Protocol Version 1.1: version submitted in the initial applications to European regulatory authorities for approval. Summary of changes from original protocol (v.1.0, 08 February 2010) 1. Change of “patient” into “subject” throughout the protocol 2. Clarify storage conditions and temperature monitoring 3. Add a 10th administration of study medication at Visit 10 4. Add Visit 11 as the final follow-up/termination visit, at Month 25 5. Table 2: Add Regular Chemistry tests at all visits where other laboratory tests are run (Visits 1, 2, 4,6, 8 and 11) 6. Added text on the purpose of the CGMS use in the study and how the participating sites will be selected 7. Clarify that regardless of CGMS measurements, regular self-measurements of blood glucose should be continued 8. Added text on the purpose of performing the HLA typing as an exploratory study and the intended analysis of outcome data 9. Correct typo in name of enzyme: aspartate aminotransferase (AST) instead of alanine aminotransferase 10. Due to extension of the overall follow-up of period per subject from 24 to 25 months and the addition of an 11th visit (see changes #2 & #3, above), the final follow-up period was shortened from 3 months to 1 month. Accordingly, the follow-up of AE after last administration of study medication 11. Contact details for i3 Drug Safety Team in North America was added 12. Clarification of the correct form for reporting pregnancy during the study 13. Clarification that the Baseline for the MMTT will be either the Qualification Visit or Pre-Visit 1, if Qualification Visit is unnecessary 14. Correct typo, Qualification Visit occurs 1-3 weeks before Visit 1, not 2-4 weeks 15. Clarification regarding Informed Consent Process: a. “A person designated by the Investigator” was added as someone who can give the explanations about the study b. Participation in both HLA typing and CGMS sub-studies is voluntary, and subjects will sign separate ICFs if they wish to participate

Protocol Version 2.1 Rationale for amendment: most of the changes introduced into the protocol were extended explanations and clarifications regarding target population, time lines and procedures; Repeated queries from the sites indicated that some sections were not clear enough and might have resulted in different interpretation by different investigators. Significant issues that had to be clarified included: - Target population: the eligibility of subjects who were first diagnosed for months or years as T2D or LADA was questioned. This was clearly addressed in this protocol amendment. - Exclusion due to other diseases: subject with resolved basal cell or cervix carcinoma were allowed, as well as subjects that require systemic inhalers to treat/prevent asthma attacks. - Screening-Qualification Period-Randomization: timelines between the visits at the start of the study, how to handle subjects who require more time to reach optimal glycemic control, or who have to be re-screened. Clear time windows were provided for each visit up to randomization/Visit 1 - The original protocol required total fasting and no insulin medication on the morning of performing a beta-cell function test: it was recognized that this could often result in blood glucose level higher than the recommended range by the time the test is administered. Detailed recommendations were provided to the investigators in the Study Reference Manual so they could instruct their patients to take some insulin in the morning, according to actual blood glucose. - At the request of the German Federal Agency (BfArM), extra focus was placed on identifying and following-up incipient autoimmune diseases and immune effects.

Protocol Version 3.0 The reason for amending the protocol was mainly to better define the secondary endpoints, the statistical analysis and the handling of missing data. These changes were due to better understanding of the clinical outcome based on the availability of the final results from the first phase 3 study, protocol 901.

Protocol Version 4.0 List of Changes: 1. Primary efficacy endpoint defined as the change from Baseline to End of Study of glucagon-stimulated C-peptide secretion, measured as area under the curve (AUC). This is a change from change in mixed-meal induced C-peptide secretion. 2. Major clinical benefit parameters were defined, as suggested by FDA communication and clinical advisory board (CAB). These included glycemic control, insulin dose and hypoglycemic events. These parameters were already defined a secondary endpoints, but then they were highlighted as the main clinical endpoints. 3. Updates were introduced to the statistical section, following recommendations received from the FDA and input from CAB. These include: a. Re-definition of Primary, Major Clinical (secondary) efficacy endpoints, as described above. b. Definition of primary population for evaluation of efficacy as ITT subjects who had at least the Baseline and one post-baseline efficacy data (FDA recommendation). c. Definition of Completers Population analysis as main sensitivity analysis for primary efficacy (FDA recommendation). d. Use of mixed-model repeated measurements (MMRM) analysis, with pre-defined co-variants, for analysis of primary and secondary beta-cell function parameters (FDA recommendation). 4. Addition of liver enzyme testing at the visits of 9, 15, 21 & 25 months, according to DSMB recommendations. 5. Definition of how to proceed with safety assessment and IMP administration in cases of elevated liver enzymes, according to DSMB recommendation.