ABSTRACT: Iron deficiency-induced anemia is generally a representative nutritional problem in most populations. We reported that the anemia due to dietary iron deficiency causes a variety of changes in nutrient metabolism, even leading to apoptosis as a result of associated endoplasmic reticulum (ER) stress in the rat liver. On the other hand, it appears that non-anemic iron-deficiency causes no serious problem because no appreciable down-regulation of hemoglobin synthesis occurs. Biochemically, iron is essential for activation of cytochrome-related enzymes and its deficiency should yield some physiological problems. We performed a comprehensive transcriptome analysis to define the effects of non-anemic iron deficiency on hepatic gene expression. Four-week-old rats were fed a low-iron diet (ca. 3 ppm iron) for 2 days. These rats were compared with those fed a control diet (48 ppm iron) by pair feeding. On day 3, the rats were sacrificed under anesthesia, and their livers were dissected for DNA microarray analysis. Rats in the iron-deficient diet group, showed that their serum ferritin and iron levels decreased with an increase in the serum total iron binding capacity (TIBC) level, while the hemoglobin level was not changed. In the DNA microarray study, we identified 91 up-regulated and 186 down-regulated probe sets that characterized the iron-deficient diet group. In the up-regulated probe sets, genes involved in glucose and lipid metabolic processes were significantly enriched, whereas genes related to organic acid metabolic process, cellular ketone metabolic process, lipid metabolic process, oxidation reduction, response to drug, response to extracellular stimulus and gas transport were significantly enriched in the down-regulated probe sets. These results suggest that even the non-anemic iron-deficiency exerts various influences on nutrient metabolisms in the liver. Three-week-old male rats (Sprague Dawley) were purchased from Charles River Japan (Kanagawa, Japan) and housed in a room maintained at 24 M-BM-1 1M-BM-0C and 40 M-BM-1 5% humidity with a 12-h light/dark cycle (light 08:00M-bM-^@M-^S20:00; dark 20:00M-bM-^@M-^S08:00). Rats were given a normal diet (Research Diets, Inc., New Brunswick, NJ, USA) as control and water for 24 h ad libitum. The composition of the control diet, 48 ppm iron, was based on the AIN-93G diet; Avicel was used in place of cellulose which may contain a trace amount of iron. On day 3, diet was removed at 18:00 and the feeding was conducted between 09:00 and 17:00 for another 4 days to synchronize the feeding behaviors. On day 8, rats were divided into two groups with similar average body weights. Rats in the one group (n = 5) were given ad libitum an iron-deficient diet, ca. 3 ppm iron, that was prepared by removal of iron (ferric citrate) from the control diet, and those in the other group (n = 5) were fed the control diet by pair feeding. On day 1 and day 3 of feeding with the experiment diet, the hemoglobin level of each rat was measured for the blood samples collected from the tail vein. On day 3, each rat was sacrificed under anesthesia after 1.5 h feeding, prior to excising the liver which was soon immersed in RNAlater.