Role of Dynamic Actin Cytoskeleton Remodeling in Foxp3+ Regulatory T Cell Development and Function: Implications for Osteoclastogenesis

Abstract

In T cells, processes such as migration and immunological synapse formation are accompanied by the dynamic reorganization of the actin cytoskeleton, which has been suggested to be mediated by regulators of RhoGTPases and by F-actin bundlers. SWAP-70 controls F-actin dynamics in various immune cells, but its role in T cell development and function has remained incompletely understood. CD4⁺ regulatory T (Treg) cells expressing the transcription factor Foxp3 employ diverse mechanisms to suppress innate and adaptive immunity, which is critical for maintaining immune homeostasis and self-tolerance. Here, we propose Swap-70 as a novel member of the Foxp3-dependent canonical Treg cell signature. We show that Swap-70^-/- mice have increased numbers of Foxp3⁺ Treg cells with an effector/memory-like phenotype that exhibit impaired suppressor function in vitro, but maintain overall immune homeostasis in vivo. Upon formation of an immunological synapse with antigen presenting cells in vitro, cytosolic SWAP-70 protein is selectively recruited to the interface in Treg cells. In this context, Swap-70^-/- Treg cells fail to downregulate CD80/CD86 on osteoclast precursor cells by trans-endocytosis and to efficiently suppress osteoclastogenesis and osteoclast function. These data provide first evidence for a crucial role of SWAP-70 in Treg cell biology and further highlight the important non-immune function of Foxp3⁺ Treg cells in bone homeostasis mediated through direct SWAP-70-dependent mechanisms.

Keywords: Foxp3; SWAP-70; T cell homeostasis; Treg cell; actin dynamics; osteoclasts.

Figure 1

Constitutive SWAP-70 expression in Foxp3⁺ Treg cells and inducible Swap-70 expression in short-term stimulated and memory-type CD4⁺ Tcon cells. (A) mRNA expression levels of IL2ra, Foxp3 and Swap-70 determined by real-time RT-PCR employing FACS-purified T cell populations (naïve Treg cells, memory-type Treg cells, naïve Tcon cells and memory-type Tcon cells). Where indicated, naïve Swap-70^+/+ and Swap-70^-/- Tcon were short-term stimulated in vitro in the presence of PMA/Ionomycin. Peripheral B cells were used as positive control. Mean values ± S.D. of relative expression determined in a representative experiment (from three independent experiments) are shown for indicated genes. (B) Flow cytometric protein level analysis of intracellular (IC) SWAP-70 expression in electronically gated thymic and peripheral (scLN, mesLN, and spleen) T cell subsets. B cells were used as positive control. MFIs of samples from 6 individual mice (n=1 for CD19⁺B220⁺ B cells) were normalized to Swap-70^-/- controls (set to 1). Data are shown as heatmap displaying Swap-70^+/+ to Swap-70^-/- fold-change values of SWAP-70 expression. See also Supplementary Figure S1. (C) Identification of putative Foxp3, NF-AT, AP1 and zinc finger binding sites in the promoter region of Swap-70 by multiple sequence alignment (left). (D) mRNA expression levels of Swap-70, IL2ra and Il2 of Treg cells (green, red) and Tcon (blue, purple) cells FACS-purified from Swap-70^+/+ (green, blue) and Swap-70^-/- mice (red, purple) were cultured for 1h in the absence or presence of cyclosporine A (CsA) prior to a 4h stimulation with PMA/Iono as indicated. Mean values ± S.D. of relative expression (n=4 from 2 independent experiments), determined in triplicate, are shown for indicated genes. (E) Representative imaging flow cytometry of co-cultures of FACS-purified Treg cells and Tcon cells expressing the transgenic 2D2 MOG-specific TCR cultured for 48h in the presence of CD11c⁺ dendritic cells (DCs) (DC: Treg:Tcon cell ratio 1:5:5) and 10 µg/ml MOG peptide. Cell clusters were analyzed for Foxp3^GFP, CD11c, TCRβ, CD3 and IC SWAP-70 expression. DC-Treg cell interactions between Swap-70^+/+ DCs and Swap-70^+/+ Treg cells are shown in both rows.

Figure 2

SWAP-70, Helios and Nrp-1 expression during tTreg cell development and in extrathymic Treg cells. (A) Postsort analysis of FACS-purified CD4SP CD25⁺ Treg precursor stages from the thymus of Foxp3^RFP/GFP mice showing Foxp3^GFP and Foxp3^RFP expression. (B) IC Helios, Nrp-1 and IC SWAP-70 expression during the instructive phase (Foxp3^RFP-Foxp3^GFP-CD25⁺ CD4SP, stage I, grey) and both stages of the consolidation phase (Foxp3^RFP+Foxp3^GFP-CD25⁺ CD4SP, stage II, red and Foxp3^RFP+Foxp3^GFP+CD25^hi CD4SP, stage III, green) of tTreg cell lineage commitment. (C) Bar charts depict representative expression levels (MFI) of CD25, Helios, Nrp-1 and SWAP-70 in CD4SP CD25^- thymocytes and during Treg cell development (stages I-III). (D) Postsort analysis of FACS-purifed tTreg (Foxp3^RFP+Foxp3^GFP+CD25⁺CD4⁺) and pTreg cells (Foxp3^RFP+Foxp3^GFP-CD25⁺CD4⁺) cells isolated from scLN of Foxp3^RFP/GFP mice showing Foxp3^GFP and Foxp3^RFP expression. (E) IC Helios, Nrp-1 and IC SWAP-70 expression in CD4⁺CD25^- Tcon, pTreg and tTreg cells. (F) Bar charts depict mean expression levels (MFI) of Helios, Nrp-1 and SWAP-70 in CD4⁺CD25^- Tcon, pTreg and tTreg cells isolated from different anatomical locations (mesLN, scLN, and spleen). (G) Representative flow cytometry of electronically gated Foxp3^GFP+CD4SP thymocytes (top) and extrathymic Foxp3^GFP+CD4⁺ Treg cells (scLN, bottom) from Foxp3^GFP mice stained for IC SWAP-70 in combination with IC Helios or Nrp-1. (H) Graph illustrates frequencies of IC SWAP-70 expressing cells among the respective Foxp3^GFP+CD4⁺CD8^- cell subsets. Symbols and horizontal lines indicate individual mice and mean values, respectively (blue symbols: SWAP-70⁺ among Helios⁺ or Nrp-1⁺ cells, green symbols: SWAP-70⁺ among Helios^- or Nrp-1^- cells) mean of 6 individual mice shown, circles: 8-9-week-old, triangles 10-13-week-old mice. Numbers in dot plots indicate representative frequencies of gated cells within the respective gate. Numbers in histograms indicate representative MFIs of gated cells. See also See also Supplementary Figure S2.

Figure 3

Impact of SWAP-70 on immune homeostasis. (A–D) Swap-70 deficiency does not affect fertility, mortality or organ cellularity. (A) Number of litters and litter sizes of 8 Swap-70^+/+ (blue) and 6 Swap-70^-/- (red) mating trios (2♀ x 1♂) within a period of 15 months. Symbols and horizontal lines indicate individual mice and mean values, respectively. (B) Distribution of genotypes and (C) mortality rate of live born Swap-70 x Foxp3^GFP pups. Pie and bar chart display data from 2616 mice monitored over a period of 36 months (Swap-70^+/+ : blue; Swap-70^-/- : red; Swap-70^+/- : grey). (D) Total organ cellularity of Swap-70^+/+ and Swap-70^-/- mice as indicator for multi-organ autoimmunity. Symbols and horizontal lines indicate individual mice and mean values, respectively (squares: 5-week-old, circles: 7-8-week-old; triangles: 9-13-week-old mice, n= 12 per genotype). (E) Representative flow cytometry of CD44, CD62L, IC Helios and Nrp-1 expression of electronically gated Foxp3^GFP+ CD4SP thymocytes from Swap-70^+/+ (blue) and Swap-70^-/- (red) Foxp3^GFP mice. (F) Graphs depict frequencies of the respective cell populations among electronically gated Foxp3^GFP+ CD4SP thymocytes and Helios/Nrp-1 protein expression levels (MFI) of electronically gated Foxp3^GFP+ CD4SP thymocytes. (G) Representative flow cytometry of CD44 and CD62L expression among gated CD4⁺ Foxp3^GFP-CD25^- Tcon, Foxp3^GFP and CD25 expression among total CD4⁺ and CD44 and CD62L expression among CD4⁺Foxp3^GFP+ Treg cells in the spleen of Swap-70^+/+ (blue) and Swap-70^-/- (red) mice. (H) Graphs illustrate frequencies of memory T cells, Treg cells and effector-memory Treg cells in mesLN, scLN, and spleen. (I) Normalized protein expression levels in Swap-70^-/- Foxp3^GFP+ Treg cells from mesLN, scLN and spleen determined by surface and/or intracellular flow cytometry staining of the indicated proteins. MFIs of 4-6 individual mice were normalized to Swap-70^+/+ Foxp3^GFP+ controls (set to 1). Mean values are shown as heatmap displaying Swap-70^-/- to Swap-70^+/+ fold-change values. The differential expression is shown as up-regulation (red) or downregulation (blue). (J) Representative flow cytometry of electronically gated CD4⁺Foxp3^GFP+ Treg cells in the spleen of Swap-70^+/+ (blue) and Swap-70^-/- (red) mice costained with CD25 and CD44, CD62L and ICOS. (K) Graphs depict CD25 expression levels (MFI) on electronically gated CD4⁺Foxp3^GFP+CD25⁺Treg cells, frequency and cell numbers of CD4⁺Foxp3^GFP+CD25^- Treg cells. (L–N) Treg cell-mediated suppression of responder T cells (Tresp) in vitro: FACS-purified CD4⁺CD62L⁺Foxp3^GFP-CD25^- Tresp cells were co-cultured in different ratios with FACS-purified Swap-70^+/+ (blue) or Swap-70^-/- (red) Foxp3^GFP+ Treg cells - or as control alone (black) - in the presence of irradiated antigen presenting cells (APCs) and 1 µg/ml α-CD3ϵ. Cells were harvested and analyzed by flow cytometry at the indicated time points. (L) Representative flow cytometry of CD25 expression and cell proliferation dye (CPD) ef670 dilution of electronically gated Foxp3^GFP- Tresp cells. (M) Representative histogram overlays to determine CPD ef670 dilution of electronically gated Tresp cells that were co-cultured with Swap-70^+/+ (blue) or Swap-70^-/- (red) Foxp3^GFP+ Treg cells at a Treg : Tresp ratio of 1:1. Arrows mark proliferation dye dilution peaks (each peak represents one cell division). (N) Graphs depict frequencies of divided Tresp cells (left) and CD25 expression levels (MFI, right) at day 3 of culture for indicated Treg : Tresp ratios (n= 8 from 3 independent experiments). See also Supplementary Figure S3. *P ≤ 0.05; **P≤ 0.01; ***P ≤ 0.001; or ****P ≤ 0.0001.

Figure 4

Treg-cell-mediated suppression of osteoclastogenesis in vitro. Bone marrow-derived macrophages (BMMs) were cultured in the absence or presence of FACS-purified Swap-70^+/- or Swap-70^-/- Treg cells. (A) Representative images of tartrate-resistant acid-phosphatase (TRAP) stained osteoclast differentiation cultures without (left) or with Treg cells (middle: Swap-70^+/- , right: Swap-70^-/- Treg cells) at day 6. (B). Graph illustrates the impact of Treg cells on osteoclast formation assessed by the analysis of osteoclast areas (the number of TRAP-positive multinucleated cells) from co-cultures with Swap-70^+/+ (blue) or Swap-70^-/- (red) Treg cells normalized to control cultures without Treg cells (black) (n= 3, mean values with SD shown). (C) Bone resorption in vitro assay: Pre-osteoclast-Treg cell co-culture was seeded on bone slices at different ratios, and carboxy-terminal collagen crosslinks (CTX) in the day 5 supernatant from one sample of each titration step were measured by ELISA. (D) The bone slices of the co-culture shown in (C) were harvested at day 7, dried, coated and analyzed by SEM. (E–G) Treg-cell-mediated downregulation of CD80/86 on pre-osteoclasts. BMMs were co-cultured with Swap-70^+/- or Swap-70^-/- Treg cells, harvested and analyzed by flow cytometry at the indicated time points. (E) Representative histogram overlays of CD80 (top), CD86 (middle) and MCSF receptor (cfms, bottom) expression on electronically gated CD4^-Foxp3^GFP-CD11b⁺ cells from osteoclast differentiation cultures cultured without (grey) or with Treg cells (Swap-70^+/- blue; Swap-70^-/- red) at indicated time points. Numbers in histograms indicate representative MFIs of gated cells. (F) Graphs show expression levels of CD80, CD86, cfms and CD11b (MFI) for indicated culture conditions and timepoints (one or two data points per condition from continuous co-culture). (G) Representative imaging flow cytometry of osteoclast-Treg cells co-culture derived cell suspensions that were analyzed for CD11b, Foxp3^GFP, CTLA-4 and CD80 expression, showing single cell (top left: Foxp3⁺ Treg cell, bottom left: pre-osteoclast) and cluster images (top right: pre-osteoclast - Swap-70^+/+ Treg cell interaction, bottom right: pre-osteoclast - Swap-70^-/- Treg cell interaction).

Figure 5

Proposed role for SWAP-70 in the direct interplay between Foxp3⁺Treg cells and osteoclast precursors. Bone homeostasis mediated through Foxp3⁺ Treg cells (via CTLA-4 interacting with CD80/CD86) underlies SWAP-70-dependent mechanisms. Adapted from . This figure was in part created with modified Servier Medical Art templates, licensed under a Creative Commons Attribution 3.0 Unported license: http://smart.servier.com.