In vitro transfer of clarithromycin and amoxicillin across the epithelial

Page created by Karen Barnett
 
CONTINUE READING
Journal of Antimicrobial Chemotherapy (2002) 50, 865–872
DOI: 10.1093/jac/dkf219

     In vitro transfer of clarithromycin and amoxicillin across the epithelial
                        barrier: effect of Helicobacter pylori

                   Tamara Matysiak-Budnik1*, Martine Heyman1, Céline Candalh1, Denis Lethuaire2 and
                                                                                                          Francis Mégraud3

              1INSERM          EMI-0212, Faculté de Médecine Necker-Enfants Malades, 156 rue de Vaugirard, 75730 Paris;
                          2Sanofi-Synthélabo,   82 avenue Raspail, 94255 Gentilly Cedex; 3Laboratoire de Bactériologie,
                            Université Victor Segalen Bordeaux 2, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France

                                    Received 6 December 2001; returned 26 May 2002; revised 15 August 2002; accepted 22 August 2002

                                                                                                                                                                                                                                                                      Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
                  The mode of absorption of amoxicillin and clarithromycin, two antibiotics used in the treatment
                  of Helicobacter pylori infection, has not been completely elucidated. The aims of this study were
                  to investigate the passage of these antibiotics across normal and infected epithelium and to
                  measure their accumulation in HT29-19A or Caco2 epithelial cell monolayers. In non-infected
                  cultures, basal-to-apical fluxes were significantly higher than apical-to-basal fluxes for both
                  antibiotics, but this difference disappeared in monolayers infected with H. pylori. In 24 h studies,
                  clarithromycin, but not amoxicillin, showed rapid intracellular accumulation. No difference was
                  found between the transepithelial passage of amoxicillin across the HT29-19A and Caco2 mono-
                  layers.

                  Keywords: Helicobacter pylori, amoxicillin, clarithromycin, Ussing chamber, absorption

Introduction                                                                                                                        detected in the gastric mucosa after oral administration.5
                                                                                                                                    However, after intravenous (iv) administration, amoxicillin
Amoxicillin and clarithromycin are two antibiotics widely                                                                           was found in gastric juice7 and led to eradication of
used as antibacterials that have been used to treat Helico-                                                                         H. pylori,8,9 suggesting that eradication of the bacteria may be
bacter pylori infection. Resistance in H. pylori to amoxicillin                                                                     achieved by a systemic action of amoxicillin.
has rarely been reported, whereas resistance to clarithromycin                                                                         In contrast to amoxicillin, which is present mainly in the
is increasing steadily.1 Treatments with each one of these                                                                          extracellular space, it has been reported that clarithromycin
drugs when used as a single antibiotic therapy led to eradica-                                                                      may accumulate in the cells, suggesting a mainly systemic
tion of H. pylori in only 10–40% of infected patients, and                                                                          effect.10 After iv injection, the plasma concentration of clari-
when associated with a proton pump inhibitor, in 20–80% of                                                                          thromycin decreases immediately, and after that it increases
patients.2,3 The relatively low efficacy of amoxicillin and                                                                         slowly with time.11 Based on this observation, it has been
clarithromycin in vivo may be due to insufficient local drug                                                                        hypothesized that the drug accumulates in the cells and is sub-
concentrations at the locus of infection after oral adminis-                                                                        sequently released gradually into the extracellular fluid.12
tration.4–6                                                                                                                         Although the absorption of amoxicillin and clarithromycin
   H. pylori is located at the surface of the gastric epithelium,                                                                   has been studied in vitro13,14 and in vivo,15 there is no direct
in the mucus layer, and thus is not easily accessible to drugs.                                                                     evidence supporting this hypothesis. For this reason, we have
The question of whether bacterial eradication is achieved by                                                                        undertaken studies to measure the transepithelial passage of
the topical or systemic effect of amoxicillin in the gastric                                                                        both drugs.
mucosa remains the subject of discussion. For a long time, the                                                                         Employing an epithelial barrier model, the absorption of
main mode of action of amoxicillin has been considered to                                                                           amoxicillin and clarithromycin was investigated as a function
be topical since, in healthy volunteers, amoxicillin has been                                                                       of the mode (mucosal or serosal) of their administration, as
..................................................................................................................................................................................................................................................................

                                     *Corresponding author. Tel: +33-1-40-61-56-34; Fax: +33-1-40-61-56-38; E-mail: matysiak@necker.fr
...................................................................................................................................................................................................................................................................

                                                                                                                             865
© 2002 The British Society for Antimicrobial Chemotherapy
T. Matysiak-Budnik et al.

well as their intracellular absorption. The effects of the          attached) bacteria, and the filter-grown monolayers (with
presence of H. pylori on the absorption of the antibiotics by       attached bacteria) were placed between two half chambers of
HT29-19A epithelial cells grown as monolayers on micro-             an Ussing chamber to study the fluxes of the antibiotics.
porous filters and presenting tight junctions were measured
using radiotracer analysis.                                         Antibiotics
                                                                    Radiolabelled [14C]clarithromycin (specific activity: 56.6
                                                                    mCi/mM; radiochemical purity: >99%, labelled with 14C in
Materials and methods
                                                                    6-O-methyl position) was provided by Abbott Laboratories
Culture of the intestinal cell lines HT29-19A and Caco2             (Abbott Park, IL, USA) and [3H]amoxicillin (specific activ-
                                                                    ity: 43.7 mCi/mM; radiochemical purity: 96.4%, labelled
The intestinal cell line HT29-19A is a clone derived from the       with 3H in position 3 of the phenol moiety) was provided by
human colonic adenocarcinoma cell line HT29, which forms            SmithKline Beecham (King of Prussia, PA, USA).
tight monolayers of differentiated cells, widely used in func-
tional studies of the epithelial barrier.16 These cells were        Short-term absorption of the antibiotics in Ussing
cultured at 37°C in Dulbecco’s modified Eagle’s medium
                                                                    chamber (Figure 1a)
(DMEM; GIBCO, Life Technologies) containing 10% heat-
inactivated fetal calf serum (Boehringer Mannheim, Grenoble,        The Ussing chamber system has already been used to study

                                                                                                                                        Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
France), 4 mM glutamine and 50 µg/mL gentamicin, in a               antibiotic transfer across the epithelium in vitro.13,19 We used
humidified atmosphere of 5% CO2/95% air. In addition, the           this system to study short-term absorption of the antibiotics
Caco2 intestinal cell line was used for amoxicillin study           across the HT29-19A epithelial monolayer. The study in the
because these cells are known to express a specific transport       Ussing chamber lasted 1 h 40 min, a period of time during
carrier for β-lactam antibiotics (PepT1),17 which is not con-       which the viability of the tissue is preserved, allowing the
stitutively expressed in HT29-19A cells.18 These cells were         study of the early absorption of the antibiotics. Each side of
cultured in DMEM supplemented with 20% heat-inactivated             the exposed monolayer (surface area 0.150 cm2) was bathed
fetal calf serum, 4 mmol/L glutamine and 1% non-essential           with 1.5 mL of Ringer solution, and oxygenated at 37°C for
amino acids.                                                        the whole duration of the experiment. The electrical para-
    For experimental purposes, HT29-19A cells were seeded           meters, potential difference (PD) and electrical resistance
at a density of 0.6 × 106 per cm2 on polyethylene terephthalate     (R), attesting the integrity of the epithelium, were checked for
filters with a pore diameter of 0.4 µm and a surface area of        2 h at 30 min intervals. Administration of [3H]amoxicillin
0.9 cm2 (Falcon; Becton Dickinson Labware, NJ, USA).                and [14C]clarithromycin started after a 30 min period of
Caco2 cells were seeded at a density of 0.6 × 106 per cm2 on        equilibration of the system. Both antibiotics were introduced
Transwell polycarbonate insert filters with a pore diameter of      together into the apical (mucosal) or basal (serosal) compart-
0.4 µm and a surface area of 1 cm2 (Costar, Corning, NY,
                                                                    ment of the Ussing chamber at a final concentration of 10 mg/L
USA).
                                                                    each. Transport of the antibiotics was evaluated by sampling
    Upon reaching confluence, both HT29-19A and Caco2               the opposite compartment at 30 min intervals. At each time
cells progressively formed confluent monolayers with apical
                                                                    point, a 500 µL sample was obtained from a receiver compart-
and basolateral domains separated by tight junctions. They
                                                                    ment and was replaced with an equivalent volume of Ringer
were studied after 21 days of culture, when the electrical
                                                                    solution.
resistance, checked on the filter-grown monolayers mounted
                                                                       The concentration of radiolabelled amoxicillin and clari-
in an Ussing chamber, reached stable values of ∼150 Ω/cm2.
                                                                    thromycin was measured on 500 µL samples by β scintillation
                                                                    counting. Unidirectional fluxes of amoxicillin (Jamox) and
Bacterial strain                                                    clarithromycin (Jclari), apical-to-basal or basal-to-apical, were
An H. pylori strain isolated from a patient with duodenal ulcer     calculated using the following equation:
and subcultured minimally (less than five times) was used in                          Jamox (or Jclari) = ∆Q/dt·1/A
this work. Bacteria were grown on Vitox chocolate agar
(Oxoid, Dardilly, France) under micro-aerobic conditions            where ∆Q represents the amount of amoxicillin or clarithro-
(CampyGen in jar) for 24–36 h at 37°C. A bacterial suspen-          mycin accumulated in the opposite compartment during the
sion, containing from 6 × 107 to 9 × 107 cfu/mL, was prepared       time interval dt, and A the exposed area of tissue. The fluxes of
and used to infect the epithelial cells on the apical side. After   both antibiotics are expressed in ng/h·cm2.
24 h incubation, the cell monolayers were washed with                  The same experiment was carried out with non-infected
phosphate-buffered saline in order to remove planktonic (not        (n = 20) and H. pylori-infected (n = 10) cell monolayers.

                                                                866
Absorption of clarithromycin and amoxicillin

Figure 1. (a) Ussing chamber. Epithelial cells grown as monolayers on filters are placed between two half chambers of the Ussing chamber separat-

                                                                                                                                                            Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
ing two compartments, basal and apical. Antibiotics are added to one of the compartments and their presence on the opposite side is measured as a
function of time. (b) Cell culture insert. Inserts with filters bearing the epithelial cells, separating the basal and apical compartments, are bathed in
cell culture medium. Antibiotics are added to the apical or basal compartment and their presence on the opposite side is measured by sampling this
compartment at different time points.

Long-term absorption of the antibiotics across                                described as bearing the specific oligopeptide transporter
filter-grown intestinal monolayers (Figure 1b)                                PepT1, which can be used by this antibiotic to cross the
                                                                              epithelium.
This part of the study was carried out in order to evaluate the
long-term (up to 24 h) absorption of antibiotics across the
                                                                              HPLC chromatographic analysis of [14C]clarithromycin
epithelium, and particularly their intracellular accumulation.
Since the cell viability in Ussing chambers cannot be main-                   or its metabolites during transport across the HT29-19A
tained for more than 2–3 h, these experiments were carried                    cell monolayers
out using the cell monolayers grown on Falcon inserts. The
                                                                              While amoxicillin is excreted mainly in an intact form,20
inserts bearing the cells were bathed in the cell culture
                                                                              clarithromycin is known to undergo substantial metabolism.11
medium and kept at 37°C under an air 95%/5% CO2 atmo-
                                                                              Therefore, this part of the study was carried out in order to
sphere, which maintained the viability of the tissue for the
                                                                              determine whether clarithromycin is metabolized during its
whole duration of the experiment. Antibiotic, [14C]clarithro-
mycin or [3H]amoxicillin, was added separately, at final                      transepithelial passage across the HT29-19A intestinal cell
concentrations of 10 mg/L, to the basal (serosal) or apical                   monolayer. HPLC chromatographic methods have been used
(mucosal) compartment of the insert (surface 0.9 cm2) bearing                 widely for the determination of macrolide antibiotics in
the cell monolayers grown on the microporous filters. The                     different biological samples.21,22 We used a gel exclusion
apical and basal compartments were sampled at 10, 30, 60 and                  HPLC method with in-line detection of radioactivity to
360 min and at 10 and 24 h, starting from introduction of the                 analyse clarithromycin and its metabolites. To this end,
antibiotic. For each time point, six separate filter-grown                    [14C]clarithromycin was added at a concentration of 10 mg/L
monolayers were studied. At the same time points, the filter-                 to the basal or apical compartment of filter-grown HT29-19A
grown cells (one filter bearing ≈106 cells) were cut out of the               monolayers. After 8 h incubation, samples from the media
inserts, washed four times with Ringer solution in order to                   in the apical and basal compartments were collected and
remove extracellular antibiotic, and placed into 1 M nitric                   assayed. The 500 µL samples were concentrated using the
acid in order to obtain the cell lysis and liberation of intra-               Speed-vac system by evaporating the excess of water during
cellular drug. After a 24 h incubation, the lysed cells were                  centrifugation at low temperature in the Speed Vac Plus
centrifuged for 3 min at 3000 rpm, and 500 µL of supernatant                  apparatus (Savant Instruments, Farmingdale, NY, USA). The
was used to measure the radioactivity.                                        resulting volume of 200 µL was injected into a steric exclu-
   Additionally, the same 24 h study was carried out using                    sion HPLC column (Superdex Peptide PE 7.5/300). The
Caco2 cells with only amoxicillin. This cell line has been                    column was eluted at 30°C for 45 min with 0.1% trifluoro-

                                                                          867
T. Matysiak-Budnik et al.

acetic acid in 30% acetonitrile at a flow of 0.3 mL/min with        centration of 10 mg/L each did not affect the integrity of the
detection of eluting compounds using a radiochemical                epithelial barrier as assessed by its electrical resistance, which
detector (Berthold HPLC radioactivity monitor LB 506 6-1).          did not differ between the amoxicillin + clarithromycin
                                                                    exposed cells (n = 8) and control cells (n = 8), whose resist-
Analysis of antimicrobial activity of clarithromycin                ances were 150 ± 20 and 135 ± 15 Ω/cm2, respectively.
after its passage across the HT29-19A intestinal cell
monolayers
                                                                    Short-term absorption of amoxicillin and clarithromycin
This experiment was carried out in order to verify whether
clarithromycin antimicrobial activity is preserved after            across the control and H. pylori-infected HT29-19A cell
passage across the epithelial monolayer. Antimicrobial              monolayers
activity in the culture media containing [14C]clarithromycin
                                                                    Time course of antibiotic absorption across control mono-
was measured before and after its apical–basal passage across
                                                                    layers. Clarithromycin and amoxicillin fluxes (means ± S.D.)
the HT29-19A cell monolayer according to a standard
                                                                    as a function of time are shown in Figure 2. The apical-to-
procedure.23 Briefly, Micrococcus luteus, inoculated on
                                                                    basal fluxes of amoxicillin rapidly achieved a steady state,
Mueller–Hinton agar (Bio-Rad, Marne la Coquette, France),
                                                                    and were 50 ± 40, 57 ± 35, 63 ± 27 and 86 ± 49 ng/h·cm2 at 10,
was used as the test organism. The culture media were tested
                                                                    40, 70 and 100 min, respectively. The clarithromycin fluxes
directly and after dilution and the results were compared with
                                                                    increased throughout the duration of the experiment and were
a standard curve made with concentrations of clarithromycin
                                                                    14 ± 24, 33 ± 50, 56 ± 65 and 127 ± 110 ng/h·cm2 at 10, 40, 70

                                                                                                                                         Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
(0.015–2 mg/L). Readings were made after 24 h incubation at
                                                                    and 100 min, respectively.
30°C in air.
                                                                       Similar profiles were observed for the basal-to-apical
Statistical analysis                                                fluxes, with a steady state rapidly achieved by amoxicillin,
                                                                    and a progressive increase in the flux of clarithromycin.
Data were analysed using the SAS package (SAS Institute,            Although clarithromycin seemed to display a different profile
Cary, NC, USA). The results are expressed as means ± S.D.           from amoxicillin, there was no significant difference between
Comparison of means was carried out by analysis of variance         the amoxicillin and clarithromycin fluxes at any time point.
and by non-parametric tests (Wilcoxon two-sample test).
Differences were considered significant for P < 0.05.
                                                                    Dependence of end-point antibiotic fluxes on the side of
                                                                    drug application and effects of the presence or absence of
Results                                                             H. pylori. This analysis was carried out by comparing the
Effect of amoxicillin and clarithromycin on the epithelial          mean fluxes at 70 and 100 min. Values of 72 ± 35 and 115 ±
                                                                    84 ng/h·cm2 were measured for apical-to-basal and basal-to-
barrier integrity
                                                                    apical amoxicillin fluxes, respectively. They were not signi-
Amoxicillin and clarithromycin applied together on the apical       ficantly different from those determined for clarithromycin
or basal side of the epithelial monolayer at a final con-           (88 ± 83 and 160 ± 107 ng/h·cm2, respectively). For both anti-

Figure 2. Short-term (100 min) amoxicillin and clarithromycin apical-to-basal and basal-to-apical fluxes measured in an Ussing chamber
(mean ± S.D.).

                                                                 868
Absorption of clarithromycin and amoxicillin

                                                                             Long-term basal-to-apical absorption of clarithromycin
                                                                             and amoxicillin across HT29-19A cell monolayers
                                                                             Clarithromycin. The basal-to-apical transport of clarithro-
                                                                             mycin across the epithelial layer as a function of time is shown
                                                                             in Figure 4. At each time point, the results are presented as
                                                                             means ± S.D. of six independent measurements. Up to 60 min,
                                                                             very little clarithromycin was found in the apical compart-
                                                                             ment; its concentration was 64 ± 5 ng/mL at 120 min and
                                                                             increased gradually to a value of 686 ± 29 ng/mL after 24 h. In
                                                                             contrast, 10 min after the basal application of clarithromycin,
                                                                             the drug was detected inside the cells at a concentration of
                                                                             70 ± 10 ng/106 cells. The intracellular concentration of
                                                                             clarithromycin increased rapidly to values of 141 ± 7, 237 ± 22
Figure 3. End-point (70 and 100 min) apical-to-basal and basal-to-           and 408 ± 35 ng/106 cells after 30, 60 and 120 min, respect-
apical fluxes of amoxicillin and clarithromycin across normal and            ively, and reached a concentration of 571 ± 55 ng/106 cells
H. pylori-infected HT29-19A monolayers measured in an Ussing                 after 24 h. The data showed that basally applied clarithro-
chamber (mean ± S.D.). *Significantly different from apical-to-basal         mycin was first captured by the epithelial cells and subse-
fluxes, P < 0.02.                                                            quently released to the apical compartment.

                                                                                                                                                           Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
                                                                             Amoxicillin. Figure 4 shows that after basal application
biotics, the basal-to-apical fluxes were higher than the apical-             of 10 mg/L amoxicillin, only a small quantity of antibiotic
to-basal ones (P < 0.04 for amoxicillin and P < 0.02 for                     (13 ± 1 ng/mL) was found in the apical compartment after
clarithromycin) (Figure 3).                                                  60 min. The concentration of antibiotic in this compartment
   The presence of H. pylori did not modify the apical-to-                   increased gradually, reaching 56 ± 4 ng/mL at 6 h, 88 ±
basal fluxes of both antibiotics. The values measured for                    9 ng/mL at 10 h and 224 ± 7 ng/mL at 24 h. Virtually no
H. pylori-infected cells were 93 ± 76 ng/h·cm2 for amoxi-                    amoxicillin (
T. Matysiak-Budnik et al.

Figure 5. Comparison of the apical-to-basal and basal-to-apical           Figure 6. HPLC chromatograms of [14C]clarithromycin before and
amoxicillin fluxes across Caco2 monolayers in cell culture inserts        after transepithelial transport. Different elution times (26.62 min for
(mean ± S.D.). No significant differences were found between apical and   native clarithromycin and 28.13 min for its metabolite) indicate that dur-
basal fluxes of the antibiotic.                                           ing the transport clarithromycin underwent a modification of its structure.

                                                                                                                                                        Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
Long-term absorption of amoxicillin across Caco2 cell                     0.25 mg/L pure clarithromycin (reference) activity, whereas
monolayers                                                                the antimicrobial activity of the medium in the basal compart-
                                                                          ment (where the antibiotic was recovered), was equivalent to
The time course of the basal-to-apical and apical-to-basal
                                                                          0.5 mg/L reference activity. These results indicate that there
amoxicillin fluxes across the Caco2 cell monolayers was
                                                                          was no loss of antimicrobial activity during the transepithelial
similar to that obtained with HT29-19A cell monolayers. No
                                                                          passage of clarithromycin.
intracellular accumulation of antibiotic was observed, and the
concentrations of amoxicillin at different time points in either
compartment after its application on the opposite side were               Discussion
not significantly different from those observed with HT29-
19A cell monolayers. In addition, the transport of amoxicillin            Amoxicillin and clarithromycin are antibiotics that are used
across the Caco2 monolayers showed no significant differ-                 widely in the treatment of H. pylori infection. In the present
ences between apical or basal drug application (Figure 5).                study, using an in vitro epithelial barrier model, it has been
                                                                          demonstrated that these two antibiotics had different modes
Analysis of [14C]clarithromycin metabolites                               of transport across the epithelium. Clarithromycin was
                                                                          initially taken up by the epithelial cells, and then gradually
The HPLC chromatogram of clarithromycin after trans-                      released into the opposite compartment of the epithelium as
epithelial passage across an HT29-19A monolayer was                       its 14-hydroxy metabolite. In contrast, amoxicillin did not
slightly different from that of the parent clarithromycin, with           accumulate in the cells and crossed the epithelium either via a
elution times of 26.62 and 28.13 min for the parent clarithro-            paracellular pathway, across the intercellular tight junctions
mycin and its metabolite after passage across the monolayer,              or via a transcellular pathway. This study also showed that the
respectively (Figure 6). No evidence of other degradation                 basal-to-apical passage of amoxicillin and clarithromycin is
products was found after transepithelial transport. This small            larger than the apical-to-basal passage across the epithelium.
difference in the elution times between the two molecules
                                                                             The presence of bacteria seemed to decrease the basal-to-
suggests that both compounds have a similar molecular
                                                                          apical fluxes of both antibiotics but had no significant influ-
weight. This result, together with the result of the micro-
                                                                          ence on their apical-to-basal fluxes. There was no significant
biological assay (see below), strongly suggests that the com-
                                                                          difference between short-term (100 min) transepithelial
pound found in the receiver compartment could correspond to
                                                                          fluxes of both antibiotics, but clarithromycin presented much
14-hydroxy-clarithromycin, the principal biologically active
                                                                          better long-term (24 h) transepithelial absorption than
metabolite of clarithromycin.
                                                                          amoxicillin.
Antimicrobial activity of clarithromycin after its passage                   The fact that the cells concentrated clarithromycin but not
                                                                          amoxicillin was already known,10,12 but the mode of absorp-
across HT29-19A cell monolayers
                                                                          tion of these two antibiotics across the epithelial barrier had
The antimicrobial activity of the medium in the apical com-               not been established. Amoxicillin, like all β-lactam anti-
partment (where clarithromycin was added) was equivalent to               biotics, uses the apically located, energy-dependent dipeptide

                                                                      870
Absorption of clarithromycin and amoxicillin

carrier system to cross the intestinal epithelium along the          follow serum concentrations. The improved eradication
transcellular pathway. This system is present in human               results obtained by administering a continuous infusion of
enterocytes and has been found in some intestinal cell lines,        amoxicillin would further support such an explanation. It has
such as Caco2 and also HT29 cells,17,24 but is reported to be        been hypothesized that eradication failure could also be due to
absent in the HT29-19A cells used in this study.18,25 For this       the persistence of a few intracellular H. pylori.30 The intra-
reason, transport of amoxicillin across the latter cells was         cellular penetration of clarithromycin, in contrast to amoxi-
compared with that of Caco2 cells, known to express the              cillin, may allow it to kill the few intracellular bacteria
transporter. The results with the HT29-19A monolayer sug-            present.
gested that an amoxicillin transporter may be absent in this            In conclusion, in this study, we have shown that amoxi-
cell line since no antibiotic was found inside the cells. On the     cillin and clarithromycin have different modes of transepi-
other hand, the possibility cannot be excluded that the drug         thelial absorption. The higher passage rate of clarithromycin
crosses the epithelium, in part at least, through the tight          and its capacity to accumulate inside the epithelial cells may
intercellular junctions, as the lower molecular weight of            explain, in part, its better efficacy when used alone, in the
amoxicillin (mol. wt 365) relative to that of clarithromycin         eradication of H. pylori.
(mol. wt 747) may favour a paracellular diffusion of the
former across the epithelium. Moreover, as a zwitterion, it is       Acknowledgements
easier for amoxicillin, compared with negatively charged
clarithromycin, to cross negatively charged tight junctions.         Excellent technical assistance from Clémence Puzin and

                                                                                                                                             Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
Similar absorption profiles were observed with the Caco2             Cecile Miot-Sertier is acknowledged. The authors also thank
cells to those seen with HT29-19A cells, despite the well-           Dr Nathalie Dhoyen from SmithKline Beecham and Dr
documented presence of a specific PepT1 transporter in this          Robert Flamm from Abbott for providing the radiolabelled
cell line.17,26,27 No intracellular antibiotic was found and there   antibiotics, and acknowledge the kind financial support of
was no significant difference between the apical-to-basal and        Sanofi-Synthélabo, France, in the realization of this study.
basal-to-apical amoxicillin fluxes across the Caco2 mono-
layer. These observations suggested that either the specific
                                                                     References
PepT1 transporter was not active, or that its presence did not
modify significantly the kinetics of transepithelial transport        1. Megraud, F. (1998). Epidemiology and mechanism of antibiotic
of amoxicillin.                                                      resistance in Helicobacter pylori. Gastroenterology 115, 1278–82.
   It was assumed that amoxicillin passed through the epi-             2. Labenz, J., Gyenes, E., Ruhl, G. H. & Borsch, G. (1993).
thelium in a non-metabolized form as it has been demonstrated        Omeprazole plus amoxicillin: efficacy of various treatment regimens
in humans that >90% of amoxicillin is excreted intact.28 In          to eradicate Helicobacter pylori. American Journal of Gastroentero-
contrast, clarithromycin is known to undergo a rapid biotrans-       logy 88, 491–5.
formation leading to the production of the microbiologically          3. Axon, A. T. & Moayyedi, P. (1996). Eradication of Helicobacter
active metabolite 14-hydroxy-clarithromycin.11 We have               pylori: omeprazole in combination with antibiotics. Scandinavian
                                                                     Journal of Gastroenterology 215, Suppl. 21, 582–9.
shown that transepithelial passage of clarithromycin slightly
modified its HPLC profile, suggesting the metabolism of the            4. al Assi, M. T., Genta, R. M., Karttunen, T. J. & Graham, D. Y.
parent compound. In all probability, the metabolite corre-           (1994). Clarithromycin–amoxycillin therapy for Helicobacter pylori
                                                                     infection. Alimentary Pharmacology and Therapeutics 8, 453–6.
sponded to its hydroxy-metabolite because it was confirmed
that it preserved the antimicrobial activity of the parent            5. Cooreman, M. P., Krausgrill, P. & Hengels, K. J. (1993). Local
compound.                                                            gastric and serum amoxicillin concentrations after different oral
                                                                     application forms. Antimicrobial Agents and Chemotherapy 37,
   Clarithromycin is known to concentrate in cells such as           1506–9.
macrophages, polymorphonuclear leucocytes and lympho-
                                                                      6. Graham, D. Y., Opekun, A. R. & Klein, P. D. (1993). Clarithro-
cytes, and has antimicrobial activity both extra- and intra-         mycin for the eradication of Helicobacter pylori. Journal of Clinical
cellularly.29 In contrast, amoxicillin has been shown to act         Gastroenterology 16, 292–4.
mainly in the extracellular milieu. These two different
                                                                      7. Goddard, A. F., Jessa, M. J., Barrett, D. A., Shaw, P. N.,
behaviours may explain the better efficacy of clarithromycin         Idstrom, J. P. & Cederberg, C. et al. (1996). Effect of omeprazole on
in the eradication of H. pylori when each one of these drugs is      the distribution of metronidazole, amoxicillin, and clarithromycin in
used as the only antibiotic. For clarithromycin, the mode of         human gastric juice. Gastroenterology 111, 358–67.
transepithelial transport may serve to maintain a constant            8. Adamek, R. J., Wegener, M., Opferkuch, W. & Ruhl, G. H.
level of the drug on the apical side of the gastric mucosa owing     (1993). Successful Helicobacter pylori eradication: a systemic effect
to its slow release from the gastric cells where it had been         of antibiotics? American Journal of Gastroenterology 88, 792–3.
concentrated. In contrast, for amoxicillin, diffusion to the          9. Lamouliatte, H., Cayla, R., Meyer, M., Demotes-Mainard, F.,
apical side of the gastric epithelium would more closely             Vincon, G., Quentin-Noury, C. et al. (1991). Pharmacokinetics of

                                                                 871
T. Matysiak-Budnik et al.

oral and intravenous amoxycillin in human gastric mucosa. Italian          21. Kees, F., Spangler, S. & Wellenhofer, M. (1998). Determination
Journal of Gastroenterology 23, Suppl. 2, 109 (Abstr.).                    of macrolides in biological matrices by high-performance liquid
                                                                           chromatography with electrochemical detection. Journal of
10. Mor, N., Vanderkolk, J. & Heifets, L. (1994). Accumulation of
                                                                           Chromatography A 812, 287–93.
clarithromycin in macrophages infected with Mycobacterium avium.
Pharmacotherapy 14, 100–4.                                                 22. Erah, P. O., Goddard, A. F., Barrett, D. A., Shaw, P. N. & Spiller,
11. Peters, D. H. & Clissold, S. P. (1992). Clarithromycin. A review       R. C. (1997). The stability of amoxycillin, clarithromycin and metro-
of its antimicrobial activity, pharmacokinetic properties and thera-       nidazole in gastric juice: relevance to the treatment of Helicobacter
peutic potential. Drugs 44, 117–64.                                        pylori infection. Journal of Antimicrobial Chemotherapy 39, 5–12.

12. Carbon, C. (1995). Clinical relevance of intracellular and extra-      23. Edberg, S. C. (1986). The measurements of antibiotics in
cellular concentrations of macrolides. Infection 23, Suppl. 1, S10–4.      human body fluids: techniques and significance. In Antibiotics in
                                                                           Laboratory Medicine (Lorian, V., Ed.), pp. 381–476. Williams &
13. Goddard, A. F. & Spiller, R. C. (1997). In vitro assessment of         Wilkins, Baltimore, MD, USA.
gastric mucosal transfer of anti-Helicobacter therapeutic agents.
Antimicrobial Agents and Chemotherapy 41, 1246–9.                          24. Dantzig, A. H. & Bergin, L. (1988). Carrier-mediated uptake of
                                                                           cephalexin in human intestinal cells. Biochemical and Biophysical
14. Hulten, K., Rigo, R., Gustafsson, I. & Engstrand, L. (1996). New       Research Communications 155, 1082–7.
pharmacokinetic in vitro model for studies of antibiotic activity
against intracellular microorganisms. Antimicrobial Agents and             25. Merlin, D., Si-Tahar, M., Sitaraman, S. V., Eastburn, K.,
Chemotherapy 40, 2727–31.                                                  Williams, I. & Liu, X. (2001). Colonic epithelial hPepT1 expression
                                                                           occurs in inflammatory bowel disease: transport of bacterial
15. Gustavson, L. E., Kaiser, J. F., Edmonds, A. L., Locke, C. S.,
                                                                           peptides influences expression of MHC class 1 molecules. Gastro-
DeBartolo, M. L. & Schneck, D. W. (1995). Effect of omeprazole on
                                                                           enterology 120, 1666–79.

                                                                                                                                                     Downloaded from http://jac.oxfordjournals.org/ by guest on March 15, 2015
concentrations of clarithromycin in plasma and gastric tissue at
steady state. Antimicrobial Agents and Chemotherapy 39, 2078–83.           26. Abe, H., Satoh, M., Miyauchi, S., Shuto, S., Matsuda, A. &
                                                                           Kamo, N. (1999). Conjugation of dipeptide to fluorescent dyes
16. Augeron, C. & Laboisse, C. L. (1984). Emergence of perman-
                                                                           enhances its affinity for a dipeptide transporter (PEPT1) in human
ently differentiated cell clones in a human colonic cancer cell line in
                                                                           intestinal Caco-2 cells. Bioconjugate Chemistry 10, 24–31.
culture after treatment with sodium butyrate. Cancer Research 44,
3961–9.                                                                    27. Fujita, T., Majikawa, Y., Umehisa, S., Okada, N., Yamamoto, A.
17. Dantzig, A. H. & Bergin, L. (1990). Uptake of the cephalosporin,       & Ganapathy, V. (1999). σ Receptor ligand-induced up-regulation of
cephalexin, by a dipeptide transport carrier in the human intestinal       the H+/peptide transporter PEPT1 in the human intestinal cell line
cell line, Caco-2. Biochimica et Biophysica Acta 1027, 211–7.              Caco-2. Biochemical and Biophysical Research Communications
                                                                           261, 242–6.
18. Berlioz, F., Maoret, J. J., Paris, H., Laburthe, M., Farinotti, R. &
Roze, C. (2000). Alpha(2)-adrenergic receptors stimulate oligo-            28. Adam, D., de Visser, I. & Koeppe, P. (1982). Pharmacokinetics
peptide transport in a human intestinal cell line. Journal of              of amoxicillin and clavulanic acid administered alone and in com-
Pharmacology and Experimental Therapeutics 294, 466–72.                    bination. Antimicrobial Agents and Chemotherapy 22, 353–7.

19. Lampen, A., Zhang, Y., Hackbarth, I., Benet, L. Z., Sewing, K.         29. Scaglione, F., Demartini, G., Dugnani, S. & Fraschini, F. (1993).
F. & Christians, U. (1998). Metabolism and transport of the                A new model examining intracellular and extracellular activity of
macrolide immunosuppressant sirolimus in the small intestine.              amoxicillin, azithromycin, and clarithromycin in infected cells.
Journal of Pharmacology and Experimental Therapeutics 285,                 Chemotherapy 39, 416–23.
1104–12.
                                                                           30. Engstrand, L., Graham, D., Scheynius, A., Genta, R. M. & El
20. Todd, P. A. & Benfield, P. (1990). Amoxicillin/clavulanic acid.        Zaatari, F. (1997). Is the sanctuary where Helicobacter pylori avoids
An update of its antibacterial activity, pharmacokinetic properties        antibacterial treatment intracellular? American Journal of Clinical
and therapeutic use. Drugs 39, 264–307.                                    Pathology 108, 504–9.

                                                                       872
You can also read