Age dependent type 1

diabetes pathogenesis

Ake Lernmark
From
Joerg Ermann &
C. Garrison Fathman
Nature Immunology
2, 759 - 761 (2001)
Insulitis
 Residual -cell function in new
onset Type 1 diabetes
Age C-peptide level within normal range
(years) At diagnosis One year Two years

0-15 20% 10% <5%

15-34 60% 55% 46%

rhw 5/99
 Type 1 diabetes genes
v HLA DQ2, DQ8, or both, represents almost 90% of all
type 1 diabetes patients younger than 20 years of age.
v The risk of DQ2/8 heterozygotes decreases with
increasing age.
v The protection of DQ6.2 is attenuated by increasing age
and is lost at about 35 years of age.
v Class I - INS VNTR -short tandem repeats - increase the
risk by about 2-5 %.

v CTLA-4 - long AT-repeats at the 5’ end UTR - increase

the risk by about 2-3%.
ENVIRONMENTAL FACTORS
* MONOZYGOTIC TWINS 20-30% CONCORDANCE.
** ONLY 10-15% OF NEW PATIENTS HAVE A
PARENT OR SIBLING WITH THE DISEASE.

VIRUS: MUMPS,COXSACKIE, RUBELLA, ECHO,

ROTA, LJUNGAN AND OTHERS.

FOOD ITEMS: NITROSAMINES, MILK PROTEINS

GESTATIONAL INFECTIONS AND ABO

INCOMPATIBILITY
 VIRUS AND TYPE 1 DIABETES
Coxsackie Human, mice (Yoon)
Rubella Human, hamster
Mumps Human
Cytomegalovirus Human
Rotavirus Human

CBV4 T cell activation (Vbeta analysis): T1DM=controls

(Varela-Calvino et al. 2001)
CBV4 T cell proliferation: T1DM > controls (Juhela et al 2000)
CBV4-specific T-cell epitopes: T1DM = controls
(Martilla et al. 2001)
No or little cross reactivity between molecular mimicry regions
(Several authors)
 ABO and hyperbilirubinemia
Autoantibody Controls ABO immunization Hyperbilirubinemia

IA-2Ab 1,4% (4/288) 6,6% (10/151)* 1,6% (5/311)

GAD65Ab 1,6% (5/320) 2,6% (4/151) 1,3% (4/311)

ICA 0,6% (2/320) 4,0% (6/151)** 4,2% (13/311)***

Difference compared to controls: * p=0.007; ** p=0.015; ***

p=0.003.

All samples are cord blood.

 Diagnostic sensitivity and specificity
of autoantibodies for Type 1 diabetes
Autoantigen Sensitivity Specificity

Insulin 40-70% 99%

GAD65 70-80% 99%

IA-2 50-70% 99%

rhw12/98
GENETIC EFFECTS ON AGE-DEPENDENT ONSET AND
ISLET CELL ANTIBODIES IN TYPE 1 DIABETES.

R PATIENTS: INCIDENT, 0-34 YEARS OF AGE

n=971
R CONTROLS: MATCHED FOR AGE AND GENDER
n=702

R HLA, INS VNTR AND CTLA-4

R GAD65A, IA-2A, IAA AND ICA

LOGISTIC REGRESSION TO MODEL THE NATURAL

R LOGARITHM OF THE ODDS
 A B
0 DQ8, Females 0 DQ8, Males

0.8

0.6

0.4 0 DQ2

0.2
0 DQ2

0 10 20 30 0 10 20 30
Age at clinical onset (years)
 A B
0 DQ8, Females 0 DQ8, Males

0.8
1 DQ2
0.6

0.4 0 DQ2
1 DQ2

0.2
0 DQ2

0 10 20 30 0 10 20 30
Age at clinical onset (years)
 A B
0 DQ8, Females 0 DQ8, Males
2 DQ2
0.8
2 DQ2
1 DQ2
0.6

0.4 0 DQ2
1 DQ2

0.2
0 DQ2

0 10 20 30 0 10 20 30
Age at clinical onset (years)
RISK FOR TYPE 1 DIABETES AS FUNCTION
OF AGE, GENDER, HLA AND GAD65A.
THE ODDS TO DEVELOP TYPE 1 DIABETES:

A FEMALE WITH GAD65Ab HAS

3 TIMES THE RISK OF A MALE.

COMPARED TO A FIVE YEAR OLD WITH

GAD65Ab BUT NO DQ2:
3 TIMES HIGHER RISK WITH ONE DQ2
8 TIMES HIGHER RISK WITH TWO DQ2

DQ2 DOES NOT AFFECT THE RISK FOR

A GAD65AB POSITIVE 34 YEAR OLD
 A B
0 DQ8, Females 0 DQ8, Males
0.8 0 DQ2 0 DQ2

0.6

0.4

0.2

0.0
0 10 20 30 0 10 20 30
Age at clinical onset (years)
 A B
0 DQ8, Females 0 DQ8, Males
0.8 0 DQ2 0 DQ2

1 1
0.6

0.4

0.2

0.0
0 10 20 30 0 10 20 30
Age at clinical onset (years)
 A B
0 DQ8, Females 0 DQ8, Males
0.8 0 DQ2 0 DQ2

1 1
0.6
2 2
0.4

0.2

0.0
0 10 20 30 0 10 20 30
Age at clinical onset (years)
RISK FOR TYPE 1 DIABETES AS FUNCTION
OF AGE, GENDER, HLA, AND IA-2Ab.
THE ODDS FOR TYPE 1 DIABETES WITH IA-2Ab
AT 5 YEARS OF AGE IS 11 TIMES THAT AT
34 YEARS OF AGE.

THE ODDS FOR EACH ADDITIONAL DQ8:

1.5 TIMES FOR ONE DQ8
3.0 TIMES FOR TWO DQ8

THE ODDS OF EACH ADDITIONAL DQ2:

DECREASES
0.27 TIMES FOR ONE DQ2
0.6 TIMES FOR TWO DQ2
Insulin autoantibodies are associated with a
combination of HLA-DQ8 and INS VNTR.

Click for larger picture

RISK FOR TYPE 1 DIABETES AS FUNCTION
OF AGE, GENDER, HLA, INS VNTR AND IAA.
THE ODDS FOR TYPE 1 DIABETES WITH IAA
AT 5 YEARS OF AGE IS 10 TIMES THAT AT
34 YEARS OF AGE.

THE ODDS FOR EACH ADDITIONAL DQ8:

1.4 TIMES FOR ONE DQ8
2.1 TIMES FOR TWO DQ8

THE ODDS OF EACH ADDITIONAL INS VNTR

CLASS I ALLELE:
1.5 TIMES FOR ONE CLASS I
2.2 TIMES FOR TWO CLASS I
 SUMMARY, SO FAR……...
* MULTIPLE ENVIRONMENTAL FACTORS.
*GESTATIONAL EFFECTS.

* HLA HAS THE MAJOR GENETIC EFFECT

- INS VNTR AND OTHER GENETIC FACTORS
CONTRIBUTE.

* AGE-DEPENDENT EFFECTS OF HLA AND ON

GAD65Ab
IAA - INS VNTR CONTRIBUTES
IA-2Ab

* USEFUL INFORMATION FOR PREDICTION?

 100

80

COMBINATIONS Three autoantibodies

OF ISLET CELL 60

AUTO- Two autoantibodies

ANTIBODIES 40

PREDICT
20

TYPE 1 DIABETES One autoantibody

0
0 2 4 6 8 10
Years of follow-up

Click for larger picture

 WHAT ABOUT CHILDREN
AND TEENAGERS?
* WASHINGTON PREDICTION STUDY:
> 4 500 14 year olds were screened
Follow up 9 years.
All 15 children developing diabetes were
predicted. No false negatives.
No false positives.
Hagopian et al. Diabetes Care 2002
* SCREENING NEWBORNS: HLA and antibodies
DIPP (Finland), TRIGR (international),
DAISY (Denver, CO), PANDA (Gainesville, FL),
ABIS (South East Sweden),
DiPiS (South Sweden)
MELBOURNE NEWBORN STUDY
TYPE 1 DIABETES IS A T-CELL MEDIATED DISEASE
* Poor antigen quality has hampered novel technologies to detect
T-cells reactive with GAD65, proinsulin (PI), and IA-2.

* The second T cell IDS workshop reported GAD65 (Diamyd Medical)

generated in insect cells that stimulate relevant clones and
does not inhibit third-party antigens.

* A PI preparation generated in bacteria was free of effects on

proliferation to third-party antigens and low in endotoxin.

* These preparations should be useful to develop robust and sensitive

assays of autoantigen-specific T cells that predict diseases.

* Peakman et al. Report of phase II of the Second International

Immunology of Diabetes Society Workshop for Standardization
of T-cell assays. Diabetes 50:1749-54, 2001.
 GAD65Ab modulate GAD65 antigen
presentation.
• T-cell hybridomas
• DRB1*0401
restricted • GAD65Ab positive
• GAD65 peptide sera from new onset
274-286 dependent children at various
• APC from end-point titers
DRB1*0401 subjects
• IL-2 release response

Reijonen et al Diabetes 2001

 GAD65Ab ENHANCE ANTIGEN
PRESENTATION
622 62 2

673 67 3

591 59 1

686 68 6

652 65 2

708 70 8

613 61 3
826 82 6
898 89 8
853 85 3
1306 1 30 6

0 .0 2 .5 5 .0 7 .5 10 .0 1 2 .5
0,0 2,5IL -2 con5,0 7,5 10,0
cen tra tio n (U /m l)
12,5 0,0
0.00
0,25
0.2 5
0,5
0 .5 0
0,75
0 .7 5

G A D 6 5 a n tib o d y in d e x
1,0
1 .00
1,25
1.2 5

IL-2 concentration(U/ml) GAD65 antibody index

ANTIBODY-MEDIATED POTENTIATION OF
ANTIGEN-PRESENTATION.

• GAD65Ab mediated • Preservation of

potentiation of conformation
antigen presentation dependent GAD65Ab
may explain: after diagnosis when
• Preservation of beta cells are gone.
conformation • Acceleration of beta
dependent GAD65Ab cell destruction by
before diagnosis. recruiting new CD4
and CD8 T cells.
 SUMMARY AND CONCLUSIONS
* HLA HAS THE MAJOR EFFECT - OTHER GENETIC
FACTORS SUCH AS INS VNTR AND CTLA-4
CONTRIBUTE.
* MULTIPLE ENVIRONMENTAL FACTORS.
*GESTATIONAL EFFECTS.
*EARLY T CELL RESPONSES ARE KEY TO INITIATION
OF BETA CELL AUTOIMMUNITY.
* AGE-DEPENDENT EFFECTS OF HLA AND ON
GAD65Ab
IAA INS VNTR CONTRIBUTE
IA-2Ab.
* CHRONIC BETA-CELL AUTOIMMUNITY MAY BE
MAINTAINED BY AUTOANTIBODY-FACILITATED
T CELL RESPONSES.
 Acknowledgement
• JINKO GRAHAM
• CHRISTIANE HAMPE • NORMAN BRESLOW
• LUO DONG
• TERRI DANIELS • HELENA REIJONEN
• GERALD T NEPOM
• LISA HAMMERLE
• SWEDISH DIABETES
• STEN-A. IVARSSON REGISTRIES FOR
• CORRADO CILIO CHILDREN AND ADULTS