Keratins as Susceptibility Genes for End-Stage Liver Disease Nam–On Ku, Joseph K. Lim, Sheri M. Krams, Carlos O. Esquivel, Emmet B. Keeffe, Teresa L. Wright, David A.D. Parry, M. Bishr Omary  Gastroenterology  Volume 129, Issue 3, Pages 885-893 (September 2005) DOI: 10.1053/j.gastro.2005.06.065 Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 1 Analysis of K8/K18 mutations. (A) The keratin central rod domain, which consists of α-helical subdomains, is flanked by non–α-helical head/tail domains. The head/tail domains are further subdivided into E, V, and H regions. The rod subdomains are connected by nonhelical linker (L1, L12, and L2) regions. The amino acid (aa) regions in black bars represent 5 domains that were previously examined for K8/K18 mutations,14–16 whereas the regions in gray bars represent the 10 exonic K8/K18 domains that are analyzed herein. (B) A representative pattern of genomic DNA polymerase chain reaction products, analyzed by using denaturing high-performance liquid chromatography, from a control and a liver disease patient with the K8 R340H variant. The control (K8 WT) is characterized by 1 major peak, whereas the K8 R340H shows a different double-peaked chromatogram due to resolution of homoduplexes from heteroduplexes, thereby suggesting the presence of a K8 variant. Electropherograms from DNA sequencing confirm the presence of a K8 R340H heterozygous missense mutation (CGT→CAT). Gastroenterology 2005 129, 885-893DOI: (10.1053/j.gastro.2005.06.065) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 2 Ethnic, liver disease, and protein backbone distribution of K8/K18 variants. (A) The ethnicities and sexes of the 58 liver disease patients with keratin mutations were 26 white (11 female and 15 male), 7 black (4 female and 3 male), 12 Hispanic (8 female and 4 male), 1 Indian (1 female), 1 Asian (1 male), and 11 unknown. **Three patients had a double mutation (ie, the 58 patients are counted only once): 1 patient with PBC had K8 R340H and K18 R260Q, 1 patient with AFH had K8 R340H and K18 T102A, and 1 patient with unknown liver disease etiology had K18 I149V and K18 T294M. †P < .0001 when comparing the keratin variant frequency in liver explants vs controls. (B) The amino acid positions of the K8 and K18 domains/subdomains are indicated, and each arrowhead represents an independently identified alteration at the indicated residue. The 30 independent K8 R340H variants are highlighted with a single large arrowhead. The most common K8/K18 variant is K8 R340H (30 of 58; approximately 52%), followed by K8 G61C (6 of 58) and K8 Y53H (5 of 58). Shaded areas at the end and beginning of the rod domain correspond to epidermal keratin mutation hot spots. Note that there is no overlap of the location of K8/K18 mutations with the major locations of epidermal keratin mutations. Gastroenterology 2005 129, 885-893DOI: (10.1053/j.gastro.2005.06.065) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 3 Conservation of K8 R340 among type II keratins and confirmation of K8 R340H protein expression in liver explants. (A) Single-letter abbreviations are used for amino acids, and bold dots indicate amino acids identical to the WT human K8 sequence. The shaded area highlights the conserved R340 of K8. h, human; m, mouse; r, rat; f, frog. (B) Baby hamster kidney cells were transiently cotransfected with K8/K18 WT or K8 R340H and K18 WT. Total cell lysates were prepared and blotted with anti–K8 R340 (ie, WT K8) or anti–K8 H340 (ie, variant K8) epitope-specific antibodies. (C) Total lysates were obtained from a liver with normal K8 (WT; lane 1) or from 5 livers harboring the K8 R340H variant (lanes 2–6). Samples were analyzed as described in (B). Arrowheads highlight degraded K8. Ab, antibody. Gastroenterology 2005 129, 885-893DOI: (10.1053/j.gastro.2005.06.065) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 4 Mutant K8/K18 expression in explanted livers and their biochemical properties. (A) Liver homogenates were prepared from 2 histologically normal livers (WT) and 2 livers with K8 R453C or K18 E275G variants, followed by separation with SDS-PAGE under reducing and nonreducing conditions and then blotting with anti-K8 antibody. Open arrow indicates degraded K8. (B) K8/K18 immunoprecipitates were obtained from normal or K18 E275G livers and then separated by isoelectric focusing followed by SDS-PAGE then blotting with anti-K18 antibody. Assignment of the isoforms was confirmed by analyzing a mixture of normal and mutant samples (not shown). The arrow points to the nonphosphorylated species of K18 E275G. The #4 isoforms of K18 E275G are more intense than in WT K18 because of keratin hyperphosphorylation after liver injury.28,29 (C) K8/K18 immunoprecipitates were prepared from WT or mutant liver explants or from cells transfected with the indicated mutant or WT K8/K18 and then analyzed by SDS-PAGE followed by Coomassie blue staining. Arrowheads highlight the faster migrating keratins due to truncation/deletion. (D) Baby hamster kidney cells were cotransfected with the indicated keratins followed by immunoprecipitation and then in vitro phosphorylation by p42 MAPK and [γ-32P]-adenosine triphosphate. Precipitates were then analyzed by SDS-PAGE, Coomassie staining, and autoradiography. HK8 represents a hyperphosphorylated form of K8. (E) Baby hamster kidney cells were transfected with WT K18 and 1 of the indicated K8 constructs and then incubated in the presence or absence of 20 mmol/L H2O2 (1 hour) followed by isolation of a detergent-free cytosolic soluble fraction or a total cell lysate. The isolates were analyzed by immunoblotting with antibodies to K8 (upper panel) or K18 (middle panel). Coomassie staining (lower panel) is also shown to demonstrate equal gel loading. Open arrow indicates degraded K8. Gastroenterology 2005 129, 885-893DOI: (10.1053/j.gastro.2005.06.065) Copyright © 2005 American Gastroenterological Association Terms and Conditions