Placenta 34 (2013) 1216e1222

Contents lists available at ScienceDirect

Placenta journal homepage: www.elsevier.com/locate/placenta

Effect of lipid composition of cationic SUV liposomes on materno-fetal transfer of warfarin across the perfused human term placenta R. Bajoria a, b, *, S. Sooranna a, R. Chatterjee b a b

Imperial College, School of Medicine, Chelsea and Westminster Hospital, London, UK University College London, Institute for Women’s Health London, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Accepted 7 October 2013

Introduction: Use of drugs that cross the placenta freely are currently avoided during pregnancy. We investigated whether cationic small unilamellar (SUV) liposomes of different lipid compositions could prevent the transfer and uptake of warfarin across human term placenta. Methods: Cationic liposomes encapsulated warfarin was prepared by using lecithin (F-SUV) or sterylamine (S-SUV) with cholesterol and stearylamine. The size distribution, encapsulation efficiency, and stability were determined in blood-based media. The transfer kinetics of free and liposomally encapsulated warfarin were studied in a dually perfused isolated lobule of human term placenta with creatinine. Concentrations of warfarin were measured by fluorimetry. Data are expressed as % of initial dose added and given as mean  sd. Results: Warfarin crossed the placenta freely (14.9  1.1%). Trans placental transfer of warfarin was significantly reduced by F-SUV (6.4  0.6%; P < 0.001) and S-SUV liposomes (5.0  0.8%; P < 0.001). Placental uptake of F-SUV (6.3  1.7%; P < 0.001) was greater than that of S-SUV liposomes (2.2  0.5%; P < 0.001). Conclusion: Our data suggest that cationic liposomes reduce trans placental transfer of warfarin. If confirmed “in vivo”, liposomes might provide an alternative non-invasive method of drug delivery to the mother. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Liposome Warfarin Surface charge Small unilamellar

1. Introduction Prescribing drugs during pregnancy is an unusual risk-benefit situation. Drugs that may be of benefit or even life saving for the mother can be harmful to the fetus. This often leads to under prescribing for pregnant women [1]. With recent advancement in medical care, more and more women of childbearing age with serious medical conditions are embarking on pregnancy [2,3]. This poses a therapeutic dilemma to both the patients and health care professionals regarding use of drugs for improvement of maternal outcome. Unfortunately much information about drug safety for the fetus is limited to sporadic case reports, as clinical trials in pregnancy are only undertaken in special circumstances [2]. Animal studies have limited applicability to humans because of speciesspecific variability. The safe therapeutic concentrations in the maternal circulation may not necessarily indicate adequate levels

* Corresponding author. Institute For Women’s Health, University College London, 86-96 Chenies Mews, London WC1E 6HX, UK. Tel.: þ44 (0) 2076796582; fax: þ44 (0) 2073809984. E-mail addresses: [email protected], [email protected] (R. Bajoria). 0143-4004/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.placenta.2013.10.005

in the fetus, because trans placental passage depends on the pharmacological properties of the drug, including protein binding, metabolism by the placenta, and pharmacogenetic effects which can modify drug disposition in the fetus [4]. Although teratogenic effects may result from different mechanisms, it is always secondary to transfer of drugs from maternal to fetal circulation [1,5]. The resulting fetal death, malformation, growth restriction and damage to a specific organ system are due to undesirable fetal exposure to potent therapeutic agents. This often limits maternal drug administration for prevention of significant morbidity [5]. Thus there is a strong need for development of a drug delivery system, which may have a potential to maintain therapeutic concentrations in the maternal circulation with minimal trans placental transfer. In the recent years, drug delivery systems (DDS) have been developed for achieving a better clinical response and tolerability of anticancer agents [6,7]. Various nanostructures, including liposomes, polymers, dendrimers, silicon or carbon materials, and magnetic nanoparticles, have been tested to improve the therapeutic properties of drugs [8e10]. Recent studies using fluorescently labeled polystyrene beads with diameter up to 240 nm

R. Bajoria et al. / Placenta 34 (2013) 1216e1222

were shown to cross the human perfused placenta without affecting the viability of the placental explant [11]. However, there is concern that NP might induce adverse physiological effects and impede embryogenesis [12] as pegylated gold nanoparticles 10e 30 nm six particles have been shown to be retained within the trophoblast layer of the placenta [13]. In contrast, liposomes are made of biodegradable lipids with minimal or no toxicity and have been investigated extensively [7,8]. Liposomes can be administered orally or parenterally, and their tissue distribution can be modulated by altering their size, surface charge, and lipid composition [14e17]. The incorporation of drug molecules into liposomes offers possibilities of targeting and controlled release. When administered intravenously, liposomes are mainly cleared from the circulation by liver [17e19]. Liposomes have been reported to increase the solubility of drugs and improve their pharmacokinetic properties, such as rapid metabolism, reduction of harmful side effects and increase the therapeutic index of chemotherapeutic agents [9]. Several clinical trials have shown that liposomal doxorubicin is effective to treat AIDS-related Kaposi’s sarcoma and ovarian cancer with improved clinical efficacy and tolerability [6]. Currently a number of anticancer drugs such as, daunorubicin, paclitaxel, vincristine, amphotericin B have been encapsulated in liposomes and their use have been approved for clinical use [6,20]. We have previously shown that size of liposomes can modulate placental transfer of inert polar molecules [21]. The uptake of small unilamellar liposomes (SUV) depends upon their surface charges in that cationic liposomes prevent uptake by the placenta while anionic liposomes enhance their uptake [22,23]. Furthermore, the placental uptake of small anionic liposomes depend upon their lipid composition in that fluid liposomes are internalized more readily than solid ones [24]. Based on this information, in this study, we evaluated transfer and uptake of positively charged SUV liposomes made either from unsaturated or saturated phospholipids in an “in vitro” model of dually perfused isolated lobule of human term placenta using warfarin as a model drug. 2. Materials and methods 2.1. Materials Chromatographically pure egg phosphatidylcholine, distearoyl phosphatidylcholine, in 2:1 chloroform-methanol (2:1) and stearylamine were purchased from Lipid Products (Nutfield, UK). Cholesterol (Chol) was obtained from Sigma Chemical Co. (Pool, UK). Sephadex G-25 was obtained from Pharmacia (Herts, UK). Carboxyfluorescein (CF) was purchased from Eastman Kodak (Hemel Hempstead, UK). Sodium warfarin was obtained as a gift from Glaxo. 2.2. Preparation of liposomes Cationic SUV liposomes were prepared from unsaturated (F-SUV; phosphatidylcholine-cholesterol-sterylamine, 5:5:0.05) and saturated phospholipids (S-SUV; distearoyl phosphatidylcholine-cholesterol-sterylamine, 5:5:0.05) by standard techniques of hydration of the dried lipid film with warfarin [23]. The lipid suspension was sonicated and free warfarin were separated from those encapsulated within liposomes by gel filtration on a Sephadex G-25 column (45  1 cm) equilibrated with Tris buffer, with a pH of 7.4. The phospholipid and cholesterol contents of each preparation were determined. The percentage of warfarin encapsulated per mol of lipid was calculated by determining the latency as described previously [21]. Concentration of total warfarin was determined by adding 1% Triton X-100 to 10 mL liposomal preparation in buffer. The encapsulation efficiency of warfarin was expressed as the percentage of the initial amount added. The specific encapsulation was expressed as the percentage of entrapped solute per nmol liposomal phospholipid. The size and number of lamellae of the liposomes were determined by negative staining with 1% ammonium molybdate under a JEOL 100 CX electron microscope (JEOL, Peabody, MA) [21].

1217

phosphate-buffered saline (PBS) at 37  C (n ¼ 10). Serial 100-mL samples were obtained at 15-min, 120 min and 180 min and centrifuged (1500 g, 10 min) in 4 mL PBS to determine the latency of warfarin in the supernatant. 2.4. Placental perfusion technique We present data obtained from 19 placental perfusion experiments. Seven experiments were undertaken with warfarin only, and 6 experiments each using FSUV and S-SUV encapsulated warfarin. Placentas were obtained immediately after vaginal or cesarean delivery from uncomplicated pregnancies of more than 37 weeks gestation. Dual closed circuit perfusion of the isolated lobule was commenced within 5e10 min at 37  C under optimal physiological conditions of oxygenation, pressure, flow, osmotic pressure, and acid/base status [21,25]. Closed circuit perfusion of the feto-placental circulation was established with a perfusion pressure of 40e50 mm Hg and a venous outflow of 6e9 mL/min. Fetal perfusates composed of autologous cord blood diluted with modified tissue culture medium 199 with a median hematocrit of 14 (range, 12e18) and a circulating volume of 110e120 mL. Materno-placental circulation was established with an arterial pressure of 15e18 mm Hg and a flow rate of 24e30 mL/min by placing five cannulas in the intervillous space. The maternal perfusates consisted of autologous blood collected from the intervillous space and diluted with modified tissue culture medium 199 [25] with a median hematocrit of 6 (range, 4e9) and a circulating volume of 150e160 mL. Both maternal and fetal perfusates were diluted to achieve the circulating volume as well as to reduce the viscosity of the autologous maternal or fetal blood to facilitate circulation through membranous oxygenator [25]. Tissue oxygenation was maintained by oxygenating the maternal circulation with 95% oxygen and 5% carbon dioxide. Creatinine was used as a marker to establish juxtaposition of the feto-maternal circulation [25]. 2.5. Experimental protocol In seven sets of control experiments, 2.1 mg sodium warfarin was added to the maternal circulation at the commencement of the experiments. This concentration of warfarin (0.1 mg/mL) was selected because it corresponded with the mean concentration of warfarin present in liposomal form. In an additional six experiments, either F-SUV liposomally encapsulated warfarin or S-SUV containing warfarin was added to the maternal circulation. Just before the perfusion experiment, liposomally encapsulated warfarin was separated from the free drugs by chromatography on Sephadex G-25. The free warfarin or liposomally encapsulated warfarin with 30 mg creatinine was administered to the maternal arterial cannula distribution head over a period of 6 min (the time required for a single maternal circulation). Two-milliliter fetal and maternal samples were taken every 15 min over 2 h. Additional samples (0.5 mL) of maternal and fetal perfusates were taken for determination of acid-base status at 30 min intervals. All volumes removed were replaced with a fresh perfusate. At the end of the perfusion period, both circuits were drained, and their volumes were measured. All samples were centrifuged (3000 g, 15 min), and the supernatants were aliquoted into 0.5- and 1.5-mL volumes. The 0.5-mL aliquot was stored at 20  C for creatinine assay. Liposomal stability was determined at each sample point by measuring warfarin latency in a 1.5-mL aliquot. Liposomal uptake was determined by homogenising the perfused placenta as described previously [21]. An aliquot of the homogenate was centrifuged (3000 g, 15 min), and warfarin concentrations were measured in the supernatant. The concentrations of warfarin in the maternal and fetal circulations and in the placenta were expressed as a percentage of the dose added after correction of background activity, the circuit volume, and the amount removed from the previous sample. At the end of the experiment, 2 mL maternal and fetal perfusates were applied to a Sephadex G-25 column (45  2 cm) to fractionate liposomally encapsulated one from free warfarin at room temperature with Tris-saline buffer. The elution rate was 0.63 mL/min. One-milliliter fractions were collected and assayed for warfarin, phosphatidylcholine, distearoyl phosphatidylcholine and cholesterol. 2.6. Analytical methods The concentration of warfarin was measured fluorometrically at excitation and emission wavelengths of 490 and 520 nm, respectively, with a sensitivity of 1 nm/mL and coefficients of variation of 4e7%. The phospholipid and cholesterol contents of the liposomes were assayed by colorimetry with a sensitivity of 5 mg/mL as described previously [21]. The creatinine concentration was determined by colorimetric assay [25], with a coefficient of variation of 7e12%. 2.7. Data analysis

2.3. Assessment of permeability of warfarin-containing liposomes in biological media The stability of liposomes over 4 h was determined by incubating 1 mL liposomal preparation (0.5 mmol phospholipid) in 5 mL diluted maternal blood or

All values were expressed, as the mean  SEM. Equilibrium between maternal and fetal circuits was determined when fetal/maternal ratios of the drug levels were close to unity. The integrated values of maternal (MAUC) and fetal (FAUC) concentrations of a drug were calculated by using the trapezoidal rule [21]. Data between

1218

R. Bajoria et al. / Placenta 34 (2013) 1216e1222

groups was compared using Student’s T test. P < 0.05 was considered statistically significant.

3. Results The SUV liposomes of both lipid compositions were unilamellar and had uniform size distribution, with a mean diameter of 73.6  3.8 nm. The size distribution was determined by transmission electron microscopy and as images have been published before no further images were taken for these experiments [22]. The percent encapsulation of warfarin per mol phospholipid was comparable to that of fluid (F-SUV) and solid (S-SUV) liposomes (Table 1). The percentage encapsulation of warfarin was more than that of carboxyfluorescein despite having comparable sizes of liposomes. This might be attributed to difference in physicochemical properties of warfarin (lipid solubility) and carboxyfluorescein (negative charge and hydrophilic). The stability of liposomal warfarin in PBS was comparable between F-SUV and SSUV liposomes but was significantly different at 120 and 180 min when incubated in blood based media (Table 2). The transfer of warfarin from maternal to fetal circulation was shown in Fig 1A and this constitutes the control group. The maternal warfarin concentration decreased from 100% at 0 min to 72.7  5.0% at 120 min. The fetal concentration increased linearly from 3.5 þ 1.6% at 15 min to 14.9  1.1% at 120 min, with fetal/ maternal ratio of 0.21  0.02 (Fig. 2A). The MAUC and FAUC values were 7967.8  559% and 890.7  95.4% dose min1, respectively (Table 3). The placental uptake of warfarin was 12.3  3.3% (Table 3). The fetal concentrations of creatinine in these experiments were 16.6  1.5% at 120 min with fetal warfarin to creatinine ratio at 120 min of 0.90  0.07 (Fig 2B). The effects of F-SUV and S-SUV liposomes on the placental transfer and uptake of warfarin are shown in Fig. 1B and C. The maternal concentration of F-SUV liposomal warfarin at 120 min was 88.4  3.0% (P < 0.001) with MAUC of 9642  244% dose min1 (P < 0.001). This was significantly higher than that of the control group (Fig 1B and Table 3). The fetal warfarin level increased linearly from 1.6  0.9% at 15 min to 6.4  0.6% at 120 min and this was lower than the control value (P < 0.001). Similarly, the FAUC (414.3  71.2% dose min1; P < 0.001), fetal/maternal ratio at 120 min (0.07  0.01; P < 0.001), and placental uptake (6.3 þ 1.7% P < 0.05) of F-SUV were lower than those in the control group (Table 3). The concentration of free warfarin in the maternal circuit was 1.6  0.5% at 15 min and 4.2  0.5% at 120 min. The fetal concentration of creatinine at 120 min was 15.4  1.4%, which was comparable to that of the control group. However, fetal warfarin to creatinine ratio at 120 min was 0.44  0.06, which was significantly lower than that of the control group (P < 0.001) (Fig 2). The maternal concentration of S-SUV warfarin at 120 min 94.3%  2.1% was higher than the control value (P < 0.001) (Fig 1C). The MAUC level in the S-SUV group (10119.7  161.9% dose min1) was higher than that of the control group (P < 0.001). The fetal concentration increased from 0.6  0.3 at 15 min to 5.0  0.8 at 120 min, which was significantly lower than the control group

Table 2 Stability of liposomes of in blood based media and buffer saline. Types of liposomes

In blood based media 15 min

120 min

180 min

60 min

In PBS 180 min

F-SUV (n ¼ 10) S-SUV (n ¼ 10)

97.9  1.1 99.0  0.8

96.5  2.4* 98  2.3

92.5  2.8*** 95.7  2.3

98  1.8 98  0.9

97  1.8 98  1.1

F-SUV ¼ phosphatidylcholine small unilamellar liposomes. S-SUV ¼ distearoyl phosphatidylcholine; PBS ¼ phosphate buffer saline. *P < 0.05; ***P < 0.001.

(P < 0.001). Similarly, FAUC 292.8  27.9% dose min1 (P < 0.001), fetal/maternal ratio of 0.05  0.8 at 120 min (P < 0.001), FAUC/ MAUC ratio of 0.03  0.003 (P < 0.001) all were significantly lower than the control group (Table 3). The placental uptake of S-SUV liposome was 2.2 þ 0.5%, which was lower than the control group (P < 0.001). The stability of liposomes in the maternal circulation at 15 min was comparable to that of F-SUV liposomes with free warfarin levels of 1.6  0.3% .At 120 min, free warfarin levels was significantly less than F-SUV at 120 min (3.0 þ 0.8%; P < 0.05). The fetal concentration of creatinine at 120 min was 16.3  1.8%, which was similar to the control and F-SUV group. But fetal warfarin to creatinine ratio at 120 min (0.31  0.05) was significantly lower than the control group (P < 0.001). The maternal concentration (P < 0.01) and MAUC levels (P < 0.01) of S-SUV liposomal warfarin were higher than those of FSUV, whereas the fetal concentration (P < 0.01), FAUC (P < 0.01), fetal/maternal ratio (P < 0.01) and fetal warfarin to creatinine ratio (P < 0.01) were significantly lower when compared to F-SUV. Placental uptake of F-SUV was markedly higher than that of S-SUV (P < 0.01) (Fig 1 and Table 3). 3.1. Chromatography The chromatogram of the maternal perfusates containing F-SUV liposomes showed two distinct peaks of warfarin. The peak immediately after the void volume was due to warfarin-entrapped liposomes, as confirmed by latency, liposomal phospholipid, and cholesterol content. The second peak was due to free warfarin (Fig 3A). The liposomal and free warfarin concentrations were 82.6  3.9 and 4.1  2.2%, respectively. In the fetal perfusate (Fig 3B) no intact liposomes were found, and there was a single peak that corresponded to free warfarin (7.0  1.8%). Fig 3C shows chromatogram of the control experiment for reference. Please note that both in maternal and fetal perfusates there was only one peak. The chromatograms of maternal and fetal, with S-SUV were essentially similar to those of F-SUV. 4. Comments This study shows that warfarin crosses the human placenta in significant quantities, but its transfer is significantly reduced by encapsulation of warfarin within liposome. Our findings that warfarin crosses the human term placenta freely is consistent with

Table 1 Characteristics of liposomes of different lipid composition. Types of liposomes

Initial conc of warfarin (mM)

S-SUV (n ¼ 10) F-SUV (n ¼ 10)

300 300

Conc. of PC/DSPC (mmol)

Conc. of cholesterol (mmol)

Added

Recovered

Added

Recovered

75 75

71.5  2.2 70.6  3.5

75 75

70  3.8 71.8  4.4

Encapsulation of warfarin (%)

Specific encapsulation of warfarin

Latency of warfarin

1.8 1.6

24.5  4.5 22  5.1

99.3 98.5

Specific encapsulation ¼ Percentage of warfarin entrapped per nmol lipids. Conc ¼ concentration; PC ¼ phosphatidylcholine DSPC ¼ distearoyl phosphatidylcholine. All values are expressed as the mean  SD.

R. Bajoria et al. / Placenta 34 (2013) 1216e1222

1219

Fig. 1. Materno-fetal transfer of warfarin at the commencement of the experiments following a single bolus dose to the maternal circulation. Maternal (e) and fetal (BeB) concentrations are expressed as percent of initial dose added. In these experiments both maternal and fetal perfusates were re-circulated. Fetal concentration of creatinine used as a marker in all experiments is also shown for reference (-D-). A shows transfer from maternal (C) to fetal (B) circulation of warfarin alone. This constitutes the control data. B and C show the transfer of F-SUV and S-SUV liposome-encapsulated warfarin respectively. Fig. 2A compares the feto-maternal ratio of warfarin in the control (e), F-SUV (-■-) and S-SUV (-☐-) groups of experiments. All values are expressed as a percentage of the initial dose added to the maternal circulation and are the mean of seven experiments.

Fig. 2. The fetal warfarin to creatinine ratio in control, F-SUV and S-SUV groups over a 2 h period. All values are expressed as a percentage of the initial dose added to the maternal circulation and are the mean of seven experiments. Control (e); F-SUV (-■-); and S-SUV (-☐-).

Table 3 Comparison of placental concentration, maternal and fetal AUC, of warfarin containing liposomes of different lipid composition with control data. Types of lLiposomes

Placental (% dose)

FAUC (% dose min1)

MAUC (% dose min1)

FAUC/MAUC ratio

F-SUV (n ¼ 7) S-SUV (n ¼ 7) Warfarin (n ¼ 7)

6.3  1.7** 2.2  0.50 12.3  3.3A

414.3  71.2** 292.8  27.9 890.7  95.4AAA

9642  244** 10119.7  161.9 7967.8  558.7AAA

0.04  0.01** 0.03  0.003 0.12  0.02 AAA

MAUC ¼ maternal area under the curve; FAUC ¼ fetal area under the curve. F-SUV vs. S-SUV **P < 0.01. F-SUV vs. Warfarin AAAP < 0.001; AP < 0.05.

1220

R. Bajoria et al. / Placenta 34 (2013) 1216e1222

Fig. 3. Chromatogram of (A) maternal, (B) fetal perfusate, and (C) maternal (e) and fetal perfusate (BeB) of control group on Sephadex G-25. Concentration of liposomal lipids is shown as dotted line in A and B. In A, the first peak is that of liposomal encapsulated warfarin and the second peak is that of free warfarin. In B, only one peak for free warfarin is found. Results for only F-SUV liposomes are shown.

the clinical observation that use of warfarin during pregnancy is associated with risk of still birth, congenital anomalies and intracranial haemorrhage [26,27]. The relatively small molecular size of warfarin (330 Da), extensive lipid solubility all favors its free transfer across the placenta by simple diffusion. Therefore we used warfarin as a model drug to test our hypothesis that liposomes can prevent transfer of a lipid soluble small molecule drug across the placenta. Furthermore, warfarin is 4e5 times more effective in preventing thrombosis than heparin [27]. This is especially applicable after recent awareness that use of heparin alone is associated with 29e33% risk of life- threatening thrombosis and 7e15% risk of mortality in mothers with mechanical heart valves [28,29]. Consequently, warfarin and other vitamin K antagonists are the preferred anticoagulants to minimize maternal morbidity and mortality from thromboembolism. We studied liposomal warfarin transport in an “in vitro” model of an isolated lobule of dually perfused human term placenta. Previously we have used the perfused placental to determine the permeability of molecules such as heparin, TSH, thyroxine and TRH and obtained results that corroborate those of “in vivo” studies [24,25,30]. Dual perfusion of a single placental lobule is the only experimental model to study human placental transfer of substances. In general, the fetal-to-maternal drug concentration ratios matched well between placental perfusion experiments and in vivo samples taken at the time of delivery of the infant [31,32].

Our data suggest that irrespective of lipid composition, liposome encapsulation significantly reduces transfer of warfarin from the maternal to the fetal circulation. Reduced transfer of liposomal warfarin could be attributed to the fact that little or no free warfarin is available at the maternal placental interface. This was unlikely to be due to failure of establishment of juxtaposition between maternal and fetal circulations, because in all of our experiments the transport of a freely diffusible molecule creatinine was within the previously published range of 11e17%. This study suggests that intact liposomes failed to cross the human placenta because intact liposomes or liposomal lipids were undetectable in the fetal circulation. These finding concord with our previous observations that intact liposomes do not cross the human placental barrier [21,24]. More recently, Kaga et al. [33], have also shown that liposome-encapsulated hemoglobin (hemoglobin-vesicle) was not transferred from the mother to the fetus at a late stage of pregnancy in the rat model. It was highly unlikely that liposomes of 70 nm size could cross the placental capillary bed as human fetal capillaries have an intracellular cleft of

Effect of lipid composition of cationic SUV liposomes on materno-fetal transfer of warfarin across the perfused human term placenta.

Use of drugs that cross the placenta freely are currently avoided during pregnancy. We investigated whether cationic small unilamellar (SUV) liposomes...
652KB Sizes 0 Downloads 0 Views