- Email: [email protected]
Expression of ras GTPase isoforms in normal and diseased pancreas
Original Paper Pancreatology 2005;5:205–214 DOI: 10.1159/000085273
Received: March 23, 2004 Accepted after revision: July 6, 2004 Published online: April 22, 2005
Expression of Ras GTPase Isoforms in Normal and Diseased Pancreas Hemant M. Kochera Ron Senkusb Jane Moorheadc Mashal Al-Nawabc Ameet G. Patela Irving S. Benjamina Bruce M. Hendryd a
Department of Surgery, b Electron Microscopy Unit, c Histopathology and d Renal Medicine, King’s College Hospital, Guy’s King’s and St Thomas’ School of Medicine, King’s College London, London, UK
Key Words Ras GTPase Pancreas Cancer Islets
Abstract Background: Ki-Ras is well studied in its oncogenic form in relation to pancreatic pathologies. However, the individual contribution of each of the wild-type Ras isoforms (Ha-, Ki-, and N-) in pancreatic cells in health and disease is unknown. Methods: Archival formalin-fixed, paraffinembedded specimens of normal (n = 6) and malignant pancreas (n = 35) were used for immuno-histochemical detection of Ras isoforms using a modified polymer system. In addition, immunogold labelling for Ras isoforms was done for subcellular localisation under electron microscopy. Results: Pancreatic ductal cells expressed HaRas in the cytoplasm, with Ki-Ras in the apical region and N-Ras (50% of cases) in a supranuclear distribution. Pancreatic acinar cells express all three isoforms with some nuclear expression of Ki-Ras and supranuclear expression of N-Ras. Islets show Ki- and Ha-Ras mainly with differential expression of Ha-Ras ( cells showing less Ha-Ras and more Ki-Ras than cells). Electron microscopy shows that Ha-Ras is mainly localised in the endoplasmic reticulum and Golgi apparatus of the acinar cells with some plasma membrane localisation of Ki-Ras in the ductal cells. There was no change in any of the Ras isoform expression in the ductal or acinar cells in various malignancies studied (Mann-Whitney U test, p 1 0.1).
© 2005 S. Karger AG, Basel and IAP 1424–3903/05/0053–0205$22.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
Conclusions: Ras isoforms have distinct and separate cellular and subcellular distribution that may persist even in the malignantly transformed state. Understanding this distinct functional distribution patterns in detail is an essential step if mutant Ki-Ras is to be targeted in the pancreas by genetic or molecular therapeutic tools. Copyright © 2005 S. Karger AG, Basel and IAP
Introduction
Oncogenic Ras is locked in the ‘on’ form (Ras-GTP) and provides a continuous uncontrolled signal to cell cycle machinery. Nearly 20–30% of all human cancers have oncogenic Ras [1]. Pancreatic cancer is unique, in that it exhibits a high prevalence (nearly 80%) of Ki-Ras mutations [1, 2]. In addition some forms of chronic pancreatitis are also known to harbour similar mutations [3, 4]. It has been proposed that Ki-Ras mutations may be one of the first steps in the multistep molecular pathway towards pancreatic carcinogenesis [5]. There is increasing knowledge that the three main Ras isoforms, namely Ha-Ras, Ki-Ras (4A and 4B) and N-Ras may be functionally different due to their structural differences and spatial distribution [6–8]. This may explain diverse signalling endpoints such as cellular proliferation, migration, and apoptosis on activation of Ras in the same cell. Seufferlein et al. [9] have shown that in 2 pancreatic cancer cell lines (MiaPaCa-2 and Panc-1) possessing oncogenic Ki-Ras
Prof. Bruce M. Hendry Cell Signalling Group, Department of Renal Medicine King’s College London, Bessemer Road London SE5 9PJ (UK) Tel. +44 20 7848 0491, Fax +44 20 7848 0515, E-Mail [email protected]
mutations, TGF-stimulated cell growth is dependent on activation of Ha-Ras [9]. It appears therefore that wildtype Ras proteins retain important cellular functions even in cells expressing oncogenic mutant Ras. The aim of this study was to study the spatial localisation of Ras isoforms in different pancreatic cell types. This would be the initial step in understanding the complex functioning of Ras proteins and interactions between different isoforms in pancreatic pathologies.
Materials and Methods Tissue Selection Normal Pancreas Specimens Light Microscopy. Six of the initial eighteen normal pancreas archival specimens obtained from patients who had a pancreatic resection were used [final diagnosis: benign stricture of bile duct (n = 2), ampullary adenoma (n = 1), cholangiocarcinoma (n = 3)]. Patients (n = 3) had mid-CBD cholangiocarcinoma and the pancreas was unaffected on histology. Twelve of these specimens were excluded because of the presence of dilated duct (n = 5, histologically or radiologically), chronic pancreatitis (n = 5) or equivocal character of the pancreatic tissue (n = 2). Pathological Pancreas Specimens. Archival core biopsy (n = 16) or resection specimens (n = 19) of patients with pancreatic ductal adenocarcinomas, mucinous cystadenocarcinomas, cholangiocarcinoma of the lower end of common bile duct involving the pancreatic duct were used. Normal Pancreas Specimens for Electron Microscopy. Seven fresh pancreas specimens were obtained at resection for suspected cancer with ethical approval and full informed consent. The normal pancreas was obtained from the resection margin of the pancreas, where there was frozen section and further H & E confirmation of normality (absence of tumour or chronic pancreatitis). Four were excluded because of the presence of malignancy (n = 1), chronic pancreatitis (n = 2) or inadequate representative sample (n = 1) at the resection margin on H & E. Three sections from each of these specimens were stained. Reagents and Instruments Light Microscopy Materials. Monoclonal antibodies (mAb) against pan-Ras (clone Ras10), Harvey (clone 235-1.7.1), Kirsten (clone 234-4.2) and Neural (clone F155-277) isoforms of Ras (Oncogene Research Products, USA), polyclonal rabbit anti-glucagon and polyclonal guinea pig anti-insulin (DAKO, USA) were obtained. Secondary antibodies were polymer system labelled with goat anti-mouse secondary antibodies (DAKO Envision™ + system, DAKO USA), goat anti-rabbit IgG (Sigma ImmunoChemicals, USA) and goat anti-guinea pig IgG (Jackson ImmunoResearch Lab. Inc., USA). Dual labelling immunohistochemistry was by Vector VIP substrate kit for peroxidase (Vector Laboratories, USA), in order to give a pink colour. Electron Microscopy Materials. Monoclonal antibodies against Ras isoforms were as above and the secondary antibody was goat anti-mouse IgG conjugated with 10-nm gold particles (British Bio-
206
Pancreatology 2005;5:205–214
cell International, UK). Embedding resin was Lowicryl K4M (Polysciences Ltd, Eppelheim, Germany). 200-mesh nickel grids (Agar Scientific) with formvar coating (polyvinyl formal, BDH, UK) were used. The cutting machine used was Ultrotome III (LKB, Sweden), the cryostat was FS/FAS (Brights, UK) and the electron microscope was JEOL 100C (Japanese Electron Optics Ltd, Japan). Methods Staining for Light Microscopy. 4-m serial sections were used and antibody dilutions were as described and optimised previously [10]. These protocols allowed selective and specific detection of the 3 Ras isoforms [10]. Negative controls (omission of primary antibody) and positive controls with skin were used for each set of experiments. For the purposes of dual staining, consecutive specimens were labelled with only Ras or only hormone antibody in order to identify the type of cell with more certainty. Dual staining with the best colour contrast was Ras staining followed by insulin/glucagon staining. The glucagon staining gave problems due to high background and the Ras staining accentuated this. Hence, all analysis of dual staining was done with Ras and insulin staining alone. Interpretation. Two independent observers who were blind to the type of antibody or the pathology of the specimen used, interpreted the staining pattern as negative (–), marginal (+\–), mild (+), moderate (++) and intense (+++) for various cell types (ductal, acinar, centro-acinar, islet, fibroblasts) for each section and specific subcellular staining pattern, if any, was recorded. These staining patterns were scored 0–4, respectively. Statistical analysis was by Mann-Whitney U test (SPSS 10.1, Chicago, USA) to compare respective cell types between normal and pathological pancreas, with a significance level at p ! 0.05. Electron Microscopy Protocol Fixation and Embedding. Fresh specimens were fixed with 2% formaldehyde with 0.5% glutaraldehyde for 2 h followed by preservation in phosphate buffer. Embedding was with freshly prepared Lowicryl K4M with progressive lowering of temperature [11]. Briefly, tissue was dehydrated at 4 ° C in increasing concentration of methanol and then infiltrated with Lowicryl K4M. The tissue was then embedded in fresh resin in gelatine capsules, immersed in ethanol baths, without air entrapment and photopolymerised in UV light (360 nm, Ultraviolet Prod. Inc., USA) at –35 ° C in a cryostat for 24 h and then at room temperature. Semi-thin (2 m) survey sections were cut and stained with 0.5% toluidine blue. Appropriate areas of the blocks were then cut in ultra-thin sections and put on formvar-coated 200-mesh nickel grids. A section was stained with 1% uranyl acetate (25 min) and lead citrate (0.4%, 10 min) to assess adequate fixation and embedding with Lowicryl. Air-dried ultra-thin sections (silver-gold interface, 70–90 nm) were washed in 1:20 normal goat serum (20 min) and incubated with monoclonal antibodies against Ras: Ha-Ras (1:50), Ki-Ras (1:1), N-Ras (1:10) and secondary antibody (1: 50) and after washes with secondary antibodies conjugated with gold (1:50, 2 h) at room temperature. Counterstaining was done with 1% uranyl acetate. Electron Microscopy and Analysis. Initial scanning at low magnification (!20,000) identified the cells by morphology. Position and counts of the gold particles in the particular subcellular compartment were studied in the nucleus, cytosol and plasma membrane (!35,000). In addition in acinar cell, mitochondria, endo-
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
plasmic reticulum, zymogen granules, Golgi apparatus, were also studied (!60,000). These settings were kept constant to keep the count of gold particles uniform (at per square centimetre: 0.08 m2 for magnification of 35,000! and 0.028 m2 for magnification of 60,000!). Quantitative analysis was performed by counting the number of gold particles per square centimetre; squares were selected randomly using a defined grid. Each photograph had 6 such readings taken for each cellular organelle; where possible, these readings were repeated thrice in other photographs. Statistical analysis was done organelle-wise using paired sample t test for comparing isoforms (SPSS 10.1, USA), with a significance level at p ! 0.01.
Results
Clinical characteristics of patients used throughout the study are summarised in table 1. Immunohistochemistry in Normal Pancreas All negative and positive controls used with each set of experiments on different days were uniformly negative or positive, respectively. Ras protein expression as detected by pan-Ras monoclonal antibody was seen in pancreatic acinar and centroacinar cells, ductal cells, and islet cells as well as stromal fibroblasts, all to a variable degree of intensity. A semi-quantitative summary of the intensity of staining according to the cell type and the isoform is shown in table 2. The expression of the Ras isoforms was cell specific and each isoform had specific subcellular distribution. Pancreatic ductal cells showed intense cytoplasmic staining with pan-Ras (fig. 1A, 3rd column). Pancreatic ductal cells expressed Ha-Ras in the cytoplasm with no
Table 2. Immunohistochemical staining for Ras isoforms in normal pancreas tissue
Fibroblast Acinar cells Centroacinar cells Islets Duct Nerve Mast cells
nuclear expression of the isoform, whilst Ki-Ras was predominantly concentrated on the luminal aspect of the ductal cellular cytoplasm with some (punctate) nuclear staining. In contrast, 50% of ductal cells in the normal pancreas showed a specific supranuclear distribution for N-Ras (fig. 1A, 3rd column). The remaining 50% of cases showed no staining with N-Ras in the ductal cells. The acinar cell complex showed mild staining with pan-Ras. It was predominantly in the cytoplasm of the centroacinar cells with some staining of the acinar cells’ cytoplasm (fig. 1A, 1st column). Specifically with Ki-Ras, there was punctate nuclear staining in 50% of cells in all specimens and some cytoplasmic staining. In addition, there was supranuclear distribution of N-Ras (50% of cases). Ha-Ras stains the acinar complex mildly.
Table 1. Clinical characteristic of patients in the study Patient
Total Male Age: median (range), years
Normal pancreas for light microscopy 6 Normal pancreas for electron microscopy1 3 Pathological pancreas Well differentiated PDA 3 Moderately differentiated PDA 26 Poorly differentiated PDA 2 Cholangiocarcinoma 2 Mucinous cystadenocarcinomas 2
4 2
50 (38–75) 70 (70–76)
3 14 0 2 2
62 (41–74) 61 (56–84) 76 (72–78) 66 (62–70) 56 (48–64)
PDA = Pancreatic ductal adenocarcinoma. None of these patients had previous pancreatitis or ethanol abuse. 1
Pan
Kirsten (Ki)
Harvey (Ha)
Neural (N)
1.1380.4 1.5080.3 1.1780.3 3.1380.3 1.6380.4 1.0280.4 0.5780.6
0.1380.1 0.6380.2 (N) 0.5680.3 1.7580.3 0.6280.2 (N) 0 2.7580.6
0.1480.1 1.0380.4 0 2.2580.21 0.7580.3 0.1780.2 0
0 0.6280.3 (SN) 0 0.8780.3 0.2580.2 (SN) 0 0
The semi-quantitative analysis of Ras isoforms (Ki, Ha, and N) expression in normal human pancreas according to cell type was scored as: negative (–) staining: 0; marginal (+\–): 1; mild (+): 2; moderate (++): 3; and intense (+++): 4. Special characteristics were noted as: SN: supranuclear stain, N: nuclear stain. 1 Differential staining pattern for islet cell subpopulation. All figures are mean 8 standard error of mean.
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
207
A
Pan 1000x
400x
1000x
1000x
400x
1000x
1000x
1000x
1000x
1000x
Ha
Ki 1000x
N
1000x
Acini
Islet
Duct
α
B
β 1000x
Islet: Insulin 208
Pancreatology 2005;5:205–214
1000x
Ha-Ras
1000x
Insulin & Ha-Ras Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
The cells of the islets of Langerhans showed marked cytoplasmic staining with pan-, Ha- and Ki-Ras mAb but mild staining with N-Ras mAb. In addition, there was a differential staining of the and islet cells with Ha-Ras (fig. 1A, 2nd column). With Ha-Ras, a few islets cells, mostly at the periphery with strands branching towards the centre of the islets (believed to be cells in distribution) showed intense cytoplasmic staining. Most of the other islet cells showed mild cytoplasmic staining with Ha-Ras and a very few cells showed no staining. Ki-Ras showed marked cytoplasmic staining. In addition, some punctate nuclear staining could also be discerned. N-Ras showed mild staining with about 75% of the cells and those cells had coarse granular type of cytoplasmic staining with a faint background staining of the cytoplasm. Fibroblasts showed staining with pan- and Ki-Ras mAb in the cytoplasm, but no staining with Ha- or N-Ras mAb. Dual Labelling in Normal Human Pancreatic Islets Both glucagon and insulin monoclonal antibodies worked very well when used alone. The distribution of cells as shown by the glucagon staining were same as that of the cells staining intensely with Ha-Ras. There was background staining with glucagon, which was accentuated with Ras staining and could not be used for analysis for dual staining. However, dual staining with insulin worked very well without background stain, but the colour differences were difficult to discern. Cells have heavier staining with Ki-Ras than cells. Cells (as shown by distribution in the islets, staining with glucagon and not staining with insulin) have intense
brown staining with Ha-Ras. In addition, cells showed some Ki-Ras and minimal staining with N-Ras. In cells (as shown by distribution in the islets, staining with insulin and not staining with glucagon), there was mild staining with Ha-Ras (as compared to cells) (fig. 1B). Ki-Ras stained cells intensely as compared to cells. Pathological Pancreas A semi-quantitative analysis with a non-parametric statistical test (Mann-Whitney U test) was used to compare the data on the respective cells (ductal and acinar) in pancreatic ductal adenocarcinoma, cholangiocarcinoma, and pancreatic mucinous cystadenocarcinomas with those from the normal pancreas. The comparison was made from the pathological area in the pancreas, i.e., malignant ductal cells and distorted acini (acinar cells acted as surrogate controls for cell type) around the malignant area. In some biopsy specimens there were not enough acini present to comment on the staining for them. There was no change in the Ras isoform expression in the malignant ducts as compared to normal ducts, when compared by histology or grade of tumour for any of the Ras isoforms (all isoforms: p 1 0.1) (fig. 2). Ha-Ras showed a trend towards an increase in staining intensity with pancreatic ductal adenocarcinoma, cholangiocarcinoma and mucinous cystadenocarcinomas. There were 2 types of controls for the type of malignancy of interest: pancreatic ductal adenocarcinoma; mucinous cystadenocarcinoma where the Ki-Ras mutation is not ubiquitous (control for Ras mutation) and cholangiocarcinoma of the lower end of the bile duct causing pancreatic ductal obstruction and therefore obstructive chronic pancreatitis (pathological control: no malignancy in the pancreas). There was no change in the expression of Ras isoforms in the acinus in malignant specimens (histology or grade of tumour did not matter) when compared with normal specimens for any of the Ras isoforms (all isoforms: p 1 0.1) (fig. 3).
Fig. 1. A Ras immunohistochemical staining in normal human pancreas. First row is staining with pan-Ras and subsequent 3 rows are with Ha-, Ki- and N-Ras respectively. Left column for staining in the acini, middle column for staining in the islets and right column for staining in the duct. Acinar staining is shown at a higher magnification of the acinar region (1,000!). Red arrows point to the centroacinar cells to differentiate them from the acinar cells. Bent arrows point to the stellate cells in the interstitium. Yellow arrows represent the supranuclear concentration of the N-Ras staining in the acinar cells. The islets of Langerhans: arrows point to the islets and arrowheads point to the cells, whilst bold arrows point to cells. Ducts, the arrowhead points to the mast cell and the yellow arrow to the supranuclear distribution of the N-Ras stain. B Immunohistochemistry and light microscopy with dual staining with insulin and Ha-Ras isoforms, to visualize the islets of Langerhans. Higher resolution (1,000!) microphotographs showing the staining with insulin alone, with Ha-Ras alone and with Ha-Ras and insulin to show the and cells.
Electron Microscopy for Normal Pancreas Ha-Ras showed significantly higher concentration (p ! 0.001, paired t test) than the other isoforms in the mitochondria, Golgi apparatus, zymogen granules and endoplasmic reticulum of the acinar cells; there was a negligible amount of Ras isoforms seen in the nucleus (fig. 4). The centroacinar cells showed more Ras in the cytosol than the nucleus. The plasma membrane of the duct cells showed some Ha-Ras as well as Ki-Ras but no N-Ras (fig. 4). In the cytosol of the duct cells, Ha-Ras was most prominent. The
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
209
3.5
3.0
3.0
2.5
2.5
2.5
2.0 1.5 1.0
2.0 1.5 1.0
1.5 1.0
0.5
0.5
0.0
0.0
0.0
0.5 N
8
31
2
2
N
PDA
MucCA
CCA
0.5 N
8
31
2
2
N
PDA
MucCA
CCA
N
3.5
3.0
3.0
2.5
2.5
2.5
2.0 1.5 1.0
N-Ras duct
3.5
3.0 Ki-Ras duct
3.5
2.0 1.5 1.0 0.0
0.0
0.5
0.5
0.5
3
25
2
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
2
MucCA
CCA
0.5
N
8 N
3
25
2
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
Fig. 2. Semi-quantitative analysis of Ras isoform expression in the ducts of patients with normal pancreas (N) compared with those having pancreatic ductal adenocarcinoma (PDA), mucinous cystadenocarcinoma (MucCA) and cholangiocarcinoma (CCA). Staining in the PDA was further studied as according to the differentiation of the tumour: well, moderately and poorly differentiated. The
method of differentiating cell from and cells is by the way of cytoplasmic granules with varying character. Cells showed more Ha-Ras than the cells, and Ha-Ras was the most prominent isoform expressed in the endocrine cells (fig. 4). Statistical Analysis for Pancreas for Electron Microscopy It will be seen from table 3 that Ha-Ras is expressed in all the cellular organelles, but more so in the acinar Golgi apparatus and the endoplasmic reticulum. Other important observations are that the plasma membranes in the duct cells show both Ha- and Ki-Ras and no N-Ras. The statistical significance and the confidence intervals of the difference in staining with various isoforms are given using the paired t test. Staining in some of the organelles could not be measured in enough sections to give confident estimates and are represented as NA.
Pancreatology 2005;5:205–214
2
PDA
1.0
0.0 8
30
N
1.5
0.5
N
8
2.0
0.5
N
210
2.0
0.5 0.5
Ha-Ras duct
N-Ras duct
3.5
3.0 Ki-Ras duct
Ha-Ras duct
3.5
N
8 N
3
24
2
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
top row amalgamates the readings of all the grades of pancreatic ductal adenocarcinoma to compare it with the other cancers and the normal gland. There is an increase in the staining of Ha-Ras in all the cancers but it does not reach statistical significance (p 1 0.05). The left column is with Ha-Ras, middle with Ki-Ras and right with N-Ras.
Discussion
This set of experiments shows for the first time that there is a definite pattern of Ras isoform distribution in normal pancreatic tissues, which differs not only from cell to cell but also at the subcellular level. Of interest is the fact that the amount of Ras expressed in the pathological pancreas (malignancies) did not change although the mutated Ki-Ras is a well-known early carcinogenic event in pancreatic ductal adenocarcinoma. The functional differences of Ras isoforms have been brought to attention by 2 key biological findings; different Ras isoforms are mutated in various human cancers and the Ras knockout mice models have varied levels of arrest of embryonic development with each isoform [1, 2, 12]. These functional distinctions between Ras isoforms may explain different endpoints of activating the Ras GTPases in the same cell [13, 14].
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
2.5
3.0
3.0
2.0
2.5
2.5
2.0 1.5 1.0 0.5
2.0 1.5 1.0 0.5
0.5 8
19
2
2
N
PDA
MucCA
CCA
N
3.0
3.0
2.5
2.5
2.0 1.5 1.0 0.5 0.0
8
19
2
2
N
PDA
MucCA
CCA
N
N
1
16
1
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
18
2
2
PDA
MucCA
CCA
2.0
2.0 1.5 1.0 0.5
1.5 1.0 0.5 0.0
0.5
0.5 8
8 N
2.5
0.0
0.5
0.5
N
N-Ras acinus
3.5 Ki-Ras acinus
3.5
1.0
0.5
0.5 N
1.5
0.0
0.0
0.0
Ha-Ras acinus
N-Ras acinus
3.5
Ki-Ras acinus
Ha-Ras acinus
3.5
8
N
N
1
16
1
2
2
N
Well Mod Poor MucCA CCA PDA PDA PDA
8
1
15
N
Well PDA
Mod PDA
1
2
Poor MucCA CCA PDA
Fig. 3. Semi-quantitative analysis of Ras expression in the acinus of patients with normal pancreas compared
with those having cancers as in figure 2. No change in the amount of Ras isoforms is seen as compared to the normal pancreas.
Table 3. Differences in staining between isoform in different cell organelles for pancreatic cells
Organelle
Acini nuclear Acini cytosol Acini mitochon Acini zymogen Acini PM Acini ER Acini Golgi Duct PM Duct cytosol Duct nucleus Cac nucleus Cac cytosol Cell Cell
Mean 8 SE (mean)1
p (95% CI)2
Ha-Ras
Ki-Ras
N-Ras
Ha-Ki
Ki-N
Ha-N
0.580.3 0.280.1 12.381.3 4.880.9 NA 2.480.3 13.381.3 1.580.4 380.6 1.380.4 0.580.3 1.881.3 5.780.6 1.880.3
0.580.2 0.180.1 0 0 0 0.380.1 0 0.580.1 0.380.1 0 0.280.2 0.380.2 0.180.1 0.880.3
0.380.2 0.180.1 NA 1.680.5 0 0.780.2 0.380.2 0 0.280.1 0.580.3 0 0 0.380.1 0.380.2
0.363 (–1.2, +0.5) 0.717 (–0.3, +0.4) 0.000 (+8.9, +15.8) 0.000 (+2.7, +6.9) NA 0.000 (+1.4, +2.7) 0.000 (+10.4, +16) 0.039 (+0.06, +1.9) 0.001 (+1.3, +4.0) 0.004 (+0.4, +0.5) 0.465 (–0.8, +1.4) 0.343 (–2.2, +5.2) 0.000 (+4.3, +7.1) 0.035 (+0.09, +2.1)
1.00 (–0.6, +0.6) 0.579 (–0.2, +0.3) NA 0.015 (–2.8, –0.4) NA 0.069 (+0.03, –1.9) 0.111 (–0.8, +0.08) 0.007 (+0.2, +0.8) 0.438 (–0.3, +0.6) 0.082 (–1.1, +0.1) 0.363 (–0.3, +0.6) 0.175 (–0.2, +0.9) 0.191 (–0.6, +0.1) 0.295 (–0.4, +1.2)
0.465 (–1.4, +0.8) 0.430 (–0.2, +0.4) NA 0.010 (+0.6, +4.1) NA 0.000 (+2.4, +5.5) 0.000 (+10.1, +15) 0.004 (+0.6, +2.4) 0.001 (+4.1, +4.7) 0.145 (–0.3, +1.8) 0.203 (–0.4, +1.3) 0.226 (–1.6, +5.2) 0.000 (+4.2, +6.6) 0.001 (+0.8, +2.2)
Mean number of gold particles counted per cm2 (0.08 m2). p values by paired t test (statistically significant values in bold). 95% CI = 95% confidence intervals for the difference in means; NA = not applicable; ER = endoplasmic reticulum; PM = plasma membrane; Cac = centroacinar cell; Mitochon = mitochondria. 1 2
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
211
Ha-Ras, Acinus, x35000
Ha-Ras, Centroacinar, x35000
Nucleus Nucleus
Ha-Ras, Acinus, x60000
Ha-Ras, Duct x35000
Nucleus Lumen
Ha-Ras, Alpha cell x35000
Ha-Ras, Beta cell x35000
Nucleus Fig. 4. Immunogold labeling with Ha-Ras and electron microscopic view of the pancreatic cell types. Top left picture shows an acinar cell and the top right picture a centroacinar cell. There is intense staining particularly in the Golgi apparatus, and mitochondria. The inset shows the staining in the Golgi apparatus at a higher magnification (arrow). There is also some staining in the cytosol of the centroacinar cell. Bar = 0.28 m. Middle left picture shows the typical staining of the Golgi apparatus, and endoplasmic reticulum,
212
Pancreatology 2005;5:205–214
with Ha-Ras at a magnification of 60,000!. Bar = 0.17 m. The middle right picture is of the duct cell along with staining with HaRas at a magnification of 35,000!. The inset shows the staining plasma membrane of the ductal cell. Bar = 0.28 m. Bottom left picture shows the staining of the granules or vacuoles of the cell intensely with Ha-Ras in comparison with those of the cell (bottom right picture). The inset shows the characteristic staining at a higher power. Bar = 0.28 m.
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
In in vitro experiments, Ras isoforms vary in their ability to activate Raf-1 serine/threonine kinase, the best characterised downstream effector of Ras, which resulted in differences in cellular activities in various cell types [15]. Ki-Ras is more effective as a recruiter and activator of Raf-1 to the plasma membrane than Ha-Ras, whilst Ha-Ras recruits PI 3-kinase efficiently [16]. Since PI 3kinases play an important role in prevention of apoptosis, Ha-Ras is believed to be a survival signalling molecule [16, 17]. In the multistep carcinogenesis model of colon cancer, it has been proposed that in the presence of already high activity of PI 3-kinase, mutated Ki-Ras activates Raf/MAP kinase pathway thus achieving synergistic activity [18]. An example of differential spatial distribution is the fact that in IL-2-deprived murine T cell line TS1, Ha-Ras localises to mitochondria, whereas in IL2-stimulated cells, Ki-Ras localises to mitochondria [19]. Thus in different circumstances such as growth promoting or apoptosis, Ras isoforms could be differentially targeted to mitochondria. Thus Ha-Ras localised to mitochondria in acinar cells as seen in this study may represent a survival signal. The main activity of Ras is at the plasma membrane. One such microdomain in the plasma membrane is formed by caveolae [20] and Ha-Ras-, but not Ki-Rasdependent activation of Raf can be blocked by a dominant negative caveolin [21]. Ras trafficking is dependent on post-translational modification. Ha- and N-Ras are palmitoylated on their C-terminal cysteine residues during their transport through the endoplasmic reticulum (and Golgi apparatus). They are co-localised with caveolin. On the other hand, Ki-Ras has lysine-rich sequence at its C-terminal end, which localises it to functionally distinct microdomains of the plasma membrane [22]. NRas passes via the Golgi apparatus, where it may undergo post-transcriptional changes [22]. Apolloni et al. [7] studied the trafficking of Ras isoforms in BHK cells. The first post-translational modification of farnesylation/geranylgeranylation takes place in the cytosol. The C-terminal AAX motif is removed and methylation takes place in the endoplasmic reticulum. The polybasic domain of the KiRas then sorts the protein away from the endoplasmic reticulum and allows it to migrate to the plasma membrane. In taxol-treated cells (disruption of microtubules), Ki-Ras fails to reach the plasma membrane and is found to align in the multivesicular and univesicular endosomal structures [23]. CAAX proteins with incomplete trafficking signal can fail to reach the plasma membrane. In real-time experiments on PtK2 cells, it has been shown that proteins go from trans-Golgi network to trans-
porting carriers (which migrate with fusion and fission) along the microtubules to reach the plasma membrane [24]. Using a transforming N-Ras in NRK cell line, these experiments suggest the role of N-Ras in the carrier transport role of trans-Golgi network TGN and microtubules. In both the pancreatic acinar and ductal cells N-Ras is seen in the supranuclear position where Golgi apparatus is present. Sharpe et al. [25, 26] have shown that in human renal fibroblasts, there are critical and distinct roles for Ki-Ras and Ha-Ras in cytokine-stimulated proliferation without a functional role for N-Ras. Seufferlein et al. [9] have shown that in 2 pancreatic cancer cell lines (MiaPaCa-2 and Panc-1) possessing oncogenic Ki-Ras mutations, TGF-stimulated cell growth is dependent on activation of Ha-Ras. Oncogenic Ki-Ras constitutively activates the MAP-kinase cascade. They have demonstrated that TGF stimulation increases the membrane translocation and activation of Ha-Ras and these cells use the Ha-Ras-ERK pathway to proliferate. This dependence on Ha-Ras in pancreas cancer is surprising in view of the clear co-expression of oncogenic Ki-Ras in these cells. Thus the Ras isoforms can synergistically promote growth and metastasis of pancreatic cancer cells. These in vitro experiments on renal fibroblasts and pancreatic cancer cells strongly support the concept of distinct functional roles for the Ha-Ras and Ki-Ras isoforms and support our findings of different roles for Ras as suggested by the present in vivo data. Interesting observation was made of the differential staining of sub-populations of islet cells, especially with Ha-Ras. This was confirmed by the technique of dual labelling and also immunogold labelling in electron microscopy. Cells producing glucagon, exhibit more Ha-Ras than the cells. Ras may have a role in the insulin-mediated negative regulation of glucagon gene transcription [27]. Gherzi et al. [27] stimulated PKA- and PKC-induced gene transcription using insulinoma (-TC1) and glucagonoma (In-R1-G9) cell lines. Using effector constructs of Ha-Ras, they showed that the PKA and the Ras signalling pathways are functionally antagonistic in these cell lines. Thus Ras has an important role in gene transcription in these human endocrine cells. Ki-Ras was seen predominantly in cells on electron microscopy with NRas present in both types of cells in minimal quantities. The Ras isoforms can be differentially targeted achieving desired and focused therapeutic endpoints. There are numerous therapeutic strategies to target Ras. The most actively used molecules are those inhibiting the posttranslational modification of Ras proteins called prenylation inhibitors such as farnesyl transferase inhibitors
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
213
and geranylgeranyl transferase inhibitor [28]. Farnesyl transferase inhibitor is being used in phase II clinical trials in patients with unresectable pancreatic cancer with or without gemcitabine. Ha-Ras antisense oligonucleotides are being used in phase II clinical trials in patients with pancreatic cancer. Thus Ras targetting, as shown by this body of work, holds an immense potential in treatment of human diseases, provided specific agents could be delivered safely.
Acknowledgements We are most grateful to Kristain Loebner who provided some of the reagents for dual labelling, Mohammed Sohail for digital capture of the microscopic slides and some analysis, Anne Clarke (Oxford) for initial help in developing the electron microscopy work. The Agostino Trappani Foundation Naples supported this work.
References 1 Bos JL: The ras gene family and human carcinogenesis. Mutat Res 1988;195:255–271. 2 Barbacid M: Ras genes. Annu Rev Biochem 1987;56:779–827. 3 Yanagisawa A, Ohtake K, Ohashi K, et al: Frequent c-Ki-ras oncogene activation in mucous cell hyperplasias of pancreas suffering from chronic inflammation. Cancer Res 1993; 53: 953–956. 4 Lemoine NR, Jain S, Hughes CM, et al: Ki-ras oncogene activation in preinvasive pancreatic cancer. Gastroenterology 1992;102:230–236. 5 Hruban RH, Wilentz RE, Kern SE: Genetic progression in the pancreatic ducts. Am J Pathol 2000;156:1821–1825. 6 Mochizuki N, Yamashita S, Kurokawa K, et al: Spatio-temporal images of growth-factor-induced activation of Ras and Rap1. Nature 2001;411:1065–1068. 7 Apolloni A, Prior IA, Lindsay M, et al: H-ras but not K-ras traffics to the plasma membrane through the exocytic pathway. Mol Cell Biol 2000;20:2475–2487. 8 Roy S, Wyse B, Hancock JF: H-Ras signaling and K-Ras signaling are differentially dependent on endocytosis. Mol Cell Biol 2002; 22: 5128–5140. 9 Seufferlein T, Van Lint J, Liptay S, et al: Transforming growth factor alpha activates Ha-Ras in human pancreatic cancer cells with Ki-ras mutations. Gastroenterology 1999; 116: 1441– 1452. 10 Kocher HM, Codd J, Sharpe CC, Dockerell MEC, Al-Nawab M, Hendry BM: Expression of Ras GTPases in normal kidney and glomerulonephritis. Nephrol Dial Transplant 2003; 14:848–854.
214
11 Armbruster BL, Carlemalm E, Chiovetti R, et al: Specimen preparation for electron microscopy using low temperature embedding resins. J Microsc 1982;126(pt 1):77–85. 12 Johnson L, Greenbaum D, Cichowski K, et al: K-ras is an essential gene in the mouse with partial functional overlap with N-ras. [erratum appears in Genes Dev 1997; 11: 3277]. Genes Dev 1997;11(19):2468–2481. 13 Hall A, Cales C, Hancock JF, et al: Analysis of mammalian ras effector function. Cold Spring Harb Symp Quant Biol 1988; 53(pt 2):855– 862. 14 Reuther GW, Der CJ: The Ras branch of small GTPases: Ras family members don’t fall far from the tree. Curr Opin Cell Biol 2000; 12: 157–165. 15 Voice JK, Klemke RL, Le A, Jackson JH: Four human ras homologs differ in their abilities to activate Raf-1, induce transformation, and stimulate cell motility. J Biol Chem 1999;274: 17164–17170. 16 Yan J, Roy S, Apolloni A, Lane A, Hancock JF: Ras isoforms vary in their ability to activate Raf-1 and phosphoinositide 3-kinase. J Biol Chem 1998;273:24052–24056. 17 Downward J: Ras signalling and apoptosis. Curr Opin Genet Dev 1998;8:49–54. 18 Phillips WA, St Clair F, Munday AD, Thomas RJ, Mitchell CA: Increased levels of phosphatidylinositol 3-kinase activity in colorectal tumors. Cancer 1998;83:41–47. 19 Rebollo A, Perez-Sala D, Martinez A: Bcl-2 differentially targets K-, N-, and H-Ras to mitochondria in IL-2 supplemented or deprived cells: Implications in prevention of apoptosis. Oncogene 1999;18:4930–4939.
Pancreatology 2005;5:205–214
20 Kurzchalia TV, Parton RG: Membrane microdomains and caveolae. Curr Opin Cell Biol 1999;11:424–431. 21 Roy S, Luetterforst R, Harding AE, et al: Dominant-negative caveolin inhibits H-Ras function by disrupting cholesterol-rich plasma membrane domains. Nat Cell Biol 1999;1:98– 105. 22 Choy E, Chiu VK, Silletti JD, et al: Endomembrane trafficking of ras: The CAAX motif targets proteins to the ER and Golgi. Cell 1999; 98:69–80. 23 Thissen JA, Gross JM, Subramanian K, et al: Prenylation-dependent association of Ki-Ras with microtubules. Evidence for a role in subcellular trafficking. J Biol Chem 1997; 272: 30362–30370. 24 Toomre D, Keller P, White J, et al: Dual-color visualization of trans-Golgi network to plasma membrane traffic along microtubules in living cells. J Cell Sci 1999;112(pt 1):21–33. 25 Sharpe CC, Dockrell ME, Scott R, et al: Evidence of a role for Ki-Ras in the stimulated proliferation of renal fibroblasts. J Am Soc Nephrol 1999;10:1186–1192. 26 Sharpe CC, Dockrell ME, Noor MI, et al: Role of Ras isoforms in the stimulated proliferation of human renal fibroblasts in primary culture. J Am Soc Nephrol 2000;11:1600–1606. 27 Gherzi R, Briata P, Fehmann HC, Goke B: Ras antagonizes cAMP stimulated glucagon gene transcription in pancreatic islet cell lines. FEBS Lett 1994;353:277–280. 28 Khwaja A, O’Connolly J, Hendry BM: Prenylation inhibitors in renal disease. Lancet 2000; 355:741–744.
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
Received: March 23, 2004 Accepted after revision: July 6, 2004 Published online: April 22, 2005
Expression of Ras GTPase Isoforms in Normal and Diseased Pancreas Hemant M. Kochera Ron Senkusb Jane Moorheadc Mashal Al-Nawabc Ameet G. Patela Irving S. Benjamina Bruce M. Hendryd a
Department of Surgery, b Electron Microscopy Unit, c Histopathology and d Renal Medicine, King’s College Hospital, Guy’s King’s and St Thomas’ School of Medicine, King’s College London, London, UK
Key Words Ras GTPase Pancreas Cancer Islets
Abstract Background: Ki-Ras is well studied in its oncogenic form in relation to pancreatic pathologies. However, the individual contribution of each of the wild-type Ras isoforms (Ha-, Ki-, and N-) in pancreatic cells in health and disease is unknown. Methods: Archival formalin-fixed, paraffinembedded specimens of normal (n = 6) and malignant pancreas (n = 35) were used for immuno-histochemical detection of Ras isoforms using a modified polymer system. In addition, immunogold labelling for Ras isoforms was done for subcellular localisation under electron microscopy. Results: Pancreatic ductal cells expressed HaRas in the cytoplasm, with Ki-Ras in the apical region and N-Ras (50% of cases) in a supranuclear distribution. Pancreatic acinar cells express all three isoforms with some nuclear expression of Ki-Ras and supranuclear expression of N-Ras. Islets show Ki- and Ha-Ras mainly with differential expression of Ha-Ras ( cells showing less Ha-Ras and more Ki-Ras than cells). Electron microscopy shows that Ha-Ras is mainly localised in the endoplasmic reticulum and Golgi apparatus of the acinar cells with some plasma membrane localisation of Ki-Ras in the ductal cells. There was no change in any of the Ras isoform expression in the ductal or acinar cells in various malignancies studied (Mann-Whitney U test, p 1 0.1).
© 2005 S. Karger AG, Basel and IAP 1424–3903/05/0053–0205$22.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
Conclusions: Ras isoforms have distinct and separate cellular and subcellular distribution that may persist even in the malignantly transformed state. Understanding this distinct functional distribution patterns in detail is an essential step if mutant Ki-Ras is to be targeted in the pancreas by genetic or molecular therapeutic tools. Copyright © 2005 S. Karger AG, Basel and IAP
Introduction
Oncogenic Ras is locked in the ‘on’ form (Ras-GTP) and provides a continuous uncontrolled signal to cell cycle machinery. Nearly 20–30% of all human cancers have oncogenic Ras [1]. Pancreatic cancer is unique, in that it exhibits a high prevalence (nearly 80%) of Ki-Ras mutations [1, 2]. In addition some forms of chronic pancreatitis are also known to harbour similar mutations [3, 4]. It has been proposed that Ki-Ras mutations may be one of the first steps in the multistep molecular pathway towards pancreatic carcinogenesis [5]. There is increasing knowledge that the three main Ras isoforms, namely Ha-Ras, Ki-Ras (4A and 4B) and N-Ras may be functionally different due to their structural differences and spatial distribution [6–8]. This may explain diverse signalling endpoints such as cellular proliferation, migration, and apoptosis on activation of Ras in the same cell. Seufferlein et al. [9] have shown that in 2 pancreatic cancer cell lines (MiaPaCa-2 and Panc-1) possessing oncogenic Ki-Ras
Prof. Bruce M. Hendry Cell Signalling Group, Department of Renal Medicine King’s College London, Bessemer Road London SE5 9PJ (UK) Tel. +44 20 7848 0491, Fax +44 20 7848 0515, E-Mail [email protected]
mutations, TGF-stimulated cell growth is dependent on activation of Ha-Ras [9]. It appears therefore that wildtype Ras proteins retain important cellular functions even in cells expressing oncogenic mutant Ras. The aim of this study was to study the spatial localisation of Ras isoforms in different pancreatic cell types. This would be the initial step in understanding the complex functioning of Ras proteins and interactions between different isoforms in pancreatic pathologies.
Materials and Methods Tissue Selection Normal Pancreas Specimens Light Microscopy. Six of the initial eighteen normal pancreas archival specimens obtained from patients who had a pancreatic resection were used [final diagnosis: benign stricture of bile duct (n = 2), ampullary adenoma (n = 1), cholangiocarcinoma (n = 3)]. Patients (n = 3) had mid-CBD cholangiocarcinoma and the pancreas was unaffected on histology. Twelve of these specimens were excluded because of the presence of dilated duct (n = 5, histologically or radiologically), chronic pancreatitis (n = 5) or equivocal character of the pancreatic tissue (n = 2). Pathological Pancreas Specimens. Archival core biopsy (n = 16) or resection specimens (n = 19) of patients with pancreatic ductal adenocarcinomas, mucinous cystadenocarcinomas, cholangiocarcinoma of the lower end of common bile duct involving the pancreatic duct were used. Normal Pancreas Specimens for Electron Microscopy. Seven fresh pancreas specimens were obtained at resection for suspected cancer with ethical approval and full informed consent. The normal pancreas was obtained from the resection margin of the pancreas, where there was frozen section and further H & E confirmation of normality (absence of tumour or chronic pancreatitis). Four were excluded because of the presence of malignancy (n = 1), chronic pancreatitis (n = 2) or inadequate representative sample (n = 1) at the resection margin on H & E. Three sections from each of these specimens were stained. Reagents and Instruments Light Microscopy Materials. Monoclonal antibodies (mAb) against pan-Ras (clone Ras10), Harvey (clone 235-1.7.1), Kirsten (clone 234-4.2) and Neural (clone F155-277) isoforms of Ras (Oncogene Research Products, USA), polyclonal rabbit anti-glucagon and polyclonal guinea pig anti-insulin (DAKO, USA) were obtained. Secondary antibodies were polymer system labelled with goat anti-mouse secondary antibodies (DAKO Envision™ + system, DAKO USA), goat anti-rabbit IgG (Sigma ImmunoChemicals, USA) and goat anti-guinea pig IgG (Jackson ImmunoResearch Lab. Inc., USA). Dual labelling immunohistochemistry was by Vector VIP substrate kit for peroxidase (Vector Laboratories, USA), in order to give a pink colour. Electron Microscopy Materials. Monoclonal antibodies against Ras isoforms were as above and the secondary antibody was goat anti-mouse IgG conjugated with 10-nm gold particles (British Bio-
206
Pancreatology 2005;5:205–214
cell International, UK). Embedding resin was Lowicryl K4M (Polysciences Ltd, Eppelheim, Germany). 200-mesh nickel grids (Agar Scientific) with formvar coating (polyvinyl formal, BDH, UK) were used. The cutting machine used was Ultrotome III (LKB, Sweden), the cryostat was FS/FAS (Brights, UK) and the electron microscope was JEOL 100C (Japanese Electron Optics Ltd, Japan). Methods Staining for Light Microscopy. 4-m serial sections were used and antibody dilutions were as described and optimised previously [10]. These protocols allowed selective and specific detection of the 3 Ras isoforms [10]. Negative controls (omission of primary antibody) and positive controls with skin were used for each set of experiments. For the purposes of dual staining, consecutive specimens were labelled with only Ras or only hormone antibody in order to identify the type of cell with more certainty. Dual staining with the best colour contrast was Ras staining followed by insulin/glucagon staining. The glucagon staining gave problems due to high background and the Ras staining accentuated this. Hence, all analysis of dual staining was done with Ras and insulin staining alone. Interpretation. Two independent observers who were blind to the type of antibody or the pathology of the specimen used, interpreted the staining pattern as negative (–), marginal (+\–), mild (+), moderate (++) and intense (+++) for various cell types (ductal, acinar, centro-acinar, islet, fibroblasts) for each section and specific subcellular staining pattern, if any, was recorded. These staining patterns were scored 0–4, respectively. Statistical analysis was by Mann-Whitney U test (SPSS 10.1, Chicago, USA) to compare respective cell types between normal and pathological pancreas, with a significance level at p ! 0.05. Electron Microscopy Protocol Fixation and Embedding. Fresh specimens were fixed with 2% formaldehyde with 0.5% glutaraldehyde for 2 h followed by preservation in phosphate buffer. Embedding was with freshly prepared Lowicryl K4M with progressive lowering of temperature [11]. Briefly, tissue was dehydrated at 4 ° C in increasing concentration of methanol and then infiltrated with Lowicryl K4M. The tissue was then embedded in fresh resin in gelatine capsules, immersed in ethanol baths, without air entrapment and photopolymerised in UV light (360 nm, Ultraviolet Prod. Inc., USA) at –35 ° C in a cryostat for 24 h and then at room temperature. Semi-thin (2 m) survey sections were cut and stained with 0.5% toluidine blue. Appropriate areas of the blocks were then cut in ultra-thin sections and put on formvar-coated 200-mesh nickel grids. A section was stained with 1% uranyl acetate (25 min) and lead citrate (0.4%, 10 min) to assess adequate fixation and embedding with Lowicryl. Air-dried ultra-thin sections (silver-gold interface, 70–90 nm) were washed in 1:20 normal goat serum (20 min) and incubated with monoclonal antibodies against Ras: Ha-Ras (1:50), Ki-Ras (1:1), N-Ras (1:10) and secondary antibody (1: 50) and after washes with secondary antibodies conjugated with gold (1:50, 2 h) at room temperature. Counterstaining was done with 1% uranyl acetate. Electron Microscopy and Analysis. Initial scanning at low magnification (!20,000) identified the cells by morphology. Position and counts of the gold particles in the particular subcellular compartment were studied in the nucleus, cytosol and plasma membrane (!35,000). In addition in acinar cell, mitochondria, endo-
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
plasmic reticulum, zymogen granules, Golgi apparatus, were also studied (!60,000). These settings were kept constant to keep the count of gold particles uniform (at per square centimetre: 0.08 m2 for magnification of 35,000! and 0.028 m2 for magnification of 60,000!). Quantitative analysis was performed by counting the number of gold particles per square centimetre; squares were selected randomly using a defined grid. Each photograph had 6 such readings taken for each cellular organelle; where possible, these readings were repeated thrice in other photographs. Statistical analysis was done organelle-wise using paired sample t test for comparing isoforms (SPSS 10.1, USA), with a significance level at p ! 0.01.
Results
Clinical characteristics of patients used throughout the study are summarised in table 1. Immunohistochemistry in Normal Pancreas All negative and positive controls used with each set of experiments on different days were uniformly negative or positive, respectively. Ras protein expression as detected by pan-Ras monoclonal antibody was seen in pancreatic acinar and centroacinar cells, ductal cells, and islet cells as well as stromal fibroblasts, all to a variable degree of intensity. A semi-quantitative summary of the intensity of staining according to the cell type and the isoform is shown in table 2. The expression of the Ras isoforms was cell specific and each isoform had specific subcellular distribution. Pancreatic ductal cells showed intense cytoplasmic staining with pan-Ras (fig. 1A, 3rd column). Pancreatic ductal cells expressed Ha-Ras in the cytoplasm with no
Table 2. Immunohistochemical staining for Ras isoforms in normal pancreas tissue
Fibroblast Acinar cells Centroacinar cells Islets Duct Nerve Mast cells
nuclear expression of the isoform, whilst Ki-Ras was predominantly concentrated on the luminal aspect of the ductal cellular cytoplasm with some (punctate) nuclear staining. In contrast, 50% of ductal cells in the normal pancreas showed a specific supranuclear distribution for N-Ras (fig. 1A, 3rd column). The remaining 50% of cases showed no staining with N-Ras in the ductal cells. The acinar cell complex showed mild staining with pan-Ras. It was predominantly in the cytoplasm of the centroacinar cells with some staining of the acinar cells’ cytoplasm (fig. 1A, 1st column). Specifically with Ki-Ras, there was punctate nuclear staining in 50% of cells in all specimens and some cytoplasmic staining. In addition, there was supranuclear distribution of N-Ras (50% of cases). Ha-Ras stains the acinar complex mildly.
Table 1. Clinical characteristic of patients in the study Patient
Total Male Age: median (range), years
Normal pancreas for light microscopy 6 Normal pancreas for electron microscopy1 3 Pathological pancreas Well differentiated PDA 3 Moderately differentiated PDA 26 Poorly differentiated PDA 2 Cholangiocarcinoma 2 Mucinous cystadenocarcinomas 2
4 2
50 (38–75) 70 (70–76)
3 14 0 2 2
62 (41–74) 61 (56–84) 76 (72–78) 66 (62–70) 56 (48–64)
PDA = Pancreatic ductal adenocarcinoma. None of these patients had previous pancreatitis or ethanol abuse. 1
Pan
Kirsten (Ki)
Harvey (Ha)
Neural (N)
1.1380.4 1.5080.3 1.1780.3 3.1380.3 1.6380.4 1.0280.4 0.5780.6
0.1380.1 0.6380.2 (N) 0.5680.3 1.7580.3 0.6280.2 (N) 0 2.7580.6
0.1480.1 1.0380.4 0 2.2580.21 0.7580.3 0.1780.2 0
0 0.6280.3 (SN) 0 0.8780.3 0.2580.2 (SN) 0 0
The semi-quantitative analysis of Ras isoforms (Ki, Ha, and N) expression in normal human pancreas according to cell type was scored as: negative (–) staining: 0; marginal (+\–): 1; mild (+): 2; moderate (++): 3; and intense (+++): 4. Special characteristics were noted as: SN: supranuclear stain, N: nuclear stain. 1 Differential staining pattern for islet cell subpopulation. All figures are mean 8 standard error of mean.
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
207
A
Pan 1000x
400x
1000x
1000x
400x
1000x
1000x
1000x
1000x
1000x
Ha
Ki 1000x
N
1000x
Acini
Islet
Duct
α
B
β 1000x
Islet: Insulin 208
Pancreatology 2005;5:205–214
1000x
Ha-Ras
1000x
Insulin & Ha-Ras Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
The cells of the islets of Langerhans showed marked cytoplasmic staining with pan-, Ha- and Ki-Ras mAb but mild staining with N-Ras mAb. In addition, there was a differential staining of the and islet cells with Ha-Ras (fig. 1A, 2nd column). With Ha-Ras, a few islets cells, mostly at the periphery with strands branching towards the centre of the islets (believed to be cells in distribution) showed intense cytoplasmic staining. Most of the other islet cells showed mild cytoplasmic staining with Ha-Ras and a very few cells showed no staining. Ki-Ras showed marked cytoplasmic staining. In addition, some punctate nuclear staining could also be discerned. N-Ras showed mild staining with about 75% of the cells and those cells had coarse granular type of cytoplasmic staining with a faint background staining of the cytoplasm. Fibroblasts showed staining with pan- and Ki-Ras mAb in the cytoplasm, but no staining with Ha- or N-Ras mAb. Dual Labelling in Normal Human Pancreatic Islets Both glucagon and insulin monoclonal antibodies worked very well when used alone. The distribution of cells as shown by the glucagon staining were same as that of the cells staining intensely with Ha-Ras. There was background staining with glucagon, which was accentuated with Ras staining and could not be used for analysis for dual staining. However, dual staining with insulin worked very well without background stain, but the colour differences were difficult to discern. Cells have heavier staining with Ki-Ras than cells. Cells (as shown by distribution in the islets, staining with glucagon and not staining with insulin) have intense
brown staining with Ha-Ras. In addition, cells showed some Ki-Ras and minimal staining with N-Ras. In cells (as shown by distribution in the islets, staining with insulin and not staining with glucagon), there was mild staining with Ha-Ras (as compared to cells) (fig. 1B). Ki-Ras stained cells intensely as compared to cells. Pathological Pancreas A semi-quantitative analysis with a non-parametric statistical test (Mann-Whitney U test) was used to compare the data on the respective cells (ductal and acinar) in pancreatic ductal adenocarcinoma, cholangiocarcinoma, and pancreatic mucinous cystadenocarcinomas with those from the normal pancreas. The comparison was made from the pathological area in the pancreas, i.e., malignant ductal cells and distorted acini (acinar cells acted as surrogate controls for cell type) around the malignant area. In some biopsy specimens there were not enough acini present to comment on the staining for them. There was no change in the Ras isoform expression in the malignant ducts as compared to normal ducts, when compared by histology or grade of tumour for any of the Ras isoforms (all isoforms: p 1 0.1) (fig. 2). Ha-Ras showed a trend towards an increase in staining intensity with pancreatic ductal adenocarcinoma, cholangiocarcinoma and mucinous cystadenocarcinomas. There were 2 types of controls for the type of malignancy of interest: pancreatic ductal adenocarcinoma; mucinous cystadenocarcinoma where the Ki-Ras mutation is not ubiquitous (control for Ras mutation) and cholangiocarcinoma of the lower end of the bile duct causing pancreatic ductal obstruction and therefore obstructive chronic pancreatitis (pathological control: no malignancy in the pancreas). There was no change in the expression of Ras isoforms in the acinus in malignant specimens (histology or grade of tumour did not matter) when compared with normal specimens for any of the Ras isoforms (all isoforms: p 1 0.1) (fig. 3).
Fig. 1. A Ras immunohistochemical staining in normal human pancreas. First row is staining with pan-Ras and subsequent 3 rows are with Ha-, Ki- and N-Ras respectively. Left column for staining in the acini, middle column for staining in the islets and right column for staining in the duct. Acinar staining is shown at a higher magnification of the acinar region (1,000!). Red arrows point to the centroacinar cells to differentiate them from the acinar cells. Bent arrows point to the stellate cells in the interstitium. Yellow arrows represent the supranuclear concentration of the N-Ras staining in the acinar cells. The islets of Langerhans: arrows point to the islets and arrowheads point to the cells, whilst bold arrows point to cells. Ducts, the arrowhead points to the mast cell and the yellow arrow to the supranuclear distribution of the N-Ras stain. B Immunohistochemistry and light microscopy with dual staining with insulin and Ha-Ras isoforms, to visualize the islets of Langerhans. Higher resolution (1,000!) microphotographs showing the staining with insulin alone, with Ha-Ras alone and with Ha-Ras and insulin to show the and cells.
Electron Microscopy for Normal Pancreas Ha-Ras showed significantly higher concentration (p ! 0.001, paired t test) than the other isoforms in the mitochondria, Golgi apparatus, zymogen granules and endoplasmic reticulum of the acinar cells; there was a negligible amount of Ras isoforms seen in the nucleus (fig. 4). The centroacinar cells showed more Ras in the cytosol than the nucleus. The plasma membrane of the duct cells showed some Ha-Ras as well as Ki-Ras but no N-Ras (fig. 4). In the cytosol of the duct cells, Ha-Ras was most prominent. The
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
209
3.5
3.0
3.0
2.5
2.5
2.5
2.0 1.5 1.0
2.0 1.5 1.0
1.5 1.0
0.5
0.5
0.0
0.0
0.0
0.5 N
8
31
2
2
N
PDA
MucCA
CCA
0.5 N
8
31
2
2
N
PDA
MucCA
CCA
N
3.5
3.0
3.0
2.5
2.5
2.5
2.0 1.5 1.0
N-Ras duct
3.5
3.0 Ki-Ras duct
3.5
2.0 1.5 1.0 0.0
0.0
0.5
0.5
0.5
3
25
2
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
2
MucCA
CCA
0.5
N
8 N
3
25
2
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
Fig. 2. Semi-quantitative analysis of Ras isoform expression in the ducts of patients with normal pancreas (N) compared with those having pancreatic ductal adenocarcinoma (PDA), mucinous cystadenocarcinoma (MucCA) and cholangiocarcinoma (CCA). Staining in the PDA was further studied as according to the differentiation of the tumour: well, moderately and poorly differentiated. The
method of differentiating cell from and cells is by the way of cytoplasmic granules with varying character. Cells showed more Ha-Ras than the cells, and Ha-Ras was the most prominent isoform expressed in the endocrine cells (fig. 4). Statistical Analysis for Pancreas for Electron Microscopy It will be seen from table 3 that Ha-Ras is expressed in all the cellular organelles, but more so in the acinar Golgi apparatus and the endoplasmic reticulum. Other important observations are that the plasma membranes in the duct cells show both Ha- and Ki-Ras and no N-Ras. The statistical significance and the confidence intervals of the difference in staining with various isoforms are given using the paired t test. Staining in some of the organelles could not be measured in enough sections to give confident estimates and are represented as NA.
Pancreatology 2005;5:205–214
2
PDA
1.0
0.0 8
30
N
1.5
0.5
N
8
2.0
0.5
N
210
2.0
0.5 0.5
Ha-Ras duct
N-Ras duct
3.5
3.0 Ki-Ras duct
Ha-Ras duct
3.5
N
8 N
3
24
2
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
top row amalgamates the readings of all the grades of pancreatic ductal adenocarcinoma to compare it with the other cancers and the normal gland. There is an increase in the staining of Ha-Ras in all the cancers but it does not reach statistical significance (p 1 0.05). The left column is with Ha-Ras, middle with Ki-Ras and right with N-Ras.
Discussion
This set of experiments shows for the first time that there is a definite pattern of Ras isoform distribution in normal pancreatic tissues, which differs not only from cell to cell but also at the subcellular level. Of interest is the fact that the amount of Ras expressed in the pathological pancreas (malignancies) did not change although the mutated Ki-Ras is a well-known early carcinogenic event in pancreatic ductal adenocarcinoma. The functional differences of Ras isoforms have been brought to attention by 2 key biological findings; different Ras isoforms are mutated in various human cancers and the Ras knockout mice models have varied levels of arrest of embryonic development with each isoform [1, 2, 12]. These functional distinctions between Ras isoforms may explain different endpoints of activating the Ras GTPases in the same cell [13, 14].
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
2.5
3.0
3.0
2.0
2.5
2.5
2.0 1.5 1.0 0.5
2.0 1.5 1.0 0.5
0.5 8
19
2
2
N
PDA
MucCA
CCA
N
3.0
3.0
2.5
2.5
2.0 1.5 1.0 0.5 0.0
8
19
2
2
N
PDA
MucCA
CCA
N
N
1
16
1
2
2
Well Mod Poor MucCA CCA PDA PDA PDA
18
2
2
PDA
MucCA
CCA
2.0
2.0 1.5 1.0 0.5
1.5 1.0 0.5 0.0
0.5
0.5 8
8 N
2.5
0.0
0.5
0.5
N
N-Ras acinus
3.5 Ki-Ras acinus
3.5
1.0
0.5
0.5 N
1.5
0.0
0.0
0.0
Ha-Ras acinus
N-Ras acinus
3.5
Ki-Ras acinus
Ha-Ras acinus
3.5
8
N
N
1
16
1
2
2
N
Well Mod Poor MucCA CCA PDA PDA PDA
8
1
15
N
Well PDA
Mod PDA
1
2
Poor MucCA CCA PDA
Fig. 3. Semi-quantitative analysis of Ras expression in the acinus of patients with normal pancreas compared
with those having cancers as in figure 2. No change in the amount of Ras isoforms is seen as compared to the normal pancreas.
Table 3. Differences in staining between isoform in different cell organelles for pancreatic cells
Organelle
Acini nuclear Acini cytosol Acini mitochon Acini zymogen Acini PM Acini ER Acini Golgi Duct PM Duct cytosol Duct nucleus Cac nucleus Cac cytosol Cell Cell
Mean 8 SE (mean)1
p (95% CI)2
Ha-Ras
Ki-Ras
N-Ras
Ha-Ki
Ki-N
Ha-N
0.580.3 0.280.1 12.381.3 4.880.9 NA 2.480.3 13.381.3 1.580.4 380.6 1.380.4 0.580.3 1.881.3 5.780.6 1.880.3
0.580.2 0.180.1 0 0 0 0.380.1 0 0.580.1 0.380.1 0 0.280.2 0.380.2 0.180.1 0.880.3
0.380.2 0.180.1 NA 1.680.5 0 0.780.2 0.380.2 0 0.280.1 0.580.3 0 0 0.380.1 0.380.2
0.363 (–1.2, +0.5) 0.717 (–0.3, +0.4) 0.000 (+8.9, +15.8) 0.000 (+2.7, +6.9) NA 0.000 (+1.4, +2.7) 0.000 (+10.4, +16) 0.039 (+0.06, +1.9) 0.001 (+1.3, +4.0) 0.004 (+0.4, +0.5) 0.465 (–0.8, +1.4) 0.343 (–2.2, +5.2) 0.000 (+4.3, +7.1) 0.035 (+0.09, +2.1)
1.00 (–0.6, +0.6) 0.579 (–0.2, +0.3) NA 0.015 (–2.8, –0.4) NA 0.069 (+0.03, –1.9) 0.111 (–0.8, +0.08) 0.007 (+0.2, +0.8) 0.438 (–0.3, +0.6) 0.082 (–1.1, +0.1) 0.363 (–0.3, +0.6) 0.175 (–0.2, +0.9) 0.191 (–0.6, +0.1) 0.295 (–0.4, +1.2)
0.465 (–1.4, +0.8) 0.430 (–0.2, +0.4) NA 0.010 (+0.6, +4.1) NA 0.000 (+2.4, +5.5) 0.000 (+10.1, +15) 0.004 (+0.6, +2.4) 0.001 (+4.1, +4.7) 0.145 (–0.3, +1.8) 0.203 (–0.4, +1.3) 0.226 (–1.6, +5.2) 0.000 (+4.2, +6.6) 0.001 (+0.8, +2.2)
Mean number of gold particles counted per cm2 (0.08 m2). p values by paired t test (statistically significant values in bold). 95% CI = 95% confidence intervals for the difference in means; NA = not applicable; ER = endoplasmic reticulum; PM = plasma membrane; Cac = centroacinar cell; Mitochon = mitochondria. 1 2
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
211
Ha-Ras, Acinus, x35000
Ha-Ras, Centroacinar, x35000
Nucleus Nucleus
Ha-Ras, Acinus, x60000
Ha-Ras, Duct x35000
Nucleus Lumen
Ha-Ras, Alpha cell x35000
Ha-Ras, Beta cell x35000
Nucleus Fig. 4. Immunogold labeling with Ha-Ras and electron microscopic view of the pancreatic cell types. Top left picture shows an acinar cell and the top right picture a centroacinar cell. There is intense staining particularly in the Golgi apparatus, and mitochondria. The inset shows the staining in the Golgi apparatus at a higher magnification (arrow). There is also some staining in the cytosol of the centroacinar cell. Bar = 0.28 m. Middle left picture shows the typical staining of the Golgi apparatus, and endoplasmic reticulum,
212
Pancreatology 2005;5:205–214
with Ha-Ras at a magnification of 60,000!. Bar = 0.17 m. The middle right picture is of the duct cell along with staining with HaRas at a magnification of 35,000!. The inset shows the staining plasma membrane of the ductal cell. Bar = 0.28 m. Bottom left picture shows the staining of the granules or vacuoles of the cell intensely with Ha-Ras in comparison with those of the cell (bottom right picture). The inset shows the characteristic staining at a higher power. Bar = 0.28 m.
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry
In in vitro experiments, Ras isoforms vary in their ability to activate Raf-1 serine/threonine kinase, the best characterised downstream effector of Ras, which resulted in differences in cellular activities in various cell types [15]. Ki-Ras is more effective as a recruiter and activator of Raf-1 to the plasma membrane than Ha-Ras, whilst Ha-Ras recruits PI 3-kinase efficiently [16]. Since PI 3kinases play an important role in prevention of apoptosis, Ha-Ras is believed to be a survival signalling molecule [16, 17]. In the multistep carcinogenesis model of colon cancer, it has been proposed that in the presence of already high activity of PI 3-kinase, mutated Ki-Ras activates Raf/MAP kinase pathway thus achieving synergistic activity [18]. An example of differential spatial distribution is the fact that in IL-2-deprived murine T cell line TS1, Ha-Ras localises to mitochondria, whereas in IL2-stimulated cells, Ki-Ras localises to mitochondria [19]. Thus in different circumstances such as growth promoting or apoptosis, Ras isoforms could be differentially targeted to mitochondria. Thus Ha-Ras localised to mitochondria in acinar cells as seen in this study may represent a survival signal. The main activity of Ras is at the plasma membrane. One such microdomain in the plasma membrane is formed by caveolae [20] and Ha-Ras-, but not Ki-Rasdependent activation of Raf can be blocked by a dominant negative caveolin [21]. Ras trafficking is dependent on post-translational modification. Ha- and N-Ras are palmitoylated on their C-terminal cysteine residues during their transport through the endoplasmic reticulum (and Golgi apparatus). They are co-localised with caveolin. On the other hand, Ki-Ras has lysine-rich sequence at its C-terminal end, which localises it to functionally distinct microdomains of the plasma membrane [22]. NRas passes via the Golgi apparatus, where it may undergo post-transcriptional changes [22]. Apolloni et al. [7] studied the trafficking of Ras isoforms in BHK cells. The first post-translational modification of farnesylation/geranylgeranylation takes place in the cytosol. The C-terminal AAX motif is removed and methylation takes place in the endoplasmic reticulum. The polybasic domain of the KiRas then sorts the protein away from the endoplasmic reticulum and allows it to migrate to the plasma membrane. In taxol-treated cells (disruption of microtubules), Ki-Ras fails to reach the plasma membrane and is found to align in the multivesicular and univesicular endosomal structures [23]. CAAX proteins with incomplete trafficking signal can fail to reach the plasma membrane. In real-time experiments on PtK2 cells, it has been shown that proteins go from trans-Golgi network to trans-
porting carriers (which migrate with fusion and fission) along the microtubules to reach the plasma membrane [24]. Using a transforming N-Ras in NRK cell line, these experiments suggest the role of N-Ras in the carrier transport role of trans-Golgi network TGN and microtubules. In both the pancreatic acinar and ductal cells N-Ras is seen in the supranuclear position where Golgi apparatus is present. Sharpe et al. [25, 26] have shown that in human renal fibroblasts, there are critical and distinct roles for Ki-Ras and Ha-Ras in cytokine-stimulated proliferation without a functional role for N-Ras. Seufferlein et al. [9] have shown that in 2 pancreatic cancer cell lines (MiaPaCa-2 and Panc-1) possessing oncogenic Ki-Ras mutations, TGF-stimulated cell growth is dependent on activation of Ha-Ras. Oncogenic Ki-Ras constitutively activates the MAP-kinase cascade. They have demonstrated that TGF stimulation increases the membrane translocation and activation of Ha-Ras and these cells use the Ha-Ras-ERK pathway to proliferate. This dependence on Ha-Ras in pancreas cancer is surprising in view of the clear co-expression of oncogenic Ki-Ras in these cells. Thus the Ras isoforms can synergistically promote growth and metastasis of pancreatic cancer cells. These in vitro experiments on renal fibroblasts and pancreatic cancer cells strongly support the concept of distinct functional roles for the Ha-Ras and Ki-Ras isoforms and support our findings of different roles for Ras as suggested by the present in vivo data. Interesting observation was made of the differential staining of sub-populations of islet cells, especially with Ha-Ras. This was confirmed by the technique of dual labelling and also immunogold labelling in electron microscopy. Cells producing glucagon, exhibit more Ha-Ras than the cells. Ras may have a role in the insulin-mediated negative regulation of glucagon gene transcription [27]. Gherzi et al. [27] stimulated PKA- and PKC-induced gene transcription using insulinoma (-TC1) and glucagonoma (In-R1-G9) cell lines. Using effector constructs of Ha-Ras, they showed that the PKA and the Ras signalling pathways are functionally antagonistic in these cell lines. Thus Ras has an important role in gene transcription in these human endocrine cells. Ki-Ras was seen predominantly in cells on electron microscopy with NRas present in both types of cells in minimal quantities. The Ras isoforms can be differentially targeted achieving desired and focused therapeutic endpoints. There are numerous therapeutic strategies to target Ras. The most actively used molecules are those inhibiting the posttranslational modification of Ras proteins called prenylation inhibitors such as farnesyl transferase inhibitors
Expression of Ras GTPase Isoforms
Pancreatology 2005;5:205–214
213
and geranylgeranyl transferase inhibitor [28]. Farnesyl transferase inhibitor is being used in phase II clinical trials in patients with unresectable pancreatic cancer with or without gemcitabine. Ha-Ras antisense oligonucleotides are being used in phase II clinical trials in patients with pancreatic cancer. Thus Ras targetting, as shown by this body of work, holds an immense potential in treatment of human diseases, provided specific agents could be delivered safely.
Acknowledgements We are most grateful to Kristain Loebner who provided some of the reagents for dual labelling, Mohammed Sohail for digital capture of the microscopic slides and some analysis, Anne Clarke (Oxford) for initial help in developing the electron microscopy work. The Agostino Trappani Foundation Naples supported this work.
References 1 Bos JL: The ras gene family and human carcinogenesis. Mutat Res 1988;195:255–271. 2 Barbacid M: Ras genes. Annu Rev Biochem 1987;56:779–827. 3 Yanagisawa A, Ohtake K, Ohashi K, et al: Frequent c-Ki-ras oncogene activation in mucous cell hyperplasias of pancreas suffering from chronic inflammation. Cancer Res 1993; 53: 953–956. 4 Lemoine NR, Jain S, Hughes CM, et al: Ki-ras oncogene activation in preinvasive pancreatic cancer. Gastroenterology 1992;102:230–236. 5 Hruban RH, Wilentz RE, Kern SE: Genetic progression in the pancreatic ducts. Am J Pathol 2000;156:1821–1825. 6 Mochizuki N, Yamashita S, Kurokawa K, et al: Spatio-temporal images of growth-factor-induced activation of Ras and Rap1. Nature 2001;411:1065–1068. 7 Apolloni A, Prior IA, Lindsay M, et al: H-ras but not K-ras traffics to the plasma membrane through the exocytic pathway. Mol Cell Biol 2000;20:2475–2487. 8 Roy S, Wyse B, Hancock JF: H-Ras signaling and K-Ras signaling are differentially dependent on endocytosis. Mol Cell Biol 2002; 22: 5128–5140. 9 Seufferlein T, Van Lint J, Liptay S, et al: Transforming growth factor alpha activates Ha-Ras in human pancreatic cancer cells with Ki-ras mutations. Gastroenterology 1999; 116: 1441– 1452. 10 Kocher HM, Codd J, Sharpe CC, Dockerell MEC, Al-Nawab M, Hendry BM: Expression of Ras GTPases in normal kidney and glomerulonephritis. Nephrol Dial Transplant 2003; 14:848–854.
214
11 Armbruster BL, Carlemalm E, Chiovetti R, et al: Specimen preparation for electron microscopy using low temperature embedding resins. J Microsc 1982;126(pt 1):77–85. 12 Johnson L, Greenbaum D, Cichowski K, et al: K-ras is an essential gene in the mouse with partial functional overlap with N-ras. [erratum appears in Genes Dev 1997; 11: 3277]. Genes Dev 1997;11(19):2468–2481. 13 Hall A, Cales C, Hancock JF, et al: Analysis of mammalian ras effector function. Cold Spring Harb Symp Quant Biol 1988; 53(pt 2):855– 862. 14 Reuther GW, Der CJ: The Ras branch of small GTPases: Ras family members don’t fall far from the tree. Curr Opin Cell Biol 2000; 12: 157–165. 15 Voice JK, Klemke RL, Le A, Jackson JH: Four human ras homologs differ in their abilities to activate Raf-1, induce transformation, and stimulate cell motility. J Biol Chem 1999;274: 17164–17170. 16 Yan J, Roy S, Apolloni A, Lane A, Hancock JF: Ras isoforms vary in their ability to activate Raf-1 and phosphoinositide 3-kinase. J Biol Chem 1998;273:24052–24056. 17 Downward J: Ras signalling and apoptosis. Curr Opin Genet Dev 1998;8:49–54. 18 Phillips WA, St Clair F, Munday AD, Thomas RJ, Mitchell CA: Increased levels of phosphatidylinositol 3-kinase activity in colorectal tumors. Cancer 1998;83:41–47. 19 Rebollo A, Perez-Sala D, Martinez A: Bcl-2 differentially targets K-, N-, and H-Ras to mitochondria in IL-2 supplemented or deprived cells: Implications in prevention of apoptosis. Oncogene 1999;18:4930–4939.
Pancreatology 2005;5:205–214
20 Kurzchalia TV, Parton RG: Membrane microdomains and caveolae. Curr Opin Cell Biol 1999;11:424–431. 21 Roy S, Luetterforst R, Harding AE, et al: Dominant-negative caveolin inhibits H-Ras function by disrupting cholesterol-rich plasma membrane domains. Nat Cell Biol 1999;1:98– 105. 22 Choy E, Chiu VK, Silletti JD, et al: Endomembrane trafficking of ras: The CAAX motif targets proteins to the ER and Golgi. Cell 1999; 98:69–80. 23 Thissen JA, Gross JM, Subramanian K, et al: Prenylation-dependent association of Ki-Ras with microtubules. Evidence for a role in subcellular trafficking. J Biol Chem 1997; 272: 30362–30370. 24 Toomre D, Keller P, White J, et al: Dual-color visualization of trans-Golgi network to plasma membrane traffic along microtubules in living cells. J Cell Sci 1999;112(pt 1):21–33. 25 Sharpe CC, Dockrell ME, Scott R, et al: Evidence of a role for Ki-Ras in the stimulated proliferation of renal fibroblasts. J Am Soc Nephrol 1999;10:1186–1192. 26 Sharpe CC, Dockrell ME, Noor MI, et al: Role of Ras isoforms in the stimulated proliferation of human renal fibroblasts in primary culture. J Am Soc Nephrol 2000;11:1600–1606. 27 Gherzi R, Briata P, Fehmann HC, Goke B: Ras antagonizes cAMP stimulated glucagon gene transcription in pancreatic islet cell lines. FEBS Lett 1994;353:277–280. 28 Khwaja A, O’Connolly J, Hendry BM: Prenylation inhibitors in renal disease. Lancet 2000; 355:741–744.
Kocher/Senkus/Moorhead/Al-Nawab/ Patel/Benjamin/Hendry