Treatment of autoimmune diseases in MRL/lpr mice by allogenic bone marrow transplantation plus adult thymus transplantation

Summary

MRL/lpr mice (H-2^k) with Fas gene mutation develop severe autoimmune diseases, and their haematolymphoid cells such as bone marrow and spleen cells showed a low apoptotic activity by irradiation. Therefore, conventional bone marrow transplantation (BMT) cannot be used to treat autoimmune diseases in these mice (chimeric resistance). In the present study, we examine the effects of additional adult thymus transplantation (TT) from the same donor on successful BMT. When the MRL/lpr mice were lethally irradiated (9·5Gy) and reconstituted with 3 × 10⁷ of C57BL/6 mouse (H-2^b) bone marrow cells (BMCs) in conjunction with TT, the mice significantly survived long term and showed a high donor-derived chimerism in comparison with those treated with BMT alone. Interestingly, the numbers of not only donor-derived T cells but also B cells increased significantly in the mice treated with BMT plus TT, even at the early phase of BMT. The number of aberrant CD3⁺B220⁺ cells decreased significantly, and the numbers of lymphocyte subsets were also normalized 4 weeks after the treatment. Finally, the autoimmune diseases in MRL/lpr mice could be cured by BMT with TT. These results indicate that the combination of BMT plus TT can overcome the chimeric resistance and treat the autoimmune diseases in MRL/lpr mice.

Introduction

In recent years, bone marrow transplantation (BMT) has become a valuable strategy for the treatment of haematological disorders (leukaemia, lymphoma, aplastic anaemia), congenital immunodeficiencies, metabolic disorders and autoimmune diseases [1]. Using various animal models we have found that allogenic BMT can be used to treat autoimmune disease, such as insulin-dependent diabetes mellitus, a certain type of non-insulin-dependent diabetes mellitus, systemic lupus erthythematosus, rheumatoid arthritis, chronic pancreatitis and chronic glomerulonephritis, by replacing the pathogenic bone marrow cells (BMCs) with normal BMCs [1–8].

However, allogenic BMT has some problems. Although T cells in allogenic BMCs facilitate the engraftment [9], they often induce graft-versus-host disease (GVHD) [10]. Conversely, if anti-host reaction by donor T cells is low, the primary disease recurs [11]. In addition, the success rate of allogenic BMT is low in elderly patients as they run the risk of several complications, including interstitial pneumonitis, GVHD and systemic infections [12–14]. It is thus extremely important to overcome these limitations of allogenic BMT.

We have developed various BMT methods to resolve these problems. To supply recipients with major histocompatibility complex (MHC)-matched bone marrow (BM) stromal cells, we performed BMT plus bone grafts from the same donor [5]. For elderly hosts with thymic involution, we carried out thymus grafts with BMT [7]. To induce extra-medullary haematopoiesis in the liver we injected the BMCs from the portal veins [15]. Finally, we developed the intrabone marrow (IBM)–BMT (IBM–BMT) method, in which BMCs are injected directly into the BM cavity [16].

MRL/lpr mice have a mutation of the Fas gene that induces apoptosis [17], and the mice show severe autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematosus [18]. They are resistant to conventional BMT (chimeric resistance), and autoimmune diseases recur frequently after conventional BMT [19]. Although lymphocytes of MRL/lpr mice show a low apoptotic activity by irradiation [20], the mechanism underlying the resistance is still unknown. In humans, patients with the Fas gene mutation who develop some autoimmune diseases have been reported as having autoimmune lymphoproliferative syndrome (ALPS) [21].

The thymus is an organ in which T cells can be induced to differentiate from precursor T cells. We have reported previously that fetal thymus transplantation in conjunction with BMT is successful for elderly hosts with thymic involution [7]. However, thymus transplantation has been applied clinically only for patients with DiGeorge syndrome or HIV infection, which elicits hypoplasia of the thymus [22,23]; its effectiveness in the treatment of other intractable diseases has not been examined.

In the present study, we attempt to treat chimeric-resistant and autoimmune prone MRL/lpr mice with allogenic BMT plus adult thymus transplantation (TT) to supply T cells continuously and to prevent rejection. We show here that this method can overcome chimeric resistance and help eradicate the diseases in MRL/lpr mice.

Materials and methods

Mice

Female 3–4-month-old MRL/MP-lpr/lpr mice (MRL/lpr) (H-2^k), age-matched MRL/Mp-+/+ mice (MRL/+) (H-2^k) and 6–8-week-old female C57BL/6 (B6) (H-2^b) mice were obtained from SLC (Shizuoka, Japan) and maintained until use in our animal facilities under specific pathogen-free conditions.

BMT and adult TT

BMCs were collected from the femurs and tibias of 6–8-week-old B6 mice; 3–4-month-old MRL/lpr mice with onset of autoimmune disease (proteinuria > 100 mg/dl and evident lymphadenopathy) or age-matched MRL/⁺ mice were lethally irradiated (9·5Gy) 1 day before BMT. Next day, 1× or 3 × 10⁷ BMCs were injected intravenously into the mice. Adult thymus was removed from the mediastinum of the B6 mice and was grafted simultaneously into muscle of the thigh of some recipients with BMT, as performed clinically [22]. TT alone was also performed in other mice.

Analysis of surface markers and radiation-induced apoptosis in lymphocytes by flow cytometry

Surface markers on lymphocytes from peripheral blood were analysed with three-colour fluorescence staining using fluorescence activated cell sorter (FACScan) (Becton Dickinson, Franklin Lakes, NJ, USA). Fluorescein isothiocyanate (FITC)-conjugated anti-H-2K^b or H-2K^k monoclonal antibodies (MoAbs) (Pharmingen, San Diego, CA, USA) were used to determine chimerism. FITC-, phycoerythrin (PE-) or biotin-conjugated CD4, CD8, CD3, CD45 or B220 (Becton Dickinson or Pharmingen) were used to analyse lymphocyte subsets. Avidin-Cy5 (Dako, Kyoto, Japan) was used as the third colour in the avidin/biotin system. Radiation-induced apoptosis in BMCs and spleen cells was examined by annexin staining using an apoptosis detection kit (Pharmingen). The cells from MRL/lpr and MRL/⁺ mice were suspended in 1 × 10⁶ cells/ml of RPMI-1640 medium supplemented with 10% heat-inactivated fetal low serum (FCS). The cells were irradiated with 10 Gy and were incubated at 37°C with 5% CO₂ overnight (16 h), which is similar to the BMT protocol. Non-irradiated cells were also prepared as a control. The cells were stained by annexin–FITC [24] and CD45, CD4, CD8, CD3-PE or/and B220-Cy5 according to the manufacturer's protocol. The apoptotic cells were identified as annexin-positive cells gated in CD45⁺ cells for BMCs, including the haematopoetic stem cells [25], or in CD4⁺, CD8⁺, CD3⁺ or CD3^– and B220⁺ cells for spleen cells at lymphocyte lesion by forward scatter (FSC)/side scatter (SSC).

Measurement of autoantibodies in sera

IgG- and IgM-type anti-ssDNA antibodies and IgG-type rheumatoid factor in the sera from the chimeric and the non-treated mice were measured using a standard enzyme-linked immunosorbent assay (ELISA), as described previously [26]. The concentration of autoantibodies was measured by absorbance at 405 nm developing the phosphatase substrate (Sigma-Aldrich, St Louis, MO, USA).

Pathological findings

The liver, small intestine, thymus and kidney from the chimeric mice were fixed in 10% formaldehyde solution and embedded in paraffin. Four µm-thick serial tissue sections were prepared and stained with haematoxylin and eosin (H&E). The histology was examined under the microscope. For study of glomerular immune-complex deposits the specimens of kidney were frozen in dry ice/acetone, and 3-µm sections were stained with FITC-conjugated anti-mouse IgG (Medical and Biomedical Laboratories, Nagoya, Japan). The deposition of the immunoglobulin in glomeruli was evaluated under the fluorescence microscope.

Proteinuria

The level of proteinuria was measured using testing papers (–; 0 mg/dl, + 30 mg/dl, + + 100 mg/dl and + + + 300 mg/dl) (Pretest 7aII; Wako, Osaka, Japan).

Statistical analysis

Non-parametric analyses (Mann–Whitney U-test and log rank-test) were performed using StatView software (Abacus Concepts, Berkley, CA, USA). Values of P < 0·05 were considered statistically significant.

Results

Radiation-induced apoptosis of BMCs and spleen cells from MRL/lpr and MRL/+ mice

First, we examined radiation-induced apoptosis not only in CD45⁺ BMCs, which contain haemopoietic stem cells [25] but also in lymphocyte subsets in the spleen from MRL/lpr and the wild counterpart of MRL/+ mice to clarify the mechanisms underlying chimeric resistance in vitro. To approximate more closely practical BMT conditions, the radiation dose was set at 10 Gy and analysis was performed the following day (overnight culture). The levels of apoptosis by irradiation in both CD45⁺ BMCs and the lymphocyte subsets (CD4 and CD8 T cells and the CD3^–B220⁺ conventional B cells in the spleen) in MRL/lpr were much lower than those in MRL/+ mice (Fig. 1). The CD3⁺B220⁺ aberrant lpr-T cells, which are detected in the spleen but not in BMCs from MRL/lpr mice, also showed apoptosis, although the level was lower than the conventional lymphocytes. This finding may explain one of the mechanisms of chimeric resistance in MRL/lpr mice, suggesting that more BMCs, as in a mega-dose BMT [27], and/or the repeated supplementation of donor lymphocytes, as in extensive donor lymphocyte infusion (DLI), are required for successful BMT.

Figure 1. Radiation-induced apoptosis in bone marrow cells (BMCs) and spleen cells from MRL/lpr and MRL/+ mice. Ten Gy-irradiated or non-irradiated BMCs (a) and spleen cells (b) from 3–4-month-old MRL/lpr and age-matched MRL/+ mice were cultured for 16 h (overnight). These cells were analysed by flow cytometry for annexin⁺ apoptotic cells gated in CD45⁺ BMCs and in CD4⁺, CD8⁺, CD3^–B220⁺ and CD3⁺B220⁺ spleen cells at lymphocyte lesion in fluorescence activated cell sorter (FACS) profile. Representative data from three independent experiments are shown. Not done: n.d.

Survival rates of MRL/lpr and MRL/⁺ mice with BMT plus TT

We compared several BMT methods using MRL/lpr and MRL/+ mice (Fig. 2). While lethally irradiated (9·5Gy) MRL/+ mice reconstituted with 1 × 10⁷ B6 BMCs (as the conventional dose of BMT) survived for a long time (80% rate survival at 5 months after BMT), MRL/lpr mice treated with the same BMT method showed a significantly low survival rate (all died within 3 weeks) (P = 0·0001) (Fig. 2a). BMT with 3 × 10⁷ BMCs only slightly improved the survival rate in MRL/lpr mice (30% survival rate at 3 months after BMT; not significant in comparison to BMT with 1 × 10⁷ BMCs).

Figure 2. Survival rate in MRL/lpr and MRL/+ mice by bone marrow transplantation (BMT) with or without thymus transplantation (TT). Lethally irradiated 3–4-month-old MRL/lpr and age-matched MRL/+ mice that received 1× or 3 × 10⁷ B6 bone marrow cells (BMCs) without (a) or with TT (b) and the survival rates (%) are shown. *P = 0·0001 compared with MRL/lpr by1 × 10⁷ BMC, †P = 0·016 compared with MRL/lpr by 3 × 10⁷ BMC, ‡P = 0·052 compared with MRL/lpr by 1 × 10⁷ BMC. §P < 0·0001 compared with MRL/lpr by 3 × 10⁷ BMC + TT.

We next attempted to carry out TT in the mice in conjunction with BMT from the same donor as an additional means to recruit donor-derived T cells (Fig. 2b). Interestingly, MRL/lpr mice with 3 × 10⁷ BMCs plus TT showed a significantly longer survival rate (about 70%) at 5 months after BMT than those without TT (P = 0·016, compared with absence of TT). There was a slightly significant difference in the survival rate of MRL/lpr mice treated with 1 × 10⁷ BMCs plus TT (P = 0·052, compared with absence of TT). As expected, lethally irradiated MRL/lpr with TT alone (without BMT) died very early due to graft failure (P < 0·0001 compared to the presence of 3 × 10⁷ BMCs in MRL/lpr mice).

Chimeric analyses of MRL/lpr and MRL/+ mice with BMT plus TT

Chimeric analyses were performed using peripheral blood from MRL/lpr and MRL/+ mice with or without TT (Fig. 3). In 9·5 Gy-irradiated MRL/+ mice reconstituted with 1 × 10⁷ B6 BMCs, donor-derived cells gradually increased, whereas in 9·5 Gy-irradiated MRL/lpr mice reconstituted with 1 × 10⁷ B6 of BMCs the cells were significantly rejected in the early phase after BMT (58·8 ± 5·2% in MRL/+versus 5·4 ± 0·6% in MRL/lpr mice at 1 week; P = 0·008, 82·8 ± 10·6% in MRL/+versus 2·6 ± 0·5% in MRL/lpr mice at 2 weeks; P = 0·016, respectively) (Fig. 3a). Although MRL/lpr mice reconstituted with 3 × 10⁷ BMCs showed a slightly higher chimerism than those reconstituted with 1 × 10⁷ BMCs at 1 week, the cells were also rejected rapidly at 2 weeks (15·4 ± 5·5% at 1 week, 7·4 ± 4·3% at 2 weeks). In contrast, TT in conjunction with 3 × 10⁷ BMC resulted in significant reconstitution in MRL/lpr mice in comparison with the absence of TT at 1 and 2 weeks (33·6 ± 4·8% at 1 week; P = 0·004, 94·8 ± 2·6% at 2 weeks; P = 0·008) (Fig. 3b). Although a slightly higher reconstitution rate was also observed in MRL/lpr with 1 × 10⁷ BMC plus TT (21·1 ± 7·7% at 1 week, 8·2 ± 7·6% at 2 weeks) than with 1 × 10⁷ BMC alone, there was no statistically significant difference in reconstituting rates: the donor cells were rejected by 2 weeks. The MRL/lpr mice with 1 × 10⁷ BMC plus TT and the mice with TT alone also significantly decreased donor-derived cells in contrast to those with 3 × 10⁷ BMC plus TT at 1 and 2 weeks (P = 0·038 at 1 week, P = 0·043 at 2 weeks by 1 × 10⁷ BMC plus TT: 16·3 ± 6·2% at 1 week, P = 0·034; 3·1 ± 2·3% at 2 weeks, P = 0·036 by TT alone, respectively).

Figure 3. Chimerism in MRL/lpr and MRL/+ mice by bone marrow transplantation (BMT) with or without thymus transplantation (TT). Lethally irradiated 3–4-month-old MRL/lpr and age-matched MRL/+ mice received 1× or 3 × 10⁷ B6 bone marrow cells (BMCs) without (a) or with TT (b). Donor derived (H-2K^b+) cells were evaluated in peripheral blood from these chimeric mice. Open circle, MRL/+ mice that received 1 × 10⁷ B6 BMCs; open square, MRL/lpr mice that received 3 × 10⁷ B6 BMCs; closed square, MRL/lpr mice that received 3 × 10⁷ B6 BMCs (a); open square, MRL/lpr mice that received 1 × 10⁷ B6 BMCs plus TT; closed square, MRL/lpr mice that received 3 × 10⁷ B6 BMCs plus TT; open triangle, MRL/lpr mice that received TT alone (B). *P = 0·008 and †P = 0·016 compared between MRL/lpr and MRL/+ mice that received 1 × 10⁷ BMCs 1 or 2 weeks after BMT. ‡P = 0·004 and §P = 0·008 compared with MRL/lpr mice that received 3 × 10⁷ BMCs alone at 1 or 2 weeks after BMT. §**MRL/lpr mice that received 3 × 10⁷ BMCs plus TT compared with 1 × 10⁷ BMCs plus TT (P = 0·038 at 1 week and P = 0·043 at 2 weeks) and with TT alone (P = 0·034 at 1 week and P = 0·036 at 1 week) after BMT; n = 5, 4, 4 and 4 in MRL/⁺ mice received 1 × 10⁷ BMCs, n = 5 and 5 in MRL/lpr mice received 1 × 10⁷ BMCs, n = 14 and 5 in MRL/lpr mice received 3 × 10⁷ BMCs, n = 9, 5, 6 and 6 in MRL/lpr mice that received 3 × 10⁷ BMCs plus TT, n = 11 and 6 in MRL/lpr mice that received 1 × 10⁷ BMCs plus TT, n = 4 and 3 in MRL/lpr mice that received TT alone at 1 week, 2 weeks, 1 month and 3 months after BMT.

Lymphocyte subsets in MRL/lpr and MRL/⁺ mice with BMT plus TT

We next investigated lymphocyte subsets in peripheral blood from these chimeric mice (Table 1). Interestingly, not only the numbers of both CD4 and CD8 T cell subsets but also B cells were significantly elevated in MRL/lpr mice treated with BMT plus TT 1 and 2 weeks after the treatment. In addition, the number of host-derived aberrant B220⁺CD3⁺ cells were reduced significantly by BMT plus TT in contrast to the absence of TT in MRL/lpr mice. At 4 weeks, the subsets reached normal levels and remained at this level for more than 3 months (data not shown).

Table 1.  Donor-derived lymphocyte subsets in peripheral blood from MRL/lpr mice by bone marrow transplantation with or without thymus transplantation.

Mice	Weeks after BMT	TT	n	% of donor-derived (H-2K^b^+)			B220^+ CD3^+*
				CD4	CD8	B220^+ CD3^-
Lymphocytes from peripheral blood were analysed using 3–4-month-old MRL/lpr mice transferred with 3 × 10^7 B6 BMCs with or without TT by flow cytometry. Data shown are mean ± s.e. *B220 CD3^+ cells were host-derived. ^aP = 0·016, ^bP = 0·035, ^cP = 0·003, ^dP = 0·004, ^eP = 0·001, ^fP = 0·015, ^gP = 0·002, ^hP = 0·001 compared with absence of TT.
BMT-treated MRL/lpr	1	+	9	7·0 ± 1·7^a	16·5 ± 5·3^b	15·4 ± 2·8^c	4·9 ± 1·7^d
		–	5	3·7 ± 1·7	9·5 ± 5·2	4·2 ± 2·4	16·4 ± 5·1
	2	+	5	13·8 ± 1·7^e	34·6 ± 4·2	36·2 ± 7·1^g	1·1 ± 0·2^h
		–	5	1·2 ± 0·3	7·2 ± 5·3	3·4 ± 1·4	9·7 ± 4·6
	4	+	4	13·5 ± 2·6	4·9 ± 1·4	74·7 ± 0·4	0·9 ± 0·4
		–		n.d.	n.d.	n.d.	n.d.
Non-treated MRl/lpr	3-4 months old		5	14·6 ± 2·2	14·0 ± 1·7	21·2 ± 1·0	37·2 ± 1·0
B6	8 weeks old		4	19·1 ± 1·8	10·9 ± 1·3	56·9 ± 7·0	0·1 ± 0·1

Histology and FACS profile in TT

The B6 adult thymi were transplanted into muscle of the thigh of MRL/lpr mice with 3 × 10⁷ BMCs from B6 mice (Fig. 4a). Three months later both the cortex and medullary areas were observed clearly in the transplanted thymus. FACS analyses also showed the normal differentiation of CD4^– 8^–, CD4⁺ 8⁺, CD4⁺ 8^– and CD4^– 8⁺ T cell subsets (as seen in the normal intact thymus) (Fig. 4b).

Figure 4. Histological findings of transplanted thymus and the fluorescence activated cell sorter (FACS) profile in MRL/lpr mice that received bone marrow transplantation (BMT) with thymus transplantation (TT). Histology (H&E staining, × 100) (a) and the FACS profile by CD4 and CD8 double-staining (b) in the transplanted thymus from muscle of the thigh in MRL/lpr mice that received 3 × 10⁷ B6 bone marrow cells (BMCs) with TT 3 months after BMT. Black arrow, cortex area; white arrow, medullary area.

Treatment of autoimmune diseases in MRL/lpr mice by BMT plus TT

Finally, we examined the effects of BMT plus TT on autoimmune diseases in MRL/lpr mice. The levels of autoantibodies (IgG- and IgM-type anti-ssDNA antibodies and IgG-type rheumatoid factors) in sera were reduced significantly, comparable to those of donor B6 mice after BMT plus TT (Fig. 5a). In histology, irregular mesangium and capillary walls with massive IgG deposition were observed in the glomeruli from non-treated 3-month-old MRL/lpr mice (Fig. 5b). However, 3 months after the treatment there were no pathological findings (with reduction in IgG deposition) in the kidney. In addition the levels of proteinurea, which were elavated significantly in MRLlpr mice compared with the donor B6 mice (166·7 ± 42·2 mg/dl in 3-month-old MRL/lpr versus 30·2 ± 0·4 mg/dl mg/dl in 8-week-old B6 mice, P = 0·004), were also reduced after BMT plus TT (53·3 ± 14·8 mg/dl; P = 0·026) (Fig. 5c).

Figure 5. Treatment of autoimmune diseases in MRL/lpr by bone marrow transplantation (BMT) (3 × 10⁷ B6 BMCs) plus thymus transplantation (TT). Autoantibodies (IgG type ss-DNA and RF, and IgM type ss-DNA antibodies) were evaluated in serum from 8-week-old B6, 3-month-old MRL/lpr and the MRL/lpr mice 3 months after BMT plus TT by enzyme-linked immunosorbent assay (ELISA) (a). Histology (H&E staining × 200, a; upper panel) and IgG deposition (immunofluorescent staining × 200, b; lower panel) in glomeruli from the 3-month-old MRL/lpr and the mice 3 months after BMT plus TT (b). Although swollen and irregular capillary walls in glomeruli with massive IgG deposition were observed in the non-treated 3-month-old MRL/lpr mice, the walls were almost normalized with reduction of IgG deposition after BMT plus TT. Level of proteinuria from 8-week-old B6, 3-month-old MRL/lpr and the MRL/lpr mice 3 months after BMT plus TT (C). *P = 0·0007 in anti-ssDNA (IgG), †P = 0·0007 in anti-ssDNA (IgM), ‡P = 0·0007 in anti-RF (IgG) from sera of 8-week-old B6 mice compared with those from 3-month-old MRL/lpr mice; §P = 0·0004 in anti-ssDNA (IgG), ¶P = 0·003 in anti-ssDNA (IgM), **P = 0·0002 in anti-RF (IgG) from sera of 3-month-old MRL/lpr compared with those from the mice 3 months after BMT plus TT. ^aP = 0·004 in proteinuria from 8-week-old B6 mice compared with those from 3-month-old MRL/lpr mice; ^bP = 0·026 in proteinuria from 3-month-old MRL/lpr compared with those from the mice 3 months after BMT plus TT; n = 8 in 8-week-old B6, n = 6 in 3-month-old MRL/lpr and n = 8 in MRL/lpr treated with BMT plus TT (a). n = 5 in 8-week-old B6, n = 6 in 3-month-old MRL/lpr and n = 6 in MRL/lpr treated with BMT plus TT (c).

Discussion

In the present study we have examined how successful BMT can be achieved in chimeric-resistant MRL/lpr mice with radioresistant haematopoietic cells. While conventional BMT alone was ineffective, the addition of TT with BMT (3 × 10⁷ BMCs) enhanced significantly reconstitution of MRL/lpr mice with donor B6 cells and the mice showed significantly prolonged survival. Analyses of lymphocyte subsets demonstrated that the numbers of both donor-derived T cells and B cells increased significantly, while the number of aberrant B220⁺ CD3⁺ cells decreased in comparison with the absence of TT. Finally, the chimeric resistance could be overcome, which resulted in the amelioration of autoimmune diseases in MRL/lpr mice. These findings indicate that BMT with TT facilitates the engraftment of donor-derived cells in chimeric-resistant hosts.

In the present study we analysed first the mechanisms underlying chimeric resistance in MRL/lpr mice. The mice showed a low apoptotic activity in both BMCs and spleen cells by irradiation (Fig. 1), as reported previously in spleen cells [20], although a different apoptotic assay was used. These findings indicate that most haematolymphoid cells in MRL/lpr mice have strong resistance to irradiation, which results in chimeric resistance. Therefore, in contrast to conventional BMT, transplantation of either further numbers of BMCs (mega-dose BMT [27]) and/or donor-derived lymphocytes (as in DLI) may be necessary to overcome host-resistant cells.

We compared various BMT methods using MRL/lpr and MRL/+ mice. While the MRL/+ mice were reconstituted and able to survive for a long time with the conventional BMT (1 × 10⁷ BMCs), MRL/lpr mice treated with 1 × 10⁷ BMCs died very early due to graft failure; the MRL/lpr mice treated with 3 × 10⁷ BMCs showed only a slightly longer survival rate. However, TT from the same donor with 3 × 10⁷ BMCs showed a significantly high reconstitution rate and long survival, although 1 × 10⁷ BMCs plus TT did not produce a satisfactory reconstitution, indicating that BMT with a high number of BMCs as well as TT are required for reconstitution in MRL/lpr mice.

In analyses of lymphocyte subsets, the numbers of both donor-type CD4 and CD8 T cell subsets in mice treated with BMT plus TT were significantly higher than those with BMT alone in the early phase after BMT in MRL/lpr. Interestingly, the number of donor-type B cells in BMT plus TT were significantly elevated. The elevated T cells or transplanted thymus itself, including thymic epithelial cells (TECs), may produce B cell proliferative cytokines, such as interleukin (IL)-4, IL-5, IL-6 or IL-7 [28]. Importantly, the numbers of aberrant CD3⁺B220⁺ T cells, which are related to the pathogenesis of autoimmune diseases in lpr mice [29], were also reduced significantly in mice treated with BMT plus TT. These findings indicate that TT with BMT facilitates the normal repopulation of lymphocytes, and also leads to a significantly high reconstitution rate in comparison with the absence of TT in chimeric-resistant MRL/lpr mice.

The MRL/lpr mice treated with BMT and TT should have about four types of T cells: (i) radiation (10 Gy)-resistant MRL/lpr-derived conventional T cells and aberrant CD3⁺B220⁺ T cells, (ii) T cells contained in the transplanted B6 thymus, (iii) B6 mouse haematopioietic stem cell-derived T cells, which have been educated in the transplanted B6 thymus and (iv) B6 mouse haematopoietic stem cell-derived T cells, which have been educated in the host MRL/lpr thymus. T cells of (i) and (ii) or (iii) should react with each other. However, T cells of (iv) should be tolerant to both the MRL/lpr and B6 mouse MHC determinants. If T cells of (i) are dominant, rejection occurs, as seen in MRL/lpr mice treated with BMT alone. In contrast, if T cells of (ii) and (iii) are dominant, strong GVHD develops, as seen in BMT plus DLI [30,31]. However, it remains to be clarified whether the T cells of (iii) can really kill the host (MRL/lpr)-derived cells or whether they are tolerant to MRL/lpt (H-2^k) determinants due to the presence of host (MRL/lpr)-derived radio-resistant cells (such as macrophages and DCs) that have entered the transplanted B6 thymus. Thus, the outcomes from this strategy (BMT + TT) seem to be dependent on the balance of these four types of T cells. Although we have thus far discussed MHC-mediated cell interaction in the mechanisms of chimeric resistance, we would like to discuss next the role of the Fas–FasL interaction to explain the mechanisms of chimeric resistance.

The MRL/lpr mice have abundant FasL, especially in aberrant CD3⁺B220⁺ cells [32]. The cells should also play an important role in the chimeric resistance by inducing apoptotsis on the donor cells expressing Fas [33]. In addition, the host radioresistant cells have few Fas [34], leading to apoptosis resistance with FasL from the donor cells. However, we were able to overcome chimeric resistance by BMT plus TT. One reason may be that the newly developed allogeneic T cells by TT are naive T cells, which show less Fas expression and more resistance to apoptosis than the activated or memory T cells with their high Fas expression [35,36]. BMT plus TT may induce early and continuous supplementation of such donor-naive T cells. In addition, although FasL-mediated apoptosis is less effective, other cytotoxic molecules such as perforin, granzyme, tumour necrosis factor (TNF)-α or TNF-related ligand (TRAIL) may be involved in the mechanisms to overcome chimeric resistance.

The serum levels of autoantibodies, IgG deposition in the glumeruli, and the degree of proteinuria were reduced significantly in MRL/lpr mice that had received BMT plus TT. The lymphadenopathy also disappeared, with a reduction in the number of aberrant CD3⁺B220⁺ T cells (data not shown).

As a strategy for supplementing donor lymphocytes, DLI is now used around the world with BMT [30,31]. Although we did not compare these two methods of BMT plus TT and DLI in the present study, the following differences may exist: (i) the low frequency of acute GVHD is due to a small number of mature T cells supplemented at the transplantation, (ii) naive T cells are exclusively supplemented, [37] (iii) additional DLI is not necessary due to continuous supplementation of T cells from the transplanted thymus and (iv) the transplantation of TECs produces significant cytokines for lymphocyte development and regulation [28]. In the development of GVHD, we have found recently that BMT plus TT induces GVHD much less than BMT with DLI in normal mouse combinations (manuscript in preparation).

Overall, BMT with TT might facilitate haematolymphoid reconstitution and overcome chimeric resistance. From the viewpoint of clinical application, severe autoimmune diseases or hereditary diseases such as ALPS with Fas mutation are possible candidates. In addition, this combination might also apply to elderly patients or patients with sublethal irradiation or with malignant tumours.

Acknowledgements

The authors wish to thank Ms Y. Tokuyama, Ms R. Hayashi, and Ms A. Kitajima for technical assistance and Ms K. Ando for secretarial assistance. This work was supported by the research grant C from Kansai Medical University.