Figure 1.
In pancreas, epiplakin colocalizes with ductular and acinar keratins and its apicolateral localization in acinar cells is dependent on keratin filaments.
(A–L) Immunofluorescence microscopy for epiplakin, K7, K8 and K19 in pancreata of wild-type mice. Nuclear staining is shown for easier distinction of ductal and acinar cells. Epiplakin is expressed in ductal cells (dc) and in the apicolateral compartment (al) of acinar cells (B, F), where it colocalizes with K8 and K19 (C, G). K7 signals were found in ductal cells but were absent from acinar cells (J). (M–P) Immunofluorescence microscopy for epiplakin and K8 in pancreata of K8-deficient (K8−/−) mice. In K8−/− pancreata, epiplakin is expressed in ductal cells but is absent from acinar cells (N). Whole acini are outlined by dashed lines for orientation. Asterisks indicate cytosolic/perinuclear areas. Scale bars, 20 µm.
Figure 2.
Epiplakin and K8 are upregulated during pancreatitis.
Immunohistochemistry for epiplakin and K8 in mouse (A–D) and human (E–H) pancreata. Epiplakin is expressed mainly in ducts of healthy murine pancreas (A and A’). 9 h after pancreatitis induction epiplakin is strongly upregulated which is most apparent in acinar cells (C). Epiplakin closely parallels K8 staining that is predominantly ductal under basal conditions (B and B’) and is robustly upregulated in acinar cells during acute pancreatitis (D). (E) Quantitative RT-PCR analysis of pancreatic RNA extracts confirmed the strong epiplakin overexpression 6 h after pancreatitis induction. Note that epiplakin transcript levels are already significantly decreased at the 12 h timepoint. 3 mice per group and timepoint were used and the values were normalized to the epiplakin levels in untreated mice, which were arbitrarily set as 1. Data are expressed as mean; error bars represent the s.e.m.; n = 3; *, p ≤ 0.05. (F–I) In healthy adult human pancreata, epiplakin is detected primarily in small ducts (arrows in F and F’), where it colocalizes with K8 as seen on consecutive sections (arrows in G and G’). K8 is also expressed in the apical compartment of acinar cells. In acute necrotizing pancreatitis, acinar overexpression of epiplakin (H) and K8 (I) occurs. Rectangles in A, B, F, and G highlight areas that are shown at a higher magnification in A’, B’, F’, and G’, respectively. Scale bars, 25 µm.
Figure 3.
Subcellular localization of epiplakin in acinar cells is altered during pancreatitis-induced keratin reorganization.
In healthy pancreas (control) epiplakin is found at the apicolateral compartment (al) of acinar cells (A) colocalizing with apicolateral K8 filaments (B, C). Faint K8 filaments are found in the cytoplasm throughout the acinar cell (B), where epiplakin is hardly detectable (A). 9 h after induction of experimental pancreatitis epiplakin and K8 undergo a similar redistribution (F), but compared to epiplakin (D), K8 is more homogeneously distributed throughout the cytoplasm of acinar cells (E). Whole acini are outlined by dashed lines for orientation. Asterisks indicate cytosolic/perinuclear areas. Scale bars, 10 µm.
Figure 4.
Epiplakin deficiency aggravates caerulein-induced murine pancreatitis.
Histological (A–H), morphometrical (I) and serological (J) analyses of caerulein-induced pancreatitis in wild-type and EPPK−/− mice. 6 h after induction of pancreatitis (C, D), EPPK−/− pancreata developed increased edema (ed) and neutrophil infiltration (ni). 9 h after the first caerulein injection (E, F), EPPK−/− pancreata displayed more cell death (cd). Control pancreata showed no obvious abnormalities (A, B) and the caerulein-treated pancreata largely recovered after 24 h in both genotypes (G, H). Scale bars, 100 µm. (I) Histopathological evaluation of wild-type (WT) and EPPK−/− mice pancreatic sections 6 h, 9 h, and 24 h after the induction of pancreatitis. EPPK−/− mice showed increased edema formation 6 h after pancreatitis induction and displayed significantly more cell death and a higher overall pancreatitis score at the 9 h timepoint compared to their wild-type littermates. 24 h after induction no significantly different histopathological scores were found for wild-type and EPPK−/− mice. Horizontal bars represent the median; each symbol represents data from one animal; ns, not significant; n = 12; *, p ≤ 0.05. (J) Serum lipase levels (units per liter) confirm the development of a stronger tissue injury in EPPK−/− mice 6 h after the induction of experimental pancreatitis. (K) Serum amylase levels (units per liter) of EPPK−/− mice during pancreatitis were, although slightly increased, not significantly elevated compared to their wild-type littermates. Data are expressed as mean; error bars represent the s.e.m.; n = 4 for untreated, n = 6 for 6 h and 24 h, and n = 12 for the 9 h timepoint; *, p≤0.05.
Figure 5.
EPPK−/− mice show no abnormalities regarding their acinar keratin organization and junctional complexes.
(A) Immunofluorescence microscopy for K8 in pancreata from wild-type and EPPK−/− mice reveal no differences in keratin network structure of acinar cells. (B–E) Immunofluorescence microscopy analysis using antibodies to desmoplakin (B, C) e-cadherin (D, E) and occludin (F, G) shows no differences in intensity and localization of these junctional proteins between wild-type and EPPK−/− mice in unstressed tissue (B, D, F) and during pancreatitis (C, E, G). Note the disassembly of desmoplakin- and occludin-positive structures during pancreatitis in both wild-type and EPPK−/− mice (B–E). Scale bars, 20 µm.
Figure 6.
Loss of epiplakin does not impair pancreatic enzyme secretion.
(A) Zymogen granules in wild-type and EPPK−/− mouse pancreata were analyzed by transmission electron microscopy. Scale bars, 5 µm. (B) Average diameter of granules was determined by quantification of at least 4 independent sections from 2 animals per genotype. At least 200 zymogen granules were measured per mouse. No significant difference in zymogen granule diameter was found. Data are expressed as mean; error bars represent the SD. (C) Spontaneous and caerulein-induced amylase release of pancreatic lobules isolated from wild-type and EPPK−/− littermates was measured at 3 timepoints after incubation. Values represent amylase release as percentage of total lobular amylase content. Note that no significant differences were found in spontaneous as well as stimulated secretion in wild-type and EPPK−/− mice. Data are expressed as mean; error bars represent the s.e.m.; n = 3 for 60′ and 90′, n = 6 for 30′.
Figure 7.
Increased numbers of keratin aggregations in pancreatic acinar cells of EPPK−/− mice during caerulein-induced pancreatitis.
(A–D) Immunofluorescence microscopy of epiplakin (A, B) and K8 (C, D) in pancreata of wild-type and EPPK−/− littermates 9 h after the first caerulein injection. Note the increased occurrence of keratin granules in pancreata from EPPK−/− mice (D, D’). Dotted lines indicate individual acini. Asterisks, cytosolic regions of acinar cells. Arrow, keratin granule. Scale bars, 50 µm (A-D); 10 µm (D’). (E) Statistical evaluation of pancreatic areas showing keratin granules during acute pancreatitis. 4.5 mm2 of each pancreas section were scored for the occurrence of acinar cells displaying keratin granules. Data are expressed as mean; error bars represent the s.e.m.; n = 6; *, p≤0.05.