JOURNAL OF BACTERIOLOGY, Nov., 1966
Copyright © 1966 American Society for Microbiology

Vol, 92, No. 5
Printed in U.S.A,

Fluorescent Method for the Detection of Excreted
Ribonuclease Around Bacterial Colonies

JANOS K. LANYI! anp JOSHUA LEDERBERG
Department of Genetics, Stanford University School of Medicine, Palo Alto, California

Received for publication 21 July 1966

ABSTRACT

Lanyl, Janos K. (Stanford University School of Medicine, Palo Alto, Calif.),
AND JosHUA LEDERBERG. Fluorescent method for the detection of excreted ribo-
nuclease around bacterial colonies. J. Bacteriol. 92:1469-1472. 1966.—A test for the
release of extracellular ribonuclease by Bacillus subtilis colonies was developed. The
method consists of incorporating acridine orange and ribonucleic acid into nutrient
agar plates and viewing the grown bacterial colonies under ultraviolet light. Regions
of ribonuclease secretion appear as dark halos around the colonies on a green fluores-
cent background. The theoretical basis and the utility of this test are discussed.

 

Nucleases are released into the growth medium
(3, 4, 9, 11, 16) by many species of bacteria. The
genetic study of these extracellular enzymes has
been hindered by the limitations of existing
methods either for detecting deviant colonies on
agar plates or for observing the time of enzyme
excretion during growth. In one method (5), the
bacteria were grown on agar plates containing
large amounts of nucleic acid; the areas of nu-
clease action around the colonies were detected,
upon flooding the plates with hydrochloric acid,
as clear halos amidst the precipitated nucleic
acid. The procedure kills the bacteria. This hin-
drance applies also to another method (1) where
a phosphate reagent is used in combination with
trichloroacetic acid. Another technique for the
detection of deoxyribonuclease (15) is based on a
color difference between toluidine blue and its
complex with deoxyribonucleic acid (DNA).

It has been our intention to make use of the
high sensitivity of fluorescence assay in extra-
cellular enzyme detection. Our efforts were di-
rected toward the extracellular ribonuclease of
Bacillus subtilis Marburg strain (6, 9). According
to previous reports, when a fluorescent dye, such
as acridine orange, is combined with a large
excess of polymeric ribonucleic acid (RNA) a
complex is formed (2, 13) in which the dye
molecules are intercalated between successive
layers of the base pairs (8). This interaction
enhances the green fluorescence of the dye. We
expected that on depolymerization by nuclease
the complex would be destroyed and the fluo-

1Present address: Biological Adaptation Branch,
Ames Research Center, Moffett Field, Calif.

rescence would be decreased. The zones of nu-
clease action around the bacterial colonies would
be detected in this way as dark halos on a green
fluorescent background.

EXPERIMENTAL

Acridine orange is known to have an emission
peak at 540 mu which is enhanced up to three-
fold by adding a large excess of nucleic acid
polymers (2, 13). This effect was explored by
taking the fluorescence spectrum of a solution
of acridine orange and RNA with a Zeiss RMQ
spectrofluorometer. The exciting light used was
broad-band ultraviolet, and the spectrum of the
emitted light was recorded manually at the highest
sensitivity setting. The concentration of acridine
orange was 2.5 <x 10-4 m; that of the RNA
(Calbiochem) was 2.5 x 10°°m (in nucleotide)
at pH 7.0. The results obtained are shown in
arbitrary fluorescence units versus wavelength in
Fig. 1, where it can be seen that the addition of a
10-fold excess of RNA does enhance fluorescence
without changing the shape of the emission
spectrum. Even though the increase in fluo-
rescence became greater upon manipulating the
relative concentrations of dye and nucleic acid,
this effect would not be great enough by itself
for a visual ribonuclease assay. However, sub-
stituted polysaccharides such as agar interact with
free acridine orange (14) and the fluorescence
of the dye is quenched. The quenching effect
is shown by adding 0.35°% agar (kept in solu-
tion before mixing by holding the temperature
between 40 and 50 C) to a solution of acridine
orange (2 x 10-4Mm) and recording the emission

1469
1470

 

T J

] I
0 2.5x104M Acridine + 1mg/mi RNA

sok 8 25 x104M Acridine |

Fluorescence units

 

 

 

| | | i |

Q
500 $20 540 560 580

Emission wavelength mu

600

® Fic. 1. Fluorescence of acridine orange in the pres-
ence and absence of RNA.

spectrum. The results (Fig. 2) indicate that in
the presence of agar the fluorescence is almost
completely eliminated.

The mechanism of this quenching is not under-
stood. To explore the possibility that an elec-
trolyte-type binding between the dye and agar
plays a part, a solution of acridine orange
(2 x 107!) was titrated by agar to the point
of quenching. In Fig. 3, the extent of quenching
is plotted as a function of the concentration of
agar. It is seen that quenching is a linear function
of the dose of agar up to a point of saturation.
For calculating the stoichiometry of the inter-
action, one needs to know the number of ioniz-
able groups in agar and on the acridine orange
molecule. It has been reported that agar contains
one sulfate ester group for approximately 53
galactose units (10). Acridine orange, on the
other hand, has three amino groups per mole-
cule. The equivalence point in this case, based
on the assumption that all three amino groups
are charged and can bind the sulfate groups, is
at about 0.0756, agar. The curve in Fig. 3 in-
dicates that this is indeed the agar concentration
where quenching approaches a maximum. A
quenching curve for acridine yellow is also given,
with similar results.

It should be noted that, even though the de-
pendence of quenching on agar concentration is
linear, it does not represent a 1:1 relationship;
the slope varies for different dyes. No explanation
jis advanced for this discrepancy at this time.
Nevertheless, it is tentatively concluded that the
quenching of fluorescence by agar is a result of
electromeric interaction between the dye and
the agar.

Complexes of acridine orange with agar are

LANY! AND LEDERBERG

J. BACTERIOL.

weaker than with nucleic acid, and the latter
therefore reverses the quenching. To show this,
a 10-fold excess of RNA was added to the acri-
dine orange-agar complex, and the fluorescence
spectrum was recorded. The results, shown in
Fig. 2, bear out the expectation; it can be seen
that fluorescence is fully restored by addition of
RNA. This increase in the fluorescence of the
acridine orange, due to the presence of RNA, is
quite dramatic, and proved suitable for visual
detection.

Preliminary to the agar-plate experiments,
the RNA-acridine orange-agar complex was
tested as an assay system for ribonuclease in
liquid state. The assay mixture consisted of 1.0
ml of acridine orange-RNA (10-*m dye, 107?M
RNA neutralized) and 3.0 ml of agar solution
(0.05¢%). This mixture exhibited a fluorescence
of 35.8 arbitrary units as compared with a control

 

 

 

 

I T T T T
02.5x104M Acridine + \mg/ml RNA
@2.5x104M Acridine
30/7 7
#2
€
a
8
€ 20 4
8
z
9
c
9
35
in 10 -
0 L 4 L J
500 520 540 560 580 600

Emission wavelength ma

Fic. 2. Fluorescence of acridine orange in 0.35°¢
agar solution in the presence and absence of RNA.

eO

    
 
  
 

Acridine, orange
($45 Mu)

Equivalence point
(3 amino groups reactive)

,

Acridine yellow
(S15 ha)

Percent quenching
o
T

 

 

1
0! 0.0f 0) \
Percent Agar

 

Fic. 3. Quenching of acridine orange and acridine
yellow fluorescence by agar (as determined at the indi-
cated wavelengths).
VOL. 92, 1966

where distilled water was substituted in place of
the agar solution, and where the fluorescence
was 37.8 units. In two subsequent experiments,
10 and 20 uliters of pancreatic ribonuclease
solution [2.0 mg/ml in tris(hydroxymethyl)-
aminomethane-HCl buffer, pH, 5 Worthington
Biochemical Co., Freehold, NJ.] was added,
and the change in fluorescence was observed.
The solution was kept at room temperature.
The two kinetic curves obtained are shown in
Fig. 4. It is clear from this graph that the amounts
of enzyme and the slopes observed for the de-
crease in fluorescence are not strictly propor-
tional. However, as our intention was merely to
develop a semiquantitative ribonuclease assay,
the approximate correlation obtained was con-
sidered satisfactory. It has been reported that
deoxyribonuclease is competitively inhibited by
acridine orange (7). However, as the concentra-
tion of the dye needs to be severalfold higher
for 50°; reduction in activity than that used
here, this effect was not considered important.

Agar plates for extracellular ribonuclease
detection were made up in the conventional
way with the following exception. Before auto-
claving, 1 ml of stock solution (1 mg/ml of acri-
dine orange and 15 mg/ml of RNA, neutralized
with 1 SX NaOH) was added to 100 ml of medium.

 

no enzyme

 

2x enzyme

20}— J

[xXenzyme

Percent Fluorescence remaining
8

DETECTION OF EXTRACELLULAR RIBONUCLEASE

 

 

{ { \ l

 

0 2 4 6 8 10
Time in minutes
Fic. 4. Decrease in the fluorescence of the acridine
orange-RNA-agar complex on addition of pancreatic
ribonuclease.

1471

After autoclaving for 20 to 25 min at 120C, the
medium was left to cool at 50C for at least 2
hr. This latter period was necessary for obtaining
maximal fluorescence. Upon solidifying, the
plates were viewed with an ultraviolet lamp
(Mineralite) in the dark; intense green fluo-
rescence was observed.

B. subtilis colonies grew normally on these
plates; cells picked from the surface of the agar
proved viable. The action of various B. subtilis
strains upon the acridine orange-RNA plates
was then studied. The full details of obtaining
these strains are being published elsewhere
(Lanyi and Lederberg, in preparation); suffice it
to say here that they were produced by ultra-
violet irradiation of the B. subtilis Marburg
strain and subsequent plating on fluorescent
plates. The objective of this study was to establish
a correlation between the size of dark (non-
fluorescent) halos around the colonies and an-
other, more quantitative, assay of ribonuclease
activity performed on the culture filtrates of the
same strains. The halo-forming property of
four strains of B. subtilis is shown in Fig. 5. The
agar medium consisted of a minimal medium
described by Spizizen (12): 1% agar, 0.05¢% acid
casein hydrolysate (neutralized; Nutritional Bio-
chemicals Corp., Cleveland, Ohio), and 20 ug/ml
of L-tryptophan. The photographs in Fig. 5,
taken 18 to 20 hr after streaking, show that,
on this agar medium, the size of the halos

 

Fic. 5. Halo production by several strains of Bacillus
subtilis. The agar plate was made and the bacterial col-
onies were grown according to the procedure described
in the text. The plate was photographed in ultraviolet
light with a yellow-green filter. Strains (clockwise,
starting at top): SB 168, SB 764, SB 765, SB 29.
1472

TABLE 1. Halo formation and ribonuclease
secretion by some strains of
Bacillus subtilis

 

Strain Halo size a counts hydvovived
ns oO mmoles
SB 168....... +++ ! 1,817 20.6
SB 29. ++ 803 9.1
SB 764....... — + 319 3.6
SB 765....... | +44 1,428 16,2

 

varies with the strains; such differences were
found to be reproducible.

The strains used above were also grown in a
liquid medium similar to that described above,
containing the following: inorganic salts (12),
0.8% yeast nitrogen base (without amino acids;
Difco), 0.1% acid casein hydrolysate (neutra-
lized; Nutritional Biochemicals Corp.), 4%
soluble starch, and 20 yg/ml of L-tryptophan.
The cultures were grown at 30C with shaking.
At 124 hr., the samples were withdrawn and
centrifuged; the supernatant fluid was assayed
for ribonuclease with tritiated RNA as substrate.
The assay mixture contained 20 yliters of H*-
uridine RNA (6.4 x 104 counts per min) in
0.2 ml of 0.05 mM glycine-NaOH buffer (pH 9.55),
to which 0.1 mi of culture supernatant fluid was
added. After incubation at 37 C for 30 min, the
samples were chilled, and 0.3 ml of high molecular
weight bacterial RNA (1.25 mg/ml) was added.
The unhydrolyzed portion of the nucleic acid was
precipitated at 0C by adding 1.0 ml of 7% tri-
chloroacetic acid; the soluble fraction was col-
lected after centrifugation, and 50 yp liters was
counted in a Packard Tricarb scintillation counter.
Zero-time digests gave 400 to 500 counts per 5
min; thees values were subtracted from the ex-
perimental data. The results obtained in these
assays are given in Table 1. Also given in Table 1
is the visual estimate of the size of the halos for
the various strains. There was a good correlation
between the results of the fluorescence assay and
the results obtained through the more conven-
tional assay described above. It was concluded,
therefore, that the fluorescent plate technique can
be used successfully in detecting differences in
the levels of extracellular ribonuclease production.
The method was found to have the advantage
that large numbers of colonies (up to 100 to 150
per plate) can be examined at a glance, and mu-
tants can be easily recognized and isolated.

ACKNOWLEDGMENTS

This investigation was supported by Public Health
Service training grant 2T1-GM-295 and research grant

LANYI AND LEDERBERG

J. BACTERIOL.

AI-5160 from National Institutes of Health and by re-
search grant G-6411 from the National Science Foun-
dation.

LITERATURE CITED

1. Berry, S. A., AND J. N. CAMPBELL. 1965. A
diffusion plate technique for the detection of
polynucleotide depolymerase and phosphatase
activity. Biochim. Biophys. Acta 99:566-569.

2. BRADLEY, D. F., AND M. K. WoLF. 1959. Aggrega-
tion of dyes bound to polyanions. Proc. Natl.
Acad. Sci. U.S. 45:944-952.

3. CUNNINGHAM, L., B. W. CaTLIN, AND M. P.
DEGARILHE, 1956. A deoxyribonuclease of
Micrococcus pyogenes. J. Am. Chem. Soc. 78:
4642-4645.

4. Eaves, G. N., AND C. D. JeFFrigs. 1963. Isolation
and properties of an exocellular nuclease of
Serratia marcescens. J. Bacteriol. 85:273-278.

5. JEFFRIES, C. D., D. F. HOLTMAN, AND D. G, GUSE.
1957. Rapid method for determining the ac-
tivity of microorganisms on nucleic acids. J.
Bacteriol. 73:590-591.

6. Kerr, I. M., E. A. Pratt, AND 1. R. LEHMAN.
1965. Exonucleolytic degradation of high-molec-
ular-weight DNA and RNA to nucleoside 3’-
phosphates by a nuclease from B. subtilis. Bio-
chem. Biophys. Res. Commun. 20:154-162.

7. Lerty, J. D., Jr. 1963. Acridine orange and acri-
flavin inhibit deoxyribonuciease action. Bio-
chim. Biophys. Acta 72:643-644.

8. Lerman, L. S. 1963. The structure of the DNA-
acridine complex. Proc. Natl. Acad. Sci. USS.
49 94-102.

9, Nakat, M., Z. MiNaMI, T. YAMAZAKI, AND A.
Tsucira. "1965, Studies on the nucleases of a
strain of Bacillus subtilis. J. Biochem. (Tokyo)
57:96-99.

10. PicMan, W. [ed.] 1957. The carbohydrates. Aca-
demic Press, Inc., New York.

11. ROTHBERG, N. W., AND M. N. Swartz. 1965.
Extracellular deoxyribonucleases in members
of the family Enterobacteriaceae. J. Bacteriol.
90:294-295.

12. Spizizen, J. 1958. Transformation of biochemi-
cally deficient strains of Bacillus subtilis by
deoxyribonucleate. Proc. Natl. Acad. Sci. U.S.
44:1072-1078.

13. Steiner, R. F., AND R. F. BEERS. 1961. Poly-
nucleotides, p. 290-320. Elsevier Publishing Co.,
New York.

14. Stone, A. L., L. G. Cuivpers, and D. F.
BRADLEY. 1963. Investigation of structural
aspects and classification of plant sulfated poly-
saccharides on the basis of the optical proper-
ties of their complexes with metachromatic dyes.
Biopolymers 1:111-131.

15. STREITFELD, M. M., E. M. HoFeMAN, AND H. M.
JANKLOW. 1962. Evaluation of ‘extracellular
deoxyribonuclease activity in Pseudomonas. J.
Bacteriol. 84:77--80.

16. WANNAMAKER, L, W. 1958. The differentiation of
three distinct desoxyribonucleases of Group A
Streptococci. J. Exptl. Med. 107:797-812.