Stanley Lasch,1 Peter Müller,1 Monika Bayer,1 Josef M. Pfeilschifter,1

Andrew D. Luster,2 Edith Hintermann,1 and Urs Christen1

Anti-CD3/Anti-CXCL10 AntibodyCombination Therapy Inducesa Persistent Remission of Type 1Diabetes in Two Mouse ModelsDiabetes 2015;64:4198–4211 | DOI: 10.2337/db15-0479

Anti-CD3 therapy of type 1 diabetes results in a temporaryhalt of its pathogenesis but does not constitute a perma-nent cure. One problem is the reinfiltration of islets ofLangerhans with regenerated, autoaggressive lympho-cytes. We aimed at blocking such a reentry by neutral-izing the key chemokine CXCL10. Combination therapyof diabetic RIP-LCMV and NOD mice with anti-CD3 andanti-CXCL10 antibodies caused a substantial remissionof diabetes and was superior to monotherapy with anti-CD3 or anti-CXCL10 alone. The combination therapyprevented islet-specific T cells from reentering the isletsof Langerhans and thereby blocked the autodestructiveprocess. In addition, the local immune balance in thepancreas was shifted toward a regulatory phenotype. Asequential temporal inactivation of T cells and blockadeof T-cell migration might constitute a novel therapy forpatients with type 1 diabetes.

Type 1 diabetes (T1D) is a serious autoimmune-mediateddisease characterized by the progressive destruction ofinsulin-producing b-cells in the islets of Langerhans in thepancreas. Several attempts have been made to block theautoimmune destruction of these b-cells. One of the mostpromising therapies targets T cells using anti-CD3 antibodiessuch as hOKT3g1, teplizumab, and otelixizumab (alsoknown as ChAglyCD3), which have been evaluated in sev-eral clinical trials of patients with new and recently diag-nosed T1D, including the DEFEND-1 study (otelixizumab)and the Protégé study (teplizumab) (1–4). Similar to pre-clinical animal models (5,6), treatment with anti-CD3 anti-bodies was effective in decelerating the pathogenesis of

T1D in clinical trials (1,7,8). Administration of anti-CD3antibodies results in the inactivation of conventional T cellsand the expansion of previously constrained regulatoryT-cell (Treg) populations (9,10). Unfortunately, in themajority of treated patients, prevention of the declinein b-cell function lasted no longer than 1–2 years(1,2,8,11,12). In addition, many patients did not respondto the treatment, for unknown reasons (13).

Several combination therapies (CTs) to achieve long-termprotection in the majority of patients with T1D have beenassessed in preclinical models. In addition to anti-CD3antibodies, several immunomodulatory agents have beenused, including administration of nasal proinsulin (14),Lactococcus lactis–secreting interleukin (IL)-10/proinsulin(15), cyclosporine A and vitamin D3 analog (TX527) (16),IL-1 receptor antagonist (17), anti-CD20 antibody (18),fingolimod (FTY720) (19), the selective sphingosine 1 phos-phate 1 modulator ponesimod (20), dipeptidyl peptidase-4inhibitor MK626 (21), and HSP60 peptide p277 (22). Manyof these CTs were superior to monotherapies. However, withthe exception of CTs with anti-CD3 antibody/fingolimodtested in the LEW.1AR1-iddm rat model (19) and anti-CD3antibody/ponesimod investigated in the NOD mousemodel (20), none included blocking cell migration intothe islets. To avoid side effects associated with traditionalimmunosuppressive drugs, such as cytostatic drugs or glu-cocorticoids, the anti-CD3 antibody therapy aimed at theshort-term deletion/inactivation of T cells. However, theT-cell repertoire regenerates; therefore, one major problemis the reentry of regenerated autoaggressive lymphocytesinto the islets of Langerhans.

1Pharmazentrum Frankfurt/ZAFES, Goethe University Hospital Frankfurt, Frankfurtam Main, Germany2Massachusetts General Hospital, Harvard Medical School, Boston, MA

Corresponding author: Urs Christen, christen@med.uni-frankfurt.de.

Received 9 April 2015 and accepted 11 August 2015.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db15-0479/-/DC1.

© 2015 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

See accompanying article, p. 3990.

4198 Diabetes Volume 64, December 2015

IMMUNOLOGY

AND

TRANSPLANTATIO

N

Here we present data from a CT with a non-Fc-bindinganti-CD3e F(ab9)2 fragment (clone 145–2C11) (aCD3) (6)and a neutralizing anti-CXCL10 antibody (clone 1F11)(aCXCL10) (23). The chemokine CXCL10 (IP-10, a 10-kDainterferon [IFN]-g–inducible protein) has been demon-strated to play a key role in the pathogenesis of T1D inanimal models (24–27) and is elevated in islets of patientswith T1D (28,29). As an inducible model for T1D, we usedthe RIP-LCMV-GP model, in which transgenic mice ex-press the glycoprotein (GP) of the lymphocytic chorio-meningitis virus (LCMV) under the rat insulin promoter(RIP) in the b-cells of the islets of Langerhans (30). Inaddition, we used the NOD mouse as a model for spon-taneous T1D. We could demonstrate in both diabetic RIP-LCMV-GP and NOD mice that partial and temporal T-cellinactivation, followed by a blockade of CXCL10-mediatedT-cell migration, resulted in the persistent remission ofT1D. CT was superior to monotherapy in both models.A detailed analysis of the frequency and activity of isletantigen–specific T cells in the spleen and pancreas ofreverted RIP-LCMV-GP mice revealed that T cells indeedrecovered after aCD3 treatment but were prevented fromislet reentry by CXCL10 neutralization. This observationwas also supported by data obtained in the NOD modelfollowing islet-antigen peptide mimotope NRP-V7–specific T cells. Our data suggest that CT with aCD3 andaCXCL10 might constitute a novel therapy for patientswith T1D.

RESEARCH DESIGN AND METHODS

Mice and VirusH-2b RIP-LCMV-GP transgenic mice were generated andscreened by PCR as previously described (30,31). CXCL10mice were generated as previously described (32) and havebeen backcrossed to C57BL/6 mice for more than 10 years.NOD mice were from The Jackson Laboratory and werebred in the local breeding facility of the Georg-Speyer Haus,Frankfurt, Germany. Approximately 60% of female NODmice developed T1D within 14 to 26 weeks of age (data notshown). LCMV Armstrong clone 53b was produced as de-scribed elsewhere (31). Animal experiments were approvedby the local Ethics Animal Review Board, Darmstadt, Ger-many (V54–19c20/15-F143/56). Blood glucose (BG) wasmonitored at weekly intervals using a dynaValeo gluco-meter from dynamiCARE. Animals with BG concentrations.300 mg/dL were considered to be diabetic (33).

AntibodiesArmenian hamster anti-mouse CXCL10 monoclonal anti-body (clone 1F11) (23) was purified from hybridoma cellsupernatant on a HiTrap Protein G HP column. Monoclo-nal Armenian hamster anti-CD3 antibody (145–2C11F(ab9)2 fragment, pepsin digested) was obtained fromBio X Cell (Be0001–1FAB 4294/0212; West Lebanon,NH). Armenian hamster IgG (MBL/Biozol, Eching, Germany)was used as an isotype-matched control monoclonal antibodyfor the anti-mouse CXCL10 antibody.

ImmunohistochemistryTissues were immersed in Tissue-Tek OCT and quick-frozenon dry ice; 7-mm tissue sections were cut, then fixed inethanol or ethanol/acetone (1:1) at 220°C. Primary anti-bodies used include rat anti-mouse CD8a and rat anti-mouseCD4 (BD Biosciences) and a polyclonal guinea pig anti-swineinsulin antibody (DakoCytomation). Images of pancreas sec-tions were acquired with an Axioscope 2 microscope (Zeiss).

Double Immunofluorescence Staining of FoxP3+ TCellsTissues were immersed in Tissue-Tek OCT and quick-frozenon dry ice; 7-mm tissue sections were cut and fixed in eth-anol at220°C. Rat anti-mouse FoxP3 (eBioscience) and goatanti-rat Alexa Fluor 594 (Invitrogen) antibodies were used fordetection of FoxP3+ cells; directly conjugated rat anti-mouseCD8a-FITC and rat anti-mouse CD4-FITC (Southern Biotech)antibodies also were used. Images were acquired with aconfocal microscope (Zeiss LSM 510 META).

Isolation of Pancreatic LymphocytesThe pancreas was extracted and 3 mL of collagenase Psolution (1.2 U/mL in RPMI 1640) were injected. After30 min of incubation at 37°C, the collagenase solution wasremoved and cold RPMI 1640 containing 20 mg/mLDNase I was added. The pancreas was shaken for 1 minand pressed through a 70-mm cell strainer. The suspen-sion was washed with RPMI 1640 containing 20 mg/mLDNase I and resuspended again in RPMI containing20 mg/mL DNase I. Then, 40% Ficoll in PBS was overlayedwith the cell suspension and a gradient was performed.The pellet was washed with RPMI 1640 and resuspendedin RPMI 1640 containing 10% FCS.

Flow CytometrySingle-cell suspensions of spleen and pancreatic draininglymph nodes (PDLNs) were stimulated overnight with 2mg/mLLCMV peptides GP33 (CD8) and GP61 (CD4), or with 2 mg/mLof the NOD islet-antigen peptide mimotope NRP-V7 (34)in the presence of Brefeldin A. Cells were stained for sur-face expression of CD8 and CD4 and fixed, permeabilizedas previously described (33), and stained for intracellularIFN-g and FoxP3. V450-conjugated rat anti-CD4 antibody,allophycocyanin-Cy7-conjugated rat anti-CD8 antibody, andallophycocyanin-conjugated rat anti–IFN-g antibody all wereobtained from BD Biosciences. The phycoerythrin-conjugated rat anti-FoxP3 antibody was obtained fromeBioscience. Samples were acquired with a FACSCanto IIflow cytometer (BD Biosciences).

In Vivo Cytotoxicity AssayIn vivo cytotoxicity assay was performed as previouslydescribed (35). Briefly, splenocytes from C57BL/6 micewere divided into two groups. One group was pulsed over-night with 2 mg/mL LCMV-GP33 peptide. Peptide-pulsedcells were labeled at a final concentration of 0.5 mmol/L(carboxyfluorescein succinimidyl ester [CFSE]lo) and unpulsedcontrol cells at 5 mmol/L (CFSEhi). Equal amounts (1.5 3107) of CFSEhi and CFSElo cells were mixed and injected

diabetes.diabetesjournals.org Lasch and Associates 4199

intravenously into recipient mice. Specific in vivo cytotox-icity was determined by collecting blood and assessing theamounts of differentially CFSE-labeled target cell popula-tions by flow cytometry (FACSCanto II; BD Biosciences).The data obtained at different times after target cell in-jection were normalized against the ratio between CFSElo

and CFSEhi cells detected 10 min after transfer.

Statistical EvaluationsDiabetes incidence curves were analyzed using the Mantel-Cox log-rank test. T-cell frequencies and counts wereanalyzed using the unpaired, two-tailed t test (Prismsoftware version 5.02; GraphPad).

RESULTS

Administration of aCD3 Reduces the Frequency ofIslet-Specific T Cells in the PancreasTo find the optimal window for aCXCL10 administration,we first performed a single round of aCD3 therapy indiabetic RIP-LCMV-GP mice. At days 7 to 9 (when T-cellresponse peaked) and days 10 to 12 (the start of diabeteswith BG concentrations .300 mg/dL) after LCMV infec-tion, RIP-LCMV-GP mice received three daily intravenousinjections of either 3 or 30 mg aCD3. One day after thefinal aCD3 injection, the frequency of CD4 T cells wassignificantly decreased in the blood, spleen, pancreaticlymph nodes, and pancreas of mice receiving aCD3 atdays 10–12 after infection (Fig. 1A and SupplementaryFig. 1). Whereas administration of 30 mg aCD3 almostcompletely depleted CD4 T cells, injection of 3 mg resultedin a reduction of 40–60% (Fig. 1A and SupplementaryFig. 1). By contrast, the reduction of CD8 T cells wasmuch less pronounced (Fig. 1A and Supplementary Fig. 1).These data confirm recent observations that differentialexpression of CD3 on T-cell subsets influences aCD3-induced T-cell depletion (36). Importantly, the aCD3treatment affected the T cells in the pancreas and reducedthe local frequency of both CD8 and CD4 T cells (Fig. 1Aand Supplementary Fig. 1). Our data also confirm that thefrequency of FoxP3+ Tregs is increased after aCD3 treat-ment (Fig. 1A). Importantly, we also found that, afteraCD3 treatment, the total number of islet antigen(LCMV-GP)–specific T cells in the pancreas was signifi-cantly decreased and their frequency was reduced in thespleen (Fig. 1A). Examination of sections of pancreas tis-sue obtained 1 day after the last aCD3 injection revealedthat administration of both 3 and 30 mg of aCD3 reducedinsulitis and preserved the production of insulin (Supple-mentary Fig. 2). Administration of aCD3 at days 7–9 afterLCMV infection had a less pronounced effect on theT cells’ frequency, possibly because of the relatively highnumber of T cells present at that time (SupplementaryFig. 1A).

CT With aCD3 and aCXCL10 Persistently Reverts T1DTo avoid a general immune suppression, we used an aCD3dose of 3 mg for the following CT: Diabetic mice received

three intravenous injections of 3 mg aCD3 on days 10–12after infection with LCMV, followed by two initial intra-peritoneal injections of 100 mg aCXCL10 on days 13 and14 after infection, then six additional intraperitonealinjections of 100 mg aCXCL10 from day 17 to 28 afterinfection (three injections per week). The last aCXCL10injection was given on day 28 after infection (Fig. 1B). CTresulted in a remission of T1D in more than 60% of di-abetic mice (Fig. 1C). This remission was significantlygreater than the reduction of T1D after single therapywith either aCD3 (38%) or aCXCL10 (36%) (Fig. 1C). Atdays 12 (last aCD3 injection) and 35 (1 week after the endof CT), the majority (70%) of mice treated with CTshowed a reduced BG concentration, and numerous dia-betic mice remitted to normoglycemia (BG ,200 mg/dL)(Fig. 1D). By contrast, BG concentrations of most micereceiving no or only aCXCL10 monotherapy increasedfrom days 12 to 35 (Fig. 1D). Mice receiving aCD3 mono-therapy displayed a mixed outcome; a large fraction ofmice (47%) showed increased BG concentrations, but a sig-nificant fraction (47%) also presented with decreased BGconcentrations (Fig. 1D).

The remission of T1D and the blockade of furtherdestruction of b-cell mass was reflected by the degree ofinsulitis at day 31 after LCMV infection. In untreatedmice, islet infiltration and b-cell destruction progressover time, leaving almost no functional islets behind(Fig. 2A and B). By contrast, mice receiving CT showeda reduced degree of insulitis and retained more functionalb-cells (Fig. 2A and B). Similarly, a single administrationof aCD3 resulted in the delay of islet destruction anddecreased insulitis (Fig. 2A and B). Assessment of isletintegrity in mice that did not develop T1D until the endof the observation period (day 182 after infection)revealed that islets of mice treated with CT were largelydevoid of cellular infiltrations, whereas islets of mice thatreceived aCD3 alone displayed substantial insulitis, al-though they did not display hyperglycemia (Fig. 2C). Ithas to be noted here that only nondiabetic mice in theaCD3 and CT groups were analyzed at day 182 after in-fection. Isotype control and aCXCL10-treated mice devel-oped severe T1D and had to be killed much earlier becauseof severe T1D. These data indicate that administration ofaCXCL10 in addition to aCD3 results in a persistent re-duction of cellular recruitment to the islets, thereby block-ing the destructive process and reversing T1D.

CT Reduces the Reoccurrence of Islet-Specific T Cellsin the PancreasThe observed T1D remission was associated with a re-duced presence of islet antigen–specific T cells in the pan-creas. T-cell frequencies were analyzed at days 20 and 31after infection (Fig. 3). At the end of CT (day 31 afterinfection), the frequency of islet antigen–specific CD8 Tcells was decreased by approximately 73% in the spleen ofCT-treated mice compared with the spleen of isotype con-trol mice (Fig. 3A and C). aCD3 monotherapy reduced the

4200 Combination Therapy of Type 1 Diabetes Diabetes Volume 64, December 2015

frequency by 56%, and aCXCL10 monotherapy had nosignificant effect (Fig. 3A and C). By contrast, total CD8and CD4 T cells recovered completely in all treatmentgroups. Importantly, in the pancreas of mice receivingCT, the total number of islet antigen–specific T cells wasalso diminished by 78% compared with isotype controlanimals (Fig. 3A and C). Monotherapy with aCD3 oraCXCL10 reduced the frequency by only 23% or had noeffect, respectively (Fig. 3A and C). Furthermore, thenumber of pancreas-infiltrating total CD8 and CD4 T cellswas lower after CT, indicating that decreased overall in-flammation of the pancreas also results in the reducedrecruitment of bystander T cells (Fig. 3A and C). Theobserved effect was already detectable at day 20 afterinfection (i.e., after four of eight aCXCL10 injections).At that time, CD4 T cells had not yet completely recoveredfrom the aCD3 treatment, whereas CD8 T-cell levels werealready similar to those in isotype control mice (Fig. 3B).Islet antigen–specific CD8 T cells were already signifi-cantly diminished in the spleen of CT- and aCD3-treatedmice (Fig. 3B). Importantly, in the pancreas of CT-treatedmice, the total number of islet antigen–specific CD8 Tcells is significantly reduced by 81% (Fig. 3B). At day 20after infection, aCD3 or aCXCL10 monotherapy resultedin a reduction of islet antigen–specific CD8 T cells by 66%or 49%, respectively. Thus, when comparing the totalnumber of LCMV-GP33–specific CD8 T cells in the pan-creas at days 20 and 31, it becomes apparent that a re-entry of cells is prevented only in CT-treated mice, not inmice receiving aCD3 or aCXCL10 monotherapy (Fig. 3Band C). LCMV-specific CD4 T cells play only a minor rolein the RIP-LCMV-GP model, and the number of LCMV-GP61–specific CD4 T cells was below 1% in these experi-ments (Supplementary Fig. 3). Nevertheless, we coulddetect a reduction in the frequency of LCMV-GP61–specific CD4 T cells after treatment with aCD3 and CT(Supplementary Fig. 3).

In contrast to the absolute number of total CD4 T cells,the number of FoxP3+ CD4 T cells was not significantlyreduced in the pancreas after treatment with eitherantibody or CT (Fig. 3B and C). Hence, calculating theratio between islet antigen–specific CD8 T cells andFoxP3 CD4 T cells reveals a clear shift toward a regulatorymilieu. In particular, at day 31 after infection, the pan-creas of CT-treated mice shows a significant three- tofourfold elevation of the FoxP3+ CD4 versus islet anti-gen–specific CD8 T cells compared with mice receivingmonotherapy or isotype control treatment (Fig. 3D). A

Figure 1—CT results in the remission of T1D. A: Administration ofthree daily injections of 3 or 30 mg aCD3 at days 10–12 after in-fection results in the partial deletion of CD4 and CD8 T cells in thespleen and the pancreas. Frequencies of total CD4 T cells, totalCD8 T cells, and FoxP3+ CD4 T cells in the spleen, and their ab-solute numbers in the pancreas, were determined in four mice pertreatment group using flow cytometry. The frequency and absolutenumber of islet antigen (LCMV-GP33)–specific CD8 T cells weredetermined by stimulating isolated lymphocytes from spleen andpancreas with LCMV-GP33, followed by intracellular IFN-g stainingand flow cytometry. Significant differences are indicated (P values).C, control. B: For CT, mice were administered three daily injectionsof 3 mg aCD3 at days (d) 10–12 and eight injections (three injec-tions/week) of aCXCL10 or an isotype-matched control antibody atdays 13–28 after LCMV infection. ip, intraperitoneal; iv, intravenous.C: BG concentrations were determined until week 26 after infection.Concentrations >300 mg/dL were considered diabetic. T1D remis-sion was defined as a stable reversion of BG to concentrations

below 300 mg/dL. The number of mice analyzed in each group isindicated in parentheses. Significant differences are indicated(P values). D: Comparison of BG concentrations at days 12 and 35after infection. Mice that displayed at day 35 an increase in BG of>20 mg/dL, a decrease in BG of >20 mg/dL, or a largely un-changed concentration (within a range of 620 mg/dL) are indicatedin red, green, and gray, respectively.

diabetes.diabetesjournals.org Lasch and Associates 4201

tendency toward a more regulatory milieu could also bedetected in the spleen of CT-treated mice (Fig. 3D).

After CT and aCD3 monotherapy, the remaining isletantigen–specific CD8 T cells found in the pancreas at day20 after infection show little activity. They produced lessIFN-g (Fig. 3E and F), and only a small fraction of IFN-g–producing T cells also generated tumor necrosis factor(TNF)-a upon stimulation with the LCMV epitope GP33(Fig. 3E and F). By contrast, in the pancreas of untreatedor aCXCL10-treated mice, a large fraction (.30%) of islet-specific CD8 T cells also produced TNF-a (Fig. 3E and F).In the spleen of isotype control mice, almost 80% of isletantigen–specific CD8 T cells generated both IFN-g andTNF-a. There was only a slight decrease of such T cellsin CT-treated mice and no significant difference in aCD3-or aCXCL10-treated mice (Fig. 3F). Such a low frequency

of TNF-a–producing T cells in mice receiving CT or aCD3monotherapy indicates that most of the islet antigen–specific CD8 T cells were indeed newly regenerated, since,in contrast to experienced T cells, newly activated CD8 Tcells predominantly express IFN-g and only smallamounts of TNF-a (33,37–39).

The overall cytotoxicity to islet antigen peptide–loadedtarget cells was reduced in CT-treated mice. At day 31after infection, we performed an in vivo cytotoxicity assayusing differentially CFSE-labeled splenocytes as targetcells. LCMV-GP33–loaded CFSElo splenocytes were mixedwith an equal amount of unloaded CFSEhi splenocytes andwere transferred into LCMV-infected mice treated withCT, aCD3, aCXCL10, or isotype control. The ratio ofCFSElo to CFSEhi splenocytes was determined at severaltime points after the adoptive transfer of the target cells.

Figure 2—Insulitis is reduced in CT-treated RIP-LCMV-GP mice. A: Consecutive pancreas tissue sections obtained from mice receivingisotype-matched control antibody, aCD3 and isotype-matched control antibody (aCD3), aCXCL10 alone, or CT at day 31 after infectionwere stained for insulin, CD4 T cells (CD4), and CD8 T cells (CD8). Images of two representative islets per group are displayed (originalmagnification, 320). B: The degree of insulitis was scored as indicated for more than 100 individual islets from pancreas sections of fourmice. C: At the end of the observation time (day 182 after infection), pancreas sections from nondiabetic mice were stained for insulin, CD4,and CD8. Images of two representative islets per group are displayed (original magnification, 320). Note that, in contrast to aCD3monotherapy, CT resulted in a permanent block of insulitis.

4202 Combination Therapy of Type 1 Diabetes Diabetes Volume 64, December 2015

Abs

olut

e T

cell

num

ber

[Cou

nt/1

04 ]A

bsol

ute

T ce

ll nu

mbe

r[C

ount

/105 ]

Abs

olut

e T

cell

num

ber

[Cou

nt/1

05 ]A

bsol

ute

T ce

ll nu

mbe

r[C

ount

/104 ]

Abs

olut

e T

cell

num

ber

[Cou

nt/1

02 ]A

bsol

ute

T ce

ll nu

mbe

r[C

ount

/103 ]

Abs

olut

e T

cell

num

ber

[Cou

nt/1

02 ]A

bsol

ute

T ce

ll nu

mbe

r[C

ount

/103 ]

Figure 3—The number of islet antigen–specific T cells is reduced in CT-treated RIP-LCMV-GP mice. At days 20 and 31 after infection, thespleen and pancreas of RIP-LCMV mice treated with isotype-matched control antibody, aCD3 and isotype-matched control antibody (aCD3),aCXCL10 alone, or CT were removed and the frequency of CD4 T cells, CD8 T cells, FoxP3+ CD4 T cells, and RIP-LCMV-GP33–specific CD8T cells were analyzed by flow cytometry. The total numbers of infiltrating cells in the pancreas, rather than their relative frequencies, weredetermined. To this end, lymphocytes were isolated from the pancreas after removal of the pancreatic lymph nodes. A: Representative dotblots from flow cytometry of splenocytes and pancreatic lymphocytes harvested at day 31 after infection. Frequencies (spleen) and totalnumbers (pancreas) were calculated for total CD4 T cells, total CD8 T cells, FoxP3+ CD4 T cells, and RIP-LCMV-GP33–specific CD8 T cells atdays 20 (B) and 31 (C) after infection. Data are mean values 6 SD (day 20: n = 10; day 31: n = 6–13). Significant differences are indicated(P values). D: Calculated ratio of FoxP3+ CD4 T cells to RIP-LCMV–GP33–specific CD8 T cells in the spleen and pancreas at day 31 afterinfection. E and F: Islet antigen (RIP-LCMV-GP33)–specific CD8 T cells were further analyzed for TNF-a expression in splenocytes andpancreatic lymphocytes at day 20 after infection. Representative dot blots (E) and the calculated ratio of TNF-a–producing islet antigen–specificCD8 T cells (F) are displayed. Data are mean values6 SD (n = 5). Significant differences are indicated (P values). C, control; Teff, effector T cell.

diabetes.diabetesjournals.org Lasch and Associates 4203

In isotype control and aCXCL10-treated mice, a consider-able fraction of epitope-loaded target cells was killedwithin 6 h, and most of the target cells were gone after24 h (Fig. 4A and B). In CT- and aCD3-treated mice, thekilling was less efficient and delayed. Almost no targetcells were lost after 6 h, and a large proportion survivedfor more than 24 h (Fig. 4A and B). Evaluation of the half-life of the epitope-loaded target cells revealed a twofoldlonger mean survival time in CT-treated mice, indicatingthe reduced overall cytotoxic potency of the islet-specificimmune response (Fig. 4B and C).

CXCL10-Deficient Mice Display Reduced T1DFrequency, and aCD3 Administration Results inComplete RemissionTo further demonstrate that a lack of the key chemokineCXCL10 improves the efficacy of anti-CD3 therapy, wetreated CXCL10-deficient RIP-LCMV-GP mice with aCD3.For this experiment, regular RIP-LCMV-GP mice (30) were

crossed with CXCL10-deficient mice (32); homozygousCXCL102/2 3 RIP-LCMV-GP and CXCL10+/+ 3 RIP-LCMV-GP littermates were infected with LCMV and weretreated at days 10–12 with three doses of 3 mg aCD3 orwere left untreated. Whereas aCD3 treatment of RIP-LCMV-GP mice resulted in protection similar to that gainedbefore (;30% remission; compare with Fig. 1), all diabeticCXCL102/23 RIP-LCMV-GPmice reverted to a nondiabeticstate within 8 weeks after aCD3 administration andremained free of diabetes until the end of the observationperiod at week 28 after infection (Fig. 5). It has to be notedhere that the frequency of T1D was also reduced in un-treated CXCL102/2 3 RIP-LCMV-GP mice and that manyof these mice displayed only a mild form of T1D, allowingremission over time after LCMV infection (Fig. 5). Thesedata indicate that the impact of a total absence of CXCL10on the course of T1D in aCD3-treated mice is even strongerthan a neutralization with aCXCL10.

Figure 4—The overall cytotoxic activity toward islet antigen–presenting target cells is reduced in CT-treated RIP-LCMV-GP mice. A: In thein vivo cytotoxicity assay, target splenocytes isolated from uninfected C57BL/6 mice were split into two groups and were either loaded withthe immunodominant CD8 peptide GP33 or left unloaded. Peptide-pulsed and unpulsed splenocytes were labeled with a low (CFSElo) orhigh (CFSEhi) concentration of CFSE, respectively. CFSElo and CFSEhi populations were mixed at a 1:1 ratio and injected into RIP-LCMV-GP mice treated with isotype-matched control antibody, aCD3 and isotype-matched control antibody (aCD3), aCXCL10 alone, or CT at day31 after infection. Killing of target cells was assessed by flow cytometry of blood lymphocytes 6 and 24 h after cell injection. A represen-tative histogram is displayed for each group. B: The ratio of CFSElo to CFSEhi cells was calculated and normalized against the ratioobtained at the start (i.e., 10 min after injection). Target cell killing from 10 min to 48 h after injection is shown (mean 6 SD; n = 2–4). C:Calculated half-life of peptide-loaded target cells after injection. Note that a significant delay in killing was observed only in CT-treatedmice. C, control.

4204 Combination Therapy of Type 1 Diabetes Diabetes Volume 64, December 2015

CT Reverts T1D in Diabetic NOD MiceThe incidence data from the inducible RIP-LCMV mousemodel were confirmed in the spontaneous NOD mousemodel. Diabetic female NOD mice were treated with aCD3,isotype control, or CT within 1 week after becoming diabetic(BG .300 mg/dL). Dose-finding studies with NOD micerevealed that, in contrast to the RIP-LCMV-GP mice,a dose of 30 mg aCD3 was required to induce remissionfrom T1D in a fraction of the mice. However, even ata dose of 30 mg, large clusters of infiltrating T cells remainedaround the islets of Langerhans (Supplementary Fig. 4),which stands in contrast to the situation in aCD3-treatedRIP-LCMV mice (Supplementary Fig. 2). Similar to datafrom earlier studies (14,15,21), aCD3 monotherapy resultedin T1D remission in 30% of female NOD mice (3 of 10) (Fig.6A). CT of diabetic NOD mice improved the outcome andresulted in T1D remission in 55% of mice (6 of 11) (Fig. 6A).Immunohistochemistry revealed that pancreata of isotypecontrol–treated NOD mice did not contain any functionalislets, and most lymphocytes had already left the pancreas,leaving behind residual islet scars (Fig. 6B). By contrast,both aCD3 and CT-treated NOD mice displayed remainingfunctional islets producing insulin (Fig. 6B). However, largecellular infiltrates were still present in pancreas of NODmice receiving either treatment regimen (Fig. 6B). Scoringof islet infiltration revealed that, in contrast to untreatedNOD mice, which displayed massive infiltrations in all islets,aCD3 and CT treatment resulted in reduced overall insulitis(Fig. 6C).

At day 21 after the first aCD3 dose, lymphocytesisolated from spleen and PDLNs were stimulated with theNOD islet antigen peptide mimotope NRP-V7 (34) to de-termine the frequency of islet antigen–specific T cells byintracellular cytokine assay for IFN-g. Because of thelower frequency of islet antigen–specific T cells, the smallernumber of NOD mice used, and the poor synchronicity of

the pathogenesis between individual NOD mice, the dataobtained were not as evident as in the RIP-LCMV model.However, we could detect a tendency toward a reducedfrequency of islet antigen–specific CD8 T cells in thePDLNs of aCD3- and CT-treated diabetic NOD mice com-pared with isotype control mice (Fig. 7A and B). In addi-tion, we found a tendency toward a higher frequency ofFoxP3+ T cells after treatment with aCD3 or CT inthe spleen and the PDLNs (Fig. 7A and B). Similar to theexperiments with the RIP-LCMV model, we assessed theimmune balance locally in the pancreas and determinedthe ratio of aggressive, islet antigen–specific CD8 T cellsand FoxP3+ T cells in the NOD mouse model. As detectedat day 21 after the first dose of aCD3, CT caused a significantshift (.17-fold increase) in the ratio of FoxP3+ versus isletautoantigen–specific T cells toward a more regulatory andless aggressive milieu in the PDLNs (Fig. 7B). In contrastto the RIP-LCMV-GP model, the change of the insulitisphenotype was dominated by the relative increase ofFoxP3+ T cells rather than the relative decrease in isletantigen–specific aggressive CD8 T cells.

Interestingly, the majority of these FoxP3+ T cells was ofthe CD8, rather than the CD4, phenotype (Fig. 7A and B).Therefore, we performed double fluorescence immunohisto-chemistry of pancreas sections obtained from CT-, aCD3-,and isotype control–treated diabetic NOD mice at day 21after the first aCD3 dose. Using confocal microscopy, wefound an increased number of FoxP3+ T cells in the infil-trated islets of Langerhans of aCD3- and CT-treated mice(Fig. 7C). Differential counting of the cells revealed thatthe absolute number of FoxP3+ CD4 T cells was about 10-to 20-fold higher than FoxP3+ CD8 T cells in all treatmentgroups (Fig. 7D). The highest frequency of FoxP3+ CD4 Tcells was found in the aCD3-treated group (Fig. 7D). Therewas no significant difference in the frequency of FoxP3+CD8 T cells. However, the highest frequency was found inthe pancreas of CT-treated mice (Fig. 7D).

DISCUSSION

In contrast to many preclinical studies of rodents, anti-CD3 monotherapy caused only a temporary halt of T1Dprogress rather than permanently curing the disease(1,2,4,8,11,12). One reason for this lack of persistenceis the regeneration of the inactivated T-cell repertoire.Hence, to achieve long-term protection in the majorityof patients with T1D, several CTs pairing anti-CD3 ther-apy with a secondary treatment have been assessed inpreclinical models (14–19,21,22). However, none of theinvestigated CTs directly targeted the reentry of auto-aggressive lymphocytes, which regenerated after inactiva-tion by anti-CD3, into the islets of Langerhans. Here wereport that a CT of aCD3 and aCXCL10 persistently blocksT1D pathogenesis by preventing the reentry of auto-aggressive T cells into the islets. We used two differentmouse models for T1D, and we treated diabetic micewith three doses of aCD3 followed by eight injections ofneutralizing aCXCL10. In the inducible RIP-LCMV-GP and

Figure 5—T1D is reduced in anti-CD3–treated, CXCL10-deficientRIP-LCMV-GP mice (the numbers in parentheses indicate the num-ber of mice in each group). The study of diabetes remission innormal RIP-LCMV-GP and CXCL10-deficient RIP-LCMV-GP micethat were treated with aCD3. For aCD3 therapy, mice were admin-istered three daily injections of 3 mg aCD3 at days 10–12 afterLCMV infection. BG concentrations were determined until week26 after infection. Concentrations >300 mg/dL were consideredto be diabetic. T1D remission was defined as a stable reversionof BG to concentrations below 300 mg/dL. Significant differencesare indicated (P values).

diabetes.diabetesjournals.org Lasch and Associates 4205

Figure 6—T1D is reduced in CT-treated NOD mice. A: Diabetes remission was studied in diabetic NOD mice. After diabetes onset(BG >300 mg/dL), female NOD mice were treated with three daily injections of 30 mg aCD3. BG concentrations were determined untilweek 26 after diabetes onset. Concentrations >300 mg/dL were considered to be diabetic. T1D remission was defined as a stable reversionof BG to concentrations below 300 mg/dL. Significant differences are indicated (P values). B: Consecutive pancreas tissue sections obtainedfrom mice treated with isotype-matched control antibody, aCD3 and isotype-matched control antibody (aCD3), aCXCL10 alone, or CT at day21 after the first dose of aCD3 were stained for insulin, CD4 T cells (CD4), and CD8 T cells (CD8). Images of two representative islets of twomice per group are displayed (original magnification, 320). The BG values (milligrams per deciliter) of the corresponding mice measured atday 0 and day 21 after the first dose of aCD3 are indicated in the lower left corners of the images in column 1. C: The degree of insulitis forindividual islets from pancreas sections of three mice was scored as indicated.

4206 Combination Therapy of Type 1 Diabetes Diabetes Volume 64, December 2015

Figure 7—The immune balance in the PDLNs is shifted toward a more regulatory milieu after CT in RIP-LCMV-GP and NOD mice. A: At day 21after the first injection of aCD3, spleen and PDLNs of NOD mice treated with isotype-matched control antibody, aCD3 and isotype-matchedcontrol antibody (aCD3), or CT were removed. The frequency of CD4 T cells, CD8 T cells, FoxP3+ T cells, and NRP-V7–specific CD8 T cells wereanalyzed by flow cytometry. Representative dot blots from flow cytometry of splenocytes and PDLNs are displayed. B: Calculated frequenciesof total CD8 T cells, total CD4 T cells, FoxP3+ CD4 T cells, FoxP3+ CD8 T cells, and NRP-V7–specific CD8 T cells in the spleen and the PDLNs.Data are mean values6 SD (n = 4–6). Note that the calculated ratio of FoxP3+ T cells to NRP-V7–specific CD8 T cells (Treg/effector T cell [Teff])is significantly higher in the PDLNs of CT-treated mice compared with those treated with aCD3 or isotype-matched control antibody (right panel,bottom row). C: Double immunofluorescence staining of pancreas sections obtained at day 21 after the first dose of aCD3 from NOD micetreated with isotype-matched control antibody, aCD3 and isotype-matched control antibody (aCD3), or CT. Sections were stained for CD4(green), FoxP3 (red), and CD8 (green). Images of representative islets of three mice per group are displayed (left; original magnification, 340).Magnified views (right) show the presence of FoxP3+ (nuclear staining) CD4 and CD8 (both membrane-staining) T cells (white arrowheads).D: Total CD4 and CD8 T cells and FoxP3+ cells were counted; the frequencies of FoxP3+ CD4 and CD8 T cells are displayed. C, control.

diabetes.diabetesjournals.org Lasch and Associates 4207

the spontaneous NOD mouse model, CT induced T1Dremission in 60% and 55% of diabetic mice, respectively.CT was superior to monotherapies with either aCD3 oraCXCL10, and the observed T1D remission was reflectedin restored insulin production and reduced insulitis.Mechanistically, we found that the sequential inactivationof T cells followed by a blockade of T-cell migrationresulted in a change in the composition of T cells in thepancreas. In the RIP-LCMV-GP model the total number ofislet antigen–specific CD8 T cells was significantly reducedin the pancreas, whereas in NOD mice we found onlya tendency toward a reduction of specific T cells. It wasrecently demonstrated by in situ MHC-peptide tetramerstaining of pancreas tissue from cadaveric donors withT1D that islet autoantigen–specific T cells are indeed pres-ent in the islets of patients with T1D up to 8 yearsafter clinical diagnosis (40). These data suggest that a

long-lasting reduction of islet autoantigen–specific T cellsin the vicinity of the islets might be important in a per-sistent abrogation of disease. In contrast to RIP-LCMV-GPmice, we found a marked increase in FoxP3+ T cells inNOD mice receiving CT. Direct analysis of infiltratedislets by double immunofluorescence staining and confocalmicroscopy revealed that the majority of FoxP3+ T cellswere CD4+, as previously reported (9,10). However, wealso detected some FoxP3+ CD8 T cells in the pancreasand, to a larger extent, in the PDLNs. Such CD8+FoxP3+T cells have also been detected in patients with T1D afteraCD3 therapy (41). Importantly, in both the RIP-LCMVand the NOD mouse models, the ratio between FoxP3+T cells and islet antigen–specific CD8 T cells was signifi-cantly higher in CT-treated mice, indicating that a shiftin the immune balance toward regulation in the proximityof the islets of Langerhans might be responsible for the

Figure 7—Continued.

4208 Combination Therapy of Type 1 Diabetes Diabetes Volume 64, December 2015

persistent blockade of the autodestructive process. In ourhands the aCD3 monotherapy was also effective. Howeverthe CT was far more effective (60% vs. 38% remission),particularly in the RIP-LCMV model. An aCD3 monother-apy was evaluated in the RIP-LCMV model before (42)and resulted in a higher rate of remission (75–100%,depending on the time of aCD3 administration). How-ever, in contrast to our studies, which used a rather lowaCD3 dose (three injections of 3 mg), five daily doses of100 mg were administered in the previous study (42).

CXCL10 has been found to be elevated in islets ofpatients with T1D (28,29). In the RIP-LCMV model,CXCL10 was induced very soon after LCMV infection(25). Neutralization of CXCL10 resulted in reduced T1Dfrequency and decreased recruitment of CXCR3-positive Tcells to the pancreas (25). Reduced incidence and delayedonset of T1D have also been reported for CXCR3-deficientRIP-LCMV mice (24). However, a recent follow-up studyusing aCXCL10-treated or CXCR3-deficient RIP-LCMV micesuggested a certain redundancy of the CXCL10/CXCR3 axis(27). Here we found distinct T1D remission after aCXCL10monotherapy. In contrast to our earlier studies of T1Dprevention (25,27), mice were treated with aCXCL10 for19 days after diabetes onset. A similar discrepancy existsfor CXCL10-deficient RIP-LCMV-GP mice, which, in a pre-vious study, developed T1D between day 9 and 14 afterLCMV infection, just like regular RIP-LCMV-GP mice (27).In our hands the majority of CXCL10-deficient RIP-LCMV-GP mice also developed T1D within 2 weeks after infection.Many of these diabetic mice reverted in the followingweeks, however, and when additionally treated withaCD3, all of the diabetic CXCL10-deficient RIP-LCMV-GP mice display full remission. Thus, one can speculatethat CXCL10 plays a critical role in retaining aggressiveT cells in the islets, whereas at earlier times other inflam-matory factors might compensate for the relative lack ofCXCL10.

Still, aCXCL10 monotherapy seems unlikely to beeffective in patients with T1D. The reason for a likelyfailure is that, even if the concept of pathogens beinginvolved in T1D etiology (43,44) holds true, at the time ofdiagnosis the responsible pathogenic infection would liein the past, and it therefore would be impossible to in-terfere with the initial chemokine burst. In the CT setting,the administration of aCXCL10 follows a precise schedule,in which diabetic mice are treated first with low-doseaCD3 to inactivate a significant portion of T cells, includingislet antigen–specific CD8 T cells. Thereafter, aCXCL10administration prevents the de novo migration of reacti-vated/regenerated T cells to the islets. Such a regimenseems to be realistic for therapeutic application inpatients with T1D to prevent the reported relapse within2 years of aCD3 treatment (1,2,8,11,12). The effect of CTin the RIP-LCMV-GP and the NOD mouse models waslong-lasting. Even 20 weeks after remission, no relapsewas detected. Insulin production was maintained; in par-ticular, islets of CT-treated RIP-LCMV mice remained

largely without insulitis. In NOD mice we detected a stateof peri-insulitis similar to that in young mice that havenot yet developed T1D. Such a steady state might be pre-served by the presence of a high frequency of FoxP3+ Tcells maintaining a long-term regulatory milieu. Interest-ingly, no further administration of aCXCL10 is needed tomaintain T1D remission after the initial eight injectionsfollowing aCD3 therapy. Note that at the end of CT, noinfectious virus particles were detected in the spleen andpancreas of RIP-LCMV-GP mice, as determined by virusplaque assay (45) (data not shown). Thus, reactivation ofthe LCMV-GP–specific islet-destructing CD8 T cells islikely to occur via presentation of b-cell–derived trans-genic LCMV-GP rather than through remaining LCMV.

Among different aCD3 CTs, it seems to be impor-tant to target two distinct mechanisms to achieve a signif-icant improvement over the corresponding monotherapy.Whereas some CTs aim at generating islet antigen–specificTregs (14,15), others use two antibodies in parallel, such asaCD3 combined with an anti-CD20 antibody (aCD20) (18).Similar to our study, the aCD3/aCD20 CT of NOD micedemonstrated that intravenous injection of two antibodiesis successful in reversing T1D. In contrast to the persis-tent effect of our aCD3/aCXCL10 CT, however, the parallelinactivation of T and B cells with aCD3/aCD20 caused T1Dremission within 1 month of treatment, but most of themice showed a relapse by month 3 after treatment (18). Inany case, the therapeutic window relative to T1D onsetmight be critical for success. We treated RIP-LCMV andNOD mice early after onset. At this stage, most of theb-cells are functionally inactivated by inflammatory stressrather than completely eradicated. Thus, stress reliefthrough the inactivation of T cells and prevention of denovo insulitis restored b-cell function. Later, when mostof the b-cells are physically desztroyed, such CT might beineffective, since b-cells’ natural regeneration might beinsufficient.

In conclusion, we showed that CT with aCD3 andaCXCL10 results in persistent T1D remission in diabeticRIP-LCMV-GP and NOD mice. Both models do notentirely reflect the pathogenesis of human T1D. On theone hand, in contrast to human T1D, the RIP-LCMV-GPmodel is independent of CD4 T-cell help but has theadvantage of being inducible by a well-defined triggeringevent (i.e., virus infection), which results in highlysynchronized pathogenic events. This allows for the precisetiming of a given therapeutic intervention in relation topathogenic status. The NOD mouse, on the other hand, isa spontaneous model in which a genetic predispositionresults in T1D-like disease independent of environmentaltriggering events. Just as in human T1D, however, thedestruction of b-cells is dependent on both CD4 and CD8 Tcells. Although our data from those two models are notcompletely identical, it is important to acknowledge thatneutralization of the critical inflammatory chemokineCXCL10 directly after the transient inactivation of T cellswith low-dose aCD3 results in the diminished presence of

diabetes.diabetesjournals.org Lasch and Associates 4209

autoaggressive T cells in the pancreas. The ongoing func-tional inactivation and destruction of b-cells is therebyhalted, and this truce is maintained by a shift in the localimmune balance toward regulation. Since the administra-tion of both low-dose aCD3 as well as aCXCL10 is onlytransient, such a CT might be highly attractive for an ap-plication in patients with T1D.

Funding. This study was supported by the Else Kröner-Fresenius Foundation(EKFS), Research Training Group Translational Research Innovation—Pharma(TRIP), the German Research Foundation (DFG), and the Goethe University HospitalFrankfurt.Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. S.L. designed the study, performed experiments,interpreted data, and drafted the manuscript. P.M. and M.B. performed experiments.J.M.P. critically revised the manuscript. A.D.L. designed the study and criticallyrevised the manuscript. E.H. performed experiments and critically revised themanuscript. U.C. conceived of and designed the study, interpreted data, and draftedthe manuscript. U.C. is the guarantor of this work and, as such, had full access to allthe data in the study and takes responsibility for the integrity of the data and theaccuracy of the data analysis.

References1. Keymeulen B, Vandemeulebroucke E, Ziegler AG, et al. Insulin needs after CD3-

antibody therapy in new-onset type 1 diabetes. N Engl J Med 2005;352:2598–26082. Sherry N, Hagopian W, Ludvigsson J, et al.; Protégé Trial Investigators.

Teplizumab for treatment of type 1 diabetes (Protégé study): 1-year results from

a randomised, placebo-controlled trial. Lancet 2011;378:487–4973. Daifotis AG, Koenig S, Chatenoud L, Herold KC. Anti-CD3 clinical trials in

type 1 diabetes mellitus. Clin Immunol 2013;149:268–2784. Aronson R, Gottlieb PA, Christiansen JS, et al.; DEFEND Investigator Group.

Low-dose otelixizumab anti-CD3 monoclonal antibody DEFEND-1 study: results

of the randomized phase III study in recent-onset human type 1 diabetes. Di-

abetes Care 2014;37:2746–27545. Chatenoud L, Thervet E, Primo J, Bach JF. Anti-CD3 antibody induces long-

term remission of overt autoimmunity in nonobese diabetic mice. Proc Natl Acad

Sci U S A 1994;91:123–1276. Chatenoud L, Primo J, Bach JF. CD3 antibody-induced dominant self tol-

erance in overtly diabetic NOD mice. J Immunol 1997;158:2947–29547. Herold KC, Hagopian W, Auger JA, et al. Anti-CD3 monoclonal antibody in

new-onset type 1 diabetes mellitus. N Engl J Med 2002;346:1692–16988. Herold KC, Gitelman SE, Masharani U, et al. A single course of anti-CD3

monoclonal antibody hOKT3gamma1(Ala-Ala) results in improvement in C-peptide

responses and clinical parameters for at least 2 years after onset of type 1

diabetes. Diabetes 2005;54:1763–17699. Chatenoud L, Bluestone JA. CD3-specific antibodies: a portal to the treat-

ment of autoimmunity. Nat Rev Immunol 2007;7:622–63210. Nishio J, Feuerer M, Wong J, Mathis D, Benoist C. Anti-CD3 therapy permits

regulatory T cells to surmount T cell receptor-specified peripheral niche con-

straints. J Exp Med 2010;207:1879–188911. Keymeulen B, Walter M, Mathieu C, et al. Four-year metabolic outcome of

a randomised controlled CD3-antibody trial in recent-onset type 1 diabetic pa-

tients depends on their age and baseline residual beta cell mass. Diabetologia

2010;53:614–62312. Hagopian W, Ferry RJ Jr, Sherry N, et al.; Protégé Trial Investigators. Teplizumab

preserves C-peptide in recent-onset type 1 diabetes: two-year results from the

randomized, placebo-controlled Protégé trial. Diabetes 2013;62:3901–390813. Herold KC, Gitelman SE, Ehlers MR, et al.; AbATE Study Team. Teplizumab

(anti-CD3 mAb) treatment preserves C-peptide responses in patients with new-onset

type 1 diabetes in a randomized controlled trial: metabolic and immunologic featuresat baseline identify a subgroup of responders. Diabetes 2013;62:3766–377414. Bresson D, Togher L, Rodrigo E, et al. Anti-CD3 and nasal proinsulincombination therapy enhances remission from recent-onset autoimmune di-abetes by inducing Tregs. J Clin Invest 2006;116:1371–138115. Takiishi T, Korf H, Van Belle TL, et al. Reversal of autoimmune diabetes byrestoration of antigen-specific tolerance using genetically modified Lactococcuslactis in mice. J Clin Invest 2012;122:1717–172516. Baeke F, Van Belle TL, Takiishi T, et al. Low doses of anti-CD3, ciclosporin Aand the vitamin D analogue, TX527, synergise to delay recurrence of autoimmunediabetes in an islet-transplanted NOD mouse model of diabetes. Diabetologia 2012;55:2723–273217. Ablamunits V, Henegariu O, Hansen JB, et al. Synergistic reversal of type 1diabetes in NOD mice with anti-CD3 and interleukin-1 blockade: evidence ofimproved immune regulation. Diabetes 2012;61:145–15418. Hu C, Ding H, Zhang X, Wong FS, Wen L. Combination treatment with anti-CD20 and oral anti-CD3 prevents and reverses autoimmune diabetes. Diabetes2013;62:2849–285819. Jörns A, Akin M, Arndt T, et al. Anti-TCR therapy combined with fingolimodfor reversal of diabetic hyperglycemia by b cell regeneration in the LEW.1AR1-iddm rat model of type 1 diabetes. J Mol Med (Berl) 2014;92:743–75520. You S, Piali L, Kuhn C, et al. Therapeutic use of a selective S1P1 receptormodulator ponesimod in autoimmune diabetes. PLoS One 2013;8:e7729621. Ding L, Gysemans CA, Stangé G, et al. Combining MK626, a novel DPP-4inhibitor, and low-dose monoclonal CD3 antibody for stable remission of new-onset diabetes in mice. PLoS One 2014;9:e10793522. Sarikonda G, Sachithanantham S, Miller JF, Pagni PP, Coppieters KT, vonHerrath M. The Hsp60 peptide p277 enhances anti-CD3 mediated diabetes re-mission in non-obese diabetic mice. J Autoimmun 2015;59:61–6623. Khan IA, MacLean JA, Lee FS, et al. IP-10 is critical for effector T cell traffickingand host survival in Toxoplasma gondii infection. Immunity 2000;12:483–49424. Frigerio S, Junt T, Lu B, et al. Beta cells are responsible for CXCR3-mediated T-cell infiltration in insulitis. Nat Med 2002;8:1414–142025. Christen U, McGavern DB, Luster AD, von Herrath MG, Oldstone MB. AmongCXCR3 chemokines, IFN-gamma-inducible protein of 10 kDa (CXC chemokineligand (CXCL) 10) but not monokine induced by IFN-gamma (CXCL9) imprintsa pattern for the subsequent development of autoimmune disease. J Immunol2003;171:6838–684526. Rhode A, Pauza ME, Barral AM, et al. Islet-specific expression of CXCL10causes spontaneous islet infiltration and accelerates diabetes development. JImmunol 2005;175:3516–352427. Coppieters KT, Amirian N, Pagni PP, et al. Functional redundancy of CXCR3/CXCL10 signaling in the recruitment of diabetogenic cytotoxic T lymphocytes topancreatic islets in a virally induced autoimmune diabetes model. Diabetes 2013;62:2492–249928. Shimada A, Morimoto J, Kodama K, et al. Elevated serum IP-10 levelsobserved in type 1 diabetes. Diabetes Care 2001;24:510–51529. Roep BO, Kleijwegt FS, van Halteren AG, et al. Islet inflammation andCXCL10 in recent-onset type 1 diabetes. Clin Exp Immunol 2010;159:338–34330. Oldstone MBA, Nerenberg M, Southern P, Price J, Lewicki H. Virus infectiontriggers insulin-dependent diabetes mellitus in a transgenic model: role of anti-self (virus) immune response. Cell 1991;65:319–33131. von Herrath MG, Dockter J, Oldstone MBA. How virus induces a rapid or slowonset insulin-dependent diabetes mellitus in a transgenic model. Immunity 1994;1:231–24232. Dufour JH, Dziejman M, Liu MT, Leung JH, Lane TE, Luster AD. IFN-gamma-inducible protein 10 (IP-10; CXCL10)-deficient mice reveal a role for IP-10 in effector T cell generation and trafficking. J Immunol 2002;168:3195–320433. Christen U, Wolfe T, Möhrle U, et al. A dual role for TNF-alpha in type 1diabetes: islet-specific expression abrogates the ongoing autoimmune processwhen induced late but not early during pathogenesis. J Immunol 2001;166:7023–7032

4210 Combination Therapy of Type 1 Diabetes Diabetes Volume 64, December 2015

34. Trudeau JD, Kelly-Smith C, Verchere CB, et al. Prediction of spontaneousautoimmune diabetes in NOD mice by quantification of autoreactive T cells inperipheral blood. J Clin Invest 2003;111:217–22335. Ehser J, Holdener M, Christen S, et al. Molecular mimicry rather thanidentity breaks T-cell tolerance in the CYP2D6 mouse model for human auto-immune hepatitis. J Autoimmun 2013;42:39–4936. Valle A, Barbagiovanni G, Jofra T, et al. Heterogeneous CD3 expressionlevels in differing T cell subsets correlate with the in vivo anti-CD3-mediated Tcell modulation. J Immunol 2015;194:2117–212737. Murali-Krishna K, Altman JD, Suresh M, et al. Counting antigen-specificCD8 T cells: a reevaluation of bystander activation during viral infection. Immunity1998;8:177–18738. Slifka MK, Rodriguez F, Whitton JL. Rapid on/off cycling of cytokine pro-duction by virus-specific CD8+ T cells. Nature 1999;401:76–7939. Slifka MK, Whitton JL. Activated and memory CD8+ T cells can be distinguishedby their cytokine profiles and phenotypic markers. J Immunol 2000;164:208–216

40. Coppieters KT, Dotta F, Amirian N, et al. Demonstration of islet-autoreactiveCD8 T cells in insulitic lesions from recent onset and long-term type 1 diabetespatients. J Exp Med 2012;209:51–6041. Bisikirska B, Colgan J, Luban J, Bluestone JA, Herold KC. TCR stimula-tion with modified anti-CD3 mAb expands CD8+ T cell population and inducesCD8+CD25+ Tregs. J Clin Invest 2005;115:2904–291342. von Herrath MG, Coon B, Wolfe T, Chatenoud L. Nonmitogenic CD3 antibodyreverses virally induced (rat insulin promoter-lymphocytic choriomeningitis virus) au-toimmune diabetes without impeding viral clearance. J Immunol 2002;168:933–94143. Christen U, Bender C, von Herrath MG. Infection as a cause of type 1 di-abetes? Curr Opin Rheumatol 2012;24:417–42344. Kondrashova A, Hyöty H. Role of viruses and other microbes in thepathogenesis of type 1 diabetes. Int Rev Immunol 2014;33:284–29545. Christen U, Benke D, Wolfe T, et al. Cure of prediabetic mice by viral in-fections involves lymphocyte recruitment along an IP-10 gradient. J Clin Invest2004;113:74–84

diabetes.diabetesjournals.org Lasch and Associates 4211