Lymphotoxin beta receptor (Lt betaR): dual roles in demyelination and remyelination and successful therapeutic intervention using Lt betaR-Ig protein

Abstract

Inflammation mediated by macrophages is increasingly found to play a central role in diseases and disorders that affect a myriad of organs, prominent among these are diseases of the CNS. The neurotoxicant-induced, cuprizone model of demyelination is ideally suited for the analysis of inflammatory events. Demyelination on exposure to cuprizone is accompanied by predictable microglial activation and astrogliosis, and, after cuprizone withdrawal, this activation reproducibly diminishes during remyelination. This study demonstrates enhanced expression of lymphotoxin beta receptor (Lt betaR) during the demyelination phase of this model, and Lt betaR is found in areas enriched with microglial and astroglial cells. Deletion of the Lt betaR gene (Lt betaR-/-) resulted in a significant delay in demyelination but also a slight delay in remyelination. Inhibition of Lt betaR signaling by an Lt betaR-Ig fusion decoy protein successfully delayed demyelination in wild-type mice. Unexpectedly, this Lt betaR-Ig decoy protein dramatically accelerated the rate of remyelination, even after the maximal pathological disease state had been reached. This strongly indicates the beneficial role of Lt betaR-Ig in the delay of demyelination and the acceleration of remyelination. The discrepancy between remyelination rates in these systems could be attributed to developmental abnormalities in the immune systems of Lt betaR-/- mice. These findings bode well for the use of an inhibitory Lt betaR-Ig as a candidate biological therapy in demyelinating disorders, because it is beneficial during both demyelination and remyelination.

Figure 1.

Upregulation of LtβR during demyelination and inflammation. A, Upregulation of LtβR during demyelination and inflammation. Quantitative real-time RT-PCR analysis demonstrates an upregulation of LtβR mRNA expression in wild-type mice during cuprizone treatment (through week 5) and a decline to baseline levels during remyelination (weeks 7–10). B, Localization of LtβR mRNA by ISH during cuprizone treatment. At the level of autoradiogram analysis, LtβR mRNA expression is not detected by in situ hybridization in brains from untreated wild-type mice (i). Induction of LtβR mRNA is weakly detected within the corpus callosum of mice treated for 3 weeks (ii), and robust induction of LtβR is seen in the same regions in brains from mice treated for 5 weeks (iii). C, Microscopic examination of the same brain sections shows LtβR expression is highest in regions and time points with a large accumulation of microglia/macrophages. In i–vi, nuclei are visualized by hematoxylin (blue) and LtβR expression by ³³P-labeled riboprobes (dark grains within the emulsion). In i–iii, microglia, blood vessels, and macrophages are visualized with tomato lectin (brown), whereas in iv–vi, astrocytes are visualized with GFAP antibodies (brown). In all panels, the focal plane is optimized to visualize the riboprobe-induced grains within the emulsion.

Figure 2.

PCR analysis of LIGHT during cuprizone treatment. A, RT-PCR for the ligand LIGHT was performed on cDNA from untreated and cuprizone-treated wild-type mouse brain. As controls, untreated thymus and spleen were analyzed. Representative results show that very low levels of LIGHT mRNA exist in the untreated and cuprizone-treated brain. B, RT-PCR on wild-type (WT) and LtβR^−/− untreated (untrt) and treated brain cDNA demonstrate that the lack of LtβR has no effect on LIGHT RNA levels.

Figure 3.

Time course of demyelination and remyelination in LtβR^−/− and wild-type mice. A, LFB–PAS-stained paraffin sections were graded on a scale from 0 (normal myelination) to 3 (complete demyelination) by three double-blinded investigators. Each circle represents an individual mouse: open circles, C57BL/6 wild-type (Wt); filled circles, LtβR^−/− mice. Horizontal lines indicate the median score of each group. Significant differences in demyelination were seen between wild-type and LtβR^−/− mice at 3 weeks (p < 0.02), 3.5 weeks (p < 0.01), and 4 weeks (p < 0.001). Significant differences during remyelination were detected at 7 weeks (p < 0.001) and 10 weeks (p < 0.02). B, Representative LFB–PAS pictures at 4 weeks of treatment demonstrate the delay in demyelination in LtβR^−/− mice (ii) compared with wild-type mice (i). Higher-magnification images of the corpus callosum of LtβR^−/− (iv) and wild-type (iii) mice demonstrate the lack of myelinated fibers and increased inflammation in wild-type mice during demyelination.

Figure 4.

Delayed loss of mature oligodendrocytes during demyelination in LtβR^−/− mice. A, Mature oligodendrocytes were detected at midline corpus callosum of wild-type and LtβR^−/− brains by GSTπ immunohistochemistry. More GSTπ⁺ cells were found in LtβR^−/− mice compared with wild-type mice at 3 weeks (p = 0.09) (wild-type, gray bars; LtβR^−/−, black bars). Significantly more GSTπ⁺ cells were found in LtβR^−/− mice at 3.5 weeks (p < 0.03), although no differences in oligodendrocytes were found at 4 and 5 weeks of cuprizone treatment. After the removal of cuprizone, no differences in oligodendrocyte repopulation of the corpus callosum were observed between wild-type and LtβR^−/− mice. B, Representative pictures of GSTπ⁺ cells from wild-type and LtβR^−/− mice at 3 and 3.5 weeks of cuprizone treatment. C, Microglial/macrophage accumulation at midline corpus callosum is unaffected by LtβR. Microglia/macrophages were detected by RCA-1 lectin staining of paraffin sections from wild-type and LtβR^−/− mice during demyelination and remyelination time points. No significant difference in numbers of RCA-1⁺ cells was observed between wild-type and LtβR^−/− mice at any time point.

Figure 5.

Therapeutic inhibition of LtβR significantly delays demyelination. A, C57BL/6 mice were treated with 0.2% cuprizone for 3.5 weeks. Mice received weekly injections (indicated by small arrows) of either LtβR–hIg or control human Ig beginning at day −1 of cuprizone treatment. B, Significant differences in demyelination were observed between LtβR–hIg-treated and control human Ig-treated mice after 3.5 weeks of cuprizone treatment (p < 0.02). LFB–PAS-stained paraffin sections were graded on a scale of 0 (normal myelination) to 3 (complete demyelination) by three double-blinded investigators. Each circle represents an individual mouse: open circles, control human Ig-treated mice; filled circles, LtβR–hIg-treated mice. Horizontal lines indicate the median score of each group. C, Representative LFB–PAS pictures at 3.5 weeks of treatment demonstrate the delay in demyelination in control human Ig-treated mice (i) compared with LtβR–hIg-treated mice (ii). High-magnification images of the corpus callosum of control human Ig-treated mice (iii) and LtβR–hIg-treated mice (iv) demonstrate the lack of myelinated fibers and increased inflammation in control human Ig-treated mice during demyelination. Immunohistochemistry for MBP confirmed the presence of more myelinated fibers in LtβR–hIg-treated mice (vi) compared with control human Ig-treated mice (v).

Figure 6.

Therapeutic inhibition of LtβR significantly enhances remyelination in a posttreatment paradigm. A, C57BL/6 mice were treated with 0.2% cuprizone for 6 weeks and allowed to remyelinate for 4 weeks before they were killed. Mice received weekly injections of either LtβR–mIg or control mouse Ig beginning 5 weeks plus 2 d after the start of cuprizone treatment (the approximate height of demyelination), as indicated by small arrows. B, Remyelination was enhanced on inhibition of LtβR signaling. LFB–PAS-stained paraffin sections were graded on a scale of 0 (normal myelination) to 3 (complete demyelination) by three double-blinded investigators. Each circle represents an individual mouse: open circles, control mouse Ig-treated mice; filled circles, LtβR–mIg-treated mice. Horizontal lines indicate the median score of each group. Significant enhancement of remyelination was observed in LtβR–mIg-treated mice compared with mice treated with control mouse Ig (p < 0.007). C, Representative LFB–PAS pictures at 10 weeks demonstrate the enhanced remyelination in LtβR–mIg-treated mice (ii) compared with control mouse Ig-treated mice (i). High-magnification images of the corpus callosum of LtβR–mIg-treated (iv) and control mouse Ig-treated (iii) mice demonstrate the lack of myelinated fibers and greater inflammation in control mouse Ig-treated mice during remyelination. Immunohistochemistry for MBP confirmed enhanced remyelination in LtβR–mIg-treated mice (vi) compared with control mouse Ig-treated mice (v). Representative GSTπ-immunostained images of the corpus callosum demonstrating more oligodendrocytes in mice treated with LtβR–mIg inhibitor (viii) than mice treated with the control mouse Ig (vii).