Note: Descriptions are shown in the official language in which they were submitted.
A need exists for a method of rapidly detecting
bacteria in fluids from many sources. Of particular significance
is the need for rapid detection of pathogenic bacteria in
physiological fluid specimens, such as blood, urine and the
like. Moreover, a need exists for a method for rapidly deter-
mining the susceptibility of such infecting bacteria.
Urine specimens in general form the major part of
the work load of the diagnostic microbiology laboratory. By far
the most common urological disease is urinary tract infection.
In fact, in many hospitals, bacteriuria is the most common form
of nosocomial infection, often following the use of in-dwelling
catheters and various surgical procedures. The volume of specimens
requiring bacteriuria screening is further increased by the need
to repeat the tests to insure accurate diagnosis where their
reliability may have been reduced due to contamination of the
specimen during collection. A further problem with diagnosis
and treatment of bacteriuria is the frequent lack of correlation
between a patient's symptomatic response to antimicrobial
treatment and successful treatment. In order to insure that
the prescribed antimicrobial agent is in fact effective,
repeated tests during therapy are required. The need for simple,
rapid bacteriuria tests is thus clear. Moreover, in view of the
frequent unsuspected asymptomatic occurrences of urinary tract
infections among children, pregnant women, diabetics and
geriatric populations, diagnosis of which may require collection
and testing of several specimens, bacteriuria tests must be
sufficiently simple and economical to permit routine performance.
A need thus exists for rapid, inexpensive screening tests to
facilitate diagnosis and insure proper treatment of urinary
tract infections.
--1--
f~ 7
Rapid tests for detection of bacteria in blood are
also needed, in view of the high mortality rate associated with
septicemia and bacteremia. Prompt detection of the disease
permits early administration of an appropriate antibiotic thus
greatly improving the chances for survival.
According to conventional techniques, bacterial
infections in specimens, such as blood, urine, spinal fluid and
the like, are detected by diluting a specimen with culture media
and incubating the diluted specimen at 36C. The appearance of
turbidity manifests bacterial growth. However, relatively
extended periods of incubation are required since turbidity due
to bacterial growth is difficult to distinguish from turbidity
due to the presence of blood cells or contaminants in the specimen
and from turbidity caused by precipitate formation. Substantial
increases in turbidity following incubation periods of about
24 hours indicate bacterial growth.
Another very important procedure in the clinical
laboratory is determination of antimicrobial susceptibilities.
The principal methods presently employed to determine suscep-
tibility of a micro-organism to an antibiotic include dilution
tests, such as the broth tube and agar plate procedures, and
agar diffusion tests, utilizing antibiotic-impregnated discs.
Typically, such methods require incubation periods of 16 to 18
hours before the inhibitory effect of an antimicrobial agent
can be accurately assessed. Furthermore, such tests often are
time consuming, relatively expensive and must be performed by
slcilled laboratory personnel.
Although staining techniques are known in clinical
microbiology, such techniques are t~-pically employed to stain
dried bacterial smears on slides rather than in fluid specimens.
In the practice of such prior art staining techniques, a dried
bacterial smear on a slide is treated with a reagent which
stains the bacteria in a manner which permits read,~ microscopic
examination thereof. Thus, expenslve equipment and skilled
microbiologists are required to perform such analyses.
In addition to bacterial examination of body fluids,
it is often necessary to analyze the bacterial content of other
fluid specimens, such as water and pharmaceutical products.
The need for rapid, simple, inexpensive and accurate methods
for detecting and analyzing bacteria in body fluids and other
fluid specimens is thus evident.
It has now unexpectedly been discovered that both
gram-negative and gram-positive living bacteria can be stained
for simple, rapid analysis by means of the composition of the
present invention. Concentrated bacteria stained with the
composition are readily visible and can thus be rapidly detected
without resort to microscopic examination or specially trained
personnel. Moreover, antimicrobial susceptibility of bacteria
can be determined rapidly and simply by means of the present
invention. Further, it was unexpectedly found that inexpensive,
simple and rapid quantitative analyses of bacteria are possible
employing the present staining composition. Finally by means
of the present invention, it is possible to differentiate gram-
negative and gram-positive bacteria.
A composition for staining both gram-negative and ~ -
gram-positive bacteria is provided. The composition comprises
- a chelating agent operative in the basic pH range and a basic
dye capable of staining bacteria at a basic pH. Bacteria are
stained when contacted with the composition at a pH above about
7Ø Bacteria which are stained with the composition and
concentrated become readily visible, and may thus be detected.
Semi-quantitative analysis of bacteria may be
accomplished by comparing the gradation of color developed in
concentrated stained bacteria, with a nomograph or other
calibrated standard. Semi-qualitative analysis of the stained
bacteria may be effected by means of an organic acid wash having
a pH of about 2.5 or 2.6, since such an acid wash will completely
decolorize only gram-positiye bacteria stained with the
composltlon .
By incubatin~ bacteria with an antimicrobial agent
prior to staining with the composition of the inyention, the
susceptibility of the bacteria to the agent can be determined.
The relative intensity of the color of stained, concentrated
bacteria, thus treated, will be related to the effectiveness
of the agent employed.
The invention is particularly useful in laboratory
screening of body fluids and other physiological fluid specimens.
This invention relates to compositions useful for
staining both gram-negative and gram-positive bacteria and to
various methods of detectin~ and analyzing bacteria in fluids.
Broadly stated, the staining composition of the invention comprises
a chelating agent and a dye. ~acteria contacted with this
composition at a p~ above about 7.0 are stained and upon con-
centration become readily visible.
Since the color intensity of stained concentrated
bacteria is correlated with the numher of bacteria in a sample,
semi-quantitative a~alysis of bacteria ~ay be accomplished by
co~paring the intensity of the color developed in the stained
concentrated bacteria with a nomograph or other known standard.
When concentration of the bacteria is effected by deposition of
bacteria on a semi-permeable membrane, dye not associated with
the bacteria, which may interfere with an accurate detection
and quantitation of bacterial presence, may be removed by means
of an organic acid wash having a pH in the range of about 2.7 to
4Ø If bacteria are incubated with an antimicrobial agent for a
brief period prior to contact with the staining composition, the
susceptibility of the bacteria to the agent is determined by
comparing the color intensity of the stained, concentrated baeteria
with a control. Differentiation of the gram-stain of baeteria
may be effected by treating the stained bacteria with an organic
acid wash having a pH of about 2.5-2.6. Gram-positive bacteria
are completely decolorized by such a wash whereas stained gram-
negative bacteria are not.
The composition and methods of the invention have
particular application to the detection and analysis of bacteria
in physiological fluid specimens, particularly urine specimens.
By means of the instant invention, rapid and economical detection
; and treatment of bacterial infection is possible.
More particularly, it has been discovered that the
combination of a chelating agent, operative in the basic p~I
range, and a dye, capable of staining bacteria at a pH above
about 7.0, results in a eomposition having the eapaeity to stain
both gram-negative and gram-positive baeteria. In the absenee
of the ehelating agent, dyes, partieularly basie dyes, fail
to stain gram-negative baeteria. Bacteria may be stained
simply by contacting either concentrated or fluidly suspended
baeteria with the ehelating agent/dye eomposition at a nearly
neutral or basie pH.
Any dye eapable of staining baeteria at a basie or
neutral pH may be employed in the eomposition and method for
staining bacteria described herein. Since the staining
operation is effected at a pH of about 7 or higher, the dyes
used must be operative in this pH range. As a general rule,
basic or cationic dyes are effective bacteria stains in the
practice of the present invention. Specifically, Safranin-O,
toluidine blue, methylene blue, crystal violet and neutral red
may be utilized in the present invention, with Safranin-O being
particularly preferred.
The chelating agents which may be employed in the
practice of the present invention are also limited to those
which are operative at the pH at which the staining is effected,
that is, about 7.0 or higher. Salts of ethylenediaminetetraacetic
acid (EDTA) and citric acid may be utilized. In particular
various sodium salts of these two acids are effective, specifically
the di- and tetrasodium salts of EDTA and the di- and trisodium
salts of citric acid. Tetrasodium EDTA is a particularly
preferred chelating agent.
The amounts of chelating agent and dye necessary to
effectively stain bacteria range from about 0.001 to about 0.1
molar (M) chelating agent and 1:1000 to 1:300,000 dilution of
dye. These amounts are calculated as final concentrations, taking
into account any dilution due to the material in which the
bacteria may be present.
The specific concentration of dye and chelating agent
utilized may be dependent in part upon the condition of the
bacteria when contacted with the staining composition. For
example, where the staining is effected on bacteria which are
relatively concentrated or free of interfering substances,
competing chemical or physical reactions will as a rule be
reduced and more concentrated compositions may be employed. On
i7
the other hand, where the bacteria are dispersed in a fluid
medium containing other materials, it may become necessary
to adjust the concentration of dye and/or chelating agent
upward or downward to compensate for reactions with these
additional materials. For example, in urine specimens, reduced
concentrations of dye should be used to avoid formation of
precipitates with urine compounds which occurs at 1:1000 dye
dilution. In general, dye dilution on the order of 1:2500 or
more is adequate to avoid such precipitate formation, but
dilution of 1:10,000 or more is preferred. In general,
particularly effective bacteria staining can be accomplished
e~ploying compositions comprising about 0.05 M chelating agent
and 1:1000 or higher dye dilution with relatively pure or
concentrated bacteria or 0.05 M chelating agent and 1:10,000
or higher dye dilution where the bacteria is fluidized with
interfering materials.
In practice, the staining composition may be stored
in concentrated form. For example, sterile Safranin-O EDTA
could be stored at the following concentrations: Safranin,
1:1000; EDTA Na4, 0.5 M. At the time of use, this mixture
could be diluted to the desired concentration. For example, 1 ml
could be added to 9 ml of test material to obtain a final
concentration of 1;10,000 Safranin and 0.05 M EDTA. The
storage stability of the staining composition is increased when
the dye used to make the composition has been solubilized in
undiluted organic media.
As indicated above, the composition is effective to
stain both concentrated bacteria and fluidly suspended bacteria.
Staining of bacteria is accomplished simply by adding the
composition to a fluid specimen believed to contain bacteria
or by contacting a solution of the composition with concentrated
bacteria. Thus, for example, bacteria in physiological fluid
specimens may be stained by simply adding the composition to
the specimen. Alternatively, the bacteria might first be
deposited on a semi-permeable membrane. Thereafter, staining
of the bacteria could be effected by pouring a solution of
the composition through the membrane.
The degree of staining is somewhat dependent upon
concentration of dye and time of contact. With higher con-
centrations, the period of contact may be reduced; converselywith lower concentrations of dye, increased holding times are
required. Further, the time of contact is inversely related
to the temperature at which the contact is effected. For example,
optimal staining of bacteria in fluid specimens with a dye
dilution of 1:5000 requires holding times of 45 minutes at 4C,
15 minutes at 25C, 5 minutes at 37C and 1 to 2 minutes at
50C. In general, at least 15 minutes at room temperature is
required to obtain maximum staining of bacteria in urine
specimens; after 30 minutes, no further staining is observed.
However, if the bacteria is concentrated on semi-permeable
membranes prior to staining, periods of as little as 15-60 seconds
are required, since staining compositions having a 1:1000 dye
dilution may be employed.
Bacteria stained in accordance with the present
invention may be readily detected if concentrated. When the
staining has been effected on concentrated bacteria in a
manner which does not result in the bacteria becomin~ fluidized,
the presence of bacteria is immediately manifested. Stained
bacteria which are fluidly dispersed will, upon concentration,
become readily visible.
The concentration of bacteria which can be detected
by this staining procedure varies somewhat with the type of
bacterium, but in general gram-negative bacteria can be detected
at levels of 105 CFU/ml, whereas detection of gram-positive
bacteria may require accumulation of 106 CFU/ml. Of course,
smaller concentrations of bacteria can be detected by con-
centrating larger quantities of fluid.
Sedimentation and filtration are examples of effective
means for concentrating bacteria. When sedimentation is
employed, bacteria present in the specimen will be manifested
by a precipitate having the color of the dye employed. With
filtration techniques, bacteria are deposited on semi-permeable
membranes whereupon their presence is evidenced by the color
of the dye developing on the membrane.
Conventional procedures, such as centrifugation may
be employed to effect sedimentation. For example, bacteria in
a 100 ml physiological fluid specimen could be sedimented at
3000 rpm for 15-30 minutes in a conventional chemical centrifuge,
after being contacted with the composition of the invention. A
pellet in the tube havin~ the color of the dye used indicates
the presence of bacteria.
- Where the present invention is practiced utilizing
filtration techniques, a semi-permeable membrane having a
pore size sufficient to retain bacteria is required. In general,
membranes having a pore dia~.eter of about 0.2 to 1.0 Am may be
employed. The membrane may contain conventional materials,
including fiberglass, epoxy, nitrocellulose, cell-~lose acetate,
asbestos or combinations thereof. Preferred are epoxy-fiberglass
filters having good flow rates and a depth such that clogging is
minimized. Flow rates and depth of a membrane are of particular
r;~
importance when dealing with very turbid specimens, such as urine.
Further, it is preferred that the membrane employed
not retain substantial amounts of dye which are not associated
with bacteria. Retention of free dye by the membrane is
preferably sufficiently low to permit differentiating the
color developed on the membrane when only free dye is present
and that developed when stained bacteria is additionally present.
In general, membranes which do not have a net negative
electrostatic surface charge must be employed. The relative
suitability of membranes can be evaluated by simply passing the
appropriate concentration of the dye being used through the
various membranes and comparing the intensity of color developed.
Preferred membranes are those which adsorb minimal
amounts or no free dye. However, if the dye/membrane combination
is such that free dye partially colors the membrane surface, the
presence of stained bacteria can be accurately detected in
accordance with the present invention simply by subtracting the
intensity of this partial coloration as background. To insure
accurate detection of bacteria, the color developed on the
membrane during a control run utilizing a fluid specimen
containing a staining composition, but no bacteria could be
compared with the color developed in a run utilizing a similar
specimen containing the staining composition and a pathological
amount of bacteria.
Some membranes which are colorized by free dye can
be decolorized partially or completely by means of an organic
acid wash. By contacting organic acids with some membranes,
the surfaces of which are colorized by the staining composition
in the absence of bacteria, at least part of the free dye adsorbed
by the membrane surface can be removed without removing dye
--10--
associated wi-th bacteria deposited on the membrane. Such an
acid wash, thus, reduces the amount of ree dye retained by
some membranes and thereby improves the accuracy with which
bacteria may be detected on such-membranes.
The degree of decolorization effected by an acid wash
will depend on a number of factors, includin~ nature of the
membrane, particular acid used, pH of the acid and material
in which the dye is solubilized. The color developed due to free
dye on membranes containin~ various materials, including fiber-
glass, nitrocellulose, cellulose acetate, asbestos and epoxy,may generally be removed to some degree by an organic acid wash.
Organic acids, including citric and acetic acid, are generally
effective to remove free dye on me~brane surfaces, without
removal of dye attached to bacteria, if the pH of the acid is
between about 2.7 and 4Ø pX's below about 2.7 should be
avoided since decolorization of stained bacteria may also occur.
Acetic acid at a pH of about 3 is a preferred wash.
The degree of attachment of free dye to membranes can
be reduced and removal thereof by an acid wash can be enhanced
if the dye utilized in the staining composition is solubilized
in organic media. Basic dyes dissolved in water or inorganic
salts tend to attach to membranes and are not generally
decolorized by an organic acid wash.
Most effective removal of free dye is accomplished
in those cases where a basic dye has been completely solubilized
in rich or~anic media. The de~ree of solubilization may be
determined by passing a test solution of the basic dye through a
cation exchange resin and thereafter filterin~ the eluent
through a membrane hayin~ a capacity to adsorb the dye. The
de~ree of solubilization will be indicated by the amount of dye
~ G ~
adsorbed on the membrane; where a dye is completely solubilized,
no dye will be evident on the membrane, whereas relatiYely
lesser degrees of solubilization will be indicated by the
relatively increased intensity of the color developed on the
membrane.
Solubilization of hasic dyes can be accomplished
in rich organic media used for culturing bacteria, such as
trypticase soy broth, tryptose phosphate broth, ~lucose or brain-
heart infusion. Preferred media include undiluted trypticase
soy broth, tryptose phosphate broth, brain-heart infusion or
media of similar nature, since not only do such.media minimize
the incidence of false positives, but additionally result in
staining compositions which exhibit a reduced tendency to
precipitate or become turbid oYer time and thus are more storage
stable.
A preferred combination for maximizing removal of
free dye adsorbed by membrane surfaces is as follows: a fiberglass-
epoxy filter having a net positive surface charge and particularly
one having the pore and flow properties of the G-2 series sold
by Finite Filter Corp. (Detroit, ~ichigan), acetic acid at a pH
of about 3 and a basic dye, preferably Safranin-O, solubilized
in undiluted bacteria culturin~ media. Substantially all free
dye on a membrane surface is decolorized when this combination
is employed in the practice of the present inyention.
The decolorizing acid wash may be effected simply
by contacting the colored surface of the.memb.rane with the
acid for a short period and thereafter suctioning or otherwise
removing the wash.from the mem~rane. The optimum time and
number of washes can be determined by simple trial and error
control runs. Typically, with an acid at pH 3, 1 to 3 washes
for a period of less than five minutes each will be suffieient.
The presence of bacteria can be semi-quantitatively
detected employing the staining composition of the invention. Such
a quantitative analysis can be accomplished by simply staining
and concentrating the bacteria as above described. The intensity
of the color of the stained, concentrated bacteria, which
correlates with the bacterial population, can then be compared
with a standard which has been calibrated using known bacterial
amounts. Conventional techniques, such as nomographic,
colorimetric and photometric procedures, may be employed to make
the quantitative analysis. Bacterial growth in fluids may be
measured using the above methodology by comparing the intensity
of bacterial stains developed in samples drawn from the fluid
at different time intervals.
Differentiation of the gram-stain of baeteria may
also be aecomplished employing the staining composition of this
invention. As noted above, organie aeid washes below a pH of
about 2.7 tend to deeolorize stained baeteria as well as free
dye on a membrane surfaee. However, if the pH of the aeid is
maintained at about 2.5 to 2.6, gram-positive bacteria are
totally decolorized; below a pH of about 2.5 both gram-positive
and gram-negative baeteria are deeolorized. It is thus possible
to differentiate gram-negative and gram-positive baeteria. Thus,
by means of an organie aeid wash, of the type used to deeolorize
free dye on a membrane, but having a pH redueed to about 2.5 to
2.6, a semi-qualitative analysis of baeteria stained with the
eomposition of the invention ean be performed.
By means of the present invention, it is also possible
to determine antimierobial suseeptibilities of baeteria. Treat-
ment of baeteria with an antimierobial agent to whieh they are
-13-
'7
susceptible prior to contact with the staining composition will
result in a diminution in number of bacteria. Consequently,
the color of the stained concentratecl bacteria thus treated
will be less intense than that of resistant cultùres or an
untreated control. The reduction in color will be roughly
parallel to the degree of susceptibility to the antimicrobial
agent. Thus, when bacteria are treated with an antimicrobial
agent prior to contact with the staining composition of the
invention, the intensity of the color of the stained, con-
centrated bacteria will be related to the susceptibility of
the bacteria to the agent. Treatment of bacteria with an
antimicrobial agent having a bacteriostatic or bactericidal
effect prior to staining will result in the color intensity
of the stained concentrated bacteria being comparatively less
than that of stained concentrated bacteria which were not
treated with the agent. By comparing the colors developed in
bacteria which have been treated with different antimicrobial
agents or different amounts of a single agent, the relative
inhibitory effects thereof can be evaluated.
Treatment of bacteria with an antimicrobial agent can
be effected simply by contacting either concentrated or fluidly
suspended bacteria with the agent generally for no more than
about 1 to 3 hours. The procedure may be employed with
bacteria in a fluid specimen or with colonies of bacteria from
a culture plate which haYe been suspended in an organic broth.
The amount of agent employed in this procedure will be in
accordance with known standards, such as standardized FDA
approved antimicrobial discs.
If desired, a bacteria sa~ple may be incubated prior
to treatment with antimicrobial agents. Incubation will
-14-
;'7
enhance the accuracy with which susceptibility to the agents
is determined due to the culture reaching log phase of growth.
Since bacteria grow at a rapid rate when incubated at 35-36C,
bacterially infected samples need be incubated for only about 30
minutes to 1 hour to insure highly accurate results. Such incuba-
tion is desirable where the relative inhibitory effects of several
antimicrobial agents having similar activities are being assessed.
The composition and methods of the invention have
particular application to the staining and analysis of bacteria
in physiological fluid specimens. For example, urine, which has
been clarified conventionally, may be treated with a solution
containing 1:10,000 Safranin-O solubilized in nutrient broth and
0.05 M tetrasodium ethylenediaminetetraacetate. The urine is
then passed through a bacteriological filter having a net positive
charge whereupon the stained bacteria are readily visible. If
desired, the filter is then washed with pH 3 acetic or citric
acid.
Alternatively, the urine may be passed through the
bacteriological membrane which results in the deposition of
the bacteria in the urine onto the membrane surface. There-
after, the deposited bacteria are treated with sufficient
1:1000 basic dye - 0.05 M EDTA salt mixture to cover the mem-
brane surface. After 15-60 seconds, or longer if desired, the
dye is drawn through the membrane by suction. If desired, the
membrane may then be washed with p~ 3 acetic acid.
In some instances, urine of patients suffering with
bacteriuria may have precipitates which clog membranes used
in the practice of the present invention. Such urine is first
clarified, for example with a 5~ m clarifier, to remove the
precipitates and enhance filtration of the urine. Occasionally,
-15-
urine may contain gram-positive bacteria in the form of
aggregates which are removed by the 5- ~m clarifier. Without
clarification, such urines would not be able to be processed
by the bacteriuria-detection method of the invention.
In order to increase the flow rates of urine through
the 0.65 ~m filters employed in the present method, the
sediments, such as urates, present in the urine may be
solubilized. Acetic acid is the optimal solvent for this
purpose. Urine specimens mixed with equal volumes of acetic
acid at pH levels of 2.0, 2.5, 3.Q, 3.5 and 4.0 exhibit increased
optical transmission at 5~0 nm only at pH 2.5 or lower. Further,
mixtures of pH 2.5 acetic acid and urine attain a final pH
between 3.5 and 4.5 in most cases and are not deleterious to
the staining reaction of bacteria (i.e., the bacteria retained
their ability to react with safranin).
The acetic acid diluent enhances the flow rates of
urines. In many instances, the staining intensity is greater
in the acetic acid diluted urines than in corresponding specimens
without acetic acid. This is believed to be due to the fact that
suspended solids which are solubilized can no lon~er impact on
the entrapped bacteria on the filter and prevent staining.
Although the acetic acid diluent described above aids
flow rates of urines which contain solids, heavily pigmented
urines containing soluble or~anics often clo~ membranes because
of the adsorption of the pi~ments to the 0.65-~4m filters. Among
anionic exchangers which remoye urine pi~ments, Exchan~er A109-D
(Diamond Shamrock, Cl char~ed) is the resin of choice since
it renders the urine almost colorless. Flow rates of urines
through 0.65- ~m filters are dramatically increased if the urine
is first passed through the anionic resin. Urines may be
-16-
processed employing such a resin as follows: ~ ml of a urine
specimen is passed through 5-gram resin column resulting in
recovery of 2.5 ml of the specimen in the resin filtrate. This
filtrate is then mixed with an equal volume of acetic acid
diluent and processed through the filter, stained and washed
as previously described.
Resin treatment in this manner enhances rapid
filtration of the sample through filters. The average flow rate
of such samples is 0.2 minutes. Also, some positive bacteriuria
samples which may appear negative without resin treatment, will
produce positive results when the resin is used. Additionally,
urines which clog filters without resin treatment will pass
them more easily after the resin treatment.
Incorporation of the resin treatment into the method
of the invention may be accomplished as follows: Elkay filters
(serum separators~, which are 10-ml plastic tubes with filters
~30-40~4m) at the butt of the tube and a skirt protruding
around the butt which forms a seal when the separator, are
loaded with 5 grams of resin suspended in water containing
1:250 formalin and are placed in 16-mm test tubes each of which
contains 2.5 ml of pH 2.5 acetic acid. The fluid phase of the
resin suspension is then drawn off by vacuum, leaving a moist
resin column within the separator. The residual formalin
maintains the sterility of the column. The separator is then
placed in the 16-mm tube containing the acetic acid by forcing
the butt of the Elkay tube into the 16-mm tube and driYing the
plastic tube into the acetic acid. Tubes may be thus prepared
prior to use and stored in this manner. ~t the time of use,
3.0 ml of urine is added to the Elkay tube (which has about
4.5 ml of reservoir volume above the resin column~. The Elkay
-17-
tube ls then gripped and removed slowLy from the test tube. This
action produces a vacuum in the test tube because of the Elkay
skirt against the sides of the tube, -thus drawin~ the urine
through the resin and into the acetic acid. The end result is
a 5-ml sample containing the urine and acetic acid, which may then
be filtered, stained and washed in accordance with the method
of the invention.
Although the acetic acid diluent and the resin
exchan~er increase the efficiency of the bacteriuria-detection
method, there are still occasional urines which giye problems
due to the presence of pigments that are not remo~ed by the
anionic resin exchanger. slood, hemoglobin, certain basic drugs,
and basic pigments present in urine of patients with certain
pathologic disorders will coat the 0.65 ~m filter and prevent
staining of the bacteria. For example, when urine containing
blood is processed, the erythrocytes pass the resin (since cells
cannot exchange with resin~. When mixed with acetic acid, the
blood cells are lysed and the basic hemoglobin is concentrated
onto the filter in the form of a ~reenish pigment, which interferes
with the staining of bacteria. However, when such filters are
treated with hydrogen peroxide, the problem is resolved. A
30-second treatment of a filter containing hemoglobin with
0.2 ml of 30% H2O2 completely clears the filter of color. Staining
of the filter with safranin-EDTA and subsequent washing indicate
that the peroxide has no effect on the stainability of the
bacteria. In fact, peroxide treatment of bacteria often enhances
the staining. Therefore, in all cases where filters manifest excess
pigment ~after processing through the resin and acetic acid~ on
O.65~4m filters, they may be treated with H202 as described above
-18-
for 30 seconds. Thereafter they are stained and washed as
described earlier in this application.
Occasionally, very turbid, bloody or dark amber urines
will deposit a precipitate or pigmented compound on the membrane.
Staining of this material may lead to false-positive results. In
those cases where urine samples are so hea~ily contaminated with
precipitates, such as phosphates, carbonates, urates or blood,
it may be possible to employ the methods of the present invention
if the specimen is centrifu~ed at low speeds whereby these
materials are sedimented without sedimentation of bacteria.
Centrifugation at speeds on the order of 500 rpm are generally
effective for this purpose. As a result of such centrifugation,
the bacteria-containing supernatant will more readily pass
through the filter. This procedure will reduce false-positives
and will uncover positives that may be masked by the excess
pigment deposits.
The composition and methods described herein may
similarly be applied to the staining, detection and analysis
of gram-negative and gram-positive bacteria in other fluids,
such as culture media, blood, spinal fluids and water, as well
as to staining bacteria from such fluids which have been deposited
on membranes.
The following examples are illustrative of the invention
and are not to be taken in a limiting sense.
EXAMPLE 1
To a 100-ml sample of normal, bacteria-free urine was
added E. coli (~ram-ne~ative bacte-rial to make a final
concentration of 106 colony-~ormin~ units ~C~/ml. Staphy-
lococcus aureus ~gram-negative bacteria) was added in a sim-
ilar manner to a second-urine sample. The urine was then
--19--
~ J'7
treated with Safranin-O (a red, hasic dye~, at a dilution of
1:200,000. A control sample of urine without added bacteria was
also treated with the dye as descrihed.
The samples were held at room temperaturc (RT) for
30 minutes and then tubes containing the lO0-ml urine samples
were centrifuged at 3000 rpm for 30 minutes. The tubes were
then inspected for stained bacteria. The tube containiny the
added E. coli manifested a pellet at the bottom of the tube,
but no red color was evident, only the typical yrayish mass
seen when unstained bacteria are pelleted. In the tube
containing the Staphylococcus aureus was a pellet having a red
color. The control tube contained no pelleted mass.
The experiment was repeated with crystal violet and
toluidine blue. In both cases, the pellet deposited in the E.
coli tube was not colored, while in the Staphylococcus aureus-
containing tube the pellet exhibited the color of the dye
used: dark blue with the toluidine blue and purple-blue with
crystal violet.
EXAMPLE 2
.
100-ml samples of normal pooled urine were treated
with 10 CFU/ml E. coli and 1:200,000 Safranin-O. The control
samples were treated with the dye but were not treated with
bacteria. All samples were treated with the tetrasodium salt
of EDTA concentrations indicated below. The samples were held
at RT for 30 minutes, and the 100-ml tubes were then centrifuged
at 3000 rpm for 30 minutes to sediment the bacteria.
The bacteria pellet at the hottom of each tube was
scored for the amount of Sa~ranin p~esent in the pellet. O = no
color to the pelleted bacte~ia. * = trace of red. ~, ++, +~+,
++++ indicate increasing amounts of dye attached to the bacteria.
-20-
In addition, the supernatant fluids after centrifuqation were
tested for absorbance at the wavelength of the dye (520 nm) to
determine the percentage of dye removed by bacteria. The results
were as follows:
TABLE 1
.. . ... _
Flnal
conc. E. coli containinq urine Control urine
EDTA in Absor- % Absor- %
urine (m) Pellet bance removed Pellet bance removed
none 0 .423 0 0 .420 0
0.01 + .390 6 0 .415 0
0.02 + .320 23 0 .424 0
0.03 ++ .280 33 0 .409 o
0.04 ++++ .190 54 0 .416 0
0.05 ++++ .185 56 0 .421 0
0.06 ++++ .196 53 0 .409 0
0.07 ++++ .199 52 0 .424 0
.. . _ . . ... . _ .. . ...
The result in Table 1 indicate that, in the presence
of EDTA, significant amounts of dye become attached to the
bacteria and are found in the pellet.
Essentially the same results were obtained upon
repetition of the above experiment with the following basic
dyes: crystal violet, toluidine blue, methylene blue and neutral
red.
E~MPLE 3
Employing the procedure set forth in Example 2,
experiments were conducted with a variety of organisms using
Safranin-O as a model dye. The results were as follows:
TABLE 2
Test organismGram~ye attachment
stainNo EDTA 0.05 M EDTA
E. coli - 0 +~++
S. aureus + +*++ ~+++
Proteus vulgaris - Q +*++
Pseudomonas - Q *~*+
Group A StrepO + ~+++ *++~
Group D Strep. + ++++ ++++
Klebsiella pn. - 0 ++++
_ .. . . _ ~
-21-
rr ~
The results indicate that gram-negative organisms require the
presence of a chelating agent to bind the basic dye to bacteria
in urine, whereas gram-positive bacteria are stained by the dye
both in the presence and absence of a chelating agent.
Further experiments revealed that 1:5000 basic dye
was more effective in producing stained pellets in urine than
the dye in dilute form, as described above (i.e., 1:20Q,OQO). Even
with the higher concentration of dye, F~TA was still required for
dye-attachment to gra~-negative organisms.
EXP~LE 4
1:500 suspensions of Safranin-O were made in the diluents
indicated below. The samples were then autoclaved at 15 psi for
30 min. After cooling, the autoclaved samples were filtered
through a 0.22- ~m Millipore fllter. The filtrate was then mixed
- with an equal volume of 0.1 M EDTA to attain a final of 1:1000
Safranin and 0.05 M EDTA.
Ten-ml samples of normal urine were than passed through
13-mm diameter, 0.65- Mm porosity fiberglass-epoxy Finite filters.
The dye-EDTA stocks indicated below were used to treat the mem-
branes through which the urine had passed, by holding the stocksin contact with the membranes for a l-min staining period. There-
after, the stain was suctioned throu~h the membrane. The membrane
was then washed twice with 5 ml of pH 3 acetic acid to decolorize
the membrane.
The membranes were then scored as follows: O, complete
decolorization of membrane, which appears white; +, faint tin~e
of red; +, definite red coating of membrane; ++, red to purple
color. All scorin~ other than "O" represents false positi~es.
In addition, the turbidity of the dye stocks was scored with
0 indicating r.o turbidity and +, ++, +~+ and ++++ indicating
-22-
increasing turbidity. The dyes were then stored at ambient tem-
perature, and the quality of the suspension was similarly
scored after 24 hours.
The results of these tests are set forth in Table 3.
TABLE 3
Diluent for dye Membrane Turbidity of dye-EDTA
score stocks at
0 hr 24 hr
Distilled water + +++ ++++
Saline + ++ ++++
Trypticase soy broth (undiluted) Q 0 0
1:10 broth in water + 0 +
1:100 broth in water + 0 ++
Tryptose phosphate broth 0 0 0
l:lO broth in water + 0 +
l:lO0 broth in water + 0 +++
Trypticase soy broth l:lO in
saline + 0 +
Tryptose phosphate broth, 1:10
in saline * 0 +
5% glucose in water 0 0 ++++
10% calf serum in water + 0 ++++
1% sodium acetate +* 0 (not done)
-
The results in Table 3 indicate that only undiluted
tryptose phosphate broth, trypticase soy broth and glucose
yielded a dye product that could be completely remo~ed from
the membrane with the acid wash. All other diluents, including
dilutions of broths in water or saline, resulted in some degree
of staining of the membrane. Further, after o~ernight storage,
all dye-EDTA mixtures, except those in undiluted broths, had
at least begun to precipitate.
E~AMpLE 5
Ten-ml urine samples were treated with 0.05 ~ EDTA
and l:10,000 Safranin-O in trypticase soy broth and were held
-23-
at RT for 30 minutes. Samples were then passed through
13-mm, 0.65- ~m bacteriological me~branes (epoxy-fiber~lass~.
The membrane was then washed with 5 ml of the acids indicated
below. The membranes were then scored with 0 indicating no
color on the membrane and +, ++, +++, ++++ indicatin~ increasing
membrane color. The results were as follows:
TABLE 4
Washlng agentNormal Urlne (no bacteria)
Color before wash Color after wash
... . .
pH 3 acetic acid ++ o
pH 3 citric acid ++ 0
pH 3 HCl ++ ++
pH 3 H2SO4 ++ +
pH 3 nltrlc acid ++ ++
Similar tests were run on samples to which bacteria
were added using pH 3 acetic acid. The results of these
tests, scored as in Table 4, are set forth in Table 5.
TABLE 5
Test bacterlaUrine + bacteria
added to urine Color before wash Color after wash
E. coli ++++ ++++
S. aureus +++ +++
Pseudomonas ++ ++
Group A Strep. ~+++ ++++
The results of these experiments indicate that dye
attached to bacteria is not removed by an organic acid wash,
whereas free dye adsorbed by the membrane is removed by such
a wash.
EXAMPLE 6
A patient's urine may be tested for bacteriuria at a
pathognomonic leYel ~i.e., bacte~ria in urine at levels of 105
CFU/ml or greater) as follows. To a 9-ml sample of the urine
1 ml of a ten fold Safranin-O/EDTA concentrate is added, mixed
-24-
, ~ !r~, 7
and held at 25C for 30 minutes and then placed in a vessel, which
is connected in series to a 25-mm diameter clarifying 5-~4m
polypropylene felt filter and a 13-mm diameter bacteria-retaining
0.65- ~m white fiberglass-epoxy filter which can be decolorized
by pH 3 acetic acid. The urine is passed through the poly-
propylene filter and then throu~h the bacteria-retainin~ filter
by negative pressure. After -the total sample passes the filters,
the filters are treated with 5 ml of pH 3 acetic acid - 1;500
formalin mixture by passing this fluid through both filters
under negative pressure. The 13-mm fiberglass-epoxy filter is
then examined for color. The color of the membrane is matched
with a nomograph which indicates the expected bacterial counts
based on color intensity of the membranes.
In an actual clinical trial using the above procedure,
the patient's treated urine rendered the membrane surface orange-
red, indicating approximately 107 CFU/ml bacteria.
EXAMPLE 7
To determine whether or not a patient has responded
favorably to an antibiotic, the procedure outlined in Example 6
may be repeated at intervals. Antibiotic effectiveness is
indicated if no bacteria are eYident or the intensity of the
color is reduced, following commencement of antibiotic therapy.
On the other hand, the same or increased intensity indicates
bacterial resistance to the antibiotic administered.
The patient described in Example 6 was placed on Keflin
(a penicillin-type antibiotic~. The followin~ day, his urine
was re-examined as described in Example 6. The results were
again positive, showing an oran~e-red color that indicated
about 10 CFU/ml bacteria and indicatin~ that the or~anism was
resistant to the Keflin antibiotic. Gentamycin was then
LZ~
prescribed and upon testin~ the patient's urine the next day,
the filter surface manifested an off-whi-te color, indicatin~
absence or very low level bacteria and that the antibiotic
most recently prescribed was ef,fective.
EXAMPLE 8
As an alternative to the method described in Example 6,
a patient's urine may be tested for bacteriuria at a pathogenic
level as follows: 10 ml sample of urine is passed throu~h a
bacteriological, 13-mm diameter, 0.65- ~m fiberylass-epoxy
filter described in Example 6 to deposit bacteria present in the
urine thereon. The membrane is then treated with a 0.5 ml of
a 1:1000 Safranin-O - 0.05 M EDTA mi-xture for 30 seconds to 1
minute. The dye is then drawn through the membrane by suction
and the membrane is washed with 3-5 ml portions of acetic acid
as described in Example 6. One portion may be removed through a
side drain to remove excess dye and the others may be removed
through the filter. The color of the membrane may then be
matched a~ainst a nomograph to determine the degree of bacteriuria.
EXA~PLE 9
.
Septicemia or bacteremia can be detected as follows:
a blood culture is made by diluting a blood sample tenfold with
culture broth. The sample is incubated at 36~C. Every hour
(starting 3 hours after the initial incubation period at 36C)
a 3-ml sample is removed from the blood culture. The sample is
passed through a 2- ~m clarifyin~-me~brane to remove blood cells
and debris. The filtrate is then treated with dye and EDTA as
described in Example 6, except that only 0.3 ml of the dye-EDTA
test solution is added to the 3-~1 sa,mple. The sample is held
at 25C for 30 minutes and then passed serially throu~h the
clarifier and fiberglass-epoxy filter as in Example 6. After a
-26-
5-ml acetic acid-formalin wash, the filter is obseryed for
color and compared with a nom~o~raph.
Results of such a test usin~ a patient's blood are
shown below.
TA~LE 6
Hours after Color of Scorin~ for
culture initiated membrane bacterial ~rowth*
3 white a
4 faint pink +
orange
6 oran~e-red +~
7 dark red ~++*
_ _ . . . _ _ _ _ _ .
*Scoring as in Example 5
These results indicate that usin~ the method of the
present invention, after only a 4-5 hour incubation period, it
was possible to determine that at least 105 C~U/ml bacteria were
present in the culture. This is a far shorter period than that
which would be required to produce turbidity even in clear
(blood-free) fluid systems, which normally would appear only
after 24 hours.
EXAMPLE 10
5 ml of spinal fluid from a patient sufferin~ from
septic meningitis was added to 50 ml of culture broth and incubated
at 36C. At the intervals indicated below, 3-ml samples were
obtained from the culture and processed as described in Example 9.
At these same intervals, the culture was observed for gross
turbidity, which would indicate bacterial growth. The results
are shown below:
TAsLE 7
Hours after Visual turbldity Color of Scorin~ ~or
culture initiated of culture membrane bacterial growth
3 0 white 0
4 0pinkish-orange +
0 orange +
6 0orange-red ++
7 + red +++
8 +dark red ~+++
The presence of bacteria was manifested usin~ the
method of the invention 3 hours before turbidity had become
evident in the culture media. This is a significant advantage
when a patient is suffering with a serious disease such as
meningitis.
E~AM2LE 11
Antimicrobial susceptibility of bacteria may be
determined as follows: A rich organic broth suspension of bacteria
is diluted and divided into aliquots--one for each antibiotic to
be tested plus a control--and placed in wells in a cuvette in
contact with an antimicrobial elution disk. A zero hour control
is made by adding formalin to the original inoculum and incubating
and reading under the same conditions as the test. The cuvettes
are agitated briefly by rotary motion at 200 rpm in a ~7~C
incubator and then incubated until several generations of gro-~th
occur, i.e.,l 1/2 to 3 hours. The cultures are then stained
and filtered as described in E2ample 6. The intensity of the
color of each antimicrobial-containing culture is compared with
the control. Resistant cultures e~hibit the same color
intensity as the control, while susceptible cultures show less
color and intermediate cultures fall in-between.
EXAMPLE 12
Le~els of antibiotics in ~lood may be determined as
follows: A patient's blood is obtained and the serum removed.
-28-
The serum is twofold serially diluted in culture media. 1 ml
samples of the different serum dilutions are placed in tubes
and then each is treated with a 0.1 ml suspension of the original
bacteria isolated from the patient ~~ 105 CFU/ml). The tubes
are then incubated at 36C for 2-3 hours.
A second set of tubes are run side by side with the ones
described above. In the second set of tubes are placed 1 ml
samples containing a ran~e of known concentrations of the anti-
biotic which is present in the patient's blood. Each serially
diluted antibiotic sample is then treated with a 0.1 ml
bacterial suspension and incubated as described above.
A control is prepared by placing a 1 ml sample containing
no antibiotic in a tube, treating it with 0.1 ml bacteria
suspension and incubating as above.
At the end of the 2-3 hours incubation period, each
sample is stained, filtered and washed as described in E~ample 8.
The colors of the membranes are scored to determine the most
dilute serum sample and the least concentrated antibiotic sample
(i.e. MIC) which have an inhibitory effect on bacterial growth,
(i.e., the least concentrated serum which has comparatively less
¢olor intensity relative to the control and the lowest concen~
tration of antibiotic which e~hibits a less intense color than
the control, respectively~. ~ultiplication of this de~ree of
dilution level by the minimum inhibitory concentration gives
the concentration of the antibiotic in the patient's blood.
E~A~PLE 13
Several urines known to be positive for bacteriuria
(based on plating and countin~ colonies) but which were
difficult to process throu~h the bacteriuria-detection device
because of suspended solids being present, were treated with
-29-
an equal volume of pH 2.5 acetic acid - a . 05 M ~lycine diluent.
Duplicate urine samples were mixed with an equal volume of sterile
water as controls. In each case the total sample (in this case
2.5 ml urine + 2.5 ml acetic acid or 2.5 ml sterile water) was
filtered through a 10-mm diameter, 0.65- ~m filter and the
flow rate recorded. The filters were then stained and washed
with pH 3 acetic acid as described above and color intensities
of membranes scored accordin~ly. The results are shown below.
.. .. _ .
10Patient Causa-ti~e agent. C~U/ml Color intensity of
no. filter~flow rate
Urine + Urine +
_ water acetic acid
Proteus lo6 clogged +/1.7
11 E. coli 10 ~/1.9 ++/0.9
12 E. coli 3 x 10 clogged +/1.4
13 Pseudomonas 5 x 105 clogged +/1.6
14 S. aureus 3 x 105 0/1.8 +/0.7
Enterococci 6 x 105 clogged 0/1.7
16 Enterococci 2 x 105 0/1.9 -/0.9
17 S. epidermidis 8 x lQ5 0/1.9 0/1.4
18 K. pneumoniae 106 +/2.0 ++/0.9
19 S. ~arcescens lo6 +/1.8 +/0.8
* Numerator indicates color intensity of filter surface.
Denominator indicates the time in minutes to filter the
total sample.
E~AMPLE 14
A urine sample which contained excess precipitates and
which was heavily pigmented was processed using the acetic acid
diluent method described in Example 13. The urine, although free
of visual precipitates, still clogged the membrane and the test
-30-
,r~; j~
could not be completed. However, the test was repeated by passing
3 ml of urine through a 5-gram anionic resin column (~la9-D,
Diamond Shamrock, Cl charged~, and 2.5 ml of the filtrate was
collected in 2.5 ml of acetic acid diluent. The 5-ml sample was
then passed through the 0.65- ~m filter, which only required
0.2 minute. ~fter staining and washing with pH 3 acetic acid as
described above, the resultant filter surface manifested a red
color indicatin~ bacteriuria.
When another urine was processed, which was known from
plating to be negative for bacteriuria, it cloy~ed the filter
when the acetic acid diluent only was used. When processed
through the resin and collected in the acetic acid diluent as
described above, and then stained and washed, the memkrane
manifested a typical off-white color indicative of a negative
result.
Thus, the combination of the pH 2.5 acetic acid diluent
and the resin exchanger resolves clo~ging problems and will allow
the technician to make an immediate determination as to positive
or negative bacteriuria, rather than having to wait for the
24-48 hours required if the sample had to be plated and examined
for growth of colonies.
EXA~PLE 15
A bloody urine was passed through the anionic resin
described in Example 14 and 2.5-ml of the resin filtrate was
collected in 2.5 ml of pH 2.5 acetic acid diluent. The total
5 ml was then passed through the l~-mm diameter, 0.65- ~m filter
and a greenish pigment coated the me~brane. $tainin~ of the
filter with safranin-EDTA and pH 3 wash resulted in a bright
green filter surface. The urine sample was then processed again
as described above except the filter was first treated with
0.2 ml of 30% hydrogen peroxide for 30 seconds prior to stainin~.
After the 30-second peroxide treatment, the peraxide was drawn
through the filter and then stain was applied and the ilter was
washed with pH 3 acetic acid. The results indicated a strong
positive bacteriuria, since the membrane surface now manifested
a red color. Plating of the sample proved the urine to contain
E. coli in an amount of 106 CF~ml.
Another bloody urine was processed as described above,
but without peraxide treatment. ~ecause of the pigments, it
was impossible to determine whether bacteria were present. After
treatment of the greenish filter surface with H2O2 as above,
there was no pigment on the~membrane. Staining and then washing
with acetic acid resulted in a typical off-white color indicating
a negative test. Platin~ failed to detect any bacteria. - -
Another sample of urine containing a fluorescent yellow
pigment manifested a yellow filter surface. The resin had failed
to remove this pigment, and staining with dye yielded a strong
yellow color on the filter surface. H2O2 treatment removed the
pigment, and the standard treatment now yielded a positive red
color test for bacteria, which was later confirmed by plating.
-32-
