World
Journal
of Microbiology
& Biotechnology
12.7-l
1
Optimization of mosquitocidal toxin synthesis from Bacillus sphaericus using gene fusions H.K. Ahmed, W.J. Mitchell and F.G. Priest* /I-Galactosidase gene fusions have been used to monitor the progress of mosquito-larvicidal-toxin gene expression in Bacillus sphclericus strain 2362. &Galactosidase estimation in cells from late-growth-phase batch cultures was compared with larvicidal toxicity after incubation for 48 h. Conditions which promoted efficient sporulation, such as plentiful trace elements and relatively crude protein sources (soybean or cottonseed flours), enhanced reporter gene expression and provided high toxicity. However, acetate, which repressed sporulation, similarly repressed binary toxin yield. Gene fusions to the binary and x00-kDa toxin genes of B. sphaericus could be useful for the rapid screening of fermentation conditions for the local production of this larvicidal bacterium but, in view of the poor correlation with toxicity at high toxicity levels, such experiments should be confirmed with bioassays. Key words: Bacillus sphaericw, biological
control, fermentation,
Two mosquito-larvicidal toxins are synthesized by some strains of Bacillus sphaericw. The binary toxin accumulates as a small crystal in early-stationary-phase cells and is highly toxic following ingestion by C&x qtrinquefasci&s or Anopheles stephensi larvae, although it is less active against Aedes aegypfi larvae (Berry et al. 1993). The 100kDa toxin (mosquitocidal toxin; Mtx) is synthesized in vegetative cells (Myers & Yousten 1980; Ahmed et al. 1995) and is also toxic to Culex and, to a lesser extent, Aedes larvae (Thanabalu ef al. 1991; 1993). Since the crystal protein generally accumulates to a far greater extent than the loo-kDa toxin, strains which produce the former are referred to as high-toxicity strains and those which synthesize only the latter as low-toxicity strains. However, these designations are misleading because the lo@kDa toxin is extremely toxic and it is only the poor accumulation of the toxin in the bacterial cell which results in low toxicity (Thanabalu et al. 1993). Bacillus sphaericus is a valuable biological control agent in areas infested by Culex mosquitoes and the associated disease of filariasis (Regis et al. 1995). It has also been
The authors are with the Department of Biological Sciences, Heriot Watt University, Edinburgh EH14 4AS. UK; fax: 44 131 451 3009. ‘Corresponding author. @ 1996 Rapid Science
gene control.
useful, either alone (Kumar ef al. 1994) or in combination with B. thtlringiensis var. israelensis (B.t.i.), for control of Anopheles larvae and malaria transmission (Xu et al. 1992), where the complementary target range in association with B.t.i. and environmental persistence of the bacterium were particularly attractive. Although B. sphaericus is available commercially, for many applications it is grown locally on a relatively small scale (Xu et al. 1992). It would probably be used more extensively if simple fermentation conditions based on locally available ingredients could be easily formulated (Bhumiratana 1990). However, fermentation development is complicated and time consuming if toxin yield is estimated by bioassay and requires specialist equipment and expertise if the yield is determined immunochemically (Anandkumar ef al. 1991). Moreover, these procedures cannot distinguish easily between the two toxins which accumulate at different stages of the batch-culture growth cycle. We have therefore investigated, for fermentation development, the use of reporter gene fusions in which the promoters of the two toxin genes are fused independently to the structural gene for P-galactosidase (IacZ). In cells containing these plasmid constructs, /Sgalactosidase, which can be assayed very simply using the calorimetric substrate ortho-nitrophenyl$galactoside (ONPG), accumulates in
Publishers World Journal of Microbiology 6 Biotechnology, Vol 12. 1996
7
H.K. Ahmed, W.]. Mitchell and F.G. Priest place of the toxin proteins and can be used to monitor the ‘progress of a fermentation (Ahmed et al. 1995). In the present study, this is shown to be a useful approach to fermentation optimization for a high-toxicity strain of B. sphaericus.
Materials Bacteria,
and Methods
Plasmids
and Growth
Conditions
The lacZ translational fusions to the toxin gene promoters in the B. sphaericw 2362 host used have been described previously (Ahmed et al. 1995). In brief, the binary-toxin gene fusion comprised the first 100 nucleotides of the structural gene for the 5IkDa component of the binary toxin and the 500 bp upstream of the translational start-point of this gene, fused to the promoterless lacZ gene in the pUBIIO-based vector pBT142 (Alonso 1988). Similarly, the IOO-kDa-toxin gene fusion comprised the first 215 bp of the structural gene and 232 bp upstream of the translational start point, fused to the promoterless 1acZ gene in vector pBT143 (Alonso 1988). Bacillus sphaerictrs 2362 was grown routinely in NYSM (nutrient broth/yeast extract/salts medium) agar and broth (Myers & Yousten 1980), containing kanamycin (lOpg/ml) to select for the plasmids, at 37°C unless otherwise stated. Spizizen’s minimal medium (Anagnostopoulos 8s Spizizen 1961) was used with a tenth of the normal concentration of phosphate buffer and was routinely supplemented with a standard trace element solution (1 ml/l) containing (mg/l): MnC1,.4H,O, 200; FeCl,.6H,O, 270; ZnS0,7H,O, 287; and CaCl,, 110. Where additional trace elements were added to minimal medium, these were added as a solution (1 ml/l medium) containing (mg/l): 120; CuC1,.6H,O, 17; NiCl,.6H,O, 24; CoC1,.6H,O, Na,M0,.2H,O, 24; and H,BO,, 300. Carbon and nitrogen sources were added as indicated in the Results section. Cultures were grown routinely in loo-ml amounts in 500.ml flasks containing stainless-steel coils to improve aeration while shaking on an orbital incubator at 120 rev/mm. Sporulation rate was assessed by phase-contrast microscopy to within 10% for routine fermentation experiments but, for more accurate estimations, cultures were plated in duplicate onto NYSM agar before and after heating at 85°C for 10 min to kill vegetative cells.
Unless otherwise stated, fermentations were repeated at least twice and mean results are given. Isolafion of Sporulafion Mutants A I-ml sample of B. sphaericus 2362 grown in NYSM broth containing 2% (w/v) sodium acetate for 24 h was incubated at 85°C for 10 min, re-inoculated into 100 ml NYSM broth containing 2% (w/v) sodium acetate and incubated for 24 h. This cycle was repeated four times to enrich for mutants able to sporulate efficiently in the presence of the catabolite-repressing carbon source, acetate (Ahmed et al. 1993). A single colony from such an enrichment was used as the mutant strain. Enzyme Assay Cells were harvested from cultures by centrifugation and washed and resuspended in 100 rnr.4 potassium phosphate buffer, pH 7.4. Toluene (I% v/v) was added and the-cell suspension mixed on a vortex mixer for 30 s. The cells were incubated at 37°C for 30 min to evaporate the toluene and j?-galactosidase was assayed using ONPG as substrate, as described earlier (Ahmed
et al. 1995).
Bioassay All bioassays were performed in duplicate. Range-finding assays were based on lo-fold dilutions (lo-’ to 10F6) of washed cell suspensions in distilled water, prepared from NYSM broth cultures incubated at 37°C for 48 h. Cells were added to two sets of beakers, each beaker containing 10, second-instar larvae of Anopheles sfephensi. Surviving larvae were counted after 48 h. Having determined the approximate toxicity of the cultures, more precise assays were conducted using duplicate preparations at four serial dilutions
within
the IO-fold
range (e.g. at lo-“, 8 x 10e4, 6 x and IOe4), again with two sets of after 48 h. Median lethal concentrations
10-4, 4 x 10-4, 2 x 10-4 10 larvae each, counted (LC,,) were calculated
from log/probit
plots of assays.
Results General Considerations When the lacZ-fusion-bearing strains of B. sphaericus are grown in NYSM broth, /?-galactosidase activity from the binary-toxin fusion peaks during the early stationary phase, after incubation for about 15 h, and that from the loo-kDa fusion peaks in the late exponential phase, after incubation for about 7 h (Ahmed et al. 1995). Thereafter, p-galactosidase activity declines, presumably due to proteolysis. We therefore routinely sampled cultures at these times for ,@-galactosidase activity, although it must be remembered that toxicity for the binary toxin will accumulate throughout sporulation and generally reach a plateau level after about 48 h. The phosphate content of the basal medium had little effect on P-galactosidase activity from either 1acZ fusion, or on sporulation (data not shown), and therefore the fairly high concentration of phosphate in Spizizen salts was reduced IO-fold in all subsequent experiments. Previous studies had indicated that the concentration of trace elements in the medium influenced toxin-gene expression (Ahmed et al. 1995). This was explored further by growing cells in media supplemented with twice the normal concentration of trace elements and with a supplementary range of trace elements. Doubling the concentration of trace elements produced a slight increase in P-galactosidase activity from the binary-toxin gene promoter and raised the sporulation rate and expression from the loo-kDa toxin gene promoter a little (Table I). In future experiments, which concentrated on the major, binary-toxin gene, the increased concentration of normal trace elements was used but the wider range of trace elements was not used since there was little improvement in the sporulation rate or level of gene expression in the presence of these salts (Table 1). Effects of Nitrogen Sources in the Growth Medium and Growth Conditions on Expression from the Toxin-gene Promoters Since B. sphaericus is unable to use carbohydrates as carbon and energy sources (Russell ef al. 1989), medium optimization was restricted to variations in proteinaceous materials and the effects of acetate, a recognized carbon source for
Toxin synthesis in B. sphaericus I Table 1. Effect of trace elements in the growth expression of /acZ fusions to toxin-gene promoters sphaericus. Sporuiation frequency (%I
Medium’
Control 2 x TE Additional
/&gaiactosidase
(Miller
lacz-binarytoxin-gene*
units)T
/acZ-1 OO-kDatoxin-genes
76 101 90
25 30 20
Salts
medium on in Bacillus
4.3 5.2 6.5
* Control-Spizizen salts supplemented with 1% peptone and 0.2% yeast extract; 2 x TE-as control but with twice the concentration of the routine trace elements; Additional saltsas control but with the additional trace element solution described in Materials and Methods. T Miller units are described by Ahmed et al. (1995). $ Estimated in early stationary phase (15 h after inoculation). g Estimated in late growth phase (7 h after inoculation).
Table 2. Effect /ILgalactosidase strain. Nitrogen (% w/v)’
of nftrogen sources in the growth activity in the binary-toxin-gene/acZ
source
PH
YE,0.25;P,0.75 YE,0.5;P.0.5 YE,l.O;P,l.O YE,1.5;P,1.5 YE,0.5;SF,0.5 YE,0.5;CF,0.5
a.3 a.7 a.7 a.5 a.2 a.4
Sporuiation frequency (%I
medium
on fusion
B-gaiactosidaset (Miller units)
20 25 25 5 50 50
a5 316 38% 54* 308* 354
* Supplements added to Spizizen salts were yeast extract (YE), peptone (P), cottonseed flour (CF) and/or soybean flour (SF). T Estimated in early stationary phase (15 h after inoculation). Miller units are described by Ahmed et al. (1995). $ ResuItsofsingle-batchculturefermentations.AIlotherfermentations conducted at least twice and results given as means, with variation always within 20% of these values.
Table 3. Sporuiation mutant of Bacillus containing 2% (w/v)
and toxicity sphaericus acetate.
Strain
of a sporuiation control 2362 grown in NYSM
Sporulatlon frequency
B. sphaericus 6. sphaericus
2362 2362 Scol
(Sco) broth
G4 (ng dry wt/mi)
3 x 10-a 2.3 x 10m4
4.0 0.4
this bacterium (White & Lotay 1980). Peptone and yeast extract, each at OS%, had a marked effect on binary-toxin gene
expression,
presumably
resulting
in balanced
growth
and optimal conditions for expression from the binary-toxin gene promoter (Table 2). Raising the overall nitrogen level to 3% total peptone and yeast extract inhibited sporulation
Figure 1. Comparison (0) between B-galactosidase activity from the binary-toxin-gene-/acZ fusion, measured after incubation for 15 h, with the reciprocal of toxicity (measured as LC,,) to Anopheles stephensi larvae, determined with washed biomass which had been incubated for 46 h.
and reduced expression from the binary-toxin gene promoter. Replacement of the peptone (0.5%) with a less refined protein source, cottonseed flour or soyaflour, at the same concentration had little effect on binary-toxin gene expression, which was maintained at a high level. Effecfs of Acefafe in the Growth Medium on Expression from the Toxin-gene Promoters Acetate is known to be a rapidly catabolized carbon source for B. sphaericus and to cause catabolite repression of sporulation (Ahmed et al. 1993). Addition of 0.5% acetate to the cottonseed-flour-containing medium reduced the expression of the binary-toxin gene fusion lo-fold, from around 350 units to about 35 units, and halved the sporulation rate from 50% to 25%. It also reduced the already low level of expression from the 100-kDa toxin gene to an undetectable amount. To investigate further the effects of acetate on toxin synthesis and sporulation, a mutant (sporulation control; Sco mutant) was prepared that was partially derepressed for acetate inhibition of sporulation and produced spores at 10,000 times the rate of the wild type in the presence of 2% (w/v) acetate (Table 3). When cultures grown for 48 h in NYSM broth containing 2% (w/v) acetate were tested against larvae of A. stephensi, the Sco mutant was 10 times more toxic than the parent strain. Comparison between 1acZ Expression from the Binary-toxin Gene Fusion and Toxicity to Larvae Several cultures were prepared on various combinations of peptone, yeast extract and cottonseed flour designed to provide different amounts of binary-toxin accumulation. Cultures were prepared in duplicate and inoculated with B. sphaericus 2362 and with the same organism containing the
World Journal of Microbiology 6 Biotechnology. Vol 12, 1996
9
H.K. Ahmed, W.J Mitchell and F.G. Priest binary-toxin gene fusion. The former cultures were incubated for 48 h and examined for toxicity against secondinstar larvae of A. stephensi. The latter cultures were sampled after 15 h and /%galactosidase activity was determined. When the reciprocal of toxicity was plotted against /Igalactosidase activity (Figure 1), enzyme activity provided an under-estimate of the toxicity in very toxic cultures (grown in the cottonseed-based medium), but at lower toxicities, P-galactosidase provided a reasonable indication of toxicity.
Discussion When interpreting these studies it is important to remember that estimation of 1acZ expression early in the stationary phase is being used for the prediction of binary-toxin protein accumulation after incubation for 48 h. In these experiments, P-galactosidase yield will be a compromise of synthesis and degradation of the foreign protein, but binary toxin, being an indigenous protein, will presumably accumulate with minimal degradation. Because of the different degradation rates, it might be expected that the lacZbinary-toxin fusion might give an under-estimate of toxin protein accumulation. This is clearly seen in the data in Figure I, which indicate that degradation of /I-galactosidase is more pronounced at the higher levels of expression. Nevertheless, the gene fusion provides a simple and rapid way to assess the potential of various medium combinations to support binary-toxin accumulation. Selected fermentation formulations should always be checked subsequently by bioassay. The regulation of synthesis of the two toxins is very different. The binary-toxin gene promoter provides robust gene expression under a variety of conditions. The 100kDa-toxin gene promoter, on the other hand, is weak and performs poorly under all conditions tested here or reported earlier (Ahmed et al. 1995). However, when placed in nontoxic cells of either B. subfilis or E scherichia coli, the IOO-kDa-toxin gene is expressed much more strongly, which indicates that there is potential for improving the accumulation of this toxin, probably by genetic manipulation rather than by optimizing the fermentation conditions. The results presented here add to our meagre knowledge of binary-toxin gene expression in B. sphaericus. There is a pronounced correlation with sporulation; conditions which enhance sporulation, such as modifying the combination of trace elements in the medium, enhance toxicity and factors which reduce sporulation, for example the addition of acetate, dramatically reduce it. This correlation was further supported by the demonstration of high toxicity in a sporulation control mutant of B. sphaericus (Table 3). The relationship between sporulation and binary-toxin accumulation has been noted before (Yousten & Davidson 1982;
10
World Jouml
of Microbiology
& Biotechnology, Vol 12, 1996
Charles et al. 1988; Andreev et al. 1994), although this is the first time a direct effect on gene expression has been demonstrated. This invites the question whether binarytoxin synthesis is dependent on sporulation or simply activated by the conditions which activate the sporulation pathway. The lack of toxin accumulation in SpoO mutants of B. sphaericw 2297 (Charles ef al. 1988) and the poor expression of the binary-toxin gene hcZ fusion in a spoOA mutant of B. subtilis (Ahmed et al. 1995) indicate a dependence on the prior expression of at least some stage 0 genes. Given that B. sphaericus does not metabolize carbohydrates (Russell et al. 1989), the scope for medium optimization for toxicity is limited. The results in Table 2 show that the nature of the nitrogen source is important and that crude preparations, such as soybean and cottonseed flours, are at least as effective at promoting sporulation and toxin-gene expression as more purified forms, such as peptone. This ease of propagation has led to effective local production on various nitrogen sources such as bean meals (Ejiofor & Okafor 1988), animal products including blood and fishmeal (reviewed in Bhumiratana 1990) and the byproducts from a monosodium-glutamate factory (Dharmsthiti et al. 1985). The balance of nitrogen sources is, however, extremely important, a factor which has previously been observed with serine protease synthesis in this bacterium (Dumusois & Priest 1993). High nitrogen levels repress sporulation and toxin-gene expression dramatically (Table 2) and similarly repress protease synthesis. It would seem, therefore, that the B. sphaericus fermentation is reasonably straightforward and highly toxic biomass could be readily mass-produced using local, inexpensive materials. The lacZ-fusion strains described here are available to any laboratory wishing to evaluate the B. sphaericus fermentation using their own materials.
Acknowledgements We are very grateful to D. Walliker, University burgh, for provision of mosquito larvae.
of Edin-
References H.K., Mitchell, W.J. & Priest, F.G. 1993 Catabolite repression of histidase biosynthesis in Bucillrrs sphaericw by acetate. FEMS Microbiology Letters 106, 71-76.
Ahmed,
Ahmed, H.K., mosquitocidal
Mitchell, W.J. & Priest, F.G. 1995 Regulation of toxin synthesis in Baci\ltcs sphaericus. Applied Microbiology and Biotechnology 43, 310-314. AJonso, J.C. 1988 New plasmid vectors for the construction of translational gene fusions in B&/w subtilis. Gene 65, 325-328. Anagnostopoulos, C. 81 Spizizen, J. 1961 Requirements for transformation in B&w subfilis. Journal of Bacteriology 81, 741-746. Anandkumar, K., Kuppusamy, M. & Balaraman, K. 1991 An enzyme linked immunosorbent assay for monitoring BaciNw sphaericus toxin. Zndiun ]oumal of Experimental Biology 29, 953-957.
Toxin synlhesis in B. sphaericus Andreev, J,, Dibrov, P.A., Klein, D. & Braun, S. 1994 Chemotaxis, sporulation and larvicide production in Bacillus sphuericlrs 2362. FEBS Letters 347, 231-234. Berry, C., Hindley, J,, Erhardt, A.F., Grounds, T., De Souza, I. & Davidson, E.W. 1993 Genetic determinants of the host range of the Bacillus sphuericus mosquitocidal toxins. ]o~rnul of Bucteriol-
ogy175,510-518. Bhumiratana, A. 1990 Local production of Bacterial Control of Mosquitoes und Blackflies, Sutherland, D. pp 256-271. New Jersey: Press. Charles, J.F., Kalfon, A., Bourgouin, C. & Bucilllrs sphuericw asporogenous mutants: pattern and larvicidal activity. Annules de
Bacillus sphuericw. In eds De Barjac, H. & Rutgers University De Barjac, H. 1988 morphology, protein I’lnstihut Pasteur, Paris
139,243-259. Dharmsthiti, S.C., Pantuwatana, S. & Bhurniratana, A. 198.5 Production of Bacillus fhuringiensis and Buciks sphuericlts 1.593 on media using a byproduct from a monosodium glutamate factory. ]ournul of Invertebrate Pathology 46, 231-238. Dumusois, C. & Priest, F.G. 1993 Extracellular serine protease synthesis by mosquito-pathogenic strains of Bacillus sphuericus. ]ournul of Applied Bacteriology 75, 416-419. Ejiofor, A.O. & Okafor, N. 1988 The production of Bacillus sphuericus using fermented cowpea (Vichua unguiculufu) medium containing minerals substitutes from Nigeria. MIRCEN Journal @Applied Microbiology and Biotechnology 4, 455-462. Kumar, A., Sharma, V.P., Somodan, P.K., Thavaselvam, D. & Kamat, R.H. 1994 Malaria control utilizing Bacillus sphuericus against Anopheles stephensi in Panji, Goa. Journal of the American Mosquito Control Association 10, 534-539. Myers, P. & Yousten, A.A. 1980 Localization of a mosquito-larval
toxin of Bacillus sphaericm 1593. Applied and Environmental Microbiology 39, 1205-1211. Regis, L., Silva-Filha, M.-H.N.L., De Oliveira, C.M.F., Rios, E.M., Da Silva, S.B. & Furtado, A.F. 1995 Integrated vector control measures against C&r quinquefasciahrs, the vector of filariasis in Recife. Mt+norias do lnstihrfo Oswaldo Cruz 90, 115-120. Russell, B.L., Jelley, S.A. & Yousten, A.A. 1989 Carbohydrate metabolism in the mosquito pathogen Bucilltrs sphuericw. Applied and Environmental Microbiology 55, 294-297. Thanabalu, T., Hindley, J. & Berry, C. 1993 Cytotoxicity and ADP-ribosylating activity of the mosquitocidal toxin from Bacillus sphuericus SSII-1: possible roles of the 27- and 70. kilodalton peptides. ]ourna/ of Bucferiology 175, 2314-2320. Thanabalu, T., Hindley, J., Jackson-Jap, J. & Berry, C. 1991 Cloning, sequencing, and expression of a gene encoding a IOOkilodalton mosquitocidal toxin from Bucillus sphuerictrs. lottrnul of Bacteriology 173, 27762785. White, P.J. & Lotay, H. 1980 Minimal nutritional requirements of Bucillrrs sphuerictts NCTC 9602 and 26 other strains of this species: the majority grow and sporulate with acetate as sole major source of carbon. lotrrnul of General Microbiology 118,
13-19. Xu, B.Z., Becker, N., Xiao, X.Q. & Ludwig, H.W. 1992 Microbial control of mosquitoes in Hubei Province, People’s Republic of China. Bulletin of the Society for Vector Ecology 17, l-10. Yousten, A.A. & Davidson, E.W. 1982 Ultrastructural analysis of spores and parasporal crystals formed in Bacillus sphuericus. Applied and Environmental Microbiology 44, 1449-1455.
(Received in revised form
20 July 1995; accepled 23 July 1995)
World Journal of Microbiology 6 Biotechnology, Vol 1.2,1996
11
Journal
of Microbiology
& Biotechnology
12.7-l
1
Optimization of mosquitocidal toxin synthesis from Bacillus sphaericus using gene fusions H.K. Ahmed, W.J. Mitchell and F.G. Priest* /I-Galactosidase gene fusions have been used to monitor the progress of mosquito-larvicidal-toxin gene expression in Bacillus sphclericus strain 2362. &Galactosidase estimation in cells from late-growth-phase batch cultures was compared with larvicidal toxicity after incubation for 48 h. Conditions which promoted efficient sporulation, such as plentiful trace elements and relatively crude protein sources (soybean or cottonseed flours), enhanced reporter gene expression and provided high toxicity. However, acetate, which repressed sporulation, similarly repressed binary toxin yield. Gene fusions to the binary and x00-kDa toxin genes of B. sphaericus could be useful for the rapid screening of fermentation conditions for the local production of this larvicidal bacterium but, in view of the poor correlation with toxicity at high toxicity levels, such experiments should be confirmed with bioassays. Key words: Bacillus sphaericw, biological
control, fermentation,
Two mosquito-larvicidal toxins are synthesized by some strains of Bacillus sphaericw. The binary toxin accumulates as a small crystal in early-stationary-phase cells and is highly toxic following ingestion by C&x qtrinquefasci&s or Anopheles stephensi larvae, although it is less active against Aedes aegypfi larvae (Berry et al. 1993). The 100kDa toxin (mosquitocidal toxin; Mtx) is synthesized in vegetative cells (Myers & Yousten 1980; Ahmed et al. 1995) and is also toxic to Culex and, to a lesser extent, Aedes larvae (Thanabalu ef al. 1991; 1993). Since the crystal protein generally accumulates to a far greater extent than the loo-kDa toxin, strains which produce the former are referred to as high-toxicity strains and those which synthesize only the latter as low-toxicity strains. However, these designations are misleading because the lo@kDa toxin is extremely toxic and it is only the poor accumulation of the toxin in the bacterial cell which results in low toxicity (Thanabalu et al. 1993). Bacillus sphaericus is a valuable biological control agent in areas infested by Culex mosquitoes and the associated disease of filariasis (Regis et al. 1995). It has also been
The authors are with the Department of Biological Sciences, Heriot Watt University, Edinburgh EH14 4AS. UK; fax: 44 131 451 3009. ‘Corresponding author. @ 1996 Rapid Science
gene control.
useful, either alone (Kumar ef al. 1994) or in combination with B. thtlringiensis var. israelensis (B.t.i.), for control of Anopheles larvae and malaria transmission (Xu et al. 1992), where the complementary target range in association with B.t.i. and environmental persistence of the bacterium were particularly attractive. Although B. sphaericus is available commercially, for many applications it is grown locally on a relatively small scale (Xu et al. 1992). It would probably be used more extensively if simple fermentation conditions based on locally available ingredients could be easily formulated (Bhumiratana 1990). However, fermentation development is complicated and time consuming if toxin yield is estimated by bioassay and requires specialist equipment and expertise if the yield is determined immunochemically (Anandkumar ef al. 1991). Moreover, these procedures cannot distinguish easily between the two toxins which accumulate at different stages of the batch-culture growth cycle. We have therefore investigated, for fermentation development, the use of reporter gene fusions in which the promoters of the two toxin genes are fused independently to the structural gene for P-galactosidase (IacZ). In cells containing these plasmid constructs, /Sgalactosidase, which can be assayed very simply using the calorimetric substrate ortho-nitrophenyl$galactoside (ONPG), accumulates in
Publishers World Journal of Microbiology 6 Biotechnology, Vol 12. 1996
7
H.K. Ahmed, W.]. Mitchell and F.G. Priest place of the toxin proteins and can be used to monitor the ‘progress of a fermentation (Ahmed et al. 1995). In the present study, this is shown to be a useful approach to fermentation optimization for a high-toxicity strain of B. sphaericus.
Materials Bacteria,
and Methods
Plasmids
and Growth
Conditions
The lacZ translational fusions to the toxin gene promoters in the B. sphaericw 2362 host used have been described previously (Ahmed et al. 1995). In brief, the binary-toxin gene fusion comprised the first 100 nucleotides of the structural gene for the 5IkDa component of the binary toxin and the 500 bp upstream of the translational start-point of this gene, fused to the promoterless lacZ gene in the pUBIIO-based vector pBT142 (Alonso 1988). Similarly, the IOO-kDa-toxin gene fusion comprised the first 215 bp of the structural gene and 232 bp upstream of the translational start point, fused to the promoterless 1acZ gene in vector pBT143 (Alonso 1988). Bacillus sphaerictrs 2362 was grown routinely in NYSM (nutrient broth/yeast extract/salts medium) agar and broth (Myers & Yousten 1980), containing kanamycin (lOpg/ml) to select for the plasmids, at 37°C unless otherwise stated. Spizizen’s minimal medium (Anagnostopoulos 8s Spizizen 1961) was used with a tenth of the normal concentration of phosphate buffer and was routinely supplemented with a standard trace element solution (1 ml/l) containing (mg/l): MnC1,.4H,O, 200; FeCl,.6H,O, 270; ZnS0,7H,O, 287; and CaCl,, 110. Where additional trace elements were added to minimal medium, these were added as a solution (1 ml/l medium) containing (mg/l): 120; CuC1,.6H,O, 17; NiCl,.6H,O, 24; CoC1,.6H,O, Na,M0,.2H,O, 24; and H,BO,, 300. Carbon and nitrogen sources were added as indicated in the Results section. Cultures were grown routinely in loo-ml amounts in 500.ml flasks containing stainless-steel coils to improve aeration while shaking on an orbital incubator at 120 rev/mm. Sporulation rate was assessed by phase-contrast microscopy to within 10% for routine fermentation experiments but, for more accurate estimations, cultures were plated in duplicate onto NYSM agar before and after heating at 85°C for 10 min to kill vegetative cells.
Unless otherwise stated, fermentations were repeated at least twice and mean results are given. Isolafion of Sporulafion Mutants A I-ml sample of B. sphaericus 2362 grown in NYSM broth containing 2% (w/v) sodium acetate for 24 h was incubated at 85°C for 10 min, re-inoculated into 100 ml NYSM broth containing 2% (w/v) sodium acetate and incubated for 24 h. This cycle was repeated four times to enrich for mutants able to sporulate efficiently in the presence of the catabolite-repressing carbon source, acetate (Ahmed et al. 1993). A single colony from such an enrichment was used as the mutant strain. Enzyme Assay Cells were harvested from cultures by centrifugation and washed and resuspended in 100 rnr.4 potassium phosphate buffer, pH 7.4. Toluene (I% v/v) was added and the-cell suspension mixed on a vortex mixer for 30 s. The cells were incubated at 37°C for 30 min to evaporate the toluene and j?-galactosidase was assayed using ONPG as substrate, as described earlier (Ahmed
et al. 1995).
Bioassay All bioassays were performed in duplicate. Range-finding assays were based on lo-fold dilutions (lo-’ to 10F6) of washed cell suspensions in distilled water, prepared from NYSM broth cultures incubated at 37°C for 48 h. Cells were added to two sets of beakers, each beaker containing 10, second-instar larvae of Anopheles sfephensi. Surviving larvae were counted after 48 h. Having determined the approximate toxicity of the cultures, more precise assays were conducted using duplicate preparations at four serial dilutions
within
the IO-fold
range (e.g. at lo-“, 8 x 10e4, 6 x and IOe4), again with two sets of after 48 h. Median lethal concentrations
10-4, 4 x 10-4, 2 x 10-4 10 larvae each, counted (LC,,) were calculated
from log/probit
plots of assays.
Results General Considerations When the lacZ-fusion-bearing strains of B. sphaericus are grown in NYSM broth, /?-galactosidase activity from the binary-toxin fusion peaks during the early stationary phase, after incubation for about 15 h, and that from the loo-kDa fusion peaks in the late exponential phase, after incubation for about 7 h (Ahmed et al. 1995). Thereafter, p-galactosidase activity declines, presumably due to proteolysis. We therefore routinely sampled cultures at these times for ,@-galactosidase activity, although it must be remembered that toxicity for the binary toxin will accumulate throughout sporulation and generally reach a plateau level after about 48 h. The phosphate content of the basal medium had little effect on P-galactosidase activity from either 1acZ fusion, or on sporulation (data not shown), and therefore the fairly high concentration of phosphate in Spizizen salts was reduced IO-fold in all subsequent experiments. Previous studies had indicated that the concentration of trace elements in the medium influenced toxin-gene expression (Ahmed et al. 1995). This was explored further by growing cells in media supplemented with twice the normal concentration of trace elements and with a supplementary range of trace elements. Doubling the concentration of trace elements produced a slight increase in P-galactosidase activity from the binary-toxin gene promoter and raised the sporulation rate and expression from the loo-kDa toxin gene promoter a little (Table I). In future experiments, which concentrated on the major, binary-toxin gene, the increased concentration of normal trace elements was used but the wider range of trace elements was not used since there was little improvement in the sporulation rate or level of gene expression in the presence of these salts (Table 1). Effects of Nitrogen Sources in the Growth Medium and Growth Conditions on Expression from the Toxin-gene Promoters Since B. sphaericus is unable to use carbohydrates as carbon and energy sources (Russell ef al. 1989), medium optimization was restricted to variations in proteinaceous materials and the effects of acetate, a recognized carbon source for
Toxin synthesis in B. sphaericus I Table 1. Effect of trace elements in the growth expression of /acZ fusions to toxin-gene promoters sphaericus. Sporuiation frequency (%I
Medium’
Control 2 x TE Additional
/&gaiactosidase
(Miller
lacz-binarytoxin-gene*
units)T
/acZ-1 OO-kDatoxin-genes
76 101 90
25 30 20
Salts
medium on in Bacillus
4.3 5.2 6.5
* Control-Spizizen salts supplemented with 1% peptone and 0.2% yeast extract; 2 x TE-as control but with twice the concentration of the routine trace elements; Additional saltsas control but with the additional trace element solution described in Materials and Methods. T Miller units are described by Ahmed et al. (1995). $ Estimated in early stationary phase (15 h after inoculation). g Estimated in late growth phase (7 h after inoculation).
Table 2. Effect /ILgalactosidase strain. Nitrogen (% w/v)’
of nftrogen sources in the growth activity in the binary-toxin-gene/acZ
source
PH
YE,0.25;P,0.75 YE,0.5;P.0.5 YE,l.O;P,l.O YE,1.5;P,1.5 YE,0.5;SF,0.5 YE,0.5;CF,0.5
a.3 a.7 a.7 a.5 a.2 a.4
Sporuiation frequency (%I
medium
on fusion
B-gaiactosidaset (Miller units)
20 25 25 5 50 50
a5 316 38% 54* 308* 354
* Supplements added to Spizizen salts were yeast extract (YE), peptone (P), cottonseed flour (CF) and/or soybean flour (SF). T Estimated in early stationary phase (15 h after inoculation). Miller units are described by Ahmed et al. (1995). $ ResuItsofsingle-batchculturefermentations.AIlotherfermentations conducted at least twice and results given as means, with variation always within 20% of these values.
Table 3. Sporuiation mutant of Bacillus containing 2% (w/v)
and toxicity sphaericus acetate.
Strain
of a sporuiation control 2362 grown in NYSM
Sporulatlon frequency
B. sphaericus 6. sphaericus
2362 2362 Scol
(Sco) broth
G4 (ng dry wt/mi)
3 x 10-a 2.3 x 10m4
4.0 0.4
this bacterium (White & Lotay 1980). Peptone and yeast extract, each at OS%, had a marked effect on binary-toxin gene
expression,
presumably
resulting
in balanced
growth
and optimal conditions for expression from the binary-toxin gene promoter (Table 2). Raising the overall nitrogen level to 3% total peptone and yeast extract inhibited sporulation
Figure 1. Comparison (0) between B-galactosidase activity from the binary-toxin-gene-/acZ fusion, measured after incubation for 15 h, with the reciprocal of toxicity (measured as LC,,) to Anopheles stephensi larvae, determined with washed biomass which had been incubated for 46 h.
and reduced expression from the binary-toxin gene promoter. Replacement of the peptone (0.5%) with a less refined protein source, cottonseed flour or soyaflour, at the same concentration had little effect on binary-toxin gene expression, which was maintained at a high level. Effecfs of Acefafe in the Growth Medium on Expression from the Toxin-gene Promoters Acetate is known to be a rapidly catabolized carbon source for B. sphaericus and to cause catabolite repression of sporulation (Ahmed et al. 1993). Addition of 0.5% acetate to the cottonseed-flour-containing medium reduced the expression of the binary-toxin gene fusion lo-fold, from around 350 units to about 35 units, and halved the sporulation rate from 50% to 25%. It also reduced the already low level of expression from the 100-kDa toxin gene to an undetectable amount. To investigate further the effects of acetate on toxin synthesis and sporulation, a mutant (sporulation control; Sco mutant) was prepared that was partially derepressed for acetate inhibition of sporulation and produced spores at 10,000 times the rate of the wild type in the presence of 2% (w/v) acetate (Table 3). When cultures grown for 48 h in NYSM broth containing 2% (w/v) acetate were tested against larvae of A. stephensi, the Sco mutant was 10 times more toxic than the parent strain. Comparison between 1acZ Expression from the Binary-toxin Gene Fusion and Toxicity to Larvae Several cultures were prepared on various combinations of peptone, yeast extract and cottonseed flour designed to provide different amounts of binary-toxin accumulation. Cultures were prepared in duplicate and inoculated with B. sphaericus 2362 and with the same organism containing the
World Journal of Microbiology 6 Biotechnology. Vol 12, 1996
9
H.K. Ahmed, W.J Mitchell and F.G. Priest binary-toxin gene fusion. The former cultures were incubated for 48 h and examined for toxicity against secondinstar larvae of A. stephensi. The latter cultures were sampled after 15 h and /%galactosidase activity was determined. When the reciprocal of toxicity was plotted against /Igalactosidase activity (Figure 1), enzyme activity provided an under-estimate of the toxicity in very toxic cultures (grown in the cottonseed-based medium), but at lower toxicities, P-galactosidase provided a reasonable indication of toxicity.
Discussion When interpreting these studies it is important to remember that estimation of 1acZ expression early in the stationary phase is being used for the prediction of binary-toxin protein accumulation after incubation for 48 h. In these experiments, P-galactosidase yield will be a compromise of synthesis and degradation of the foreign protein, but binary toxin, being an indigenous protein, will presumably accumulate with minimal degradation. Because of the different degradation rates, it might be expected that the lacZbinary-toxin fusion might give an under-estimate of toxin protein accumulation. This is clearly seen in the data in Figure I, which indicate that degradation of /I-galactosidase is more pronounced at the higher levels of expression. Nevertheless, the gene fusion provides a simple and rapid way to assess the potential of various medium combinations to support binary-toxin accumulation. Selected fermentation formulations should always be checked subsequently by bioassay. The regulation of synthesis of the two toxins is very different. The binary-toxin gene promoter provides robust gene expression under a variety of conditions. The 100kDa-toxin gene promoter, on the other hand, is weak and performs poorly under all conditions tested here or reported earlier (Ahmed et al. 1995). However, when placed in nontoxic cells of either B. subfilis or E scherichia coli, the IOO-kDa-toxin gene is expressed much more strongly, which indicates that there is potential for improving the accumulation of this toxin, probably by genetic manipulation rather than by optimizing the fermentation conditions. The results presented here add to our meagre knowledge of binary-toxin gene expression in B. sphaericus. There is a pronounced correlation with sporulation; conditions which enhance sporulation, such as modifying the combination of trace elements in the medium, enhance toxicity and factors which reduce sporulation, for example the addition of acetate, dramatically reduce it. This correlation was further supported by the demonstration of high toxicity in a sporulation control mutant of B. sphaericus (Table 3). The relationship between sporulation and binary-toxin accumulation has been noted before (Yousten & Davidson 1982;
10
World Jouml
of Microbiology
& Biotechnology, Vol 12, 1996
Charles et al. 1988; Andreev et al. 1994), although this is the first time a direct effect on gene expression has been demonstrated. This invites the question whether binarytoxin synthesis is dependent on sporulation or simply activated by the conditions which activate the sporulation pathway. The lack of toxin accumulation in SpoO mutants of B. sphaericw 2297 (Charles ef al. 1988) and the poor expression of the binary-toxin gene hcZ fusion in a spoOA mutant of B. subtilis (Ahmed et al. 1995) indicate a dependence on the prior expression of at least some stage 0 genes. Given that B. sphaericus does not metabolize carbohydrates (Russell et al. 1989), the scope for medium optimization for toxicity is limited. The results in Table 2 show that the nature of the nitrogen source is important and that crude preparations, such as soybean and cottonseed flours, are at least as effective at promoting sporulation and toxin-gene expression as more purified forms, such as peptone. This ease of propagation has led to effective local production on various nitrogen sources such as bean meals (Ejiofor & Okafor 1988), animal products including blood and fishmeal (reviewed in Bhumiratana 1990) and the byproducts from a monosodium-glutamate factory (Dharmsthiti et al. 1985). The balance of nitrogen sources is, however, extremely important, a factor which has previously been observed with serine protease synthesis in this bacterium (Dumusois & Priest 1993). High nitrogen levels repress sporulation and toxin-gene expression dramatically (Table 2) and similarly repress protease synthesis. It would seem, therefore, that the B. sphaericus fermentation is reasonably straightforward and highly toxic biomass could be readily mass-produced using local, inexpensive materials. The lacZ-fusion strains described here are available to any laboratory wishing to evaluate the B. sphaericus fermentation using their own materials.
Acknowledgements We are very grateful to D. Walliker, University burgh, for provision of mosquito larvae.
of Edin-
References H.K., Mitchell, W.J. & Priest, F.G. 1993 Catabolite repression of histidase biosynthesis in Bucillrrs sphaericw by acetate. FEMS Microbiology Letters 106, 71-76.
Ahmed,
Ahmed, H.K., mosquitocidal
Mitchell, W.J. & Priest, F.G. 1995 Regulation of toxin synthesis in Baci\ltcs sphaericus. Applied Microbiology and Biotechnology 43, 310-314. AJonso, J.C. 1988 New plasmid vectors for the construction of translational gene fusions in B&/w subtilis. Gene 65, 325-328. Anagnostopoulos, C. 81 Spizizen, J. 1961 Requirements for transformation in B&w subfilis. Journal of Bacteriology 81, 741-746. Anandkumar, K., Kuppusamy, M. & Balaraman, K. 1991 An enzyme linked immunosorbent assay for monitoring BaciNw sphaericus toxin. Zndiun ]oumal of Experimental Biology 29, 953-957.
Toxin synlhesis in B. sphaericus Andreev, J,, Dibrov, P.A., Klein, D. & Braun, S. 1994 Chemotaxis, sporulation and larvicide production in Bacillus sphuericlrs 2362. FEBS Letters 347, 231-234. Berry, C., Hindley, J,, Erhardt, A.F., Grounds, T., De Souza, I. & Davidson, E.W. 1993 Genetic determinants of the host range of the Bacillus sphuericus mosquitocidal toxins. ]o~rnul of Bucteriol-
ogy175,510-518. Bhumiratana, A. 1990 Local production of Bacterial Control of Mosquitoes und Blackflies, Sutherland, D. pp 256-271. New Jersey: Press. Charles, J.F., Kalfon, A., Bourgouin, C. & Bucilllrs sphuericw asporogenous mutants: pattern and larvicidal activity. Annules de
Bacillus sphuericw. In eds De Barjac, H. & Rutgers University De Barjac, H. 1988 morphology, protein I’lnstihut Pasteur, Paris
139,243-259. Dharmsthiti, S.C., Pantuwatana, S. & Bhurniratana, A. 198.5 Production of Bacillus fhuringiensis and Buciks sphuericlts 1.593 on media using a byproduct from a monosodium glutamate factory. ]ournul of Invertebrate Pathology 46, 231-238. Dumusois, C. & Priest, F.G. 1993 Extracellular serine protease synthesis by mosquito-pathogenic strains of Bacillus sphuericus. ]ournul of Applied Bacteriology 75, 416-419. Ejiofor, A.O. & Okafor, N. 1988 The production of Bacillus sphuericus using fermented cowpea (Vichua unguiculufu) medium containing minerals substitutes from Nigeria. MIRCEN Journal @Applied Microbiology and Biotechnology 4, 455-462. Kumar, A., Sharma, V.P., Somodan, P.K., Thavaselvam, D. & Kamat, R.H. 1994 Malaria control utilizing Bacillus sphuericus against Anopheles stephensi in Panji, Goa. Journal of the American Mosquito Control Association 10, 534-539. Myers, P. & Yousten, A.A. 1980 Localization of a mosquito-larval
toxin of Bacillus sphaericm 1593. Applied and Environmental Microbiology 39, 1205-1211. Regis, L., Silva-Filha, M.-H.N.L., De Oliveira, C.M.F., Rios, E.M., Da Silva, S.B. & Furtado, A.F. 1995 Integrated vector control measures against C&r quinquefasciahrs, the vector of filariasis in Recife. Mt+norias do lnstihrfo Oswaldo Cruz 90, 115-120. Russell, B.L., Jelley, S.A. & Yousten, A.A. 1989 Carbohydrate metabolism in the mosquito pathogen Bucilltrs sphuericw. Applied and Environmental Microbiology 55, 294-297. Thanabalu, T., Hindley, J. & Berry, C. 1993 Cytotoxicity and ADP-ribosylating activity of the mosquitocidal toxin from Bacillus sphuericus SSII-1: possible roles of the 27- and 70. kilodalton peptides. ]ourna/ of Bucferiology 175, 2314-2320. Thanabalu, T., Hindley, J., Jackson-Jap, J. & Berry, C. 1991 Cloning, sequencing, and expression of a gene encoding a IOOkilodalton mosquitocidal toxin from Bucillus sphuerictrs. lottrnul of Bacteriology 173, 27762785. White, P.J. & Lotay, H. 1980 Minimal nutritional requirements of Bucillrrs sphuerictts NCTC 9602 and 26 other strains of this species: the majority grow and sporulate with acetate as sole major source of carbon. lotrrnul of General Microbiology 118,
13-19. Xu, B.Z., Becker, N., Xiao, X.Q. & Ludwig, H.W. 1992 Microbial control of mosquitoes in Hubei Province, People’s Republic of China. Bulletin of the Society for Vector Ecology 17, l-10. Yousten, A.A. & Davidson, E.W. 1982 Ultrastructural analysis of spores and parasporal crystals formed in Bacillus sphuericus. Applied and Environmental Microbiology 44, 1449-1455.
(Received in revised form
20 July 1995; accepled 23 July 1995)
World Journal of Microbiology 6 Biotechnology, Vol 1.2,1996
11
