The authors would like to apologize for any inconvenience caused. Characterization of Xk(−/−) and Kel(−/−)/Xk(−/−) mice. The construct map of the targeted disruption of mouse Xk(−/−) is shown in the Figure S1. The targeting strategy of Xk was to replace a wild type 806-bp segment that includes partial 5′ end of exon 3 and its flanking intron 2 with a neomycin resistant gene cassette (1.85 kb). The neomycin resistant gene cassette contains an EcoRV site that wild type

Xk does not have resulting in different EcoRV restriction map in a Southern Blot analysis (Fig. S2). The wild type Xk yields check details two bands, 5.6 and 2.2 kb in size and the disrupted Xk gene yields 5.6 and 2.2 kb bands. The 2.2 kb-band is common in both genes, which could be used as an internal control for the southern blot analysis. The probe used for the Southern blot analysis was prepared from the fragment that includes only the middle EcoRV site shown in Fig. 1 as a filled oval circle. The description for Kel(−/−) gene and its Southern blot analysis was reported previously (1); Kel-yields 15 kb band and Kel + band yields 8 kb band upon digestion of genomic DNA with EcoRV in the Southern blot analysis. The mouse Xk has 80% amino acid similarity with human XK and is organized Selleck INK-128 in 3 exons as the human counterpart. The mouse Xk(−/−) gene and the wild type Xk gene are shown in the supplemental

figure S3. To produce Kel(−/−) or Xk(−/−) mice to have homogeneous C57BL/6 background, female Kel(−/−) or male Xk(−/y) mice were mated to C57BL/6 mice (Charles River Laboratories) http://www.selleck.co.jp/products/Adrucil(Fluorouracil).html and backcrossed for 10 generations by breeding heterozygous or hemizygous offspring with C57BL/6 mates. To generate double-knockout [Kel(−/−)/Xk(−/−)] mice, male Kel(−/−) mice with C57BL/6 background and female Xk(−/−) mice with C57BL/6 background were used in the initial mating. The phenotypes of the red cell

ghosts of the three knockout mouse lines with Xk(−/−), Kel(−/−) or Kel(−/−)/Xk(−/−) were analyzed by Western blot and compared with the results of the wild type mouse to confirm the absence of XK, Kell or both in the red blood cells of Xk(−/−), Kel(−/−) or Kel(−/−)/Xk(−/−) double knockout mice, respectively. The results are shown in the supplemental figure S4. As expected XK (lane 3 of left panel), Kell (lane 2 of right panel) or both XK and Kell (lanes 4 of both panels) are absent in the red blood cells of Xk(−/−), Kel(−/−) or Kel(−/−)/Xk(−/−) double knockout mice, respectively. Similar to human Kell null red blood cells and McLeod red blood cells, mouse Kel(−/−) red blood cells have markedly reduced level of XK protein (lane 2 of left panel probed with anti-XK) and Xk(−/−) red blood cells have markedly reduced level of Kell protein (lane 3 of right panel probed with anti-Kell), respectively. References 1.) X. Zhu, A. Rivera, M.S. Golub, et al.