Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 15.
Published in final edited form as: Int J Radiat Oncol Biol Phys. 2022 Oct 23;115(4):945–956. doi: 10.1016/j.ijrobp.2022.10.024

Long-Term Immunological Consequences of Radiation Exposure in a Diverse Cohort of Rhesus Macaques

Matthew J French *, Ryan Wuerker *, Greg Dugan †, John D Olson †, Brittany R Sanders *, Janet A Tooze ‡, David L Caudell †, J Mark Cline †, Gregory D Sempowski *, Andrew N Macintyre *
PMCID: PMC9974872  NIHMSID: NIHMS1850954  PMID: 36288757

Abstract

Purpose:

The aim of this study was to develop an improved understanding of the delayed immunologic effects of acute total body irradiation (TBI) using a diverse cohort of nonhuman primates as a model for an irradiated human population.

Methods and Materials:

Immune recovery was evaluated in 221 rhesus macaques either left unirradiated (n = 36) or previously irradiated (n = 185) at 1.1 to 8.5 Gy TBI (median, 6.5 Gy) when aged 2.1 to 15.5 years (median, 4.2 years). Blood was drawn annually for up to 5 years total between 0.5 and 14.3 years after exposure. Blood was analyzed by complete blood count, immunophenotyping of monocytes, dendritic cells (DC) and lymphocytes by flow cytometry, and signal joint T-cell receptor exclusion circle quantification in isolated peripheral blood CD4 and CD8 T cells. Animals were categorized by age, irradiation status, and time since irradiation. Sex-adjusted means of immune metrics were evaluated by generalized estimating equation models to identify cell populations altered by TBI.

Results:

Overall, the differences between irradiated and nonirradiated animals were subtle and largely restricted to younger animals and select cell populations. Subsets of monocytes, DC, T cells, and B cells showed significant interaction effects between radiation dose and age after adjustment for sex. Irradiation at a young age caused transient increases in the percentage of peripheral blood myeloid DC and dose-dependent changes in monocyte balance for at least 5 years after TBI. TBI also led to a sustained decrease in the percentage of circulating memory B cells. Young irradiated animals exhibited statistically significant and prolonged disruption of the naïve/effector memory/central memory CD4 and CD8 T-cell equilibrium and exhibited a dose-dependent increase in thymopoiesis for 2 to 3 years after exposure.

Conclusions:

This study indicates TBI subtly but significantly alters the circulating proportions of cellular mediators of adaptive immune memory for several years after irradiation, especially in macaques under 5 years of age and those receiving a high dose of radiation.

Introduction

Total body irradiation (TBI) causes acute damage to multiple tissues including the reproductive, vascular, gastrointestinal, and immune systems.1 Although these immediate effects of radiation exposure are well described, the immunologic consequences of TBI years after the initial injury are poorly characterized. This is an important preparedness concern given the risk of widespread irradiation from a nuclear accident or device and the limited availability and efficacy of medical countermeasures.2

TBI has been used clinically for over 50 years for myeloablation before hematopoietic stem cell transplant. Hematopoietic stem cell transplant recipients exhibit immune dysregulation, abnormal T-cell function,3 and low-grade inflammation for years posttransplant.4 Although grafted tissue and medical management make it impossible to confirm TBI as the causative factor, low-grade inflammation correlates with TBI, rather than chemotherapeutic, ablation.5 Similarly, patients with Hodgkin lymphoma treated with radiation therapy were found to exhibit abnormal in vitro T-cell responses a decade after treatment compared with nonirradiated patients.6

Additional indications of long-term immune consequences of irradiation have come from survivors of the 1945 atomic bombings in Japan and the 1986 Chernobyl nuclear accident. Individuals irradiated during the 1945 bombings had decreased cellular immune function decades after exposure, including reduced CD4 T-cell numbers, altered CD4 memory T-cell antigen receptor (TCR) gene usage, and diminished CD4 T-cell responses to in vitro stimulation.7–9 Reports of immune damage in those exposed at Chernobyl are contradictory, particularly with regard to T-cell effects, and are incongruous with the atomic bomb studies.10–13 The Chernobyl studies also are confounded by elevated levels of lead, an immunotoxin, in some participants, likely as a consequence of lead use during the emergency response.14

Given the limitations of observational studies in humans, the rhesus macaque nonhuman primate (NHP) model often is used to study the immune consequences of radiation. NHPs are physiologically, genetically, and immunologically similar to humans, and NHPs are susceptible to many human pathogens and simian pathogens with human equivalents.2,15,16 Furthermore, NHPs are prone to age- and inflammation-associated morbidities, such as hypertension and obesity common in older humans.17 The acute immune effect of TBI on macaques is similar to that in humans, with rapid loss of circulating granulocytes, B, and T lymphocytes days after TBI and delays in CD4 T-cell recovery leading to a transient shift in the CD4/8 T-cell ratio.18–21 Prospective studies have found that although gross lymphopenia resolves within 5 weeks of exposure and both B and T cells return to preirradiation counts within 4 months, the balance of naïve and memory T cells remain different from age-matched controls at least 18 months after irradiation.18,22 Similarly, a recent study of 29 irradiated macaques at a single time point suggested that there was a dose-dependent increase in the memory to naïve CD8 T-cell ratio 5 years after exposure at 6.5 to 8.4 Gy.23

To assess the likely long-term immunologic consequences of TBI, a cohort of more than 200 rhesus macaques that were either left unirradiated or previously irradiated at a range of doses and ages was used to mimic a diverse human population. The animals were immunophenotyped annually for up to 5 years total depending on when they entered the cohort, and the data interrogated using generalized estimating equation models.

Methods and Materials

Additional details are provided in the supplemental.

Animal husbandry and TBI

Two hundred twenty-one rhesus macaques (Macaca mulatta) were studied. All animal procedures including medical management and blood collections were conducted at Wake Forest University (WFU) School of Medicine with prior approval from the WFU Institutional Animal Care and Use Committee. The WFU Laboratory Animal Care Program complies with the “Guide for the Care and Use of Laboratory Animals,”24 the “Principles for Use of Animals,” and has been accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International since 1966. WFU has an assurance with the Office for Protection from Research Risks, Office of the Director, National Institutes of Health, who accepts responsibility for the humane care of animals (OPRR #A-339101).

The control and irradiated animals used for this study were obtained from 10 institutions and transferred to WFU for long-term monitoring. Irradiated animals received 1.1 to 8.5 Gy TBI as either simultaneous bilateral exposure or split dose a few seconds apart. All irradiations were performed under Institutional Animal Care and Use Committee oversight at WFU or prior institution. One hundred and one of the animals were irradiated as part of natural history acute radiation response studies and 84 of the animals were irradiated as part of mitigator efficacy studies and treated with experimental mitigators during the acute postirradiation phase. None of the irradiated animals in this study were part of radiation cancer therapy studies. Irradiation procedures and animal care are detailed in the supplemental data.

Blood collection and processing

Animals were sampled annually with approximately 52 weeks between sampling dates for the same animal unless medical management necessitated schedule alterations. Animals were sedated with intramuscular ketamine and femoral vein blood was drawn into sodium heparin-coated collection tubes (BD Bioscience). Complete blood counts (CBC) with differentials were collected at time of draw and then blood transferred to Duke by courier. The following day, plasma and peripheral blood mononuclear cells (PBMC) were separated using System-Histopaque-1077 separation media (Sigma). PBMC were washed in RPMI-1640 (Sigma), immunophenotyped and the remaining cells cryopreserved in 90% fetal bovine serum 10% dimethyl sulfoxide (Sigma).

Immunophenotyping

PBMC were labeled for 40 minutes at room temperature with 4 cocktails of fluorescently conjugated antibodies (targets, clones, and suppliers are listed in Table E1) in PBS 1% BSA, washed in PBS 1% BSA and then PBS alone, and stained with live/dead fix aqua (Invitrogen). Stained cells were fixed in 4% paraformaldehyde (ThermoFisher) and analyzed using a LSRII (BD Bioscience) or iQue3 (Sartorius) flow cytometer. Data were gated using FlowJo version 10 (BD Bioscience) or ForeCyt version 7.02 (Sartorius). Gating strategies are shown in Fig. E7.

Signal joint TCR excision circles quantification

Cryopreserved PBMC were thawed, counted, and labeled with CD4-binding magnetic microbeads (Miltenyi Biotec). Labeled cells were isolated using magnetic columns (Miltenyi Biotec). Non-CD4+ cells were then labeled with CD8-binding magnetic microbeads (Miltenyi Biotec) and CD8+ cells isolated as described previously. Cells were counted and then lysed at 10,000 cells/μL in 96 μg/mL proteinase K (Roche) (1 hour 56°C then 10 minutes 95°C). Signal joint TCR excision circles (sjTREC) in the lysates and a quantified, serially diluted sjTREC plasmid25 were amplified using 0.125 μL Platinum Taq DNA polymerase, 0.2 mM dNTPs, 3.5 mM MgCl2 (all Invitrogen) 0.5 μM primers (5’-CACATCCCTTTCAACCATGCT-3’ and 5’-GCCAGCTGCAGGGTTTAGG-3’), and 2 μM probe (5’-/56-FAM/ACGCCTCTG/ZEN/GTTTTTGTAAAGGTGC TCACT/3IABkFQ/−3’) (all IDT) on a CFX96 or CFX Connect real-time PCR instrument (Bio-Rad) (95°C for 10 minutes; 45 cycles: 95°C for 15 seconds, 60°C for 1 minute). Sample sjTREC per 100,000 cells was extrapolated from the plasmid standard.

Statistical analyses

Subject characteristics were summarized using descriptive statistics. Generalized estimating equation (GEE) models26,27 with a linear link and exchangeable correlation matrix that adjusted for repeated measures were used to evaluate differences by irradiation group and age using 3 sequential models: first, a model was fit with sex, irradiation (yes, no), age category (3–4, 5, 6–7, 8–11, 12–23 years) and the interaction of irradiation and age category; second, sex, radiation dose categories (unirradiated, lethal dose (LD) under 10 percent (<LD10), LD10–50, LD50–90, >LD90), age category, and dose category by age category interaction were examined. Least squares (adjusted) means were estimated for each age category by dose category group; these means were adjusted for sex by assuming balance. Finally, among irradiated animals only, sex, radiation dose category, age (as continuous), time since irradiation category (0-<1, 1-<2, 2-<3, 3-<5, ≥5 years), and the interaction of radiation dose category and time since irradiation category were examined. Least squares means were estimated for each radiation dose category by time since irradiation category group; these means were adjusted by sex by assuming balance and for age by holding it at the mean value. Main effects of irradiation and interaction effects were also evaluated to indicate if the relationship between age category and irradiation (models 1 and 2) or time since irradiation category and irradiation group (model 3) differed over time. Linear contrasts were used to evaluate differences at each age category for models 1 and 2 and at the time since irradiation category in model 3. All analyses were performed in SAS version 9.4 (SAS Institute Inc) at a 2-sided alpha level of 0.05. All plots show adjusted means with 95% confidence interval, and the asterisk indicates P < .05 for differences between adjusted means.

Results

NHP radiation survivor cohort as a model of an irradiated population

Blood was drawn from 221 animals annually for up to 5 years depending on when each individual entered the cohort. Of the 221 subjects, 185 were previously irradiated (126 male and 59 female) and 36 were never irradiated (33 male, 3 female). The irradiated animals were exposed to between 1.1 Gy and 8.5 Gy TBI (median, 6.5 Gy) depending on the needs of the original experiments. Irradiated animals were grouped by the approximate LD of the exposure (under 5.5 Gy equaled <LD10, 5.5 to <6.75 Gy equalled LD10-LD50, 6.75 to <8 Gy equaled LD50-LD90, over 8 Gy equaled >LD90).20 Animals were irradiated at between 2.3 and 15.5 years of age (median, 4.2 years; Fig. E8A) and blood was drawn between 0.5 and 14.3 years after exposure. A total of 706 blood draws were performed, with draws being obtained from animals irradiated at a range of doses, ages, and years since irradiation (Fig. E8B). Five hundred sixty-one of these blood draws were collected from irradiated animals at ages from 3.5 to 22.9 years, and 145 were from nonirradiated animals at ages from between 4.2 and 21.0 years. The diversity of blood draw participants was, therefore, similar to those that might be obtained from a human population of mixed age, sex, and radiation exposure in the years after an exposure event.

After each blood draw, CBC were determined and then PBMC were isolated. One PBMC aliquot was used immediately for immunophenotyping via flow cytometry and the remainder were cryopreserved for sjTREC analysis to assess thymopoiesis.

The objective of this study was to determine whether TBI altered immune cell population distributions years after recovery from acute radiation injury. Immunosenescence induces measurable changes in immune cells over this timespan,28 and so CBC and other immune metrics were visualized against age at blood draw (Fig. E9). Three statistical analyses were performed on the immune metrics using GEE models that adjust for repeated measures. GEE offers a population-level approach suitable for modeling longitudinal, unbalanced data sets.27 Significance tables (P values) of groups by radiation exposure expressed (1) as a binary (yes/no), (2) as a categorical variable grouped by dose (none, <LD10, LD10–50, LD50-90, >LD90), and (3) by time since irradiation among irradiated animals only with adjustment for sex are presented in Tables E2 to E13. Figures 1 to 6 show the adjusted mean from the statistical model and are adjusted for sex by assuming balance. The unadjusted population percentages for Figs. 1 to 6 are shown in Figs. E1 to E6.

Fig. 1.

Fig. 1.

Open in a new tab

Long-term effect of total body irradiation on monocytes. The percentages of classical, intermediate, and nonclassical monocytes in the peripheral blood mononuclear cells of irradiated and unirradiated nonhuman primates were determined annually for up to 5 years. Means were adjusted for sex and then plotted against age at sample collection. Results were grouped by (A-C) irradiation status or (D-F) dose. (G-I) Means for irradiated animals were adjusted for age at sample collection and plotted against time since irradiation. All adjusted group means were compared using generalized estimating equation models taking into account repeated measures. P values are in Tables E4 and E5, and percentages without adjustments are shown in Fig. E1.

Fig. 6.

Fig. 6.

Open in a new tab

Long-term effect of total body irradiation on CD4 T-cell subsets. CD4 T cells in the peripheral blood mononuclear cells of irradiated and unirradiated nonhuman primates were immunophenotyped and assayed for signal joint T-cell antigen receptor exclusion circle content annually for up to 5 years. Results were grouped by (A-D, M) irradiation status or (E-H, N) dose and means were adjusted for sex and plotted against age at sample collection. (I-L, O) Means for irradiated animals were then adjusted to correct for age at sample collection and plotted against time since irradiation. All adjusted group means were compared using generalized estimating equation models taking into account repeated measures. P values are in Tables E10 to E13, and percentages without adjustments are shown in Fig. E6.

Total white blood cell (WBC) counts in the peripheral blood decreased as animals aged regardless of irradiation status (Fig. E9A). Irradiated animals had slightly higher counts regardless of age, but this was only marginally significant (P = .042, Table E2). It was most apparent in adult animals aged 8 to 11 years and showed a strong dose-dependency, with increased dose correlating with increased WBC count (Fig. E9D, Table E2) This dose-dependent stratification was evident for at least 5 years after irradiation after adjustment for age, sex, and repeated measures (Fig. E9G, Table E3). There was a dose-dependent increase in WBC and total lymphocytes in irradiated animals aged 6 to 11, with higher dose correlating with increased counts (Fig. E9B, E9C, E9E, E9F; Table E2). After adjustment for age and sex there was a transient, dose-dependent increase in total lymphocyte counts in the first 2 years after irradiation (P = .016 for interaction between dose and time since irradiation; Fig. E9H, Table E3).

Having determined that circulating counts of lymphocytes and monocytes returned to levels similar to those observed in age-matched, nonirradiated controls, flow cytometry was next used to immunophenotype the freshly isolated PBMC to determine whether radiation led to changes in the proportions of the major lymphocyte, monocyte and dendritic cell subpopulations.

TBI causes subtle, long-term changes to monocyte subpopulations

In both humans and NHP, monocytes can be classified as classical (CD14++CD16−), intermediate (CD14++CD16+), and nonclassical (CD14+/lowCD16++). Classical monocytes, the major monocyte population, are associated with inflammatory responses and wound healing; intermediate monocytes process and present antigens to T cells; and nonclassical monocytes patrol the endothelium and stimulate CD4 T-cell responses.29 Elevated intermediate monocyte polarization has been implicated as a key mediator of pathogenic fibrosis in some contexts, including after acute radiation injury.29,30 Thus, radiation-induced chronic changes in monocyte balance may have immunologic and pathologic consequences.

The percentage of classical, intermediate, and nonclassical populations within the monocytes of the study cohort were grouped by irradiation status and the means plotted against age after adjustment for sex (Fig. 1A–C, Table E4). In nonirradiated animals, the proportion of intermediate monocytes dropped sharply before adolescence and then remained stable, while classical monocytes peaked at early adulthood (age 6–7 years) and then declined. Conversely, the proportion of nonclassical monocytes gradually increased as animals aged (Fig. 1A–C). Irradiated animals tended to have a lower percentage of classical monocytes at most ages, and there was an interaction effect between irradiation status and age on the classical monocyte percentage (P = .006). The effect was most significant in animals aged 6 to 7 years (P = .001; Table E4), and was significantly dose-dependent, being most pronounced in the highest dose (≥LD90) group (Fig. 1D, 1G). Young irradiated animals (aged 3–4 years) also exhibited a reduction in the percentage of intermediate monocytes (P = .046) that was evident in all dose groups (Fig. 1B, 1E). Older animals (>12 years of age) had a concomitant increase in nonclassical monocytes compared with animals of similar age regardless of dose (Fig. 1F).

Monocyte distributions were then tested for time-effects of irradiation by calculating the time since irradiation and adjusting the results for sex and age at time of collection (Fig. 1G–I, Table E5). The percentages of both classical and intermediate monocytes gradually increased in the years after irradiation, and more than 5 years after irradiation there was a clear stratification of effects on monocytes by dose, with the percentage of both classical and intermediate monocytes increasing as dose increased (over LD90>LD50-90>LD10–50> under-LD10; P = .027 for classical and P = .003 for intermediates at ≥5 years). Thus, while TBI did not cause a sustained change in absolute number of monocytes (Fig. E3C), it did lead to subtle, long-term changes to the monocyte subpopulation ratios.

Irradiation at a young age causes transient increases in myeloid DC

Age- and irradiation-induced changes were next investigated in dendritic cell (DC) populations within the PBMC. Both human and macaque DCs can be defined as CD3−CD20−HLD-DR+CD14− and then classified by expression of CD11c and CD123. In humans, there are 2 major DC populations: CD11c+CD123lo myeloid DC (mDC) and CD11c−CD123hi plasmacytoid DC (pDC). There is also a minor less-differentiated subset that expresses neither marker. Macaque DC can also be grouped as mDC and pDC using these definitions,31 although the CD11c−CD123− population typically forms a larger proportion of DC.32,33 In both species, pDC sense viral infection and induce type I and III interferon responses, and mDC present antigens to T cells and release T-cell regulatory cytokines.

DC populations were expressed as a percentage of live CD45+ PBMC, grouped by dose, adjusted for sex and plotted against age (Fig. 2A–C). The pDC population showed no differences between irradiated and unirradiated animals (Fig. 2A, Table E6). Young irradiated animals (age 3–4 years) exhibited an increase in the proportion of cells that were mDC (P < .001) in comparison to controls; however, in older animals the proportion of mDC was similar regardless of irradiation status (Fig. 2B, Table E6). The effect on young animals was dose-dependent, being most evident in the animals that received LD50-90 and lowest in the animals that received LD10–50 (Fig. 2B). After correcting for age at sample collection, the dose dependent effect of irradiation was only significant in animals within 2 years of irradiation (Fig. 2E, Table E7). The percentage of the PBMC that were less-differentiated DC dropped as animals aged regardless of irradiation status (Fig. 2C), and the effect of radiation exposure on this population was subtle and varied by both age and dose. Young adult macaques (aged 5–7 years) irradiated at over LD10 had a higher proportion of this subset in comparison to age-matched controls, but these differences were not apparent in older animals (Fig. 2C). After correction for age at time of collection, there was a significant (P < .001) interacting effect of time since irradiation and dose on this population, with dose shaping the percentage of less-differentiated DC up to 3 years after exposure (Fig. 2F, Table E7).

Fig. 2.

Fig. 2.

Open in a new tab

Long-term effect of total body irradiation on dendritic cells (DC). The percentages of plasmacytoid, myeloid, and nondifferentiated DC in the peripheral blood mononuclear cells of irradiated and unirradiated nonhuman primates were determined annually for up to 5 years. Results were grouped by (A-C) dose and means were adjusted for sex then plotted against age at sample collection. (D-F) Means for irradiated animals were then adjusted to correct for age at sample collection and plotted against time since irradiation. All adjusted group means were compared using generalized estimating equation models taking into account repeated measures. P values are in Tables E6 and E7, and percentages without adjustments are shown in Fig. E2.

TBI causes a sustained decrease in circulating memory B cells

B cell-mediated humoral responses and T cell-mediated cellular responses are critical for adaptive immunity. Radiation did not cause a sustained change in total lymphocyte count (Fig. E3C, Table E2), and so flow cytometry was used to determine whether TBI changed the lymphocyte subpopulations. PBMC were gated by viability, CD45+ and then lymphocytes identified as CD3+ (T cells) or CD20+ (B cells).

The effect of irradiation on the overall proportion of B cells in the lymphocyte gate was small and showed a strong interaction effect between age and both irradiation status (P = .003) and dose (P < .001). The most significant increase in B-cell percentage was observed in young (3–4 years of age) animals irradiated at >LD90 (Table E8, Fig. 3A). Although radiation exposure status had no significant overall long-term effect on the percentage of activated B cells (CD80+), there was a short-term increase in activated B cells in young animals that received <LD50 compared with animals irradiated at higher doses (Fig. 3B, 3F; Tables E8 and E9). One year after irradiation, all dose groups showed similar activated B-cell percentages (Fig. 3F). The percentages of memory (CD27+) B cells were also evaluated; nearly all ages and doses of irradiation exposure reduced the percentage of memory B cells compared with nonirradiated controls (Fig. 3C). Curiously, one exception was young animals in the highest dose category (aged 3–4 years, >LD90); they tended to have a higher memory B-cell proportion than similarly aged nonirradiated animals (Fig. 3C). Regardless of dose, there was a highly significant interacting effect of age and irradiation status on the percentage of memory B cells (Table E8, Fig. 3C).

Fig. 3.

Fig. 3.

Open in a new tab

Long-term effect of total body irradiation on B cells. B cells in the peripheral blood mononuclear cells of irradiated and unirradiated nonhuman primates were immunophenotyped annually for up to 5 years. Results were grouped by (A-D) dose and means were adjusted for sex and plotted against age at sample collection. (E-H) Means for irradiated animals were then adjusted to correct for age at sample collection and plotted against time since irradiation. All adjusted group means were compared using generalized estimating equation models taking into account repeated measures. P values are in Tables E8 and E9, and percentages without adjustments are shown in Fig. E3.

B-cell surface expression of MHC Class II receptor HLA-DR, used for antigen presentation to T cells, fell as animals aged (Fig. 3D), and there were negligible differences between irradiated and nonirradiated animals at most ages and doses. Curiously, B-cell HLA-DR expression rebounded in older irradiated adults (12–23 years) to above that of nonirradiated age-equivalent controls. It is possible that any effect of irradiation on B-cell HLA-DR expression is transient, as after correcting for age there was a significant interacting effect of dose and time since irradiation (P < .001) and within all dose groups there was drop in HLA-DR expression 1 to 3 years after exposure followed by a rebound (Table E9, Fig. 3H).

TBI at a young age causes a transient, dose-dependent increase in thymopoiesis

Aging led to reduction in the proportion of T cells in the PBMC, and this was accelerated by irradiation in younger animals (Fig. 4A, 4E; Table E10). All animals showed a skewing from CD8 to CD4 T cells as they aged, and this was accelerated in the LD50-90 dose group (Fig. 4B, 4C, 4F, 4G). The proportion of CD4+CD8+ double positive cells T cells increased in irradiated animals regardless of age (P = .017; Fig. 4D, 4H).

Fig. 4.

Fig. 4.

Open in a new tab

Long-term effect of total body irradiation on major T-cell populations. T cells in the peripheral blood mononuclear cells of irradiated and unirradiated nonhuman primates were immunophenotyped annually for up to 5 years. Results were grouped by (A-D) irradiation status or (E-H) dose and means were adjusted for sex and plotted against age at sample collection. All adjusted group means were compared using generalized estimating equation models taking into account repeated measures. P values are in Tables E10 to E13, and percentages without adjustments are shown in Fig. E4.

T cells can be classified as naïve (CD28+CD95−), effector memory (Tem) (CD28−CD95+), or central memory (Tcm) (CD28+CD95+).31 As observed in other rhesus cohorts,31 CD8 T cells skewed from a naïve to Tem phenotype as animals aged (Fig. 5A–C). Irradiation at a young age had a dramatic effect on this balance, leading to a reduction in the percentage of CD8 Tcm and naïve T cells and an increase in Tem. The Tem and Tcm effects were significant across age categories (P = .011 and P = .016, respectively), and there were significant interacting effects between age and irradiation (P < .02 for all) (Fig. 5E–G, Table E10). Radiation dose had a highly significant effect on the proportions of all 3 CD8 T-cell subpopulations regardless of age at sample collection, and there were clear interacting effects between dose and time since irradiation (P < .001 for all, Table E11). The effects on CD8 Tcm and Tem were most apparent within a year of irradiation, and the decrease in naïve and increase in Tem were directly proportional to dose received. The dose-dependency of the effect on naïve CD8 T-cell proportions was only significant a year or more after irradiation. Overall, the dose-dependent effects of irradiation on the CD8 T-cell balance were significant for at least 3 to 5 years after exposure (Fig. 5I–K, Table E11).

Fig. 5.

Fig. 5.

Open in a new tab

Long-term effect of total body irradiation on CD8 T-cell subsets. CD8 T cells in the peripheral blood mononuclear cells of irradiated and unirradiated nonhuman primates were immunophenotyped and assayed for signal joint T-cell antigen receptor exclusion circle content annually for up to 5 years. Results were grouped by (A-D, M) irradiation status or (E-H, N) dose and means were adjusted for sex and plotted against age at sample collection. (I-L, O) Means for irradiated animals were then adjusted to correct for age at sample collection and plotted against time since irradiation. All adjusted group means were compared using generalized estimating equation models taking into account repeated measures. P values are in Tables E10 to E13, and percentages without adjustments are shown in Fig. E5.

In contrast to CD8 T cells, the CD4 population in nonirradiated animals skewed away from a Tem phenotype as animals aged (Fig. 6A–C). Irradiation had a clear effect on CD4 T cells in younger animals, with TBI leading to an increase in the percentage of naïve CD4 T cells and a concomitant reduction in Tcm until around 7 years of age (Fig. 6A–C, 6E–G; Table E10). The effects of radiation on CD4 Tcm and naive proportions were significant (both P < .001 when radiation exposure treated as a binary variable; Table E10), and there were significant interacting effects between age and dose. There were dose-dependent effects on the naïve CD4 T-cell proportion for at least 2 years after TBI (Fig. 6I, Table E11), with the highest dose (>LD90) animals exhibiting the largest reduction in naïve CD4 T cells for at least 5 years after irradiation (Fig. 6I). Young animals receiving LD10 to LD90 showed the greatest increase in naïve CD4 T cells (Fig. 6E), and these animals also exhibited a faster recovery in the naïve CD4 T-cell population after irradiation, within 1 to 2 years of TBI, compared with the <LD10 and >LD90 groups (Fig. 6I). This may reflect a rheostatic response wherein low-dose TBI is insufficient to trigger aggressive homeostatic regeneration while the physiological damage caused by doses over LD90 precludes recovery until a few years after exposure.

Macaque memory T cells can be further categorized by CCR7 expression as Tcm (CD95+CD28+CCR7+), Tem (CD95+CD28−CCR7−), or transitional memory (CD95 +CD28+CCR7−).31 The proportions of CD4 and CD8 transitional cells were reduced in irradiated animals regardless of age (P < .001 and P = .024, respectively; Table E10). These differences were especially acute in animals under 7 years of age (Fig. 5D, 6D), and the effect of radiation on CD8 transitional cells was highly dose dependent (P = .001; Fig. 5L, Table E11). Collectively, these data indicate that young irradiated animals exhibited a multiyear destabilization of T-cell subsets. In contrast, older irradiated animals showed little difference in T-cell balance compared with their nonirradiated counterparts.

The resurgence in T cells in irradiated animals in the years after acute radiation induced lymphopenia18,19 may be mediated by thymopoiesis and/or peripheral expansion. This is a critical distinction as the generation of novel TCR only occurs within the thymus, and so only thymopoiesis enhances TCR diversity. TCR gene recombination in thymocytes generates circularized signal joint TCR excision circles (sjTREC) of DNA. These remain within recent thymic emigrants but are not reproduced during mitosis34 and so peripheral T-cell sjTRECs are a readout of thymopoiesis. The sjTREC in T cells isolated from cryopreserved PBMC were therefore quantified to determine whether thymopoiesis contributed to repopulation. After adjusting for sex, there were significant effects of TBI on sjTREC numbers in both CD8 (P = .025) and CD4 (P = .03) T cells and strong interacting effects of age with irradiation and dose in both populations (all P < .02) (Fig. 5M, 6M; Table E12). These changes were most significant in animals aged under 5 years. After adjusting for age at collection and sex, there was a highly significant (P < .001) interaction effect of dose and years since irradiation on sjTREC numbers in both populations (Table E13). sjTREC in animals 1 to 2 years out from irradiation exhibited dose-dependency (P < .001), with animals receiving doses lower than LD10 showing the lowest numbers of sjTREC (Fig. 5O, 6O). Collectively, these results demonstrate that thymopoiesis contributes to peripheral T-cell reconstitution in young animals for around 1 to 2 years after TBI in a dose-dependent manner. In contrast, the sjTREC numbers and naïve T-cell proportions were similar in older irradiated and nonirradiated animals were similar, suggesting thymic output does not drive extensive repopulation in older animals.

Discussion

Collectively, this 5-year study of a cohort of irradiated NHPs demonstrated that radiation causes subtle but multiyear changes in monocyte populations, transient shifts in DC distribution, and potentially life-long reductions in the proportions of circulating memory B cells and central memory T cells. Some of the outcomes measured were altered in a dose-dependent fashion, although the relationships between dose and immune outcome were sometimes directly proportional to dose and sometimes parabolic with animals receiving low (<LD10) or high (>LD90) being similar, but both extremes differing from animals receiving >LD10 to <LD90. More commonly, strong interacting effects were found between dose and both age at sample collection and time since irradiation. For many populations (eg, intermediate monocytes, mDC, total B and T cells, and all CD4 and CD8 T-cell subsets) the differences between irradiated and control were most apparent in the youngest animals in the cohort (age 3–4 years at time of sample collection) and were negligible in the eldest animals.

Although this study revealed several highly statically significant and long-lasting effects of radiation on the immune system, there were some limitations. First, the cohort consisted of animals irradiated as part of other research studies, and so the dosage, mitigator usage and age at irradiation distributions were skewed by the requirements of the underlying studies and not evenly distributed. Second, the sex distribution was skewed toward males making the study insufficiently powered for detection of sex differences in radiation responses. Third, this observational study did not assess cell function or appraise the numbers, phenotypes, or organization of the tissue resident cells necessary for immune responses in vivo. In addition, neutrophils and other granulocytic WBC were not immunophenotyped as they do not fractionate with PBMC. Neutrophil recovery is important for survival after TBI35 and will be a focus of future study in this cohort. Finally, logistical constraints meant PBMC had to be isolated and stained the day after blood collection. This may have affected some immunophenotyping markers or subsets; however, all samples were treated the same way and so any effects of the delay would be consistent across the study.

Young irradiated animals showed a decrease in the percentage of T cells in peripheral blood along with an increase in the proportion of naïve CD4 T cells, and both CD4 and CD8 naïve T-cell percentages showed strong interaction effects between irradiation status and age. Young (3–5 years of age) irradiated animals exhibited increases in CD4 and CD8 T-cell sjTREC after irradiation that were proportional to dose. The high degree of variability in sjTREC counts shortly after irradiation, especially within the LD>90 group, meant no significant dose-dependent effects were identified in the first year after exposure after correcting for age; however, 1 to 2 years after TBI the variability reduced and both CD4 and CD8 sjTREC numbers were dose dependent. There was no evidence for dose dependency in sjTREC numbers beyond this time point. Thus, younger animals responded to T-cell lymphopenia by greatly increasing thymic output in a dose-dependent manner for at least 2 years. The failure of older animals to continue to respond in the same way may be a consequence of age-induced thymic involution.28 Radiation exposure could also be causing thymic damage in older animals given that atomic bomb survivors,36 and experimentally irradiated mice37 and cynomolgus macaques21 all exhibit thymic abnormalities and decreased thymic function.

The age-dependent differences in the path for T-cell repopulation mean that while the short-term immunophenotypic consequences of TBI were more dramatic in younger NHP, the consequences for T-cell immunity may be more significant in older animals. TCR diversity naturally decreases with age,38 and our results indicate that in comparison to young animals, older NHP rely more on peripheral expansion than thymopoiesis after irradiation. This may mean older irradiated individuals are more likely to have gaps in immune coverage, particularly individuals irradiated at either low (<LD10) or high (>LD90) dose extremes. Although a recent study of 5- to 6-year-old NHP irradiated at 4 Gy did not identify defects in TCR alpha or beta sequence diversity 9 months after irradiation,22 the effect of TBI on TCR diversity in later years or at other doses remains to be investigated. Peripheral expansion is skewed by antigen exposure,39 and so the virome and microbiome may play an outsized role in shaping T-cell recovery in older irradiated animals.

In addition to radiation-induced T-cell defects, this study found that TBI led to a long-term reduction in the proportion of CD27+ memory B cells in irradiated adolescent and adult animals (5–11 years of age). The loss of lymphocyte CD27 after TBI has previously been observed in studies of irradiated mice,40 NHP,41 and humans,42,43 and has been proposed as a biodosimetry readout of acute radiation injury. However, the immunologic consequences of long-term circulating memory B-cell loss are unclear as long-lived plasma cells, which are more radiation resistant,44 and tissue resident memory B cells also contribute to maintenance of antibody-mediated memory. Primary humoral responses years after irradiation appear grossly normal based on studies of influenza vaccination in elderly atomic bomb survivors,45 West Nile virus challenge in old mice that were irradiated at a young age,46 and NHP vaccinated for tetanus, rabies or influenza years after irradiation.22,23 Antibody recall responses after irradiation have not been fully evaluated; however, atomic bomb survivors have not died of infectious disease at significantly higher rates, suggesting antibody responses in this population are still effective47 even if antibody titers against specific pathogens are reduced.48

Conclusions

This study found that years after acute TBI, Rhesus macaques had generally recovered near normal proportions of the evaluated immune cell populations; however, some key cellular mediators of adaptive immunity underwent subtle long-lasting phenotypic changes. These changes were most striking in young animals within the first few years of exposure and in animals exposed to near LDs of radiation (>LD90). Young irradiated animals responded to radiation-induced lymphopenia by increasing thymic output whereas older irradiated animals responded by increasing peripheral T-cell expansion.

Supplementary Material

Supplemental methods and data
supplemental figures

Acknowledgments—

The authors thank Jean Gardin, the staff of the Cline Laboratory and the Comparative Pathology Laboratory, Kristina Riebe, James Devore for technical assistance, and Dr Laura Hale for constructive discussions. Flow cytometry data were acquired in the Duke Human Vaccine Institute Flow Cytometry Facility, Durham, North Carolina, under the direction of Dr Derek Cain.

This research was supported by Department of Defense grant W81XWH-15–1-0574 (J.M.C.), National Institutes of Health (NIH)/National Institute of Allergy and Infectious Diseases (NIAID) grant U01 AI150578 (J.M.C.), and the NIH/NIAID Radiation Countermeasures Center of Research Excellence grant U19 AI67798 (principal investigator: Dr Nelson Chao). The Duke Regional Biocontainment Laboratory received partial support for construction from the NIH NIAID (UC6-AI058607, G20-AI167200 GDS).

Footnotes

Disclosures:: none.

Data sharing statement: Research data are stored in an institutional repository and will be shared upon request to J.M.C. (jmcline@wakehealth.edu).

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijrobp.2022.10.024.

References

  • 1.Kiang JG, Olabisi AO. Radiation: A poly-traumatic hit leading to multi-organ injury. Cell Biosci. 2019;9:25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.DiCarlo AL. Scientific research and product development in the United States to address injuries from a radiation public health emergency. J Radiat Res. Sep 13 2021;62:752–763. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Daniel S, Nylander V, Ingerslev LR, et al. T cell epigenetic remodeling and accelerated epigenetic aging are linked to long-term immune alterations in childhood cancer survivors. Clin Epigenetics. 2018;10:138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Ketterl TG, Chow EJ, Leisenring WM, et al. Adipokines, inflammation, and adiposity in hematopoietic cell transplantation survivors. Biol Blood Marrow Transplant. 2018;24:622–626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Muhic E, Mathiesen S, Nielsen MM, et al. Metabolic syndrome in male survivors of pediatric allogeneic hematopoietic stem cell transplantation: Impact of total body irradiation, low-grade inflammation, and hypogonadism. Transplant Cell Ther. 2021;27 778.e1–778.e8. [DOI] [PubMed] [Google Scholar]
  • 6.Fuks Z, Strober S, Bobrove AM, Sasazuki T, McMichael A, Kaplan HS. Long term effects of radiation of T and B lymphocytes in peripheral blood of patients with Hodgkin’s disease. J Clin Invest. 1976;58:803–814. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Yamaoka M, Kusunoki Y, Kasagi F, Hayashi T, Nakachi K, Kyoizumi S. Decreases in percentages of naive CD4 and CD8 T cells and increases in percentages of memory CD8 T-cell subsets in the peripheral blood lymphocyte populations of A-bomb survivors. Radiat Res. 2004;161:290–298. [DOI] [PubMed] [Google Scholar]
  • 8.Ilienko IM, Golyarnik NA, Lyaskivska OV, Belayev OA, Bazyka DA. Expression of biological markers induced by ionizing radiation at the late period after exposure in a wide range of doses. Probl Radiac Med Radiobiol. 2018;23:331–350. [DOI] [PubMed] [Google Scholar]
  • 9.Kusunoki Y, Yamaoka M, Kasagi F, Hayashi T, MacPhee DG, Kyoizumi S. Long-lasting changes in the T-cell receptor V beta repertoires of CD4 memory T-cell populations in the peripheral blood of radiation-exposed people. Br J Haematol. 2003;122:975–984. [DOI] [PubMed] [Google Scholar]
  • 10.Titova LD, Oradovskaia IV, Sharova NI, Iarilin AA. Sravnitel’naia otsenka soderzhaniia subpopuliatsiĭ T-limfotsitov, al’fa 1-timozina i autoantitel k épitelial’nym kletkam timusa u personala 30-kilometrovoĭ zony kontrolia avarii na Chernobyl’skoĭ AES [A comparative evaluation of the content of T-lymphocyte subpopulations, alpha 1-thymosin and autoantibodies to epithelial thymic cells in the personnel in the 30-kilometer control zone of the accident at the Chernobyl Atomic Electric Power Station]. Radiats Biol Radioecol. 1996;36:601–609 [in Russian]. [PubMed] [Google Scholar]
  • 11.Kurjane N, Bruvere R, Shitova O, et al. Analysis of the immune status in Latvian Chernobyl clean-up workers with nononcological thyroid diseases. Scand J Immunol. 2001;54:528–533. [DOI] [PubMed] [Google Scholar]
  • 12.Bazyka D, Anatolii C, Byelyaeva N, et al. Immune cells in Chernobyl radiation workers exposed to low dose irradiation. Int J Low Radiat. 2003;1:63–75. [Google Scholar]
  • 13.Kuzmenok O, Potapnev M, Potapova S, et al. Late effects of the Chernobyl radiation accident on T cell-mediated immunity in cleanup workers. Radiat Res. 2003;159:109–116. [DOI] [PubMed] [Google Scholar]
  • 14.World Health Organization. Health effects of the Chernobyl accident and special health care programmes: Report of the UN Chernobyl Forum Expert Group “Health.” Available at: https://www.who.int/publications/i/item/9241594179. Accessed February 16, 2022.
  • 15.Singh VK, Olabisi AO. Nonhuman primates as models for the discovery and development of radiation countermeasures. Expert Opin Drug Discov. 2017;12:695–709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Kennedy RC, Shearer MH, Hildebrand W. Nonhuman primate models to evaluate vaccine safety and immunogenicity. Vaccine. Jun 1997;15:903–908. [DOI] [PubMed] [Google Scholar]
  • 17.Havel PJ, Kievit P, Comuzzie AG, Bremer AA. Use and importance of nonhuman primates in metabolic disease research: Current state of the field. ILAR J. 2017;58:251–268. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Farese AM, Hankey KG, Cohen MV, MacVittie TJ. Lymphoid and myeloid recovery in rhesus macaques following total body x-irradiation. Health Phys. 2015;109:414–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.MacVittie TJ, Bennett AW, Cohen MV, Farese AM, Higgins A, Hankey KG. Immune cell reconstitution after exposure to potentially lethal doses of radiation in the nonhuman primate. Health Phys. 2014;106:84–96. [DOI] [PubMed] [Google Scholar]
  • 20.MacVittie TJ, Farese AM, Jackson 3rd W. The hematopoietic syndrome of the acute radiation syndrome in rhesus macaques: A systematic review of the lethal dose response relationship. Health Phys. 2015;109:342–366. [DOI] [PubMed] [Google Scholar]
  • 21.DeBo RJ, Register TC, Caudell DL, et al. Molecular and cellular profiling of acute responses to total body radiation exposure in ovariectomized female cynomolgus macaques. Int J Radiat Biol. 2015;91:510–518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Macintyre AN, French MJ, Sanders BR, et al. Long-term recovery of the adaptive immune system in rhesus macaques after total body irradiation. Adv Radiat Oncol. 2021;6 100677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Hale LP, Rajam G, Carlone GM, et al. Late effects of total body irradiation on hematopoietic recovery and immune function in rhesus macaques. PLoS One. 2019;14:e0210663. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Council NR. Guide for the Care and Use of Laboratory Animals: Eighth Edition. Washington, DC: The National Academies Press; 2011:246. [Google Scholar]
  • 25.Lynch HE, Sempowski GD. Molecular measurement of T cell receptor excision circles. Methods Mol Biol. 2013;979:147–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Liang KY, Zeger SL. Longitudinal data-analysis using generalized linear-models. Biometrika. 1986;73:13–22. [Google Scholar]
  • 27.Wang M.Generalized estimating equations in longitudinal data analysis: A review and recent developments. Advances in Statistics. 2014;2014. [Google Scholar]
  • 28.Crooke SN, Ovsyannikova IG, Poland GA, Kennedy RB. Immunosenescence: A systems-level overview of immune cell biology and strategies for improving vaccine responses. Exp Gerontol. 2019;124 110632. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ozanska A, Szymczak D, Rybka J. Pattern of human monocyte subpopulations in health and disease. Scand J Immunol. 2020;92:e12883. [DOI] [PubMed] [Google Scholar]
  • 30.Michalson KT, Macintyre AN, Sempowski GD, et al. Monocyte polarization is altered by total-body irradiation in male rhesus macaques: Implications for delayed effects of acute radiation exposure. Radiat Res. 2019;192:121–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Messaoudi I, Estep R, Robinson B, Wong SW. Nonhuman primate models of human immunology. Antioxid Redox Signal. 2011;14:261–273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Asquith M, Haberthur K, Brown M, et al. Age-dependent changes in innate immune phenotype and function in rhesus macaques (Macaca mulatta). Pathobiol Aging Age Relat Dis. 2012;2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Meyer C, Walker J, Dewane J, et al. Impact of irradiation and immunosuppressive agents on immune system homeostasis in rhesus macaques. Clin Exp Immunol. 2015;181:491–510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Sodora DL, Douek DC, Silvestri G, et al. Quantification of thymic function by measuring T cell receptor excision circles within peripheral blood and lymphoid tissues in monkeys. Eur J Immunol. 2000;30:1145–1153. [DOI] [PubMed] [Google Scholar]
  • 35.Gluzman-Poltorak Z, Vainstein V, Basile LA. Association of hematological nadirs and survival in a nonhuman primate model of hematopoietic syndrome of acute radiation syndrome. Radiat Res. 2015;184:226–230. [DOI] [PubMed] [Google Scholar]
  • 36.Ito R, Hale LP, Geyer SM, et al. Late effects of exposure to ionizing radiation and age on human thymus morphology and function. Radiat Res. 2017;187:589–598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Xiao S, Shterev ID, Zhang W, et al. Sublethal total body irradiation causes long-term deficits in thymus function by reducing lymphoid progenitors. J Immunol. 2017;199:2701–2712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Britanova OV, Putintseva EV, Shugay M, et al. Age-related decrease in TCR repertoire diversity measured with deep and normalized sequence profiling. J Immunol. 2014;192:2689–2698. [DOI] [PubMed] [Google Scholar]
  • 39.Mackall CL, Bare CV, Granger LA, Sharrow SO, Titus JA, Gress RE. Thymic-independent T cell regeneration occurs via antigen-driven expansion of peripheral T cells resulting in a repertoire that is limited in diversity and prone to skewing. J Immunol. 1996;156:4609–4616. [PubMed] [Google Scholar]
  • 40.Ossetrova NI, Stanton P, Krasnopolsky K, Ismail M, Doreswamy A, Hieber KP. Biomarkers for radiation biodosimetry and injury assessment after mixed-field (neutron and gamma) radiation in the mouse total-body irradiation model. Health Phys. 2018;115:727–742. [DOI] [PubMed] [Google Scholar]
  • 41.Blakely WF, Bolduc DL, Debad J, et al. Use of proteomic and hematology biomarkers for prediction of hematopoietic acute radiation syndrome severity in baboon radiation models. Health Phys. 2018;115:29–36. [DOI] [PubMed] [Google Scholar]
  • 42.Klucinski P, Mazur B, Sedek L, et al. Assessment of selected B cells populations in the workers of X-ray departments. Int J Occup Med Environ Health. 2014;27:467–473. [DOI] [PubMed] [Google Scholar]
  • 43.Rehakova Z, Sinkora J, Vlkova M, et al. CD27(+) peripheral blood B-cells are a useful biodosimetric marker in vitro. Physiol Res. 2008;57:589–600. [DOI] [PubMed] [Google Scholar]
  • 44.Slifka MK, Antia R, Whitmire JK, Ahmed R. Humoral immunity due to long-lived plasma cells. Immunity. 1998;8:363–372. [DOI] [PubMed] [Google Scholar]
  • 45.Hayashi T, Lynch HE, Geyer S, et al. Impact of early life exposure to ionizing radiation on influenza vaccine response in an elderly Japanese cohort. Vaccine. 2018;36:6650–6659. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Pugh JL, Foster SA, Sukhina AS, et al. Acute systemic DNA damage in youth does not impair immune defense with aging. Aging Cell. 2016;15:686–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Palmer JD, Tsang DS, Tinkle CL, et al. Late effects of radiation therapy in pediatric patients and survivorship. Pediatr Blood Cancer. 2021;68 (suppl 2):e28349. [DOI] [PubMed] [Google Scholar]
  • 48.Hakoda M, Kasagi F, Kusunoki Y, et al. Levels of antibodies to microorganisms implicated in atherosclerosis and of C-reactive protein among atomic bomb survivors. Radiat Res. 2006;166:360–366. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental methods and data
NIHMS1850954-supplement-Supplemental_methods_and_data.pdf (356.9KB, pdf)
supplemental figures
NIHMS1850954-supplement-supplemental_figures.pdf (2MB, pdf)

RESOURCES