Our work demonstrates that two distinct activities are elicited by high-glucose-induced activation of IRE1a in pancreatic b-cells. Likewise, in the AREDS cohort, cigarette smoking was associated with development of geographic atrophy, which is characterized by atrophy of the RPE, and cell death from apoptosis. Cigarette smoke is a strong oxidant generated by 4700 chemical components. The most obvious ultrastructural sign of injury to the RPE was enlargement and loss of basolateral infoldings, which is an established marker of epithelial cell injury from a number of etiologies including oxidative stress. We used vacuole formation as a second sign of RPE change because it is known to occur in RPE cells overlying drusen deposits. We presume that the degree of oxygen free radicals generated from cigarette smoke was involved in ultrastructural damage to the RPE. Our TUNEL experiments showed a clear increase in RPE cell apoptosis in mice exposed to cigarette smoke. The Espinosa-Heidmann et al Lucidenic-acid-LM1 protocol had higher levels of total suspended particulate and carbon monoxide compared to our levels, respectively. In addition, they used significantly older mice than in our study. We selected our protocol based on evidence that this model induces emphysema in mice, and that AMD lesions are thought to develop over a long period of time. The younger age allows us to isolate the effect of cigarette smoke on the RPE from the complex factors related to chronological aging, which remains the most common risk factor for AMD. It is difficult to determine what factors caused preferential injury to the RPE over Bruch membrane in our study. Interestingly, Espinosa-Heidmann et al did not find compelling ultrastructural evidence of RPE cell injury. The simpler ��filtering�� Silicristin techniques classify the subgroups by maximizing the ratio of between-group to within-group variance. Examples of filtering techniques include the Wilcoxon��s rank sum test, Fisher��s Discriminant Analysis, discriminative partial least squares or genetic algorithm – based classification and clustering.