CD4-positive T cell-mediated neuroprotection requires dual compartment antigen presentation

Abstract

Our laboratory discovered that CD4-positive (CD4+) T cells of the immune system convey transitory neuroprotection to injured mouse facial motoneurons (FMNs) (Serpe et al., 1999, 2000, 2003). A fundamental question in the mechanisms responsible for neuroprotection concerns the identity of the cell(s) that serves as the antigen-presenting cell (APC) to activate the CD4+ T cells. Here, we first establish that CD4+ T cells reactive to non-CNS antigen fail to support FMN survival and, second, demonstrate a two-compartment model of CD4+ T cell activation. Mouse bone marrow (BM) chimeras were developed that discriminate between resident antigen-presenting host cell and BM-derived antigen-presenting donor cell expression of major histocompatibility complex II within central and peripheral compartments, respectively. After facial nerve transection, neither compartment alone is sufficient to result in activated CD4+ T cell-mediated FMN survival. Rather, CD4+ T cell-mediated neuroprotection appears to depend on both resident microglial cells in the central compartment and a BM-derived APC in the peripheral compartment. This is the first in vivo report demonstrating a neuroprotective mechanism requiring APC functions by resident (i.e., parenchymal) microglial cells.

Figure 1.

CD4⁺ T cells reactive to non-CNS antigen fail to support FMN survival. a–d, Thionin-stained FMNs of the control and axotomized facial motor nuclei 4 weeks after facial nerve axotomy. a, WT mice without reconstitution (none); inset shows a thionin-labeled FMN with clear nucleus used for counting. b, RAG-2 KO without reconstitution (none). c, WT-derived CD4⁺ T cell-reconstituted (WT-CD4⁺ T) RAG-2 KO. d, DO11.10-derived CD4⁺ T cell-reconstituted (DO11.10 CD4⁺ T) RAG-2 KO. e, Average percentage survival ± SEM of FMNs from right (axotomized) side of WT (black), RAG-2 KO (gray), WT-derived CD4⁺ T cell-reconstituted RAG-2 KO, and DO11.10-derived CD4⁺ T cell-reconstituted RAG-2KO mice, relative to the unoperated left-side controls [the asterisk and the number sign represent a significant difference from WT (none) and RAG-2KO (none), respectively, at p < 0.01; n = 6 for all groups].

Figure 2.

Immunohistochemistry of MAC-1-labeled microglia shows that microglia are present and reactive in MHC II KO mice. MAC-1-labeled microglia in the control and axotomized facial motor nuclei 7 d after facial nerve transection. a, Control side of WT, MHC II KO, and RAG-2 KO mice (4× and 40× magnification). b, Axotomized side of WT, MHCII KO, and RAG-2 KO mice (4× and 40× magnification). Arrows indicate phagocytic cluster.

Figure 3.

FMN survival requires both peripheral and central APC function. Thionin-stained FMNs of the control and axotomized facial motor nuclei 4 weeks after facial nerve axotomy. a, WT mice without reconstitution (none). b, MHC II KO chimera mice without reconstitution (none). c, Naive CD4⁺ T cell-reconstituted MHC II KO chimera mice. d, Activated CD4⁺ T cell reconstituted MHC II KO chimera mice. e, WT chimera without reconstitution (none). f, Naive CD4⁺ T cell reconstituted WT chimera mice. g, Activated CD4⁺ T cell-reconstituted WT chimera mice.

Figure 4.

FMN survival requires both peripheral and central APC function. Average percentage survival ± SEM of FMNs from right [axotomized (Ax)] side of WT (black), MHC II KO chimera (gray), and WT chimera (hatched) mice either without reconstitution (–) or with naive or activated CD4⁺ T cell reconstitution (+) relative to the unoperated left-side controls (C) (points represent individual animals in each group, and asterisks represent a significant difference at p < 0.001; n = 6–9 for all groups).