- Email: [email protected]
Difference in transcription levels of cap genes for γ-polyglutamic acid production between Bacillus subtilis IFO 16449 and Marburg 168
JOURNALOFBIOSCIENCEANDBIOENGINEERING Vol. 93, No. 2,252-254. 2002
Difference in Transcription Levels of cap Genes for y-Polyglutamic Acid Production between Bacillus subtilis IF0 16449 and Marburg 168 YUJI URUSHIBATA,’
SHINJI TOKUYAMA,’
AND
YASUTAKA
TAHARA’*
Department ofApplied Biological Chemistry, Faculty ofAgriculture, Shizuoka Universi@, 836 Ohya, Shizuoka 422-8529, Japan’ Received
18September 2001/Accepted
12 November
2001
In a strain carrying capB-1acZ fusion of Bacillus subtilis IF016449, which produces a large amount of y-polyglutamic acid (PGA), P-galactosidase activity was enhanced by about five times with the addition of L-glutamic acid. This increase was also confirmed by Northern blot analysis. On the other hand, the activity was not detected in a strain carrying capB-1acZ fusion of B. subtilis Marburg 168. However, when the cap genes (capBC4 andywtC) were fused to the IPTG-inducible spat promoter, B. subtilis Marburg 168 produced PGA. These results suggest that the inability of B. subtilis Marburg 168 to produce PGA is due to defective expression of the cap genes. [Key words: y-polyglutamic acid, Bacillus subtilis, cap genes] y-Polyglutamic acid (PGA), a polymer that consists of Dand L-glutamic acids through y-glutamyl bonds, is produced in some strains of Bacillus species (l-6), and the genes required for PGA synthesis have been cloned as capBCA from B. anthracis (7) and as pgsBCA from B. subtilis IF03336 (8). Previously, we reported the cloning and sequencing of the capBCA and ywtC genes (ywsC and ywtABC, respectively) of B. subtilis IFO16449, a strain isolated from a commercially available fermented food natto (9). The nucleotide sequences of the four genes were almost identical to those of the ywsC and ywtABC genes of B. subtilis Marburg 168 (10). Despite the fact that B. subtilis Marburg 168 has nearly the same genes responsible for PGA synthesis, it does not produce PGA, suggesting that this strain may be defective in function or expression of the cap genes. In this study, we investigated the difference in the transcriptional levels of the cap genes between B. subtilis strains IF0 16449 and Marburg 168. In addition, we constructed a strain of B. subtilis Marburg 168 in which the transcription of cap genes is controlled by the spat promoter, and determined the PGA productivity. For construction of a strain carrying capB-/acZ i%sion of B. subtilis IF0 16449 and Marburg 168, we first constructed the plasmid pBR322-LC (Fig. 1) according to the following procedures. A 3 .O-kb XbaI-DraI-digested fragment containing the 1acZ gene and the 5’ region containing the spoVG ribosome-binding site was excised from pMUTin4MCS (1 1), and then inserted into the HincII site of pUCl9 after both termini had been blunt-ended. The constructed plasmid was named pUC-L. On the other hand, a l.O-kb %oII-Naeldigested fragment containing the chloramphenicol acetyltransferase gene (cat) was excised from pC194 and then
blunt-ended. The fragment was inserted into the blunt-ended KpnI site downstream of the 1acZ gene in pUC-L, generating pUC-LC. A 4.0-kb DNA fragment containing 1acZ and cat genes was excised with EcoRI from pUC-LC and then blunt-ended. The fragment was inserted into the blunt-ended Sac11 site 150-bp downstream of the initiation codon of the capB gene in pBFl (9). A 5.2-kb fragment containing capB1acZ and a l.O-kb region upstream of the capB gene was excised with PstI and inserted into the EcoRV site of pBR322 after both termini had been blunt-ended, generating pBR322-LC. The constructed plasmid was transformed in B. subtilis IF016449 and Marburg 168 competent cells according to the method of Cutting and Horn (12). The recombinant strains were selected on an LB plate containing 5 &ml chloramphenicol. Recombinants were expected to result from a double-crossover event on the chromosome at the capB locus (Fig. 1). Integration of the plasmid was verified by Southern hybridization (data not shown). B. subtilis IF0 16449, Marburg 168 and their derivatives were grown in Spizizen minimal medium (13) containing 0.5 pg/ml bio-
pBR322-LC
Chromosome
FIG. 1. Integration of pBR322-LC into B. suhtilis chromosome. The chromosomal DNA and plasmid sequences are indicated by thick and dotted lines, respectively.
* Corresponding phone/fax:
-
author. e-mail: [email protected] +81-(0)54-238-4878 252
NOTES
var.. 93, 2002 800
1
2
3
4
253
5
--cr--
0
2
4
6
Cultivation
8
10
12
time (h)
FIG. 2. Expression of capB-IacZ in B. subtilis IFOl6449. Cells were grown in Spizizen minimal medium in the presence (triangles) or absence (circles) of L-glutamic acid. Cell growth and P-galactosidase activity are shown by open symbols and closed symbols, respectively. One unit of P-galactosidase activity is equivalent to 1000x A420 lOD600/ml/min, where A420 is the absorbance at 420 nm.
tin at 37°C with rotary shaking. Cell growth was monitored by measuring the optical density at 600 nm. P-Galactosidase activity was assayed by the method of Nicholson and Setlow(14). Figure 2 shows the cell growth and the expression of cupB-ZucZ fusion in B. subtilis IF016449. P-Galactosidase activity increased in the middle log phase and a maximum level at the late log phase. When L-glutamic acid was added at a final concentration of 2%, the expression level increased 5 times that of the cells grown in the absence of glutamate, indicating that the expression of the cup genes was stimulated by L-glutamic acid. Northern blot analysis also showed high expression levels of the cap genes after addition of L-glutamic acid into the medium (Fig. 3). Despite the fact that the hybrid 3.0-kb transcript of capBCA and ywtC was less observed in cells grown in a glutamatefree medium, the amount of the hybrid transcript increased in the presence of L-glutamic acid. These results strongly suggest that L-glutamic acid enhances the expression of the cup genes of B. subtilis IF016449. On the other hand, we also constructed a strain carrying cupB-1ncZ fusion of B. subtilis Marburg 168 using pBR322-LC. No P-galactosidase activity was detected when this strain was grown in the presence of glutamate, although the strain had an upstream region of the capB gene of B. swbtilis IF016449 (data not shown). Next, we constructed a strain in which the cup promoter was fused with the spat promoter of B. subtilis Marburg 168 as follows. The OX-kb fragment containing the 5’ region, including a Shine-Dalgamo sequence, of the cupB gene was amplified by PCR using a sense primer, 5’-GCC AAGCTTCAATATAGAAGGAGATGTCG-3’, an antisense primer, 5’-TGCGAATTGCTGTGTGCCGATCG-3’ and pBF 1 (9) as a template DNA. The amplified fragment was digested with EcoRI and inserted into the EcoRl site of pMUTin4MCS generating pMUT0.8, which was transformed in B. subtilis Marburg 168. The transformant was selected on an LB plate containing 2 pg/ml erythromycin.
FIG. 3. Effects of L-glutamic acid addition on capB-1acZ expression. Total RNA was extracted from cells grown in Spizizen minimal medium containing 0% (lane l), 0.5% (lane 2), I .O% (lane 3), I .5% (lane 4), or 2.0% (lane 5) L-glutamic acid. Ten micrograms of total RNA was electrophoresed, and Northern blot analysis was performed using a DIG-labeled capB-specific probe described previously (9).
Integration of the pMUT0.8 into the chromosome DNA took place at the capB locus, and the cap genes were placed under the control of the spat promoter (Fig. 4A). Southern hybridization verified that the plasmid had been correctly integrated (data not shown). Cells were grown in a chemically defined medium (15) containing 50 ug/ml L-tryptophan for 4 h under the same conditions as described above. IPTG was added at a final concentration of 1 mM after 4-h cultivation. Cells were grown for an additional 15 h for
A Pcap
Pspac i&Z lucl
B
ori
em
12345
FIG. 4. SDS-PAGE analysis of PGA production in B. subtilis. (A) A schematic of the construct at the capB locus. The spat promoter (Pspac) and the putative cap promoter (Pcup) are indicated by bent arrows. (B) SDS-PAGE analysis of PGA produced by Marburg I68 (lanes 2, 3) and Marburg I68 with capB under the control of the spat promoter (lanes 4, 5) in the presence (lanes 3, 5) or absence (lanes 2,4) of IPTG. Lane I, purified PGA.
254
URUSHIBATA ET AL.
PGA production. PGA was purified from the supernatant by the ethanol precipitation method described previously (9). As shown in Fig. 4B, a small amount of PGA was produced by the addition of IPTG, but no PGA was detected in the medium without IPTG. These results suggest that Cap proteins of B. subtilis Marburg 168 are active in PGA synthesis, and that the inability of the bacterium to produce PGA is due to defective expression of the cap genes. This work was partly supported by a grant-in-aid for science research from the Ministry of Education, Science, and Culture of Japan (grant 11660081).
REFERENCES 1. Bruckner, V. and Kovacs, J.: Structure of poly-D-glutamic
2.
3.
4.
5.
6.
acid isolated from capusulated strains of B. anthracis. Nature, 172,508 (I 953). Thorne, C. B., Gomez, C. G., Blind, G. R., and Housewright, R.D.: Synthesis of glutamic acid and glutamyl polypeptide by BaciZlus anthrucis. J. Bacterial., 65, 472478 (1953). Birrer, G.A., Cromwick, A., and Gross, R.A.: yPoly(glutamic acid) formation by Bacillus lichenzjkmis 9945a: physiological and biochemical studies. Int. J. Biol. Macromol., 16,265-275 (1994). Ito, Y., Tanaka, T., Ohmachi, T., and Asada, Y.: Glutamic acid independent production of poly(y-glutamic acid) by Bacillus subtilis TAM-4. Biosci. Biotech. Biochem., 60, 1239-1242 (1996). Ogawa, Y., Yamaguchi, F., Yuasa, K., and Tahara, Y.: Eff~cient production of y-polyglutamic acid by Bacillus subtilis (natto) in jar fermenters. Biosci. Biotech. Biochem., 61, 1684-1687 (1997). Torii, M., Kurimura, O., Utsumi, S., NOZU, H., and Amano, T.: Decapsulation of Bacillus meguterium. Biken’s
.I. BIOS~I.BIOENG.,
J., 2,265-276 (1959). 7. Makino, S., Uchida, I., Terakado, N., Sasakawa, C., and Yoshikawa, M.: Molecular characterization and protein analysis of the cap region, which is essential for encapsulation in Bacillus unthracis. J. Bacterial., 171,722-730 (1989). 8. Ashiuchi, M., Soda, K., and Misono, H.: A poly-yglutamate synthetic system of Bacillus subtilis IF03336: gene cloning and biochemical analysis of poly-y-glutamate produced by Escherichia coli clone cells. Biochem. Biophys. Res. Commun., 263,6-l 2 (1999). 9. Urushibata, Y., Tokuyama, S., and Tabara, Y.: Characterization of the Bacillus subtilis yvsC gene, involved in y-polyglutamic acid production. J. Bacterial., 184, 337-343 (2002). 10. Presecan, E., Moszer, I., Boursier, L., Ramos, H. C., de la Fuente, V., Hullo, M.-F., Lelong, C., Schleich, S., Sekowska, A., Song, B. H., Villani, G., Kunst, F., Danchin, A., and Glaser, P.: The Bacillus subtilis genome from gerBC (3 11 degrees) to 1icR (334 degrees). Microbiology, 143,3313-3328 (1997). 11. Vagner, V., Dervyn, E., and Ehrlich, S.D.: A vector for systematic gene inactivation in Bacillus subtilis. Microbiology, 144,3097-3 104 (1998). 12. Cutting, S. M. and Vander Horn, P. B.: Genetic analysis, p. 27-74. In Hatwood, C. R. and Cutting, S. M. (ed.), Molecular biological methods for Bacillus. John Wiley & Sons Ltd., Chichester, UK (1990). 13. Anagnostopoulos, C. and Spizizen, J.: Requirements for transformation in Bacillus subtilis. J. Biol. Chem., 81, 741746 (1960). 14. Nicholson, W. L. and Setlow, P.: Sporulation, germination and outgrowth, p. 91450. In Harwood, C. R. and Cutting, S. M. (ed.), Molecular biological methods for Bacillus. John Wiley & Sons Ltd., Chichester, UK (1990). 15. Kunioka, M. and Goto, A.: Biosynthesis of poly(y-glutamic acid) from L-glutamic acid, citric acid and ammonium sulfate in Bacillus subtilis IF03335. Appl. Microbial. Biotechnol., 40,867-872 (1994).
Difference in Transcription Levels of cap Genes for y-Polyglutamic Acid Production between Bacillus subtilis IF0 16449 and Marburg 168 YUJI URUSHIBATA,’
SHINJI TOKUYAMA,’
AND
YASUTAKA
TAHARA’*
Department ofApplied Biological Chemistry, Faculty ofAgriculture, Shizuoka Universi@, 836 Ohya, Shizuoka 422-8529, Japan’ Received
18September 2001/Accepted
12 November
2001
In a strain carrying capB-1acZ fusion of Bacillus subtilis IF016449, which produces a large amount of y-polyglutamic acid (PGA), P-galactosidase activity was enhanced by about five times with the addition of L-glutamic acid. This increase was also confirmed by Northern blot analysis. On the other hand, the activity was not detected in a strain carrying capB-1acZ fusion of B. subtilis Marburg 168. However, when the cap genes (capBC4 andywtC) were fused to the IPTG-inducible spat promoter, B. subtilis Marburg 168 produced PGA. These results suggest that the inability of B. subtilis Marburg 168 to produce PGA is due to defective expression of the cap genes. [Key words: y-polyglutamic acid, Bacillus subtilis, cap genes] y-Polyglutamic acid (PGA), a polymer that consists of Dand L-glutamic acids through y-glutamyl bonds, is produced in some strains of Bacillus species (l-6), and the genes required for PGA synthesis have been cloned as capBCA from B. anthracis (7) and as pgsBCA from B. subtilis IF03336 (8). Previously, we reported the cloning and sequencing of the capBCA and ywtC genes (ywsC and ywtABC, respectively) of B. subtilis IFO16449, a strain isolated from a commercially available fermented food natto (9). The nucleotide sequences of the four genes were almost identical to those of the ywsC and ywtABC genes of B. subtilis Marburg 168 (10). Despite the fact that B. subtilis Marburg 168 has nearly the same genes responsible for PGA synthesis, it does not produce PGA, suggesting that this strain may be defective in function or expression of the cap genes. In this study, we investigated the difference in the transcriptional levels of the cap genes between B. subtilis strains IF0 16449 and Marburg 168. In addition, we constructed a strain of B. subtilis Marburg 168 in which the transcription of cap genes is controlled by the spat promoter, and determined the PGA productivity. For construction of a strain carrying capB-/acZ i%sion of B. subtilis IF0 16449 and Marburg 168, we first constructed the plasmid pBR322-LC (Fig. 1) according to the following procedures. A 3 .O-kb XbaI-DraI-digested fragment containing the 1acZ gene and the 5’ region containing the spoVG ribosome-binding site was excised from pMUTin4MCS (1 1), and then inserted into the HincII site of pUCl9 after both termini had been blunt-ended. The constructed plasmid was named pUC-L. On the other hand, a l.O-kb %oII-Naeldigested fragment containing the chloramphenicol acetyltransferase gene (cat) was excised from pC194 and then
blunt-ended. The fragment was inserted into the blunt-ended KpnI site downstream of the 1acZ gene in pUC-L, generating pUC-LC. A 4.0-kb DNA fragment containing 1acZ and cat genes was excised with EcoRI from pUC-LC and then blunt-ended. The fragment was inserted into the blunt-ended Sac11 site 150-bp downstream of the initiation codon of the capB gene in pBFl (9). A 5.2-kb fragment containing capB1acZ and a l.O-kb region upstream of the capB gene was excised with PstI and inserted into the EcoRV site of pBR322 after both termini had been blunt-ended, generating pBR322-LC. The constructed plasmid was transformed in B. subtilis IF016449 and Marburg 168 competent cells according to the method of Cutting and Horn (12). The recombinant strains were selected on an LB plate containing 5 &ml chloramphenicol. Recombinants were expected to result from a double-crossover event on the chromosome at the capB locus (Fig. 1). Integration of the plasmid was verified by Southern hybridization (data not shown). B. subtilis IF0 16449, Marburg 168 and their derivatives were grown in Spizizen minimal medium (13) containing 0.5 pg/ml bio-
pBR322-LC
Chromosome
FIG. 1. Integration of pBR322-LC into B. suhtilis chromosome. The chromosomal DNA and plasmid sequences are indicated by thick and dotted lines, respectively.
* Corresponding phone/fax:
-
author. e-mail: [email protected] +81-(0)54-238-4878 252
NOTES
var.. 93, 2002 800
1
2
3
4
253
5
--cr--
0
2
4
6
Cultivation
8
10
12
time (h)
FIG. 2. Expression of capB-IacZ in B. subtilis IFOl6449. Cells were grown in Spizizen minimal medium in the presence (triangles) or absence (circles) of L-glutamic acid. Cell growth and P-galactosidase activity are shown by open symbols and closed symbols, respectively. One unit of P-galactosidase activity is equivalent to 1000x A420 lOD600/ml/min, where A420 is the absorbance at 420 nm.
tin at 37°C with rotary shaking. Cell growth was monitored by measuring the optical density at 600 nm. P-Galactosidase activity was assayed by the method of Nicholson and Setlow(14). Figure 2 shows the cell growth and the expression of cupB-ZucZ fusion in B. subtilis IF016449. P-Galactosidase activity increased in the middle log phase and a maximum level at the late log phase. When L-glutamic acid was added at a final concentration of 2%, the expression level increased 5 times that of the cells grown in the absence of glutamate, indicating that the expression of the cup genes was stimulated by L-glutamic acid. Northern blot analysis also showed high expression levels of the cap genes after addition of L-glutamic acid into the medium (Fig. 3). Despite the fact that the hybrid 3.0-kb transcript of capBCA and ywtC was less observed in cells grown in a glutamatefree medium, the amount of the hybrid transcript increased in the presence of L-glutamic acid. These results strongly suggest that L-glutamic acid enhances the expression of the cup genes of B. subtilis IF016449. On the other hand, we also constructed a strain carrying cupB-1ncZ fusion of B. subtilis Marburg 168 using pBR322-LC. No P-galactosidase activity was detected when this strain was grown in the presence of glutamate, although the strain had an upstream region of the capB gene of B. swbtilis IF016449 (data not shown). Next, we constructed a strain in which the cup promoter was fused with the spat promoter of B. subtilis Marburg 168 as follows. The OX-kb fragment containing the 5’ region, including a Shine-Dalgamo sequence, of the cupB gene was amplified by PCR using a sense primer, 5’-GCC AAGCTTCAATATAGAAGGAGATGTCG-3’, an antisense primer, 5’-TGCGAATTGCTGTGTGCCGATCG-3’ and pBF 1 (9) as a template DNA. The amplified fragment was digested with EcoRI and inserted into the EcoRl site of pMUTin4MCS generating pMUT0.8, which was transformed in B. subtilis Marburg 168. The transformant was selected on an LB plate containing 2 pg/ml erythromycin.
FIG. 3. Effects of L-glutamic acid addition on capB-1acZ expression. Total RNA was extracted from cells grown in Spizizen minimal medium containing 0% (lane l), 0.5% (lane 2), I .O% (lane 3), I .5% (lane 4), or 2.0% (lane 5) L-glutamic acid. Ten micrograms of total RNA was electrophoresed, and Northern blot analysis was performed using a DIG-labeled capB-specific probe described previously (9).
Integration of the pMUT0.8 into the chromosome DNA took place at the capB locus, and the cap genes were placed under the control of the spat promoter (Fig. 4A). Southern hybridization verified that the plasmid had been correctly integrated (data not shown). Cells were grown in a chemically defined medium (15) containing 50 ug/ml L-tryptophan for 4 h under the same conditions as described above. IPTG was added at a final concentration of 1 mM after 4-h cultivation. Cells were grown for an additional 15 h for
A Pcap
Pspac i&Z lucl
B
ori
em
12345
FIG. 4. SDS-PAGE analysis of PGA production in B. subtilis. (A) A schematic of the construct at the capB locus. The spat promoter (Pspac) and the putative cap promoter (Pcup) are indicated by bent arrows. (B) SDS-PAGE analysis of PGA produced by Marburg I68 (lanes 2, 3) and Marburg I68 with capB under the control of the spat promoter (lanes 4, 5) in the presence (lanes 3, 5) or absence (lanes 2,4) of IPTG. Lane I, purified PGA.
254
URUSHIBATA ET AL.
PGA production. PGA was purified from the supernatant by the ethanol precipitation method described previously (9). As shown in Fig. 4B, a small amount of PGA was produced by the addition of IPTG, but no PGA was detected in the medium without IPTG. These results suggest that Cap proteins of B. subtilis Marburg 168 are active in PGA synthesis, and that the inability of the bacterium to produce PGA is due to defective expression of the cap genes. This work was partly supported by a grant-in-aid for science research from the Ministry of Education, Science, and Culture of Japan (grant 11660081).
REFERENCES 1. Bruckner, V. and Kovacs, J.: Structure of poly-D-glutamic
2.
3.
4.
5.
6.
acid isolated from capusulated strains of B. anthracis. Nature, 172,508 (I 953). Thorne, C. B., Gomez, C. G., Blind, G. R., and Housewright, R.D.: Synthesis of glutamic acid and glutamyl polypeptide by BaciZlus anthrucis. J. Bacterial., 65, 472478 (1953). Birrer, G.A., Cromwick, A., and Gross, R.A.: yPoly(glutamic acid) formation by Bacillus lichenzjkmis 9945a: physiological and biochemical studies. Int. J. Biol. Macromol., 16,265-275 (1994). Ito, Y., Tanaka, T., Ohmachi, T., and Asada, Y.: Glutamic acid independent production of poly(y-glutamic acid) by Bacillus subtilis TAM-4. Biosci. Biotech. Biochem., 60, 1239-1242 (1996). Ogawa, Y., Yamaguchi, F., Yuasa, K., and Tahara, Y.: Eff~cient production of y-polyglutamic acid by Bacillus subtilis (natto) in jar fermenters. Biosci. Biotech. Biochem., 61, 1684-1687 (1997). Torii, M., Kurimura, O., Utsumi, S., NOZU, H., and Amano, T.: Decapsulation of Bacillus meguterium. Biken’s
.I. BIOS~I.BIOENG.,
J., 2,265-276 (1959). 7. Makino, S., Uchida, I., Terakado, N., Sasakawa, C., and Yoshikawa, M.: Molecular characterization and protein analysis of the cap region, which is essential for encapsulation in Bacillus unthracis. J. Bacterial., 171,722-730 (1989). 8. Ashiuchi, M., Soda, K., and Misono, H.: A poly-yglutamate synthetic system of Bacillus subtilis IF03336: gene cloning and biochemical analysis of poly-y-glutamate produced by Escherichia coli clone cells. Biochem. Biophys. Res. Commun., 263,6-l 2 (1999). 9. Urushibata, Y., Tokuyama, S., and Tabara, Y.: Characterization of the Bacillus subtilis yvsC gene, involved in y-polyglutamic acid production. J. Bacterial., 184, 337-343 (2002). 10. Presecan, E., Moszer, I., Boursier, L., Ramos, H. C., de la Fuente, V., Hullo, M.-F., Lelong, C., Schleich, S., Sekowska, A., Song, B. H., Villani, G., Kunst, F., Danchin, A., and Glaser, P.: The Bacillus subtilis genome from gerBC (3 11 degrees) to 1icR (334 degrees). Microbiology, 143,3313-3328 (1997). 11. Vagner, V., Dervyn, E., and Ehrlich, S.D.: A vector for systematic gene inactivation in Bacillus subtilis. Microbiology, 144,3097-3 104 (1998). 12. Cutting, S. M. and Vander Horn, P. B.: Genetic analysis, p. 27-74. In Hatwood, C. R. and Cutting, S. M. (ed.), Molecular biological methods for Bacillus. John Wiley & Sons Ltd., Chichester, UK (1990). 13. Anagnostopoulos, C. and Spizizen, J.: Requirements for transformation in Bacillus subtilis. J. Biol. Chem., 81, 741746 (1960). 14. Nicholson, W. L. and Setlow, P.: Sporulation, germination and outgrowth, p. 91450. In Harwood, C. R. and Cutting, S. M. (ed.), Molecular biological methods for Bacillus. John Wiley & Sons Ltd., Chichester, UK (1990). 15. Kunioka, M. and Goto, A.: Biosynthesis of poly(y-glutamic acid) from L-glutamic acid, citric acid and ammonium sulfate in Bacillus subtilis IF03335. Appl. Microbial. Biotechnol., 40,867-872 (1994).