Connexin32 knockout mice (Cx32-KO) exhibit increased chemical-and radiation-induced liver and lung tumor formation with many lung tumors demonstrating decreased levels of the tumor suppressor p27 KIP1 . To determine if p27 deficiency alters Cx32-influenced tumorigenesis, we have generated a Cx32/p27 double-deficient mouse strain (DKO) and show here that exposure of these mice to X-ray radiation resulted in an increase or decrease in tumorigenesis depending on the tissue. Several tissues were highly sensitive to loss of p27 tumor suppressor function (intestine, adrenal, pituitary) resulting in an increased overall tumor burden in DKO mice compared to both wild-type (Po0.005) and Cx32-KO mice (P ¼ 0.066). However, additional deletion of p27 in a Cx32-KO background resulted in a statistically significant decrease in the liver tumor incidence suggesting that Cx32 and p27 pathways mechanistically interact. Immunohistochemical analysis revealed an increased percentage of Cx32-KO liver and lung tumors harboring active mitogen-activated protein kinase (Erk1, Erk2) pathways in contrast to lower percentages of activated wild-type (Po0.005) and DKO tumors (P ¼ 0.027). Increased MAPK activation in liver tumors did not correlate with Ha-ras codon-61 mutation status. This study demonstrates that tissues dependent on Cx32 tumor suppression, such as the liver and lung, exhibit altered tumorigenesis and tumor biology (MAPK pathway activation) related to p27 status.

The results suggest that the FFQ provides a useful marker of soy food consumption and isoflavone exposure in this population.

Background-β-glucuronidase, an acid hydrolase that deconjugates glucuronides, may increase cancer risk; however, little is known about factors associated with human β-glucuronidase.

No abstract

A radioimmunoassay for parathyroid hormone is described that circumvents some previous difficulties. Parathyroid hormone binds strongly to other serum components. Consequently, a "control" curve is set up for each specimen, the specimen consisting of serum, diluent, and radiolabel, but no antibody. Final calculations are then based on the difference between the control curve and the curve obtained by incubating serum, diluent, radiolabel, and antibody. The lack of parallelism between various sample curves makes it almost mandatory that curves derived from at least 12 dilutions of serum be included for each specimen.