JAC
Journal of Antimicrobial Chemotherapy (1997) 39, 453–459
Effect of pH and buffer system on the in-vitro activity of five
antifungals against yeasts
I. Gadea, M. Cuenca, M. I. Gegúndez, J. Zapardiel, M. L. Valero and F. Soriano*
Department of Medical Microbiology, Fundación Jiménez Díaz, Avenida Reyes Católicos 2,
28040 Madrid, Spain
We have compared the effect of various media on the in-vitro activity of amphotericin B,
flucytosine, fluconazole, itraconazole and ketoconazole against 93 clinical yeast isolates by a
micro-broth dilution technique. The media used were: RPMI 1640 with 2% glucose, buffered
with 0.165 M MOPS at pH 7.0; the same medium, but buffered at pH 7.4; and the same
medium, but buffered at pH 7.4 with 0.15% sodium bicarbonate. The three media gave similar
results with azole antifungals and flucytosine, but the medium buffered at pH 7.0 failed to
detect different populations of yeasts with respect to amphotericin B susceptibility. In the
case of the media buffered at pH 7.4, Candida krusei was significantly less susceptible to
amphotericin B than Candida albicans or Torulopsis glabrata. We could not evaluate the
results obtained with Candida parapsilosis and Cryptococcus neoformans since these
species did not grow adequately in all three media.
Introduction
The widespread use of broad-spectrum antibacterial
agents and the increased susceptibility of some of the
population, such as immunocompromised people has led
to an increase in the incidence of serious yeast infections.1,2 The treatment of these infections with classical
and newer antifungals has created an increased demand
for in-vitro antifungal susceptibility testing.
Considerable effort has been made to define the conditions necessary to achieve reproducible in-vitro susceptibility testing of antifungal agents.2–16 As a result of many
collaborative studies, consensus within the National Committee for Clinical Laboratory Standards (NCCLS) has
been achieved and a reference broth macrodilution
method for antifungal susceptibility testing of yeasts has
been developed.17 Reports of comparative studies have
documented excellent agreement between the NCCLS
reference macrodilution method and microdilution testing, suggesting that the microdilution test is an adequate
tool for antifungal susceptibility testing of yeasts when
performed following the NCCLS for macrodilution
susceptibility testing.2,6,13
Despite this progress, some problems still remain.
Partial inhibition of fungal growth in vitro often takes
place over a range of flucytosine or azole concentrations
*Corresponding author. Tel:
and the reference method may not detect clinically relevant amphotericin B resistance.18 Some of these findings
could be related to the variation in MICs obtained under
different test conditions,19,20 for instance, because of the
differing pH values and buffer systems used.
The aim of this study was to determine the in-vitro
activity of five antifungal drugs against different yeasts
using RPMI medium with three different buffer systems.
Materials and methods
Organisms
Ninety-three clinical isolates and four control strains were
tested. The 93 clinical isolates included 38 isolates of
Candida albicans, 15 of Candida parapsilosis, 17 of Torulopsis glabrata, 13 of Candida krusei and 10 of Cryptococcus
neoformans. Candida albicans ATCC 64551, C. albicans
ATCC 14053, C. albicans ATCC 64548 and Candida kefyr
ATCC 28838 were used as controls in all susceptibility
tests.
Antifungal agents
Fluconazole powder (Pfizer S.A., Madrid, Spain) and
flucytosine (Sigma Chemical Co., St Louis, MO, USA)
34-1-5447387; Fax:
453
© 1997 The British Society for Antimicrobial Chemotherapy
34-1-5494764.
I. Gadea et al.
were dissolved in sterile distilled water at a concentration
of 10 g/L and sterilized by filtration through 0.22 m
Millipore filters (Millipore S.A., Molsheim, France).
Itraconazole and ketoconazole (Janssen S.A., Madrid,
Spain) were rendered soluble in 100% dimethyl sulphoxide (Sigma) at a concentration of 10 g/L. Amphotericin
B deoxycholate (Fungizone, Squibb Industria Farmacéutica S.A., Esplugues-Barcelona, Spain) was dissolved
in sterile distilled water at a concentration of 5 g/L. The
five drug solutions were frozen at 70°C until used.
Assay media
For RPMI 1640 medium with 2% glucose and buffered
with 0.165 M morpholinepropanesulphonic acid (MOPS)
at pH 7.0 or pH 7.4, we used an RPMI 1640 broth
described earlier.14 Briefly, RPMI 1640 medium powder
with L-glutamine and without sodium bicarbonate (ICNFlow, Aurora, OH, USA) was used. Preweighed amounts
(10.4 g) of medium were supplemented with 18 g of
glucose and 34.53 g of MOPS (ICN-Flow), mixed with
600 mL of distilled water and then adjusted to pH 7.0 or
7.4 using 10 M NaOH. Distilled water was added to
the pH-adjusted medium up to a final volume of 1 L.
The medium was sterilized by filtration through 0.22 m
Millipore filters.
For RPMI 1640 medium with 2% glucose, buffered
with 0.15% sodium bicarbonate at pH 7.4, we used the
same RPMI 1640 medium powder with L-glutamine and
without sodium bicarbonate as above. Preweighed
amounts (10.4 g) of medium were supplemented with 18 g
of glucose and mixed with 600 mL of distilled water and
10 mL of 7.5% sodium bicarbonate (ICN-Flow). Distilled
water was then added to a final volume of 1 L. The
medium was sterilized by filtration through 0.22 m Millipore filters. The pH was 7.4 when the plates were incubated in 5% CO2 atmosphere.
Microtitre plates
We used sterile 96-well plastic assay plates (with roundbottomed wells) with matching covers (Greiner, Frickenhausen, Germany). The fluconazole concentration ranged
from 50.0 to 0.05 mg/L, the ketoconazole concentration
from 25.0 to 0.025 mg/L, the amphotericin B concentration from 25 to 0.025 mg/L, the flucytosine concentration
ranged from 50 to 0.05 mg/L and the itraconazole concentration from 12.5 to 0.012 mg/L. We prepared plates containing, in each row, 11 two-fold serial dilutions of
antifungal and, in the twelfth well, drug-free medium for
growth control. The plates with media buffered with
MOPS were prepared and handled as previously
described.3 The plates with RPMI buffered with sodium
bicarbonate were preincubated in a 5% CO2 incubator for
1 h so that they had reached the final pH (7.4) before they
were inoculated.
Inoculum
The yeast isolates were grown on Sabouraud dextrose
agar (Oxoid, Unipath Ltd, Basingstoke, UK) for 24 h at
30°C. Suspensions were prepared by picking five colonies
from the cultures. The yeast cells were suspended in 5 mL
sterile 0.85% saline. The resulting yeast suspension was
vortexed, and the turbidity adjusted to 1
106–5
106
cfu/mL using spectrophotometric methods as recommended by the NCCLS.17 The suspensions were diluted
1 : 10 with the same drug-free medium and 10 L were dispensed in each well. Confirmation of the inoculum size
was determined by diluting the final inocula 1 : 10. Ten
microlitres of the latter solution was spread on to a
Sabouraud dextrose agar plate with a sterile, bent glass
rod. Plates were incubated at 35°C until growth was
evident. This allowed counts of 10–50 colonies per plate.
Incubation
Inoculated plates with both media buffered with MOPS
were incubated at 35°C in air. The plates with medium
buffered with sodium bicarbonate were incubated at 35°C
with 5% CO 2. The incubation time was 48 h for the plates
inoculated with C. neoformans and 24 h for the other
yeasts.
Endpoint criteria
The MIC endpoint for amphotericin B was defined as the
minimum concentration in which no growth could be
observed after agitation of the plates. Readings for the
other antifungals were made, both visually and spectrophotometrically, after agitation of the plates with a
microtitre shaker. If there were bubbles in the wells, the
plates were centrifuged at 1000g for 10 s, and then the
plates were shaken again. Spectrophotometric readings
were performed with Dynatech automatic plate readers
set at both 405 and 550 nm. The MIC endpoint (IC50) was
calculated as previously described (the lowest drug concentration that met the criterion %T %Tk 0.5(100
%Tk), where T is transmission and k is the drug-free
well).14 The visual reading of the MIC, according to the
NCCLS, was also performed (a prominent decrease in
turbidity with respect to control wells corresponded to an
80% drop in the turbidity of the respective growth control).6 All the visual readings were taken by the same
person.
Analysis of the data
Each of the 93 clinical isolates, in the three media tested,
were read both spectrophotometric and visually. Discrepancies between the MIC endpoints with the reference
medium and the two modified media of no more than
454
Effect of pH and buffer on antifungal activity
455
I. Gadea et al.
two dilutions were used to obtain the per cent values. The
antifungal susceptibilities of the different species of yeasts
were compared with each of the three media used: MICs
obtained with every species of yeast were compared with
each other; the differences in the MICs were evaluated
with the Mann–Whitney non-parametric test. Pearson
linear correlation coefficients were calculated with the
results obtained with each medium.
In order to study the reproducibility of the MICs
obtained with the three media, the MICs for the four control strains of the five antifungals were determined six
times in each medium. The global differences among the
variances calculated with the binary logarithms of the
MICs obtained with the three media of the five antifungals were compared with the Cochran and Bartlett–
Box tests. The standard deviations obtained with each
antifungal in each medium were compared two by two
using the F-test; all of them are included in SPSSPC 4.0
statistics application.21
Results
MIC determination
C. albicans, C. krusei and T. glabrata isolates produced
clearly detectable growth at 24 h in all three RPMI 1640
media, and the determination of MICs was possible at this
time. C. parapsilosis produced sufficient growth for MIC
determination at 24 h only in RPMI buffered with MOPS
at pH 7.0. C. neoformans grew enough to determine MICs
at 48 h in the two media buffered with MOPS, but not
enough to determine the MIC clearly in the medium
buffered with sodium bicarbonate.
The spectrophotometric value ranges, MIC50s and
MIC90s of the three azole antifungals and flucytosine and
optical ones of amphotericin B, with the three species that
grew in the three media, are shown in Table I.
Determination of different populations of yeasts with
the three media
The frequency curves obtained with each antifungal for
each species (C. albicans, C. krusei and T. glabrata), with
the three media used, were quite similar.
Correlation determination of the MICs obtained with
the three media
Correlation coefficients among the results obtained with
the three media with the five antifungals are shown in
Table II. With the azole antifungals and flucytosine, the
correlation coefficients for the results in the three different media were very good, so that, whatever the medium
used, the results were similar. MICs of amphotericin B, in
Table II. Correlation coefficients among the different
media
BIC (7.4)
Flucytosine
BIC (7.4)
MOPS (7.0)
MOPS (7.4)
Amphotericin B
BIC (7.4)
MOPS (7.0)
MOPS (7.4)
Ketoconazole
BIC (7.4)
MOPS (7.0)
MOPS (7.4)
Fluconazole
BIC (7.4)
MOPS (7.0)
MOPS (7.4)
Itraconazole
BIC (7.4)
MOPS (7.0)
MOPS (7.4)
Buffer (pH)
MOPS (7.0) MOPS (7.4)
1.00
0.94
0.93
0.94
1.00
0.95
0.93
0.95
1.00
1.00
0.47
0.55
0.47
1.00
0.50
0.55
0.50
1.00
1.00
0.84
0.82
0.84
1.00
0.90
0.82
0.90
1.00
1.00
0.91
0.94
0.91
1.00
0.96
0.94
0.96
1.00
1.00
0.71
0.61
0.71
1.00
0.79
0.61
0.79
1.00
BIC, sodium bicarbonate buffer.
For all values less than 1.00, P 0.001.
contrast, were more influenced by the conditions of the
test, and so were not so well correlated.
Detection of interspecies differences in antifungal
susceptibility
To ascertain if any of the three media could detect any
difference among the susceptibility of the different
species to the five antifungals, MICs obtained with the
different media for each species were compared with
the non-parametric Mann–Whitney U-test. As expected,
the results were very similar in the three media with fluconazole and flucytosine, the two antifungals with better
correlation coefficients. The results were as follows: C.
krusei was less susceptible to flucytosine, ketoconazole
and fluconazole than C. albicans and T. glabrata (P
0.01) and less susceptible to itraconazole than C. albicans
(P 0.01) irrespective of the medium used. Although T.
glabrata was more sensitive to fluconazole than C. krusei,
the fact that two T. glabrata strains highly resistant to
fluconazole were included makes the MIC90 value for this
species higher than that for C. krusei, as may be seen in
Table I. Torulopsis glabrata was more susceptible to flucytosine than C. albicans in the three media used, being
456
Effect of pH and buffer on antifungal activity
Table III. Percentages of value agreement and number of exact matches between spectrophotometric and visual
readings
Organism (no. of isolates)
MOPS pH 7.0
C. albicans (37)
C. krusei (12)
T. glabrata (17)
MOPS pH 7.4
C. albicans (37)
C. krusei (10)
T. glabrata (17)
Sodium bicarbonate pH 7.4
C. albicans (36)
C. krusei (11)
T. glabrata (17)
% Value agreement
ketoconazole fluconazole itraconazole
Number of exact matches
ketoconazole fluconazole itraconazole
94.5
83.3
100
97.3
100
100
100
91.6
100
34
2
11
28
7
14
28
3
10
97.3
90.0
100
100
100
100
100
80.0
100
35
4
11
22
3
10
18
4
12
100
100
100
100
100
100
100
100
100
33
3
9
20
9
12
17
4
7
more susceptible in the two media buffered at pH 7.4 (P
0.146 for medium buffered with MOPS at pH 7.0; P
0.0001 with the other two). Torulopsis glabrata
appeared less susceptible than C. albicans to fluconazole
and itraconazole with the three media (P 0.001); it was
also less susceptible to ketoconazole, but only with the
two media buffered with MOPS, irrespective of the pH
(P 0.001). No significant differences in the susceptibility
of T. glabrata and C. albicans to ketoconazole were seen
in the medium buffered with sodium bicarbonate (P
0.238). With amphotericin B, no differences in susceptibility were detected among the different species with the
medium buffered with MOPS at pH 7.0, but with the
other two media buffered at pH 7.4, C. krusei was
significantly less susceptible to amphotericin B than both
C. albicans or T. glabrata (P 0.001).
Optical and spectrophotometric reading differences
The correlations obtained with visual readings and the
spectrophotometric ones with the three media were compared, calculating percentages of value agreement, i.e.
percentages of values with differences of fewer than two
dilutions among the two different reading methods. With
flucytosine, visual readings were easily performed so that
a clear-cut drop in turbidity was always seen. On the
other hand, with azole drugs, a clear-cut endpoint was
more difficult to determine visually, but there were good
correlations with both visual and spectrophotometric
readings. Subjectively, the visual reading was easier with
the bicarbonate buffered medium but, as may be seen in
Table III, the results were similar.
Intra-laboratory reproducibility with the three media
Intra-laboratory reproducibility performed with the control yeasts was good with the three media, with no significant global differences among them (Cochran C
0.38;
P
0.075; Bartlett-Box F
2.18; P
0.088). When the
values obtained with each antifungal in the three media
were compared two by two, there were significant differences between the reproducibility obtained with amphotericin B; it was more reproducible with medium buffered
with sodium bicarbonate than with MOPS both at pH 7.0
and pH 7.4 (F values 320.99 and 161.79, corresponding to
P 0.001 and P 0.002, respectively).
Discussion
Collaborative studies to date have focused on optimizing
intra-laboratory agreement and have only addressed indirectly the issue of detecting strains with unusually high
levels of drug resistance and their in-vivo susceptibility.
Several reports suggest a good correlation of the in-vitro
MICs of fluconazole or flucytosine and the in-vivo susceptibility of C. albicans, T. glabrata and C. neoformans.4,8
That is not, however, the situation with amphotericin B.
The reference method may not detect clinically relevant
amphotericin B resistance.22
Modifications of pH or buffering systems affected the
values of azole or flucytosine MICs, but did not alter the
utility of detecting populations with different susceptibility. The three methods were of comparable reproducibility
and results could be easily read, both spectrophotometrically or visually, to give similar readings.
With the three micromethods, we found, as expected,
457
I. Gadea et al.
that C. krusei was less susceptible to azole antifungals and
flucytosine than other yeasts, and that T. glabrata was less
susceptible than C. albicans to fluconazole (Table I). The
medium buffered at pH 7.0 with MOPS was superior in
supporting the growth of yeasts.
We were not able to detect any difference in the susceptibility of overall yeast isolates to amphotericin B with
the microdilution adaptation of the reference method, the
medium buffered at pH 7.0. As may be seen in Table I,
C. krusei, which has a poorer clinical response to amphotericin B, was significantly less susceptible to amphotericin
B than C. albicans or T. glabrata (P
0.001, Mann–
Whitney non-parametric test) when the two media
buffered at pH 7.4 were used. As we did not have any
strain of C. albicans or T. glabrata isolated from a patient
treated with amphotericin B where treatment failed, we
could not check any correlation between in-vitro and invivo results. Five out of 13 isolates of C. krusei were blood
isolates from patients with haematologic malignancies,
who had disease progression in spite of amphotericin B
treatment. The other eight strains came from untreated
patients with no invasive disease. This led us to believe
that a pH higher than 7.0 could be better in detecting
amphotericin B resistance. Hence to improve the detection of amphotericin B resistance, further studies will be
needed to determine whether modifications, perhaps
higher pH and new formulations to support yeast growth
at this pH, should be carried out.
Acknowledgements
We thank G. Ayala, Statistics Professor of Valencia
University, for valuable suggestions concerning statistical
studies. Part of this work was presented at the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, California, 17–22 September 1995.23
References
1. Ng, T. T. C. & Denning, D. W. (1993). Fluconazole resistance in
Candida in patients with AIDS—a therapeutic approach. Journal of
Infection 26, 117–25.
2. Pfaller, M. A. & Barry, A. L. (1994). Evaluation of a novel colorimetric broth microdilution method for antifungal susceptibility testing of yeast isolates. Journal of Clinical Microbiology 32, 1992–6.
3. Anaissie, E., Paetznick, V. & Bodey, G. P. (1991). Fluconazole
susceptibility testing of Candida albicans: microtiter method that is
independent of inoculum size, temperature, and time of reading.
Antimicrobial Agents and Chemotherapy 35, 1641–6.
4. Cameron, M. L., Schell, W. A., Bruch, S., Bartless, J. A.,
Waskin, H. A. & Perfect, J. R. (1993). Correlation of in vitro fluconazole resistance of Candida isolates in relation to therapy and
symptoms of individuals seropositive for human immunodeficiency
virus type 1. Antimicrobial Agents and Chemotherapy 37, 2449–53.
5. Cook, R. A., McIntyre, K. A. & Galgiani, J. N. (1990). Effects of
incubation temperature, inoculum size, and medium on agreement
of macro- and microdilution broth susceptibility test results for
yeasts. Antimicrobial Agents and Chemotherapy 34, 1542–5.
6. Espinel-Ingroff, A., Kish, C. W., Kerkering, T. M., Fromtling, R. A.,
Bartizal, K., Galgiani, J. N. et al. (1992). Collaborative comparison
of broth macrodilution and microdilution antifungal susceptibility
tests. Journal of Clinical Microbiology 30, 3138–45.
7. Espinel-Ingroff, A., Kerkering, T. M., Goldson, P. R. &
Shadomy, S. (1991). Comparison study of broth macrodilution and
microdilution antifungal susceptibility tests. Journal of Clinical
Microbiology 29, 1089–94.
8. Fromtling, R. A., Galgiani, J. N., Pfaller, M. A., Espinel-Ingroff,
A., Bartizal, K. F., Bartlett, M. S. et al. (1993). Multicenter evaluation of a broth macrodilution antifungal susceptibility test for
yeasts. Antimicrobial Agents and Chemotherapy 37, 39–45.
9. Ghannoum, M. A., Ibrahim, A. S., Fu, Y., Shafiq, M. C.,
Edwards, J. E. & Criddle, R. S. (1992). Susceptibility testing of
Cryptococcus neoformans: a microdilution technique. Journal of
Clinical Microbiology 30, 2881–6.
10. Gordon, M. A., Lapa, E. W. & Passero, P. G. (1988). Improved method for azole antifungal susceptibility testing. Journal
of Clinical Microbiology 26, 1874–7.
11. Pfaller, M. A., Burmeister, L., Bartlett, M. S. & Rinaldi, M. G.
(1988). Multicenter evaluation of four methods of yeast inoculum
preparation. Journal of Clinical Microbiology 26, 1437–41.
12. Pfaller, M. A., Rinaldi, M. G., Galgiani, J. N., Bartlett, M. S.,
Body, B. A., Espinel-Ingroff, A. et al. (1990). Collaborative investigation of variables in susceptibility testing of yeasts. Antimicrobial
Agents and Chemotherapy 34, 1648–54.
13. Pfaller, M. A., Bale, M., Buschelman, B., Lancaster, M.,
Espinel-Ingroff, A., Rex, J. H. et al. (1994). Selection of candidate
quality control isolates and tentative quality control ranges for in
vitro susceptibility testing of yeast isolates by National Committee
for Clinical Laboratory Standards proposed standard methods.
Journal of Clinical Microbiology 32, 1650–3.
14. Rodriguez-Tudela, J. L. & Martinez-Suarez, J. V. (1994).
Improved medium for fluconazole susceptibility testing of Candida
albicans. Antimicrobial Agents and Chemotherapy 38, 45–8.
15. Sewell, D. L., Pfaller, M. A. & Barry, A. L. (1994). Comparison
of broth macrodilution, broth microdilution, and E test antifungal
susceptibility tests for fluconazole. Journal of Clinical Microbiology
32, 2099–102.
16. Shawar, R., Paetznick, V., Witte, Z., Ensign, L. G., Anaissie,
E. & LaRocco, M. (1992). Collaborative investigation of broth
microdilution and semisolid agar dilution for in vitro susceptibility
testing of Candida albicans. Journal of Clinical Microbiology 30,
1976–81.
17. National Committee for Clinical Laboratory Standards. (1995).
Reference Method for Broth Dilution Antifungal Susceptibility Testing of Yeasts; Tentative Standard M27-T. NCCLS, Villanova, PA.
18. Galgiani, J. N. (1987). Antifungal susceptibility tests. Antimicrobial Agents and Chemotherapy 31, 1867–70.
19. Rinaldi, M. G. (1992). Laboratory evaluation of antifungal
agents: a brief overview. Clinical Infectious Diseases 14,Suppl. 1,
S130–3.
20. Working Party of the British Society for Antimicrobial
Chemotherapy. (1995). Antifungal drug susceptibility testing.
Journal of Antimicrobial Chemotherapy 36, 899–909.
458
Effect of pH and buffer on antifungal activity
21. Norussis, M. J. (1986). SPSS/PC
Chicago SPSS, Inc.
TM: for the IBM XT. AT.
22. Galgiani, J. N. (1993). Susceptibility testing of fungi: current
status of the standardization process. Antimicrobial Agents and
Chemotherapy 37, 2517–21.
23. Gadea, I., Gegundez, M. I., Cuenca, M., Valero, M. L., Zapardiel,
J. & Soriano, F. (1995). Effect of pH and buffer system on the in-
vitro activity of amphotericin B against yeasts. In Abstracts of the
35th Interscience Conference on Antimicrobial Agents and
Chemotherapy, San Francisco, CA, 1995. Abstract E75, p. 90.
American Society for Microbiology, Washington, DC.
Received 14 May 1996; returned 5 August 1996; revised 23
August 1996; accepted 18 October 1996
459