- Email: [email protected]
An assay for the detection of superoxide dismutase in individual Escherichia coli colonies
ANALYTICAL
168,455-461
BIOCHEMISTRY
(1988)
An Assay for the Detection of Superoxide Dismutase Eschefichia co/i Colonies’ JOAN R. SCHIAVONE Departments
in Individual
AND HOSNI M. HASSAN*
of Food Science and Microbiology, Raleigh, North Carolina
North Carolina 27695- 7624
State University,
Received July 27, 1987 A method for detecting superoxide dismutase activity in individual colonies of Escherichia was developed. The assayinvolves the lysis of individual cells in colonies on filter papers by a series of lysozyme, chloroform, and freeze-thaw treatments. Filters are placed on agar plates to allow diffusion of cellular enzymes into a solid matrix. A nitroblue tetrazolium overlay is applied to detect superoxide dismutase activity. Colonies possessing activity produce achromatic zones against a dark Formazan background. The assaycan detect the presence of superoxide dismutase and the relative amount of enzyme as well. This assay provides a method for screening a population of cells for mutants deficient in or overproducing superoxide dismucoli
tZ+Z.
0 1988 Academic
Press, Inc.
KEY WORDS: SOD; whole cells; NBT-plate
assay;Escherichia
Superoxide dismutases (SODS)~ are metalloenzymes that protect cells against oxygen toxicity by scavenging the superoxide radical (O;), the first intermediate produced during the univalent reduction of molecular oxygen (1). Escherichia coli possesses three dimeric forms of superoxide dismutase-a manganese-containing homodimer (MnSOD), an iron-containing homodimer (FeSOD), and a heterodimer (HySOD) containing one subunit of the Mn enzyme and one subunit of the Fe enzyme (2). In E. coli, the FeSOD is a ’ This work was supported in part by Grants DMB-8609239 from the National Science Foundation and 86-G-007 19 from the NC Biotechnology Center. This is Paper No. 11192 of the Journal Series of the North Carolina Agricultural Research Service, Raleigh, NC 27695. The use of trade names in this publication does not imply endorsement by the North Carolina Research Service of the products named, nor criticism of ones not mentioned. * To whom all correspondence should be addressed. 3 Abbreviations used: SOD, superoxide dismutase; NBT, nitroblue tetrazolium; TEMED, N, N, N’,N’-tetramethylethylenediamine; LB, Luria-Bertani.
coli;
SOD-mutants.
constitutive enzyme, whereas MnSOD is inducible upon exposure to oxygen (2) to redox active compounds that generate 0: in the presence of oxygen (3,4), and to iron chelators (5). The regulation of MnSOD biosynthesis is currently under investigation in several laboratories, but efforts have been hampered by the inability to identify and isolate regulatory mutants. In this paper we describe a rapid and inexpensive method for detecting SOD activities in a single colony. The method can identify mutants lacking superoxide dismutase activity as well as strains that overproduce the enzyme. MATERIALS
AND METHODS
Materials. Kanamycin, chloramphenicol, ampicillin, tetracycline, lysozyme, NBT (nitroblue tetrazolium), riboflavin, and TEMED (N,N,N’,W-tetramethylethylenediamine) were purchased from Sigma. Commercially prepared MnSOD from E. coli 455
0003-2697/88 $3.00 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any fom reserved.
456
SCHIAVONE
AND TABLE
STRAINS OF
1
Escherichia coli K- 12 AND F%ASMIDS USED IN THIS STUDY
Strain
GC4468 QC78 1 QC773 QC779
HASSAN
Description Alac U169 rpsL GC4468 but (sodA::MudPRl3)25 CmR CC4468 but ~j@odB-kun) 1 -A2 KmR QC773 but (sodA::MudPR 13)25 CmR
AB2463 AB2463A+ AB2463B+
F as as as F as as
Plasmids pDTl-5 pHSl-4
sodA + ApR sodB+ TetR
Reference
or source
(6) D. Touati
(6) (6)
leu thr proA argE his lac gal rpsL recA
(7)
AB2463 AB2463
(8)
(pDTl-5) (pHSl-4)
(9) (9) (10)
(Sigma) was resuspended in 0.05 M phos- isozymes were quantitated via linear scanphate buffer to a final concentration of 5 ning densitometry (14). mg/ml. Strains and growth conditions. The strains RESULTS used in this study are listed in Table 1. Cultures were grown in LB media containing, Detection of SOD in individual colonies. In per liter, 10 g Bacto peptone (Sigma), 5 g order to screen thousands of colonies for the yeast extract (Difco), and 10 g NaCl (Sigma). absence or overproduction of SOD, a simple The strains were grown overnight at 37°C on and reliable assay was required. Most assays a rotary shaker (200 rpm) in liquid LB media for SOD depend on the presence of a source containing the appropriate antibiotics, trans- or a generator of superoxide radical (0:) and ferred into microtiter plates, and then replica an indicator molecule that can react with 0: to develop a characteristic color (15). Xanplated onto LB agar plates. The resulting colthine/xanthine oxidase and riboflavin/light onies were used for the development of the are often used as 0: generators, while cySOD colony assay. tochrome c3+ and NBT serve as detector Cultures used for the determination of molecules ( 15). Unfortunately, these SOD activity via liquid assays were grown as a lawn on LB plates and then scraped off and methods are not applicable to intact cells beresuspended in 0.05 M phosphate buffer con- cause 0: does not cross the cytoplasmic taining 0.1 mM EDTA. This protocol was membrane (16). Therefore, the rationale befollowed in order to mimic the colony hind this new method was to allow cells to grow into individual colonies before transgrowth conditions. Assays.Cell-free extracts were prepared as ferring and lysing them onto filter papers. described previously ( 10). Protein was deter- The colony lysates may then be subjected to mined by the method of Lowry et al. (11) a modification of the NBT assay (13) for deusing bovine serum albumin as a standard. tecting SOD activity. We tried several different protocols for Superoxide dismutase activity was estimated lysing and assaying SOD in individual coloby the cytochrome c method (1). Superoxide dismutase isozymes were separated on 10% nies and found the following protocol (Fig. 1) polyacrylamide gels ( 12) and visualized via to yield the best results. Individual colonies the NBT method (13). Superoxide dismutase were grown overnight and then transferred
SUPEROXIDE
DISMUTASE
via blotting to Whatman No. 1 filters with the orientation marked. The filters were inverted and placed in glass petri dishes containing 0.50 ml of a 1 mg/ml lysozyme solution. After 30 min of incubation at room temperature, the plates were incubated (30 min) in a glass desiccator having a chloroform atmosphere to further effect lysis. The plates were then removed from the desiccator and subjected to three ( lo-min) freezethaw cycles. The filters were then transferred, colony side up, to petri plates containing 10 ml of solidified 1% agar dissolved in 0.05 M phosphate/O.1 mM EDTA buffer. Care was taken to avoid the formation of air bubbles between the filters and the agar surface. The colony orientation was marked on the bottom of the petri plates. The plates were incubated for 3 h at room temperature to allow diffusion of cell-free extracts into the agar matrix. The filters were removed and a NBT overlay solution ( 15 ml) was applied to each plate. The overlay was a modification of (1) and consisted of 0.55 mM NBT, 66.0 PM ri-
Growth Individual Colonies
of
rI
Pick
Single
: #
Grow
in
I I , L
Colonies
+
Microtiter
Replica
Plate
Transfer
to
Plates Onto
Filter
Solid
Media
Paper
i Lysoryme
!lmg/ml) Treatment, 30 min I Treatment, 30 min + Freeze-Thaw Treatment, 3 x 10 min I
Chlorofot?
I
Blotting
f I I L
Transfer Extracts
Add
i
of
Colonies'
to 1% Agar II
Cell-free Plates,
3 h
i
NBT Overlay f Incubate in Dark, t Expose to Light
3 h
FIG. 1. Flow diagram illustrating the method developed for the detection of SOD activity in individual colonies.
PLATE ASSAY
457
boflavin, 0.5% TEMED, and 1.0% agar dissolved in 0.05 M phosphate/O.1 mM EDTA buffer. The NBT, riboflavin, and TEMED were mixed together and added simultaneously to the tempered (55Y) agar solution directly prior to pouring the overlay solution. The plates were incubated at room temperature in the dark for 3 h and then exposed to light from a 15-W incandescent lamp for 5- 10 min or until full color development occurred. Colonies possessing SOD activity produced achromatic zones against a dark Formazan background. A 7% acetic acid solution was poured onto the plates to stop the reaction. Results in Fig. 2 demonstrate the use of the plate assay to detect SOD activities in strains of bacteria that possess different amounts of the enzyme. Figure 2A shows the differences in superoxide dismutase activity between the parental wild type (sodA+sodB+), the MnSOD-deficient mutant (sodA-), the FeSOD-deficient mutant (sodB-), and the double negative strain (sodA-sodB-). Note that the strains lacking one isozyme displayed less activity than the parental control, whereas very little or no activity was detected in the double-negative mutant. The changes in super-oxide dismutase activity as seen by the plate assay correlated well with the differences seen in superoxide dismutase activity assayed by the cytochrome c method (Table 2). For example, QC773 (sodB-) produced more superoxide dismutase activity than did QC781 (sodA-). The activity seen on the plate assay indicated that the parental strain, GC4468 (sodA+sodB+), produced the most activity, followed by QC773, QC781, and QC779 (sodA-sodB-), respectively. The small amount of activity seen by the double-negative mutant presumably resulted from growing the mutant on LB plates without any antibiotic stress, thus allowing some degree of reversion to occur. Results shown in Fig. 2B demonstrate that the plate assay may be used to detect overproducers of superoxide dismutase as well.
458
SCHIAVONE
FIG. 2. The seven strains described in Table 1 were grown overnight in LB media containing the appropriate antibiotics. Aliquots (1~1) of the cultures were applied to LB plates and incubated overnight. The resulting colonies were used to assaysuperoxide dismutase activity via the plate assay as described under Materials and Methods. (A) Lane 1, E. coli Gc4468; 2, E. coli QC78 1; 3, E. coli QC773; 4, E. coli QC779. (B) Lane 1, E. coli AB2463; 2, E. coli AB2463A+ (pDTl-5); 3, E. coli AB2463B+ (pHS l-4).
The two strains carrying the sodA or sodB genes on a multicopy plasmid (lanes 2 and 3, respectively) exhibited much more activity than the parental control. These differences were also observed when activity was mea-
AND HASSAN
sured by the cytochrome c assay (Table 2). Both the plate assay and the liquid assay illustrated that the strain harboring the multicopy plasmid that contains FeSOD, pHS l-4, exhibited more activity than either the parental wild type or the strain containing the MnSOD plasmid, pDTl-5. These results confirm previous findings (8,9) which show that overproducers of FeSOD lack the regulatory control found in overproducers of the Mn enzyme. Note, however, that AB2463Bf exhibited greater activity than the parental control in the plate assay even though the distinction was not representative of the fold induction seen via the liquid assay. This was due to the fact that the NBT stain cannot detect a linear increase in activity beyond a certain concentration. Sensitivity of the plate assay. Figure 3 illustrates the sensitivity of the NBT plate assay with different concentrations of pure MnSOD. Each protein concentration was delivered in a lo-~1 sample volume. This assay method was sensitive to 0.05-hg quantities of pure superoxide dismutase. This degree of sensitivity was slightly less than the sensitivity of the NBT staining method applicable to polyacrylamide gels which may detect as little as 0.0 16 pg of bovine erythrocyte superoxide dismutase (13). The decrease in sensitivity was attributed to the fact that the plate assay does not confine the enzyme to a small, sharp zone as is the case when the enzyme is electrophoresed on polyacrylamide gels and then stained for activity. The achromatic zones on the plates became larger and clearer as the concentration of protein increased as commonly observed in polyacrylamide gels stained for superoxide dismutase activity. The plate assay did have an upper limit, however, for a point was reached where a certain enzyme concentration yielded a sharp achromatic zone that did not become any sharper or clearer as the concentration of enzyme was increased (data not shown).
SUPEROXIDE
DISMUTASE
DISCUSSION
Efforts to isolate and identify superoxide dismutase mutants have been hampered because the enzyme does not impart readily identifiable traits such as auxotrophy, and the exact phenotype expressed by such mutations is not fully understood. Carlioz and Touati (17) generated and isolated E. coli mutants deficient in Mn-, Fe-, or both isozymic forms of superoxide dismutase. Their method of choice for the detection of these mutants involves immunoblotting analysis of crude extracts followed by a confirmation via the cytochrome c assay. Even though effective, the selection of super-oxide dismutase mutants based on their protocol can be expensive and time consuming. The plate assay that we have developed incorporates the principle of detecting superoxide dismutase activity via the NBT method with the lysis of whole cells from individual colonies and subsequent leakage of cellular enzymes into an agar matrix. This procedure allows one to rapidly screen many colonies simultaneously for superoxide dismutase activity. As illustrated in Figs. 2 and 3, the assay can discriminate between mu-
PLATE
459
ASSAY
tants lacking superoxide dismutase activity and mutants which overproduce the enzyme. This procedure does not require the lengthy preparation of cell-free extracts, the preparation of antibodies specific for one isozymic form of superoxide dismutase, or expensive laboratory equipment. Another positive asset to the assay is that the physiological state of the cells is not an important factor for the detection of superoxide dismutase mutants, as it is when the selection procedure is based on paraquat resistance (18) or oxygen sensitivity ( 17). The plate assay may also be modified to differentiate between the isozymic forms of superoxide dismutase in E. coli. The assay may be used to isolate mutants in FeSOD biosynthesis by simply growing the cells anaerobically prior to lysis. The Mn enzyme is not synthesized anaerobically, so any activity present is due to the FeSOD. The isolation of MnSOD-deficient mutants can be done easily in a FeSOD-negative background by growing the cells aerobically and then screening for superoxide dismutase activity. The isolation of MnSOD-deficient mutants in a FeSOD-positive background is not as
TABLE 2 SUPEROXIDE
DISMUTASE
ACTIVITY
IN VARIOUS
STRAINS OF Escherichia
coli
K12
Superoxide dismutase (U/mg) Strain
Phenotype
GC4468 QC78 I QC773 QC779
Wild type sodAsodBsodA-sodB-
AB2463 AB2463A+ AB2463B+
Wild type sodA++ sodB++
Total
Mn-
HY-
Fe-
25.0 5.6 20.5 1.2
20.2 0.0 20.5 1.2
2.1 1.3 0.0 0.0
2.7 4.4 0.0 0.0
30.8 41.7 244.1
21.8 37.1 104.7
3.3 3.1 31.5
5.7 1.5 107.9
Note. The strains outlined in Table 1 were grown overnight in LB media with the appropriate antibiotics. Approximately 200 ~1 of each overnight culture was applied to LB plates without antibiotics present. The bacterial lawns were scraped off each plate and resuspended in 0.05 M phosphate buffer containing 0.1 mM EDTA. Cell-free extracts were prepared and superoxide dismutase activity measured as described under Materials and Methods.
460
SCHIAVONE
FIG. 3. Wells (6 mm diameter) were bored into the bottom agar surface and the plugs were removed. The bottom of the wells were subsequently sealed with a few microliters of molten agar. Serial dilutions of manganese guperoxide dismutase from Escherichia coli were ap plied to each well in a total volume of 10 ~1. The enzyme was allowed to diffuse into the agar layer for 3 h before the NBT overlay was applied. The staining procedure that followed was as described under Materials and Methods. The following amounts of superoxide dismutase (ag) were applied to the wells: (1) 0.05, (2) 0.10, (3) 0.50, (4) 1, (5) 5, (6) 10, (7) 25, and (8) 50.
AND HASSAN
biosynthesis, or work in a catalasedeficient background. The SOD assay described in this paper or a modification thereof may be applicable to detecting SOD activities in other prokaryotic organisms or eukaryotic systems such as yeast. The principle of lysing whole cells onto a solid support may also be applied for the detection of other enzymes that do not exhibit readily identifiable phenotypic markers but for which a chromogenic assay is known. The plate assay is presently being used in our laboratory to isolate regulatory mutants of superoxide dismutase in E. coli. The current model for the regulation of MnSOD biosynthesis in E. coli suggests that a negative control operon exists whereby the activity of an iron-containing repressor protein is sensitive to changes in the redox state of the cell (21; J. R. Schiavone and H. H. Hassan, manuscript submitted for publication). The plate assay will allow us to further elucidate the mechanism of regulation by providing the means to screen a large population of cells for mutants derepressed for superoxide dismutase biosynthesis. ACKNOWLEDGMENT
straightforward. Asada et al. (19) showed that the FeSOD is sensitive to hydrogen peroxide, and the isozyme may be inactivated in a polyacrylamide gel by adding hydrogen peroxide to the staining solution; but polyacrylamide gels treated with hydrogen peroxide and then stained for superoxide dismutase show achromatic zones that comigrate with catalase (20). This anomaly can be overlooked in a polyacrylamide gel because the catalase and FeSOD are separated during electrophoresis. This is not the case, however, in the plate/colony assay. The catalase and FeSOD are not separated so that any hydrogen peroxide treatment results in the production of achromatic zones and oxygen bubbles in the agar. One must take precautions to inactivate any catalases present, alter the growth conditions to minimize catalase
We thank the late J. S. Siwecki for his generous gift of strains AB2463, AB2463A+, and AB2463B+ and D. Tovati for her generous gift of strains QC78 1, QC773, and Qc779.
REFERENCES 1. McCord, J. M., and Fridovich, I. (1969) J. Biol. Chem. 244,6049-6055. 2. Hassan, H. M., and Fridovich, I. ( 1977) J. Bucteriol. 129, 1574-1583. 3. Hassan, H. M., and Ftidovich, I. (1977) J. Biol. Chem. 252,7667-7672. 4. Hassan, H. M., and Fridovich, I. (1979) Arch. Biothem. Biophys. 1%,385-395. 5. Hassan, H. M., and Moody, C. S. (1984) FEMS Microbial. L&t. 25,233-236. 6. Farr, S. B., D’Ari, R., and Touati, D. (1986) Proc. Natl. Acad. Sci. USA 83,8268-8272. 7. Howard-Flanders, P., and Boyce, R. P. ( 1966) Radiat. Rex 6(Suppl.), 156- 184. 8. Touati, D. (1983) J. Bucteriol. 155, 1078-1087. 9. Sakamoto, H., and Touati, D. (1984) J. Bucteriol 159,418-420.
SUPEROXIDE
DISMUTASE
10. Moody, C. S., and Hassan, H. M. (1982) Proc. NutI. Acad. Sci. USA 79,2855-2859. 11. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275.
12. Davis, B. J. (1964) Ann. N.Y. Acad. Sci. 121,
ASSAY
461
16. Hassan, H. M., and Fridovich, I. (1979) J. Biol. Chem. 254, 10846-10852. 17. Carlioz, A., and Touati, D. (1986) EMBO J. 5, 623-630.
18. Bloch, C. A., and Ausebel, F. M. (1986) .I. Bucferiol. 168,795-798.
404-421.
13. Beauchamp, C., and Fridovich, I. (197 1) Anal. Biothem. 44,276-287. 14. Hassan, H. M., and Fridovich, I. (1977) J. Bacterial. 129, 1574-1583. 15. McCord, J. M., Crapo, J. D., and Fridovich, I. ( 1977) in Superoxide and Superoxide Dismutases (Michelson, A. M., McCord, J. M., and Fridovich, I., Eds.), pp. I l- 17, Academic Press, New York.
PLATE
19. Asada, K., Yoshikawa, K., Takahashi, M., Maeda, Y., and Enmanji, K. (1975) J. Biol. Chem. 250, 2801-2807. 20. Clare, D. A., Duong, M. N., Darr, D., Archibald, F., and Fridovich, I. (1984) Anal. Biochem. 140, 532-537. 21.
Moody, C. S., and Hassan, H. M. (1984) J. Eiol. Chem. 259, 12821-12825.
168,455-461
BIOCHEMISTRY
(1988)
An Assay for the Detection of Superoxide Dismutase Eschefichia co/i Colonies’ JOAN R. SCHIAVONE Departments
in Individual
AND HOSNI M. HASSAN*
of Food Science and Microbiology, Raleigh, North Carolina
North Carolina 27695- 7624
State University,
Received July 27, 1987 A method for detecting superoxide dismutase activity in individual colonies of Escherichia was developed. The assayinvolves the lysis of individual cells in colonies on filter papers by a series of lysozyme, chloroform, and freeze-thaw treatments. Filters are placed on agar plates to allow diffusion of cellular enzymes into a solid matrix. A nitroblue tetrazolium overlay is applied to detect superoxide dismutase activity. Colonies possessing activity produce achromatic zones against a dark Formazan background. The assaycan detect the presence of superoxide dismutase and the relative amount of enzyme as well. This assay provides a method for screening a population of cells for mutants deficient in or overproducing superoxide dismucoli
tZ+Z.
0 1988 Academic
Press, Inc.
KEY WORDS: SOD; whole cells; NBT-plate
assay;Escherichia
Superoxide dismutases (SODS)~ are metalloenzymes that protect cells against oxygen toxicity by scavenging the superoxide radical (O;), the first intermediate produced during the univalent reduction of molecular oxygen (1). Escherichia coli possesses three dimeric forms of superoxide dismutase-a manganese-containing homodimer (MnSOD), an iron-containing homodimer (FeSOD), and a heterodimer (HySOD) containing one subunit of the Mn enzyme and one subunit of the Fe enzyme (2). In E. coli, the FeSOD is a ’ This work was supported in part by Grants DMB-8609239 from the National Science Foundation and 86-G-007 19 from the NC Biotechnology Center. This is Paper No. 11192 of the Journal Series of the North Carolina Agricultural Research Service, Raleigh, NC 27695. The use of trade names in this publication does not imply endorsement by the North Carolina Research Service of the products named, nor criticism of ones not mentioned. * To whom all correspondence should be addressed. 3 Abbreviations used: SOD, superoxide dismutase; NBT, nitroblue tetrazolium; TEMED, N, N, N’,N’-tetramethylethylenediamine; LB, Luria-Bertani.
coli;
SOD-mutants.
constitutive enzyme, whereas MnSOD is inducible upon exposure to oxygen (2) to redox active compounds that generate 0: in the presence of oxygen (3,4), and to iron chelators (5). The regulation of MnSOD biosynthesis is currently under investigation in several laboratories, but efforts have been hampered by the inability to identify and isolate regulatory mutants. In this paper we describe a rapid and inexpensive method for detecting SOD activities in a single colony. The method can identify mutants lacking superoxide dismutase activity as well as strains that overproduce the enzyme. MATERIALS
AND METHODS
Materials. Kanamycin, chloramphenicol, ampicillin, tetracycline, lysozyme, NBT (nitroblue tetrazolium), riboflavin, and TEMED (N,N,N’,W-tetramethylethylenediamine) were purchased from Sigma. Commercially prepared MnSOD from E. coli 455
0003-2697/88 $3.00 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any fom reserved.
456
SCHIAVONE
AND TABLE
STRAINS OF
1
Escherichia coli K- 12 AND F%ASMIDS USED IN THIS STUDY
Strain
GC4468 QC78 1 QC773 QC779
HASSAN
Description Alac U169 rpsL GC4468 but (sodA::MudPRl3)25 CmR CC4468 but ~j@odB-kun) 1 -A2 KmR QC773 but (sodA::MudPR 13)25 CmR
AB2463 AB2463A+ AB2463B+
F as as as F as as
Plasmids pDTl-5 pHSl-4
sodA + ApR sodB+ TetR
Reference
or source
(6) D. Touati
(6) (6)
leu thr proA argE his lac gal rpsL recA
(7)
AB2463 AB2463
(8)
(pDTl-5) (pHSl-4)
(9) (9) (10)
(Sigma) was resuspended in 0.05 M phos- isozymes were quantitated via linear scanphate buffer to a final concentration of 5 ning densitometry (14). mg/ml. Strains and growth conditions. The strains RESULTS used in this study are listed in Table 1. Cultures were grown in LB media containing, Detection of SOD in individual colonies. In per liter, 10 g Bacto peptone (Sigma), 5 g order to screen thousands of colonies for the yeast extract (Difco), and 10 g NaCl (Sigma). absence or overproduction of SOD, a simple The strains were grown overnight at 37°C on and reliable assay was required. Most assays a rotary shaker (200 rpm) in liquid LB media for SOD depend on the presence of a source containing the appropriate antibiotics, trans- or a generator of superoxide radical (0:) and ferred into microtiter plates, and then replica an indicator molecule that can react with 0: to develop a characteristic color (15). Xanplated onto LB agar plates. The resulting colthine/xanthine oxidase and riboflavin/light onies were used for the development of the are often used as 0: generators, while cySOD colony assay. tochrome c3+ and NBT serve as detector Cultures used for the determination of molecules ( 15). Unfortunately, these SOD activity via liquid assays were grown as a lawn on LB plates and then scraped off and methods are not applicable to intact cells beresuspended in 0.05 M phosphate buffer con- cause 0: does not cross the cytoplasmic taining 0.1 mM EDTA. This protocol was membrane (16). Therefore, the rationale befollowed in order to mimic the colony hind this new method was to allow cells to grow into individual colonies before transgrowth conditions. Assays.Cell-free extracts were prepared as ferring and lysing them onto filter papers. described previously ( 10). Protein was deter- The colony lysates may then be subjected to mined by the method of Lowry et al. (11) a modification of the NBT assay (13) for deusing bovine serum albumin as a standard. tecting SOD activity. We tried several different protocols for Superoxide dismutase activity was estimated lysing and assaying SOD in individual coloby the cytochrome c method (1). Superoxide dismutase isozymes were separated on 10% nies and found the following protocol (Fig. 1) polyacrylamide gels ( 12) and visualized via to yield the best results. Individual colonies the NBT method (13). Superoxide dismutase were grown overnight and then transferred
SUPEROXIDE
DISMUTASE
via blotting to Whatman No. 1 filters with the orientation marked. The filters were inverted and placed in glass petri dishes containing 0.50 ml of a 1 mg/ml lysozyme solution. After 30 min of incubation at room temperature, the plates were incubated (30 min) in a glass desiccator having a chloroform atmosphere to further effect lysis. The plates were then removed from the desiccator and subjected to three ( lo-min) freezethaw cycles. The filters were then transferred, colony side up, to petri plates containing 10 ml of solidified 1% agar dissolved in 0.05 M phosphate/O.1 mM EDTA buffer. Care was taken to avoid the formation of air bubbles between the filters and the agar surface. The colony orientation was marked on the bottom of the petri plates. The plates were incubated for 3 h at room temperature to allow diffusion of cell-free extracts into the agar matrix. The filters were removed and a NBT overlay solution ( 15 ml) was applied to each plate. The overlay was a modification of (1) and consisted of 0.55 mM NBT, 66.0 PM ri-
Growth Individual Colonies
of
rI
Pick
Single
: #
Grow
in
I I , L
Colonies
+
Microtiter
Replica
Plate
Transfer
to
Plates Onto
Filter
Solid
Media
Paper
i Lysoryme
!lmg/ml) Treatment, 30 min I Treatment, 30 min + Freeze-Thaw Treatment, 3 x 10 min I
Chlorofot?
I
Blotting
f I I L
Transfer Extracts
Add
i
of
Colonies'
to 1% Agar II
Cell-free Plates,
3 h
i
NBT Overlay f Incubate in Dark, t Expose to Light
3 h
FIG. 1. Flow diagram illustrating the method developed for the detection of SOD activity in individual colonies.
PLATE ASSAY
457
boflavin, 0.5% TEMED, and 1.0% agar dissolved in 0.05 M phosphate/O.1 mM EDTA buffer. The NBT, riboflavin, and TEMED were mixed together and added simultaneously to the tempered (55Y) agar solution directly prior to pouring the overlay solution. The plates were incubated at room temperature in the dark for 3 h and then exposed to light from a 15-W incandescent lamp for 5- 10 min or until full color development occurred. Colonies possessing SOD activity produced achromatic zones against a dark Formazan background. A 7% acetic acid solution was poured onto the plates to stop the reaction. Results in Fig. 2 demonstrate the use of the plate assay to detect SOD activities in strains of bacteria that possess different amounts of the enzyme. Figure 2A shows the differences in superoxide dismutase activity between the parental wild type (sodA+sodB+), the MnSOD-deficient mutant (sodA-), the FeSOD-deficient mutant (sodB-), and the double negative strain (sodA-sodB-). Note that the strains lacking one isozyme displayed less activity than the parental control, whereas very little or no activity was detected in the double-negative mutant. The changes in super-oxide dismutase activity as seen by the plate assay correlated well with the differences seen in superoxide dismutase activity assayed by the cytochrome c method (Table 2). For example, QC773 (sodB-) produced more superoxide dismutase activity than did QC781 (sodA-). The activity seen on the plate assay indicated that the parental strain, GC4468 (sodA+sodB+), produced the most activity, followed by QC773, QC781, and QC779 (sodA-sodB-), respectively. The small amount of activity seen by the double-negative mutant presumably resulted from growing the mutant on LB plates without any antibiotic stress, thus allowing some degree of reversion to occur. Results shown in Fig. 2B demonstrate that the plate assay may be used to detect overproducers of superoxide dismutase as well.
458
SCHIAVONE
FIG. 2. The seven strains described in Table 1 were grown overnight in LB media containing the appropriate antibiotics. Aliquots (1~1) of the cultures were applied to LB plates and incubated overnight. The resulting colonies were used to assaysuperoxide dismutase activity via the plate assay as described under Materials and Methods. (A) Lane 1, E. coli Gc4468; 2, E. coli QC78 1; 3, E. coli QC773; 4, E. coli QC779. (B) Lane 1, E. coli AB2463; 2, E. coli AB2463A+ (pDTl-5); 3, E. coli AB2463B+ (pHS l-4).
The two strains carrying the sodA or sodB genes on a multicopy plasmid (lanes 2 and 3, respectively) exhibited much more activity than the parental control. These differences were also observed when activity was mea-
AND HASSAN
sured by the cytochrome c assay (Table 2). Both the plate assay and the liquid assay illustrated that the strain harboring the multicopy plasmid that contains FeSOD, pHS l-4, exhibited more activity than either the parental wild type or the strain containing the MnSOD plasmid, pDTl-5. These results confirm previous findings (8,9) which show that overproducers of FeSOD lack the regulatory control found in overproducers of the Mn enzyme. Note, however, that AB2463Bf exhibited greater activity than the parental control in the plate assay even though the distinction was not representative of the fold induction seen via the liquid assay. This was due to the fact that the NBT stain cannot detect a linear increase in activity beyond a certain concentration. Sensitivity of the plate assay. Figure 3 illustrates the sensitivity of the NBT plate assay with different concentrations of pure MnSOD. Each protein concentration was delivered in a lo-~1 sample volume. This assay method was sensitive to 0.05-hg quantities of pure superoxide dismutase. This degree of sensitivity was slightly less than the sensitivity of the NBT staining method applicable to polyacrylamide gels which may detect as little as 0.0 16 pg of bovine erythrocyte superoxide dismutase (13). The decrease in sensitivity was attributed to the fact that the plate assay does not confine the enzyme to a small, sharp zone as is the case when the enzyme is electrophoresed on polyacrylamide gels and then stained for activity. The achromatic zones on the plates became larger and clearer as the concentration of protein increased as commonly observed in polyacrylamide gels stained for superoxide dismutase activity. The plate assay did have an upper limit, however, for a point was reached where a certain enzyme concentration yielded a sharp achromatic zone that did not become any sharper or clearer as the concentration of enzyme was increased (data not shown).
SUPEROXIDE
DISMUTASE
DISCUSSION
Efforts to isolate and identify superoxide dismutase mutants have been hampered because the enzyme does not impart readily identifiable traits such as auxotrophy, and the exact phenotype expressed by such mutations is not fully understood. Carlioz and Touati (17) generated and isolated E. coli mutants deficient in Mn-, Fe-, or both isozymic forms of superoxide dismutase. Their method of choice for the detection of these mutants involves immunoblotting analysis of crude extracts followed by a confirmation via the cytochrome c assay. Even though effective, the selection of super-oxide dismutase mutants based on their protocol can be expensive and time consuming. The plate assay that we have developed incorporates the principle of detecting superoxide dismutase activity via the NBT method with the lysis of whole cells from individual colonies and subsequent leakage of cellular enzymes into an agar matrix. This procedure allows one to rapidly screen many colonies simultaneously for superoxide dismutase activity. As illustrated in Figs. 2 and 3, the assay can discriminate between mu-
PLATE
459
ASSAY
tants lacking superoxide dismutase activity and mutants which overproduce the enzyme. This procedure does not require the lengthy preparation of cell-free extracts, the preparation of antibodies specific for one isozymic form of superoxide dismutase, or expensive laboratory equipment. Another positive asset to the assay is that the physiological state of the cells is not an important factor for the detection of superoxide dismutase mutants, as it is when the selection procedure is based on paraquat resistance (18) or oxygen sensitivity ( 17). The plate assay may also be modified to differentiate between the isozymic forms of superoxide dismutase in E. coli. The assay may be used to isolate mutants in FeSOD biosynthesis by simply growing the cells anaerobically prior to lysis. The Mn enzyme is not synthesized anaerobically, so any activity present is due to the FeSOD. The isolation of MnSOD-deficient mutants can be done easily in a FeSOD-negative background by growing the cells aerobically and then screening for superoxide dismutase activity. The isolation of MnSOD-deficient mutants in a FeSOD-positive background is not as
TABLE 2 SUPEROXIDE
DISMUTASE
ACTIVITY
IN VARIOUS
STRAINS OF Escherichia
coli
K12
Superoxide dismutase (U/mg) Strain
Phenotype
GC4468 QC78 I QC773 QC779
Wild type sodAsodBsodA-sodB-
AB2463 AB2463A+ AB2463B+
Wild type sodA++ sodB++
Total
Mn-
HY-
Fe-
25.0 5.6 20.5 1.2
20.2 0.0 20.5 1.2
2.1 1.3 0.0 0.0
2.7 4.4 0.0 0.0
30.8 41.7 244.1
21.8 37.1 104.7
3.3 3.1 31.5
5.7 1.5 107.9
Note. The strains outlined in Table 1 were grown overnight in LB media with the appropriate antibiotics. Approximately 200 ~1 of each overnight culture was applied to LB plates without antibiotics present. The bacterial lawns were scraped off each plate and resuspended in 0.05 M phosphate buffer containing 0.1 mM EDTA. Cell-free extracts were prepared and superoxide dismutase activity measured as described under Materials and Methods.
460
SCHIAVONE
FIG. 3. Wells (6 mm diameter) were bored into the bottom agar surface and the plugs were removed. The bottom of the wells were subsequently sealed with a few microliters of molten agar. Serial dilutions of manganese guperoxide dismutase from Escherichia coli were ap plied to each well in a total volume of 10 ~1. The enzyme was allowed to diffuse into the agar layer for 3 h before the NBT overlay was applied. The staining procedure that followed was as described under Materials and Methods. The following amounts of superoxide dismutase (ag) were applied to the wells: (1) 0.05, (2) 0.10, (3) 0.50, (4) 1, (5) 5, (6) 10, (7) 25, and (8) 50.
AND HASSAN
biosynthesis, or work in a catalasedeficient background. The SOD assay described in this paper or a modification thereof may be applicable to detecting SOD activities in other prokaryotic organisms or eukaryotic systems such as yeast. The principle of lysing whole cells onto a solid support may also be applied for the detection of other enzymes that do not exhibit readily identifiable phenotypic markers but for which a chromogenic assay is known. The plate assay is presently being used in our laboratory to isolate regulatory mutants of superoxide dismutase in E. coli. The current model for the regulation of MnSOD biosynthesis in E. coli suggests that a negative control operon exists whereby the activity of an iron-containing repressor protein is sensitive to changes in the redox state of the cell (21; J. R. Schiavone and H. H. Hassan, manuscript submitted for publication). The plate assay will allow us to further elucidate the mechanism of regulation by providing the means to screen a large population of cells for mutants derepressed for superoxide dismutase biosynthesis. ACKNOWLEDGMENT
straightforward. Asada et al. (19) showed that the FeSOD is sensitive to hydrogen peroxide, and the isozyme may be inactivated in a polyacrylamide gel by adding hydrogen peroxide to the staining solution; but polyacrylamide gels treated with hydrogen peroxide and then stained for superoxide dismutase show achromatic zones that comigrate with catalase (20). This anomaly can be overlooked in a polyacrylamide gel because the catalase and FeSOD are separated during electrophoresis. This is not the case, however, in the plate/colony assay. The catalase and FeSOD are not separated so that any hydrogen peroxide treatment results in the production of achromatic zones and oxygen bubbles in the agar. One must take precautions to inactivate any catalases present, alter the growth conditions to minimize catalase
We thank the late J. S. Siwecki for his generous gift of strains AB2463, AB2463A+, and AB2463B+ and D. Tovati for her generous gift of strains QC78 1, QC773, and Qc779.
REFERENCES 1. McCord, J. M., and Fridovich, I. (1969) J. Biol. Chem. 244,6049-6055. 2. Hassan, H. M., and Fridovich, I. ( 1977) J. Bucteriol. 129, 1574-1583. 3. Hassan, H. M., and Ftidovich, I. (1977) J. Biol. Chem. 252,7667-7672. 4. Hassan, H. M., and Fridovich, I. (1979) Arch. Biothem. Biophys. 1%,385-395. 5. Hassan, H. M., and Moody, C. S. (1984) FEMS Microbial. L&t. 25,233-236. 6. Farr, S. B., D’Ari, R., and Touati, D. (1986) Proc. Natl. Acad. Sci. USA 83,8268-8272. 7. Howard-Flanders, P., and Boyce, R. P. ( 1966) Radiat. Rex 6(Suppl.), 156- 184. 8. Touati, D. (1983) J. Bucteriol. 155, 1078-1087. 9. Sakamoto, H., and Touati, D. (1984) J. Bucteriol 159,418-420.
SUPEROXIDE
DISMUTASE
10. Moody, C. S., and Hassan, H. M. (1982) Proc. NutI. Acad. Sci. USA 79,2855-2859. 11. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275.
12. Davis, B. J. (1964) Ann. N.Y. Acad. Sci. 121,
ASSAY
461
16. Hassan, H. M., and Fridovich, I. (1979) J. Biol. Chem. 254, 10846-10852. 17. Carlioz, A., and Touati, D. (1986) EMBO J. 5, 623-630.
18. Bloch, C. A., and Ausebel, F. M. (1986) .I. Bucferiol. 168,795-798.
404-421.
13. Beauchamp, C., and Fridovich, I. (197 1) Anal. Biothem. 44,276-287. 14. Hassan, H. M., and Fridovich, I. (1977) J. Bacterial. 129, 1574-1583. 15. McCord, J. M., Crapo, J. D., and Fridovich, I. ( 1977) in Superoxide and Superoxide Dismutases (Michelson, A. M., McCord, J. M., and Fridovich, I., Eds.), pp. I l- 17, Academic Press, New York.
PLATE
19. Asada, K., Yoshikawa, K., Takahashi, M., Maeda, Y., and Enmanji, K. (1975) J. Biol. Chem. 250, 2801-2807. 20. Clare, D. A., Duong, M. N., Darr, D., Archibald, F., and Fridovich, I. (1984) Anal. Biochem. 140, 532-537. 21.
Moody, C. S., and Hassan, H. M. (1984) J. Eiol. Chem. 259, 12821-12825.