2 out of 5 type 1 diabetes pancreases and one pancreatic transplant had VP1+ Islets Dotta et al 2007
2
3
More recent evidence for enteroviral infection in Type 1 diabetes • Epidemiological studies demonstrate that EV infection associated with: 1. Development of autoantibodies in at risk individuals 2. Accelerated progression from autoantibody positivity to clinical diabetes • Type 1 diabetes is associated with enterovirus infection of the gut mucosa (Oikarinen, Diabetes, 2012) • A recent meta-analysis showed a significant association between enterovirus infection and type 1 diabetes related autoimmunity and clinical type 1 diabetes (Yeung, BMJ 2011)
Insulitis and beta-cell loss in human type 1 diabetes displays a lobular pattern
Strengths and limitations of UK cohort • The largest single collection in the world. • Many cases have residual insulin-containing islet & insulitis
• • • • • •
Patients died between 30-50y ago All collected from the UK Varied fixation times and techniques Sometimes only one block available to study Limited patient data Limited variety of other tissues available
8
A new initiative in the USA network of Pancreatic Organ Donors (nPOD) Benefits include • A variety of samples from cadaveric organ donors; including some nondiabetic individuals deemed to be at risk of developing T1D. • Recovery and fixation are uniform between cases. • Information is shared between all nPOD investigators via an online informatics system.
BUT - THERE ARE VERY FEW RECENT-ONSET CASES
9
The Cohorts UK Cohort
JDRF nPOD cohort
72
17
Mean Age
12.65±1.1y
25.7±2.9y
Age Range
1-42y
4-50y
Mean Time Since Diagnosis
8.2±4.1mths
11.9±2.3y
Range Time Since Diagnosis
0-6y
1-35y
Scotland, England and Wales
USA
1959-1983
2007 onwards
Autopsy and organ donors
Organ donors
Variable fixation types and times
Strictly controlled and uniform
Number of cases
Geography Sample Collection Sample type Sample Processing
10
nPOD Cohort
UK Cohort
Enteroviral VP1 (5D8/1) is expressed in the islets of type 1 diabetes patients from both cohorts
11
Controls
Non-diabetic control (UK)
Non-diabetic control (nPOD)
12
Enteroviral VP1 (5D8/1) is expressed in the insulin-containing islets of type 1 diabetes patients in both cohorts Percentage of cases with islet VP1+ cells
90 80 70
Control Type 1 diabetes (ICIs) Type 1 diabetes (IDIs)
60 50 40 30 20
10 0
UK
nPOD
13
nPOD Cohort
Enteroviral VP1 expression co-localises with insulin in both cohorts
UK Cohort
Insulin
Enteroviral VP1
Merge
14
Specificity of the Dako VP1 (5D8/1) antibody Hansson et al, J Pathology 2013 • Suggested that 5D8/1 can bind two mitochondrial proteins, creatine kinase B (CKB) and ATP synthase beta (ATP5B) under denaturing conditions
15
Detection of proteins by Western blotting in HepG2 cells VP1 (5D8/1) Mw (kDa)
CKB
CVB3 C
CVB3 C
0h 2h 4h 6h 8h
0h 2h 4h 6h 8h
ATP5B Mw (kDa)
Mw (kDa)
170 130
170 130
170 130
100
100
100
70
70
70
55
55
55
40
40
40
35
35
35
25
25
25
15
15
15
Actin
Actin
CVB3 C
0h 2h 4h 6h 8h
Actin
16
Peptide ELISA confirms 5D8/1 can bind CKB but not ATP5B 1.00
Absorbance at 405nm
0.80
0.60 pATP5B pCKB
0.40
pVP1
0.20
0.00 1
10
100 5D8/1 [ng/ml]
1000
10000
17
Dako 5D8/1 does not label paraffin-embedded, mitochondrial rich tissues or 78 of 80 human cells and tissues on a microarray. Skeletal muscle (2)
Normal gastric body (2) Oes squamous carcinoma (2) Pancreatic adenocarcinoma (2) Lung NE Ca (2)
Gastric antrum (1)
Normal breast (2)
Normal prostatic acini (1) Prostatic adenocarcinoma (2) Normal ovary surface (2) Ov Pap ser adenoc (2) Normal duodenum villi (2) Normal Bruner Gland (2) Normal thyroid (2)
Normal cartilage (2) Peripheral nerve tumour (2) Squamous cell carcinoma skin (2) Normal spleen (2) Melanoma (2)
sh Chondrosarcoma (1)
Normal pleura (2) Mesothelioma (1) Lung SCC (1) Fibroid (2) Breast Lob Ca (2)
A. Liver
B. Kidney
Normal epidermis (2) Normal tonsil (1) Transitional cell carcinoma (2) Normal oes sq ep (2) Colonic adenocarcinoma (2) Normal omentum (2) Lung adenocarcinoma (2) Normal alveoli (1)
Merkel cell tumour (2) Normal urothelium (2)
Gastric adenocarcinoma (2) Normal nipple (2) Breast ductal Ca (2)
Normal salivary acini (1) Ov muc adenoca (2)
Med Ca breast (2)
Phylloids (2)
Ectocervix (2)
Endocervix (2)
Cx adenoca (1) Fallopian tube (2)
Endomet adenoca (2) Endometrium (2)
Appendix carcinoid (1) Normal appendix (1)
Normal SI muscoa (2)
Normal colon (2)
POC (2)
Normal pancreatic acini (2) Normal parathyroid (1)
Thyroid Hurtle cell (2)
Med Ca thyr (2)
Normal thymus (1)
Hepatocellular Ca (1)
Normal renal glomeruli (2) Normal testis (2)
Normal renal tubule (2)
Clear cell carcinoma (2) Liver parenchyma (2)
Liver portal tract (2)
Adrenal cortex (2)
Adrenal medulla (2)
Seminoma (2)
18
Detection of ATP5B and VP1 in human pancreas sections
A. Insulin
B. ATP5B
C. VP1
VP1
ATP5B
19
Detection of VP1 and CKB in PANC1 cells
CKB
VP1
Mock-infected PANC1
CKB CVB3-infected PANC1
VP1
20
Dako VP1 (5D8/1) Specificity • Under denaturing conditions, we have confirmed that 5D8/1 can recognise CKB but not ATP5B • Under carefully optimised non-denaturing IHC conditions, 5D8/1 does not recognise CKB or ATP5B • 5D8/1 remains an extremely useful tool for detecting enterovirus infections in FFPE tissue
The presence of virus needs to be confirmed using other methods
The nPOD-V Group • The nPOD-V group began activities last year recognising that demonstrating a pathogenic role for one or more viruses in T1D could have very important therapeutic implications
• This effort has culminated in a grant proposal submitted to JDRF in April 2012 and awarded in August 2012
The nPOD-V Group • Aim is to develop a pipeline for comprehensive and integrated understanding of the role of enteroviruses in disease pathogenesis • The sharing of tissues and coordinated analysis by multiple investigators affords a key unifying element in science and the rare opportunity to coordinate studies that take into account multiple approaches and design input from multiple investigators
Coordinated analysis of viral antigens in 30 nPOD samples Section Number: 1-2; UK 3 IHC-Viral protein; UK 4 IHC-Viral protein; Finland 5 ISH-Viral genome; Finland 6 Insulin/ Glucagon; UK 7 IHC Class I MHC; UK 8 IHC PKR; UK 9-10; Finland
Key Info: 1. 96% concordance with viral protein IHC staining between laboratories 2. 69% concordance of IHC and ISH (20/29).
24
Dako VP1 Concordance Section 3
Section 4
UK PH 6088-03
Finland
nPOD 6088-03 Pan Head Images from nPOD Online Pathology website Class I MHC
Based on VP1/ ISH and MHC hyper expression we predicted 26 of the 30 case classifications correctly – this included all of the T1Ds with residual ICIs and 6 of the 7 controls
Summary of Concordance Study • IHC correctly identified all the 12 T1D or AAb+ cases with residual ICIs - 6 of these were also ISH+ - All had class I MHC hyper expression • ISH positivity was identified in 8 T1D, 1 AAb+ and 1 control • MHC hyper-expression was seen only in T1Ds or AAb+ cases • ONLY 92 VP1+ cells were found among a total of almost 3000 islets examined across 29 cases → IT WILL BE HARD TO DETECT & IDENTIFY SPECIFIC SEROTYPE (Proteomics/ RNA analysis / virus isolation) • 1 exception; Autoantibody+ case (ISH+ & MHC h/e)
Pathogen Recognition Receptors Class I MHC
VP1
IFN alpha
PKR
Hyper-expression of Class I MHC CD45
Cytokines/ Chemokines Recruitment of immune cells
Autoimmunity Insulin
Destruction of beta cells
Viral footprint?
29
UK Cohort
nPOD Cohort
Enteroviral VP1 co-localises with Protein Kinase R (PKR) in both cohorts
Enteroviral VP1
Protein kinase R
Merge
30
Summary • Small numbers of enteroviral VP1+ cells are observed in the ICIs of type 1 diabetes cases in both the nPOD collection and in a separate (older) UK cohort • The presence of VP1 was dependent on the presence of beta cells, regardless of duration of disease
• Enterovirus presence was confirmed by enterovirus specific in situ hybridisation • In both cohorts, enteroviral VP1 expression correlated with increased expression of the pathogen-recognition receptor, PKR and hyper expression of class I MHC
31
Conclusions These results imply that enteroviral infection occurs commonly in type 1 diabetes and that an anti-viral response is mounted in infected islet cells This anti-viral response has the potential to sensitise beta cells to immune-mediated destruction
mda5
Viral detection?
Normal ICI
IFNα
VP1
Insulin
Defence initiation
Acute viral infection PKR
Reactivation of virus
Death of β cell Persistent silent infection with no active virus produced
MHC I Insulin
Destruction of β cells CD45
Insulitis
Reduced Mcl1 expression THEREFORE Increased sensitivity to apoptosis
Insulin
IDI
Induction of autoimmunity?
↑ Adhesion molecules/ chemokines
Release and uptake of β cell antigens by APCs
University of Exeter (Plymouth) Noel Morgan Shalinee Dhayal Pia Leete David Hilton Glasgow Alan Foulis Maura Farquharson Andrew Hamilton University of Brighton Adrian Bone nPOD V members
JDRF nPOD Martha Campbell-Thompson Alberto Pugliese Mark Atkinson University of Tampere Heikki Hyöty Sami Oikarinen Maarit Oikarinen Jutta Laiho Uppsala University Gun Frisk Therese Rosenling Karolinska Instituet Malin Flodstrom-Tullberg Katharina Lind Emma Svedin
34
Protein Kinase R – is it active? • Promotes anti-viral responses by inhibiting protein synthesis P
• Levels of rapidly turned over proteins are quickly lost from the cell
PKR
PKR
P eIF2α
• One such protein is the antiapoptotic Bcl2 family member – Mcl-1
protein synthesis
P
PKR Insulin eIF2α
P
• Mcl-1 is a rapidly turned over protein that functions to promote cell survival
Protein synthesis
• Mcl-1 is expressed in β-cells
Mcl-1 Mcl-1
Death cascade
• Mcl-1 knockdown increases stress induced β-cell apoptosis Cell Death
Overlay
Myeloid Leukemia Sequence 1 (Mcl-1)
Mcl-1 is reduced in VP1+ β-cells
insulin
VP-1
Mcl-1
overlay
Conclusion PKR
eIF2α
P
P
Protein synthesis
Mcl-1
Increased sensitivity to
Cell Death
The data suggest that the PKR which is induced during enteroviral infection is activated.
Additionally, the decrease in Mcl-1 levels in infected β-cells may alter their sensitivity to pro-apoptotic stimuli released from infiltrating immune cells
39
Immunodetection of enteroviral capsid protein, VP-1 in the heart and pancreas of a neonate with Coxsackie B viral pancreatitis (Dako clone 5D8/1)
Heart
Pancreas
40
but not in controls
Heart
Pancreas
Concordance Study Identifier
Section 3 Dako VP1 UK
Section 4 Dako VP1 Finland
Section 5 ISH Finland
Section 7 Class I MHC
No. of indicators Positive
Reveal
PB 6070-02
4 islets (11 cells)
pos
neg
Hyper expression
3
T1D
PT 6070-02
5 islets (11 cells)
pos
pos
Hyper expression
4
T1D
PB 6073-04
neg
neg
neg
Normal (exocrine +)
0
Control
PB 6081-02
1 islet (1 cell)
pos
pos
h/e (exocrine +)
4
T1D
PB 6084-01
7 islets (26 cells)
pos
neg
Hyper expression
3
T1D
PH 6088-03
1 islet (8 cells)
pos
pos
Hyper expression
4
T1D
PB 6088-08
3 islets (4 cells)
pos
pos
Hyper expression
4
T1D
PB 6095-04
neg
neg
neg
Normal (exocrine +)
0
Control
PB 6096-01
neg
neg
neg
Normal
0
Control
Poth 6121-04
6 islets (8 cells)
pos
pos
Hyper expression
4
T1D
Figure :Progression of Insulitis at various stages of T1D 18
Mean No. of cells per islet
16
14 12 10
CD4 CD8
8
CD20 CD68
6 4 2 0 Ins ++ve
Ins +ve Insulin Level
Ins -ve
43
Diabetes Incidence / Growth Predictions In 2011 there were an estimated 2.9 million patients with diabetes in the UK This is expected to rise to 5
million people by 2025.
Diabetes UK estimates that
Up to 15 per cent of these patients may have Type 1 Diabetes
The prediction is that new cases of type 1 diabetes in children younger than 5 years will double by 2020. (Patterson et al, 2009)