Saccharomyces cerevisiae was engineered to express different amount of heavy (H)-and light (L)-chain subunits of human ferritin by using a low-copy integrative vector (YIp) and a high-copy episomal vector (YEp). In addition to pep4::HIS3 allele, the expression host strain was bred to have the selection markers leu2 ؊ and ura3 ؊ for YIplac128 and YEp352, respectively. The heterologous expression of phytase was used to determine the expression capability of the host strain. Expression in the new host strain (2805-a7) was as high as that in the parental strain (2805), which expresses high levels of several foreign genes. Following transformation, Northern and Western blot analyses demonstrated the expression of H-and L-chain genes. The recombinant yeast was more iron tolerant, in that transformed cells formed colonies on plates containing more than 25 mM ferric citrate, whereas none of the recipient strain cells did. Prussian blue staining indicated that the expressed isoferritins were assembled in vivo into a complex that bound iron. The expressed subunits showed a clear preference for the formation of heteropolymers over homopolymers. The molar ratio of H to L chains was estimated to be 1:6.8. The gel-purified heteropolymer took up iron faster than the L homopolymer, and it took up more iron than the H homopolymer did. The iron concentrations in transformants expressing the heteropolymer, L homopolymer, and H homopolymer were 1,004, 760, and 500 g per g (dry weight) of recombinant yeast cells, respectively. The results indicate that heterologously expressed H and L subunits coassemble into a heteropolymer in vivo and that the iron-carrying capacity of yeast is further enhanced by the expression of heteropolymeric isoferritin.

If "gene" contributes to disease risk only in the presence of exposure, the existence of the gene-environment interaction can be efficiently inferred from a deliberately mis-specified "gene-only" disease model in nested case-control studies.

A survey of monthly average concentrations of sulfur dioxide (SO2) and hydrogen sulfide (H2S) at rural locations in western Canada (provinces of Alberta, British Columbia, and Saskatchewan) was conducted in 2001-2002, as part of an epidemiological study of the effects of oil and gas industry emissions on the health of cattle. Repeated measurements were obtained at some months and locations. We aimed to develop statistical models of the effect of oil and gas infrastructure on air concentrations. The regulatory authorities supplied the information on location of the different oil and gas facilities during the study period and, for Alberta, provided data on H2S content of wells and flaring volumes. Linear mixed effects models were used to relate observed concentrations to proximity and type of oil and gas infrastructure. Low concentrations were recorded; the monthly geometric mean was 0.1-0.2 ppb for H2S, and 0.3-1.3 ppb for SO2. Substantial variability between repeated measurements was observed. The precision of the measurement method was 0.005 ppb for both contaminants. There were seasonal trends in the concentrations, but the spatial variability was greater. This was explained, in part, by proximity to oil/gas/bitumen wells and (for SO2) gas plants. Wells within 2 km of monitoring stations had the greatest impact on measured concentrations. For H2S, 8% of between-location variability was explained by proximity to industrial sources of emissions; for SO2 this proportion was 18%. In Alberta, proximity to sour gas wells and flares was associated with elevated H2S concentrations; however, the estimate of the effect of sour gas wells in the immediate vicinity of monitoring stations was unstable. Our study was unable to control for all possible sources of the contaminants. However, the results suggest that oil and gas extraction activities contribute to air pollution in rural areas of western Canada.

The group-based exposure assessment has been widely used in occupational epidemiology. When the sample size used to estimate group means is "large", this leads to negligible attenuation in the estimation of odds ratio. However, the bias is proportional to the between-subject variability and is affected by the difference in true group means. We explore a Bayesian method, which adjusts in a natural way for the extra uncertainty in the outcome model associated with using the predicted values as exposures. We aim to improve the estimate obtained in naïve analysis by exploiting the properties of Berkson type error structure. We consider cases where differences in the proximity of group means and the between-subject variance are both large. The results of the simulations show that our Bayesian measurement error adjustment method that follows group-based exposure assessment improves estimates of odds ratios when the between-subject variance is large and group means are far apart.