Differences between CD8+ T cells in lupus-prone (NZB x NZW) F1 mice and healthy (BALB/c x NZW) F1 mice may influence autoimmunity in the lupus model

Abstract

Immunization with portions of a murine antibody to DNA induced Ig peptide-reactive peripheral CD8+ inhibitory T (Ti) cells in non-autoimmune (BALB/c x NZW) F1 (CWF1) mice. Those Ti suppressed in vitro production of IgG anti-DNA by lymphocytes from MHC-matched, lupus-prone (NZB x NZW) F1 (BWF1) mice, primarily via secretion of transforming growth factor-beta (TGF-beta). However, splenic CD8+ cells from immunized BWF1 mice failed to suppress anti-DNA. Therefore, BWF1 mice were studied for defects in peripheral CD8+ T cells. The potential to suppress autoimmunity mediated by activated CD4+ helper T and B cells in BWF1 mice was assessed. As BWF1 mice aged, peripheral CD8+ T cells expanded little; fewer than 10% displayed surface markers of activation and memory. In contrast, quantities of splenic CD4+ T and B cells increased; high proportions displayed activation/memory markers. In old compared to young BWF1 mice, splenic cell secretion of two cytokines required for generation of CD8+ T effectors, IL-2 and TGF-beta, was decreased. Immunizing BWF1 mice activated peptide-reactive CD8+ T cells, but their number was decreased compared to young BWF1 or old normal mice. While peptide-reactive splenic CD8+ T cells from immunized BWF1 mice did not survive in short-term cultures, similar CD8+ T cell lines from immunized CWF1 mice expanded and on transfer into BWF1 mice delayed autoimmunity and prolonged survival. Therefore, CD8+ T cells in old BWF1 mice are impaired in expansion, acquisition of memory, secretion of cytokine, and suppression of autoimmunity. Understanding these defects might identify targets for therapy in systemic lupus erythematosus.

Fig. 1

Examples of CD8⁺ CWF1 T cell lines that were peptide-specific, nonspecific, and non-stimulatory. Each cell line was tested at least two times, each experiment in triplicate, with the two most suppression-stimulating peptides (A6.31 and A6.41) and with at least two non-stimulatory peptides (A6.114, HYHEL.31, or HELp106). CWF1 females were immunized with either A6.31 or A6.41, boosted once, and CD8⁺ T cells were isolated from spleens 10–14 days later. Lines were developed by repeated stimulation with Con A supernatant, BWF1 APC, and immunizing peptides. Each of these lines was developed after at least two 14-day cycles of stimulation (none of the CD8⁺ BWF1 lines survived beyond that time) and harvested 12 days after rest from stimulation. Lines (10⁵ cells) were co-cultured with BWF1 B cells (1 × 10⁵) and T cells (1 × 10⁶) with or without 20 μM of peptides. In all figures, a result of zero means that peptide was not tested. (A) Examples of two BWF1 CD8⁺ T cell lines (not suppressive) and two CD8⁺ CWF1 T cell lines (suppressive) tested for their abilities, when activated by their immunizing peptide (A6.31), to suppress the numbers of BWF1 B cells producing IgG anti-dsDNA (on the x axis). Data are expressed as mean AFC per 10⁵ B cells. Vertical lines indicate SEM. (B) Example of six additional cell lines that were suppressive and peptide-specific (white: no peptide added; horizontal lines: peptide A6.1 31–45 added; vertical lines: peptide A6.1 41–54 added; diagonal lines: peptide A6.1 114 added; gray: peptide HYHEL.31 added; black: peptide HEL106 added). * Significantly different from B+T+peptide controls in far left column, p<0.05 by Mann-Whitney U-test. Note that lines 3, 7, 10, and 11 are activated specifically by A6.31 and lines 4 and 1 are activated specifically by A6.41. (C) Examples of three lines that are suppressive, but not strictly peptide-specific. One line is activated to suppress by every peptide tested; it is autoreactive. (D) Examples of three cell lines that were not suppressive.

Fig. 2

Numbers of lymphocytes (total spleen mononuclear cells, CD4⁺, B, and CD8⁺) change over time in CWF1 and BWF1 mice. y: young mice aged 11–12 weeks; o: old mice aged 35–38 weeks (all BWF1 with proteinuria ≥2+). Spleen cells pooled from three to five mice in each group were analyzed by FACS after appropriate surface staining. Means ± SEM are shown for four experiments for each group. (A) In old BWF1 mice (BWo) compared to young (BWy) total mononuclear spleen cell numbers increased significantly (*p<0.01 by ANOVA). (B) In old BWF1, numbers of CD4⁺ T cells increased significantly (***p<0.001), as did B cells (***p<0.001) but not CD8⁺ T cells, as compared to 12-week-old mice. (C) As CWF1 mice aged, total numbers of CD4⁺ T cells declined significantly (*p=0.05), as did CD8⁺ T cells (**p<0.01), as compared to young CWF1.

Fig. 3

Surface phenotype of splenic CD4⁺ T cells. Spleen cells pooled from three to six mice, groups as described in Fig. 2, were analyzed by flow cytometry. Results are shown as the means ± SEM from four experiments. (A) Activation and memory indices in various groups. All indices were significantly higher in old BWF1 mice compared to all other groups (*p<0.05 by ANOVA). (B) Surface expression of CD69 occurred on a higher proportion of CD4⁺ T cells from old BWF1 compared to other groups (*p<0.05). (C, D) Expression of CD45RB on CD4⁺ T cells from young (C) vs. old (D) mice. Expression of CD45RB fell dramatically with age. (E, F) Expression of CD44 on surface of CD62L⁺ or CD62L⁻ CD4⁺ T cells in young (E) vs. old (F) mice. Proportions of CD44^highCD62L^low cells increased dramatically with age. Altogether, these data indicate that greatly increased numbers/proportions of CD4⁺ T cells express markers of activation and memory in old BWF1 mice compared to young.

Fig. 4

Surface phenotype of splenic CD8⁺ T cells in young and old CWF1 and BWF1 mice. Panels are in same order as for Fig. 4. Note that in contrast to CD4⁺ T cells, activation/memory markers are not elevated in the CD8⁺ T cells of old BWF1 mice compared to other groups.

Fig. 5

Characteristics of CD8⁺ T cells from old BWF1 and old CWF1 mice after activation in vivo or in vitro. (A) Higher proportion of splenic CD8⁺ T cells from immunized old BWF1 express activation markers 1 week after a single booster immunization with peptide A6.1 31–45. Each point represents a single mouse. n: naive; i: immunized. p<0.02 by ANOVA for old BWF1 immunized compared to all other groups. Apoptosis of T cells following activation with plate-bound anti-CD3 and anti-CD28 after 12 h in culture was measured by annexin V staining in flow cytometry experiments on co-cultures of CD4⁺, CD8^+, and B cells from splenocytes of mice immunized with A6.31. The table shows statistical analysis of comparison of means of mean fluorescence intensity for annexin V staining in three to five separate experiments in each mouse group. Results of representative individual experiments are shown in (B–G). Cells from young mice are shown in black, from old mice in gray. (B–D) Annexin V staining of CD4⁺-gated cells in lupus-prone (BWF1) mice, normal mice (BALB/c), and normal MHC-matched control mice (CWF1). (E–G) Comparison of annexin V staining among CD8⁺ T cells in a representative experiment of five on groups of three to five mice per group. Measurement of cell death by 7-AAD staining gave analogous results (not shown).

Fig. 6

In vivo transfer of CWF1 CD8⁺ Ti cell line #3 into young BWF1 mice (two injections 2 weeks apart of 4×10⁶ cells activated with an A6.1 peptide and syngeneic APC), compared to injections of saline or of identical numbers of activated CWF1 CD4⁺ Th cells from another peptide-specific line, suppressed lupus-like disease. (A, B) Mean serum IgG anti-DNA (± SEM) was significantly decreased at 24 and 32 weeks of age in Ti-transferred recipients; *p<0.05 and p<0.001, respectively, by ANOVA. (C) Transfer of CD8⁺ Ti delayed the onset of proteinuria (≥2+) (p<0.01 compared to either saline or Th). (D) Transfer of CD8⁺ Ti prolonged survival compared to saline or Th cell transfer (*p=0.01 comparing group receiving Ti cells to both other groups by ANOVA). Numbers of mice were 5 receiving CD8⁺ Ti, 21 receiving saline, 16 receiving CD4⁺ Th.