (ii) Perhaps the C proteins regulate the expression of their own operon. If so, an alteration in C protein levels might alter the level of the complete nutrient receptor. However, there is no evidence for such autoregulation, and the different effects of the C proteins with and without diacylglycerylation upon GerA and GerB receptor function are difficult to fit into such a model. (iii) Perhaps some of the effects of overexpressed C proteins are due to competition for a limiting amount of Lgt, thus reducing diacylglycerylation of other C proteins and altering nutrient receptor function. However, previous work has shown that the overexpression of gerB from P_sspB has no effect on GerA receptor function (8), suggesting that there is sufficient Lgt to diacylglycerylate overexpressed C proteins. In addition, the overexpression of GerAC^Cys18 increased germination via the GerA receptor, while this germination pathway was inhibited by the overexpression of GerAC^Ala18. (iv) Some of the C proteins may be unstable when expressed alone and thus do not accumulate to high levels. This is a reasonable possibility, and as a consequence, the interpretation of experiments in which an overexpressed C protein has no effect may be ambiguous. However, this possibility should not affect the interpretation of experiments in which significant alterations in spore germination accompanied the overexpression of a C protein. (v) The effects of C protein overexpression on spore germination may be masked because the nutrient receptors are not rate limiting for spore germination. We used DPA release as a measure of spore germination, since this is an early event in germination that is independent of later events, such as hydrolysis of the spore cortex (21, 26). In addition, DPA release is the earliest spore germination-associated process that is triggered by nutrients binding to their receptors and is easily measured (26). However, until the precise function of the nutrient receptors is understood and the proteins these receptors interact with are known, this will remain an area of uncertainty.

Conclusions.

With the provisos noted above in mind, the results in this communication lead to a number of conclusions. First, since the expression of gerAC^Ala18 from P_gerA had a much greater effect when gerAC^Ala18 was in the gerA operon than when it was expressed monocistronically at amyE, GerAA, GerAB, and GerAC may assemble in a complex as these proteins are translated from a polycistronic mRNA. Presumably, GerAC made from mRNA transcribed at amyE is less able to compete for binding to GerAA-GerAB than is GerAC translated from polycistronic gerA mRNA. However, since wild-type GerAC translated from a monocistronic mRNA transcribed from P_gerA at amyE was able to suppress the effect of a gerAC^Ala18 mutation in the intact gerA operon, there must be some mixing of GerAC proteins made from various locations, although GerAC lacking diacylglycerol appears to be less able to compete for GerAA-GerAB than the diacylglycerylated protein. However, we assume that P_gerA is transcribed equally well at the gerA and amyE loci, and this may not be the case. Transcription of gerA is by RNA polymerase with the sporulation-specific σ factor σ^G, and under some conditions P_gerA can be utilized in vivo by Eσ^F (5, 29). In addition, the amyE locus enters the developing spore well before the gerA locus (30). Consequently, P_gerA at amyE may be more frequently transcribed, perhaps even to some degree by Eσ^F, than is P_gerA at gerA (29, 31). We also cannot be sure that gerAC mRNA made as a monocistronic transcript at amyE and as a part of the polycistronic transcript at gerA are translated equally well, since translational coupling may affect the translation of gerAC in the polycistronic mRNA and there are differences between the ribosome-binding regions and translation start codons of gerAC mRNA at amyE and gerA.

The second conclusion follows from the observation that the overexpression of GerAC^Cys18 significantly increased the rate of spore germination via the GerA receptor. This increase was also seen when the gerA operon was overexpressed (1), but it was surprising to obtain this result when only GerAC was overexpressed, since GerAA and GerAB are essential for l-alanine triggering of spore germination via the GerA receptor (13, 14, 15, 21). One possible conclusion is that the levels of these three proteins (normally quite low) (5, 7, 20) and their affinities for one another are such that there is significant dissociation of GerAC from GerAA-GerAB in wild-type spores. Consequently, the overexpression of GerAC may increase GerAA-GerAB-GerAC complex formation, thus increasing GerA-mediated germination. This conclusion is obviously consistent with GerAC physically interacting with GerAA and GerAB (see also below). Since the overexpression of GerAC^Ala18 or GerAC^Cys18 altered GerA receptor function, overexpressed GerAC might have interfered with GerB or GerK receptor function, but this was not the case.

The observations noted above for spores with overexpressed GerAC^Ala18 held true to some degree for overexpressed GerBC^Ala20, as germination via the GerB or GerB* receptor was decreased by high levels of GerBC^Ala20. However, the effects of overexpressed GerBC^Ala20 were small, suggesting that this protein competes poorly for GerBA-GerBB. High levels of GerBC^Cys20 also decreased germination, in particular via the GerB* receptor, in contrast to the increase in GerA receptor-mediated germination by high levels of GerAC^Cys18. This finding suggests that the interaction of GerBC with GerBA-GerBB is much stronger than is the comparable interaction in the GerA receptor. However, we have no good explanation for the inhibition of GerB* receptor function by the overexpression of GerBC^Cys20. Since the level of gerB expression is significantly lower than that of gerA (2, 5), the ratio of overexpressed GerBC to GerBA-GerBB will likely be much higher than the ratio of overexpressed GerAC to GerAA-GerAB; perhaps the extremely high relative level of GerBC^Cys20 allows formation of higher oligomers of GerBC with GerBA-GerBB that are inactive. Indeed, the expression of GerBC^Cys20 from the much weaker P_gerA did not reduce significantly the germination via the GerB or GerB* receptors (Table 3). In any case, the finding that the overexpression of GerBC^Cys20 in particular decreased spore germination due to the GerB* receptor is consistent with the physical interaction of GerBC with GerBA-GerBB.

A third conclusion is that in at least one case, the C protein from one GerA receptor homolog can interact with the A and B proteins of another receptor. While this was not the case for GerAC and GerBA-GerBB, high levels of GerBC^Cys20 significantly decreased GerA receptor function. The simplest explanation for this observation is that high levels of GerBC^Cys20 compete well with GerAC^Cys18 for binding to GerAA-GerAB, and the presence of GerBC^Cys20 in this complex generates an inactive receptor. This explanation is consistent with the interaction of GerAC with GerAA-GerAB being not extremely strong, as suggested above. Presumably, GerBC interaction with GerBA-GerBB is much stronger, which is why GerAC overexpression does not alter GerB receptor function. However, the interaction of GerBC with GerAA-GerAB may not be physiologically significant, as it was seen only with high levels of GerBC^Cys20. The possibility of interactions among different GerA receptor homologs is also worth noting. Recent work on the plasma membrane chemoreceptors that modulate prokaryotic chemotaxis has shown that formation of large multireceptor complexes is important in the function of these receptors (11, 27, 28). Perhaps spore nutrient receptors are another example of this phenomenon.

The fourth conclusion is that gerAC^Ala18, gerBC^Cys20, and perhaps gerBC^Ala20 can act as dominant negative mutants, as was seen in the marked reduction in the GerA-mediated germination of spores of the strain that overexpresses gerAC^Ala18 at amyE and in the significant decrease in GerB*-mediated germination of spores with high levels of GerBC^Cys20. The simplest interpretation of these results and those supporting the other conclusions is that GerAC physically interacts with GerAA-GerAB, while GerBC physically interacts with GerBA-GerBB. As noted above, there is evidence for direct interaction between GerAA and GerAB and between GerBA and GerBB (15, 18). It is possible that the three GerA and GerB proteins do not form a long-lived complex, but associate only transiently as they diffuse through the dormant spore's inner membrane. However, since lipid probes appear to be largely immobile in the dormant spore's inner membrane (3), it is difficult to imagine that proteins readily diffuse in this membrane, although this could be the case in developing forespores. Consequently, it appears likely that GerAA-GerAB-GerAC and GerBA-GerBB-GerBC form a long-lived complex in the spore's inner membrane. Given the unusual properties of this membrane (3), further understanding of the function of the nutrient receptors will require knowledge of their structure and knowledge of the proteins with which the receptors interact.