Institute of Medical Microbiology, Solvegatan 23, 223 62 Lund, Sweden. Kronvall, G. Sr Myhre, E. Differential staining of bacteria in clinical specimens using ...
Acta path. microbiol. scand. Sect. B, 85: 249-254, 1977
DIFFERENTIAL STAINING O F BACTERIA IN CLINICAL SPECIMENS USING ACRIDINE ORANGE BUFFERED AT LOW pH GORANKRONVALL and ERLING MYHRE Institute of Medical Microbiology, Solvegatan 23, 223 62 Lund, Sweden
Kronvall, G. Sr Myhre, E. Differential staining of bacteria in clinical specimens using acridine orange buffered at low pH. Acta path. microbiol. scand. Sect. B, 85: 249-254, 1977. Optimal conditions for acridine orange staining of air dried and methanol fixed bacteria on glass slides were studied. The pH of the staining buffer did not influence the fluorescence of an S. UUTCUS and an E. coli strain at dye concentrations of 25-50 mg per litre. 81 bacterial strains representing 15 different species were stained with acridine orange under standard conditions, all strains showing orange fluorescence. The p H of the buffer influenced markedly the staining patterns of human cells and tissue materials, as represented by smears of peripheral blood, buccal scrapings, urethral secretions and tracheal exudates. The fluorescence obtained ranged from low intensity green at low p H values to bright orange at neutral and alkaline pH. This variability indicated a possibility of designing conditions for a differential staining method for the detection of bacteria in clinical specimens. The differential staining effect with a low p H in the buffer was confirmed on smears of buccal scrapings, cerebrospinal fluid samples and urethral secretions, showing orange fluorescence of the bacteria present and green-to-yellow fluorescence of background material, cells and tissue debris. Key words: Differential staining ; bacteria; acridine orange; low pH.
G. Kronvall, Institute of Medical Microbiology, Solvegatan 23, 223 62 Lund, Sweden.
Received 4.ii.77
Accepted 4.ii.77
The Gram stain was originally developed as a differential staining method for the detection of pneumococci in clinical specimens, this giving a dark bluish coloration of the bacteria in good contrast with the tissue material. However, this differential staining effect is obviously limited to Gram positive microorganisms. No staining method is yet available which will give one colour for bacteria and a contrasting colour for background material. Experiments reported here indicate that fluorochrome staining with acridine orange can
be modified in such a way that a differential staining effect can be achieved. Techniques for the staining of microorganisms with fluorochromes are in common use for the detection of acid-fast mycobacteria ( 12). Fluorochrome staining has been employed for studies of bacterial content of water and soil samples (2, 3, 5, 9, 10). The application of acridine orange staining for the detection of malarial parasites in blood smears has been reported (4, 6). Acridine orange is usually preferred in microbiological work. Its use has been hampered, however, by a green249
to-red variation in the staining pattern (1, 7, 8, 11) ,a variation presumed to correlate with viability (2, 9). We have studied some parameters in order to define optimal conditions for the staining of bacteria and tissue components in dried smears on glass slides. The main objective was to find conditions where all bacteria gave uniform staining, also possibly of another colour than the background tissue material. Such differential fluorescence was achieved when a low p H in the buffer was used, resulting in an orange staining of bacteria in contrast to green-to-yellow staining of human cells and tissue debris in the sample. MATERIALS AND METHODS
Microorganisms. Eighty-one strains of Gram positive and Gram negative bacterial species were obtained from routine cultures. The bacterial species were as follows: Staphylococcus aureus (13 strains), S . epidermidis ( 4 strains), S. saprophyticus (4 strains), group A, B, and C streptococcus ( 1 2 strains), enterococcus (7 strains), diphtheroid bacillus ( 1 strain), Ncisseria meningitidis ( 1 strain), N . gonorrhocae (7 strains), Eschcrichia coli ( 1 1 strains), Klcbsiella pneumoniae (10 strains), Protcus mirabilis, P . morganii and P . vulgaris (14 strains), Pscudomonas aeruginosa ( 1 strain). Clinical specimens. Duplicate smears of urethral secretions from male patients with suspected gonococcal infection were obtained from the Department of Dermatology, Lund. Smears of tracheal aspirates fmm patients with artificial respiration were made from samples sent for routine microbie logical examination. These samples contained inflammatory and phagocytic cells in various stages of degradation. Thin smears of blood and buccal scrapings from an apparently healthy male volunteer were used to evaluate the staining of less activated inflammatory cells. Very viscous samples were homogenized by adding 5 per cent acetylcystein (MukomysrB) , followed by centrifugation and suspension in phosphate buffered saline (PBSA; 0.12 M naC1, 0.03 M phosphate, p H 7.3, 0.02 per cent sodium azide). No untoward effects of this treatment on the staining results were seen. Staining solutions. Acridine orange (E. Merck A.G., Darmstadt, No. 1333) was dissolved in 0.15 M buffers ranging fmm p H 3.5 to p H 9.0.Acetate buffers covered p H values from 3.5 to 5.5, phosphate buffers p H 6.0 to 8.0, and Tris-HC1 p H 8.5 and 9.0. Staining was performed in troughs, each taking 8 slides in 50-80 ml of staining solution. The slides were then washed with tap water and dried.
250
Microscopical examination. A Leitz Ortholux I1 microscope equipped with Ploemopak 2.2, filter combination K, for incident light fluorescence was used throughout the studies. Microphotograps were taken with Nikkormat EL and FT2 cameras fitted on a Leitz shutter attachment on the microscope, using Kodak plus-X and Panatomic-X black and white films and Kodachrome 64 film for colour slides. RESULTS
Influence of p H on Acridine Orange Staining Patterns The effect of the hydrogen concentration during staining was studied on isolated bacteria and on inflammatory cells from various sources. Smears of Staphylococcus aureus and Escherichia coli were fixed with methanol for two minutes and stained for five minutes with 20 mg per litre acridine orange in buffers ranging from p H 3.5 to 9.0 with 0.5 p H unit intervals. Both strains showed intense orange fluorescence throughout the whole range of pH values, with very little variation in colour and intensity. Smears of normal blood, mucopurulent tracheal exudate from patients with artificial respiration, and urethral secretions from patients with culture positive Gc-infections were fixed in methanol and stained with 20 mg per litre acridine orange in buffers of different pH, and examined in the fluorescence microscope. The staining pattern of polymorphonuclear cells ( PMNs) showed clear correlation with the pH of the staining buffer (Table 1). At low pH, only the nucleus of blood PMNs was visible with faint green fluorescence. At higher p H values, the nucleus stained more strongly with green-to-yellow fluorescence and a green cytoplasm (Table 1) . Polymorphonuclear cells from exudates in general gave stronger staining, apparently related to their degree of activation (Fig. 1, Table 1). At low pH the cells were still green, whereas a t neutral pH or higher the orange staining of PMNs made the detection of bacilli in the smears more difficult (Fig. 1A and 1B). Free nuclei from disintegrated PMNs showed a reddish staining also at low pH. Mononuclear cells were similar to PMNs in their staining pattern. Lymphocyte nuclei
TABLE 1. Acridine Orange Staining Patterns of Polymorphonuclear Leukocytes from Peripheral Blood and from Gc-positive Urethral Secretions at Various p H Values PH Nucleus
3.0-5.0 5.5-8. o 8.5-9.0
Blood PMN Cytoplasm
Faint green Green Yellow
Unstained Faint green Green
Urethr. PNN Nucleus Cytoplasm Green Yellow Orange
Faint green Green Yellow
Bacteria in secretions and from cultures showed orange fluorescence throughout the pH range.
Fig. 1. Acridine orange staining (25 mg per 1) of human urethral secretions containing N . gonorrhoeae. When stained at pH 4.0, as shown in Fig. lA, the cocci stand out with an orange colour against a green background. At pH 8.5, Fig. lB, also the PMNs in the smear stain yellow to orange. x 1,500.
gave orange fluorescence already at low pH values. Erythrocytes in blood smears followed the same staining pattern as the background material, e.g. no to faint green fluorescence at low pH to clear green at high pH. Bacteria present in tracheal exudates and urethral secretions were also examined with regard to their staining patterns. Irrespective of the pH of the staining buffer, the bacteria always showed strong orange fluorescence (Fig. 1) . They were, however, more easily detectable in smears stained at low pH which had the least staining effect on the surrounding cells and background material (Fig. 3A). Intracellular bacteria including bacteria over-
laying the nucleus were easily detected, thus indicating the potential advantages of acridine orange staining at low pH. Fading of fluorescence was not so marked with orange staining of the bacteria, whereas the green background faded rapidly.
Standardization of Staining Conditions The influence of various parameters on the staining of bacteria was examined in order to ensure reproducib!e staining. Furthermore the influence of dye concentration on the staining of bacteria was studied. Acridine orange solutions ranging from 2 to 100 mg per litre in a
25 1
neutral buffer (PBSA) were used for 5-minute staining of S. aureus and E. coli. From 100 mg down to 20 mg acridine orange per litre gave strong orange fluorescence of the bacteria. Ten mg per litre and less of acridine orange gave only weak green-to-yellow staining. Concentrations between 25 and 50 mg per litre seemed suitable for bacterial microscopy. By varying the staining time between half a minute and 10 minutes with this concentration of acridine orange, full uptake of dye developed already after 1 minute, with no further changes up to the 10 minutes tested. A staining time of 2 minutes was selected as optimal for routine staining which was always performed in troughs to ensure proper buffering. Various methods for fixation of s. aureus and E. coli smears were also tested. Methanol, chloroform, ethyl ether: ethanol 1:1, and 10 per cent formaldehyde were used as fixatives for 10 minutes and compared with unfixed air-dried smears. No differences were seen between any of the procedures and the control. Similarly, varying the time for methanol fixation between 1 and 30 minutes did not change the staining properties of the bacteria. Methanol fixation for 2 minutes is used rou-
tinely for direct smears of human material, and was therefore retained in further experiments.
Acridine Orange Staining of Various Isolated Microorganisms The standard staining conditions selected were tested further on 81 bacterial strains from routine cultures. The bacterial strains represented 15 different species covering both Gram positive and Gram negative cocci and bacilli. The staining method consisted of methanol fixation for 2 minutes followed by brief washing in tap water and then immersion in acridine orange, 25 mg per litre acetate buffer, 0.15 M, pH 4.0, for 2 minutes. The slides were washed in tap water and dried and then examined in the fluorescence microscope with immersion oil applied to the smears. I n all cases, the bacteria showed strong orange fluorescence after staining with acridine orange. On a few slides there were occasional areas with about 10 per cent yellow-stained bacteria. By increasing the concentration of acridine orange to 50 mg per litre, this percentage decreased. This concentration should thus be preferred for routine use.
Fig. 2. Acridine orange staining at low pH. Fig. 2A shows an epithelial cell from buccal scraping with bacteria adhering. x 500. Fig. 2B shows a smear of a cerebrospinal fluid specimen from a patient with pneumococcal meningitis. x 800. Bacteria in both 2A and 2B stain orange against a green background of human cells.
252
cells, leucocytes and bacteria were examined. The orange fluorescence of the bacteria contrasted well with the surrounding human cells, as shown in Fig. 2 A. The staining pattern at low pH described above for leucocytes was also true for these specimens. The epithelial cells showed faint green cytoplasmic fluorescence and a nucleus with a green-toyellow colour. The bacteria were easily detected also when they adhered to the cells. Smears of cerebrospinal fluid samples from patients with suspected bacterial meningitis were stained with acridine orange at p H 4.0, in addition to routine Gram and methylene blue staining (Fig. 2 B, 3 B, 3 C). Results of a small series indicated that acridine orange was superior to the other two methods for detecting small numbers of bacteria, for instance when only a few meningococci were present on each smear (Fig. 3 B). Acridine orange therefore supplemented the Gram and methylene blue staining methods. The staining method was also tested for use in venerological practice by comparing it with methylene blue staining of urethral secretions from cases of suspected gonococcal infections (Fig. 3 A). I n 16 out of 17 cases the ,two staining methods were in agreement, 4 being Gc-negative and 12 Gc-positive. In one case the acridine orange staining was positive, methylene blue staining negative, and the culture showed growth of N. gonorrhoeae. Re-examination of the methylene blue stained smear revealed occasional diplococci. Fig. 3. Acridine orange staining of clinical specimens using an acidic staining buffer, pH 3.0. Fig. 3A, Exsudate from male urethra showing Neisseria gonorrhocae. Fig. 3B, Cerebrospinal fluid specimen from a case of meningococcal meningitis. Fig. 3C, Cerebrospinal fluid specimen from a patient with meningitis caused by Pscudomonas aeruginosa. A: x 1O00, B and C: x 500.
Acridine Orange Staining of Clinical Specimens
The standardized acridine orange staining was also applied to smears of clinical specimens. Buccal scrapings containing epithelial
DISCUSSION
Acid-fast staining methods using fluorochromes seem to be superior to Ziehl-Neelsen staining techniques for detecting mycobacteria in clinical specimens (12). This is attributed to the excellent contrast obtained between the stained bacteria and the dark background, thus making the examination of such slides both easy and thorough. Direct fluorochrome staining of other, non-acid-fast bacteria will present difficulties because of the uptake of dye by other cells as well as by the back-
253
ground material in smears. We present here a staining technique for the detection of bacteria in clinical specimens which gives a differential coloration of these structures. The method utilizes the p H dependence of acridine orange staining patterns of human cells. Our results have indicated that both human epithelial and inflammatory cells show greento-yellow fluorescence of the nucleus and no or green fluorescence of the cytoplasm at low pH. Since bacteria give strong orange fluorescence at these p H values, the conditions for a differential staining method are obtained. A p H dependence for acridine orange has been described by Strugger ( 7 ) for plant cells. Differences were noted in acridine orange staining of polymorphonuclear cells from various sources. Blood PMNs were stained green throughout the whole pH range, whereas PMNs from inflammatory exudates gave yellow-to-orange fluorescence already at neutral pH. In some cases of tracheal exudates the PMNs were almost red already at acid p H in the same way as free nuclei. The tendency to give a more orange staining seemed to correlate with the degree of activation and subsequent autolysis of PMNs. Acridine orange staining at low p H of clinical specimens, particularly cerebrospinal fluid samples and urethral secretions, seems to offer a valuable complement to other staining methods. Apart from its simplicity, its main advantage 'is the marked differential staining effect with orange colouring of bacteria in contrast to green-to-yellow staining of human cells and background material. REFERENCES 1. Adams, L. R . & Kamentsky, L. A.: Machine characterization of human leukocytes by acridine orange fluorescence. Acta Cytologica 15: 289-290, 197 1.
254
2. DaIcy, R . J . & Hobbie, J . E.: Direct counts of aquatic bacteria by a modified epifluorescence technique. Limnol. Oceanogr. 20: 875-882, 1975. 3. Flicrmans, C. G. & Schmidt, E. L.: Fluorescence microscopy: Direct detection, enumeration and spatial distribution of bacteria in aquatic systems. Arch. Hydrobiol. 76: 33-42, 1975. 4. Hansen, D . W., Hunter, D . T., Richards, D . F. & Allred, L.: Acridine orange in the staining of blood parasites. J. Parasitology 56: 3 8 6 3 8 7 , 1970. 5. Jones, J . G. & Simon, B. M . : An investigation of errors in direct counts of aquatic bacteria by epifluorescence microscopy, with reference to a new method for dyeing membrane filters. J. appl. Bach. 39: 317-329, 1975. 6. Seitz, H . M . : Vergleichende Untersuchung von Akridinorange und Wright-Farbung zur Malariadiagnose. Z. Tropenmed. Parasit. 22: 136-138, 1971. 7. Strugger, S.: FluoreszenzmikroskopischeUntersuchungen uber die Aufnahme und Speicherung des Akridinorange durch Iebende und tote Pflanzenzellen. Jenaische Zeitschrift fur Naturwissenschaft 73: 97-134, 1940. 8. Struggcr, S.: Neues uber die Fluoreszenzfarbung toter und lebender Bakterien. 1. Zur Theori der Bakterienfarbung. 2. Ein neues Verfahren zur mikroskopischen Unterscheidung lebender und toter Bakterienzellen. Deut. Tierantl. Wochschr. 50: 51-53, 1942. 9. Struggcr, S.: Fluorescence microscope examination of bacteria in soil. Can. J. Research Sect. C, 26: 188-193, 1948. 10. Strugger, S. & Hilbrich, P.: Die fluoreszenzmikroskopische Unterscheidung lebender und toter Bakterienzellen mit Hilfe des Akridinorangefarbung. Deut. Tierantl. Wochschr. 50: 121-130, 1942. 11. Traganos, F., Adams, L. R., Kamcntsky, L. A . & Melamed, M . R . : Critical effect of dye concentration on acridine orange fluorescence of fixed thvmocytes. Acta Cytologica 16: 281283, 1972. 12. Truant, J . P., Brett, W . A . & Thomas, Jr., W.: Fluorescence microscopy of tubercle bacilli stained with auramine and rhodamin. Henry Ford Hosp. Med. Bull. 10: 287-296, 1962.
