Research ArticleAging Open Access | 10.1172/JCI158122

Xiaoping Sun,1 Thomas Nguyen,1 Achouak Achour,1 Annette Ko,1 Jeffrey Cifello,1 Chen Ling,1 Jay Sharma,1 Toyoko Hiroi,1 Yongqing Zhang,2 Chee W. Chia,3 William Wood III,2 Wells W. Wu,4 Linda Zukley,5 Je-Nie Phue,4 Kevin G. Becker,2 Rong-Fong Shen,4 Luigi Ferrucci,5 and Nan-ping Weng1

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Sun, X. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Nguyen, T. in: JCI | PubMed | Google Scholar |

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Achour, A. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Ko, A. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Cifello, J. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Ling, C. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Sharma, J. in: JCI | PubMed | Google Scholar |

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Hiroi, T. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Zhang, Y. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Chia, C. in: JCI | PubMed | Google Scholar |

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Wood, W. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Wu, W. in: JCI | PubMed | Google Scholar |

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Zukley, L. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Phue, J. in: JCI | PubMed | Google Scholar |

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Becker, K. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Shen, R. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Ferrucci, L. in: JCI | PubMed | Google Scholar

1Laboratory of Molecular Biology and Immunology,

2Gene expression and Genomics Unit, Laboratory of Genetics and Genomics, and

3Laboratory of Clinical Investigation, National Institute on Aging, NIH, Baltimore, Maryland, USA.

4Facility for Biotechnology Resources, Center for Biologics Evaluation and Research, Food and Drug Administration, Silver Spring, Maryland, USA.

5Translational Gerontology Branch, National Institute on Aging, NIH, Baltimore, Maryland, USA.

Address correspondence to: Nan-ping Weng, Laboratory of Molecular Biology and Immunology, National Institute on Aging, NIH, 251 Bayview Blvd. Suite 100, Baltimore, Maryland 21224, USA. Email: Wengn@mail.nih.gov.

Authorship note: XS, TN, and AA contributed equally to this work.

Find articles by Weng, N. in: JCI | PubMed | Google Scholar

Authorship note: XS, TN, and AA contributed equally to this work.

Published June 16, 2022 - More info

Published in Volume 132, Issue 17 on September 1, 2022
J Clin Invest. 2022;132(17):e158122. https://doi.org/10.1172/JCI158122.
© 2022 Sun et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Published June 16, 2022 - Version history
Received: January 4, 2022; Accepted: June 14, 2022 View PDF Abstract

A diverse T cell receptor (TCR) repertoire is essential for protection against a variety of pathogens, and TCR repertoire size is believed to decline with age. However, the precise size of human TCR repertoires, in both total and subsets of T cells, as well as their changes with age, are not fully characterized. We conducted a longitudinal analysis of the human blood TCRα and TCRβ repertoire of CD4+ and CD8+ T cell subsets using a unique molecular identifier–based (UMI-based) RNA-seq method. Thorough analysis of 1.9 × 108 T cells yielded the lower estimate of TCR repertoire richness in an adult at 3.8 × 108. Alterations of the TCR repertoire with age were observed in all 4 subsets of T cells. The greatest reduction was observed in naive CD8+ T cells, while the greatest clonal expansion was in memory CD8+ T cells, and the highest increased retention of TCR sequences was in memory CD8+ T cells. Our results demonstrated that age-related TCR repertoire attrition is subset specific and more profound for CD8+ than CD4+ T cells, suggesting that aging has a more profound effect on cytotoxic as opposed to helper T cell functions. This may explain the increased susceptibility of older adults to novel infections.

Graphical Abstract Introduction

Naive T cells produced by the thymus have the potential to recognize any pathogen, whereas memory T cells are generated from a past immunological response and offer long-lasting protection against pathogens during a subsequent encounter (1–4). Production of naive T cells substantially declines after puberty, creating a challenge to maintaining a T cell system throughout a lifetime that balances the numbers of naive and memory T cells (5, 6). Memory T cells continuously accumulate with various degrees of selective clonal expansion after new or repeated immune responses (7, 8). There are 2 general types of T cells: CD4+ cells primarily offer a helper function via the release of cytokines to promote and regulate functions of both B cells (humoral response) and CD8+ T cells; and CD8+ cells use their cytotoxic pathways to kill virally infected or cancerous cells. With age, the reduction in naive T cells in circulating blood is more severe for CD8+ than CD4+ T cells, although the rate of naive CD8+ T cell loss varies tremendously among healthy adults (9–13). Studying the dynamics of naive and memory CD4+ and CD8+ T cells throughout the adult lifetime is important for understanding immunity and aging. Details at the level of T cell receptor (TCR) clonotypes are necessary to understand the age-associated changes of TCR repertoire, which is measured by T cell richness, meaning the number of unique TCR sequences in an individual’s T cell repertoire. Currently, the dynamics of naive and memory CD4+ and CD8+ T cells throughout the adult lifetime at the level of TCR sequence are not known.

The repertoire of αβ TCR — the TCR variable segments that recognize specific pathogens — is determined in adult humans by both genetic events, such as recombination of variable gene regions and α-β chain pairing, and the number of mature T cells in the body. The estimated number of T cells in circulation in an adult is approximately 4 × 1011 (14), with estimates of αβ TCR repertoire size based on genetic elements as high as 1 × 1015 (15–17). Experimental analyses of TCRβ sequences from small numbers of T cells (~1 × 106) suggest that the predicted TCRβ repertoire size range is 1 × 106–108 (18, 19). In addition, TCRβ repertoires appear to have reduced richness with age for both total T cells (20–22) and naive and memory CD4+ and CD8+ cells (23). Research shows that TCRβ repertoire size is highly diverse in human adults (20–22) and is substantially different between naive and memory T cells (23–25), but longitudinal analysis of actual TCR repertoire changes in naive and memory T cells with age using human samples are lacking. Furthermore, reports on TCRβ repertoire size as it corresponds with age were performed with the assumption that all study participants had equal number of T cells without considering the substantial individual differences in T cell numbers and their changes with age. Finally, little is known about size and age changes in the TCRα repertoire or the actual αβ TCR repertoire in humans.

Naive T cells are long-lived cells (26), and the main route for maintaining the naive T cell pool throughout adulthood in humans is homeostatic proliferation (5). The survival of naive T cells depends on having received maintenance signals through the TCR as well as having been exposed to cytokines such as IL-7 in lymphoid organs (27, 28). This naive T cell maintenance mode appears unbiased in early adulthood, but selective expansion of certain naive T cell clones is reported in older humans (23, 29). However, when the uneven expansion of naive T cells starts in an adult life and whether this uneven expansion continues or occurs randomly with age are unknown. Cumulative homeostatic proliferation has 2 known consequences: (a) altered activation thresholds of naive T cells to antigenic activation (30, 31), and (b) loss of naive phenotype and gain in memory phenotype, which is not because of differentiation induced by cognate antigen stimulation (32–35). These alterations are largely characterized by their phenotypes and activation-induced response, but the clonal evidence of these changes has not been determined.

In this study, we conducted a longitudinal assessment of the TCRα and TCRβ repertoires in naive and memory CD4+ and CD8+ T cells from healthy adults. We applied RNA-seq with a TCR-mRNA-marking method using unique molecular identifiers (UMI) to reduce the errors of sequencing read–based methods. We determined longitudinal changes in TCR repertoire and projected TCR repertoire size using the actual circulating T cell numbers from participants’ blood provided at each of 2 donations. We developed equations to calculate αβ TCR repertoire size from TCRα and TCRβ sequences. Our study demonstrated that increasing age is associated with (a) reduced αβ TCR repertoire richness in CD4+ and CD8+ T cells, particularly in naive CD8+ T cells; (b) increased clonal expansion of memory CD8+ T cells; (c) increased overlap in TCR sequences in longitudinal samples for both CD4+ and CD8+ T cells, particularly memory CD8+ T cells; and (d) reduced distinction of TCR sequences between naive and memory CD4+ and CD8+ T cells as well as between CD4+ and CD8+ T cells. These findings, based on actual T cell numbers in individual healthy adults, reveal the dynamic in vivo changes with age in naive and memory CD4+ and CD8+ T cells at the resolution of TCRα and TCRβ sequences.

Results

Reduction of TCRα and TCRβ repertoires in CD4+ and CD8+ T cells with age. To determine changes in αβ TCR repertoires with age, we isolated CD4+ and CD8+ T cell subsets from cryopreserved PBMCs of 30 healthy humans. Using samples taken an average of 9.2 years apart, we determined TCRα and TCRβ repertoires using a UMI-based RNA-seq method (Figure 1A and Supplemental Table 1; supplemental material available online with this article; https://doi.org/10.1172/JCI158122DS1) (36–38). The age of study donors at first visit was late 20s to early 80s, with equal numbers of male and female participants (Figure 1B). To accurately measure TCR repertoire changes with age, we first determined the numbers of circulating CD4+ and CD8+ T cells in the blood of each participant at each donation. The total number of T cells and their subsets in an individual were calculated by (a) determining the percentages of CD4+ and CD8+ T cells and their subsets by flow cytometry of lymphocytes; (b) calculating the counts of CD4+ and CD8+ T cells and their subsets in a microliter of blood based on complete blood cell counts (CBCs); (c) determining total blood volume based on donor height and weight at donation using Nadler’s Equation (39); and (d) calculating the number of CD4+ and CD8+ T cells and their subsets in the donor’s total blood. We observed a significant reduction in numbers of lymphocytes, CD4+ and CD8+ T cells, and naive CD4+ and CD8+ T cells with age (Figure 1, C–F, Supplemental Figure 1, and Supplemental Table 1). These actual numbers of T cells in the blood for each donor were used for projected TCR repertoire richness.

Figure 1

Experimental scheme. (A) Experimental design. Thirty healthy adults were selected from participants of the Baltimore Longitudinal Study of Aging (BLSA). At each of 2 visits, weight and height were measured and fasting blood was drawn, and PBMCs were isolated and cryopreserved. From each sample, CBC counts were analyzed. For experiments, PBMCs were thawed and stained for CD4, CD8, CD45RA and CD28. Naive and memory CD4+ and CD8+ T cells were isolated by cell sorting for T cell receptor α (TCRα) and TCRβ repertoire analysis. PBMCs were isolated from 3 additional healthy adults and naive and memory CD4+ and CD8+ T cells were sorted in 3–4 aliquots as reproducibility controls. UMI, unique molecular identifier. (B) Age and sex of participants at first and second donation. Each line represents 1 donor, and the length of line indicates years between donations. (C–F) Numbers of lymphocytes, total, naive, and memory CD4+ and CD8+ T cells in samples from 2 donations, with change with age. Cell numbers were based on (a) lymphocyte counts per microliter of blood; (b) percentage of CD4+ and CD8+ T cells and naive and memory cells in lymphocytes, calculated from flow cytometry; and (c) blood volume calculated from donor weight and height adjusted by sex (39). Thin short lines link 2 donations from 1 participant. The thick long line is the trend from MLE analysis. The colored shade around the trend line indicates 95% confidence interval. Unless otherwise noted, values were transformed with log10 for presentation and statistical analysis. Values for slope (S) of the trend line and P values (≤ 0.05 was considered significant) are presented. N, naive; M, memory T cells.

After analysis of 1.9 × 107 individual TCRα and TCRβ mRNA molecules (UMI counts) from 1.9 × 108 isolated T cells from 30 donors with an average sequencing depth of approximately 30 sequencing reads per UMI, we calculated (a) TCR repertoire richness that measures the number of unique TCRs in a donor by rarefaction equations that project to the actual numbers of circulating T cells in the blood (1% of total T cell counts) (40); and (b) the Inverse Simpson’s Index (ISI)) that measures both the number of different TCRs and their clonal expansion (Supplemental Table 2). The results showed that the TCRα and TCRβ repertoire richness of both CD4+ and CD8+ T cells varied greatly, ranging from 1 × 104–106 for both TCRα and TCRβ, with varying changes with age among the donors (Figure 2, A and C). For both CD4+ and CD8+ T cells, we found a significant reduction in richness with age for TCRβ but not TCRα and a significant reduction with age for both TCRα and TCRβ measured by ISI (indicating increased clonal expansion) using mixed linear effects (MLE) analysis (Figure 2, B and D). These findings suggested that repertoire changes with age affected both richness and clonal expansion, and that reduction in TCR repertoire richness was more rapid in CD8+ than in CD4+ T cells.

Figure 2

Reductions in αβ TCR repertoires in CD4+ and CD8+ T cells with age. (A) Age-associated reduction in projected richness of T cell receptor α (TCR α) and TCRβ repertoires of CD4+ T cells. TCRα and TCRβ sequences were calculated for each donor and projected to 1% of total circulating CD4+ T cells (in log10-based values) (see C). (B) Age-associated reduction of TCRα and TCRβ diversity of CD4+ T cells measured by ISI. (C) Age-associated reduction in projected richness of TCRα and TCRβ repertoires of CD8+ T cells. (D) Age-associated reduction of TCRα and TCRβ diversity of CD8+ T cells measured by ISI. The colored shade around the trend line indicates the 95% confidence interval. S, slope of the trend line.

Reductions in TCRα and TCRβ repertoires with age in naive and memory CD4+ and CD8+ T cells. The clonal distribution and expansion of naive T cells is an important determinant of T cell immunity (25). To determine whether the observed reductions with age in TCRα and TCRβ repertoires in CD4+ and CD8+ T cells occurred in naive or memory T cells, we measured TCRα and TCRβ repertoires of naive and memory CD4+ and CD8+ T cells isolated by cell sorting, with CD45RA+CD28+ cells sorted as naive and all other cells sorted as memory cells and determined changes in TCR repertoires with age. The richness of TCR repertoires was projected to 1% of the total actual naive and memory CD4+ and CD8+ T cells in the blood of donors, and TCR clonal expansion was calculated by ISI (Supplemental Table 3). We observed significant reductions with age in TCR repertoire richness, especially TCRβ richness, in CD4+ and CD8+ naive T cells, but not CD4+ and CD8+ memory T cells (Figure 3, A and C). Again, the reductions in TCR richness were more rapid in naive CD8+ T cells (TCRα = –2.19 %/year and TCRβ = –3.48%/year) than in naive CD4+ T cells (TCRα = –0.66%/year and TCRβ =–2.27%/year). Age also led to increased clonal expansion in the naive TCRβ repertoire of CD8+ but not CD4+ T cells (Figure 3, B and D). In contrast, reductions with age in TCRβ repertoire richness were not significant for CD4+ and CD8+ memory T cells (Figure 3C), but reduction of ISI with age was significant for TCRβ but not TCRα of memory CD4+ and CD8+ T cells (Figure 3D). Together, these findings demonstrated that age led to a more profound reduction in richness of TCRα and TCRβ repertoires in naive than in memory T cells. Age also resulted in a significant clonal expansion of TCRβ repertoires in both CD4+ and CD8+ naive and memory T cells.

Figure 3

CD4+ and CD8+ T cell subset–specific reductions of αβ TCR repertoires with age. (A) Age-associated reduction in richness of TCRα and TCRβ repertoires of naive CD4+ and CD8+ T cells. TCRα and TCRβ sequences were calculated for each donor and projected for 1% of total circulating naive CD4+ or CD8+ T cells (in log10 -based values), for (A) and (C) in this figure. (B) Age-associated reduction in ISI for TCRα and TCRβ diversity of naive CD4+ and CD8+ T cells. (C) Age-associated reduction in richness of TCRα and TCRβ repertoires of memory CD4+ and CD8+ T cells. (D) Age-associated reduction in ISI of TCRα and TCRβ diversity of memory CD4+ and CD8+ T cells. Thin short lines link 2 donations from 1 participant. Thick lines are trends from MLE analysis. The colored shade around the trend line indicates the 95% confidence interval. S, slope of trend line.

Next, we analyzed TCR richness changes with age in naive and memory CD4+ and CD8+ T cells for each donor, comparing samples provided at different ages. To determine the true age-associated changes, we first measured TCR richness variation in samples collected at the same time but measured independently. The SDs of projected TCR richness of naive and memory CD4+ and CD8+ T cells were calculated using samples from 3 healthy adults (Supplemental Figure 2). We defined an age-associated change in TCR richness as greater than 1 SD in the estimated richness of each type of T cell subset. Using this criterion, we found the following changes in naive TCR repertoire richness among cell subsets. Reduced richness was observed in 59% of donors (average of TCRα and TCRβ for both naive and memory CD4+ and CD+ T cells); 11% had no obvious changes; and 30% had increased richness (Table 1). Further analysis showed that there was no statistical significance between the average age of donors in which their TCRα and TCRβ richness increased versus those in which their richness decreased (Supplemental Figure 3).

Table 1

Type of age-associated changes in TCR richness in study donors

Predicting paired αβ TCR repertoires and their age-associated changes. Studies have reported methods for pairing TCRα and TCRβ from bulk TCRα and TCRβ sequences using statistical modeling and frequencies (41, 42). We analyzed the relationship between separated TCRα and TCRβ sequences and their αβ-paired TCR using paired αβ TCR sequences (from 745,182 CD4+ and 158,305 CD8+ T cells) from single-cell RNA-seq studies and observed a linear relationship between the number of unique TCRα and TCRβ sequences and the number of paired αβ TCR clones (Supplemental Figure 4A). The numbers of TCRα and TCRβ and the numbers of their pairs reveal a mathematical principle that allows for direct estimation of αβ-paired TCR repertoires from individual TCRα and TCRβ sequences. Because some T cells have 2 functional TCRα sequences (43–45), we used the same data sets to calculate the average percentage of T cells with only single TCRα sequences and used this information to adjust the bulk TCRα sequences in calculations of paired αβ TCR richness (Supplemental Figure 4B). The TCR repertoire is larger for CD4+ T cells than for CD8+ T cells (19, 46), so we used separate equations to estimate the paired αβ TCR richness for CD4+ and CD8+ T cells (Supplemental Figure 4C). We found that projected paired αβ TCR richness was larger for CD4+ than CD8+ T cells, specifically an average 1.6-fold of the average TCRα and TCRβ richness for CD4+ cells and 1.5-fold of the average for CD8+ T cells (Figure 4 and Supplemental Table 2). Paired αβ TCR richness showed significant reductions with age for total CD8+ (–2.36%/year, P = 0.003), naive (–2.84%/year, P = 0.001), and memory (–2.04%/year, P = 0.028) CD8+ T cells, but not total, naive, or memory CD4+ T cells (Figure 4, A–C, and Supplemental Tables 2 and 3). To overcome the problem of a small number of T cells used for predicting the total TCR repertoire, we combined TCR sequences for CD4+ (1.26 × 108) and CD8+ (6.07 × 107) T cells for all 30 donors and projected αβ TCR richness for CD4+ and CD8+ T cells to 1% of the average of total cells in the blood for all donors. We found that the paired αβ TCR richness of 1% of average total blood was 3.0 × 106 for CD4+ T cells and 7.9 × 105 CD8+ T cells (Figure 4D). Thus, the αβ TCR repertoire richness in the total blood of an adult human was estimated, at the lower end, to be approximately 3.8 × 108.

Figure 4

Age-associated decline of predicted αβTCR repertoires based on TCRα and TCRβ sequences. (A) Reduction with age of paired αβ T cell receptor (TCR) repertoire richness of CD4+ and CD8+ T cells. Paired αβ TCR repertoire richness of CD4+ and CD8+ T cells was estimated based on projected TCRα and TCRβ richness (1% of total circulating cells in blood) via linear regression of single-cell αβ TCR sequences (Supplemental Figure 4) (A–C of this figure). (B) Reduction with age in paired αβ TCR repert