Munich Cluster for Systems Neurology
print


Breadcrumb Navigation


Content

Reply to 'Comment on: Repositioning TH cell polarization from single cytokines to complex help'.

Nat Immunol. 2022 Feb 21. doi: 10.1038/s41590-022-01142-0. Epub ahead of print. PMID: 35190719.

Authors/Editors: Tuzlak S, Ginhoux F, Korn T, Becher B.
Publication Date: 2022

Tuzlak et al. reply: We thank Jankovic et al.1 for their interest in our Perspective2. The authors suggest that our proposition of repositioning T helper cell classifications based on the help they provide would result in overlapping classifications as TH cells function on several target cells. Indeed, these are exactly the arguments that led us to propose that a ‘help-based’ classification would be physiologically meaningful. In contrast to naming cells ‘IL-xy producing TH-z-like cells’, the classification can be reduced to ‘TH cells that induce type 1 and 3 responses by IL-xy…’. Thereby, the complexity of TH cell phenotypes is captured, while simultaneously focusing on the physiological consequences of T cell help.

With our proposed TH cell grouping, we did not intend to diminish the worth of the past 35 years of TH research; the prevailing helper subset paradigm has hitherto been of great utility. Nonetheless, we believe that a classification based on the physiological and pathophysiological effects of TH cells is helpful in light of the increasing technical capabilities of phenotyping TH cells. Our proposal was not to dismiss the existing nomenclature, which could easily be incorporated into the proposed classification, but to relate the respective TH cells to a broader system in line with what was proposed by Eberl and Pradeu3. In other words, although it is still important to characterize the cytokine profiles of TH cells in detail, we propose that it makes sense to refrain from automatically pigeonholing TH cells based on the expression of the cytokines measured at one point in time. Instead, the nature of the antigen (for example, intracellular (type I) versus non-phagocytosable (type II)) or the site of the response (for example, epithelial barriers (type III)) defines the players that need to be helped and dictates the type of T cell help that is appropriate. Although we think that such a concept is meaningful (and in principle distinct), we agree that it will certainly not eliminate any overlap.

The references cited by Jankovic et al.1 do, in our opinion, strongly support a reconsideration of the ever-expanding TH nomenclature. In brief, Tibbit et al.4 analyzed single-cell RNA sequencing (scRNA-seq) data from 695 TH cells purified from the bronchoalveolar lavage in a dust mite model. They identified clusters resembling TH1 and TH2 cells based on their transcription factor profile. However, although both clusters expressed Il4 mRNA, Ifng mRNA was not detected in the ‘TH1 cluster’, which underscores our arguments not to base classification solely on cytokines.

Wu et al.5 compared the metabolic signature of TH17 cells from the lamina propria to inflammatory TH17 cells from the spinal cord of experimental autoimmune encephalitis (EAE) mice, which they found to have characteristics of both TH1 and TH17 cells, and were therefore termed “pathogenic TH17/TH1” cells.

Ma et al.6 also underscore our argument by showing the complexity of human TH2 cells and cytotoxic CD4+ T cells isolated from the nasal polyps of patients with chronic rhinosinusitis.

Lönnberg et al.7 used a computational approach to show that TH1 and follicular helper T (TFH) cells can develop from the same CD4+ T cell clone in a mouse model of malaria. In their follow up study (Soon et al.8), malaria-specific CD4+ T cells were analyzed in more detail by single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq). The authors realized that not a single gene accurately marked all TH1 or TFH cells. Furthermore, the ability to assign effectors to TH1 or TFH cells diminished along pseudotime analysis, arguing for a transient effector state of the different TH subsets. Ciucci et al.9 analyzed CD44hi T cells from the spleens of mice infected with lymphocytic choriomeningitis virus (LCMV) for 7 days by scRNA-seq. Within 2,782 CD4+ T cells, they identified TH1 cells, TFH cells, regulatory T (Treg) cells and memory T cells. However, a substantial percentage (23%) of CD4+ T cells could not be assigned to a specific signature, which highlights the difficulty in classifying CD4+ T cells to the existing nomenclature even by scRNA-seq from a single time point of a well-studied mouse model.

To understand allergen-specific TH cells, Seumois et al.10 analyzed activated CD40L-expressing TH cells isolated from peripheral blood mononuclear cells (PBMCs) of patients with asthma and control groups. In addition to TH1, TH2 and TH17 cells, the authors also identified a subset of TH cell that was enriched for type I and II interferon response genes and was therefore termed ‘THIFNR’, as well as three different clusters that were enriched for genes linked with cell activation and were termed ‘THACT1–THACT3’. This study clearly shows that the resolution we obtain by scRNA-seq is not compatible with the existing TH classification and rather enforces the definition of ever new TH subsets that are most likely reflecting a snapshot in time of transitional states or ‘mixed’ priming.

All these examples naturally identify TH cells that can be attributed to the current classification but at the same time they also clearly illustrate its limitations and the need to consider a more inclusive and physiologically meaningful view on TH cell function and the outcomes of their actions.

Finally, we would like to thank Jankovic and colleagues for having initiated this open discussion about the future of TH cell classification, which was the primary purpose of our perspective.

 

Related Links