Th2 single-cell heterogeneity and clonal interorgan distribution in helminth-infected mice

Th2 cells provide effector functions in type 2 immune responses to helminths and allergens. Despite knowledge about molecular mechanisms of Th2 cell differentiation, there is little information on Th2 cell heterogeneity and clonal distribution between organs. To address this, we performed combined single-cell transcriptome and TCR clonotype analysis on murine Th2 cells in mesenteric lymph nodes (MLN) and lung after infection with Nippostrongylus brasiliensis (Nb) as a human hookworm infection model. We find organ-specific expression profiles, but also populations with conserved migration or effector/resident memory signatures that unexpectedly cluster with potentially regulatory Il10posFoxp3neg cells. A substantial MLN subpopulation with an interferon response signature suggests a role for interferon-signaling in Th2 differentiation or diversification. Further RNA-inferred developmental directions indicate proliferation as a hub for differentiation decisions. We also link long noted Cxcr3 expression in the Th2 compartment to a population of Il4pos NKT cells. Although the TCR repertoire is highly heterogeneous, we identified expanded clones and CDR3 motifs. Clonal relatedness between distant organs confirmed effective exchange of Th2 effector cells, although locally expanded clones dominated the response. These results provide new insights in Th2 cell subset diversity and clonal relatedness in distant organs.


34
Th2 cells are part of the adaptive immune response against helminths and in allergic diseases.

35
They are recruited and differentiate from a pool of naïve CD4 T cells with a wide variety of T cell 36 receptors (TCR) that are formed during T cell development and provide clonotypic specificity to 37 antigens. Differentiated Th2 cells produce the key type 2 cytokines IL-4, IL-5, and IL-13 that elevate 38 type 2 immune responses and thereby promote allergic inflammation but also mediate protection 39 against helminths (Walker & McKenzie, 2018). In recent years, several IL-4 producing Th2   (Urban et al., 1992). L3 stage larvae are injected subcutaneously and then first migrate into the 3 lung before they are coughed up, swallowed and reposition to the small intestine where they mature 61 to adult worms. (Urban et al., 1992). Using this model, we have previously shown that Nb infection 62 induces a Th2 response with a broad T cell receptor (TCR) repertoire required for effective worm 63 expulsion (Seidl, Panzer, & Voehringer, 2011). Development of single cell sequencing technology 64 now allowed us to gain a deeper understanding of Th2 cell subsets, TCR clonality and tissue 65 distribution.

66
Here, we performed single-cell sequencing of T cell receptor (TCR) genes combined with 67 transcriptome profiling of Th2 cells isolated from lung and mesenteric lymph nodes (MLN) at day 68 ten after Nb infection. By this approach, we revealed heterogeneity and differentiation paths within

91
In order to restrict the analysis after sequencing to high quality Th2 cells, we included in total 92 4710 cells with detected, functional TCR αand β-chains that also passed our additional QC filters 93 (see methods section) (Suppl. Fig. 1). We used an unbiased high dimensional clustering approach

107
In line with enhanced effector function, most lung cells and some proximal MLN cells express the 108 gene for the IL-33 receptor ST2 (Il1rl1) that recognizes the alarmin IL-33 and induces IL-5 and IL-109 13 production. Mice that lack IL-33 are not able to effectively clear intestinal helminths, likely due 110 to defects in the T cell and ILC2 compartments (Hung et al., 2013). We also find an elevated gene 111 signature for IL-33-stimulated T cells in the lung, which suggests active signaling via the ST2

138
We first screened the cells for known markers of T helper cell subsets. The analyzed cells from 139 MLN and lung express the Th2 hallmark genes Il4, Gata3 and Stat6 (Fig. 2B). However, only the 140 L1 cluster expresses IL-5 which promotes eosinophil development, recruitment and survival.

141
Similarly, IL-13 is expressed a bit broader in L1 but additionally in MLN4. IL-13 elicits a broad 142 spectrum of effector type 2 immune functions including eosinophilic inflammation, mucus secretion

144
According to the pro-inflammatory IL-5 and IL-13 production, double producers are thought to be a 145 strong or pathogenic effector subset of Th2 cells that includes highly differentiated CD27 low , PD-146 1(Pdcd1) high memory cells (Upadhyaya et al., 2011), which is also reflected on gene expression 147 level in our data. Enhanced Rgs16 expression of the IL-5 + /IL-13 + cells is also associated with higher

159
The classical marker for Treg cells FOXP3 was hardly found on gene expression level in our

174
Cluster L2 and MLN5 visually form a "bridge" between the MLN and lung compartments. Indeed 175 cells in these clusters both express genes coding for CD62L (Sell), CCR7 and S1PR1 involved in

183
However, there are also differences between the lung-and MLN-associated "bridge" clusters. Cells

245
In conclusion, RNA velocity supports proliferation as a hub for differentiation in the Th2 246 compartment and supports that migratory Th2 cells rather leave secondary lymphatic organs and 247 enter peripheral organs while the reverse migration path is a rare event.  Next, we find that strongly expanded clones are effectively spread over organs (Fig. 3C). The organ also show enhanced TCR repertoire relatedness to that organ confirms significant inter-283 organ migration and that the UMAP efficiently displays real relatedness of clusters.

284
As a next step, we visualized the five most strongly expanded clones determined for each organ 285 separately or we combined counts to get the most strongly expanded clones in the total data set 286 (Fig. 3E). Members of such clones in the MLN tend to be preferentially found in the MLN1 cluster

325
In conclusion, expanded clones in the Th2 effector population show no evidence for preferential 326 usage of particular TCRα or TCRβ chains or chain combinations.

329
T cell antigen-specificity and affinity is mainly confined by variable regions of the TCR and specific 330 T cells are selected and expanded (Rock et al., 1994). We chose the ten most abundant CDR3 331 sequences on amino acid level that potentially include motifs relevant for anti-helminth immunity 332 and highlight their abundance in different organs and mice (Fig. 4C). TCR data of naïve peripheral 333 blood T cells served as a reference to identify germline-associated CDR3 regions. The most 334 abundant sequence CVVGDRGSALGRLHF was found in all samples (Fig. 4C), including the 335 peripheral blood and represents the CDR3 motif found in the invariant TCRα chain (Vα14-Jα18) of 336 NKT cells (Lantz & Bendelac, 1994). This chain is co-expressed with a variety of different TCRβ 9 chains (not shown) and cells that express such TCRs are primarily found in the innate-like/NKT 338 cluster (MLN6) ( Fig. 2A, C). As this cluster also contains some cells with expression of TCRγ chain

348
Of the remaining 9 most abundant CDR3 motifs of TCRα or TCRβ chains, seven are not found 349 in the naïve peripheral blood sample (Fig. 4C), which implies an increased probability for them to

371
In addition to CDR3 motifs found frequently in combinations (expanded clones), we also find 372 abundant CDR3 motifs combined with various other unique TCR chains. (present in several non-373 expanded clones) (Fig. 4D right panel). These could include CDR3 motifs that provide anti-Nb 374 specificity but failed to induce substantial expansion or accumulation of Th2 cells expressing such 375 TCRs.

376
In line with a slightly preferential usage of the Trbv1 gene in expanded compared to non-

478
TCR associated genes (VDJ and constant region genes for α, β, γ, δ chains) were excluded but 479 kept as metadata to avoid clustering by TCR genes. To be included, cells needed to be defined as            Suppl. Figure 3. CDR3 length of Th2 TCRα and TCRβ chains ten days post Nb infection. A) TCRα and TCRβ CDR3 amino acid length. Data of MLN and lung cells were combined. CDR3 length was determined separately for expanded and non-expanded clones. One cell of each clone was considered to avoid expansion bias. B) CDR3 amino acid length distribution for all used TCRα variable segments or C) all used TCRβ variable segments.