IL-6 is one of the key factors in the formation of gut tissue resident memory T cells from Naïve T cells
Han G. Kim, Amanda Chan, Sinmanus Vimopatranon, Alexandre Girard, Andrew Jiang, Samuel Wertz, Il-Young Hwang, John H. Kehrl, Hana Schmeisser, Madelyn M. Seemiller, Paolo Lusso, Dawei Huang, Danlan Wei, Livia R. Goes, Marcelo Soares, Elena Martinelli, James Arthos, Claudia Cicala
TL;DR
This study identifies IL-6 as a key factor in the development of gut tissue resident memory T cells, which are important for long-term immunity and may help in creating better vaccines and treating gut inflammation.
Contribution
The study reveals that IL-6, along with MAdCAM-1, TGF-β, and RA, is essential for differentiating naïve T cells into gut TRMs.
Findings
IL-6 induces TRM-associated markers like CD69, CD103, and CCR5 in naïve CD4+ T cells.
TRM differentiation occurs via JAK/STAT signaling and is suppressed by JAK/STAT antagonists.
MAdCAM-1, TGF-β, RA, and IL-6 work together to promote TRM formation in mucosal environments.
Abstract
Tissue resident memory CD4+ T cells (TRMs) populate mucosal sites and play a critical role in local immune responses. Gut TRM cells persist for extended periods in the gut mucosa where they rapidly respond to invading pathogens and provide long lasting protection. This study investigates the factors that mediate differentiation of naïve CD4+ T cells into cells presenting a gut TRM phenotype. Naïve CD4+ T cells were cultured under conditions that mimicked mucosal environments. This included signaling through MAdCAM-1 in the presence of Retinoic Acid (RA) and TGF-β. This combination of stimuli primed naïve CD4+ T cells to adopt a TRM phenotype. However, to fully differentiate into TRMs an additional soluble factor provided by memory T cells was required. Our results identified IL-6 as one of the key factors that induces the expression of TRM -associated markers, including CD69, CD103 and…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Fig 1
Fig 2
Fig 3
Fig 4- —http://dx.doi.org/10.13039/100006492Division of Intramural Research, National Institute of Allergy and Infectious Diseases
- —http://dx.doi.org/10.13039/501100004586Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsT-cell and B-cell Immunology · Immune Cell Function and Interaction · Psoriasis: Treatment and Pathogenesis
Introduction
Tissue resident memory T (T_RM_) cells [1,2] reside in tissues, including skin, lung and gut mucosa, for extended periods of time (up to years) [2–6]. These cells exhibit a distinct transcriptional program [7,8] and serve as sentinels that initiate rapid secondary immune responses against invading pathogens. CD8^+^ T_RM_ cells have been extensively described, while CD4^+^ T_RM_ cells remain less well characterized. CD103 and CD69 are the canonical cell surface markers used to identify T_RM_ cells. CD103 is the α chain of α_E_β_7_ which functions as a mucosal retention receptor through binding to E-cadherin [9]. The persistent expression of CD69 on T_RM_ cells antagonizes sphingosine-1-phosphate receptor-1 (S1PR1), a receptor that promotes lymphocyte migration out of tissues [10–12].
Two soluble factors, transforming growth factor β (TGF-β) and retinoic acid (RA) play key roles in T_RM_ differentiation. TGF-β is a pleiotropic cytokine that suppresses lymphocyte activation. It also upregulates CD103 expression [13,14]. RA is a vitamin A metabolite that is produced by dendritic cells but may also be produced by intestinal epithelial cells [15]. RA drives the expression of integrin α_4_β_7_ (α_4_β_7_) and CCR9, both of which mediate lymphocyte homing to gut tissues [16]. Of note, RA works synergistically with TGF-β to generate intestinal T_RM_ cells [17,18].
Constant exposure to dietary and bacterial antigens presents a unique challenge to the gut immune system, which requires a finely balanced control between inflammation and tolerance [19,20]. One way in which gut lymphocytes regulate immune responses involves utilizing an array of costimulatory receptors that initiate distinct cellular programs. A key costimulatory interaction involves mucosal vascular addressin cell adhesion molecule 1 (MAdCAM-1) [21–24] binding to α_4_β_7_, an integrin receptor expressed on various lymphocyte subsets. α_4_β_7_ is expressed uniformly on naïve CD4^+^ T cells, but variably on memory CD4^+^ T cells subsets. The high endothelial venules of Peyer’s patches, mesenteric lymph nodes, and intestinal lamina propria express MAdCAM-1, which is upregulated in response to inflammation [25,26]. It is this localized expression of MAdCAM-1 that underlies the assignment of α_4_β_7_ as the principal lymphocyte gut homing receptor. Both RA and interleukin-7 induce upregulation of α_4_β_7_, thus conferring upon T cell’s gut-homing properties [16,27]. In addition to its trafficking function, MAdCAM-1 delivers costimulatory signals to T cells through α_4_β_7_, which promotes cellular proliferation [23,28].
IL-6 has a dual role in GI inflammation. While it can contribute to chronic inflammation in IBD, it also plays a role in maintaining gut health and promoting tissue repair. By regulating the balance between pro- and anti-inflammatory responses in T cells, IL-6 regulates chronic GI inflammatory environment and these effects depend on the target cells [29,30]. Understanding the specific context and timing of IL-6’s actions is crucial for developing effective therapies for GI disorders.
Variations in local tissue-milieu cause T_RM_ cells to differentiate into diverse T_RM_ cell subsets [4,6,31]. Along with different costimulatory signals, tissue specific soluble factors (e.g., cytokines, chemokines, metabolites) determine the differentiation pathways that T_RM_ cells follow. How various tissue-specific factors influence this process is a subject of active investigation. Initial reports indicated that CD8^+^ T_RM_ cells derive from effector memory CD8^+^ T cells. While more recent reports suggest that they can also emerge directly from naïve CD8^+^ T cells, significant gaps remain in our understanding of their ontogeny [4,6,32]. In this report, we describe a distinctive CD4^+^ T_RM_ cells differentiation pathway originating from naïve CD4^+^ T cells. Differentiation through this pathway is mediated by MAdCAM-1 and involves an apparent synergy between a number of soluble factors including IL-6, RA and TGF-β. These findings hold the potential to inform new therapeutic approaches to the treatment of diseases involving gut immunity.
Results
Memory CD4+ T cell soluble factors promote the differentiation of MAdCAM-1 costimulated CD4+ naïve T cells into TRMs
In a previous study we demonstrated that costimulating primary CD4^+^ T cells with MAdCAM-1 (combined with plate immobilized anti CD3 mAb) in the presence of RA for 7 days with the addition of TGF-β for the last 3 days, generates cells with a CD69^+^/CD103^+^ T_RM_ phenotype [24]. In these experiments we noted that in cultures containing only purified CD45RO^-^ cells (naïve), MAdCAM-1 costimulation inefficiently generated CD69^+^/CD103^+^ T_RM_ cells relative to cultures that included both CD45RO^+^ and naïve cell subsets (S1 Fig). CD45RO expression is commonly used to define memory T cells. Although not perfect in all circumstances, CD45RO is largely consistent with memory and sufficient in this study. We subsequently examined how variation in relative frequencies of naïve and memory CD4^+^ T cell subsets among donor PBMCs influenced the efficiency of T_RM_ cell generation. We purified bulk CD4^+^ T cells from 43 independent donors and costimulated them with MAdCAM-1 in the presence of RA and TGF-β as described above. For each of these donors we plotted the starting frequency, before culture, of naïve CD4^+^ T cells at day 0 on the x-axis and the frequency of CD69^+^/CD103^+^ T_RM_ cells at day 7 on the y-axis. This analysis revealed a weak positive correlation (R^2^ = 0.053) (Fig 1A), suggesting that the naïve/memory ratio may affect T_RM_ differentiation. We confirmed this relationship by combining different ratios of purified naïve T cells and purified memory T cells from the same donor (Fig 1B). As the ratio of naïve to memory cells on day 0 increased, the frequency of T_RM_ cells on day 7 also increased. However, when we removed memory cells and employed purified naïve cell cultures, we observed a sharp reduction in the frequency of T_RM_ cells (Fig 1B). We repeated this experiment with purified naïve cultures from 8 donors which yielded an average of 9.40 ± 8.35 CD69^+^/CD103^+^ cells (median 5.98 ± 12.3) (Fig 1C). In contrast, costimulation of matched bulk cultures (n = 8) yield an average of 25.3 ± 12.2 double positive cells (median 25.8 ± 23.9). These findings suggest that MAdCAM-1 costimulation can generate T_RM_ cells from naïve CD4^+^ T cells, but a small number of memory T cells facilitates this process.
Soluble factors promote MAdCAM-1 costimulated naïve CD4+ T cells to express TRM cell surface markers.(A) The frequency (%) of CD45RO- CD4+ T cells on day 0 (x-axis) and the frequency of CD69+/CD103+ cells on day 7 in multiple independent donors (y-axis) following costimulation of bulk CD4+ T cells with MAdCAM-1 (upper), VCAM-1 (middle) or CD28 mAb (lower) in the presence of RA and TGF-β (closed circles, n = 43). The solid black line and equation represent a linear regression of that includes only bulk cultures. The dashed line represents a quadratic polynomial regression that includes both bulk and purified CD45RO- cells. Linear correlation coefficients are provided for each panel. (B) CD69/CD103 coexpression from a representative donor following MAdCAM-1 costimulation of CD45RO-/CD45RO+ CD4+ T cell cultures reconstituted at different ratios. Left panels indicate the approximate ratio of CD45RO- to CD45RO+ cells on day 0. Right panels indicate the coexpression of CD69 (y-axis) and CD103 (x-axis) on day 7, with the frequency of double positive cells indicated. (C) Frequency of CD69+/CD103+ cells following MAdCAM-1, VCAM-1, and CD28 mAb costimulation of matched bulk (B) and naïve (N) CD4+ T cell cultures (n = 8) harvested at day 7. Y-axis indicates % double positive cells. (D) Coexpression of CD69 and CD103 following MAdCAM-1 costimulation of purified CD45RO- cells in the absence (UL panel) or presence of VCAM-1 (UR), MAdCAM-1 (LL) or CD28 mAb (LR) bulk culture supernatants, from a representative donor. (E) Treatment of purified CD45RO- cells as in panel D (n = 11) harvested at day 7. Error bars indicate standard deviation (: p < 0.05, **: p < 0.01, ***: p < 0.001, ***: p < 0.0001).
We next asked how the relative frequencies of naïve and memory CD4^+^ T cells impacted T_RM_ differentiation in the context of other costimulatory signals. Cells were costimulated with VCAM-1, which binds both α_4_β_1_ and α_4_β_7_, and CD28 mAb. For bulk CD4^+^ T cell cultures (n = 43), VCAM-1 and CD28 mAb induced similar frequencies of T_RM_ cells compared to MAdCAM-1 (Fig 1C). Unlike MAdCAM-1, a positive correlation was not observed in VCAM-1 and CD28 mAb treated cells (R^2^ = 0.009 and 0.022, respectively) (Fig 1A). VCAM-1 and CD28 mAb also exhibited stronger costimulatory signaling with greater T cell viability compared to MAdCAM-1 cultures (n = 10) (average MAdCAM-1 = 29.15% ± 10.45, VCAM-1 = 50.72% ± 11.18, CD28 mAb = 52.45% ± 12.97) (S2 Fig). When comparing bulk cultures to purified naïve CD4^+^ T cell cultures from the same donors (n = 8), VCAM-1 exhibited a greater capacity to generate T_RM_ cells (average naïve: 24.3 ± 10.1 vs bulk 27.9 ± 10.3) while CD28 mAb -mediated T_RM_ cells formation was reduced, (average naïve:12.0 ± 8.39 vs bulk: 24.6 ± 8.05) (Fig 1C). To determine if memory T cells provided a soluble factor(s), we added supernatants from day 4 stimulated bulk CD4^+^ T cells (stimulated with either MAdCAM-1, VCAM-1, or CD28 mAb) to MAdCAM-1 costimulated naïve CD4^+^ T cell cultures. All 3 bulk supernatants provided a soluble factor(s) that significantly “rescued” T_RM_ cell formation from naïve cell cultures (Fig 1D, 1E). We noted that CD28 mAb and VCAM-1 supernatants were more efficient (Fig 1E). We concluded that MAdCAM-1 provides a costimulatory signal that prompts naïve CD4^+^ T cells to differentiate into T_RM_ cells, and is facilitated by a factor(s) provided by memory CD4^+^ T cells.
MAdCAM-1 costimulation in the presence of TGF-β promotes TRM cell formation despite low level proliferation
We next investigated to what extent MAdCAM-1 -mediated T_RM_ differentiation of naïve cells depended upon the proliferation of either subset of CD4^+^ T cells. While α_4_β_7_, the MAdCAM-1 receptor, is uniformly expressed on naïve CD4^+^ T cells, it appears on only a small subset of memory CD4^+^ T cells [33,34]. We labelled purified naïve and memory CD4^+^ T cells with CFSE and Far Red respectively and recombined the two cell types at a ratio favorable for T_RM_ cell formation (2:1, naïve to memory). Because TGF-β exerts an antiproliferative effect on lymphocytes [13,35,36], we cultured the cells in the absence or presence of exogenous TGF-β. Cultures were stimulated as described above and analyzed for dye dilution proliferation on day 7 using flow cytometry. Proliferation was detected in both naïve and memory cell subsets, with or without TGF-β (Fig 2A). We obtained average division indexes for naïve (Fig 2B) and memory (Fig 2C) cell subsets from 4 different donors. MAdCAM-1 promoted proliferation to a greater degree in naïve cells (naïve: 0.65 ± 0.2 vs memory: 0.24 ± 0.1), while TGF-β significantly suppressed this proliferation in both subsets (Fig 2B, 2C). For comparison we stimulated cells with VCAM-1 and CD28 mAb. Proliferation was induced in both naïve and memory subsets, and was not suppressed by the addition of TGF-β. To better understand how proliferation was linked to T_RM_ differentiation we plotted the percent expression of CD103 and CD69 against successive cell divisions. MAdCAM-1 costimulation, in the absence of TGF-β, induced few cells (< 5%) to express CD103 (Fig 2D). The addition of TGF-β significantly upregulated CD103 expression in cells undergoing one or more division cycles (Figs 2D and S3A). CD69 expression was somewhat reduced by TGF-β (~17.0%), but cells remained CD69^+^ throughout their successive divisions (Figs 2D and S3B). MAdCAM-1 in the presence of RA and TGF-β uniquely upregulated the expression of CCR5 (Figs 2D and S3C). This effect on CCR5 expression was not observed with VCAM-1 or CD28 mAb (Figs 2D and S3C). We concluded that the combination of MAdCAM-1, RA and TGF-β was distinct in its capacity to upregulate the expression of two key gut T_RM_ markers (CD103 and CCR5) on naïve CD4^+^ T cells. This occurred despite low level cell proliferation.
Naïve and memory CD4+ T cell proliferation in response to MAdCAM-1, VCAM-1, and CD28 mAb costimulation. (A) Flow cytometric dot plots of CD4+ T cell proliferation by dye dilution, from a representative donor, costimulated with MAdCAM-1, VCAM-1 or CD28 mAb without (upper) and with (lower) TGF-β. Reconstituted cultures included naïve cells (CD45RO-) labelled with CFSE (y-axis), and memory cells labelled with Far Red (x-axis). Division index (DI) for each population is indicated. (B and C) DI of naïve and memory cells from 4 donors (y-axis), costimulated as in panel A with and without TGF-β, as indicated (: p < 0.05). (D) CD4+ T cells costimulated as in panel A. Average frequency of CD103, CD69 and CCR5 expression on naïve cells in each division cycle, in the presence or absence of TGF-β as indicated (n = 3). d0 represents non-divided cell populations with each subsequent dn + 1 representing an additional division. Comparisons between MAdCAM-1 + RA and MAdCAM-1 + RA + TGF-β are provided. Error bars indicate standard deviation (p < 0.0001).*
IL-6 promoted MAdCAM-1 mediated TRM cells differentiation of naïve CD4+ T cells
As described above, MAdCAM-1 costimulation through α_4_β_7_ induces naïve CD4^+^ T cells to differentiate into T_RM_s. Additionally, unidentified soluble factors present in bulk culture supernatants promote the expression of canonical T_RM_ surface markers (Fig 1E). As candidate soluble factors we tested 3 proinflammatory cytokines: TNF-a, IFN-γ, and IL-6, and an anti-inflammatory cytokine, IL-10. We purified naïve CD4^+^ T cells from 6 independent donors and stimulated them as described in Fig 1E except that we added recombinant cytokines on day 4 in place of culture supernatants. In the absence of any added factor, naïve cells stimulated with MAdCAM-1, RA and TGF-β poorly generated T_RM_ cells (Fig 3A). While we did not rescue T_RM_ cell generation with the addition of TNF-a, IFN-γ or IL-10 (Fig 3A), when we added IL-6, naïve cells differentiated into T_RM_ cells at a level comparable to those observed when adding bulk cultures supernatants (Fig 3A). CD4^+^ CD45RA^+^ naïve T cells also express higher levels of both IL6 receptor (IL-6Rα) and the signal transducer gp130 (IL-6Rβ) when compared to CD45RO^+^ memory T cells (S4 Fig). To further define the role of IL-6, we tested whether the IL-6 receptor antagonist tocilizumab [37], inhibited T_RM_ cell formation. Purified naïve cells were costimulated with MAdCAM-1, RA and TGF-β, and bulk culture supernatants were added on day 4 and tocilizumab was added on both day 0 and day 4. Tocilizumab addition resulted in a significant reduction in the frequency of CD69^+^/CD103^+^ cells in all donors tested (mean percent reduction = 23.34 ± 10.89) (Fig 3B). The failure to completely inhibit the supernatant activity suggested that additional factors present in the bulk supernatants also promoted T_RM_ cell formation (Fig 3B).
IL-6 promotes MAdCAM-1 mediated TRM differentiation of naïve CD4+ T cells.(A) Flow cytometric analysis comparing CD69+/CD103+ upregulation on naïve CD4+ T cells costimulated with MAdCAM-1, RA and TGF-β in the absence vs presence of exogenous IL-6, IL-10, TNF-α, and IFN-γ, as indicated. Treatment with day 4 culture supernatant included for comparison (n = 6). (B) Coexpression of CD69 and CD103 following MAdCAM-1 costimulation (with RA and TGF-β) of bulk CD4+ T cells in the absence or presence of tocilizumab (added on day 0 and day 4) as indicated (n = 6). (C) Coexpression of CD69 and CD103 following VCAM-1 or CD28 mAb costimulation in the absence or presence of exogenous IL-6 (added on day 0 and day 4). MAdCAM-1 costimulation included for comparison (n = 8). (D) Average CD69/CD103 coexpression on MAdCAM-1 costimulated cells treated with IL-6 on day 4 vs day 0 and day 4 (n = 12). Tocilizumab addition included as a control. (E) Flow cytometric time course (days 0-5) (x-axis) of α4β7, CCR9 and CCR5 expression on purified naïve CD4+ T cells costimulated with MAdCAM-1 in the absence or presence of IL-6, RA, or IL-6 + RA. Average % positive cells (n = 3) is reported (y-axis). Error bars indicate standard deviation (: p < 0.05, **: p < 0.01, ***: p < 0.001, ***: p < 0.0001).
Next we determined whether IL-6 impacted T_RM_ cell generation following VCAM-1 and CD28 mAb costimulation.
Naïve cells were costimulated with each ligand in the presence or absence of exogenous IL-6 and the frequency of CD69^+^/CD103^+^ cells was measured on day 7. IL-6 promoted VCAM-1 mediated upregulation of CD69^+^/CD103^+^ cells (Fig 3C), but it had no significant effect on CD28 mAb mediated generation of T_RM_ cells (Fig 3C). Thus, IL-6 promoted T_RM_ differentiation in response to the two costimulatory ligands that signal through integrins containing the α_4_ chain, while having no significant effect on CD28 mAb -mediated induction of T_RM_ cells. To determine whether adding IL-6 earlier in the culture would further enhance T_RM_ differentiation we modified our protocol and added IL-6 on day 0 and again on day 4. This modification increased the frequency of T_RM_ cells in 12 independent donors compared to adding IL-6 only on day 4 (Fig 3D). Addition of tocilizumab on day 0 significantly suppressed the generation of T_RM_ cells. These observations suggested that IL-6 priming of naïve CD4^+^ T cells began during the early stages of T_RM_ differentiation.
Finally, we asked how IL-6, in the context of MAdCAM-1 costimulation, influences the expression of three gut associated receptors (α_4_β_7_, CCR5 and CCR9). Naïve CD4^+^ T cells from 3 independent donors were costimulated with MAdCAM-1 alone, MAdCAM-1 + IL-6, MAdCAM-1 + RA, and MAdCAM-1 + IL-6 + RA and surface receptor expression was measured over time by flow-cytometry. IL-6 combined with MAdCAM-1 (without TGF-β and RA) had little effect on the expression of these markers (Fig 3E). However, the combination of IL-6 and RA induced α_4_β_7_ and CCR5 expression to a greater extent than either factor alone (CCR5 p = 0.0002; α_4_β_7_ p = 0.0392), suggesting a synergistic effect between these two soluble factors (Figs 3E, S5). In contrast, CCR9 expression was enhanced primarily by RA, whilst IL-6 addition had little impact. Overall, these findings indicated that, in the context of MAdCAM-1 costimulation in combination with RA and TGF-β, IL-6 induces naïve CD4^+^ T cells to adopt a T_RM_ phenotype. Additionally IL-6 appears to work cooperatively with RA in promoting the expression of gut tissue receptors.
MAdCAM-1 bulk supernatants prime naïve CD4+ T cells to differentiate into TRM cells through JAK/STAT signaling
Many cytokines including IL-6 signal by activating the JAK/STAT pathway. IL-6 signaling specifically triggers STAT1 and STAT3 phosphorylation [38] (Fig 4A). To address the involvement of JAK/STAT signal transduction pathway in the generation of T_RM_ cells, we briefly costimulated naïve cells with MAdCAM-1 and RA in the absence or presence of bulk culture supernatants. We prepared lysates that were then analyzed by Western blot using a panel of anti phospho STAT mAbs. Cells treated with IL-6 were tested as a control. The bulk supernatants rapidly (~10’) induced STAT1, STAT2 and STAT3 phosphorylation (Fig 4B). We next analyzed the same lysates with total STAT1 and STAT3 mAbs along with an actin mAb, and determined that the overall levels of STAT1 and STAT3 did not change during the time course. In contrast and as expected, IL-6 induced STAT1 and STAT3 phosphorylation [39] but failed to induce STAT2 phosphorylation. This finding is consistent with the inability of tocilizumab to fully neutralize the activity of the bulk culture supernatants and suggested the presence of additional T_RM_ differentiation factors responsible for the STAT2 phosphorylation.
JAK/STAT signaling and TRM CD4+ T cell formation.(A) Schematic of JAK/STAT signaling pathways. (B) Western blot analysis of lysates from naïve CD4+ T cell cultures costimulated with MAdCAM-1 with the addition of day 4 bulk CD4+ T cell supernatants (left) or exogenous IL-6 (right). Lysates harvested from 0-60 min post treatment were reacted with anti phospho -STAT1, -STAT2 and -STAT3 mAbs and total anti -STAT1 and -STAT3 mAbs. Actin mAb staining included as a control. (C) Flow cytometric analysis of naïve CD4+ T cells (n = 5) costimulated with MAdCAM-1 (+ RA and TGF-β) in the absence or presence of either IFN-β or IL-6, as indicated. Cytokines added on both day 0 and day 4. (D) Flow cytometric analysis of CD69/CD103 coexpression of bulk CD4+ T cells (n = 11) following MAdCAM-1, VCAM-1 or CD28 mAb costimulation (+ RA and TGF-β) in the absence or presence of ruxolitinib. Ruxolitinib added on day 0 or day 0 and 4, as indicated. (E) Flow cytometric analysis of CD69/CD103 coexpression of bulk CD4+ T cells (n = 6) following MAdCAM-1, VCAM-1 or CD28 mAb costimulation (+ RA and TGF-β) in the absence or presence of AZD1480. AZD1480 added on day 0 or day 0 and 4, as indicated. Error bars indicate standard deviation (: p < 0.05, **: p < 0.01, ***: p < 0.001, ***: p < 0.0001).
A good candidate was Interferon-β (IFN-β), one of the cytokines that delivers signals through STAT2 phosphorylation [40,41] and is reported to modulate CD103 expression in mice [42]. We tested whether IFN-β could facilitate the formation of T_RM_ cells from naïve CD4^+^ T cells costimulated with MAdCAM-1. Purified naïve cells from five independent donors were stimulated with MAdCAM-1 in the presence of RA and TGF-β with and without the addition of exogenous IFN-β. IL-6 stimulation was added as a control. On day 7, we analyzed the cells for CD69/CD103 coexpression by flow cytometry. Similar to IL-6, IFN-β induced the coexpression of these two canonical T_RM_ markers (Fig 4C), a finding that is suggestive of STAT2 involvement. However, because IFN-β mediates phosphorylation of multiple STATs, we cannot exclude the possibility that upregulation of T_RM_ markers was solely mediated by STAT1 and STAT3 phosphorylation.
To firmly establish a role of JAK/STAT signaling in T_RM_ differentiation we employed two JAK phosphorylation inhibitors, ruxolitinib [43] and AZD1480 [44]. Bulk CD4^+^ T cells from 6 or more donors were treated with MAdCAM-1 in combination with TGF-β, and analyzed for CD69/CD103 coexpression by flow cytometry, as described above. VCAM-1 and CD28 mAb were tested as controls. Ruxolitinib and AZD1480 were added on both day 0 and day 4 or only on day 4. For MAdCAM-1 costimulation both inhibitors, when added at day 0, resulted in a near complete inhibition of T_RM_ cell formation (Fig 4D and 4E). The addition of ruxolitinib or AZD 1480 on day 4 only resulted in a 50.62% ± 24.07%, and 38.70% ± 23.42% inhibition respectively. For VCAM-1 and CD28 mAb costimulations, the pattern of inhibition differed. The addition on day 4 resulted in a significant increase in T_RM_ cells generation, whereas addition on both day 0 and day 4 resulted in a partial inhibition. Overall, these findings suggested that MAdCAM-1 costimulation primes the cells to differentiate into T_RM_ cells, while soluble factors that signal through a JAK/STAT pathway bring this process to completion. Moreover, MAdCAM-1 employs a pathway that is distinct from that used by VCAM-1 and CD28 mAb, underscoring that different pathways can lead to T_RM_ cell differentiation. Overall, we found that MAdCAM-1 in combination with RA primes naïve CD4^+^ T cells to express markers of gut T_RM_ cells. The addition of IL-6 and TGF-β promotes this process even further. The implications and limitations of our findings are discussed below.
Discussion
We previously reported that MAdCAM-1, when combined with RA, provides a costimulatory signal that primes a subset of peripheral CD4^+^ T cells to express canonical T_RM_ markers (CD69 and CD103), along with α_47_β, CCR5 and CCR9, three receptors associated with trafficking to the gut associated lymphoid tissue (GALT) [24]. The addition of TGF-β further upregulated the expression of T_RM_ markers. From our initial observations it was unclear whether these cells originated from naïve or memory cell subsets (or both). Previous studies suggested that T_RM_ cells originate from activated effector T cells [6,45,46]. However, recent studies have illuminated an additional pathway wherein naïve T cells can be primed to differentiate into T_RM_ cells [6,32,47]. Results presented herein support this concept of priming and provide evidence that signaling through α_4_β_7_ plays a central role in this process. A better understanding of T_RM_ cell ontogeny holds the promise to enhance our ability to develop treatments against mucosal infections and cancer, and to develop vaccines that target mucosal tissues [4,32,48].
We previously noted that the proportion of naïve and memory cells in our starting cultures impacted the efficiency with which T_RM_ cells were generated. In this study we addressed the potential of both cell subsets to adopt a T_RM_ phenotype. We found that bulk cultures with higher proportions of naïve cells tended to generate higher frequencies of T_RM_ cells following MAdCAM-1 costimulation. However, from purified naïve cell cultures, MAdCAM-1 inefficiently generated T_RM_ cells. These contrasting observations suggested that naïve cells can differentiate into T_RM_ cells, but that memory CD4^+^ T cells provided an essential soluble factor(s). We tested 4 potential factors: TNFa, IFN-γ, IL-10 and IL-6. IL-6 consistently promoted T_RM_ cell formation in purified naïve cell cultures. We conclude that IL-6 is one of the soluble factors provided by memory CD4^+^ T cells. However, incomplete inhibition of T_RM_ cell formation by tocilizumab, an IL-6 antagonist, suggested the involvement of other factors. We found that IFN-β, which signals through multiple STATs including STAT2 [40,49], could also promote T_RM_ cell formation. However this finding left unanswered the specific roles of different STATs in T_RM_ differentiation, and the identity of additional factors remains to be determined. Nevertheless, while other factors may contribute, we are able to conclude that IL-6 helps mediate T_RM_ cell formation through JAK/STAT signaling. Of note, two JAK1/2 inhibitors, ruxolitinib and AZD1480, completely inhibited T_RM_ cell formation. Overall, our findings suggest that IL-6 mediates the formation of a distinct subset of T_RM_ cells and that other factors may give rise to alternative T_RM_ subsets.
The proportion of naïve and memory cells in starting cultures had less impact on T_RM_ cell formation when those cultures were stimulated with VCAM-1 or CD28 mAb. We found a weak positive correlation between the frequency of naïve cells in starting bulk cultures and T_RM_ cell formation when MAdCAM-1 was employed, but no correlation when either VCAM-1 or CD28 mAb were employed. A likely explanation for these differences involves the capacity of these two ligands to stimulate memory CD4^+^ T cells. Their counterreceptors, α_4_β_1_ and CD28, are expressed on most memory CD4^+^ T cells. In contrast, α_4_β_7_ is expressed on all peripheral naïve CD4^+^ T cells but on only a small fraction of memory CD4^+^ T cells (5%-15%) [33,34]. Consistent with this explanation, both VCAM-1 and CD28 mAb induced memory cell proliferation in a more efficient way than MAdCAM-1 (Fig 2A, 2C). While T_RM_ cells were initially thought to originate mainly from effector T cells we, like others [6,32,47], have shown that naïve T cells can also give rise to CD4^+^ T_RM_ cells, highlighting the diversity of pathways that generates this T cell subset.
The naïve CD4^+^ T cells in our starting cultures express a near uniform level of α_4_β_7_, typically referred to as “intermediate” expression, relative to α_4_β_7_ expression levels on effector memory CD4^+^ T cells, which are typically described as α_4_β_7_^high^ cells. As naïve cells respond to MAdCAM-1, α_4_β_7_ expression increases in a process that is facilitated by RA, a vitamin A metabolite that plays a central role in mucosal immune responses [16,18,50]. Ultimately, these cells upregulate α_4_β_7_ expression to a level comparable to that observed on effector memory CD4^+^ T cells. They also upregulate CCR9 and CCR5, two chemokine receptors associated with gut tissue homing. As such, combining MAdCAM-1 with IL-6, RA and TGF-β promotes a phenotype that facilitates retention in gut tissues. Although MAdCAM-1 primes CD4^+^ T cells, TGF-β drives T_RM_ differentiation further. TGF-β is a pleiotropic immunosuppressive cytokine that dampens cell activation [13,35,36]. Previous studies report that TGF-β is linked to tissue residency through direct upregulation of CD103 [6,14,51,52]. Of note, RA has been reported to act in synergy with TGF-β in upregulating CD103 expression on T cells [53]. We find that TGF-β also upregulates CCR5 [22,24]; however, this only occurs in the context of MAdCAM-1 costimulation, underscoring the unique way in which MAdCAM-1 promotes gut T_RM_ differentiation. Several groups report that, with respect to CD8^+^ T cells, TGF-β and RA work synergistically in promoting tissue residency [17,18]. A similar dynamic likely holds true for CD4^+^ T cells. Because TGF-β exerts a suppressive effect on CD4^+^ T cells, MAdCAM-1 -mediated differentiation into T_RM_ cells proceeds despite limited cell division (Figs 2 and S3). This is consistent with the unique milieu of the gut which favors a regulatory environment that controls inflammatory responses [19,20,31].
We previously reported that the CD4^+^ T_RM_ cells we generate in vitro are highly susceptible to HIV infection, due in part to the expression of α_4_β_7_ and CCR5 [23,24]. These cells, which present a phenotype that favors retention in gut tissues, show a reduced proliferative capacity, which may facilitate the formation of latently infected cells that constitute persistent HIV-1 reservoirs. It has been established that long-lived reservoirs form in the gut within the first weeks of infection. In the acute phase of infection, α_4_β_7_^high^/CCR5^+^ CD4^+^ T cells are preferentially infected [54]. High-level viral replication in gut inductive tissue is a second feature of acute infection [55]. As such, we speculate that infected α_4_β_7_^+^ cells differentiate into T_RM_ cells and subsequently establish residency in gut tissues. Of note, galunisertib, a TGF-β antagonist, reduces the size of viral reservoirs in vivo [56]. Additionally, vedolizumab, tocilizumab and JAK inhibitors may provide novel ways to prevent the formation of long-lived HIV reservoirs that reside in gut tissues [57].
Our results, as described above, indicate that factors other than IL-6 (e.g., IFN-β) can also promote T_RM_ cell differentiation, and we speculate that these factors promote the differentiation of distinct T_RM_ subsets. However, we have not yet identified the full complement of these factors and this represents a limitation of this study. Additionally, we assume that memory CD4^+^ T cells are the source of IL-6, however we cannot exclude the possibility that other cells present in gut tissues provide an alternative source in vivo.
In conclusion, in this report we showed that MAdCAM-1 can prime naïve CD4^+^ T cells to differentiate into cells presenting a gut T_RM_ phenotype (S6 Fig). We further show that IL-6 can promote this process. We previously determined that vedolizumab, an α_4_β_7_ antagonist employed as a therapeutic for inflammatory bowel diseases (IBDs), can suppress the formation of MAdCAM-1 -generated T_RM_ cells [24]. Here we report that an IL-6 antagonist, tocilizumab, along with JAK inhibitors can also target T_RM_ cell differentiation. These findings point to new therapeutic approaches that can potentially target immune responses in gut tissues.
Methods and materials
Ethics statement
All primary CD4^+^ T cells used in this study were isolated from PBMCs collected anonymously from healthy donors from a National Institutes of Health Department of Transfusion Medicine protocol approved by the Institutional Review Board of the National Institute of Allergy and Infectious Diseases, National Institutes of Health. All study participants provided written informed consent.
Human blood processing and CD4+ T cell isolation
PBMCs were first isolated from the whole blood using lymphocyte separation medium (MP Biomedicals, Santa Clara, CA, USA). PBMCs then underwent negative isolation for both bulk CD4^+^ T cells and purified naïve CD4 T cells (CD45RO^-^/CD45RA^+^)(Stem Cell Technologies, Vancouver, Canada). Following isolation, purity of CD4^+^ isolation and percentage of CD45RO/CD45RA subsets were determined by flow cytometry.
CD4+ T cell co-stimulation
Recombinant human MAdCAM-1-Ig and VCAM-1-Ig were purchased from R&D Systems (Minneapolis, MN, USA). Anti-CD28 mAb was purchased from Invitrogen (Carlsbad, CA, USA). Co-stimulatory ligands were biotinylated per the manufacturer’s instructions using a LYNX Rapid Plus Biotin (type 2) Antibody Conjugation Kit (Bio-Rad, Hercules, CA, USA). All co-stimulation assays were carried out as previously described with the following modifications [22,23]. 96-well flat bottom cell culture-treated plates (Costa cat. # 3596) were pre-coated with 10 ng of anti-CD3 (clone OKT3) (eBioscience, San Diego, CA, USA) at 4°C for 2 hours, followed by 200 ng of neutravidin (Invitrogen, Waltham, MA, USA) at 4°C overnight in 100 µL of sterile HEPES-buffered saline (HBS). After rinsing, 200 ng of biotinylated co-stimulatory ligand was added for 1 hour at 37°C. 2 x 10^5^ purified CD4^+^ T cells were then added to the coated wells in complete Roswell Park Memorial Institute (RPMI) 1640 medium with 2% L-glutamine-penicillin-streptomycin and 10% fetal bovine serum (FBS) (both from Gibco Laboratories, Gaithersburg, MD, USA) (2 × 10^6^ cells/mL) at 37°C, 5% carbon dioxide. Depending on the experimental condition the following soluble factors were added on either day 0 or day 4: 10nm all-trans RA (Sigma-Aldrich, St. Louis, MI, USA), TGF-β (1ng/mL), IL-6 (10ng/mL), IFN-γ (10ng/mL), IL-10 (10ng/mL), TNF-α (10ng/mL), tocilizumab (5μg/mL), ruxolitinib (1μM), AZD1480 (1μM), IFN- β (10 ng/ml). On day 4, cultures were replaced with fresh complete media. For time-course experiments, 5 wells of naïve CD4^+^ T cells (2 × 10^6^ cells/mL) were prepared per experimental condition. Cells were removed at day 1–5 with no washes during this period.
Antibodies and flow cytometry
A catalogue of mAbs used in this study is provided in S1 Table. Antibody staining for flow cytometry was carried out in 2% FBS in 1 × PBS and employed standard protocols. Data were collected on a FACS Symphony A3 (BD Biosciences, San Diego, CA, USA) and analyzed using FlowJo (BD Life Sciences).
Proliferation by dye-dilution
Naïve and memory CD4^+^ T cells were negatively selected (Stem Cell Technologies, Vancouver, Canada) from PBMCs. The naïve population was labeled with 0.5μΜ of CellTrace CFSE while the memory population was labeled with 4μM of CellTrace Far Red. Fluorescent dyes were purchased from the CellTrace Cell Proliferation Kit (Invitrogen, Carlsbad, CA, USA). Stained cells were recombined into a ratio of [2] CFSE-Naïve to [1] Far Red-Memory, and then plated per costimulation methods listed above.
Western blot assay
CD45RO^-^ naїve cells were isolated and incubated with supernatants from bulk CD4^+^ T cells co-stimulated with MAdCAM-1 and RA for 4 days. Cell pellets were harvested at desired time points (10, 20, 30, 60 minutes) post stimulation with MAdCAM-1 and RA supernatant or IL-6, washed with PBS, and frozen. The 0 min condition received no RA, CD3 mAb, MAdCAM-1 or treatment. Cells pellets were analyzed as previously described [58]. Cell lysates were prepared using lysis buffer (20mM TRIS, 100mM NaCl, 1mM EDTA, 1% Triton X-100, 1mM DTT) with protease and phosphatase inhibitors (Thermo Scientific, Waltham, MA). 5 μL of each sample were loaded into wells of 10–20% Tris-Glycine SDS gels (Thermo Scientific, Waltham, MA) and run under reducing conditions, followed by transfer onto nitrocellulose membranes (Thermo Scientific, Waltham, MA) using the iBlot2TM system (Invitrogen, IB 1001). Antibodies used in Western blot are specified in S1 Table. Membranes were developed using SuperSignal West Femto Maximium Sensitivity Kit (Thermo Scientific, Waltham, MA) with images visualized by an LAS-3000 imaging system (GE Healthcare Biosciences). Post-exposure image processing was limited to linear changes in brightness, contrast, and color balance that were applied to the entire image.
Statistics
Statistical analyses was conducted using GraphPad Prism software (GraphPad Software, Inc., La Jolla, CA, USA). For (Fig 1A), equations were derived using a linear fit and nonlinear fit (second order polynomial) analysis. For analyses comparing 2 experimental groups (Fig 3A, 3B, and 3C), a Wilcoxon matched pairs signed ranked test was performed. For analysis involving 3 or more experimental groups, a Kruskal-Wallis test post-hoc Dunn’s multiple comparisons test was performed. For the CFSE proliferation and RA + IL-6 time course analysis, we performed a two-way ANOVA by comparing the means of the different treatment conditions (i.e., MAdCAM-1 + RA vs MAdCAM-1 + RA + TGF-β).
Supporting information
S1 FigInefficient formation of CD69^+^/CD103^+^ T_RM_s following MAdCAM-1 costimulation of purified naïve CD4^+^ T cell cultures.(A) Representative flow cytometry dot plots of PBMC derived naïve (upper) and bulk (lower) CD4^+^ T cells costimulated with MAdCAM-1 (+ RA and TGF-β). Y-axis indicates CD69, x-axis indicates CD103. Frequencies of CD69^+^/CD103^+^ cell as indicated.(TIF)
S2 FigVCAM-1 and CD28 mAb exhibit stronger costimulatory signaling and viability compared to MAdCAM-1.(A) Representative flow cytometry dot plots and lymphocyte gating of bulk CD4^+^ T cells that were costimulated with MAdCAM-1, VCAM-1 or CD28 mAb (+ RA and TGF-β). (B) Flow cytometric analysis of the lymphocyte gated population compared between the costimulatory ligands (n = 10). Error bars indicate standard deviation (*: p < 0.05, **: p < 0.01, ***: p < 0.001, ****: p < 0.0001).(TIF)
S3 FigNaïve cell proliferation and gut surface marker expression following MAdCAM-1 costimulation.Flow cytometric dot plots from a representative donor naïve CD4^+^ T cell proliferation by dye dilution (y-axis) and (A) CD103, (B) CD69 and (C) CCR5 expression (x-axis) following MAdCAM-1 costimulation in the presence of RA and TGF-β. Division 0 (d0) is indicated by a black arrow. Division index as indicated.(TIF)
S4 FigIL-6R expression levels in CD4 + bulk cultures.(A) CD45RO/CD45RA gating by flow cytometry from a representative donor. (B) IL-6Rα levels of each gated population. B) IL-6Rβ levels of each gated population.(TIF)
S5 FigChanges in protein expression between D0 Naïve T cells and D5 T cells after treatment with RA + IL-6.Measured proteins include (A) α_4_β_7_, (B) CCR5 and (C) CCR9. The gates used in Fig 3E are indicated. All plots are from the same representative donor.(TIF)
S6 FigSchematic of naïve CD4^+^ T cells priming by MAdCAM-1 in the presence of IL-6, Retinoic Acid and TGF-β.Schematic representation of the differentiation of naïve CD4 ⁺ T cells primed through engagement of α_4_β_7_ by MAdCAM-1. Integrin signaling in the presence of Retinoic acid (RA), TGF-β, and cytokine-driven activation of the JAK/STAT pathway promotes tissue-resident memory T cell (T_RM_) differentiation with upregulation of CD69 and CD103 (α_E_β_7_).(TIF)
S1 TableList of antibodies used in the study.(PDF)
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Sallusto F, Lenig D, Förster R, Lipp M, Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999;401(6754):708–12. doi: 10.1038/44385 10537110 · doi ↗ · pubmed ↗
- 2Szabo PA. Axes of heterogeneity in human tissue-resident memory T cells. Immunol Rev. 2023;316(1):23–37. doi: 10.1111/imr.13210 37211646 · doi ↗ · pubmed ↗
- 3Farber DL, Yudanin NA, Restifo NP. Human memory T cells: generation, compartmentalization and homeostasis. Nat Rev Immunol. 2014;14(1):24–35. doi: 10.1038/nri 3567 24336101 PMC 4032067 · doi ↗ · pubmed ↗
- 4Masopust D, Soerens AG. Tissue-Resident T Cells and Other Resident Leukocytes. Annu Rev Immunol. 2019;37:521–46.30726153 10.1146/annurev-immunol-042617-053214 PMC 7175802 · doi ↗ · pubmed ↗
- 5Schenkel JM, Pauken KE. Localization, tissue biology and T cell state - implications for cancer immunotherapy. Nat Rev Immunol. 2023;23(12):807–23. doi: 10.1038/s 41577-023-00884-8 37253877 PMC 11448857 · doi ↗ · pubmed ↗
- 6Xu H, Zhou R, Chen Z. Tissue-resident memory T cell: Ontogenetic cellular mechanism and clinical translation. Clin Exp Immunol. 2023;214(3):249–59.37586053 10.1093/cei/uxad 090PMC 10719502 · doi ↗ · pubmed ↗
- 7Christo SN, Evrard M, Park SL, Gandolfo LC, Burn TN, Fonseca R, et al. Discrete tissue microenvironments instruct diversity in resident memory T cell function and plasticity. Nat Immunol. 2021;22(9):1140–51. doi: 10.1038/s 41590-021-01004-1 34426691 · doi ↗ · pubmed ↗
- 8Kumar BV, Ma W, Miron M, Granot T, Guyer RS, Carpenter DJ. Human tissue-resident memory T cells are defined by core transcriptional and functional signatures in lymphoid and mucosal sites. Cell Reports. 2017;20(12):2921–34.28930685 10.1016/j.celrep.2017.08.078PMC 5646692 · doi ↗ · pubmed ↗