Skeletal Changes in Transgenic Male Mice Expressing Human Cytochrome P450 Aromatase^†

INTRODUCTION

ESTROGEN IS AN important hormone in maintaining normal bone metabolism. Estrogen deficiency is known to induce osteoporosis in women^1-3 and lead to abnormal bone development and osteoporosis in men with mutations in genes encoding either the estrogen receptor⁴ or aromatase enzyme.^{5, 6} Accordingly, treating osteoporotic patients with estrogen can improve their bone quality by increasing bone mass and decreasing the fracture rate both in females and males.¹,^6-9

In addition to the gonads, the biosynthesis of estrogen in other tissues could be an important source of local and circulating estrogen. Aromatization of androgens into estrogens is catalyzed by the enzyme P450 aromatase (P450arom),^10-12 and this process has been shown in several tissues including adipose tissue,¹³ skin,¹⁴ muscle,¹⁵ and bone.^{16, 17} Although the presence of aromatase in bone suggests that local estrogen production may play an important role in the maintenance of bone tissue, postmenopausal bone loss indicates that it is clearly not enough to prevent bone loss in older women.

Experimental treatment of rats with aromatase inhibitors induces a decrease in BMD and impaired skeletal modeling in growing and aged male animals.^{18, 19} Comparable results were obtained by others when they studied bone changes in mice with targeted disruption of the aromatase gene (ArKO mice).^20-22 The results showed that ArKO mice had osteopenia in the lumber spine, significant decreases in trabecular bone volume and thickness, and increased bone turnover compared with their wildtype (WT) controls. The bone changes in ArKO mice can be reversed by estradiol (E₂) treatment and markedly higher concentrations of testosterone (T), but similar levels of estrogen in the sera of ArKO mice versus WT animals suggest that decreased local estrogen biosynthesis induces bone loss in these animals. It also suggests that, probably, upregulation of local estrogen levels by aromatase could be an effective way to maintain or improve the bone tissues.

We have recently produced transgenic P450 aromatase-positive (AROM⁺) mice expressing human P450 aromatase under the human ubiquitin C promoter.²³ The mice show low universal expression of P450arom, and the male mice have increased concentrations of E₂ and reduced concentrations of T in their blood circulation. The imbalanced androgen to estrogen ratio induces severe changes in structure and function of the male reproductive system.^{23, 24} In this study, we investigated bone changes in AROM⁺ mice to obtain more understanding of the importance of a prolonged imbalance of the E₂ to T ratio regarding bone structure.

MATERIALS AND METHODS

pQCT measurements

Right tibias of the young mice were thawed at room temperature for 1 h. After removing the soft tissues, the bone was fixed in plastic tube (8 mm diameter) with a spring and scanned with pQCT equipment (XCT 540; Stratec). For the measurement levels in tibia, the reference line was placed at the proximal end of the bone. Three cross-sections, at 0.3-mm intervals, were analyzed 1.8 mm from the reference line. Measurements were also taken from two sections separated by 1 mm, starting 2.5 mm above a reference line at the tibiofibular junction. Special Software version 5.40 (Stratec) was used to analyze the images of each section, with a voxel size of 0.10 mm. Standardized analysis (peel mode 2, cort mode 1, contour mode 1, threshold 0.25 mg/cm³ for trabecular bone and 0.71 mg/cm³ for cortical bone) was applied. After measuring the BMD and length, the tibias were burned at 500°C for 18 h to obtain individual bone ash weights. The right tibias and femora from the aged mice were also analyzed. Because the thickness of the epiphysis of the proximal tibia was thinner than that of young animals, three sections were measured, starting 1.5 mm from the proximal end of the tibia. As regards the femur, three sections were measured 1.8 mm above the distal end, and another two sections were measured 6 mm from the distal end. All section levels are presented in Fig. 1. Before the right tibias were burned, the bending strength was tested, following our earlier mechanical testing method.²⁶ The span of the supporter was 10 mm.

RESULTS

Table 1 shows body weight, bone length, and ash weight of tibias of the animals at the age of 40 days. Both body weight gain and long bone growth were retarded in AROM⁺ mice. AROM⁻ animals were lighter than WT animals but clearly heavier than AROM⁺ animals. AROM⁺ mice had significantly shorter femora and tibias in both sexes compared with WT animals, but there were no differences in total ash weight. In contrast, the ash weight of tibial epiphysis was higher in male AROM⁺ mice than in other mouse groups. Bone ash weight showed no difference between WT and AROM⁻ animals. All results including pQCT measurement and histomorphometric analysis showed that the differences were more marked in the males.

Table Table 1.. Body Weight, Bone Length, and Bone Ash Weight of Mice at the Age of 40 Days

BMDs in the young mice are presented in Table 2. As expected based on ash weight data, AROM⁺ male mice showed increased density of both trabecular and cortical bone. The tissue area of cortical bone was markedly reduced, leading to a decreased polar moment of inertia at the tibial shaft. In female mice, there was no difference in BMD between control and AROM⁺ mice. However, in females also, there was a smaller tissue cross-sectional area in tibias of AROM⁺ mice versus the controls. In both AROM⁻ males and females, most of pQCT measurements fell between the values in WT and AROM⁺ animals.

Table Table 2.. BMD of the Tibia of Young Mice at the Age of 40 Days

Table Table 3.. Bone Histomorphometric Analysis in Young Mice at the Age of 40 Days

Results from bone histomorphometric analyses of young animals are presented in Table 3. Similarly to the BMD analysis, the data showed that there were significant changes in histomorphometric parameters in males but not in females. Owing to increased trabecular thickness, trabecular bone volume in the proximal tibias of AROM⁺ males was much higher than in the WT mice (Fig. 2A). The number of osteoclasts on the trabecular bone surface was decreased in AROM⁺ mice. Bone formation rate per trabecular bonevolume (BFR/BV) was significantly decreased, but it was significantly increased when calculated as the bone formation rate per tissue volume (BFR/TV). The longitudinal growth rate of the proximal tibial growth plate was decreased in AROM⁺ males, while there was no difference in the thickness of the growth plate between groups. The reduced bone formation rate in cortical bone was also evident after measuring the tetracycline and calcein labels in the cross-sections of the femoral shafts (Fig. 2B). BFR at the endocortical surface was equal in all animal groups. All other histomorphometric values, except number of osteoclasts, were similar between WT and AROM⁻ animals.

We have previously reported that 4-month-old AROM⁺ males present increased E₂ and reduced T concentrations in serum.²³ In this study, we also measured the intratesticular concentrations of E₂ and T in 4-month-old animals. As indicated in Table 4, concentrations of T in the serum, as well as in intratesticular fluid, gradually decreased with age in WT mice. However, concentrations of T in AROM⁺ mice were much lower in young mice, but much higher in aged ones versus WT animals. On the other hand, whereas concentrations of E₂ slightly increase in the serum and decrease in intratesticular fluid in male WT mice with age, E₂ concentrations in male AROM⁺ animals were markedly increased in both the serum and intratesticular fluid. The ratio of E₂ in the sera of AROM⁺ versus WT mice was 1.6 at the age of 40 days and 11.3 at the age of 9 months. In intratesticular fluid, it was 3.2 in young mice and 27.8 in aged ones.

Table Table 4.. Concentration of Testosterone and Estradiol in Serum and in Intratesticular Fluid

Figure 3 shows the bone changes in aged, 9-month-old AROM⁺ male mice. In contrast to the situation found at 40 days of age, at the age of 9 months, the body weight of the AROM⁺ mice was increased by 21.2% versus that of WT animals. The difference in femoral length between AROM⁺ and WT mice did not change over more than 7 months of growth, but the difference in tibial length changed dramatically. In AROM⁺ males versus WT males, the tibia was 23.2% shorter at the age of 9 months, whereas it was 5.26% shorter at 40 days of age. In addition, the tibias were significantly heavier and had a higher bending strength in AROM⁺ males compared with the WT animals.

BMD measurements from aged males (9 months of age) are presented in Table 5. The densities of trabecular and cortical bone, cortical bone area, and polar moment of inertia of tibias were significantly higher in AROM⁺ mice than in WT animals. In contrast to the situation found in young AROM⁺ mice, no differences in tissue areas of proximal tibias or tibial shafts were found in the aged animals. In the femur, all values were markedly higher in AROM⁺ than in WT animals except for trabecular BMD of the distal femur.

Table Table 5.. BMD of the Tibia and Femur of Aged Male Mice at the Age of 9 Months

Histomorphometric data in aged mice were consistent with the BMD measurements (Table 6). Furthermore, the data showed that, in contrast to the situation found in young mice, the BFRs at the trabecular bone surfaces and endocortical bone surfaces were significantly higher in the AROM⁺ mice compared with the WT mice (Figs. 2C and 2D).

Table Table 6.. Bone Histomorphometric Analysis of Aged Male Mice at the Age of 9 Months

DISCUSSION

Aromatization of androgens is one of the two last steps in the biosynthetic pathway of E₂ formation from cholesterol. Similarly, testicular T represents a precursor of E₂ production in AROM⁺ males, with enhanced expression of human P450 aromatase. This results in increased circulating and intratesticular E₂ concentrations in AROM⁺ male mice (this study).²³ The expression of P450 aromatase in AROM⁺ mice is, however, low, and no obvious reproductive phenotype has been found in AROM⁺ female mice, which show normal ovarian and uterine histology and are fertile. However, this study shows a clear impact of P450 aromatase expression in some of the parameters analyzed regarding the female bones at 40 days of age, such as bone growth and tissue area. This suggests that low local E₂ formation as a result of the action of P450 aromatase has an impact in female bone physiology. This is in line with the results of previous studies suggesting that, in postmenopausal women and in men, P450 aromatase activity can generate physiologically significant levels of E₂ locally without affecting circulating levels of E₂.^{28, 29}

As expected, the effects on AROM⁺ female bone physiology and structure were markedly milder than those observed in the males. AROM⁺ male mice have previously been shown to present severe structural and functional alterations in their reproductive tissues, such as increased circulating and intratesticular E₂ to T concentration ratios combined with Leydig cell hyperplasia, gynecomastia, and undeveloped accessory sex glands.^{23, 24} Several of these phenotypes are also common features that are caused by perinatal estrogen exposure in males.^30-32 A typical feature of fetal estrogen exposure is hypoandrogenization.³³ Thus, the low level of T in young adult AROM⁺ males is not only a result of the enhanced conversion of T to E₂, but is also a result of the fact that estrogen action in the males suppresses T production in the testis. Our recent data indicate that a putative reason for the enhanced production of T in 9-month-old AROM⁺ mice is the activation of testicular macrophages that could produce 25-hydroxycholesterol, this potentially being a direct substrate for T production in the Leydig cells (X Li and M Poutanen, unpublished observations, 2003).

AROM⁻ and AROM⁺ mice were littermates, and their genotypes were verified by PCR. Some bone parameters of AROM⁻ animals were significantly different from those of WT animals. This is most likely because of the fact that, during the pregnancy and suckling period of 21 days, AROM⁻ mice got more estrogen from their AROM⁺ mothers than the WT mice.

In this study, we carried out detailed analysis of bone structure and properties in AROM⁺ male mice and observed marked differences between AROM⁺ and WT males. Both the young and aged AROM⁺ males have more trabecular bone in the metaphyseal region and a higher cortical BMD, suggesting that an increased estrogen to androgen ratio increases the bone mass in transgenic males. Estrogens are considered as “antiresorptive” agents, and they inhibit bone resorption by decreasing the activation frequency of the bone remodeling cycle (e.g., in postmenopausal women).³⁴ However, studies with ovariectomized mice showed that low-dose estrogen treatment maintained bone mass after reduction of bone turnover,^{35, 36} whereas treatment with high doses of estrogens increases bone mass by increasing bone formation rate and can even lead to osteosclerosis.^36-38 The AROM⁺ mouse model shows both of the abovementioned actions of estrogens at different periods of life. In young animals, bone mass was maintained by the enhanced antiresorptive effect of estrogens, which exceeded the estrogen-mediated suppression of bone formation. In aged mice, the effects on bone formation became profound. The reason for these differences, however, is unclear. In young adult AROM⁺ mice, the level of T is low and that of E₂ is increased (this study).²³ However, as shown in this study, circulating concentrations of T are increased at the age of 9 months, and accordingly, the level of circulating E₂ is highest at this age.

Estrogen influences bone growth at various phases of skeleton maturation. In previous studies, it has been shown that both low and high doses of estrogens possess an inhibitory effect on bone growth at the growth plate in experimental animals.^{36, 37} Interestingly, in this study, we found that, in young mice (40 days of age), the differences in bone length between AROM⁺ and WT males were similar in the tibia and femur, but the tibias of aged AROM⁺ males were relatively much shorter than the femora. The tibias of the AROM⁺ males were 23.2% shorter than in WT males, whereas the femur length was only 2.7% shorter. This suggests that the tibia and femur respond differently to the hormonal changes between the ages of 40 days and 9 months. The difference could also be related to higher P450arom activity or higher estrogen receptor expression in the tibia, which remains to be studied.

Chondrocytes play an important role in longitudinal bone growth, and it has been shown previously that estrogen treatment reduces the synthetic activity of the chondrocytes.^{1, 39, 40} However, in this study, we found no differences in the thickness of the whole growth plate or in the thickness of calcified cartilage in the growth plate between the different study groups. Interestingly, the fluorescence double-labeled images clearly showed inhibition of the longitudinal growth rate in the primary spongiosa of the proximal tibia and distal femur of AROM⁺ mice compared with WT mice. Thus, the results suggest that the growth and maturation of growth plate cartilage is not affected, but the altered bone growth is a result of an effect in the primary spongiosa below and above the cartilage of the proximal tibia and distal femur, respectively.

In summary, transgenic mice bearing the human ubiquitin C promoter/human P450 aromatase fusion gene present increased estrogen biosynthesis in peripheral tissues and increased circulating estradiol levels in male mice. The changes in bone structure in these mice are consistent with the hypothesis that low-dose estrogen treatment inhibits bone resorption and maintains bone mass, whereas high doses of estrogens increase bone mass, mainly through increased bone formation.