Poster Presentation The 13th International Congress of the Immunology of Diabetes Society 2013

Doubly reactive autoantibodies define INS-IGF2 as an autoantigen largely in HLA-DQ2/8 patients newly diagnosed with autoimmune (type 1) diabetes. (#164)

Norio Kanatsuna 1 2 , Ahmed Delli 1 , Cecilia Andersson 1 , Anna-Lena Nilsson 1 3 , Fariba Vaziri-Sani 1 , Karin Larsson 4 , Anneli Carlsson 1 , Elisabeth Cedervall 5 , Björn Jönsson 6 , Jan Neiderud 7 , Helena Elding Larsson 1 , Sten-Anders Ivarsson 1 , Carina Törn 1 , Malin Fex 1 , Donald F. Steiner 8 , Åke Lernmark 1
  1. Department of Clinical Sciences, Lund University, Malmö, Sweden
  2. First Department of Internal Medicine, Osaka Medical College, Takatsuki, Japan
  3. Department of Pediatrics, Östersund Hospital, Östersund, Sweden
  4. Department of Pediatrics, Kristianstad Hospital, Kristianstad, Sweden
  5. Department of Pediatrics, Ängelholm Hospital, Ängelholm, Sweden
  6. Department of Pediatrics, Ystad Hospital, Ystad, Sweden
  7. Department of Pediatrics, Helsingborg Hospital, Helsingborg, Sweden
  8. Department of Medicine, University of Chicago, Chicago, IL, USA


INS-IGF2, the splice variant of preproinsulin and the untranslated ORF of the IGF2 gene, was recently reported to be an autoantigen in autoimmune (type 1) diabetes. We hypothesized that autoantibodies against human islet INS-IGF2 that contains the preproinsulin signal peptide, the B chain and 8 amino acids of the c-peptide in addition to 138 unique amino acids are associated with autoimmune (type 1) diabetes. The aim was to determine whether levels of specific INS-IGF2 autoantibodies (INS-IGF2A), specifically blocked by both cold insulin and INS-IGF2 (doubly reactive sera), were related to islet autoantibodies, HLA-DQ, or both, in newly diagnosed patients (n=676) and controls (n=363).


Patients, 0-18 years of age, diagnosed with type 1 diabetes in 1996-2005 were analyzed for specific doubly reactive INS-IGF2 autoantibodies (INS-IGF2A) after displacement with both insulin and INS-IGF2 to correct for non-specific binding. Islet autoantibodies (GADA, IA-2A, IAA, ICA, ZnT8RA, ZnT8WA, and ZnT8QA) and HLA-DQ genotypes were also determined.


The median levels of INS-IGF2A were higher in the patients compared to controls (p<0.001). Irrespective of age at diagnosis, 68% (458/676) patients had INS-IGF2A using the 75th percentile of the controls as a cut off at 586 U/ml (p<0.001). Similarly, 19 % (126/676) of the patients had INS-IGF2A when the cut-off was the 95th percentile of the controls (1198 U/ml) (p<0.001). Levels of INS-IGF2A and GADA correlated (p=0.024). The risk to be diagnosed with INS-IGF2A was increased among HLA-DQ2/8, but not 2/2 and 8/8, patients (OR=1.509; 95th CI 1.011, 2.252; p=0.045). Levels of INS-IGF2A correlated with antibodies against Ljungan virus (p=0.008). 


Doubly reactive INS-IGF2A sera at diagnosis support the notion that INS-IGF2 is an autoantigen in type 1 diabetes. The association specifically with HLA-DQ2/8 suggests that this autoantigen primarily is presented on HLA-DQ heterodimers in trans, rather than cis, transcription.