Human osteoblast-like cells (HOB) produce vascular endothelial growth factor (VEGF), the steady state level of which is stimulated by 1,25-dihydroxyvitamin D3 [1,25-(OH)2D3]. As osteoblasts and endothelial cells are proximally located in skeletal tissue, we investigated the anabolic effects of 1,25-(OH)2D3 and VEGF on HOB cocultured with endothelial cells. When HOB with high alkaline phosphatase (Al-P) activity and human umbilical vein endothelial cells (HUVEC) with little activity were cultured together, Al-P activity increased, accompanied by an increase in cell number. When HOB and HUVEC were cultured separately, 1,25-(OH)2D3 did not directly stimulate [3H]thymidine incorporation into HUVEC, but stimulated it in the presence of HOB. VEGF did not directly stimulate the Al-P activity of HOB but stimulated it in the presence of HUVEC. The conditioned medium of HOB stimulated the proliferation of HUVEC, and this was partially blocked by anti-VEGF antibody. Conversely, the conditioned medium of HUVEC increased Al-P activity and [3H]thymidine incorporation into HOB, and this was partially blocked by antiinsulin-like growth factor I antibody and BQ-123, a specific antagonist of the endothelin-1 (ET-1) receptor. 1,25-(OH)2D3 stimulated the release of VEGF and ET-1 from HOB and HUVEC, respectively. Furthermore, the 1,25-(OH)2D3-induced release of VEGF was enhanced in HOB cocultured with HUVEC. A quantitative reverse transcription-PCR study revealed that genes for VEGF receptors (Flt-1 and KDR) were expressed in HUVEC, but not in HOB, and that 1,25-(OH)2D3 increased the levels of expression of VEGF receptor genes in endothelial cells only when cocultured with HOB. In summary, we demonstrated that 1,25-(OH)2D3 exerts an anabolic effect on osteoblasts by enhancing their production of VEGF, which stimulates its receptors on endothelial cells, followed by increased production of osteotropic growth factors, such as insulin-like growth factor I and ET-1. These in vitro findings suggest that the VEGF/VEGF receptor system may be involved in both bone formation and bone remodeling in vivo.

To elucidate the pathogenesis of thyroid gland hypervascu-

No abstract

Four patients with adrenocorticotropic hormone (ACTH)-independent bilateral adrenocortical macronodular hyperplasia (AIMAH) were examined. All of them were men whose ages ranged from 37 to 52 years. Plasma cortisol levels were high, with a loss of diurnal rhythmicity, and plasma ACTH was undetectable. Adrenal cortisol secretion was not suppressed by dexamethasone, but it was ACTH responsive. Test results for corticotropin-releasing hormone (CRH) also were negative. Image analyses revealed a normal sella turcica and significantly enlarged adrenal glands, which showed enhanced uptake of isotope. Both adrenal glands in all cases were between 72 and 176 g in combined weight and were composed of, and distorted by, yellow nodules. Histologically, small cortical cells with or without lipid, occasional clear cells, and rare compact cells of the usual size were increased in number in the glandular cords. Enzyme histochemically, cortical cells showed weaker activity for 3 beta hydroxysteroid dehydrogenase and other enzymes than did usual cortisol-producing adenomas. Ultrastructurally, they had moderately to poorly developed smooth endoplasmic reticulum. Nonnodular areas of the cortex consisting of nonproliferating cells were atrophic and contained no compact cell zone. This is similar to the adrenal cortices attached to cortisol-producing adenomas. These features are unique to AIMAH and suggest the presence of a distinct subtype of Cushing's syndrome.

In order to investigate the risk factors, pathogenesis and natural course of the osteoporosis frequently seen in anorexia nervosa, we measured the bone mineral density (BMD) of the lumbar spine using dual X-ray absorptiometry in 51 Japanese female patients with anorexia nervosa, and followed the change in BMD of 29 patients for 11 to 46 months. We also evaluated the serum osteocalcin and the urinary CrossLaps, degradation products of collagen I, in 103 samples obtained from 51 patients. There was a significant correlation between the spinal BMD and the duration of emaciation below a body mass index (BMI) of 15 kg/m 2 (r ¼ ¹0.652, P < 0.0001) and 16 kg/m 2 (r ¼ ¹0.647, P < 0.0001). The increase in BMD per year in the 29 patients significantly correlated with the BMI at the time of entry of each follow-up period (r ¼ 0.712, P < 0.0001). The critical BMI for a positive increase in BMD was 16.4 Ϯ 0.3 kg/m 2 (mean Ϯ S.E.M.). The serum osteocalcin declined, while the urinary CrossLaps increased in proportion to a decrease in BMI. Both markers were normalized in patients whose BMI was between 16.4 and 18.5 kg/m 2 . The ratio of urinary CrossLaps to serum osteocalcin correlated with BMI (r ¼ ¹0.664, P < 0.0001).We conclude that the body weight history is the most important predictor of the presence of osteoporosis as well as of recovery. The BMD of patients does not increase to the normal range even several years after the recovery from this disorder, and they remain a high-risk group for osteoporosis in the future.