Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Mar 2;217(3):jem.20190580.
doi: 10.1084/jem.20190580.

The human fetal thymus generates invariant effector γδ T cells

Affiliations

The human fetal thymus generates invariant effector γδ T cells

Paola Tieppo et al. J Exp Med. .

Abstract

In the mouse thymus, invariant γδ T cells are generated at well-defined times during development and acquire effector functions before exiting the thymus. However, whether such thymic programming and age-dependent generation of invariant γδ T cells occur in humans is not known. Here we found that, unlike postnatal γδ thymocytes, human fetal γδ thymocytes were functionally programmed (e.g., IFNγ, granzymes) and expressed low levels of terminal deoxynucleotidyl transferase (TdT). This low level of TdT resulted in a low number of N nucleotide insertions in the complementarity-determining region-3 (CDR3) of their TCR repertoire, allowing the usage of short homology repeats within the germline-encoded VDJ segments to generate invariant/public cytomegalovirus-reactive CDR3 sequences (TRGV8-TRJP1-CATWDTTGWFKIF, TRDV2-TRDD3-CACDTGGY, and TRDV1-TRDD3-CALGELGD). Furthermore, both the generation of invariant TCRs and the intrathymic acquisition of effector functions were due to an intrinsic property of fetal hematopoietic stem and precursor cells (HSPCs) caused by high expression of the RNA-binding protein Lin28b. In conclusion, our data indicate that the human fetal thymus generates, in an HSPC/Lin28b-dependent manner, invariant γδ T cells with programmed effector functions.

PubMed Disclaimer

Figures

None
Graphical abstract
Figure 1.
Figure 1.
Fetal γδ thymocytes express programmed effector functions. (A) γδ thymocytes (without the Vγ9Vδ2 subset) were sorted from FT (n = 3) or PNT (n = 3) samples, and gene expression profiles were determined by RNAseq. GSEA plots based on the RNAseq dataset of FT versus PNT γδ thymocytes with the gene sets of activated (act.; versus naive) human (GSE28726) and mouse (GSE10239) T cells (top) and the gene sets of IL-15–stimulated (stim.; versus unstimulated) human (GSE22886) and mouse (GSE7764) NK cells (bottom). Gene sets are from the C7 immunological gene signature from the MSigDB. (B) Volcano plot illustrating differentially expressed genes between FT and PNT γδ thymocytes. (C) Heatmap of selected genes in FT and PNT γδ and αβ thymocytes. (D) Flow cytometry analysis of granzyme A (top) and IFNγ (bottom) expression by FT and PNT γδ and αβ thymocytes. Each symbol represents an individual donor: square symbols indicate CyToF data, and round symbols for γδ thymocytes indicate γδ thymocytes without the Vγ9Vδ2 subset. Horizontal lines indicate the median, and data were analyzed by Mann Whitney U test. *, P < 0.05; **, P < 0.01. (E) Flow cytometry plots of FT and PNT γδ thymocytes (without the Vγ9Vδ2 subset) illustrating the granzyme A+ cells within CD1a cells (top) and IFNγ+ cells within CD1a cells (bottom); data in the top and bottom panels are obtained from different subjects. The top panels are representative of five (FT) and six (PNT) independent experiments. The bottom panels are representative of one (FT) and seven (PNT) independent experiments. NES, normalized enrichment score.
Figure S1.
Figure S1.
Flow cytometry data on FT and PNT thymocytes. (A) Flow cytometry plots of FT and PNT γδ thymocytes illustrating the expression of granzyme A+ cells (left) and IFNγ+ cells (right) within mature CD27+ cells. (B) Flow cytometry plots illustrating the inverse relation of the expression of CD27 and CD1a maturation markers in FT and PNT γδ thymocytes. (C) Flow cytometry analysis of CD1a (left) and CD27+ (right) expression by fetal and postnatal γδ and αβ thymocytes. (D) Cytometry plot (based on CyToF data) of fetal γδ thymocytes illustrating IFNγ+ cells within Tbethigh cells. Representative of five (FT) and six (PNT) independent experiments (A, left; and B) or representative of one (FT) and seven (PNT) independent experiments (A, right). Horizontal lines indicate medians. Flow cytometry plots (A and B) and data (C) represent FT and PNT γδ thymocytes without the Vγ9Vδ2 subset.
Figure 2.
Figure 2.
Human fetal γδ thymocytes express invariant germline-encoded CDR3γ and CDR3δ repertoires. (A) Flow cytometry analysis of the expression of Vγ9Vδ2+ γδ T cell subset in FT and PNT γδ thymocytes (left); representative flow cytometry plot of fetal γδ thymocytes (right). (B) HTS analysis of the V gene segment usage TRGV (left) and TRDV (right) in FT and PNT γδ thymocytes. TRG “Others” groups: TRGV1-TRGV3, TRGV5, TRGV5P, TRGV7, and TRGV11 variable chains; TRD “Others” groups: TRDV4, TRDV6, and TRDV7 variable chains. (C) Tree map representations of the CDR3γ and CDR3δ repertoires of FT and PNT γδ thymocytes (colors used are random; no correspondence between graphs). Representative of three independent experiments. (D) Accumulated frequencies (freq.) of the 10 most abundant clonotypes in CDR3γ and CDR3δ repertoires of γδ thymocytes in FT and PNT. (E) Analysis of the “counts per 20 million” (CP20M; RNAseq data) of the DNTT gene of γδ thymocytes in FT and PNT. (F and G) Mean number of N insertions (F) and CDR3 length (G; number of nucleotides) in CDR3γ and CDR3δ repertoires of γδ thymocytes in FT and PNT. (H) Overlap analysis showing the percentage of shared CDR3γ and CDR3δ clonotypes among three FT γδ thymocytes and among three PNT γδ thymocytes. (I) Percentage of particular clonotype amino acid sequences in CDR3γ and CDR3δ repertoires in FT and PNT γδ thymocytes. Each symbol (A, E–G, and I) represents an individual donor. Graphs show the means ± SEM (B, D, and H) or the mean (A, E–G, and I). Data were analyzed by Student’s t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001. n = 8 (FT) and 17 (PNT; A) or n = 3 (FT and PNT; B–I).
Figure S2.
Figure S2.
TCR/CDR3 data of FT and PNT γδ thymocytes. (A and B) Number of N insertions (A) and CDR3 length (B; number of nucleotides) of CDR3 containing the indicated TRGV and TRDV, from sorted FT and PNT γδ thymocytes. (C) J and D gene segment usage in TRG and TRD of sorted FT and PNT γδ thymocytes. (D) J gene segment usage in TRGV2, TRGV3, TRGV4, TRGV5, TRGV8, TRGV10, TRDV1, TRDV2, TRDV3, TRDV5, and TRDV8 of sorted FT and PNT γδ thymocytes. n = 3 for FT and PNT nonVγ9Vδ2 sorted γδ thymocytes. Horizontal lines in A and B indicate means; means ± SEM are shown in C and D. Data were analyzed by Student’s t test; **, P < 0.01; ***, P < 0.001.
Figure S3.
Figure S3.
CDR3 data on αβ thymocytes. (A and B) Number of N insertions (A) and CDR3 length (B; number of nucleotides) in CDR3α and CDR3β repertoires of αβ thymocytes derived from FTs and PNTs. Colors represent replicates from the same subject with 10,000 αβ thymocytes (circles [FTs] and squares [PNTs] with black border) or 100,000 αβ thymocytes (circles [FTs] and squares [PNTs]). Horizontal lines indicate medians. Data were analyzed by Mann Whitney test; **, P < 0.01.
Figure 3.
Figure 3.
Human fetal HSPC-derived γδ T cells share a functional transcriptional profile with fetal γδ thymocytes. γδ T cells were sorted from OP9DL1 cultures either derived from FL HSPCs (n = 3) or AB HSPCs (n = 3), and gene expression profiles were determined by RNAseq. (A) GSEA plots quantifying the shared profile of FL-derived and AB-derived γδ T cells with FT and PNT thymocytes, respectively. (B) GSEA plots quantifying the shared profile of FL-derived and AB-derived γδ T cells with activated (act.; versus naive) human (GSE28726) and mouse (GSE10239) T cells (top) and with IL-15–stimulated (stim.; versus unstimulated) human (GSE22886) and mouse (GSE7764) NK cells (bottom). Gene sets are from the C7 immunological gene signature from the MSigDB. (C) Volcano plot illustrating differentially expressed genes between FL- and AB-derived γδ T cells. NES, normalized enrichment score.
Figure 4.
Figure 4.
Human fetal HSPCs efficiently generate Vγ9Vδ2+γδ T cells. (A) Kinetics of the percentage of γδ T cells (of CD7+ cells), Vδ2+, and Vδ1+ γδ T cells (top) and of the absolute number (bottom) analyzed by flow cytometry and derived from FL, FB, CB, and AB HSPCs during co-culture with OP9DL1 cells. For each time point, n ranges from 2 to 3 (FL), 6 to 9 (FB), 7 to 11 (CB), and 3 to 4 (AB). (B) Ratio between Vδ2+ versus Vδ1+ γδ T cells at day 30 ± 2 of OP9DL1 co-culture (n = 3, 9, 11, and 4, respectively, for FL-, FB-, CB-, and AB-derived HSPCs). (C) Flow cytometry analysis of the expression of Vδ2 and Vδ1 chains of γδ T cells derived from FB HSPCs and of the expression of the Vδ9 chain on the Vδ2+ γδ T cells at day 30 ± 2. Numbers indicate the percentage of the corresponding population. Data are representative of 11 independent experiments. On the right, percentage of the Vγ9-Vδ2+ γδ T cell subset (gated on Vδ2+ γδ T cells) from FL, FB, CB, and AB HSPCs at day 30 ± 2 of OP9-DL1 co-culture (n = 3, 9, 11, and 4, respectively, for FL, FB, CB, and AB). (D) HTS analysis of the V gene segment usage TRGV (left) and TRDV (right) of sorted γδ T cells derived from FL (n = 3), FB (n = 4), CB (n = 6), and AB (n = 3) HSPCs. TRG “Others” groups: TRGV1-TRGV3, TRGV5, TRGV5P, TRGV7, and TRGV11 variable chains; TRD “Others” groups: TRDV4, TRDV6, and TRDV7 variable chains. Graphs show the means ± SEM (A and D) or the medians ± range (B and C). Data were analyzed by Kruskal-Wallis ANOVA followed by Dunn’s multiple comparison test; *, P < 0.05; **, P < 0.01.
Figure 5.
Figure 5.
Human fetal HSPCs efficiently generate invariant CDR3γ and CDR3δ sequences. (A) Tree map representations of the CDR3γ and CDR3δ repertoires of OPDL1 cultures derived from FL, FB, CB, and AB HSPCs (colors in the tree maps are random; no correspondence between graphs). Graphs are representative of 3 (FL), 8 (FB), 10 (CB CDR3γ), 9 (CB CDR3δ), and 4 (AB) independent experiments. (B) Accumulated frequencies (freq.) of the 10 most abundant clonotypes of the CDR3γ and CDR3δ repertoire derived from total cells. FL (n = 3), FB (n = 8), CB (n = 10 for CDR3γ and 9 for CDR3δ), and AB (n = 4). (C) CP20M of the RNAseq data of DNTT (TdT) of γδ T cells derived from FL and AB HSPCs. (D) Mean number of N insertions in CDR3γ and CDR3δ repertoires of total cells, sorted γδ T cells, and sorted progenitor cells derived from FL, FB, CB, and AB HSPCs after co-culture with OP9DL1 cells. (E) Overlap analysis showing the percentage of shared CDR3γ and CDR3δ clonotypes among three FL, four FB, six CB, and three AB sorted γδ T cells derived from HSPCs and among these groups with three FT and three PNT thymus nonVγ9Vδ2 sorted γδ thymocytes. (F) Percentage of particular CDR3 clonotype amino acid sequences present in total cells, sorted γδ T cells, and sorted progenitor cells derived from FL, FB, CB, and AB HSPCs. Each symbol (C, D, and F) represents an individual donor. Graphs show the means ± SEM (B and E). Data were analyzed by Student’s t test (C) and Kruskal-Wallis ANOVA followed by Dunn’s multiple comparison test (D and F); *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Figure S4.
Figure S4.
TRDV10 RNA contains the leader intron introducing a stop codon. Shown are the R1 and R2 raw data reads of a TRGV10-containing sequence that possesses the invariant germline-encoded CDR3 CAAWDTTGWFKIF.
Figure 6.
Figure 6.
Short homology repeats at the VDJ junctions. (A) Left, table showing the nucleotide sequences of the variable gene segments TRGV8 and TRGV10 (3′ ends, CDR3) and of the five joining gene segments TRGJ (J1/J2, JP, JP1, and JP2; 5′ ends, CDR3). Right, table showing the nucleotide sequences of the variable gene segments TRDV1 and TRDV2 and of the three diversity gene segments TRDD (D1, D2, D3). Short homology repeats are shown in red or in red and underlined. (B) Recombination using short homology repeats between TRGV8 and TRGJP1. Overlapping nucleotides are shown in red. (C) Recombination using short homology repeats between TRDV1 and TRDD3 and between TRDV2 and TRDD3. Overlapping nucleotides are shown in red (TRDV1-TRDD3) and in underlined red (TRDV2-TRDD3).
Figure 7.
Figure 7.
Lin28b induces an effector program, inhibits TdT expression, and enhances the formation of germline-encoded invariant CDR3γ and CDR3δ sequences. (A) γδ T cells were sorted from OP9DL1 cultures either derived from Lin28b-transduced CB HSPCs (Lin28b; n = 3) or control-transduced CB HSPCs (Con.; n = 3). GSEA plots based on RNAseq data quantifying the shared profiles of Lin28b-derived versus control-derived γδ T cells with FL and AB HSPC-derived γδ T cells as gene sets. (B) GSEA plots quantifying the shared profile of Lin28b- versus control-transduced CB HSPC-derived γδ T cells with activated (act.; versus naive) human (GSE28726) and mouse (GSE10239) T cells (top) and with IL-15–stimulated (stim.; versus unstimulated) human (GSE22886) and mouse (GSE7764) NK cells (bottom). Gene sets are from the C7 immunological gene signature from the MSigDB. NES, normalized enrichment score. (C) Top, ratio between Vδ2+ and Vδ1+ γδ T cells derived from Lin28b-transduced CB HSPCs (Lin28b) and from control-transduced CB HSPCs (control) at day 30 ± 2 of OP9DL1 co-culture as determined by flow cytometry (n = 9 for Lin28b and control). Bottom, HTS analysis of the TRDV usage on sorted γδ T cells from Lin28b- and control-transduced CB HSPCs. “Others” groups: TRDV4, TRDV6, and TRDV7 variable chains (n = 5). (D) Paired comparison analysis of the CP20M RNAseq data of DNTT (TdT) expressed in γδ T cells derived from Lin28b- versus control-transduced CB HSPCs. (E) Mean number of N insertions in CDR3γ and CDR3δ repertoires of total cells, sorted γδ T cells, and sorted progenitor cells derived from Lin28b- versus control-transduced CB HSPCs. (F) Overlap analysis showing the percentage of shared CDR3γ and CDR3δ clonotypes among sorted γδ T cells (n = 6) derived from Lin28b- versus control-transduced CB HSPCs and among these samples with three FL, four FB, six CB, and three AB sorted γδ T cells derived from HSPCs. (G) Percentage of particular CDR3 clonotype amino acid sequences present in total cells, sorted γδ T cells, and sorted progenitor cells derived from Lin28b- versus control-transduced CB HSPCs. Lines between symbols link samples from the same subject (D, E, and G). Data are means ± SEM (C and F). Data were analyzed by paired Student’s t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Figure S5.
Figure S5.
TCR/CDR3 and RNAseq data of Lin28b- versus control-transduced CB HSPC-derived γδ T cells. (A) Paired comparison analysis of the CP20M (RNAseq data) for a selection of genes expressed in γδ T cells derived from Lin28b- versus control-transduced CB HSPCs. (B) Tree map representations (colors used are random; no correspondence between graph; top) and accumulated frequencies (freq.; bottom) of the 10 most abundant clonotypes of the CDR3γ and CDR3δ repertoires of total cells derived from OP9DL1 culture. Tree maps are representative of seven (Lin28b and control) independent experiments. Graphs show the means ± SEM. (C) Percentage of particular CDR3δ clonotype amino acid sequences present in total cells, sorted γδ T cells, and sorted progenitor cells derived from Lin28b- versus control-transduced CB HSPCs. n = 7 (for total) and n = 5 (for sorted γδ T cells and progenitors). Lines between symbols link samples from the same subject. Data were analyzed by paired Student’s t test; *, P < 0.05; ***, P < 0.001.

References

    1. Asarnow D.M., Cado D., and Raulet D.H.. 1993. Selection is not required to produce invariant T-cell receptor γ-gene junctional sequences. Nature. 362:158–160. 10.1038/362158a0 - DOI - PubMed
    1. Beaudin A.E., Boyer S.W., Perez-Cunningham J., Hernandez G.E., Derderian S.C., Jujjavarapu C., Aaserude E., MacKenzie T., and Forsberg E.C.. 2016. A transient developmental hematopoietic stem cell gives rise to innate-like B and T cells. Cell Stem Cell. 19:768–783. 10.1016/j.stem.2016.08.013 - DOI - PMC - PubMed
    1. Bolotin D.A., Poslavsky S., Mitrophanov I., Shugay M., Mamedov I.Z., Putintseva E.V., and Chudakov D.M.. 2015. MiXCR: software for comprehensive adaptive immunity profiling. Nat. Methods. 12:380–381. 10.1038/nmeth.3364 - DOI - PubMed
    1. Boyden L.M., Lewis J.M., Barbee S.D., Bas A., Girardi M., Hayday A.C., Tigelaar R.E., and Lifton R.P.. 2008. Skint1, the prototype of a newly identified immunoglobulin superfamily gene cluster, positively selects epidermal gammadelta T cells. Nat. Genet. 40:656–662. 10.1038/ng.108 - DOI - PMC - PubMed
    1. Boyerinas B., Park S.-M., Shomron N., Hedegaard M.M., Vinther J., Andersen J.S., Feig C., Xu J., Burge C.B., and Peter M.E.. 2008. Identification of let-7-regulated oncofetal genes. Cancer Res. 68:2587–2591. 10.1158/0008-5472.CAN-08-0264 - DOI - PubMed

Publication types

Substances