|Home | About | Journals | Submit | Contact Us | Français|
We previously found that the magnitude of skeletal deficits caused by GH deficiency varied during different growth periods. To test the hypothesis that the sensitivity to GH is growth period dependent, we treated GH-deficient lit/lit mice with GH (4 mg/kg body weight·d) or vehicle during the prepubertal and pubertal (d 7–34), pubertal (d 23–34), postpubertal (d 42–55), and adult (d 204–217) periods and evaluated GH effects on the musculoskeletal system by dual energy x-ray absorptiometry (DEXA) and peripheral quantitative computed tomography. GH treatment during different periods significantly increased total body bone mineral content, bone mineral density (BMD), bone area, and lean body mass and decreased percentage of fat compared with vehicle; however, the magnitude of change varied markedly depending on the treatment period. For example, the increase in total body BMD was significantly (P < 0.01) greater when GH was administered between d 42–55 (15%) compared with pubertal (8%) or adult (7.7%) periods, whereas the net loss in percentage of body fat was greatest (−56%) when GH was administered between d 204 and 216 and least (−27%) when GH was administered between d 7 and 35. To determine whether GH-induced anabolic effects on the musculoskeletal system are maintained after GH withdrawal, we performed DEXA measurements 3–7 wk after stopping GH treatment. The increases in total body bone mineral content, BMD, and lean body mass, but not the decrease in body fat, were sustained after GH withdrawal. Our findings demonstrate that the sensitivity to GH in target tissues is growth period and tissue type dependent and that continuous GH treatment is necessary to maintain body fat loss but not BMD gain during a 3–7 wk follow-up.
Osteoporosis is a disease characterized by low bone mineral density (BMD) leading to increased bone fragility, with fractures occurring even after minimal trauma (1). As BMD is one of the main determinants of fracture risk in osteoporotic patients, the peak BMD, which is acquired during the pubertal period and adolescence and is achieved in late adolescence, is a main determinant of osteoporosis in adulthood (2). About 60–70% of variance in peak BMD is determined genetically (3), and 40–50% of peak BMD is accumulated during puberty (4). As relatively small increases in peak BMD of 5–10% could lead to a decreased risk for fracture later in life (5), it is important to elucidate the molecular mechanisms that contribute to acquisition of peak BMD during puberty.
The dramatic accumulation of bone mass during the prepubertal and pubertal periods is caused by changes in both modeling and remodeling that occur simultaneously. With regard to the potential signaling molecules that could contribute to the skeletal changes that occur during postnatal growth, sex steroids and GH/IGF-I axis have been proposed to play a major role (6–10). The role of GH/IGF axis in the acquisition of peak BMD is evident from a number of clinical and animal studies, including: 1) Adults with childhood onset of GH deficiency exhibit decreased bone mass and increased incidence of fractures compared with age-matched controls (11–15); 2) GH/IGF production increases during puberty and correlates with skeletal changes in girls (16–19); 3) Overexpression of GH in erythroid cells using β-globin promoter increases bone size and BMD in transgenic mice (20); and 4) GH-deficient lit/lit mice and GH receptor knockout mice exhibit significant deficits in peak BMD (21–23).
Our recent studies evaluating the relative importance of GH in mediating skeletal changes that occur during postnatal growth demonstrated that the deficit in femoral BMD at the end of the postpubertal growth period was 4-fold greater compared with that of the prepubertal growth period (23). Similarly, the deficit in bone size caused by GH deficiency was greater at the end of the postpubertal growth period compared with the prepubertal growth period, suggesting that the effect of GH on bone may be growth period dependent. Consistent with this concept, Choi and Waxman (24) have recently shown that GH administration to hypophysectomized adult rats, but not prepubertal rats, led to stimulation of liver genes, thus suggesting that liver factors necessary for mediating GH effects are absent in prepubertal rats. Based on these findings, we proposed the hypothesis that the effect of GH to increase BMD is, in part, growth period dependent. To test this hypothesis, we administered GH during different growth periods (i.e. prepubertal, pubertal, postpubertal, or adult) and evaluated its effects on bone, muscle, and body fat. For our studies, we used lit/lit mice as a model because we predicted that these mice with no detectable endogenous GH levels should respond to exogenous GH much more robustly compared with wild-type mice with normal endogenous GH levels.
The recombinant human GH was a kind gift from Biosidus Co. (Buones Aires, Argentina).
Breeding pairs of GH-deficient lit/lit mice were kindly provided by Dr. L. R. Donahue (The Jackson Laboratory, Bar Harbor, ME). Due to a spontaneous mutation in the GHRH receptor molecule, the lit/lit mice (C57BL/6J) are deficient in GH and are 50% smaller than wild-type mice (25, 26). GHRH-deficient lit/lit mice were also identified by 60% lower serum IGF-I levels compared with control mice. The homozygous of GHRH-deficient lit/lit mice were used for breeding and lit/lit mice from these pairs were used for the experiments. The animals were housed in a controlled environment with 12-h light, 12-h dark cycles at 21 C with food and water ad libitum with standard rat/mouse diet containing normal calcium.
GH was injected sc three times a day (0800, 1200, and 1600 h). This method of treatment was chosen based on our previous finding that three times daily administration of IGF-I showed greater effects on BMD than a single administration or continuous administration through osmotic pump in the IGF-I deficient midi mice during puberty (27). The dose of GH administered was 4 mg/kg body weight·d equally divided among the three injections. This dose was chosen because it has been shown to be effective in producing anabolic effects in bone and other organs (28–31). We also chose the pubertal period as d 21–35 based on the previous findings that serum estradiol increases on d 26 and vaginal opening occurs by d 31 in normal mice (32) or by d 35 in GH receptor knockout mice (33) and based on our previous findings that BMD increased by 40% during this period in lit/lit mice. Fifty-seven GH-deficient lit/lit mice were divided into eight groups: groups 1 and 2 were treated with PBS (group 1; n = 8) or GH (group 2; n = 8), respectively, for 28 d from d 7–34 as a prepubertal through pubertal period; groups 3 and 4 were treated with PBS (group 3; n = 8) or GH (group 4; n = 7), respectively, for 14 d from d 21–34 as a pubertal period; groups 5 and 6 were treated with PBS (group 5; n = 8) or GH (group 6; n = 8) respectively for 14 d from d 42–55 as a postpubertal period; and groups 7 and 8 were treated with PBS (group 7; n = 5) or GH (group 8; n = 5), respectively, for 14 d from d 204–217 as an adulthood period. At the end of treatment (d 35 for groups 1–4; d 56 for groups 5 and 6; or d 218 for groups 7 and 8), a dual energy x-ray absorptiometry (DEXA) measurement was performed under anesthesia using Ketamine (Phoenix Pharmaceutical Inc., St. Joseph, MO) and Xylazine (Lloyd Laboratories, Shenandoah, IA) by ip injection. The DEXA measurement was also performed in groups 1–4 at 3 wk after withdrawal of treatment (d 56). The body weights of the mice were measured every week with an electronic balance (Scout SC2020, OHAUS, Florham Park, NJ). On d 84 for groups 1–6 or d 239 for groups 7 and 8, the mice were euthanized by CO2 inhalation followed by cervical dislocation. After DEXA measurements of total-body BMD were performed, bilateral femurs were also collected for bone densitometry and volumetric bone densitometry measurement by peripheral quantitative computed tomography (pQCT). The femur length was measured using a caliper (Dial Caliper; Mitutoyo Corp., Kawasaki, Japan) before bone density measurement. The experimental procedures performed in this study are in compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Animal Studies Subcommittee at the Jerry L. Pettis Veterans Administration Medical Center (Loma Linda, CA).
Total-body BMD, bone mineral content (BMC), bone area, percentage of body fat (% body-fat), and lean body mass were determined by DEXA (PIXImus instrument, LUNAR Corp., Madison, WI). The precision for BMD and BMC measurements was ± 1% coefficient of variation in vitro and ± 2% coefficient of variation in vivo (27).
Volumetric bone density and geometric parameters were determined by (Stratec XCT 960M, Norland Medical Systems, Ft. Atkinson, WI). Routine calibration was performed daily with a defined standard (cone phantom) containing hydroxyapatite embedded in Lucite (Norland Medical Systems, Inc.). Analysis of the scans was performed using the manufacturer-supplied software program (STRATEC MEDIZINTECHNIC GmbH Bone Density Software, version 5.40 C; Norland, Madison, WI). Volumetric bone density and geometric parameters were estimated with Loop analysis. The thresholds were set at 230– 630 mg/cm3 for total BMD measurement. The voxel size was set a 0.07 mm and a half-millimeter-thick slice was scanned through the entire length of bone. The reference line as a center of scanning was set at the midpoint of femur, thereafter the nine slices were scanned symmetrically from the reference line. The total volumetric BMD (vBMD), periostal circumference, and endosteal circumference of each three slices from distal, middle, or proximal were taken as an average, and were expressed as distal metaphysis, mid-diaphysis, or proximal metaphysis. The coefficient of variation for total vBMD, periosteal circumference, and endosteal circumference for repeat measurements of four femurs (two to five measurements) were 3, 1, and 2%, respectively (4, 34, 35).
Serum IGF-I levels were measured by a RIA as previously described (36). IGFs were separated from IGF binding proteins by an acid gel filtration protocol using BioSpin column (Bio-Rad Laboratories, Hercules, CA) as described previously (36).
Results are expressed as mean ± sem. Statistical analysis of the data was performed by Student’s unpaired t test for the comparison of GH treatment to vehicle treatment at each time point or Fisher’s protected least significant difference method (post hoc test) for multiple comparisons in a one-way ANOVA as appropriate for the comparison of GH or vehicle treatment at each time point and for the comparison of net changes of DEXA and pQCT parameters during GH treatment. P values less than 0.05 were considered significant.
The initial body weights among treatment groups were similar at the beginning of experiment. GH treatment at four different periods (groups 2, 4, 6, and 8) increased the body weight at the end of treatment and the femoral bone length at the end of the study compared with the corresponding control group (Figs. 1 and and2).2). However, the magnitude of GH effect on body weight was growth period dependent. For example, GH treatment caused a less than 20% increase in body weight when it was administered between wk 1 and 3 compared with a 70% increase in body weight when it was given during pubertal (d 21–34) or postpubertal (d 41–55) growth periods.
GH treatment caused significant increases in the total-body BMD, BMC, bone area, and lean body mass irrespective of the treatment period. However, the magnitude of increase in these parameters varied depending on the treatment period. For example, 2 wk GH treatment between d 42 and 55 significantly increased total-body BMD by 15% compared with 7–9% increases in total-body BMD seen when GH was administered between d 21–34 and 204–217 (Table 1). Similarly, the increase in total-body BMC was significantly greater when GH was administered between d 42–55 compared with d 204–217. In contrast to skeletal changes, percent increase in lean body mass did not vary considerably when GH was administered during different growth periods. Consistent with the increase in lean body mass, GH administration caused a significant reduction in % body-fat. The reduction in % body-fat (56%) was, however, significantly greater when GH was administered in adult compared with prepubertal (26%) or pubertal (35%) growth periods (Table 1).
To determine whether the GH-induced skeletal and muscle changes are maintained after withdrawal of GH treatment, we measured total-body BMD, BMC, lean body mass, and % body-fat at 3–7 wk after withdrawal of GH treatment. While the increases in total-body BMD, BMC, and lean body mass were maintained after withdrawal of GH treatment, the changes in percentage of body fat were not sustained after withdrawal of GH treatment (Table 1).
Because BMD measurement by DEXA is influenced by bone size, we performed pQCT measurements in the femur to evaluate changes in vBMD and bone size in response to GH treatment.
Consistent with the total-body BMD changes measured by DEXA, GH treatment between d 42 and 55 significantly increased the total vBMD (10–15%) at three different regions of the femur (proximal metaphysis, mid-diaphysis, and distal metaphysis) compared with the corresponding control group. The increase in total vBMD cannot be explained solely on the basis of an increase in size since cortical vBMD at mid diaphysis was also significantly increased in this GH-treated group compared with control group (888 ± 24 vs. 785 ± 31 mg/cm3; P < 0.001). However, GH treatment during the other growth periods did not cause significant increases in the total vBMD in the femur except at the mid-diaphysis of GH-treated adult mice.
GH treatment caused a significant increase in the periosteal and endosteal circumferences irrespective of the treatment period compared with corresponding control mice. The increases were significant in all three regions of the femur (proximal metaphysis, mid-diaphysis, and distal metaphysis) with the exception that the periosteal circumference at mid-diaphysis in group 2 and the endosteal circumference at proximal metaphysis in group 4 was not significant.
To determine whether GH-induced changes in various skeletal parameters, lean body mass, and % body-fat are dependent on the period when GH was administered, we evaluated net changes during treatment in various parameters in different groups. Figure 3 shows that the net gain in total-body BMD was significantly higher when GH was administered between d 42 and 55 compared with the other treatment periods. Although the net gain in lean body mass did not vary between pubertal, postpubertal, and adult periods, the net loss of % body-fat was the greatest when GH was administered during adult period (Fig. 3). Figure 4 shows the net changes in several parameters measured by pQCT at the mid-diaphysis of femur. GH treatment between d 42 and 55 caused the greatest net gain in total vBMD compared with the other periods (Fig. 4A). GH treatment between d 42–55 and 204–217 exerted greater net gain in periosteal circumference compared with the other periods (Fig. 4B). The net gain in endosteal circumference was not significantly different among different GH-treated groups (Fig. 4C).
To evaluate whether the greater increase in bone accretion caused by GH treatment during the postpubertal growth period is caused by greater increase in IGF-I production, we measured serum IGF-I levels in groups 3–6. Figure 5 shows that GH treatment caused a significantly greater increase in serum IGF-I levels when administered during the postpubertal growth period compared with the pubertal growth period in lit/lit mice. Although serum IGF-I level was nearly doubled after GH treatment in lit/lit mice, the values are still lower than the age-matched control mice of the same genetic background with normal GH levels (160 ± 29 ng/ml in lit/lit mice vs. 258 ± 73 ng/ml in C57BL mice).
Consistent with the human clinical data using GH-deficient subjects (7, 8, 11–15, 37), we have found that GH-deficient lit/lit mice exhibit a significant decrease in BMD and that treatment of these mice with GH caused a significant increase in BMD. In terms of the magnitude of BMD change caused by GH treatment, it is remarkable that 2 wk of GH treatment increased total body BMD by as much as 15% when it was administered between d 42–55 in lit/lit mice. Our study demonstrates for the first time that the magnitude of GH effect in GH-deficient mice is dependent on the age when GH is administered to these mice. We found that 2 wk of GH treatment during the postpubertal growth period (d 42–55) caused nearly 2-fold greater increase in BMD (15% vs. 8%) compared with 2 wk of GH treatment during 2 wk of pubertal growth period. Thus, the net gain in total body BMD was significantly greater when GH was administered during the post pubertal growth period compared with any other growth periods tested in this study.
It is now well established that the areal BMD measurements by DEXA can be influenced by size differences. In this regard, our findings that both the total vBMD and cortical vBMD were significantly increased in mice treated with GH during the postpubertal period suggest that the increase in BMD in GH-treated lit/lit mice cannot be solely explained on the basis of increased bone size. This conclusion needs to be further verified by microCT and histomorphometric analyses because lit/lit mice have smaller bones and pQCT measurements may be subject to influence by partial volume effect.
In contrast to the findings in this study, Rosen et al. (28) found that 4 wk of GH treatment did not significantly increase femoral BMD or serum IGF-I level in 10- to 12-wk-old rats. There are several potential explanations for the observed differences in GH response between the two studies that include: 1) differences in the model (GH-deficient mice vs. normal growing rats) used; 2) differences in the dose and number of injections per day; and 3) differences in serum IGF-I response between the two studies.
Our findings also reveal that GH administration during the prepubertal growth period has little effect on general body growth or on bone accretion. This conclusion is based on the findings that the magnitude of increase in body weight was much smaller when GH was administered between d 7–21 compared with between d 21–34 (< 20% vs. 70%) and that overall changes in many of the parameters studied were similar in mice treated with GH between d 7 and 34 vs. d 21–34. Accordingly, we have recently reported that the skeletal deficit is minimal in GH-deficient lit/lit mice compared with control mice at the end of the prepubertal growth period (23). If these findings can be extrapolated to humans, this would suggest that GH administration during the prepubertal growth period might not be efficacious compared with pubertal or postpubertal growth periods in GH-deficient children.
In addition to its effects on the skeleton, the effects of GH on lean body mass and body fat have been well established. Numerous clinical studies have shown that lean body mass increases in GH-treated, GH-deficient adults compared with corresponding controls (11, 38–40). We obtained similar results in lit/lit mice treated with GH. The net gain in lean body mass was significantly lower in mice treated with GH between d 7 and 34 compared with other treatment periods. In contrast to lean body mass, % body-fat was higher in GH-deficient mice and decreased upon GH treatment. The net loss of % body-fat increased as the age of the animals increased and was 4-fold greater when GH was administered in the adults compared with GH administration during the pubertal growth period. It remains to be determined whether the greater net loss of body fat in adult mice is related to the higher fat content in these mice. The observed lipolytic effects of GH in lit/lit mice were similar to the effects of GH on fat mass in human studies (11, 39–41).
The potential molecular mechanisms that contribute to growth period-dependent GH response in target tissues remain to be established. In this regard, it has previously been shown that GH-responsive, sexually dimorphic hepatic genes are expressed at a low level or not at all in prepubertal male rat liver and that the unresponsiveness of prepubertal rat liver to signal transducer and activator of transcription-5-stimulated gene expression is probably attributable to the apparent lack of liver-enriched transcription factors and/or androgen-dependent factors that are necessary for signal transducer and activator of transcription-5-stimulated gene expression (24). Our findings that GH treatment caused a much greater increase in serum IGF-I level in postpubertal mice compared with pubertal mice are consistent with the idea that the sensitivity to GH may be dependent on the age of the animals. The issue of whether the age-dependent differences in GH sensitivity in bone and other tissues in lit/lit mice is relevant to animals and humans that may become GH-deficient only in the postnatal age remains to be established. In this regard, lit/lit mice are genetically deficient in GH due to mutation in GHRH receptor gene, and therefore GH deficiency exists from conception. The lack of GH from early embryogenesis may predispose these animals to have GH sensitivity by influencing expression of GH receptor or other components of GH/IGF axis.
In summary, the results of this study indicate that GH treatment increased bone accretion and lean body mass but decreased % body-fat in GH-deficient mice and that net change in bone accretion was greatest when GH was administered immediately after puberty while the net change in % body-fat was greatest when GH was administered during adulthood in GH-deficient lit/lit mice. The molecular pathways that contribute to growth period-dependent GH sensitivity in target tissues may lead to a better understanding of GH action and the development of more effective therapies for treatment of GH-deficient patients.
We are grateful to Dr. Leah Rae Donahue (Jackson Laboratory, Bar Harbor, ME) for providing us breeding pairs of lit/lit mice for our studies. We would also like to acknowledge the secretarial assistance provided by Sean Belcher.
This material is based upon work supported in part by the NIH (AR31062), the National Medical Test Bed, and the United States Department of the Army under Cooperative Agreement DAMD17-97-2-7016. The view, opinions, and/or findings contained in this report are those of the author(s) and should not be construed as a position, policy, decision, or endorsement of the Federal Government or the National Medical Technology Testbed, Inc. All work was performed in facilities provided by the Department of Veterans Affairs.