Durable CD4⁺ T cell immunity: cherchez la stem

Abstract

Mammalian stem cells govern development, tissue homeostasis and regeneration. Following years of study, their functions have been delineated with increasing precision. The past decade has witnessed heightened widespread use of stem cell terminology in association with durable T cell responses to infection, anti-tumor immunity, and autoimmunity. Interpreting this literature is complicated by the fact that descriptions are diverse and criteria for labeling “stem-like” T cells are evolving. Working under the hypothesis that conceptual frameworks developed for actual stem cells can be used to better evaluate and organize T cells described to have stem-like features, we outline widely accepted properties of stem cells and compare these to different “stem-like” CD4⁺ T cell populations.

“Stemness” in stem cells and T cells

T cells are critical for pathogen defense and the elimination of malignant cells but can also cause tissue damage during uncontrolled infection and autoimmunity. Although the stem cell and T cell fields are distinct, there is increasing application of stem cell terminology to T cell subpopulations. Here we focus on CD4⁺ T cells, where the stem cell definition has been applied broadly to both effector and memory cells. We outline widely accepted properties of adult stem cells, then review relevant literature to identify commonalities and differences. We conclude that naïve CD4⁺ T cells (Tn) are most akin to a true stem cell. Stem cell memory and central memory T cells (Tscm and Tcm) also clearly possess “stem-like” features, but the ascription of stem-like properties to many other populations is frequently based on a limited number of stem cell-associated characteristics. We propose a means for a more accurate delineation of “progenitor-like” vs. “stem-like” T cells.

What makes a stem cell?

Although mammalian stem cells have been described extensively in the germline and in developing embryonic and extra-embryonic tissues, for purposes of comparison, we focus on adult somatic stem cells, e.g. of the gut, bone marrow, hair follicle, lung, and skeletal muscle. Amalgamating different tissues necessitates a high-level overview, with details and nuances omitted. More thorough reviews of individual stem cell compartments and their properties are available [1–4], but in general, adult somatic stem cells can be defined by several key features (Box 1). The main function of a stem cell is to generate differentiated cells that maintain tissue homeostasis and/or mediate regeneration. To do this continuously, they must be able to self-renew, i.e. divide and produce at least one daughter cell that retains stem cell features. Stem cells also home to specific niches, where they are sensitive to, and dependent on, signals such as Wnt and Notch that maintain self-renewal [4]. Beyond this, the stem cell definition is loose. For example, they can be either multipotent, giving rise to multiple cell lineages, or unipotent, giving rise to one. Regardless, they are developmentally plastic. Skeletal muscle stem cells for example, are unipotent and quiescent for long periods of time, but when needed, proliferate and differentiate into fused, terminally differentiated myotubes [4]. As a further complication, the field has moved to view the stem/differentiated distinction as more of a continuum [5,6]. This change has been accelerated by the application of single-cell RNA sequencing (scRNAseq) technologies, which showed that stem cell compartments are often neither homogeneous nor completely discrete from their differentiated counterparts.

Box 1: Seven key features of adult somatic stem cells.

1. Capacity to seed a tissue with more differentiated cells, termed multipotency or unipotency depending on the number of differentiated cell types. Muscle stem cells are unipotent, while hair follicle, intestinal, and bone marrow stem cells are multipotent [1–4].

2. Ability to rapidly induce silent but developmentally poised genes upon reception of differentiation signals.

3. Distinct metabolic parameters relative to more differentiated counterparts.

4. Long life, usually buttressed by the capacity for self-renewal, i.e. to express telomerase and undergo symmetric cell divisions that help maintain cell numbers.

5. Homing to specific niches.

6. Sensitivity to trophic and/or metabolic signals supplied by niche cells, e.g. Wnt/Notch.

7. Quiescence punctuated by asymmetric cell divisions that give rise to rapidly proliferating and differentiating daughter cells while preserving stem cells in the niche.

Stem cells reside in both stable and self-renewing tissues. In stable tissues undergoing less turnover, they are quiescent until required to replace damaged/lost cells. Stem cells in this situation tend to be more discrete and easily defined based on specific marker expression and scRNAseq [6,7]. In tissues with higher turnover such as the hematopoietic and gastrointestinal systems, stem cells fulfill multiple functions. They must continuously generate differentiated cells, self-renew to maintain themselves, and rapidly regenerate differentiated cell types to repopulate tissue following damage or other forms of stress. In mouse models of gastrointestinal damage, this latter activity has been shown functionally to reside in “reserve” stem cells (Figure 1) but can also arise through de-differentiation of more mature cells [8–10]. Based on their higher quiescence and mobilization during stressed conditions, reserve stem cells conceptually resemble stem cells in stable tissues subject to low turnover. Both reserve stem cells and stem cells in tissues with low turnover can be considered more “stem like” compared to cells that replenish tissue under homeostatic conditions (Figure 1). A “progenitor” label has sometimes been applied to populations that continuously generate differentiated cells, e.g. in the hematopoietic system (see below). This will become important when considering T cells below.

Figure 1. Model of typical differentiation in mammalian adult somatic tissues.

Cells with more stem-like features are shown in the center, and increasingly differentiated cells are shown radiating outward. The progression is generally from stem cell (e.g., Lgr5⁺, MMP2, and LT-HSC) to multipotent/unipotent progenitor, both of which can have self-renewal capacity (circular arrows), to lineage-restricted progenitor (also known as transit-amplifying cells) that are still proliferative but thought to be committed to differentiation in homeostatic conditions; and terminally differentiated cells that are largely post-mitotic but make up most of the biomass [1–4]. Although the diagram does not convey a temporal component, cells generally move from an inward to an outward position. This trajectory is overcome in certain conditions in which cells de-differentiate and acquire stem cell properties (red arrow). In actuality, the cells exist along continua, the concentric lines showing somewhat arbitrary cutoffs, and the depicted cells representing centroids of what are not appreciated to be more continuous and fluid populations. On the right is shown the stage at which certain stem cell attributes are progressively lost. For example, stem cells lose their virtually unlimited proliferative capacity relatively early in differentiation. This figure was created using Biorender.com

Evolving views of stem cells

The view of what constitutes a stem cell has been evolving. For example, Lgr5⁺ cells of the mouse and human small intestine were long viewed as the stem cell for this organ [11], but reserve stem cells marked by high clusterin expression were later described, replenishing tissue following damage [5]. Additionally, scRNA seq demonstrated that after radiation damage, differentiated mouse intestinal epithelial cells could de-differentiate into stem cells that replenished damaged tissue [8]. The historical maturation of hematopoietic stem cells (HSCs) followed a reverse trajectory, with the reserve cell type described first. Known as long-term repopulating HSCs (LT-HSCs), when transplanted, these cells stably recreate all hematopoietic lineages [12–15]. In mice, they can be tracked using combinations of markers enriched in stem cells such as c-kit, Sca1 and CD150, lack of lineage-specific markers such as Ter119, CD19, CD3 and NK1.1, or reporters of, e.g., Hoxb5 expression [12–15]. However, LT-HSCs elude rigid definition, with no combination of markers completely encompassing all functional LT-HSCs. Even optimal methods still capture cells lacking activity and fail to enrich cells with activity. The lack of rigidity in even the best-defined stem cell populations parallels the fluidity inherent in T cells. Recently, a population of murine multipotent progenitors (MPPs, specifically a population known as MPP2), originally considered to being committed to differentiation, supported hematopoiesis under homeostatic conditions, lacking any damage or transplant; the study was based on a combination of multiple assays including scRNAseq followed by trajectory analysis and lineage tracing [9,10]. Conceptually, these cells resemble Lgr5⁺ cells that maintain intestinal homeostasis. LT-HSCs are rightfully considered more “stem-like” than MPPs, but the presence of more differentiated cells that maintain hematopoiesis suggests a continuum of more and less stem-like differentiation states in many tissues (Figure 1). The separation of cells into stem and progenitor types in hematopoiesis represents a useful paradigm that could help clarify descriptions of T cells, as detailed below.

Stem-like properties in memory CD4⁺ T cells

The first use of stem cell terminology in the context of T cells was with graft-vs-host disease (GVHD), termed T stem cell memory cells (Tscm) [16] (Figure 2). In humans and mice, they differentiate from naïve T cells (Tn) upon antigen encounter but retain some naïve markers, e.g. CD44^loCD62L^hi. They can be distinguished from Tn by CD95, CD122, and CXCR3 expression, suggesting antigen experience [16,17]. As in cancer and autoimmunity, antigen persists in GVHD and the “memory” moniker does not apply by strict functional definitions [18]. Nevertheless, CD8⁺ and CD4⁺ central memory cells (Tcm) are specified soon after acute infection when antigen is still present, suggesting something similar may happen in GVHD to form memory-like Tscm [19,20]. Recent mouse experiments using reporters and scRNAseq establish that CD4⁺ Tscm express the transcription factor TCF1 [21].

Figure 2 (Key Figure). Model of mammalian CD4⁺ T cell differentiation.

Naïve T cells (Tn) are shown at the center and terminal effectors at the bottom. In between are different populations, e.g. T stem cell memory (Tscm), T helper 17 cells with stem-like properties (stem-like Th17), central memory T cells (Tcm), long-lived T follicular helper cells (long-lived Tfh), effector memory T cells (Tem) and resident memory T cells (Trm) with intermediate levels of differentiation. As with Figure 1, the diagram does not show a specific temporal situation, though cells generally move downward with time to form regulatory and effector cells, e.g. regulatory T (Treg), T follicular helper (Tfh), T helper 17 (Th17), T helper 2 (Th2) and T helper 1 (Th1) cells. Exceptions to this trend (red arrows) potentially include subsets of effector cells that may de-differentiate early during acute infection and form memory, i.e. memory precursor T cells (Tmp), and populations of Trm-like cells with a potential to give rise to Tem and Tcm. The question mark is present because these pathways are better demonstrated in CD8⁺ T cells and therefore here, are based on analogy and somewhat speculative in CD4⁺ T cells. As with stem cells, the T cell populations exist along continua, the concentric lines showing arbitrary cutoffs, and the depicted cells representing centroids of continuous populations. This figure was created using Biorender.com

CD4⁺ Tscm descriptions are also reasonably established in tumor immunity where they are marked by expression of TCF1, the anti-apoptotic molecule Bcl2, the chemokine receptors CXCR3 and CCR7, and by their ability to differentiate into classical memory and effector populations [17,22]. In mouse models of melanoma adoptive cell therapy, flow cytometric analysis identified antigen-specific CD4⁺ Tscm cells localized to the tumor draining lymph nodes that produced effector populations which infiltrated the tumors and mediate antitumor responses [23]. Moreover, CD8⁺ and CD4⁺ Tscm-like cells can be enriched directly from Tn in vitro using TCR, IL-7, and IL-15 stimulation, suggesting that Tscm may be highly dependent on survival signals [23,24].

Enrichment of cells akin to CD4⁺ Tscm has been described in a variety of autoimmune diseases such as rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and type-1 diabetes, based on expression of markers such as CCR7, CD27, CD62L, CD95 and Ki67 [25–30]. In some cases, the presence of these cells has been associated with disease severity. For instance, stimulated CD4⁺ Tscm derived from the peripheral blood of SLE patients demonstrated self-renewal and capacity to differentiate into memory and effector T cell subsets capable of eliciting strong autologous B cell antibody responses, a hallmark of SLE [26]. In humanized mouse models of autoimmune vasculitis, “stem-like” CD4⁺ T populations marked by TCF1 and IL-7R expression survived serial transplantation, homed to tertiary lymphoid structures surrounding diseased arteries, and generated effectors with increased Ifng and Gzmb mRNA expression [31]. In a CD4⁺ T cell transfer mouse model of colitis, stem-like CD4⁺ T cells marked by the expression of Tcf7, Bach2, and Id3, initiated and sustained disease, as evidenced from histological analysis of intestinal inflammation. In contrast, IFNγ-expressing effector cells could not recapitulate disease upon transplant [32], suggesting that though effector cells can directly mediate disease, sustained activity may require smaller populations of Tscm-like cells. While CD4⁺ T cells with stem-like properties (e.g. generating terminally differentiated effector populations) are present in multiple chronic diseases, their specific definitions vary with tissue and disease. Commonalities include TCF-1 expression and lymphoid homing. Further comparative metabolic and epigenetic characterization of these populations can identify how stem-like cells vary between normal and disease contexts, and if therapeutics targeting their common characteristics might be effective against certain pathologies.

Work on the formation of memory following acute infection has identified stem cell features in different long-lived memory T cell populations. The findings are again more developed for CD8⁺ T cells, but CD4⁺ cells have been stratified into different memory subpopulations based on their gene expression patterns, surface protein expression, and tissue homing [33]: Tscm, Tcm, effector memory (Tem), and tissue-resident memory (Trm, Figure 2). Flow cytometric and RNAseq analysis comparing these populations indicated that these cells progressively correlate with diminished “stemness” and proliferative potential by decreasing the expression of naïve-associated genes such as Lef1 and Foxp1 [17,34]. Mouse and human Tcm for example, have high replicative capacity following restimulation with cognate antigen, yet also maintain themselves through infrequent homeostatic proliferation in the absence of stimulation [35,36]. These properties are shared with somatic stem cells (Table 1). Tem by contrast, have an intermediate phenotype, with faster effector function (minutes not hours) compared to Tcm [33]. Under specific conditions, mouse CD8⁺ Tem possess unlimited proliferative capacity [37]. Whether CD4⁺ Tem have similar proliferative potential is unknown. CD103⁺ fate-mapping experiments indicated that differentiated mouse CD8⁺ Trm have low proliferative potential and plasticity upon secondary infection in the small intestine [38,39], though Hobit transcription factor fate-mapping and adoptive transfer showed that some of these cells may retain high proliferative capacity, reenter circulation, and differentiate into Tem, Tcm, and Trm upon antigenic rechallenge [40–42]. While currently unknown, something similar may occur in CD4⁺ T cells (Figure 2, red arrow).

Table 1.

Stem-like properties of select CD4⁺ T cell populations

Stem cell features
Cell type	Lineage plasticity	Self renewal	Metabolism	Poised genes	Niche homing	Signal sensitivity	Proliferative potential	Refs
Tn	+++	++	+++	+++	++	+++	+++
Tscm	++	++	+	++	+	+++	+++	17, 34
Th17 stem-like	++	+	+	++	+	++	++	43, 52
Tcm	+	++	++	++	+	++	++	35, 36
“Stem-like”
“Progenitor”
Long-lived Tfh	+	+	-	+	+	++	++	55
Tprog	+	+	-	+	+	++	++	62
Tem		+	-	+	+	+	+	33
Trm	-	+	-	+	+	+	+	33
Tmp	++	-	ND	++	+	ND	++	19, 20
Teff		-	-	-		-

¹ND = not determined

Stem-like Th17 cells

Different human and mouse Th17 cell subpopulations manifest long life, self-renewal, and expression of Tcf7 and Bcl2 genes associated with self-renewal and survival (Figure 2) [43,44]. Ex-vivo polarization of CD4⁺ Tscm cells suggests that these stem-like Th17 cells may be a CCR6⁺ subset of Tscm, as only this CCR6⁺ population could produce IL-17 after Th17 polarization [45]. Th17 cells expressing stem-associated markers can differentiate into multiple T helper subsets including T helper 1 (Th1), T helper 2 (Th2), type I regulatory (TR1), T follicular helper (Tfh) and regulatory T (Treg) cells, highlighting their plasticity [46–49]. Th17 cells play complex roles within the tumor immune response because their developmental plasticity and flexible cytokine secretion allows for both pro- and antitumor activities [50].

In experimental autoimmune encephalomyelitis mouse models of multiple sclerosis (e.g. experimental autoimmune encephalomyelitis (EAE)), an extensive scRNAseq study showed that homeostatic intestinal stem-like Th17 cells, defined by Tcf7, Il17, and Slamf6 expression, could become pathogenic in response to IL-23, access the central nervous system (CNS), and mediate disease [51]. These cells are enriched within CNS-draining lymph nodes and possess open chromatin signatures at Tcf7, Ccr7, and Sell loci [52]. Similarly, comparing genome-wide H3K4 and H3K27 histone trimethylation in naïve CD4⁺ T cells with in vitro differentiated T helper subsets showed increased epigenetic plasticity in Th17 compared to Th1 and Th2 cells, suggesting that Th17 cells are less lineage-restricted [53]. Flow cytometric profiling of CD4⁺ peripheral blood Tscm from RA patients unveiled a CD4⁺ Tscm Th17 polarization bias, highlighting similarities between stem-like Th17 and Tscm [25]. Compared to Th17 effectors, stem-like Th17 cells also show decreased mTORC1 activity and Myc expression, as well as decreased anabolic metabolism (from metabolic profiling) [54]. However metabolic comparison of stem-like Th17 cells to Tscm and Tn has yet to be conducted. Further research into the plasticity, longevity, proliferative capacity, and metabolism of these cells can reveal distinct and overlapping roles for these populations in disease.

Long-lived Tfh cells

Recent work described long-lived Tfh cells marked by Tcf7, folate receptor 4 (FR4), and high CD27, and that are transcriptionally distinct from Tcm [55]. These cells undergo self-renewal and display limited lineage plasticity, generating both Th1 and Tfh effectors (Figure 2) [55]. Long-lived Tfh are highly glycolytic with high mTORC1 phosphorylation and glucose uptake (a metabolic profile typically associated with effectors), but which can survive for long periods of time without antigen stimulation, instead requiring ICOS signals to maintain their identity [55]. The presence of both long-lived Tfh and distinct Tcm with Tfh features [56] reflects the fluidity of CD4⁺ T cell states and suggests that CD4⁺ T cells exist along a continuum of differentiated states with periodic stable populations.

Stem-like T cells vs. somatic stem cells

Comparing the properties of CD4⁺ T cell populations to bona fide stem cells (Figure 1), Tn emerge as the most stem-like (Table 1, Figure 2). Tn are the primary antigen-recognizing sentinels of T cell immunity [57]. They home to the T cell zones of secondary lymphoid tissues and are largely quiescent and long-lived (particularly in humans); however, they have potentially unlimited proliferative capacity [35,36,57]. In the absence of stimulation, they undergo infrequent homeostatic proliferation to maintain their numbers [35,36]. They rapidly remodel their metabolism upon antigen encounter and activation [58]. Lastly, although committed to the T cell lineage and lacking multipotency, they are the most developmentally plastic, giving rise to all T cell types [57]. They thus closely resemble unipotent stem cells.

Although current evidence places antigen-experienced CD4⁺ T cell subsets in more differentiated positions relative to Tn (Figure 2), Tscm, stem-like Th17 and Tcm also have clear stem-like features. Human Tscm for instance, accumulate beta-catenin and are developmentally plastic yet have increased proliferation and cytokine expression compared to Tn upon stimulation [17,59]. Tcm, like naïve cells, home to secondary lymphoid organs and have potentially unlimited proliferative capacity upon restimulation [60]. Parallels between CD8⁺ memory T cell self-renewal and hematopoietic stem cells have been highlighted before [61], and further characterization of stem-like CD4⁺ T cell populations and comparison with bona fide stem cells will likely reveal additional similarities.

The evidence above places many other “stem-like” cell types in more differentiated positions relative to Tn, Tscm/stem-like Th17, and Tcm (Figure 2). By analogy with the hematopoietic system’s MPP, a “progenitor” or “precursor” descriptor may be more accurate for these cells. By definition, MPPs produce multiple differentiated cell types. They also have varying stem-like features [9,10], with most committed to differentiation while some indefinitely sustain homeostasis. This terminology has been applied to memory precursor effectors (Tmp, Figure 2), which arise in mice early during acute lymphocytic choriomeningitis virus (LCMV) infection and can give rise to memory [19,20]. Tmp are by nature transient (Table 1). They may be most akin to cells that dedifferentiate from differentiated cells to repopulate stem cells under conditions of tissue damage (Figure 1). A precursor terminology has also been applied to CD4⁺PD-1⁺TCF1⁺ progenitor cells with self-renewal capacity that arise during chronic infection and continuously generate both Tfh and Th1 effectors [62]. These cells rapidly dissipate upon antigen loss [62]. While long-lived, Tfh cells express stem-associated genes and display some plasticity, their glycolytic metabolic profiles place them in a more differentiated and activated position compared to Tcm [55]. With careful application, the use of “progenitor” rather than “stem cell” terminology applied to cells such as long-lived Tfh, might help clarify the field.

Linking stem-like T cell populations to true stem cells Journal Pre-proof is confounded by the fact that both populations are fluid and heterogeneous. Within the last several years, more long-lived Tscm, Tfh, and Th17 subpopulations have all been associated with stem-like gene expression [43,51,52,63]. Although the more stem-like subsets show characteristics similar to Tn, they all have more limited lineage plasticity and appear more differentiated in terms of metabolism, tissue homing, and functional capacity [43,51,52,63]. Tscm, stem-like Th17, and Tcm, share many characteristics with Tn, justifying a “stem-like” terminology. Other populations such as long-lived Tfh, Tem, and Trm, possess some stem-like characteristics, but differ in one or more key stem-like features warranting, in our view, a “progenitor” moniker (Table 1).

Concluding remarks

Table 1 provides a demarcation between CD4⁺ T cell subsets that warrant a stem cell-like description vs. subsets that are better described as progenitors. The fluidity of these cell types and lack of certain functional and comparative analyses may ultimately require refinement in this demarcation. Based on their common features with stem cells, we consider Tn to be the most stem-like CD4⁺ T cell, with Tscm, stem-like Th17 cells, and Tcm, also showing significant similarity to adult stem cells. Long-lived Tfh appear to be more developmentally restricted and effector-like, indicative of a progenitor- rather than stem-like cell. Tem and Trm are relatively long-lived and, potentially, proliferative, but appear to be more lineage-restricted with epigenetic bias towards a specific T helper lineage, though some Tem and Trm may retain high proliferative capacity and plasticity. These findings highlight the fluidity of T cells and the difficulty in applying rigid definitions to what increasingly appears to be a continuum of states [64]. Even subsets of effector T cells may de-differentiate into memory cells, at least at early stages of infection (Figure 2, red arrow, [19,20].

Key unanswered questions in the field include what activities endow stem- and progenitor-like cells with different degrees of lineage plasticity and proliferative potential, and how these properties are regulated. In both mouse and human, the transcription factor TCF1 appears to be an important regulator, but its mechanisms are incompletely understood in peripheral T cells (see Outstanding questions). There may also be unique chromatin signatures and epigenetic states in these cells that endow them with these characteristics, and emerging work supports this notion [52,53]. The next few years will likely reveal exciting advances in these areas.

Outstanding questions.

What pathways and mechanisms endow T cells with varying stem-like qualities? Stem cells respond to trophic signals, e.g. Wnt and Notch. In T cells, Tcf7 expression is frequently used to define a T cell as “stem-like”. Tcf7 encodes TCF1, a terminal effector of the Wnt pathway, but work in T cells thus far indicates that engagement of the Wnt pathway in T cells is limited. In the absence of Wnt, TCF1 is largely repressive through interactions with TLE-class co-repressors. The overlap between TCF1’s functions in stem cells and stem-like T cells, and the roles played by Wnt signals and TLEs are poorly understood but are being investigated [65]. In the case of Notch, induced Tscm (iTscm) with high proliferative capacity, long-term survival, and potent antitumor activity have been described, and are generated through culture of activated CD4⁺ T cells with OP9 cells expressing the Notch ligand DL1 plus anti-IFNy and IL-7 [66].

Are there molecular cues that promote differentiation of stem-like T cells into more differentiated populations? Stem-like Th17 cells for example can mature into pathogenic and non-pathogenic subsets based on recognition of IL-23 signals for example [51]. These are now known to regulate expression of another transcription factor, Bach2 [52]. A similar level of detail is needed for other populations.

How do the characteristics, origins, destinies, and lineage relationships of stem-like T cells change in neonates and during aging? Work has been conducted on general T cell populations including CD4⁺ T cells, memory cells, etc., but with specific stem-like populations.

Lastly, can stem-like T cell populations be specifically targeted for clinical benefit? This remains an open question.

Highlights.

• Different CD4⁺ T cell subsets have been described to have “stem-like” features, frequently without a clear indication of what that means functionally or devoid of a systematic set of criteria for how to apply the terminology.

• Bona fide stem cells are themselves varied and their definitions have been evolving recently.

• CD4⁺ T cell types with clear stem-like features are naïve cells and subsets of memory cells (central memory, Tcm; and stem cell memory, Tscm). These also display stability over time, specific niche homing, high replicative potential, and the capacity to differentiate into many other T cell types. Other cells fall along a continuum of more vs. less stem-like features.

• In some cases, more functional characterization of these cells is required before appending a “stem-like” moniker is justified. For others, substituting “progenitor” or “precursor” terminology along the lines of the hematopoietic system is likely to be more accurate.

Significance.

T cells with stem-like features mediate durable immunity to pathogens and tumors but can also promote tissue damage. The complexity of the literature labeling T cells as “stem-like” suggests that both the criteria for ascribing stem cell character to T cells, and the nomenclature used to relate them to one another, need refinement. This would help systemize the field, allowing better inference of function from name.

Glossary

Graft-vs-host disease

Pathology in allogenic transplant patients where a cell from a bone marrow graft recognizes the host tissues as foreign and attack them.

Lgr5

R-Spondin receptor that stabilizes Frizzled receptors and Wnt signaling; regulates somatic stem cell function in multiple compartments such as gut and hair follicle.

Long-term repopulating HSCs

population of mammalian cells residing in the bone marrow in adult mammals, defined by their ability to indefinitely reconstitute all blood cell lineages upon transplant to new hosts.

Stem-like T cells

Populations of T cells broadly but variably analogous to different stem cell populations based on characteristics such as developmental plasticity and proliferative potential. Their varying descriptions make them difficult to easily define.

Reserve stem cells

Also termed “facultative” or “revival” stem cells, a population of highly quiescent stem cells that are largely inert in homeostatic conditions but become mobilized and repopulate tissues during conditions of stress or damage.

Tn

CD4⁺ or CD8⁺ T cell that has undergone thymic development and selection but has not yet encountered cognate antigen in the periphery. Tn reside within secondary lymphoid organs and are marked by the expression of CCR7, CD62L, and IL7R, and lack expression markers associated with current or previous activation.

Tscm

CD4⁺ or CD8⁺ T cell that shares many markers with Tn but also shows hallmarks of activation e.g. CD45RO; often distinguished from Tn cells by CD95 expression. Tscm might be an antigen-experienced precursor to all other T cell subsets.

Tem

resting CD4⁺ or CD8⁺ memory T cell with a more activated phenotype compared to Tcm, often with more immediate proliferative and cytokine expression potential upon antigen reencounter. Frequently defined by high CD44 expression and low CD62L and CCR7 expression.

Trm

CD4⁺ or CD8⁺ memory T cell residing within organ tissue without reentering circulation under static conditions. Trm rapidly proliferate and secrete cytokines upon reactivation; act as a first line of defense by the adaptive immune system to pathogen reencounter.

Stemness

describes several molecular processes shared among stem cell populations such as self-renewal and the ability to generate daughter cells with more differentiated characteristics (Table 1).

Homeostatic proliferation

process of T cell proliferation in the absence of TCR engagement. Often balances with normal attrition to keep cell numbers stable in homeostatic conditions.

Th17 cell

CD4⁺ T helper subset generally characterized by IL-17 cytokine secretion and expression of the Th17 master transcription factor Rorγt. They are considered more plastic vis-à-vis Th1 and Th2 cells; provided the right stimuli, can adopt the cytokine and gene expression of several other T helper subsets.

Th1

CD4⁺ T helper subset typically characterized by IFNγ cytokine secretion, the expression of the Th1 master transcription factor Tbet; associated with macrophage activation and inflammation.

Th2

CD4⁺ T helper subset typically characterized by IL-4, IL-5, and IL-13 cytokine secretion and the expression of the Th2 master transcription factor GATA3; associated with humoral immune responses.

TR1

regulatory CD4⁺ T cell that is important for long-term immune tolerance; characterized by IL-10, TGF-β, and IFN𝛄 cytokine secretion, and the transient expression of FOXP3, only upon activation and bearing the surface markers CD49b and Lag3.

Tfh

CD4⁺ T helper subset typically residing within secondary lymphoid organs; assists in germinal center formation and supports B cell maturation. Tfh cells are often identified through the expression of the transcription factor Bcl-6 and the chemokine receptor CXCR5.

Treg

Inhibitory regulatory T cells, largely CD4⁺, that downmodulate immune responses through the action of membrane-bound and secreted mediators.

Tmp

transitory memory precursor T cell identified within the effector T cell population early in acute infection; thought to ultimately give rise to memory T cells.