Journal of Antimicrobial Chemotherapy (2000) 46, 909–915

JAC
In vitro development of resistance to ceftriaxone, cefprozil and
azithromycin in Streptococcus pneumoniae

Kensuke Nagaia, Todd A. Daviesa, Bonifacio E. Dewassea, Glenn A. Pankucha, Michael R. Jacobsb
and Peter C. Appelbauma*
a
Departments of Pathology (Clinical Microbiology), Hershey Medical Center, 500 University Drive,
Hershey, PA 170331, USA; bCase Western Reserve University, Cleveland, OH 441062, USA

Approval of ceftriaxone for the treatment of otitis media has led to fear of selection of resistant
mutants owing to widespread use. To test this, we examined the ability of sequential sub-
cultures in sub-MICs of ceftriaxone, cefprozil and azithromycin to select resistant mutants in 12
pneumococci. Daily subculturing was performed 50 times or until mutants with raised ceftriax-
one, cefprozil or azithromycin MICs were selected. Of eight ceftriaxone-susceptible parents,
ceftriaxone did not select for any resistant mutants, while cefprozil selected for four mutants
(MICs 2–4 mg/L after 21–50 subcultures). Among four ceftriaxone-resistant parents, subcultur-
ing in ceftriaxone selected for one stable mutant with raised ceftriaxone MIC (>16 mg/L after 21
subcultures) and subculturing in cefprozil selected for one mutant with raised cefprozil MIC
(64 mg/L after 44 subcultures). Mutations were observed in pbp2x and pbp1a. Among six
azithromycin-susceptible parents, subculturing in azithromycin selected for five resistant
mutants (MIC 0.5–32 mg/L after 10–42 passages) and among six azithromycin-resistant strains,
subculturing selected for mutants with raised azithromycin MICs in all six strains (MIC 16–32
mg/L after 4–18 passages). All azithromycin-resistant mutants derived from azithromycin-
susceptible parents had mutations in domain V of 23S rRNA while all azithromycin-resistant
parents and derived mutants had mefE. Single-step mutation rates among the 12 strains at the
MIC ranged from 1.5  10–6 to <6.2  10–10 for ceftriaxone, >1.3  10–5 to 8.9  10–8 for cefprozil
and >1.1  10–6 to 6.7  10–10 for azithromycin. Multi-step and single-step testing showed that
ceftriaxone selected for resistant mutants less often than cefprozil and azithromycin.

Introduction and Moraxella catarrhalis). Ceftriaxone has good time-

The worldwide incidence of infections caused by pneumo-
cocci resistant to penicillin and other antimicrobials has
increased at an alarming rate during the past two decades
and, in particular, during the past 5 years.1–3 There is an
urgent need for antimicrobial agents that can be used for
infections such a pneumonia, bronchitis, sinusitis and otitis
media caused by penicillin-resistant pneumococci.
Ceftriaxone (one intramuscular dose) has recently been
approved for the treatment of acute otitis media in the
USA and is currently in use in Europe using three intra-
muscular doses over 3 days.4 Ceftriaxone, a third-generation
cephalosporin, has broad-spectrum antibacterial activity
including the three major causative agents of acute otitis
media (Streptococcus pneumoniae, Haemophilus influenzae
dependent bacterial killing because of its long half-life in
serum and middle ear fluid.5 The use of ceftriaxone has
been shown in several studies to be as effective as 10 days of
treatment with oral antibiotics such as amoxycillin  clavu-
lanate, cefaclor and trimethoprim–sulphamethoxazole for
the treatment of acute otitis media in children.6–9 There is a
concern that overuse or misuse of this drug could lead to an
increase in resistance.
In a previous study10 designed to determine if the recent
dramatic increase in incidence of drug-resistant pneumo-
cocci may be due in part to abuse of oral drugs such as
macrolides and cephalosporins, we found that sequential
subcultures in subinhibitory concentrations of azithro-
mycin (used to represent the macrolide group), cefuroxime
and cefaclor led to increased pneumococcal MICs. In this

*Corresponding author. Tel: 1-717-531-5113; Fax: 1-717-531-7953; E-mail: pappelbaum@psu.edu

909
© 2000 The British Society for Antimicrobial Chemotherapy
 K. Nagai et al.
study we compare ceftriaxone with two other drugs,
azithromycin and cefprozil, commonly used for the treat-
ment of otitis media. Specifically, we repeatedly exposed 12
strains of S. pneumoniae to subinhibitory concentrations
of ceftriaxone, cefprozil and azithromycin to determine if
resistance developed. Gene sequencing of pbp2x and
pbp1a and penicillin-binding protein (PBP)-labelling were
performed on some strains to determine the mechanism
of resistance to the cephalosporins and gene sequencing of
23S rRNA and L4 ribosomal protein was carried out to
determine the mechanism of resistance to azithromycin.
Materials and methods
Bacteria and antimicrobial agents
Twelve clinical strains of S. pneumoniae isolated within
the past 5 years were randomly selected. Organisms were
identified by optochin susceptibility and classified by
serotyping. Six were susceptible to ceftriaxone (MIC  0.5
mg/L) and azithromycin (MIC  0.5 mg/L); two were
susceptible to ceftriaxone and resistant to azithromycin
(MIC  2 mg/L); and four were resistant to ceftriaxone
(MIC  2 mg/L) and azithromycin. The six azithromycin-
resistant strains had mefE. Strains were stored at –70°C in
double strength skimmed milk (Difco Laboratories, Detroit,
MI, USA) before testing. Antimicrobials were obtained
as follows: ceftriaxone (Roche Pharmaceuticals, Nutley,
NJ, USA) cefprozil (Bristol-Myers Squibb, Princeton, NJ,
USA), azithromycin (Hoechst Marion Roussel, Romain-
ville, France).
MIC determination
MICs were determined by standardized microdilution
methodology in Mueller–Hinton broth (Difco Labora-
tories) supplemented with 5% lysed horse blood.11
Breakpoints for the following compounds were those
approved by the NCCLS12 with susceptibility breakpoints
as follows: ceftriaxone 0.5 mg/L, cefprozil 2 mg/L and
azithromycin 0.5 mg/L.
Serial passages
Serial passaging was performed as described previously
by Pankuch et al.10 For strains initially susceptible to the
selecting drug, passaging was stopped when the strains
became resistant to the selecting drug, except for cefprozil
which was stopped when the MIC reached the upper limits
of the susceptibility breakpoint (2 mg/L). Jacobs et al.3 have
shown that a more realistic susceptibility breakpoint for
ceprozil based on pharmacokinetic/pharmacodynamic para-
meters is 1 mg/L. For the strains that were initially resistant
to the selecting drug, passaging was stopped when the MIC
increased four-fold, irrespective of the number of sub-
910
cultures. Strains were then subcultured in antibiotic-free
medium for 10 serial passages and MICs of all drugs were
retested to determine whether the phenotype was stable.
Single-step mutational frequency
Frequency of spontaneous single-step mutation was deter-
mined by spreading c. 1  1010 cfu/ml in 100 ml aliquots on
brain–heart infusion agar (Difco) plates supplemented
with 10% lysed horse blood containing 1 , 2 , 4 , 8 
and 16  MIC of each drug. Plates were incubated aero-
bically for 48–72 h and the numbers of mutants counted.
Frequency rate was determined by dividing the number of
mutants by the total number of cells spread on to the plate.
Resistant mutants were confirmed by rechecking the MIC
to the selecting drug.
Time–kill studies
Time–kill studies were performed on some parent and
mutant strains as described previously.13
Serotyping
Serotyping of parent and passaged strains was performed
by the standard Quellung method using sera from Statens
Seruminstitut (Copenhagen, Denmark).
Pulsed-field gel electrophoresis
To determine whether resistant isolates obtained at the end
of serial passages were derived from those tested at the
beginning of the study, the parent strains and the strains
with increased MICs obtained after the last passage were
tested by pulsed-field gel electrophoresis (PFGE) using a
CHEF DR III apparatus (Bio-Rad, Hercules, CA, USA) as
described previously.10
PCR of macrolide resistance determinants
Template DNA for polymerase chain reaction (PCR) was
prepared using the Prep-A-Gene kit (Bio-Rad) as recom-
mended by the manufacturer. Strains were checked for the
presence of ermB and mefE by amplification by the PCR
using primers and cycling conditions described by Sutcliffe
et al.14 Ribosomal mutations (23S rRNA and L4) were
determined as described previously by Tait-Kamradt
et al.15
PCR and gene sequencing of pbp2x and pbp1a
A 1.2 kb segment of pbp1a was amplified by PCR using the
primers and cycling conditions as described by Asahi et al.16
A 2 kb segment of pbp2x was amplified using primers based
on those described previously by Laible et al.17 as
follows: pbp2x for 5-478CGTGGGACTATTTATGACC-
 Table I. Multi-step resistance selection results

Initial MIC (mg/L) before exposure Selected resistance after exposure to MIC (mg/L) after
to drug drug 10 drug-free subculturese

Strain pena ceftriaxb cefprozc azithrod drug MIC (mg/L) no. of passages ceftriax cefproz azithro Resistance mechanism of resistant clonesf

1 0.015 0.015 0.125 0.06 ceftriax NRg –h – – – –

cefproz >2 24 0.03 0.125 0.06 PBP1a and PBP2x: same as parent
azithro 2 21 0.015 0.125 1 23S rRNA: C2613 to T (3/4 copies)i
2 0.015 0.03 0.125 0.06 ceftriax NR – – – – –
cefproz NR – – – – –
azithro >2 21 0.015 0.125 0.5 –
3 0.015 0.015 0.125 0.03 ceftriax NR – – – – –
cefproz NR – – – – –

Pneumococcus in vitro selection of resistance

azithro 2 21 0.015 0.125 2 23S rRNA: C2613 to A (4/4 copies)
4 0.015 0.015 0.125 0.015 ceftriax NR – – – – –
cefproz NR – – – – –
azithro NR – – – – –
5 0.25 0.125 0.25 0.03 ceftriax NR – – – – –
cefproz NR – – – – –
azithro >2 42 0.125 0.25 32 23S rRNA: C2058 to G (2/4 copies)
6 2 0.25 0.5 0.06 ceftriax NR – – – – –
911

cefproz 2 50 0.25 2 0.06 –

azithro >0.5 10 0.25 0.5 1 23S rRNA: C2613 to A (4/4 copies)
7 0.25 0.125 0.5 4 ceftriax NR – – – – –
cefproz 4 21 0.125 4 4 PBP1a and PBP2x: same as parent
azithro 32 18 0.125 0.5 8 mefE
8 0.125 0.25 0.5 2 ceftriax NR – – – – –
cefproz 2 50 0.25 2 2 –
azithro 32 11 0.25 0.5 2 mefE
9 2 2 8 2 ceftriax >16 21 16 32 2 PBP1a: same as parent, PBP2x: A464T, Y524C
cefproz 64 44 2 64 2 PBP1a and PBP2x: same as parent
azithro 32 4 2 16 256 mefE
10 4 2 16 2 ceftriax 16 27 4 32 2 –
cefproz NR – – – – –
azithro 16 5 2 16 2 mefE
11 2 2 16 2 ceftriax >16 27 2 32 2 –
cefproz NR – – – – –
azithro 16 10 2 16 2 mefE
12 2 2 16 1 ceftriax >16 27 2 32 1 PBP1a: S457 to P, PBP2x: same as parent
cefproz NR – – – – –
azithro 16 44 2 16 1 mefE
a
Penicillin, bceftriaxone, ccefprozil, dazithromycin. eSelected strains with increased MICs.
f
Several ceftriaxone and cefprozil-selected mutants were screened for mutations in PBP1a and PBP2x compared with parent strain. Azithromycin-selected mutants derived from azithromycin-susceptible strains were
screened for changes in 23S rRNA and ribosomal protein L4 while azithromycin mutants derived from azithromycin-resistant parents were screened for the presence of ermB and mefE.
g
NR, did not become resistant or MIC did not increase 4. h –, not determined. iDenotes the number of mutated copies of the four 23S rRNA genes.
 K. Nagai et al.

Table II. Single-step mutational frequency

Strain MIC multiplea Ceftriaxone Cefprozil Azithromycin

1 1 1.5  10–6 b (0.015)c 4.7  10–7 (0.125)c >1.1  10–6 (0.06)c

2 2.0  10–8 1.3  10–7 >5.6  10–7
4 <5.0  10–10 <4.0  10–10 9.5  10–9
8 <5.0  10–10 <4.0  10–10 2.7  10–9
16 <5.0  10–10 <4.0  10–10 <4.0  10–10
2 1 3.1  10–8 (0.015) 8.9  10–8 (0.125) >1.6  10–7 (0.06)
2 3.7  10–9 4.2  10–10 5.3  10–10
4 <5.3  10–10 <5.3  10–10 <5.3  10–10
8 <5.3  10–10 <5.3  10–10 <5.3  10–10
16 <5.3  10–10 <5.3  10–10 <5.3  10–10
3 1 >6.0  10–7 (0.015) 5.5  10–7 (0.125) >4.9  10–7 (0.03)
2 >6.0  10–7 3.8  10–9 1.6  10–10
4 1.4  10–9 <2.0  10–10 <1.6  10–10
8 <2.0  10–10 <2.0  10–10 <1.6  10–10
16 <2.0  10–10 <2.0  10–10 <1.6  10–10
4 1 7.3  10–8 (0.015) 1.5  10–6 (0.125) <6.7  10–10 (0.015)
2 1.7  10–8 1.3  10–8 <6.7  10–10
4 <6.7  10–10 <6.7  10–10 <6.7  10–10
8 <6.7  10–10 <6.7  10–10 <6.7  10–10
16 <6.7  10–10 <6.7  10–10 <6.7  10–10
5 1 1.2  10–6 (0.015) >2.1  10–6 (0.25) 1.3  10–6 (0.03)
2 1.3  10–9 >2.1  10–6 1.9  10–8
4 <7.0  10–10 1.2  10–6 <6.5  10–10
8 <7.0  10–10 <7.6  10–10 <6.5  10–10
16 <7.0  10–10 <7.6  10–10 <6.5  10–10
6 1 <6.7  10–10 (0.25) >2.0  10–6 (0.5) >2.0  10–6 (0.06)
2 <6.7  10–10 5.3  10–8 6.7  10–9
4 <6.7  10–10 <6.7  10–10 1.3  10–9
8 <6.7  10–10 <6.7  10–10 <6.7  10–10
16 <6.7  10–10 <6.7  10–10 <6.7  10–10
7 1 2.9  10–8 (0.125) >6.1  10–6 (0.5) 1.4  10–6 (4)
2 <2.0  10–9 5.5  10–7 4.1  10–8
4 <2.0  10–9 <2.0  10–9 <2.0  10–9
8 <2.0  10–9 <2.0  10–9 <2.0  10–9
16 <2.0  10–9 <2.0  10–9 <2.0  10–9
8 1 <1.2  10–9 (0.25) >3.6  10–6 (0.5) >3.6  10–6 (2)
2 <1.2  10–9 >3.6  10–6 8.4  10–9
4 <1.2  10–9 <1.2  10–9 1.2  10–9
8 <1.2  10–9 <1.2  10–9 –d

16 <1.2  10–9 <1.2  10–9 –

9 1 2.3  10–7 (2) >1.3  10–5 (8) 2.6  10–7 (2)
2 <4.3  10–9 9.1  10–8 8.7  10–9
4 <4.3  10–9 8.7  10–9 4.3  10–9
8 <4.3  10–9 – –
16 <4.3  10–9 – –
10 1 5.0  10–7 (2) >1.3  10–6 (16) >1.3  10–6 (2)
2 <8.4  10–10 >1.3  10–6 8.3  10–8
4 <8.4  10–10 <4.2  10–10 1.3  10–9
8 <8.4  10–10 <4.2  10–10 <4.1  10–10
16 <8.4  10–10 <4.2  10–10 <4.1  10–10

912
 Pneumococcus in vitro selection of resistance

Table II. (Continued)

Strain MIC multiplea Ceftriaxone Cefprozil Azithromycin

11 1 <6.2  10–10 (2) >1.9  10–6 (16) >1.9  10–6 (2)

2 <6.2  10–10 1.1  10–6 2.5  10–8
4 <6.2  10–10 – 1.3  10–8
8 <6.2  10–10 – –
16 <6.2  10–10 – –
12 1 2.0  10–8 (2) >7.1  10–7 (16) >3.0  10–6 (1)
2 <9.4  10–10 >7.1  10–7 >3.0  10–6
4 <9.4  10–10 <2.4  10–10 >3.0  10–6
8 <9.4  10–10 <2.4  10–10 1.0  10–9
16 <9.4  10–10 <2.4  10–10 –
a
Frequency rates of existing resistant mutants in original inoculum were determined at 1 , 2 , 4 , 8  and 16  MIC.
b
Frequency was calculated as the number of resistant colonies per inoculum.
c
MIC in mg/L.
d
–, not determined.

GAAATGG503-3 and pbp2x rev 5-2533ATTCCAGCA- Results

CTGATGGAAATAAACATATTA2503-3. PCR prod-
ucts were purified from primers and excess nucleotides
using QIAquick PCR purification kit (Qiagen, Valencia,
CA, USA) as recommended by the manufacturer and
sequenced directly using an Applied Biosystems model
373A DNA sequencer using the following additional
primers:
pbp1a-2, 5-2384GTGAAAAAATGGCTGCTGCT2403-3;
pbp1a-3, 5-2403AGCAGCAGCCATCTTTTCAC2384-5;
pbp1a-4, 5-2963GATGAGCTTGAACTTTCAGC2944-3;
pbp2x-1, 5-958TATGAAAAGGATCGTCTGGG997-3;
pbp2x-2, 5-991GGAACAGAACAGTTTCCCAAC1112-3;
pbp2x-3, 5-1354ATGCCACGATTCGAGATTGGG1375-3;
pbp2x-4, 5-1488CAGGTAGCATCTCCCAT1471-3;
and pbp2x-5, 5-2105AGAGAGTCTTTCATAGCTGAA-
GC2083-3. Genes were sequenced three times each (twice
in the forward direction and once in the reverse direction)
on products of independent PCRs.
Detection of PBPs and PBP assay
One millilitre of early log phase culture was spun down for
30 s at 10 000g. Cells were resuspended in 0.1 M phosphate
buffer and incubated for 15 min at 37°C in various concen-
trations of unlabelled cefprozil or ceftriaxone. Two micro-
curies of [3H]benzylpenicillin (Amersham Life Sciences,
Piscataway, NJ, USA) was added and incubated for 15 min
at 37°C. Cells were lysed by the addition of 0.1 M phos-
phate buffer with 0.1% Triton X-100. A 10-fold excess of
unlabelled benzylpenicillin was added and incubated for
10 min at 37°C. Labelled PBPs were separated by SDS–
PAGE using a 7.5% gel and were visualized by fluoro-
graphy using preflashed Hyperfilm MP (Amersham Life
Sciences).
Subculturing in subinhibitory concentrations of
antibiotic
MIC results from subculturing in subinhibitory concentra-
tions of antibiotics are summarized in Table I. Of the eight
ceftriaxone-susceptible parents, ceftriaxone did not select
for any resistant mutants, while cefprozil selected for four
mutants (cefprozil MIC 2–4 mg/L after 21–50 subcultures).
One of these mutants (strain 1) had the cefprozil MIC revert
back to the baseline MIC of 0.125 mg/L from 2 mg/L after
10 serial subcultures in the absence of antibiotic. Among
four ceftriaxone-resistant parents, subculturing in ceftriax-
one selected for mutants in all cases with raised ceftriaxone
MICs (16 mg/L after 21–27 subcultures). However, after
subculturing for 10 passages on drug-free media the ceftri-
axone MICs reverted back to its initial levels for three of the
four mutants. Of the four ceftriaxone-resistant parents, sub-
culturing in cefprozil selected for one mutant with raised
cefprozil MIC (64 mg/L after 44 subcultures).
Among six azithromycin-susceptible parents, subcultur-
ing in azithromycin selected for five resistant mutants
(azithromycin MIC  0.5–32 mg/L after 10–42 passages) and
one for which the MIC changed from 2 mg/L to 0.5 mg/L
after 10 serial subcultures in the absence of antibiotic.
Among six azithromycin-resistant strains, subculturing in
azithromycin selected for mutants with raised azithromycin
MICs in all six strains (MIC 16–32 mg/L after 4–18 pas-
sages). However, for five of the six azithromycin-selected
mutants, the azithromycin MICs reverted back to baseline
(or within one doubling dilution) after 10 passages on anti-
biotic-free media.
Azithromycin-selected mutants did not have higher
ceftriaxone and cefprozil MICs and vice versa. Cefprozil-
selected mutants did not have raised ceftriaxone MICs

913
 K. Nagai et al.
while cefprozil MICs were typically two to four times
higher in ceftriaxone-selected mutants.
Single-step mutational frequency
Single-step mutational frequency rates are summarized in
Table II. At 1  MIC the ceftriaxone single-step muta-
tional frequency rates were less than or very similar to
(same log) the cefprozil and azithromycin single-step muta-
tional frequency rates in 10 of the 12 strains and 11 of the
12 strains, respectively. At 2  MIC the ceftriaxone single-
step mutational frequency rates were less than or very
similar to (same log) the cefprozil and azithromycin single-
step mutational frequency rates in 10 of the 12 and nine of
the 12 strains, respectively. At 4  MIC the ceftriaxone
single-step mutational frequency rates were less than or
very similar to (same log) the cefprozil and azithromycin
single-step mutational frequency rates in 11 of the 12
strains in both cases. In almost all cases at 8  and 16 
MIC, the single-step mutational frequency rates were very
similar.
Time–kill studies
Time–kill studies were performed on some parent strains
and derived mutants from serial passaging in subinhibitory
concentrations of the antibiotics. Specifically, three ceftri-
axone mutants (derived from parent strains 2, 4 and 9),
three cefprozil mutants (derived from parent strains 2, 4
and 6) and three azithromycin mutants (derived from
parent strains 1, 2 and 6) were tested (data not shown). The
time–kill kinetics between the parent and derived mutants
were essentially the same showing 99% killing at 2  and
4  MIC after 24 h (data not shown).
β-Lactam resistance mechanisms
Penicillin-binding protein genes pbp1a and pbp2x from
parent strains 1, 7, 9, 12 and their derived mutants were
sequenced to determine if resistance to cefprozil or ceftri-
axone was associated with point mutations in the genes.
Mutations were found in PBP1a at S457 to P (see Table I).
Mutations in PBP2x occurred at A464 to T, and Y524 to C.
PBP competition experiments with ceftriaxone or cef-
prozil were also performed on parent strain 9 and its
derived ceftriaxone and cefprozil mutants. No significant
differences were observed between parent and mutants in
the binding of cefprozil or ceftriaxone to PBP1a or PBP2x.
Serotyping, PFGE and macrolide resistance
determinants
The 12 pneumococcal strains used comprised serogroups/
types 1, 6, 9, 14, 19 and 23. All mutants had identical
serogroups/types and PFGE patterns to the parent strains
from which they were derived.
914
The six azithromycin-susceptible parent strains did not
contain mefE or ermB nor did the mutants derived from
them. The mutants were checked for ribosomal mutations,
specifically for mutations in domain V of 23S rRNA and
ribosomal protein L4. All four mutants had mutations in
domain V of 23S rRNA with two strains having four copies
of the 23S rRNA gene mutated at position C2613 to A, one
strain having three copies mutated at C2613 to T and one
strain having two copies mutated at position C2058 to G (see
Table I). The six azithromycin-resistant parent strains
contained mefE as did the mutants derived from them.
Discussion
This study has shown that the development of resistance
to ceftriaxone and cefprozil after repeated exposure to
subinhibitory concentrations only occurs after numerous
subcultures (20, range 21–44) even among penicillin-
resistant strains. Similar results were seen among azithro-
mycin-selected mutants derived from azithromycin-suscep-
tible parents (21–42 subcultures except for one strain which
only needed 10 subcultures). However, azithromycin
mutants derived from parent strains containing mefE
developed raised MICs very quickly (usually in 10 sub-
cultures). These results are in agreement with a previous
study of ours, which showed that development of resistance
(if it occurred at all) occurred more slowly with β-lactams
(especially amoxycillin  clavulanate) than the macro-
lides.10
Resistance to extended-spectrum cephalosporins is
mediated primarily by a reduced affinity in PBP2x and to a
lesser extent PBP1a.18 Mutations in or around S337TMK,
S395SN or K547SG of PBP2x have been observed in resistant
laboratory mutants and clinical isolates.15 Mutations in
PBP1a at or around ST371MK (particularly T371 to A or to
S) have been observed in resistant laboratory mutants and
clinical isolates.16 Comparison of some PBP1a and PBP2x
sequences between parent strains and their derived cef-
prozil- and ceftriaxone-selected mutants showed that some
contained mutations; however, not in or around the import-
ant motifs described above. Mutations could have occurred
in other PBPs that were not sequenced. Some strains of
S. pneumoniae resistant to cefotaxime have been described
that contain mutations in PBP3.18
The mechanism of resistance in the azithromycin-selected
mutants derived from azithromycin-susceptible parent
strains (i.e. ermB and mefE negative) was similar to the
mechanism of resistance described in a previous study in
which mutations were observed in domain V of 23S rRNA
or ribosomal protein L4.15 Mutations were observed only in
domain V of 23S rRNA in this study. Additionally, we
recently found similar mutations among some macrolide-
resistant clinical pneumococcal strains from Central and
Eastern Europe (L4 mutations) and the USA (23S rRNA
mutations).19
 Pneumococcus in vitro selection of resistance
In summary, in this study we attempted to ascertain the
potential of ceftriaxone, cefprozil and azithromycin to
select for resistance by exposure of S. pneumoniae to sub-
inhibitory and superinhibitory concentrations of the drugs.
Resistant mutants were selected after exposure to all three
drugs with ceftriaxone selecting for the least number of
mutants and having the overall lowest single-step mutation
frequencies compared with cefprozil and azithromycin.
These in vitro data indicate that ceftriaxone has a lower
potential to select for resistance than cefprozil and azithro-
mycin.
Acknowledgements
This study was supported by a grant from Roche Pharma-
ceuticals, Nutley, NJ, USA.
References
1. Appelbaum, P. C. (1992). Antimicrobial resistance in Strepto-
coccus pneumoniae: an overview. Clinical Infectious Diseases 15,
77–83.
2. Block, S., Harrison, C. J., Hedrick, J. A., Tyler, R. D., Smith, R. A.,
Keegan, E. et al. (1995). Penicillin-resistant Streptococcus pneu-
moniae in acute otitis media: risk factors, susceptibility patterns and
antimicrobial management. Pediatric Infectious Disease Journal 14,
751–9.
3. Jacobs, M. J., Bajaksouzian, S., Zilles, A., Lin, G., Pankuch, G. A.
& Appelbaum, P. C. (1999). Susceptibilities of Streptococcus pneu-
moniae and Haemophilus influenzae to 10 oral antimicrobial agents
based on pharmacodynamic parameters: 1997 U.S. Surveillance
Study. Antimicrobial Agents and Chemotherapy 43, 1901–8
4. Gehanno, P., Nguyen, L., Barry, B., Derriennic, M., Pichon, F.,
Goehrs, J. M. et al. (1999). Eradication by ceftriaxone of Strepto-
coccus pneumoniae isolates with increased resistance to penicillin
in cases of acute otitis media. Antimicrobial Agents and Chemo-
therapy 43, 16–20.
5. Gudnason, T., Gudbrandsson, F., Barsanti, F. & Kristinsson, K.
G. (1998). Penetration of ceftriaxone into the middle ear fluid of
children. Pediatric Infectious Disease Journal 17, 258–60.
6. Green, S. M. & Rothrock, S. G. (1993). Single-dose intramus-
cular ceftriaxone for acute otitis media in children. Pediatrics 91,
23–30.
7. Chamberlain, J. M., Boenning, D. A., Waisman, Y., Ochsen-
schlager, D. W. & Klein, B. L. (1994). Single-dose ceftriaxone versus
10 days of cefaclor for otitis media. Clinical Pediatrics 33, 642–6.
8. Barnett, E. D., Teele, D. W., Klein, J. O., Cabral, H. J. & Kharasch,
S. J. (1997). Comparison of ceftriaxone and trimethoprim–
sulfamethoxazole for acute otitis media. Pediatrics 99, 23–8.
915
9. Cohen, R., Navel, M., Grunberg, J., Boucherat, M., Geslin, P.,
Derriennic, M., Pichon, F. et al. (1999). One dose ceftriaxone vs. ten
days of amoxicillin/clavulanate therapy for acute otitis media: clinical
efficacy and change in nasopharyngeal flora. Pediatric Infectious
Disease Journal 18, 403–9.
10. Pankuch, G. A., Jueneman, S. A., Davies, T. A., Jacobs, M. R.
& Appelbaum, P. C. (1998). In vitro selection of resistance to four -
lactams and azithromycin in Streptococcus pneumoniae. Antimicro-
bial Agents and Chemotherapy 42, 2914–18.
11. National Committee for Clinical Laboratory Standards. (1998).
Performance Standards for Antimicrobial Susceptibility Testing.
M100-S8. NCCLS, Villanova, PA.
12. National Committee for Clinical Laboratory Standards. (2000).
Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria
that Grow Aerobically. M7-A5. NCCLS, Villanova, PA.
13. Pankuch, G. A., Lichtenberger, C., Jacobs, M. R. & Appelbaum,
P. C. (1996). Antipneumococcal activities of RP 59500 (quinu-
pristin–dalfopristin), penicillin G, erythromycin, and sparfloxacin
determined by MIC and rapid time–kill methodologies. Antimicrobial
Agents and Chemotherapy 40, 1653–6.
14. Sutcliffe, J., Grebe, T., Tait-Kamradt, A. & Wondrack, L. (1996).
Detection of erythromycin-resistant determinants by PCR. Anti-
microbial Agents and Chemotherapy 40, 2562–6.
15. Tait-Kamradt, A., Davies, T., Jacobs, M., Appelbaum, P. & Sut-
cliffe, J. (1999). Mutations in 23S rRNA and L4 ribosomal proteins
account for resistance in pneumococcal strains selected in vitro by
macrolide passage. In Abstracts of the Thirty-Ninth Interscience
Conference on Antimicrobial Agents and Chemotherapy, San
Francisco, CA, 1999. Abstract C-0842, p.116. American Society for
Microbiology, Washington, DC.
16. Asahi, Y. & Ubukata, K. (1998). Association of a Thr-371 sub-
stitution in a conserved amino acid motif of penicillin-binding protein
1A with penicillin resistance of Streptococcus pneumoniae. Anti-
microbial Agents and Chemotherapy 42, 2267–73.
17. Laible, G., Spratt, B. G. & Hakenbeck, R. (1991). Interspecies
recombinational events during the evolution of altered PBP2x genes
in penicillin-resistant clinical strains of Streptococcus pneumoniae.
Molecular Microbiology 5, 1993–2002.
18. Hakenbeck, R. (1999). β-Lactam-resistant Streptococcus
pneumoniae: epidemiology and evolutionary mechanism. Chemo-
therapy 45, 83–94.
19. Tait-Kamradt, A., Davies, T., Brennan, L., Depardieu, P., Cour-
valin, P., Duigan, J. et al. (1999). Two new mechanisms of resistance
to macrolides in clinical isolates of Streptococcus pneumoniae. In
Final Program, Abstracts and Exhibits Addendum of the Thirty-Ninth
Interscience Conference on Antimicrobial Agents and Chemo-
therapy, San Francisco, CA, 1999. Abstract LB-8, p. 15. American
Society for Microbiology, Washington, DC.
Received 11 April 2000; returned 27 June 2000; revised 18 July
2000; accepted 14 August 2000