Raphael J. Witorsch

An Updated Weight of the Evidence Evaluation of Reproductive and Developmental Effects of Low Doses of Bisphenol A

There is controversy over whether low doses of bisphenol A (BPA, CAS no. 80-05-7) cause reproductive and developmental effects in humans. We update the 2004 weight-of-evidence assessment of an expert panel convened by Harvards Center for Risk Analysis by critically evaluating over 50 additional studies published between April 2002 and February 2006 that examine in vivo reproductive and developmental toxicity in mammals at doses ≤5 mg/kg-d. Our findings are consistent with the Harvard study: some statistically significant findings in rats and mice exist but they are generally countered by more numerous studies showing no effect for similar endpoints. No effect is marked or consistent across species, doses, and time points. Some mouse studies report morphological changes in testes and sperm and some non-oral mouse studies report morphological changes in female reproductive organs. Owing to lack of first-pass metabolism, results from non-oral studies are of limited relevance to oral human exposure. Human biomonitoring indicates exposures lower than the “low” doses in the reviewed animal studies. Reports of human health impact are very limited and inconsistent. Taken together, the weight of evidence does not support the hypothesis that low oral doses of BPA adversely affect human reproductive and developmental health.

Weight-of-Evidence Evaluation of Reproductive and Developmental Effects of Low Doses of Bisphenol A

Recent public concern has focused on potential reproductive and developmental effects from exposure to low levels of bisphenol A (BPA, CAS number 80-05-7). Two previous published reviews (; ) conducted weight-of-evidence evaluations of in vivo reproductive/developmental toxicity from BPA exposure ≤ 5 mg/kg-d based on studies published through February 2006. Here, an update of those analyses presents additional relevant studies that were published through July 25, 2008, and a weight-of-evidence analysis of the studies evaluated in all three reviews. As with the earlier literature, positive findings: (1) are countered by null findings in more numerous studies; (2) have not been replicated; (3) do not exhibit coherence and plausibility; (4) do not show consistency across species, doses, and time points; and/or (5) were from studies using non-oral exposure routes. Owing to the lack of first-pass metabolism, results from non-oral studies are of limited relevance to human exposure. Exposure levels in most of the low-dose oral and non-oral animal studies are generally much higher than those experienced by even the most exposed people in the general population. The weight of evidence does not support the hypothesis that low oral doses of BPA adversely affect human reproductive and developmental health.

Comparison of Glucocorticoid-Induced Effects in Prolactin-Dependent and Autonomous Rat Nb2 Lymphoma Cells

Summary Abstract. Cultured Nb2 node rat lymphoma cells require lactogenic hormone for their proliferation. We reported previously that dexamethasone (Dex) inhibits prolactin (PRL)-induced mitogenesis and, in the absence of mitogen, induces apoptosis of Nb2 cells. Both antiproliferative and cytolytic effects of Dex on Nb2 cells appear to involve glucocorticoid (Type II) receptor mediation. In this study, we compared Dex effects in PRL-dependent Nb2 cells (Nb2) with SFJCD1 (SF), a clone of Nb2 cells that proliferates independently of exogenous PRL. Proliferative assays involved a 72-hr incubation in a chemically defined, serum-free medium where ovine PRL (1 ng/ml) was added to Nb2 cells but not to SF cells. Both cell lines were responsive to the antiproliferative effects of Dex in a dose (6.25-200 nM)-dependent fashion of comparable sensitivity and magnitude. Co-incubation with the glucocorticoid receptor antagonist, RU 486, prevented the antiproliferative effect of Dex in both cell lines. In the same medium devoid of PRL, Dex was cytolytic to Nb2 cells and fragmented DNA in a fashion reflective of apoptosis, but was ineffective in SF cells. A dual chamber incubation system revealed no evidence that SF cells produced cytokines that were mitogenic or anticytolytic to Nb2 cells. Both Nb2 and SF cells fragmented DNA in a fashion indicative of apoptosis in the presence of the Ca2+ ionophore, A23187 (1 μM). These studies reveal a basic difference in glucocorticoid responsiveness between the PRL-dependent Nb2 cell line and its PRL-independent subclone, SF. While both cell lines exhibit functional glucocorticoid receptors and the necessary intranuclear machinery for apoptosis, the pathway mediating the latter is inhibited or dysfunctional in SF cells. [P.S.E.B.M. 1993, Vol 203]

Proteolytic modification of rat prolactin by subcellular fractions of the lactating rat mammary gland

The current study explored prolactin proteolysis by rat lactating mammary gland. 125I-labelled rat prolactin was incubated with tissue fractions of lactating mammary gland and the extent of prolactin degradation and fragment formation was visualized and densitometrically quantitated from autoradiographs derived from SDS-polyacrylamide gel electrophoresis under reducing conditions. At pH 4.5, the 25 000 X g pellet of mammary gland converted intact prolactin (23 kDa band) to proteolytic fragments (8-16 kDa bands) in a time- and tissue concentration-dependent fashion similar to that reported previously for rat ventral prostate. The prolactin-degrading and -fragmenting activity in lactating mammary gland was 5-10-times that observed for ventral prostate, the most active male tissue. This activity at acid pH was also demonstrable in other fractions of mammary gland but appeared to predominate in the cytosol. The above activities in mammary gland virtually disappeared at pH 7.4, appeared sensitive to aspartate and sulfhydryl proteinase inhibitors, and insensitive to serine and metalloenzyme proteinase inhibitors. The distribution of this activity could not be correlated with a particular enzyme marker. These characteristics of mammary gland activity differed significantly from those reported previously for prostate. When electrophoresis was conducted under non-reducing conditions, prolactin proteolysis in prostate and mammary gland was primarily associated with the formation of a more slowly migrating product (24 kDa band) with little spontaneous 8-16 kDa fragment formation. Re-electrophoresis of the 24 kDa band under reducing conditions resulted in the appearance of the 8 and 16 kDa fragments. In conclusion, prolactin is proteolytically modified by prostate and lactating mammary gland to a variant of intact hormone (24 kDa band) with a cleavage site in its large loop, by two or more widely distributed, acid-dependent proteinases. Lactating mammary gland, the principal target for prolactin, has the capacity to cleave the hormone in its loop at rates higher than any other tissue examined to date.

Effects of oral erythrosine (2′,4′,5′,7′-tetraiodofluorescein) on thyroid function in normal men

Erythrosine (Er), a tetraiodinated derivative of fluorescein, is a coloring agent widely used in foods, cosmetics, and pharmaceutical products. Because of its high iodine content and previous reports demonstrating an inhibitory effect of erythrosine on hepatic 5-monodeiodination, we studied the effects of this compound on thyroid function and serum and urinary iodide concentrations in normal subjects. Thirty normal men, equally divided into three treatment groups, each received a 14-day course of oral Er in doses of 20, 60, or 200 mg/day. Serum thyroxine (T4), triiodothyronine (T3), reverse T3 (rT3), thyroid stimulating hormone (TSH), protein-bound iodide (PBI), and total iodide concentrations, serum T3-charcoal uptake, and 24-hour urinary iodide excretion were measured on Days 1, 8, and 15. Thyrotropin-releasing hormone (TRH) tests were performed on Days 1 and 15. There were no significant changes in serum T4, T3, rT3, and T3-charcoal uptake values at any dose. In men receiving 200 mg Er/day, the mean basal serum TSH concentration increased significantly from 1.7 +/- 0.1 (SE) on Day 1 to 2.2 +/- 0.1 microU/ml on Day 15 (p less than 0.05), and the mean peak TSH increment after TRH increased from 6.3 +/- 0.5 to 10.5 +/- 1.0 microU/ml (p less than 0.05). There were no significant changes in basal or peak TSH responses in the men receiving 20 or 60 mg Er/day. Significant dose-related increases in serum total iodide and PBI concentrations occurred during all three doses, and significant dose-related increases in urinary iodide excretion occurred during the 60 and 200 mg/day Er doses. These data suggest that the increase in TSH secretion induced by Er was related to the antithyroid effect of increased serum iodide concentrations, rather than a direct effect of Er on thyroid hormone secretion or peripheral metabolism.

Effects of oral erythrosine (2',4',5',7'-tetraiodofluorescein) on the pituitary-thyroid axis in rats

Erythrosine (FD&C Red Dye No.3) is a tetraiodinated derivative of fluorescein. Rats fed a 4% erythrosine diet for 30 months beginning in utero have an increased incidence of thyroid adenomas and adenocarcinomas. These tumors may be secondary to increased stimulation of the thyroid gland by TSH. This study was undertaken to determine if dietary erythrosine disrupts the pituitary-thyroid axis thereby altering serum thyroid hormone levels. TSH levels, or the pituitarys response to TRH. Rats were fed diets containing erythrosine (0.5, 1.0, 4.0%), sodium iodide (0.16%), or fluorescein (1.6%) for 3 weeks after which TRH testing was performed in vivo. Erythrosine produced a dose-dependent increase in serum T4 levels. With the 4% erythrosine diet, serum T4 and T3 levels and the free-T4 index were significantly increased, whereas the free-T3 index were significantly increased, whereas the free-T3 index was unchanged. Rats fed the 4.0% erythrosine diet had an exaggerated TSH response to TRH; 10 min after the TRH injection, serum TSH levels were 80% greater than TSH levels of control rats. Short-term administration of erythrosine to rats decreased hepatic T3 production by decreasing its conversion of T4 to T3, indicating that erythrosine decreases hepatic 5-deiodinase activity. These data demonstrate that dietary ingestion of 4% erythrosine disrupts the pituitary-thyroid axis as evidenced by an increased TSH response to TRH. This effect is mediated by erythrosine or an iodinated metabolite, since ingestion of its fluorescein nucleus had no effect. Erythrosines effects were not likely mediated by iodide, because serum T4 and T3 levels were elevated and iodide administration did not increase the TSH response to TRH. These data suggest that erythrosine increases the pituitarys TSH response to TRH by altering thyrotroph cell conversion of T4 to T3. Chronic erythrosine ingestion may promote thyroid tumor formation in rats via chronic stimulation of the thyroid by TSH.

Modulation of prolactin binding sites in vitro by membrane fluidizers. II: Age-dependent effects on rat ventral prostatic membranes

The objectives of this study were to determine (i) if the age-related changes in 125I-labeled ovine prolactin specific binding of rat ventral prostate was correlated with changes in membrane lipid microviscosity and (ii) if membrane fluidizers produced age-dependent effects on prolactin binding of prostatic membranes. The degree of fluidization was monitored by a fluorescence polarization method using 1,6-diphenylhexatriene. Membrane preparations of ventral prostate glands obtained from immature (24-25 days old), young-adult (80-90 days old) and aged (550-610 days old) male rats were used for prolactin binding and membrane lipid microviscosity measurements. Relative to immature rats, prostatic prolactin binding decreased approximately 50% in young-adult rats and 75% in aged rats. Membrane lipid microviscosity, relative to immature rats, was increased 72% in young-adult rats and 140% in aged rats. Prostatic membranes obtained from immature animals exhibited no significant effects of in vitro alcohol treatment on prolactin binding, whereas, those obtained from aged animals exhibited maximal increase in prolactin binding. The value of the microviscosity parameter, after in vitro alcohol exposure, exhibited no significant changes in immature animals, whereas, this parameter was decreased approximately 15% in young-adults and approximately 30% in aged animals. These data suggest that in vitro fluidization of prostatic membrane exhibits an age-dependent modification of prolactin binding.

Modulation of prolactin binding sites in vitro by membrane fluidizers I. Effects on adult rat ventral prostatic membranes

The objective of this study was to determine if aliphatic alcohols, known fluidizers of certain membranes, could increase in vitro the apparent fluidity and prolactin binding capacity of membrane preparations obtained from ventral prostate glands of adult male rats. The degree of fluidization was monitored by a fluorescence polarization method using 1,6-diphenylhexatriene. Membrane preparations were either incubated with varying concentrations of ethanol, 1-propanol or 1-butanol and 125I-oPRL overnight at room temperature or were exposed to the alcohols for 15 min at room temperature and washed prior to the overnight incubation with ligand. Regardless of the conditions of incubation, alcohol exposure produced the same effects, a dose-dependent elevation and then decline in specific prolactin binding. Butanol produced a maximal 37-42% increase in prolactin binding at a concentration of 1.0%, propanol produced a maximal 40-56% increase in prolactin binding at a concentration of 3.8%, and ethanol produced a maximal 54-77% increase in prolactin binding at concentration of 4.8%. Scatchard analysis of the oPRL binding of ventral prostatic membranes indicated that the in vitro treatment of these membrane fluidizers increased the number of oPRL binding sites rather than the apparent affinity constant. The value of the microviscosity parameter decreased by 10-13%, 13-15% and 21-25%, after a 15 min exposure of prostatic membranes to 1.0% butanol, 3.8% propanol and 4.8% ethanol, respectively. These data suggest that in vitro fluidization of prostatic membrane modifies prolactin binding capacity and are consistent with in vivo prostatic prolactin receptor level-membrane fluidity relationships observed in earlier studies.

The application of immunoperoxidase methodology for the visualization of prolactin binding sites in human prostate tissue

In previous studies the immunoperoxidase method was used to detect intracellular prolactin binding sites in epithelial cells of normal and neoplastic rat prostate. As an extension of this work, the same approach was used to test for and to localize prolactin binding sites in autopsy and biopsy specimens of formalin fixed, paraffin embedded human prostate. Rehydrated tissue sections were exposed to varying concentrations of human placental lactogen or human prolactin and then to human placental lactogen or human prolactin antisera. The loci of hormone binding were visualized by an immunoperoxidase staining sequence. Hormone pretreatment produced immunospecific, dose related staining inside epithelial cells of normal, hyperplastic, and neoplastic human prostate indicative of intracellular prolactin binding. The patterns of intracellular hormone binding in human prostate were similar to those seen previously in rat prostate tissue. The possible involvement of prolactin in prostatic cancer and the potential diagnostic value of the immunoperoxidase approach to hormone binding are discussed.

Biological, immunological and biochemical characterization of cleaved prolactin generated by lactating mammary gland

We have previously shown that rat prolactin is proteolytically cleaved in its loop by peripheral tissues of the rat. Of the tissues examined to date, lactating mammary gland exhibits the highest prolactin-cleaving activity. The objective of this study was to characterize cleaved prolactin, biologically, immunologically and chemically. By modifying an established analytical method, we were able to generate large (micrograms) amounts of cleaved rat prolactin from cell fractions of rat mammary gland which could then be assayed for biological and immunological activity relative to intact hormone. The cleaved product showed no significant difference relative to the intact rat prolactin when assayed for its ability to compete with 125I-labelled ovine prolactin for the prolactin receptor and for its ability to stimulate the proliferation of rat Nb2 lymphoma cells. Cleaved rat prolactin, however, did show a 50-60% reduction in activity relative to intact rat prolactin when assayed by radioimmunoassay. Using Edman degradation and partial amino acid analysis, we determined that the second N-terminus of the cleaved rat prolactin begins at amino acid 149. The divergence of biological and immunological activity produced by proteolytic cleavage in the loop of rat prolactin suggests that biological and immunological sites differ in location. The possible physiological implications of a cleaved rat prolactin molecule generated by target tissue with maintained biological activity and reduced immunological activity are discussed.