Characterization Studies of the Multiple Autoimmune Disease (MAD) Rat:

Type 1 Diabetes

S Duclos, B.S., M Gardner, S Champagne, B.S., RLAT, C Hogan, M.S., H Johnson, K Guberski, B.S.

Biomedical Research Models Inc., Worcester, MA

.

This research and development project was supported

in part by NIH SBIR grant R43DK085910.

Many thanks to the BRM technical staff.

Biomedical Research Models, Inc. has developed an inbred rat strain

susceptible to Multiple Autoimmune Diseases (MAD). The MAD rats were

derived from congenic LEW1.WR1 rats. During inbreeding (F55) selection

was made for autoimmune features. The goal was to develop a reliable

model to test the efficacy of treatments designed to prevent or reverse

autoimmune disease and for screening for adverse effects of these

therapies.

Immunological (TLR ligands) and viral perturbants have been identified

which increase the penetrance of disease in the several rat models of type

1 diabetes (T1D) (e.g. BBDP, BBDR and KDP rats). The MAD rat has a

normal immunological phenotype but develops spontaneous autoimmune

diabetes with a frequency of ≤ 1%.The hypothesis was made that the MAD

rat strain is genetically predisposed to the development of autoimmune

diseases that may progress to overt disease in concert with appropriate

environmental perturbants. To test this hypothesis we evaluated the

induction of T1D when treated with Toll-Like receptor (TLR) ligands. A dose

response curve was also generated for virus diabetogenicity. The age of

susceptibility was also investigated to further characterize the model and

potentially enhance future application of regimens aimed at

prevention/reversal of diabetes

Biomedical Research Models, Inc.

508-459-7544

www.brmcro.com

A panel of TLR agonists was tested for their ability to induce T1D in MAD

rats. We hypothesized that appropriate activation of innate immune cells

that initiate T1D in genetically susceptible hosts may lead directly to overt

disease. The doses were chosen based on previous studies that

demonstrated diabetogenicity of certain TLR agonists in concert with

Kilham rat virus (KRV) infection in the MHC-related rat diabetes resistant

BB rat strain (BBDR) 2,3. Standard commercially available TLR agonists

were administered at 0.5, 1, and 2 times the doses previously known to

synergize with virus infections in BBDR/Wor rats, but that are not

diabetogenic in the absence of virus. We also tested both high and low

molecular weight forms of poly I:C (TLR3 agonist). Rats were dosed 3

times weekly by IP injection beginning at 21 - 24 days of age. Compounds

were administered to 6 ♀ and 6 ♂ rats per compound or dose level, as

appropriate. Rats were treated over a 30 day period and were monitored

for diabetes from 7 to 40 days after the initiation of treatment.

In addition to screening for T1D, rats treated with TLR ligands were also

examined for evidence of inflammatory arthritis, the presence of which

would be an “off-target” effect of the induction procedure.

Diabetes can be induced in the class II MHC-identical BBDR/Wor rat

strain only until 24 - 28 days of age, after which they show resistance to

induction (unpublished data). We proposed to investigate the window of

age susceptibility in the MAD rat strain to further characterize the model

and potentially enhance future application of regimens aimed at

prevention/reversal of diabetes. We chose to use our reference

diabetogenic compound HMW poly I:C (Table 1 Group 1 and Table 3) for

these studies. Groups of 6 male and 6 female rats were administered poly

I:C (1 mg/kg) three times weekly IP until diabetes onset or until 30 days

after the initiation of treatment. Treatments were initiated at ages ranging

from 21 to 44 days.

All experimental animal procedures were carried out in accordance with

the governmental guidelines and approved by the Institutional Animal

Care and Use Committee.

• Administration of our historical standard poly I:C (primarily HWM) dose

of 1.0 mg/kg resulted in a high incidence of T1D in MAD rats in rats of

both genders (Table 1).

• The MW composition of the poly I:C material has no apparent effect on

T1D onset in this model (Table 2). It has been reported that HMW poly

I:C signals through pathways different from those activated by the low

molecular weight (LMW) form, with potential implications in autoimmune

disease (4,5). Since both HMW and LMW preparations induced

comparable frequencies of T1D at similar doses and with similar kinetics

in both genders, this data suggests that a broad spectrum of double

stranded RNA molecular species may be able to trigger autoimmune

diabetes in susceptible hosts (HMW through MDA5 signaling pathway

and LMW through RIG-I signaling pathway).

• Zymosan (Table 3) and R848 (Table 4) showed a substantial ability to

induce diabetes in this rat strain, whereas CpG (Table 5) failed to cause

glycosuria or elevated blood glucose levels in MAD rats.

• These TLR study results are presented in Table 6.

• Diabetes can be induced in the class II MHC-identical BBDR/Wor rat

strain only until 24 - 28 days of age, after which they show resistance to

induction (unpublished data). We proposed to investigate the window of

age susceptibility in the MAD rat strain to further characterize the model

and potentially enhance future application of regimens aimed at

prevention/reversal of diabetes. As shown in Table 6, induction of T1D

with standard dose poly I:C is highly efficient when treatment is started

up to 34 days of age (66 - 75%); disease penetrance then diminishes

significantly at 41 - 44 days of age (27%) (Table 7). These findings will

allow for a larger window for pretreatment with potential T1D therapies

for more flexible experimental designs in future studies.

The TLR agonists studied during this characterization are known to

synergize with KRV virus infection in the BBDR rat, but, except for poly

I:C, they do not induce diabetes or insulitis by themselves when

administered in these dose ranges in short-term dosing regimens (3

times) in BBDR rats (3,6). Furthermore, administration of these agents

to the spontaneously diabetic NOD mouse either prevents or delays the

onset of T1D (6,7). These results distinguish the MAD rat model from

the NOD mouse model. The MAD rat exhibits minimal spontaneous

disease in clean housing, but is readily induced to become diabetic with

selected purified TLR agonists. In contrast, the NOD mouse model has

a high frequency of spontaneous disease that is prevented by these

TLR agonists, as well as an extensive repertoire of other interventions

(8). The interventions proven useful in NOD mice have, to date, not

proven useful in human clinical trials (9), making the MAD rat a more

attractive competitive research tool for the evaluation of candidate

interventions.

1. Mordes JP, Poussier P, Rossini AA, Blankenhorn EP, Greiner DL:

Rat models of type 1 diabetes: Genetics, environment, and

autoimmunity. In Animal Models of Diabetes: Frontiers in

Research. 2 ed. Shafrir E, Ed. Boca Raton, CRC Press, 2007, p. 1-39

2. Mordes,JP, Guberski,DL, Leif,JH, Woda,BA, Flanagan,JF,

Greiner,DL, Kislauskis,EH, Tirabassi,RS: LEW.1WR1 rats develop

autoimmune diabetes spontaneously and in response to

environmental perturbation. Diabetes 54:2727-2733, 2005

3. Zipris,D, Lien,E, Xie,JX, Greiner,DL, Mordes,JP, Rossini,AA: TLR

activation synergizes with Kilham rat virus infection to induce

diabetes in BBDR rats. J Immunol 174:131-142, 2005

4. Kato,H, Takeuchi,O, Mikamo-Satoh,E, Hirai,R, Kawai,T,

Matsushita,K, Hiiragi,A, Dermody,TS, Fujita,T, Akira,S: Length-

dependent recognition of double-stranded ribonucleic acids by

retinoic acid-inducible gene-I and melanoma differentiation-

associated gene 5. J Exp Med 205:1601-1610, 2008

5. Pichlmair,A, Schulz,O, Tan,CP, Rehwinkel,J, Kato,H, Takeuchi,O,

Akira,S, Way,M, Schiavo,G, Reis e Sousa: Activation of MDA5

requires higher-order RNA structures generated during virus

infection\. J Virol 83:10761-10769, 2009

6. Lien,E, Zipris,D: The role of Toll-like receptor pathways in the

mechanism of type 1 diabetes. Curr Mol Med 9:52-68, 2009

7. Serreze,DV, Hamaguchi,K, Leiter,EH: Immunostimulation

circumvents diabetes in NOD/Lt mice. J Autoimmun 2:759-776,

1989

8. Atkinson,M, Leiter,EH: The NOD mouse model of insulin

dependent diabetes: As good as it gets? Nature Med 5:601-604,

1999

9. Michels,AW, Eisenbarth,GS: Immune intervention in type 1

diabetes. Semin Immunol 23:214-219, 2011

Group TLR Ligand TLR Frequency of T1D Frequency of Arthritis Gender Bias

1 Poly I:C (primarily HMW) TLR3a Positive (++++)

2 LMW Polu I:C (purified) TLR3a Positive (++++)

3 HMW Poly I:C (purified) TLR3a Positive (+++)

4 Zymosan TLR2 Positive (++)

5 R848 TLR7 Positive (++)

6 CpG oligodeoxynucleotide TLR9 Negative

None ♀ = ♂

Frequency of Diabetes in MAD Rats

Rats were treated with the reagents indicated three times weekly as described in the text. aHMW: High molecular weight poly I:C, which signals through MDA5.

LMW: Low molecular weight poly I:C, which singals via RIG-I.

Group Age at Start of Treatment Frequency (%) of Diabetes Latency To Diabetes OnsetA

1 41-44 days 3/11 (27%) 20, 22, 22 days

2 31-34 days 9/12 (75%) 13, 15, 15, 15, 18, 18, 18, 18, 18 days

3 21-24 days 8/12 (66%) 11, 13, 15, 15, 18, 18, 25, 25 days

Frequency of diabetes in MAD rats treated at different ages

ALatencies are relative to the start of poly I:C treatment on day 1.

Treatment Frequency of Diabetes % Diabetes Latency To Onset (Days)

Vehicle 0/12 0 N/A

Poly I:C 0.5 mg/kg 5/12 42 14,14,21,21,23

Poly I:C 1.0 mg/kg 11/12 92 14,14,14,14,14,14,14,18,21,21,32

Poly I:C 2.0 mg/kg 7/12 58 11,14,14,18,18,25,32

Induction of Diabetes in MAD Rats by Poly I:C

Legend: MAD rats were treated with commercial poly I:C (Sigma, St. Louis, MO) as described in the text.

Polymer length of this product is not specified, but is generally of high molecular weight.

There were no statistically significant differences between genders.

Group Treatment Frequency (% Males) Latency (Males) Frequency (% Females) Latency (Females)

1 Vehicle 0/5 0% N/A 0/5 0% N/A

2 1 mg/kg Sigma Poly I:C 3/5, 60% 16, 23, 28 0/5 0% N/A

3 2.5 mg/kg Sigma Poly I:C 2/5, 40% 12, 16, 39 1/5, 20% 19

4 5 mg/kg Sigma Poly I:C 1/5, 20% 42 2/5, 40% 12, 16

5 1 mg/kg LMW Poly I:C 2/5, 40% 12, 12 1/5, 20% 21

6 2.5 mg/kg LMW Poly I:C 2/5, 40% 19, 28 4/5, 80% 14, 14, 16, 28

7 5 mg/kg LMW Poly I:C 4/5, 80% 12, 23, 23, 40 5/5, 100% 9, 12, 12, 19, 33

8 1 mg/kg HMW Poly I:C 3/5, 60% 14, 23, 40 2/5, 40% 19, 21

9 2.5 mg/kg HMW Poly I:C 4/5, 80% 12, 14, 16, 37 5/5, 100% 12, 14, 14, 19, 37

10 5 mg/kg HMW Poly I:C 5/5, 100% 12, 14, 27, 33, 37 5/5, 100% 12, 16, 19, 23, 37

Induction of T1D in MAD Rats by Poly I:C Polymers of Differing Molecular Weight

Legend: Commercial poly I:C (Sigma) and purified poly I:C of high and low molecular weight (InVivoGen) was

given to MAD rats as described in the text. There were no statistically significant differences between genders.

Treatment Frequency of Diabetes % Diabetes Latency To Onset

Vehicle 0/12 0 N/A

Zymosan 1.0 mg/kg 5/12 42 18,20,25,25,32

Zymosan 2.0 mg/kg 9/12 75 15,20,20,20,20,25,25,29,32

Zymosan 4.0 mg/kg 10/12 83 20,20,20,20,25,25,25,25,29,29

Induction of Diabetes in MAD rats by Different Doses of the TLR2 Agonist Zymosan

Legend: MAD rats were treated with the TLR9 agonist Zymosan at the indicated doses.

There were no statistically significant differences between genders.

Treatment Frequency of Diabetes % Diabetes Latency To Onset

Vehicle 0/12 0 N/A

R848 1.0 mg/kg 3/12 25 16, 18, 28

R848 2.0 mg/kg 5/12 42 18,18,18,21,28

R848 4.0 mg/kg 1/12 8 28

Legend: Purified R848 was injected intraperitoneally into MAD rats at the

Induction of Diabetes in MAD rats by Different Doses of the TLR7 Agonist R848

indicated doses. There were no statistically significant differences between genders.

Treatment Frequency of Diabetes % Diabetes

Vehicle 0/12 0

CpG 1.0 mg/kg 0/12 0

CpG 2.0 mg/kg 0/12 0

CpG 4.0 mg/kg 0/12 0

Induction of Diabetes in MAD rats by the TLR9 Agonist CpG Oligodeoxynucleotide

Legend: MAD rats were injected with the indicated doses of the TLR9 agonist

CpG Oligodeoxynucleotide at the doses indicated. There were no statistically significant

differences between genders.