Transgenic Research 12: 369–373, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

369

Short Communication

Placental anomalies and fetal loss in mice, after administration of doxycycline in food for Tet-system activation Ren´e Moutier, Françoise Tchang, St´ephane M. Caucheteux & Colette Kanellopoulos-Langevin∗ Laboratory of Immune Regulations and Development, Jacques Monod Institute, UMR 7592 (CNRS and Universities Paris 6 and 7), 75251 Paris Cedex 05, France Received 26 September 2002; revised 22 November 2002; accepted 28 November 2002

Key words: developmental toxicity, doxycycline in food, placental defect, Tet-system, transgenic mice

Abstract During the course of a study aiming to obtain a tetracycline (Tet)-inducible transgene expression restricted to the placenta, we have observed a toxicity of doxycycline (dox) given in the food at doses of 2.5–10 mg/g to pregnant mice from two different inbred strains. During the second half of gestation, dox-fed non-transgenic mice presented placental anomalies and impaired fetal development proportional to the dose of antibiotic. Thus, dox administered in commonly used food doses can have an adverse effect on pregnancy. These observations are important for studies of placental or fetal development using inducible gene promoters. Introduction Transgenic animals have provided invaluable tools to elucidate gene functions and regulations. More recent refinements have included tissue-restricted transgene expression and inducible/repressible gene expression via an exogenous molecule. One of the first described and most widely used systems so far, the tetracycline (Tet)-system (Kistner et al., 1996) has allowed an efficient control of transgene expression in various mammalian tissues (Lewandoski, 2001). Doxycycline (dox), a water-soluble tetracycline derivative, readily crosses the placenta and via the Tet-systems, can efficiently regulate transgene expression during mouse development (Shin et al., 1999; Fedorov et al., 2001; Perl et al., 2002a,b). Our long-standing interest in fetal–maternal interactions has led us to use the Tet ‘on’ system to trigger the expression of relevant transgenes in the fetoplacental unit, and observe the consequences on maternal tolerance as well as fetal survival and development. In a systematic study, dox was given to nontransgenic pregnant females, either in their drinking ∗ Author for correspondence

E-mail: [email protected]

water or in their food at doses cited in the literature for gene induction (Mansuy & Bujard, 2000; Hasan et al., 2001; Perl et al., 2002a,b). While administration in drinking water at doses up to 5 mg/ml had no adverse effect, we observed that dox given from 2.5 to 10 mg/g of wet food during the second half of gestation resulted in a dose-dependent fetoplacental toxicity.

Materials and methods Dox treatments Mice were maintained according to institutional regulations in either conventional or specific pathogen-free environments. Six- to eight-week-old CBA/Ca and FVB females were mated overnight with males of the same strain. The next morning, females with a vaginal plug were identified. Two groups of control FVB and CBA/ca mice were left untreated (dox 0). One group of FVB mice received dox at 5 mg/ml from days 8 to 16 post coitum (pc) in drinking water, protected from light, with 7.5% sucrose added to mask the bitter taste. Other groups received dox at various times of pregnancy (days 6–8, 8–12, 8–16 and 12–16 pc) at concentrations of 2.5, 5 or 10 mg/g of wet food.

370 Dox and sucrose were dissolved in sterile water and mixed with the appropriate amount of standard food pellets to obtain a wet mixture renewed every day. Fetoplacental unit analysis Fetoplacental units were removed from each treated or control female on day 17 pc and classified as normal (N) or abnormal (AN) according to the size and appearance of both fetus and placenta. Statistical analysis AN and N fetoplacental units were scored and percentages of AN fetoplacental units were compared using the χ2 test. Histological analysis Placentae were processed for histology by fixation in Bouin’s solution or 4% paraformaldehyde in phosphate-buffered saline, followed by paraplast embedding and sectioning. Sections 8–10 µm thick were stained with Mayer’s hematoxylin-eosin (HE) or phloxine B, hematoxylin, tannin, azure B (PHTA) stain (Kurz & Wittekind, 2001).

Results and discussion Developmental toxicity of dox given in food Groups of non-transgenic FVB or CBA females received increasing doses (0, 2.5, 5 or 10 mg/g) of dox in food only. The frequency of AN fetoplacental units in each group was evaluated on day 17 pc. Results are presented in Table 1. Untreated FVB and CBA females (dox 0) had less than 5% AN. FVB females receiving dox at 5 mg/ml in drinking water from days 8 to 16 pc had percentages similar to controls. Females treated from days 8 to 16 pc with 2.5 mg/g of dox in the food produced an average of 20% AN (p < 0.01). At the dose of 5 mg/g, this percentage increased to 78%. Moreover, the same dose of dox for 5 days (days 8–12 pc), was sufficient to get a comparably high percentage of AN (89%), reaching 100% with 10 mg/g. Interestingly, when the intermediate dosage was given for 5 days later in gestation (days 12–16 pc), this regimen produced only 8% AN. The same treatment given early, from days 6 to 8 pc, did not interfere with gestation. Finally, this toxicity was not restricted

Table 1. Percentage of AN fetoplacental units in pregnant females treated with dox Strain

Femalea

Days pc

Dox

Totalb

% ANc

FVB FVB FVB FVB FVB FVB FVB FVB CBA CBA

5 7 3 4 3 3 3 3 4 4

– 8–16 8–16 8–16 8–12 6–8 12–16 8–12 – 8–12

0 5d 2.5e 5e 5e 5e 5e 10e 0 5e

49 67 25 36 28 32 25 32 31 36

2 3 20∗∗ 78∗∗ 89 NSf 0 8 100 3 25∗

a Number of females examined. b Total number of fetoplacental units. c Percentage of AN or resorbed fetoplacental units. d Doxycycline in drinking water: mg/ml. e Doxycycline in food: mg/g of wet food. f Not significantly different. ∗ p < 0.05. ∗∗ p < 0.01.

to the FVB strain, as female mice from the CBA/Ca strain, fed with 5 mg/g dox during 5 days (from 8 to 12 pc) produced 25% AN (p < 0.05). Altogether, these results demonstrate that dox is deleterious to fetal development in mice when it is given in the food in doses of 2.5–10 mg/g, between days 8 and 12 of gestation. It is clear from numerous reports that dox given at 1–5 mg/ml in drinking water yields no adverse effect during gestation (Shin et al., 1999; Fedorov et al., 2001; Hasan et al., 2001; Perl et al., 2002a,b). Our results confirm this finding. However, doses of 2.5– 10 mg/g of dox in food, which had been used safely in non-pregnant mice to enhance reporter gene induction in the brain (Hasan et al., 2001), cause significant fetal anomalies and loss in the second half of gestation. Our data also suggest that the FVB strain is more sensitive than CBA. This toxicity is observed with equal frequency and reproducibility under conventional or specific pathogen-free conditions of animal housing. Such a regimen had been used safely in mice to enhance reporter gene induction in the brain (Hasan et al., 2001). To the best of our knowledge, few studies have previously reported gestational problems associated with tetracyclines in rodents: one study in mice reports problems early in gestation after sub-cutaneous tetracycline injections (Boucher, 1969), and two other papers describe the toxicity of intra-peritoneal injections

371

Figure 1. Morphological and histological fetoplacental abnormalities in mice treated with dox. (A, B) Fetoplacental units (day 17 pc) in the uterus of one FVB female treated with dox (5 mg/g of food, days 8–12 pc), scale bar = 3.5 mm: (A) Phenotypically normal fetoplacental unit, embryo length E = 18 mm, placenta diameter P = 6 mm. (B) Phenotypically AN fetoplacental units. Embryo length E1 = 8 mm, E2 = 3.5 mm, E3 = 6.5 mm. (C–E) Placental sections (day 17 pc) from FVB females treated with dox (5 mg/g of food, days 8–12 pc), L = labyrinth, Sp = spongiotrophoblast, HE staining, scale bar = 500 µm. (C) Phenotypically normal placenta from normal embryo (18 mm), (D) moderate lesions in placenta from embryo (16 mm), (E) more extensive lesions in placenta from embryo (8 mm), (F, G) details of labyrinthine zone, HE staining, scale bar = 100 µm. (F) Control day 17 placenta (G) abnormal day 17 placenta from FVB female treated with dox, arrowheads show apoptotic bodies, (H, I) details of labyrinthine zone, PHTA staining, scale bar = 100 µm. (H) Control day 14 placenta, (I) abnormal day 14 placenta from FVB female treated with dox (5 mg/g of food, days 8–12 pc).

in pregnant rats (Steiner et al., 1965; Rechtman et al., 1999). Although tetracyclines are not teratogenic and have been used in antibiotherapies in humans

including pregnant women, they are now contraindicated throughout pregnancy because of noxious effects on fetal bone and teeth (Schwarz, 1981).

372 Characterization of fetoplacental unit anomalies The morphological aspect of fetoplacental units (day 17 pc) from FVB females treated with an intermediate dose of dox (5 mg/g of food from days 8 to 12 pc) is illustrated in Figure 1(A) and (B). Normal fetoplacental units were observed jointly with abnormal fetoplacental units affected at various degrees. Placentae corresponding to the smallest embryos were also reduced in size (Figure 1(B)). Abnormal fetuses appeared entirely white or with patches of white and pink areas, most of these were clearly arrested in development. With a higher dose of 10 mg/g of food, 80% of complete resorptions were observed. Histology of abnormal placentae Several lines of evidence suggested that placental development could be affected primarily by dox: (1) small placentae (diameter <5 mm) were always found with embryos arrested early in development, (2) normal sized placentae were connected to discolored but otherwise normal looking embryos ranging from 16 (normal size) to 6 mm in length, suggesting some functional placental defect, (3) the window of dox sensitivity (days 8–12 pc) corresponds to the time of placental differentiation and maturation. As shown in Figure 1(D) and (E), both small and normal sized placentae from abnormal concepti appeared markedly affected in their histological structure. On sections from paraffin-embedded day 17 placentae, degenerative lesions were reproducibly observed in the labyrinthine zone of the trophoblast. Their extent was related to the size of the embryo: present at a minor degree in placentae of embryos measuring 14–16 mm (Figure 1(D)) and more extensive with smaller embryos, making a remarkable lacelike pattern with numerous apparently empty spaces (Figure 1(E)). On the border of these spaces, many cells appear reduced in size and flattened, with groups of multiple smaller nuclei highly reminiscent of apoptotic bodies (Figure 1(G), arrowheads). In placental sections from day 14 pc abnormal embryos (5 mg/g dox, days 8–12 pc), the labyrinth appeared deprived of red blood cells (Figure 1(I)) compared to control unexposed placentae (Figure 1(H)), while at the same time presenting a comparable extension (see labyrinth in Figure 1(C)–(E)). This observation, made prior to the onset of degenerative lesions, appears in favor of the hypothesis that dox perturbs the later development of the labyrinth or causes its destruction, rather

than disturbing its early formation. Taken altogether, our results support the conclusion that dox affects placental formation and hence fetal development. Interestingly, Steiner et al. (1965) have reported that intraperitoneal injections of tetracycline in female rats during 5 days at the end of gestation produced abortions and 75% of abnormal placentae with degenerative changes chiefly in the labyrinth. This is in keeping with our own findings of major labyrinth anomalies in mice. Our observations of discolored, anemic but otherwise normal fetuses suggest a mechanism of placental dysfunction leading to blood deprivation. These features were not found in placentae from spontaneous resorptions observed at low frequency (<5%), in the FVB or CBA strains (data not shown). As a possible mechanism, several recent papers (Rechtman et al., 1999; Shlopov et al., 2001) pointed out the inhibitory effect of dox on the production of proinflammatory cytokines, and matrix metalloproteinases known to play a role in tissue remodeling. Intense tissue remodeling occurs during trophoblast invasion and uterine decidualization, which could explain how tetracyclines affect implantation. In our experiments performed at later stages of placental development, tissue remodeling could also be the target of dox. Dox is now available in dry food pellets at 25 mg/g, making it more stable and easier to use in long-term treatments. Our results strongly suggest that such doses are likely to perturb experiments requiring the Tet-induction of gene promoters in pregnant mice or in fetuses. Acknowledgements The collaboration of Guillaume Desanti, Sébastien Paturance and his staff, and Gérard Lefèvre, Michel Thomas, Arancha Pi-Ruiz is gratefully acknowledged. This work has been supported in part by the Fondation pour la Recherche Médicale and the Association pour la Recherche contre le Cancer. References Boucher D (1969) Abortive properties of tetracyclines in mice. Hormone treatments. Arch Sci Physiol (Paris) 23: 481–516 (in French). Fedorov LM, Tyrsin OY, Krenn V, Chernigovskaya EV and Rapp UR (2001) Tet-system for the regulation of gene expression during embryonic development. Transgenic Res 10: 247–258.

373 Hasan MT, Schönig K, Berger S, Graewe W and Bujard H (2001) Long-term, noninvasive imaging of regulated gene expression in living mice. Genesis 29: 116–122. Kistner A, Gossen M, Zimmermann F, Jerecic J, Ullmer C, Lübbert H and Bujard H (1996) Doxycycline-mediated quantitative and tissue-specific control of gene expression in transgenic mice. Proc Natl Acad Sci USA 93: 10933–10938. Kurz H and Wittekind D (2001) A new panoptic stain for developmental biology – the mouse placenta paradigm. Ann Anat 183: 437–441. Lewandoski M (2001) Conditional control of gene expression in the mouse. Nat Rev 2: 743–755. Mansuy IM and Bujard H (2000) Tetracycline-regulated gene expression in the brain. Curr Opin Neurobiol 10: 593–596. Perl AT, Tichelaar JW and Whitsett JA (2002a) Conditional gene expression in the respiratory epithelium of the mouse. Transgenic Res 11: 21–29.

Perl AK, Wert SE, Nagy A, Lobe CG and Whitsett JA (2002b) Early restriction of peripheral and proximal cell lineages during formation of the lung. Proc Natl Acad Sci USA 99: 10482–10487. Rechtman MP, Zhang J and Salamonsen LA (1999) Effect of inhibition of matrix metalloproteinases on endometrial decidualization and implantation in mated rats. J Reprod Fertil 117: 169–177. Schwarz RH (1981) Considerations of antibiotic therapy during pregnancy. Obstet Gynecol 58: 95S–99S. Shin MK, Levorse JM, Ingram RS and Tilghman SM (1999) The temporal requirement for endothelin receptor-B signalling during neural crest development. Nature 402: 496–501. Shlopov BV, Stuart JM, Gumanovskaya ML and Hasty KA (2001) Regulation of cartilage collagenase by doxycycline. J Rheumatol 28: 835–842. Steiner G, Bradford W and Craig JM (1965) Tetracycline-induced abortion in the rat. Lab Invest 14: 1456–1463.