Valproate Inhibits the Conversion of Testosterone to Estradiol and Acts as an Apoptotic Agent in Growing Porcine Ovarian Follicular Cells

Authors


Address correspondence and reprint requests to Dr. E. Taubøll at Department of Neurology, Rikshospitalet, University of Oslo, 0027 Oslo, Norway. E-mail: erik.tauboll@klinmed.uio.no

Abstract

Summary: Purpose: Long-term valproate (VPA) treatment has been associated with hyperandrogenism and polycystic ovaries in women with epilepsy. The exact mechanisms of action of the drug on sex steroid hormone function are still unsettled. The aim of the present study was to investigate the action of VPA on basal and gonadotropin-stimulated steroid secretion in porcine ovarian follicular cells and to measure the conversion of testosterone to estradiol. Second, the action of VPA on proliferation and apoptosis of follicular cells was investigated.

Methods: Small and medium follicles were obtained from pig ovaries on days 8–10 and 14–16 of the estrus cycle. Both follicular compartments, theca and granulosa cells, were cultured as a coculture resembling follicles in vivo. VPA in concentrations of 100 and 250 μg/ml was added to the control or gonadotropin-stimulated cultures.

Results: VPA caused a significant increase in basal and luteinizing hormone (LH)-stimulated testosterone secretion from small follicles, whereas in medium follicles, an increased basal but decreased LH-stimulated testosterone secretion was found. VPA caused decreased basal and follicle-stimulating hormone (FSH)-stimulated estradiol secretion by small follicles, whereas only the higher concentration decreased estradiol secretion in medium follicles. The conversion of testosterone to estradiol by small follicles was decreased under the influence of VPA in testosterone-alone and in testosterone-plus-FSH–stimulated cultures, whereas this was seen at only the higher VPA concentration in medium follicles. VPA had no effect on cell proliferation and viability, whereas in a dose-dependent manner, VPA increased caspase-3 activity.

Conclusions: VPA affected steroidogenesis in both unstimulated and gonadotropin-stimulated porcine ovarian follicular cells and inhibited the conversion of testosterone to estradiol. In addition, VPA may act as an apoptotic agent in both small and medium-sized follicles.

Valproate (VPA) has been used for >30 years in the treatment of epilepsy and is now one of the most frequently prescribed antiepileptic drugs (AEDs) worldwide. It has wide-spectrum antiepileptic activity and is effective in both partial and generalized epilepsies and in specific epileptic syndromes (1). In addition, it is increasingly used in other diseases such as bipolar psychiatric disorders (2) and migraine (3).

Although VPA has good efficacy in prevention of various types of epileptic seizures, its use may be compromised by severe side effects. Hyperandrogenism, obesity, and polycystic ovaries have been described in some studies in ≤60% of women taking VPA for epilepsy (4–6). An intense debate concerns whether this is related to the underlying epileptic disorder (7–10) or to the drug itself (6,10,11). Recent studies showing endocrine changes and polycystic ovaries also in patients treated with VPA for bipolar disorders (12) and in nonepileptic animals (13–15) indicate a drug-induced effect.

The precise mechanism of action of the drug on steroidogenesis is difficult to ascertain in human studies and in whole-animal in vivo models. We therefore previously studied the effect of VPA on steroidogenesis by using cultures of isolated porcine ovarian follicular cells (16,17). These studies demonstrated that even short-term VPA treatment disrupted follicular steroidogenesis, resulting in increased testosterone and decreased estradiol secretion. It was not possible to reverse the steroidogenic effects of VPA by removing the drug from the cell cultures. The elevated ratio of testosterone to estradiol suggested that VPA inhibits conversion of testosterone to estradiol.

The aim of the present study was to investigate mechanisms by which VPA affects ovarian steroidogenesis in female subjects. By use of in vitro cell cultures of porcine follicular cells, we investigated the effect of VPA on both basal and gonadotropin-stimulated steroidogenesis and on the conversion of testosterone to estradiol. Second, we studied the effects of VPA on cell proliferation and apoptosis.

MATERIALS AND METHODS

Reagents

Parker medium (M199), calf serum, trypsin, and phosphate-buffered saline (PBS) were obtained from Biomed (Lublin, Poland). Antibiotic antimycotic solution (1%), luteinizing hormone (LH), follicle-stimulating hormone (FSH), Ac-DEVD-AMC (7-amino-4 methyl coumarin), CHAPS {3-[(-cholamidopropyl) dimethylammonio]-1-propanesulfonate}, dithiothreitol, EDTA (ethylenediaminetetraacetic acid), glycerol, and HEPES were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Sodium valproate (VPA) was obtained from Desitin Pharma, Oslo, Norway.

Cell cultures

Porcine ovaries obtained from a local abattoir were collected into a bottle filled with sterilized ice-cold saline and transported to the laboratory. Approximately 1.5 h elapsed from slaughter to collection in the laboratory. Small (1–3 mm in diameter) and medium (4–6 mm in diameter) follicles were obtained from ovaries collected at days 8–10 and 14–16, respectively, of the estrus cycle, as described previously (18) and according to Liu et al. (19). In each experiment, six ovaries from three animals were selected for cell preparation. Because each ovary yielded four to six follicles, the total number of the follicles varied between 24 and 36. This procedure was chosen to minimize possible variations existing between follicles and animals. Granulosa cells (Gcs) and theca interna cells (Tcs) were subsequently prepared according to the technique described by Stoklosowa et al. (20). Gcs were scraped from the follicular wall with round-tipped ophthalmologic tweezers and rinsed several times with PBS. After collection, Gcs were washed 3 times in M199 and were recovered by centrifugation (10 min at 200 g). The average yield of Gcs was 5.0 × 105 cells/ml for small and 7 × 10 5 cells/ml for medium follicles. The Tcs from the same follicles were prepared as previously described in detail by Stoklosowa et al. (20). In brief, the theca layers were placed in a drop of saline under the dissection microscope. A theca interna was manually separated from the underlying theca externa. Isolated theca interna tissue was then washed, cleaned, cut with scissors, and exposed to trypsinization with 6–7 ml, 0.25% trypsin in PBS for 10 min at 37°C. The cells were separated by decantation, and the procedure was repeated 3 times. Finally, the cells were spun and resuspended in 24–36 ml of M 199 medium supplemented with 10% calf serum. The average yield of cells was 3.0 × 104 cell/ml for small and 2.5 × 105 cell/ml for medium follicles. In coculture experiments, 1 ml each of suspended Gc and Tc cells was placed in the same well, giving a total of 2 ml of culture medium per well. The cultures were maintained at 37°C in a humidified atmosphere of 5% CO2/95% O2.

Experimental procedure

Cells were isolated and initially cultured without reagents in serum containing (10% calf serum) M199 for 24 h at 37°C in a humidified atmosphere of 5% CO2/95% O2 to allow cell attachment to the plates. At 24 h, serum-containing M199 was discarded, and reagents were added for the next 48 h.

Experiment 1

To investigate the effect of VPA (100 or 250 μg/ml) on basal and gonadotropin-stimulated steroid hormone secretion, a coculture of Gcs and Tcs was prepared and cultured in basal medium with or without addition of either LH (100 ng/ml) or FSH (100 ng/ml). Forty-eight hours later, media were collected and frozen for analysis of testosterone levels in LH-stimulated cells and estradiol levels in FSH-stimulated cells.

Experiment 2

To investigate the action of VPA (100 or 250 μg/ml) on the P450 aromatase, Gcs and Tcs were cultured with testosterone as a substrate for estradiol secretion. Additionally, cells were cultured in basal medium with or without the addition of FSH, which is known to stimulate the aromatase activity in granulosa cells (21,22).

Experiment 3

This experiment was conducted to show possible action of VPA on cell proliferation. Gcs and Tcs were cocultured with 100 or 250 μg/ml VPA. Alamar Blue test was used to assess cell proliferation (23).

Experiment 4

This experiment was conducted to show possible action of VPA as an apoptotic agent. Gcs and Tcs were cocultured with 100 or 250 μg/ml VPA. Caspase-3 activity was used as a measure of the possible action of VPA as an apoptotic agent (24).

Every treatment was conducted in quadruplicate, and each experiment was repeated 3 times, giving the overall number of 12 wells per each treatment.

Steroid analysis

Testosterone (T) and 17β-estradiol (E2) were determined by radioimmunoassay by using Spectra kits (Orion, Diagnostica, Finland), supplied by Polatom (Świerk, Poland). For testosterone, the limit of the assay sensitivity was 5 pg/ml. The coefficients of variation within and between assays were 5.4% and 5.3%, respectively. The mean recoveries were 84.2–121.7%. The cross-reaction with 5α-dihydrotestosterone was 4.5%. All other tested steroids (methyltestosterone, androstenedione, progesterone, 17β-estradiol) showed less than 0.5% cross-reaction.

For 17β-estradiol, the detection limit of assay was 5 pg. The coefficients of variation between and within assays were 10.3 and 2.9%, respectively. The mean recoveries were 85.6–108.9%. The cross-reaction with ethinylestradiol was 1.4%. All other tested steroids (estrone, estriol, progesterone, testosterone, corticosterone) showed <1% cross-reaction.

Alamar Blue test

Alamar Blue assay is designed to measure the proliferation of different cell types and also cytotoxicity of agents within various chemical classes (25–27). The assay is based on detection of metabolic activity. Alamar Blue contains an oxidation–reduction indicator. Cellular proliferation induces chemical reduction of the media with addition of Alamar Blue, which results in a change in color from blue to red. The intensity of red color (reduced form indicator) reflects the extent of cellular proliferation.

The stock solution of Alamar Blue was aseptically added after 48 h to the culture wells in an amount equal to 10% of the incubation volume according to the technique described by Ahmed et al. (23). The cells were then incubated with Alamar Blue for another 6–8 h. The plates were then processed for absorbance analysis at wavelengths of 570 to 600 nm in an enzyme-linked immunosorbent (ELISA) plate reader.

Every measurement was repeated twice in every well, and each sample was repeated 3 times.

Caspase-3 activity

A fluorometric assay for caspase-3 was performed as described by Nicholson et al. (24). After replacing the culture media with Caspase Assay Buffer (50 mM HEPES, pH 7.4, 100 mM NaCl, 0.1% CHAPS, 1 mM EDTA, 10% glycerol, and 10 mM dithiothreitol), cell lysates were incubated with the caspase-3 substrate Ac-DEVD-AMC (7-amino-4 methyl coumarin; Sigma) at 37°C. The amounts of fluorescent products were monitored continuously for 90 min with a spectrofluorometer (Fluoroscan Ascent, Labsystems) at excitation wavelength of 355 nm and an emission wavelength of 460 nm. Because the initial substrate concentration was saturating (50 μM), the initial slope provided us a true and consistent measure of caspase-3 activity. Additionally, to confirm the correlation between signal detection and caspase-3 activity, we used Ac-DEVD-CHO inhibitor. Data were analyzed with Ascent software, normalized to the fluorescence in vehicle-treated cells and expressed as a fluorescence from four to eight separate samples ± SEM. Background fluorescence, determined for a no-enzyme control, was subtracted from each value.

Statistical analysis

All data points are expressed as means ± SEM from at least three different experiments (n = 3), each in quintuple. Differences between the concentrations of progesterone, estradiol, and testosterone in the control and experimental cultures were compared with analysis of variance and by using Duncan's new multiple range test.

RESULTS

Influence of VPA on basal and LH-stimulated testosterone secretion

VPA added to the culture medium caused a significant increase in basal testosterone secretion by both small (44.0 and 48.0 ng/ml, respectively, after 48-h exposure to 100 and 250 μg/ml VPA, vs. 31.2 ng/ml in the control culture) and medium-sized follicles (43.2 and 45.2 ng/ml, respectively, after 48-h exposure to 100 and 250 μg/ml, vs. 26.1 ng/ml in control culture; p < 0.05; Fig. 1).

Figure 1.

The influence of 100 μg/ml and 250 μg/ml of valproic acid on basal (open bars) and luteinizing hormone (LH)-stimulated testosterone secretion (black bars) by (a) small and (b) medium-sized follicles. Significant differences compared with untreated controls: *p < 0.05; **p < 0.01.

In LH-stimulated cells from small follicles, an additional increase in testosterone secretion was found (106.5 and 100.0 ng/ml, respectively, after 48-h exposure to 100 and 250 μg/ml VPA, vs. 70.2 ng/ml in LH-stimulated culture without drug; p < 0.05; Fig. 1a). Conversely, a decrease in LH-stimulated testosterone secretion was found in medium-sized follicles (123.0 and 126.0 ng/ml, respectively, after 48-h exposure to 100 and 250 μg/ml VPA, vs. 156.6 ng/ml in the LH-stimulated culture without drug; p < 0.05; Fig. 1b).

Influence of VPA on basal and FSH-stimulated estradiol secretion

In small follicles, VPA caused a significant decrease in estradiol secretion (20.1 and 19.1 ng/ml, respectively, after 48-h exposure to 100 and 250 μg/ml VPA, vs. 31.0 ng/ml in the control culture; p < 0.05; Fig. 2a). In medium-sized follicles, only the higher VPA dose significantly reduced estrogen secretion (47.1 ng/ml after 48-h exposure to 250 μg/ml VPA, vs. 57.3 ng/ml in control culture; p < 0.01; Fig. 2b). Similar results were found in FSH-stimulated cells (Figs. 2a and b).

Figure 2.

The influence of 100 μg and 250 μg of valproic acid on basal (open bars) and follicle-stimulating hormone–stimulated estradiol secretion (black bars) by (a) small and (b) medium-sized follicles. Significant differences compared with untreated controls: *p < 0.05; **p < 0.01.

Influence of VPA on conversion of testosterone to estradiol

In testosterone-supplemented control cultures, estradiol secretion increased to 794 ng/ml and 901 ng/ml in small and medium-sized follicles, respectively. VPA significantly reduced this increase in estradiol secretion by cultures originating from small follicles (534 and 481 ng/ml, respectively, after 48-h exposure to 100 and 250 μg/ml VPA, vs. 794 ng/ml in testosterone-supplemented cultures without drug; p < 0.05). In medium-sized follicles, a significant reduction in the conversion of testosterone to estradiol was found only at the higher doses of VPA (p < 0.01; Fig. 3).

Figure 3.

Valproic acid action on the conversion of testosterone to estradiol by small (SF) and medium-sized (MF) follicles. Testosterone in a dose of 10–7M was added to the culture medium. Significant differences compared with untreated controls: *p < 0.05; **p < 0.01.

Influence of VPA on conversion of testosterone to estradiol in FSH-stimulated cultures

FSH additionally stimulated estradiol secretion in testosterone-supplemented cells to 811 and 987 ng/ml, vs. 794 and 901 ng/ml in testosterone-supplemented cells collected from small and medium-sized follicles, respectively (Figs. 3 and 4). In small follicles, VPA caused a significant decrease in estradiol secretion in testosterone-plus-FSH–supplemented cultures (656 and 450 ng/ml, respectively, after exposure to 100 and 250 μg/ml VPA vs. 811 ng/ml in FSH-plus-testosterone without drug; p < 0.05 and p < 0.01; Fig. 4). In medium follicles, a significant decrease in conversion of testosterone to estradiol was noted only at the highest VPA dose (p < 0.05; Fig. 4).

Figure 4.

Valproic acid action on the conversion of testosterone to estradiol by small- (SF) and medium-sized (MF) follicles in follicle-stimulating hormone–stimulated cultures. Testosterone in a dose of 10−7M was added to the culture medium. Significant differences compared with untreated controls: *p < 0.05; **p < 0.01.

Influence of VPA on cell proliferation and apoptosis

VPA had no effect on follicular cell proliferation and viability evaluated by the Alamar Blue test (Fig. 5). However, VPA had a dose-dependent positive effect on the caspase-3 activity, suggesting an apoptotic effect of the drug (Fig. 5).

Figure 5.

The influence of valproic acid (VPA) on cell viability and cell apoptosis. Cells collected from small- (SF) and medium-sized (MF) follicles were cultured for 48 h in control medium (C) or with VPA (100 or 250 μg/ml). Cell viability was estimated by Alamar Blue test, and cell apoptosis, by measuring caspase-3 activity and expressed as percentage of control as 100%. Significant differences compared with untreated controls: *p < 0.05; **p < 0.01.

DISCUSSION

The results showed that VPA added to the culture medium caused a significant effect on steroidogenesis in both unstimulated and gonadotropin-stimulated porcine ovarian follicular cells. Reduced conversion of testosterone to estradiol indicates an effect of VPA on P450 aromatase activity, which is the main enzyme responsible for converting testosterone to estradiol in ovarian follicles (28).

In the present study, VPA increased not only basal, but also LH-stimulated testosterone secretion in small follicles. These results confirmed our previous findings of increased testosterone secretion from unstimulated in vitro cultured porcine follicular cells after VPA treatment (17). These data suggest that in small follicles, the steroidogenic effect of VPA is to a large extent mediated by an action on the P450 aromatase, even after short-term exposure. Normally androgens diffuse into the granulosa cells where the P450 aromatase converts testosterone to estradiol under the influence of FSH. In our study, this process was disrupted by VPA, as indicated by a decreased conversion of testosterone in basal and FSH-stimulated cells under the influence of the drug (Figs. 3 and 4).

The situation was somewhat different in cells originating from medium-sized follicles. As opposed to that in small follicles, basal estradiol concentration is higher in medium-sized follicles. Medium-sized follicles were harvested at the onset of luteolysis. At this stage of the estrus cycle, the ratio between FSH and LH receptors present in follicular cells is about to change toward an increase in LH receptors. This implies that medium-sized follicles have higher potential for steroid production, indicated by increased concentration of steroids in the follicular fluid. We found that in cells from both small- and medium-sized follicles, VPA increased basal testosterone secretion. However, in medium-sized follicles, VPA caused a decrease in LH-stimulated testosterone secretion, and at the highest dose, VPA had a negative effect on the conversion of testosterone to estradiol. These results suggest that VPA may as well influence other steps of the follicular steroidogenesis in defined stages of follicular development. One possibility is an effect on the activity of 17β-HSD (17β-hydroxysteroid dehydrogenase), which preferentially catalyzes the oxidation or reduction of specific steroid substrates and participates in the formation of androgens and estrogens (30). Elevated ovarian 17β-HSD activities have been associated with increased levels of circulating testosterone in human patients with polycystic ovarian syndrome (31,32).

Our findings are in general agreement with previous animal and human data demonstrating endocrine changes, including hyperandrogenism after long-term VPA treatment. In female rats fed with VPA for 3 months, a marked increase in testosterone-to-estrogen ratio and reduced estrogen levels were noted (15). Similarly, long-term VPA treatment in women with epilepsy has been associated with elevated testosterone and unchanged estrogen levels (4–6,33). These endocrine changes are associated with the development of polycystic ovaries in both humans (4–6) and animals (13,14). However, an intense, ongoing debate concerns whether the endocrine changes seen after VPA treatment is related to the drug itself or to ongoing epileptic activity, as epilepsy itself may induce endocrine dysfunction and polycystic ovaries (7,34–37).

These findings show a direct effect of VPA on steroidogenesis independent of epileptic activity, and indicate at least some similarities between the effects of VPA treatment and mechanisms at the ovarian level responsible for cyst formation. Hyperandrogenism is a hallmark in the development of the polycystic ovary syndrome (38). VPA increased both basal and LH-stimulated testosterone secretion and reduced conversion of testosterone to estrogen, thereby markedly increasing the testosterone-to-estrogen ratio in our in vitro cell cultures. This would support clinical evidence of an androgen-dominant microenvironment in the ovary promoting the development of polycystic ovaries. Further, Tamura et al. (39) demonstrated that the granulosa cells of the follicles in polycystic ovary syndrome in women were mainly negative for P450 aromatase, which would similarly lead to reduced conversion of testosterone to estrogen. Isojärvi et al. (4) showed that serum estradiol concentrations were not elevated in the VPA-treated women despite their elevated serum testosterone concentrations, suggesting that VPA inhibits the conversion of testosterone to estradiol. Women treated with VPA also have increased frequency of menstrual disturbances (4–6,12). Changes in testosterone and estrogen levels after VPA treatment may be of importance in this respect because the gonadal hormones exert important feedback effects at the level of the anterior pituitary gland and the hypothalamus. In the preovulatory phase, estrogen is critical to the initiation of the mid-cycle LH surge and hence to ovulation. The changes in peripheral sex steroid hormone levels with increased testosterone-to-estrogen ratio observed in our study may therefore interfere with the function of the hypothalamic–pituitary–gonadal axis and thereby affect cycle regularity, leading to menstrual disturbances. Several similarities seem to occur between findings in women with polycystic ovary syndrome and the effect of VPA. To evaluate the possible clinical implications of drug effects observed in animal studies, drug concentrations must be applicable to those used in humans, and the animal model itself must be clinically relevant. Regarding drug concentrations, therapeutic serum levels of VPA in humans range from 50 to 100 μg/ml, which is equivalent to the lowest concentration used in our study. However, in humans, therapeutic serum concentrations are measured as trough morning levels, thus representing minimal values. It is well known that diurnal fluctuations of several hundred percent occur (40,41). From toxicologic studies, it has been shown that serum concentrations ≤450 μg/ml produce limited toxicity (42). We therefore consider the concentrations used in our study to be clinically relevant when discussing possible effects of the drug related to humans.

We find cocultures of porcine follicular cells to be a valuable model for evaluation of possible drug effects on human ovarian steroidogenesis. To our knowledge, no evidence in the literature indicates that ovarian follicular steroidogenesis in humans differs markedly from that found in other mammals. Apparently the sow is a particularly good model animal because the follicular dynamics during the luteal phase of the oestrous cycle show clear similarities to those of humans (43). The estrus cycle in the pig is also quite long, of ∼20 days compared with 28 days in humans. Cocultures of theca and granulosa cells would better reflect an in vivo situation compared with models using isolated cell cultures. This is because in vivo, theca and granulosa cells work closely together as a functional unit. Our experimental model also provides the possibility of studying direct effects of AEDs on ovarian steroidogenesis in cells excised from healthy animals showing natural estrus cycles. The relevance of our model is further supported by the similarities with clinical observations of hyperandrogenism also in women treated with VPA (4–6,12,33).

VPA increased the caspase-3 activity in a dose-dependent manner (Fig. 5), suggesting an apoptotic effect of the compound (44,45). Apoptosis is an important cellular process by which superfluous or unwanted cells are deleted from an organism during tissue remodelling and differentiation. Apoptosis can be triggered by diverse stimuli ranging from intracellular stress to extracellular receptor signalling. A central component of this apoptotic machinery is the family of caspase enzymes (46,47). Activation of caspase-3 is a central event in the apoptotic process (44,45). The possibility of VPA being an apoptotic agent has wide clinical implications. Caspase-3 activity is, however, only one way of investigating this issue. It is therefore important to have this finding verified in other models of apoptosis (48).

In the ovary, estrogens are of prime importance for the survival and development of preovulatory follicles, whereas androgens are apoptotic factors (49). Evidence suggests a direct interference of estrogens with apoptotic processes. Recently estradiol was found to prevent caspase-6–mediated neuronal cell death, probably by inducing a caspase inhibitory factor (CIF) through a receptor-mediated nongenomic pathway (50). Another example is the stimulation of the expression of antiapoptotic proteins, like Bcl-2 or Bcl-XL, which prevents cytochrome c efflux from mitochondria to cytoplasm and, finally, results in inhibition of caspase-dependent apoptotic cell death (51). It also is possible that estrogens stabilize mitochondrial function either by scavenging free radicals or by affecting adenosine triphosphatase (ATPase) FOF1 (52).

The most documented effect of estrogens is related to their endogenous antioxidant capacity, which depends on the presence of hydroxyl group in the C3 position on the A ring of the steroid molecule and allows estrogens to suppress the oxidative stress in neurons, thus preventing neuronal cell death (53). Our present and previous (16,17) experiments have shown a disruption of follicular steroidogenesis induced by VPA, resulting in decreased estradiol levels. The possibility that an apoptotic action of VPA might be linked to the steroidogenic effects of the drug should receive further attention.

In conclusion, our study showed that VPA acted on both small- and medium-sized follicles and had a direct effect on steroidogenesis. The results suggest that the action of VPA may, at least in part, be due to an effect on aromatase activity but also is modified by pituitary support of gonadotropins. Moreover, our data indicate that, in both types of follicles, VPA could act as an apoptotic agent, not only by increasing testosterone as an apoptotic hormone, but also by activation of caspase-3 activity. We suggest that the proapoptotic action of VPA might be due to the removal of estradiol from the system.

Acknowledgments

Acknowledgment:  We thank M. Mika, Ph.D., from the Department of Animal Physiology, Academy of Agriculture, Kraków, Poland, for radioimmunologic determinations of steroid hormones. The work was supported by grants from Department of Neurology, Rikshospitalet, University of Oslo, Oslo, Norway, and by DS./IZ/2002.

Ancillary