In Utero and in Vitro Exposure to β-Lactams Impair Kidney Development in the Rat : Journal of the American Society of Nephrology

Journal Logo

Molecular Medicine, Genetics, and Development

In Utero and in Vitro Exposure to β-Lactams Impair Kidney Development in the Rat


Author Information
Journal of the American Society of Nephrology 11(5):p 874-884, May 2000. | DOI: 10.1681/ASN.V115874
  • Free


β-Lactams are the oldest and major class of antibiotics used in the treatment of infections. Their lack of toxicity has been frequently reported in the adult (1,2). They are widely prescribed during pregnancy because they are considered safe (reviewed in references (2) and (3). However, like most drugs, β-lactam antibiotics cross the human placenta and therefore the fetus is exposed to these molecules (4,5). Although it is well known that these drugs are not teratogenic, the absolute lack of toxicity of β-lactam antibiotics to the fetus remains to be determined. The period of organogenesis, i.e., the first trimester, is of greatest concern because any adverse effect of these drugs on developing tissues may impair their formation (6). Besides severe malformations that are usually diagnosed very early, mild impairments may also be induced that remain undetected at birth if not associated with abnormal function. To this extent, the kidney may represent a sensitive target because it accumulates antibiotics the most (7,8,9,10).

Renal organogenesis relies on inductive interactions between the ureteric bud, an epithelial outgrowth of the Wolffian duct, and the metanephrogenic mesenchyme. Dichotomous divisions of the ureteric bud will form the collecting tubules, and epithelium-mesenchyme interactions at its terminal buds will induce the mesenchyme to condense and differentiate into glomerular and tubular structures (reviewed in reference (11). The general features of the events that lead to metanephros formation are known, but the underlying mechanisms—especially those controlling the branching morphogenesis and the formation of a given nephron number—are far from fully understood (12,13,14,15). However, inborn nephron deficits have been reported, and, according to their severity, they may initiate a progressive renal disease or influence the rate of progression of acquired renal disease (16). Little attempt has been made to investigate their origin, although intrauterine growth retardation and vitamin A deficiency might be major causes of inborn nephron deficit (17,18). In rat, in utero exposure to aminoglycoside antibiotics has been shown to induce a permanent nephron deficit (19,20). This particular adverse effect on the developing kidney, leading to oligonephronia, was not predictable from what was known of aminoglycoside nephrotoxicity in the adult. And even if mild, the bilateral nephron deficit observed was sufficient to accelerate the development of glomerular lesions in adulthood (21). Concerning the mechanisms of this oligonephronia, we have demonstrated that an impaired branching capacity of the ureteric bud, but not of its elongation ability, was the initiating event of this drug-induced nephron deficit (22).

No attention has been paid to the effect of β-lactams on the embryonic kidney. The aim of this study was to determine whether they impair renal organogenesis. We studied two penicillins, ampicillin and amoxicillin, and one third-generation cephalosporin, ceftriaxone. Two experimental approaches were used. The first set of experiments was performed in vitro using whole metanephros organ culture. This culture system combined with specific labeling of glomerular structures has been shown to be well suited for recapitulating the main steps of renal organogenesis and quantifying in vitro nephrogenesis (23,24). Its use was aimed at testing the direct effect of a wide range of antibiotic concentrations on the ability of the kidney rudiment to grow and differentiate in vitro (25). The second experimental design consisted of in utero exposure of pregnant rat to these drugs, at a period overlapping early renal organogenesis. Histology and determination of the nephron mass in pups was performed to determine the potential embryo toxicity of these drugs.

Materials and Methods

Animals and Antibiotic Treatment

Sprague Dawley female rats of known mating date (day 0 of pregnancy was the day after mating) were housed individually and had free access to standard laboratory chow and tap water. For in vivo experiments, pregnant females were treated with one antibiotic, from days 11 to 15 of gestation, i.e., during the period overlapping the first stages of renal organogenesis in the fetus. Ampicillin and amoxicillin were administered using small pellets (6 mm in diameter) implanted under the skin on the neck of the animal. Each pellet was manufactured to release the antibiotic continuously for 5 d (Innovative Research of America, Sarasota, FL). The load of 150 mg in the pellet corresponded to treatment at a dose of 100 mg/kg per d. For ceftriaxone, daily intramuscular injections were performed either at 50 or 500 mg/kg. All females were allowed to deliver spontaneously. Newborn pups were weighed within 4 h of birth and litters were reduced to eight pups. Kidneys of supernumerary pups were used for light microscopy examination. At 14 d, nephron mass was determined in the left kidney, and the right kidney was processed for routine histology. For in vitro experiments, pregnant females were anesthetized and laparotomized at embryonic day 14 (E14). Fetuses were then aseptically removed and the embryonic kidneys were collected.

Metanephros Organ Culture

Metanephros organ culture was performed as described previously (23,24). After collection, metanephroi were freed of exogenous tissue and placed onto a 0.8-μm Millipore AA filter (Millipore, Saint-Quentin-en-Yvelines, France), floating on a defined serum-free medium and incubated for 6 d in 35-mm Petri dishes at 37 ± 0.5°C in a humidified incubator (5% CO2). The following defined medium was used: Dulbecco's modified Eagle's medium/F12 (vol/vol) supplemented with 15 mM Hepes, 45 mM sodium bicarbonate (pH 7.45), transferrin (6.2 × 10-8 M), selenium (6.8 × 10-9 M), insulin (8.3 × 10-7 M), triiodothyronin (2 × 10-9 M), and prostaglandin E1 (7 × 10-8 M). Culture medium was changed daily, and no antibiotic or fungicide was present throughout the control experiment. Under these conditions, metanephroi can differentiate and nephrogenesis proceeds through the usual stages (23). We performed paired experiments using both metanephroi of the same fetus: One was grown as a control, and the other was grown in the same medium supplemented with the antibiotic. For each set of experiments, 10 to 15 pairs of metanephroi were used, collected from at least three different litters. Antibiotic stock solutions were prepared as follows: Amoxicillin was made at a concentration of 4 mg/ml in the minimum medium (Dulbecco's modified Eagle's medium/F12), and ampicillin and ceftriaxone were prepared in phosphate-buffered saline (PBS) at concentrations of 10 and 30 mg/ml, respectively. All solutions were stored at -20°C, and serial dilutions were performed daily in the defined medium. Antibiotics were used at 10, 100, or 1000 μg/ml.

Determination of Differentiation and Growth States of Cultured Metanephroi

Differentiation was assessed by counting the glomerular structures present within the whole explanted metanephroi, as described previously (24). Briefly, filter-grown metanephroi were fixed individually in 2% paraformaldehyde in PBS supplemented with 0.1 mM CaCl2 and 1 mM MgCl2 (PBS/calcium magnesium [CM]) for 2 h at 4°C. Then, the metanephroi were carefully detached from the filter and rinsed sequentially in PBS/CM and PBS/CM containing 50 mM NH4Cl. After permeabilization with saponin, the explants were treated with vibrio cholera neuraminidase and labeled with fluorescein-coupled helix pomatia agglutinin and rhodamine-coupled peanut agglutinin. The samples were then mounted in PBS/glycerol mixed with anti-bleaching agents.

Overall, growth of the cultured metanephroi was determined by protein content measurement performed after glomeruli counting. Labeled metanephroi were rinsed in distilled water and sonicated for 15 s in 0.5 ml of distilled water. Protein content was measured according to Lowry's procedure, modified by Larson using serum bovine albumin as standard (26,27). Data are expressed as micrograms of protein per explant.

To assess the effect of antibiotics on the branching pattern of the ureteric bud, pairs of metanephroi were grown for 2 d with or without the test antibiotic and stained in toto with fluorescein-coupled dolichos biflorus agglutinin. The whole branching morphogenesis of the ureteric bud was visualized, and the number of tips was counted.

Nephron Mass Determination

The total number of nephrons was determined in the entire kidney of 14-d-old pups. Whole kidneys were incubated in 6N hydrochloric acid for 45 to 60 min at 37°C according to the kidney weight. After overnight storage in water at 4°C, macerated kidneys were placed in a 100-ml gauged flask. Skillful shaking leads to a suspension of tubular structures and unbroken glomeruli. Two to four aliquots of 0.5 ml were pipetted and used for glomerular counting.

Histologic Studies

Cultured metanephroi were fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) and washed overnight in the same buffer. They were then post-fixed in 1% osmium tetroxide, dehydrated through a graded ethanol series, infiltrated, and flat-embedded in Epon. Serial sections of 1 μm thickness were cut and stained with toluidine blue. To detect apoptotic cells in metanephroi grown with or without penicillin, two approaches were used. First, spatial visualization of apoptosis was performed using a whole-mount in situ DNA end-labeling technique. Metanephroi were fixed in 2% paraformaldehyde and submitted to a protocol of indirect Apoptag® in situ apoptosis detection kit (Oncor, Gaithersburg, MD) modified by Smith and Cartwright (28). Briefly, samples were dehydrated through a graded ethanol/Tris buffer saline series, treated with proteinase K (20 μg/ml for 20 min at 37°C), and incubated with the terminal deoxynucleotidyl transferase solution containing digoxigenin nucleotides for 1 h at 37°C. Digoxigenin-labeled DNA ends were recognized by alkaline phosphatase-conjugated anti-digoxigenin Fab fragments (Roche Diagnostics, Meylan, France), previously preabsorbed on rat embryonic tissues, and visualized accordingly without counterstaining. In the second approach, we determined apoptotic index according to morphologic criteria (29,30) following light microscopy examination of semithin sections of Epon-embedded metanephroi. At a final magnification of × 1000, apoptotic cells can clearly be distinguished from necrotic cells using the following stereotyped criteria: occurrence in scattered cells, presence of nuclear condensation, nondamaged plasma membrane, and cell shrinkage. Two plastic sections taken 25 μm apart were viewed for each group. Examination was focused on the induced (condensed) metanephric mesenchyme facing the ureteric bud ends and on the surrounding uninduced (loose) mesenchyme. Six samples were analyzed in control and penicillin groups, representing more than 700 cells in each group.

Renal morphology of newborn rats was examined on 1-μm-thick sections. A median slice of 1 mm thickness was first cut and treated for Epon embedding. Serial sections were collected for each animal and stained with toluidine blue. The entire cortex and medulla were visible on each section. In 14-d-old pups, the right kidney was removed, cut parallel to the short axis, and immersed in Dubosq Brazil fixative (4% picric acid in acetic acid/ethanol 80%/formaldehyde, 1:10:4). Paraffin/paraplast 3-μm sections were stained with Masson's trichrome or with Jone's reticulin. Sections were examined with a Nikon Optiphot microscope, and photographs were taken on T-Max Kodak films. For morphologic measurements, video images were transferred in a PowerMac station and analyzed using NIH Image software.

Statistical Analyses

Data are reported as means ± SEM. Results obtained from paired experiments were analyzed using the nonparametric Wilcoxon t test. χ2 tests were used to compare the various degrees of severity observed in vitro after antibiotic exposure. Comparison between various groups of treated females was performed by variance analysis (ANOVA) combined with the Fisher test. Significance was determined at P < 0.05.


In Vitro Development of Rat Metanephros in the Presence of Ampicillin and Amoxicillin

The morphology of explanted metanephroi after 6 d of culture is depicted in Figure 1. All tubular and glomerular segments were visualized using fluorescein-coupled helix pomatia agglutinin (Figure 1, A and B), and only glomerular structures appeared using rhodamine-coupled peanut agglutinin (Figure 1, C and D). As shown in Figure 1, A and C, fully developed metanephroi were observed in control experiments. It is of note that at the time of explantation, no glomeruli were present within the embryonic kidney at E14, and very few renal anlagen had been induced (24). Using glomerular labeling and counting as an index of renal differentiation in vitro, we showed that the presence of ampicillin at doses varying from 10 to 1000 μg/ml in the culture medium seemed to interfere with optimal in vitro development. Addition of 100 μg/ml of this antibiotic for 6 d diminishes in vitro metanephros development. Some areas along the periphery are free of tubular and glomerular structures (Figure 1B). At the concentration of 1000 μg/ml, the paucity of induced nephrons is evident (Figure 1D). Quantification of in vitro differentiation with ampicillin is shown in Figure 2A. A weak but significant reduction of the number of nephrons exists from the dose of 10 μg/ml, despite the occurrence of a similar protein content (20.0 ± 4.4 versus 18.0 ± 3.6 μg/explant, n = 10 pairs). With increasing concentrations of ampicillin, a dose-response effect is observed. At the highest dose, growth was impaired by 25% and nephron formation was reduced by about two-thirds. Examination of the severity and frequency of the nephron deficit in each set of experiments revealed that at the lowest dose we tested, half of the metanephros organ cultures show a normal phenotype. However, some explanted metanephroi displayed a nephron deficit that can reach 40% compared to paired control. At 100 μg/ml, only one-quarter of cultured metanephroi remained unaffected, and at 1000 μg/ml of ampicillin, almost all cultures were severely affected.

Figure 1:
In vitro metanephros differentiation in the presence of ampicillin. Pairs of metanephroi (A and B, C and D) were grown for 6 d and subsequently processed for lectin histochemistry. One metanephros was grown as control (A and C), and the other was cultured with 100 or 1000 μg/ml ampicillin (B and D, respectively). Glomerular and tubular structures were visualized with helix pomatia agglutinin (A and B) and glomeruli only with peanut agglutinin (C and D). Bar, 250 μm.
Figure 2:
Quantitative analysis of the effect of β-lactams on in vitro nephrogenesis. The number of glomeruli was determined in 10 to 15 pairs of metanephroi grown for 6 d without (□) or with ([UNK]) β-lactam at three different concentrations. (A) Ampicillin. (B) Amoxicillin. (C) Ceftriaxone. Below each pair of bars for every concentration of β-lactam is reported the distribution of the nephron deficit severity with the following code: white, <20%; stippled, 20 to 40%; gray, 40 to 60%; black, >60%. * P < 0.05, ** P < 0.01, and *** P < 0.001, respectively, compared to paired controls.

Quantitative data concerning amoxicillin are reported in Figure 2B. Observation of the whole cultured metanephroi revealed less severe features at low doses than for ampicillin. At the concentration of 10 μg/ml, no effect on both growth and differentiation occurred. A 10-fold increase in the concentration leads to a significant deficit of in vitro nephron formation of about 20% without significant reduction in the protein content. However, further increase of amoxicillin dose in the culture medium had severe toxic effects. The overall growth of the explanted metanephroi was reduced by 62% (20.8 ± 2.3 versus 7.8 ± 1.6 μg protein/explant, n = 10). In vitro differentiation was even more impaired as judged by the reduction of 75% of the number of nephrons. Distribution of the nephron deficit severity for each dose indicated that a higher dose led to a more severe and more frequent nephron deficit.


Metanephros organ cultures that grew in the presence of 10 or 100 μg/ml of ceftriaxone showed no major histologic alteration, as shown in Figure 3. Quantification of the number of nephrons formed in vitro revealed a mild reduction of 14 and 17% for each concentration, respectively (Figure 2C). No difference in the protein content was measured compared with controls. Surprisingly, by using ceftriaxone at the dose of 1000 μg/ml in the medium, a complete failure of metanephros development occurred in vitro (Figure 3D). The same results were observed with 500 μg/ml (data not shown). The nephron deficit severity for each dose is also reported in Figure 2C.

Figure 3:
In vitro metanephros differentiation in presence of cephalosporin antibiotics. Pairs of metanephroi (A and B, C and D) were grown for 6 d and subsequently processed for lectin histochemistry. One metanephros was grown as control (A and C), and the other was cultured with 100 or 1000 μg/ml ceftriaxone (B and D, respectively). Fewer glomerular structures have apparently been induced in B. At the highest dose, a complete failure of in vitro development occurred (D). Bar, 250 μm.

In Utero Exposure to β-Lactams

Effect on Pregnancy Outcome. Several parameters were monitored following birth. The delivery of most of the females was observed during day time, allowing us to determine the duration of gestation. Pups were counted and weighed 4 h after birth. None of these parameters was affected by ampicillin or amoxicillin treatment (Table 1). In the ceftriaxone group, females delivered 6 h earlier, leading to slightly lower birth-weight pups.

Table 1:
Effects of the antibiotic administration on the duration of the gestation, the mean number of newborn per litter, and the mean birthweight

Effects on Postnatal Development and Renal Ontogeny. After normalization of each litter to eight pups, newborn rats were kept with their mothers for 2 wk. Their body weight, kidney weight, and the total number of nephrons were then measured. The same postnatal growth was observed among control, ampicillin, and amoxicillin groups (Table 2). The number of nephrons was significantly reduced in pups delivered from the ampicillin group, and to a lower extent in those from the amoxicillin group. However, one pup out of four had a nephron deficit higher than 20% in the ampicillin group (28,656 ± 404, n = 9). By doubling the dose of ampicillin in vivo, we observed the same degree of oligonephronia, but all of the pups were concerned (29,120 ± 720, n = 5). The 14-d-old pups born to ceftriaxone-treated mothers unexpectedly were found to be larger than the controls (Table 2). In addition, the presence of an even larger kidney was noted (+28%), as confirmed by the increased kidney weight to body weight ratio. No sign of obvious morphologic alteration was observed during kidney collection. Determination of the nephron mass did not reveal any reduction of the number of nephrons. Expressed per gram of body weight, ceftriaxone offspring have a lower number of nephrons than controls (Table 2). Because we were quite surprised by the complete block of nephrogenesis induced by ceftriaxone at the highest dose used in vitro, we conducted an additional series of in utero ceftriaxone exposure experiments by treating pregnant rats with 500 mg/kg for 5 d. None of the parameters we measured at birth was affected by this treatment (data not shown). No nephron deficit was detected in 2-wk-old pups, but the kidney was enlarged by 35% compared with controls, leading to a kidney-to-body mass ratio above 1%.

Table 2:
Body weight and kidney weight of 14-d-old pups born to control or antibiotic-treated mothersa

Histologic Findings after β-Lactam Exposure

From in vitro studies, semithin sections of β-lactam-exposed metanephroi showed no major pathologic findings when used at 10 or 100 μg/ml. In an attempt to investigate the mechanism of the nephron deficit induced at the dose of 100 μg/ml, first we counted the number of terminal branches of the ureteric bud after 2 d of culture. No difference was observed between the pairs of metanephroi, whatever the amount of β-lactam we used (data not shown). Second, we examined the localization of cells showing features of apoptosis in cultured metanephroi with or without 100 μg/ml β-lactam. Particular attention has been paid to the nephrogenic zone. As shown in Figure 4, nick end-labeled cells were more numerous in this area following penicillin exposure than in controls. To confirm these findings and to specify which mesenchymal cells were the most susceptible to apoptosis, a morphologic approach was performed on plastic sections. In controls, the percentage of apoptotic cells in the induced and uninduced mesenchyme was 1.1 ± 0.4 and 7.9 ± 1.3, respectively. After ampicillin and amoxicillin exposure, a marked increase was observed within the condensed mesenchyme surrounding the ureteric bud ends (5.1 ± 0.7 and 3.4 ± 0.5, respectively; P < 0.05). The uninduced mesenchyme shows the same percentage of apoptotic figures compared to controls (7.7 ± 1.3 and 9.3 ± 0.9 for ampicillin and amoxicillin, respectively). At the highest dose, dilation of the intercellular spaces in ureteric bud extremities was a common feature in all of the drug-exposed metanephroi (data not shown).

Figure 4:
Visualization of apoptotic cells within metanephros organ culture. Control (A and B) and ampicillin (C)-treated embryonic kidneys were grown for 6 d. Omission of TdT enzyme served as a negative control and produced no detectable signal (A). In the absence of penicillin, few apoptotic figures were present within the nephrogenic zone (B). With ampicillin, more cells were labeled with the whole-mount TUNEL technique (C). Dotted line depicts the ureteric bud (UB) limits. Arrows point to apoptotic cells presumably in the induced and uninduced mesenchyme. Bar, 50 μm.

From in vivo studies, according to the class of β-lactam we used, two major findings were observed on renal sections from pups born to β-lactam-treated mothers. In ampicillin or amoxicillin groups, kidneys of newborn rats and 14-d-old pups displayed enlarged tubular segments (Figures 5 and 6). In newborn rats, both the collecting and the proximal tubules were affected, exhibiting slightly or severely dilated tubular lumens (Figure 5). Most of the cysts were predominantly localized in the juxtamedullary cortex, i.e., where the first nephrons have been formed. But they were also observed in the subcapsular zone in connection with ureteric bud ends (Figure 5D) or with Bowman's capsule of glomeruli lying underneath S-shaped bodies (Figure 5E). Enlargement of collecting tubules was noted in the papilla (Figure 5F). Two weeks later, cysts were still present in focal areas in all viewed sections, apparently affecting collecting tubules (Figure 6, B and C). They were more pronounced in animals in utero exposed to ampicillin than to amoxicillin. Concerning the kidneys of pups born to ceftriaxone-treated mothers, a punctuate dark material was readily observed as soon as the histologic blocks were trimmed (Figure 6E). This feature affected most of the medullary areas except the outer stripe. An interstitial edema with some leukocytic inflammation was observed (Figure 6F). Using polarization equipment, microcrystal-like structures were found within these inflammatory cells (data not shown). Occasionally, microcysts and focal areas of tubular necrosis were also observed in the cortex. In all groups, a similar cortical thickness was measured at 2 wk of age.

Figure 5:
Renal histology of newborn pups after in utero exposure to penicillin. Enlarged tubular segments (asterisk) were present in pups born of ampicillin-treated (B, D through F) or amoxicillin-treated (C) mothers, but not in controls (A). Cystic tubules affected predominantly juxtamedullary nephrons (B and C). Cysts in connection (D, arrow) with ureteric bud end (D, arrowhead) or with glomeruli (G in Panel E) were observed. Medullary cysts could also be detected in the papilla (F). Bars, 100 μm.
Figure 6:
Renal histology of 14-d-old pups born of β-lactam-treated mothers. Control (A and D) and ampicillin (B) or amoxicillin (C) in utero-exposed kidneys are illustrated. Cortex of metanephroi that have developed in the presence of β-lactams show tubular cystic dilation. Ceftriaxone-exposed fetuses displayed at 2 wk of age an edematous interstitium in the medulla (E). Transillumination of the ceftriaxone histologic blocks reveals punctuate dark materials within the medulla, but not in the outer area. Enlargement of a section corresponding to the white rectangle in E is depicted in F. Masson's trichrome stain. Bars, 300 μm in A through C and 50 μm in F.


In this study we provide experimental evidence that β-lactam antibiotics may impair renal organogenesis. First, we show that a short administration of penicillin antibiotics to pregnant rats when the metanephros forms in the fetus may lead to a permanent nephron deficit. Second, we found that pups born to β-lactam-treated mothers exhibit histologic damage within their kidneys, either focal cystic tubule dilation or interstitial inflammation. Third, using metanephros organ cultures, we demonstrate that β-lactam antibiotics exert a dose-dependent inhibitory effect on in vitro nephrogenesis. These results indicate that drugs which are associated with few or no side effects in adult kidney may exert toxic effects on the developing metanephros at therapeutic doses. The potential developmental toxicity of these drugs should therefore be kept in mind when they are administered during pregnancy at the time of renal organogenesis.

The observation of a permanent nephron deficit induced in utero by amino-penicillin exposure, combined with evidence of cystic tubule dilation, was unexpected. These compounds are widely believed to have no developmental toxicity. Nevertheless, they do possess some nephrotoxicity, since reports have mentioned acute renal failure in pediatric patients receiving supratherapeutic doses of amoxicillin (31,32,33). In our study, therapeutic doses have been used to treat pregnant females. The critical parameter was likely to be the timing of exposure, as already reported (20). Organogenesis is a period when drugs are known to have the greatest potential to cause malformations (6). The mechanism of this oligonephronia was not completely elucidated. Apparently, it did not rely on a defect of ureteric bud branching, as previously shown for an antibiotic of the aminoglycoside family (22). However, the presence of cysts in cortical ducts of the developing collecting system, which is still involved in nephron induction at its tips, may have disturbed nephrogenesis. Alternatively, the increased rate of apoptotic cells among the mesenchyme may have caused the nephron deficit. During kidney development, largescale death is a normal feature in the nephrogenic zone, reaching up to 3% of cells in rat metanephroi (30,34). Here we showed that in vitro exposure to ampicillin or amoxicillin induced more apoptotic cells in condensed mesenchymal cells adjacent to ureteric bud ends. These cells are usually rescued from apoptosis to undergo epithelial morphogenesis upon induction (35). Our findings indicate that a smaller fraction of the mesenchyme will participate in nephrogenesis, which is consistent with previous reports showing that extensive apoptosis reduces the number of nephrons (36,37,38). Another mechanism could be based on the occurrence of hypokalemia, a common feature of penicillin antibiotics (39,40). Despite the fact that such electrolyte imbalance occurring during fetal development remains hypothetical, it has been demonstrated that the early metanephros development is extremely sensitive to potassium concentration variations (41). Decreased amounts of potassium in embryonic kidneys led to abnormal development that was characterized by failure of nephron induction and occasional cystic dilations of the ureteric bud. This has been reported both in murine and human kidneys (41,42). Interestingly, it has been demonstrated recently that a decrease in intracellular potassium is an early event in programmed cell death (43). It is tempting to speculate that a similar phenomenon might be the leading event of the penicillin-induced oligonephronia.

A moderate nephron mass reduction is likely to have no functional consequences at birth due to hyperfiltration of the remaining nephrons. Thus, at first, a mild oligonephronia is unlikely to be detected. However, even if mild, an inborn nephron deficit is susceptible to favor the development of glomerulosclerosis in adulthood, as already reported in rats (20) and mice (44). Some have also suggested that a reduced number of nephrons favor the development of hypertension (45). In this study of in utero β-lactam exposure, none of these drugs had an effect on fetal growth. This was carefully checked since intrauterine growth retardation is known to induce permanent nephron mass reduction (46,47,48). In the present study, oligonephronia is combined with the development of renal cysts. The presence of enlarged tubular segments may raise some question of the normal differentiation of the juxtamedullary nephrons that have been induced during the peak exposure to the amino penicillin. But most of all, the presence of renal cysts in pups of 2 wk of age, if it persists, is likely to worsen the renal dysfunction in adulthood.

One of the difficulties in assessing potential drug developmental toxicity is to determine the maximal concentration to be used. As mentioned earlier, this was one of the reasons to use metanephros organ culture. Within 1 wk, this model, combined with quantitative approaches of nephron formation, allowed us to test various concentrations of β-lactams. This confirms its usefulness as a screening test for potentially toxic drugs (25). In humans, intravenous administration of penicillin yields a mean plasma concentration of approximately 50 mg/L (49). We therefore assumed that concentrations of 10 and 100 μg/ml of culture medium were close to physiologic levels, in agreement with the very few data available on concentration measurements of these drugs in the immature kidney (7). For both ampicillin and amoxicillin, data gained from in vivo experiments were consistent with the in vitro data obtained with the dose of 10 μg/ml, and to a lesser extent with the concentration of 100 μg/ml. By contrast, the ceftriaxone data from metanephros organ culture were different from the in vivo results. This discrepancy may be explained by the protein-binding capacity of this cephalosporin. In the plasma, its degree of protein binding can reach 98%. In metanephros organ culture, a weak binding had probably occurred, leading to an increase of free ceftriaxone within the embryonic kidney and to increased potential toxic effects, particularly striking at high concentrations. This highlights a developmental toxicity for ceftriaxone that is unlikely to occur in vivo. Another typical feature of penicillins and cephalosporin-related renal injury is tubulointerstitial nephritis (40,50,51). A clear interstitial inflammation was detected in the medulla of young rats born to ceftriaxone-treated mothers. This β-lactam is known to have a prolonged elimination half-life compared to amino penicillins, especially within the kidney, which may account for this feature (52). The edema was prominent as confirmed by the increased kidney weight. Whether this interstitial inflammation will be subsequently reversed remains to be demonstrated. However, due to frequent hypersensibility reaction development, reexposure to this drug should be avoided (31,40,51).

Penicillins cross the human placenta by simple diffusion and are well absorbed by the fetal tissues (i.e., the percentage of absorbed drug is considerably greater than in the adult) (8). Consequently, these drugs may persist for prolonged periods within the fetus. During organogenesis, a period of rapid embryonic differentiation, very low concentrations of toxic drugs may have devastating effects (53). And this is favored by a weak capacity for renal elimination and hepatic degradation in the fetus. Penicillins and cephalosporins have long been considered safe in pregnancy due to the absence of known fetal toxicity (2). However, none of these drugs has been classified in the class A of Food and Drug Administration fetal risk drug categories, therefore indicating that their absolute lack of developmental toxicity has not been demonstrated (54,55,56). Even if it is clear that they are not teratogenic, they may induce permanent, although silent at birth, renal defects. It has been proposed that a significant portion of human renal pathologies may have their origins in insults that occurred in utero (17,45).

In conclusion, we acknowledge that it is difficult to resolve the question of a β-lactam-mediated fetal nephrotoxicity in humans by using animal models or cell/organ cultures. Species differences in handling of the drugs and thresholds of toxicity make the applicability of such studies to humans unclear. However, comparisons of rat and human developmental toxicity databases have yielded frequent overlaps, and laboratory animals may carry weight in predicting human developmental toxicity (57,58). The assessment of potential drug-induced renal maldevelopment in the fetus is urgently needed, but unfortunately, it remains a difficult task (59). Therefore, caution is needed, and the use of β-lactam antibiotics in obstetrics should be carefully monitored.

Dr. Nathanson was supported by fellowships from the Société des Eaux Minérales Evian and from the Fondation pour la Recherche Médicale. The authors thank D. Droz for expert anatomo-pathology diagnosis.

This work was presented in part at the French Society of Pediatric Nephrology in Brest and at the 30th annual meeting of the American Society of Nephrology, November 2-5, 1997, San Antonio, TX, and has been published in abstract form (J Am Soc Nephrol 8: 359A, 1997).

American Society of Nephrology

1. McNamara P, Stoeckel K, Ziegler W: Pharmacokinetics of ceftriaxone following intravenous administration of a 3-g dose. Eur J Clin Pharmacol 22:71 -75, 1982
2. Chow AW, Jewesson PJ: Pharmacokinetics and safety of antimicrobial agents during pregnancy. Rev Infect Dis7: 287-313,1985
3. Heikkila[Combining Diaeresis] A, Erkkola R: Review of β-lactam antibiotics in pregnancy. Clin Pharmacokinet 27:49 -62, 1994
4. Nau H: Clinical pharmacokinetics in pregnancy and perinatology. II. Penicillins. Dev Pharmacol Ther10: 174-198,1987
5. Pacifici G, Nottoli R: Placental transfer of drugs administered to the mother. Clin Pharmacokinet28: 235-269,1995
6. Vallance P: Drugs and the fetus: Caution is needed in all women of childbearing age. Br Med J 312:1053 -1054, 1996
7. Biro L, Ivan E, Elek E, Arr M: Data on the tissue concentration of antibiotics in man: Tissue concentrations of semi-synthetic penicillins in the fetus. Int J Clin Pharmacol 3:321 -324, 1970
8. Mirkin B: Perinatal pharmacology. Anaesthesiology 43:156 -170, 1975
9. Bernard B, Abate M, Thielen PF, Attar H, Ballard CA, Wehrle PF: Maternal-fetal pharmacological activity of amikacin. J Infect Dis 135: 925-932,1977
10. Fabre J, Blanchard P, Rudhardt M: Pharmacokinetics of ampicillin, cephalothin and doxycycline in various tissues of the rat. Chemotherapy 23:129 -141, 1977
11. Saxén L: Organogenesis of the Kidney, Cambridge, Cambridge University Press,1987
12. Bard JBL, McConnell JE, Davies JA: Towards a genetic basis for kidney development. Mech Dev48: 3-11,1994
13. Vainio S, Muller U: Inductive tissue interactions, cell signaling, and the control of kidney organogenesis. Cell90: 975-978,1997
14. Lechner MS, Dressler GR: The molecular basis of embryonic kidney development. Mech Dev 62:105 -120, 1997
15. Merlet-Bénichou C, Gilbert T, Vilar J, Moreau E, Freund N, Lelie[Combining Grave Accent]vre-Pégorier M: Nephron number: Variability is the rule. Causes and consequences. Lab Invest 79:515 -527, 1999
16. Brenner BM: Nephron adaptation to renal injury or ablation. Am J Physiol 249:F324 -F337, 1985
17. Merlet-Bénichou C, Vilar J, Lelie[Combining Grave Accent]vre-Pégorier M, Moreau E, Gilbert T: Fetal nephron mass: Its control and its deficit. In: Advances in Nephrology from the Necker Hospital, edited by Gru[Combining Diaeresis]nfeld J, Bach J, Kreis H, St. Louis, Mosby-Year Book, 1997, pp20 -45
18. Merlet-Bénichou C, Vilar J, Lelie[Combining Grave Accent]vre-Pégorier M, Gilbert T: Role of retinoids in renal development: Pathophysiological implication. Curr Opin Nephrol Hypertens8: 39-43,1999
19. Gilbert T, Nabarra B, Merlet-Bénichou C: Light and electron microscopic analysis of the kidney in newborn rats exposed to gentamicin in utero. Am J Pathol 130:33 -43, 1988
20. Gilbert T, Lelie[Combining Grave Accent]vre-Pégorier M, Merlet-Bénichou C: Immediate and long-term renal effects of fetal exposure to gentamicin. Pediatr Nephrol 4:445 -450, 1990
21. Gilbert T, Lelie[Combining Grave Accent]vre-Pégorier M, Merlet-Bénichou C: Long-term effects of mild oligonephronia induced in utero by gentamicin in the rat. Pediatr Res 30:450 -456, 1991
22. Gilbert T, Cibert C, Moreau E, Géraud G, Merlet-Bénichou C: Early defect in branching morphogenesis of the ureteric bud in induced nephron deficit. Kidney Int. 50:783 -795, 1996
23. Avner ED, Ellis D, Temple T, Jaffe R: Metanephric development in serum free organ culture. In Vitro Cell Dev Biol18: 675-682,1982
24. Gilbert T, Gaonach S, Moreau E, Merlet-Bénichou C: Defect of nephrogenesis by gentamicin in rat metanephric organ culture. Lab Invest 70:656 -666, 1994
25. Merlet-Bénichou C, Gilbert T, Vilar J, Moreau E: Rat metanephric organ culture in terato-embryology. Cell Biol Toxicol 12:305 -311, 1996
26. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Protein measurement with the Folin phenol reagent. J Biol Chem193: 265-275,1951
27. Larson E, Howlett B, Jagendorf A: Artificial reductant enhancement of the Lowry method for protein determination. Anal Biochem 155:243 -248, 1986
28. Smith SM, Cartwright MM: Spatial visualization of apoptosis using a whole-mount in situ DNA end-labeling technique. Bio-Techniques 22:832 -834, 1997
29. Williams G, Smith C, McCarthy N, Grimes E: Apoptosis: Final control point in cell biology. Trends Cell Biol2: 263-267,1992
30. Savill J: Apoptosis and the kidney. J Am Soc Nephrol 5: 12-21,1994
31. Jones D, Gaber L, Nilsson G, Brewer E, Stapleton F: Acute renal failure following amoxicillin overdose. Clin Pediatr32: 735-739,1993
32. Primack W, Rahman S, Pullman J: Acute renal failure associated with amoxicillin and ibuprofen in an 11-year-old boy. Pediatr Nephrol 11:125 -126, 1997
33. Belko J: Amoxicillin overdose manifested by hematuria and acute renal failure. Pediatr Infect Dis J14: 917-919,1995
34. Coles HS, Burne JF, Raff MC: Large-scale normal cell death in the developing rat kidney and its reduction by epidermal growth factor. Development 118:777 -784, 1993
35. Koseki C, Herzlinger D, Al-Awqati Q: Apoptosis in metanephric development. J Cell Biol 119:1327 -1333, 1992
36. Veis DJ, Sorenson CM, Shutter JR, Korsmeyer SJ: Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75:229 -240, 1993
37. Sorenson CM, Rogers SA, Korsmeyer SJ, Hammerman MR: Fulminant metanephric apoptosis and abnormal kidney development in bcl-2-deficient mice. Am J Physiol 268:F73 -F81, 1995
38. Nagata M, Nakauchi H, Nakayama K, Nakayama K, Loh D, Watanabe T: Apoptosis during an early stage of nephrogenesis induces renal hypoplasia in bcl-2-deficient mice. Am J Pathol148: 1601-1611,1996
39. Tattersall M, Battersby G, Spiers A: Antibiotics and hypokalaemia. Lancet 1:630 -631, 1972
40. Appel GB, Neu HC: The nephrotoxicity of antimicrobial agents. N Engl J Med 296:663 -670, 1977
41. Crocker J, Vernier R: Fetal kidney in organ culture: Abnormalities of development induced by decreased amounts of potassium. Science 169:485 -487, 1970
42. Crocker J: Human embryonic kidneys in organ culture: Abnormalities of development induced by decreased potassium. Science181: 1178-1179,1973
43. Bortner CD, Hughes FM Jr, Cidlowski J: A primary role for K+ and Na+ efflux in the activation of apoptosis. J Biol Chem 272:32436 -32442, 1997
44. He C, Zalups R, Henderson D, Striker G, Striker L: Molecular analysis of spontaneous glomerulosclerosis in Os/+ mice, a model with reduced nephron mass. Am J Physiol 269:F266 -F273, 1995
45. Brenner BM, Garcia DL, Anderson S: Glomeruli and blood pressure: Less of one, more of the other? Am J Hypertens1: 335-347,1988
46. Hinchliffe SA, Lynch MRJ, Sargent PH, Howard CV, Van Velzen D: The effect of intrauterine growth retardation on the development of renal nephrons. Br J Obstet Gynaecol99: 296-301,1992
47. Merlet-Bénichou C, Leroy B, Gilbert T, Lelie[Combining Grave Accent]vre-Pégorier M: Retard de croissance intra-utérin et déficit en néphrons. Médecine-Sciences9: 777-780,1993
48. Merlet-Bénichou C, Gilbert T, Muffat-Joly M, Lelie[Combining Grave Accent]vre-Pégorier M, Leroy B: Intrauterine growth retardation (IUGR) leads to a permanent nephron deficit in the rat. Pediatr Nephrol8: 175-180,1994
49. Ostri P, Frimodt-meller C: Concentrations of mecillinam and ampicillin determined in serum and renal tissue: A single-dose pharmacokinetic study in patients undergoing nephrectomy. Curr Med Res Opin 10: 117-121,1986
50. Appel GB, Garvey G, Silva F: Acute interstitial nephritis due to amoxicillin therapy. Nephron27: 313-315,1981
51. Ruley EJ, Lisi LM: Interstitial nephritis and renal failure due to ampicillin. J Pediatr 84:878 -881, 1974
52. Kuhn H, Angehrn P, Havas L: Autoradiographic evidence for penetration of 3H-ceftriaxone (rocephin) into cells of spleen liver and kidney of mice. Chemotherapy 32:102 -112, 1986
53. Palmisano PA, Polhill RB: Fetal pharmacology. Pediatr Clin North Am 19:3 -20, 1972
54. Friedman J, Little B, Brent R, Cordero J, Hanson J, Shepard T: Potential human teratogenicity of frequently prescribed drugs. Obstet Gynecol 75:594 -599, 1990
55. Hedstrom S, Martens M: Antibiotics in pregnancy. Clin Obstet Gynecol 36:886 -892, 1993
56. Ledger W: Use and abuse of antibiotics in obstetric practice. In: Obstetrics and Perinatal Infections, St. Louis, Mosby-Year Book, 1993
57. Jelovsek F, Mattison D, Chen J: Prediction of risk for human developmental toxicity: How important are animal studies for hazard identification? Obstet Gynecol74: 624-636,1989
58. Ghanooni M, Mattison DR, Zhang YP, Macina OT, Rosenkranz HS, Klopman G: Structural determinants associated with risk of human developmental toxicity. Am J Obstet Gynecol176: 799-805,1997
59. Koren G, Pastuszak A, Ito S: Drugs in pregnancy. N Engl J Med 338:1128 -1137, 1998
Copyright © 2000 The Authors. Published by Wolters Kluwer Health, Inc. All rights reserved.