1,25-Dihydroxyvitamin D3 restores sensitivity to cyclophosphamide-induced apoptosis in non-obese diabetic (NOD) mice and protects against diabetes

Abstract

The activated form of vitamin D, 1,25(OH)2D3, and its analogues can prevent type I diabetes in NOD mice. Protection is achieved without signs of systemic immunosuppression and is associated with a restoration of the defective immune regulator system of the NOD mice. The aim of the present study was to investigate whether this restoration of regulator cell function is the only mechanism in the prevention of diabetes by 1,25(OH)2D3. We tested therefore if 1,25(OH)2D3 could prevent cyclophosphamide-induced diabetes, since diabetes occurring after cyclophosphamide injection is believed to be due to an elimination of suppresser cells. NOD mice treated with 1,25(OH)2D3 (5 microg/kg every 2 days) from the time of weaning were clearly protected against diabetes induced by cyclophosphamide (200 mg/kg body wt at 70 days old) (2/12 (17%) versus 36/53 (68%) in control mice, P < 0.005). By co-transfer experiments it was demonstrated that cyclophosphamide had indeed eliminated the suppresser cells present in 1,25(OH)2D3-treated mice. Since cyclophosphamide injection did not break the protection offered by 1,25(OH)2D3, it was clear that diabetogenic effector cells were affected by 1,25(OH)2D3 treatment as well. This was confirmed by the finding that splenocytes from 1,25(OH)2D3-treated mice were less capable of transferring diabetes in young, irradiated NOD mice, and by the demonstration of lower Th1 cytokine levels in the pancreases of 1,25(OH)2D3-treated, cyclophosphamide-injected mice. This better elimination of effector cells in 1,25(OH)2D3-treated mice could be explained by a restoration of the sensitivity to cyclophosphamide-induced apoptosis in both thymocytes and splenocytes, in normally apoptosis-resistant NOD mice. Altogether, these data indicate that the protection against diabetes offered by 1,25(OH)2D3 may be independent of the presence of suppresser cells, and may involve increased apoptosis of Th1 autoimmune effector cells.

Fig. 1

Incidence of cyclophosphamide-induced diabetes and insulitis. A significant protection against diabetes was present in all groups receiving 1,25(OH)₂D₃ from the time of weaning (□). While 36 of 53 control mice developed diabetes, only three of 16 mice treated with 1,25(OH)₂D₃ from day 21 until day 50 (group 1) and two of 12 mice treated from day 21 until day 69 (group 2) developed diabetes. The incidence of insulitis was investigated as well (▪): five of 12 mice in group 2 developed insulitis in contrast to the control group, where 42 of the 53 animals had insulitis (P < 0.01). In group 1, 10 of 16 animals developed insulitis (NS compared with controls). ***P < 0.005; **P < 0.01.

Fig. 2

Incidence of diabetes after cell transfer. Mice (6–8 weeks, irradiated with 7.50 Gy) receiving only splenocytes from overtly diabetic NOD mice developed diabetes (100%, 6/6, □). A clear protective effect was seen when a co-transfer was performed with 20 × 10⁶ splenocytes from 1,25(OH)₂D₃-treated (day 21 until day 69) mice (○, 4/8, P < 0.05, *). No protective effect was seen when a co-transfer was performed with 20 × 10⁶ splenocytes from control or 1,25(OH)₂D₃-treated mice that received cyclophosphamide (▪ (8/8) and • (11/11), respectively).

Fig. 3

Pancreatic mRNA levels of IL-10, IL-12, IFN-γ and transforming growth factor-beta (TGF-β) before and 2 and 9 days after cyclophosphamide (CY) treatment both in control (Co, Co2d, Co9d) and 1,25(OH)₂D₃-treated mice (D₃, D₃2d, D₃9d). Cytokine mRNA levels of IL-12 and IFN-γ were lower in mice treated with 1,25(OH)₂D₃ compared with control mice (Fig. 3a,b). This difference became significant 9 days after CY. No significant differences were seen for IL-10 and TGF-β (Fig. 3c,d). Cytokine mRNA levels are expressed relative to the β-actin polymerase chain reaction (PCR) product amplified from the same sample ((cytokine copies/β-actin copies) × 10 000). Mean ± s.e.m. values are shown for each group. *P < 0.05.

Fig. 4

A mean of 20 mice were analysed for cell number in spleen and thymus. Cyclophosphamide (CY) was injected intraperitoneally 16 h before analysis. The percentage of remaining cells compared with the untreated controls is indicated (a). Splenocytes and thymocytes were harvested 16 h after CY injection and stained with the terminal deoxynucleotidyl transferase (TDT)-mediated FITC-dUTP nick end labelling (TUNEL) reaction. Apoptosis is expressed as percentage TUNEL-positive cells (b). *P < 0.05 versus C57Bl/6; **P < 0.05 versus NOD; ***P < 0.001 versus NOD; ****P < 0.0001 versus C57Bl/6; *****P < 0.00001 versus NOD.

Fig. 5

Flow cytometry of representative cyclophosphamide (CY)-treated mice. The upper histograms show apoptosis levels in splenocytes (a) from C57Bl/6 mice (left), NOD control (centre) and NOD mice treated with 1,25(OH)₂D₃ (right). The lower histograms show apoptosis levels in thymocytes (b) from C57Bl/6 mice (left), NOD control (centre) and NOD mice treated with 1,25(OH)₂D₃ (right). Apoptotic cells are detected in the FL1 channel (FITC-dUTP). The histograms shown are representative of separate experiments (n ≥ 8).