If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Ionizing radiation causes acute damage to hematopoietic and immune cells, but the long-term immunologic consequences of irradiation are poorly understood. We therefore performed a prospective study of the delayed immune effects of radiation using a rhesus macaque model.
Methods and Materials
Ten macaques received 4 Gy high-energy x-ray total body irradiation (TBI) and 6 control animals received sham irradiation. TBI caused transient lymphopenia that resolved over several weeks. Once white blood cell counts recovered, flow cytometry was used to immunophenotype the circulating adaptive immune cell populations 4, 9, and 21 months after TBI. Data were fit using a mixed-effects model to determine age-dependent, radiation-dependent, and interacting effects. T cell receptor (TCR) sequencing and quantification of TCR Excision Circles were used to determine relative contributions of thymopoiesis and peripheral expansion to T cell repopulation. Two years after TBI, the cohort was vaccinated with a 23-valent pneumococcal polysaccharide vaccine and a tetravalent influenza hemagglutinin vaccine.
Results
Aging, but not TBI, led to significant changes in the frequencies of dendritic cells, CD4 and CD8 T cells, and B cells. However, irradiated animals exhibited increased frequencies of central memory T cells and decreased frequencies of naïve T cells. These consequences of irradiation were time-dependent and more prolonged in the CD8 T cell population. Irradiation led to transient increases in CD8+ T cell TCR Excision Circles and had no significant effect on TCR sequence entropy, indicating T cell recovery was partially mediated by thymopoiesis. Animals that were irradiated and then vaccinated showed normal immunoglobulin G binding and influenza neutralization titers in response to the 4 protein antigens but weaker immunoglobulin G binding titers to 10 of the 23 polysaccharide antigens.
Conclusions
These findings indicate that TBI causes subtle but long-lasting immune defects that are evident years after recovery from lymphopenia.
Introduction
Total body irradiation (TBI) causes acute, dose-dependent injury to multiple organs.
This is an important knowledge gap given the use of controlled radiation exposure for clinical purposes, such as bone marrow ablation, and the risk of unforeseen exposure due to clinical error, nuclear accident, or weaponization of nuclear devices. The late effects of TBI on the adaptive immune system are of particular concern given the system’s critical role in pathogen defense and vaccine-induced protection.
Analyses of patients with cancer treated with targeted radiation therapy and bone marrow transplant recipients undergoing TBI for pretransplant ablation have established that acute irradiation causes circulating white blood cell (WBC) numbers to drop sharply within days and then gradually recover over months.
However, recovery of cellularity does not equate to recovery of immune function. Indeed, follow-up of the populations irradiated during the 1945 atomic bombings (A-bomb) and the 1986 Chernobyl nuclear power plant accident indicates that TBI has long-term immunologic consequences. These include a decrease in the ratio of naïve to memory T cells, skewed T cell antigen receptor (TCR) repertoire,
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.
[The epidemiological analysis of monitoring of the immune status in liquidators of consequences of the Chernobyl accident for early identification of risk groups and diagnostics of oncological diseases. Report 1].
T cells of atomic bomb survivors respond poorly to stimulation by Staphylococcus aureus toxins in vitro: Does this stem from their peripheral lymphocyte populations having a diminished naive CD4 T-cell content?.
In contrast, a study of 2011 to 2013 seasonal influenza vaccinations in A-bomb survivors found no defects in serologic responsiveness compared with age-matched controls.
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. Oct 29 2018;36(45):6650-6659.
One caveat is that many survivors of the 1945 and 1986 exposures are now of the age where immunosenescence is likely to weaken their adaptive immunity.
More generally, studies of adaptive immunity in irradiated humans are limited by survivor bias, differences in ages at time of exposure, pre-existing morbidities, varied exposures and doses, and confounding factors postexposure.
Given these complications, the rhesus macaque nonhuman primate (NHP) model is often used for evaluation of immune function and radiation countermeasures.
Importantly, the rhesus immune system closely recapitulates that of humans, and NHPs are susceptible to many human pathogens and simian pathogens that are similar to human pathogens. Further, NHPs are inherently susceptible to inflammation-associated morbidities found in humans such as obesity and hypertension.
rhesus macaques show an acute loss of circulating granulocytes and B, CD4+, and CD8+ T cells days after TBI, with delays in CD4+ T cell recovery leading to a prolonged shift in the CD4/8 T cell ratio.
Twenty male rhesus macaques aged 5.8 ± 0.5 years were obtained from World Wide Primates Incorporated (Miami, FL). Animals were seronegative for simian retroviruses. All procedures were performed in accordance with the Guide for Care and Use of Laboratory Animals and conducted in compliance with the Wake Forest University Institutional Animal Care and Use Committee requirements in a facility accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care. Complete blood counts (CBCs) with differentials were conducted weekly.
Irradiation
Animals were randomized to receive 4 Gy TBI (n = 14) or sham irradiation (n = 6). This dose was selected to be below the LD10/30 (lethal dose to 10% of the population in 30 days) of 5.5 Gy previously described for male macaques.
On Day 0, animals were sedated via intramuscular (IM) injection of 15 mg/kg ketamine and then irradiated on a clinical linear accelerator using 2 lateral fields of 6 MV x-rays (174-cm source-axis-distance, 32 × 40 cm2 field size @ 100 cm from the source, nominal dose rate of 0.8 Gy/min @ 174 cm source-axis-distance, 2.7 minute irradiation time per field). Sham-irradiated animals underwent all procedural steps except that ionizing radiation was not delivered.
Vaccination
One hundred eighteen weeks after TBI, animals were vaccinated with 0.5 mL IM polyvalent polysaccharide vaccine against 23 Streptococcus pneumoniae strains (Pneumovax23; Merk) and 0.5 mL IM quadrivalent northern hemisphere 2018/19 seasonal influenza vaccine (FluLaval tetra; GlaxoSmithKline). A second 0.5 mL IM dose of the influenza vaccine was administered 4 weeks later.
Peripheral blood collection and processing
Animals were sedated by IM injection with 15 mg/kg ketamine and femoral vein blood was drawn into sodium heparin-coated collection tubes (BD Bioscience) then shipped overnight at room temperature. The following day, plasma and peripheral blood mononuclear cells (PBMCs) were separated using System-Histopaque-1077 tubes (Sigma). PBMCs were immunophenotyped and remaining cells and plasma cryopreserved.
Immunophenotyping
Fresh PBMCs were labeled for 40 minutes with fluorescently conjugated antibodies (targets and clones listed in Table E1) in phosphate buffered saline (PBS), 1% bovine serum albumin (BSA) (Invitrogen), washed, stained for viability with live-dead fix aqua (Invitrogen) in PBS (Sigma) and fixed in 4% paraformaldehyde (ThermoFisher). Fixed cells were analyzed using an LSRII flow cytometer (BD Bioscience) running Diva software (BD Bioscience) and data gated using FlowJo software (BD Bioscience) (Fig. E1). Baseline data were collected 5 months before irradiation.
Influenza microneutralization
Heat-inactivated (1 hour 56°C) sera were serially diluted in virus diluent (Dulbecco’s modified Eagle medium, 100 U/mL penicillin-streptomycin, 2 mM L-glutamine, 7.5% BSA, 1 μg/mL L-(tosylamido-2-phenyl) ethyl chloromethyl ketone (TPCK)-treated Trypsin; ThermoFisher) and mixed 1:2 with virus diluent containing 100 TCID50 of each influenza virus (A/Michigan/45/2015 [H1N1]pdm09, A/Singapore/INFIMH-16-0019/2016 [H3N2], B/Colorado/6/2017 [Victoria Lineage], or B/Phuket/3073/2013 [Yamagata Lineage]; all International Reagent Resource). Madin-Darby Canine Kidney (MDCK) cells (London Strain) (International Reagent Resource) were added and plates incubated overnight at 37°C 5% CO2. Plates were then washed in PBS, fixed in -20°C acetone (VWR) and air-dried. Plates were washed in PBS, 0.1% Tween-20 (Sigma) and incubated for 1 hour with mouse anti-Influenza A nucleoprotein clones A1 and A3 (Millipore) or mouse anti-Influenza B nucleoprotein clones B2 and B4 (Millipore) diluted 1:4000 in PBS, 7.5% BSA, 0.1% Tween-20, and 1% dry nonfat milk (ThermoFisher). After washing, plates were incubated for 1 hour with goat antimouse conjugated to Horseradish peroxidase (HRP), washed, and then incubated with O-phenylenediamine dihydrochloride HRP substrate (Sigma). The reaction was stopped using 0.5 N sulfuric acid (Sigma) and quantified using a Synergy H1 reader (BioTek).
Assay plates were coated with 2 μg/mL hemagglutinin (HA) (B/Phuket/3073/2013 [aa 16-546], [A/Michigan/45/2015] [H1N1]-pd09-like, B/Colorado/6/2017 [aa 16-546], or A/Singapore/ INFIMH-16-0019/2016X [H3N2]; all eEnzyme) washed in PBS 0.1% and Tween-20 (Sigma) and then blocked with 3% nonfat milk (Quality Biological). Sera serially diluted in PBS, 10% fetal bovine serum, 5% goat serum, 0.05% Tween-20 (all Sigma), and 1% nonfat milk, were applied to the plates and then incubated at 4°C overnight. After washing, plates were incubated with antihuman IgG Fc-HRP clone H2 (Sothern Biotech) for 2 hours, washed, and then incubated with tetramethylbenzidine HRP substrate (Sigma). The reaction was stopped using 1 M sulfuric acid (Sigma) and quantified using a Synergy H1 reader.
Signal Joint T Cell Receptor Excision Circle (sj(TREC)) Quantification
CD4+ and CD8+ cells were isolated from cryopreserved PBMC using magnetic-bead positive selection (Miltenyi Biotec). Cells were counted and then lysed in proteinase K (Roche). sjTREC in lysates and serially diluted, quantified NHP sjTREC standard plasmid
and NHP study sera were precleared using pneumococcal cell wall polysaccharide adsorbents and then assayed using a custom multiplex bead array similar to those described elsewhere.
Briefly, samples were incubated with fluorescent magnetic microspheres (Luminex Corp) labeled with pneumococcal polysaccharides (FXImmune Diagnostics). After sample binding, microspheres were washed in PBS, 1% BSA (Invitrogen) and incubated for 30 minutes at room temperature with 4 μg/mL biotinylated antihuman/rhesus IgG Fc clone H2 (Southern Biotech). Microspheres were washed and labeled with 8 μg/mL streptavidin-phycoerythrin (Invitrogen). Beads were read on a BioPlex 200 multiplex reader (Bio-Rad) and analyzed using BioPlex Manager software v6.2 (Bio-Rad). NHP test sera IgG-binding of each polysaccharide was then calculated relative to that of 007sp reference serum.
TCR sequencing
RNA was extracted using the RNeasy Mini Kit (Qiagen). TCR α/β libraries were generated with rhesus-specific reverse primers (Takara) in combination with the SMARTer Human TCR Profiling Kit. Libraries were purified using SPRIselect Beads (Beckman Coulter) then sequenced on a MiSeq instrument (Illumina). Sequence annotation and clonal assignments were performed using MiXCR.
A threshold of 1000 reads per clone was used, and clones were considered identical if their CDR3 sequence matched within 3 base pairs.
Statistical analyses
Except as noted, data were analyzed using Prism (GraphPad) and plotted as mean ± standard deviation. For results past 100 days postirradiation, data from ± 1 week were grouped together. Unless stated, data were analyzed using a mixed-effects repeated measures model with Greenhouse-Geisser correction for sphericity as implemented in Prism. Significant results were reported as F(degrees of freedom numerator, degrees of freedom denominator) = F statistic followed by the P value for the F statistic. Variables found to change significantly (P < .05) with time were interpreted as changing due to aging. Variables found to change significantly with radiation were interpreted as being altered by TBI irrespective of time. Variables found to change significantly in response to the interaction of time and radiation were interpreted as being temporally changed by TBI.
Multiple comparisons were performed by Fisher’s least significant difference without correction for repeated measures. Heatmaps were generated using RStudio software (RStudio) with the pheatmap package (Ravio Kolde).
Results
Fourteen animals received 4 Gy TBI, whereas 6 control animals were subject to a sham irradiation. All received supportive care (enteral feeding gavage, subcutaneous fluids, antibiotics, pain medication, blood transfusions) as required. Four animals in the TBI group were euthanized within 3 months due to opportunistic systemic infectious diseases (3 with Spironucleus sp. and 1 with Klebsiella pneumoniae). As the current study focused on the long-term effect of irradiation, data from these animals were excluded.
TBI causes transient neutropenia and lymphopenia
CBC and differentials were collected from the 16 animals at regular intervals. CBCs 200 days postirradiation were previously reported by Michalson et al.
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. Aug 2019;192(2):121-134.
TBI led to an acute decrease in circulating WBC, with mean count dropping to a nadir of 8.8% of baseline after 16 days (Fig 1A). WBC counts recovered over the next 40 days, with no significant difference in total cellularity being detected from 58 days post-TBI onwards. When fit to a mixed-effect model, total WBC count changed significantly with time (F[4.42,61.7] = 3.02, P = .02) and there was a significant interaction between irradiation and time (F[22,307] = 3.62, P < .0001), indicating that both aging and time since irradiation affected the total WBC.
Figure 1Acute loss and gradual recovery of nonhuman primate (NHP) white blood cell populations after total body irradiation (TBI). Counts of (A) white blood cells, (B) neutrophils, (C) lymphocytes, (D) eosinophils, and (E) monocytes in rhesus macaques sham-irradiated or irradiated at day 0. One outlier eosinophil count in the sham group was excluded. ∗P < .05 by Fisher’s least significant difference test.
Neutrophils and lymphocytes make up the bulk of WBCs in NHP, and both populations dropped within days of irradiation (Fig 1B,C). Neutrophil numbers recovered to those of controls within 4 weeks. Lymphocyte counts in the control animals were marginally higher than the irradiated group before irradiation. Regardless, it took 8 weeks after irradiation for lymphocyte numbers to return to baseline or match control animals. There was a transient over-recovery of lymphocytes before numbers stabilized. Both neutrophil and lymphocyte counts changed significantly over time (neutrophils: F[4.42,61.7] = 3.02, P = .02; lymphocytes: F[2.67,37.3] = 11.1, P < .0001) and showed significant interaction between irradiation and time (neutrophils: F[22,307] = 3.62, P < .0001; lymphocytes: F[22,307] = 8.62, P < .0001). Together, these results indicated that absolute numbers of circulating neutrophils and lymphocytes were affected by both TBI and aging and that the effect of radiation decreased over time.
In addition to neutrophils, the numbers of 2 other populations pivotal to innate immunity, eosinophils and monocytes, dropped dramatically after irradiation and then rapidly recovered (Fig 1D,E). Both populations exhibited time-dependent decreases (monocytes: F[4.25,59.3] = 3.44, P = .012; eosinophils F[4.57,63.6] = 2.48, P = .045) although only monocytes exhibited an interaction between irradiation and time (F[22,307] = 3.47, P < .0001).
Overall, these results are consistent with previous studies in mice, NHPs, and humans showing that high-dose TBI leads to an acute loss of circulating WBCs followed by recovery over several weeks. They also reaffirm previous findings that the major WBC populations decrease in number with age. The prospective model reported here is thus consistent with published data and represents a relevant model for investigation of the effects of TBI on immunity.
TBI perturbs the long-term balance between naïve and memory T cells
Circulating lymphocytes are an equilibrium of naïve, effector, and memory B and T cells. We therefore immunophenotyped PBMC to characterize the balance of these populations plus the frequency (%) DC, in the 21 months after TBI. Importantly, the first postradiation timepoint for this analysis was approximately 3.5 months after exposure, when absolute numbers of WBC had recovered (Fig 1A).
To evaluate changes in DC, we measured the percentage of myeloid (mDC) and plasmacytoid DCs (pDC)
within PBMC. Although the percentage of cells classified as mDC and pDC decreased over time (mDC: F[1.046,14.64] = 8.124, P = .0117; pDC F[1.048,22.00] = 14.77, P = .0008) (Fig 2A,B), there was no significant effect of irradiation or interaction between time and irradiation. This suggested that although aging caused a decrease in frequency of circulating DC, TBI did not have any long-term effect.
Figure 2Homeostatic rebalancing of B, T, and dendritic cells in the peripheral blood of rhesus macaques after total body irradiation (TBI). Peripheral blood mononuclear cells (PBMCs) were collected at baseline and/or 122, 282, 664 days (4, 9, 21 months) after irradiation. PBMCs were stained for viability and CD45 plus (A, B) CD3, CD20, HLA-DR, CD14, CD123, and CD11c (dendritic cells), (C-E) CD20, CD27, and CD80 (B cells), or (F-H) CD3, CD4, and CD8 (T cells) and analyzed by flow cytometry. Data are expressed as percentages of the parent population noted on the y-axis. ∗P < .05 by Fisher’s least significant difference test.
The frequency of circulating B cells fell significantly as time passed (F[2.284, 31.98] = 12.41, P < .0001) (Fig 2C). Radiation exposure did not have a significant effect on this frequency. To further evaluate the long-term effect of radiation on B cells, we measured the percentage of B cells expressing CD80 at baseline and after irradiation and CD27 after irradiation only. CD80 provides costimulatory signals from T cells and is a marker of B cell activation. CD27 expression correlates with B cell somatic hypermutation and is a marker of memory B cells.
The percentage of CD80high B cells decreased significantly with time (F[1.229,22.94] = 9.203, P = .0039) (Fig 2D). The percentage of CD27+ B cells changed significantly with time (F[1.428,19.99] = 56.06, P < .0001) (Fig. 2E), but irradiation had no significant effect on the expression of either marker. Collectively, these results imply that although aging led to a drop in the percentage of circulating B cells, radiation did not have a long-term effect on B cell frequency or naïve/memory balance.
Although the percentage of circulating B cells fell as the animals aged, the percentage of T cells increased (F[2.18,30.5] = 22.2, P < .001) (Fig 2F). Peripheral T cells in primates are primarily CD4+ or CD8+ plus a small percentage of CD4+ CD8+ cells of unclear function.
Although the ratio of CD4+ to CD8+ T cells changed over time (F[1.213,16.99] = 50.67, P < .0001) (Fig 2G), there was no significant effect of irradiation or interaction between time and irradiation on either the percentage of T cells or the CD4/CD8 ratio. There was no significant difference in CD4+ CD8+ frequency between the control and irradiated groups (Fig 2H). When considered in combination with the lymphocyte absolute counts over time (Fig 1C), these data indicated that lymphocyte recovery post-TBI did not lead to long-term changes in T cell numbers beyond those caused by aging.
To appraise the long-term effect of TBI on T cell subpopulations we evaluated CD28 and CD95 expression for identification of naïve (CD28midCD95low), effector memory (TEM: CD28low/-CD95mid/high), and central memory (TCM: CD28highCD95high) populations.
The relative proportion of CD4+ and CD8+ naïve cells changed with time (CD4+: F[1.912,26.76] = 7.155, P = .0036; CD8+: F[11.891,26.47] = 15.55, P < .0001) and with radiation exposure (CD4+: F[1,14] = 8.898, P = .0099; CD8+: F[1,14] = 6.389, P = .0241) (Fig 3A,D). The percentage of CD4+ naïve T cells, but not CD8+ naïve T cells, also showed a significant interaction between time and irradiation (F[2,28] = 12.74, P < .0001). The proportions of naïve T cells in irradiated animals were significantly lower than in age-matched controls at the first post-TBI timepoint tested (CD4+: P = .0002; CD8+: P = .0083). Thus, aging led to a drop in the frequency of naïve CD4+ T cells and an increase in the frequency of naïve CD8+ T cells. Further, irradiation led to significant decreases in the frequencies of both CD4+ and CD8+ naïve T cells. The effect of radiation on naïve CD4+, but not naïve CD8+, T cells decreased over time.
Figure 3Homeostatic rebalancing of CD4+ but not CD8+ T cell populations in the peripheral blood of rhesus macaques after total body irradiation (TBI). Peripheral blood mononuclear cells (PBMCs) were collected at baseline and/or 122, 282, 664 days postirradiation. PBMCs were stained for viability and CD45 plus CD3, CD4, CD8, CD45RA, CD95, and CD28 and gated on (A-C) CD4 or (D-F) CD8 plus, (A, D) naïve (CD28midCD95low), (B, E) effector memory (CD28low/-CD95mid/high), and (C, F) central memory (CD28highCD95high) populations. Data are expressed as percentages of the parent population noted on the y-axis. ∗P < .05 by Fisher’s least significant difference test.
The frequencies of both TEM populations were significantly affected by time (CD4+: F[1.701,23.81] = 6.021, P = .0102; CD8+: F[1.968,27.55] = 10.45, P = .0004). The frequency of CD8+, but not CD4+, TEM was also significantly disturbed by irradiation (CD8+: F[1,14] = 25.66, P = .0002) (Fig 3B,E). These data indicated that aging led to an increase in TEM frequency, with radiation causing a sustained increase in the frequency of CD8+ TEM.
CD8+, but not CD4+, TCM were subject to significant time-dependent effects (CD8+: F[1.828,25.60] = 4.277, P = .0278) and were sensitive to irradiation (CD8+: F[1,14] = 17.31, P = .001) (Fig 3C,F). In both CD4+ and CD8+ TCM populations there was significant interaction between time and radiation exposure (CD4+: F[2,28] = 4.538, P = .0196; CD8+: F[2,28] = 4.490, P = .0204). This indicated that aging did not affect the frequency of CD4+ TCM over the 21-month study, but that the frequencies of both CD4+ and CD8+ TCM were affected by radiation exposure, with the effect on CD8+ TCM being more prolonged.
Thymopoiesis and peripheral expansion reconstitute the peripheral T cell compartment after TBI
The number of circulating T cells dropped dramatically a few days after TBI and recovered over several months (Fig 1C). This reconstitution could be mediated by increased thymic output of naïve cells (thymopoiesis) and/or by peripheral expansion of naïve or memory cells. This is an important distinction, as thymic radiation damage may corrupt TCR selection,
while peripheral expansion cannot generate TCR diversity.
During thymopoiesis, excised TCR DNA forms TCR rearrangement excision circles (TRECs). TRECs are not replicated during proliferation, and so TRECs become diluted with subsequent cell divisions. TREC numbers are thus highest in populations of recent thymic emigrants and lowest in cell populations that have undergone extensive peripheral expansion.
To determine whether TBI increased thymic output, we quantified TREC in CD4+ and CD8+ T cells isolated from the control and irradiated animals (Fig 4A,B). We did not observe a statistically significant change in CD4+ or CD8+ TREC numbers with respect to either time or irradiation. However, CD4+ T cells in 1 irradiated animal exhibited a transient 70-fold increase in sjTREC at the first timepoint after irradiation (122 days; outlier shown in plot), and several of the irradiated animals showed a transient spike in CD8+ TREC numbers after irradiation, as evidenced by the larger standard deviations of the irradiated samples 9 months (~282 days) after irradiation. Interpreting these data is complicated by the animals in the irradiated group having higher and more variable CD4+ and CD8+ sjTREC counts than the control group preirradiation. However, they suggest that if the thymus contributes to CD4+ T cell repopulation after irradiation then it primarily does so acutely (earlier than 3 months after exposure) and that the contribution to CD8+ T cell repopulation is transient and variable.
Figure 4Thymopoiesis contributes to T cell repopulation after total body irradiation (TBI). (A, B) CD4+ and CD8+ T cells were isolated by positive selection from peripheral blood mononuclear cells (PBMCs) collected at baseline, 122, 282, and 664 days postirradiation and their T cell receptor (TCR) rearrangement excision circle (TREC) contents quantified by quantitative polymerase chain reaction (qPCR). (C-H) TCRα and TCRβ chains were sequenced from PBMC at baseline and 9 months postirradiation. Panels (C, D) show cumulative frequency distribution of top 10 clones in each animal at each timepoint. Each color represents a distinct clone, and clones found in the same animal at both timepoints are numbered. Panels (E, F) show entropy for each chain in each animal at baseline and 9 months postexposure. Lines link samples from the same animal. Panels (G, H) show the mean and per-animal cumulative frequencies of overlapping clones between baseline and 9 months postirradiation. (A color version of this figure is available at https://doi.org/10.1016/j.adro.2021.100677.)
To determine whether TBI caused long-term changes to TCR diversity, we sequenced TCRα and β mRNA from PBMC collected at baseline and 9 months postirradiation. The 9-month timepoint was selected as T cell numbers had recovered to near baseline (Figs 1C and 2F). Cryopreserved PBMC from 5 control and 6 irradiated animals yielded sufficient high-quality RNA for baseline and postexposure TCRα and/or TCRβ sequencing. We first evaluated the top 10 α and β clones in each animal at baseline and 9 months post-TBI. The distribution and identity of the immunodominant clones in each of the nonirradiated animals changed considerably between timepoints, revealing the dynamic nature of the TCR repertoire. We observed similar dynamism in immunodominant clones in the irradiated group (Fig 4C,D). To evaluate global TCR diversity beyond the top 10 clones, we calculated the TCRα and TCRβ Shannon entropy index for each animal at both timepoints. This index incorporates both the number of clones and frequency of each clone, and the higher the index, the greater the diversity of T cell clones in an animal.
Although entropies in each animal changed over time, there were no consistent trends (Fig 4E,F), and mixed-effects modeling did not indicate any effect of time or radiation on the entropies. Thus, radiation did not significantly constrain or enhance overall T cell repertoire diversity. Next, we determined the overlap in clones between baseline and 9 months post-TBI for each animal. The cumulative frequencies of overlapping TCRα and TCRβ clones over time were less in irradiated animals, although these differences did not reach statistical significance (Fig 4G,H). There was, therefore, a trend for irradiated animals to have a higher proportion of clones at 9 months that were not present at baseline.
Collectively, these data revealed that although dominant TCR clones changed with time, single high-dose irradiation did not lead to a significant long-term loss of TCR diversity or sustained domination of the repertoire through expansion of a small number of clones. New TCR clones only arise during thymopoiesis, and so these findings further support the model that long-term repopulation and homeostasis of circulating T cells after TBI were partially driven by thymic output rather than solely by peripheral expansion.
TBI leads to long-term deficiencies in response to select antigens
These findings do not mean that the adaptive immune systems of irradiated animals fully recovered. We therefore evaluated in this cohort the immune response to 2 multivalent vaccines: a 23-serotype pneumococcal polysaccharide vaccine (23vP) and a tetravalent 2018/2019 seasonal influenza vaccine containing 4 influenza HA proteins. The pneumococcal vaccine and the first dose of the influenza vaccine were administered approximately 118 weeks (826 days) after TBI or sham irradiation. A booster influenza vaccine dose was administered 4 weeks later. Serum was collected from each animal immediately before vaccination (baseline) and 1-, 4-, 5-, 8-, and 16-weeks post.
HA-specific IgG titers for each animal were normalized to baseline and plotted against time. The area under each curve (AUC) was calculated as shown in Figure 5A for a representative animal/HA. The IgG responses for all animals and all HAs are summarized in Figure 5B. With the exception of irradiated animal r6, all vaccinated animals had detectable IgG responses to at least 3 of the vaccine HAs. The majority of control (n = 4 of 6) and irradiated (n = 6 of 10) animals had IgG responses to all 4 HAs, and there were no significant differences between the control and irradiated groups in the AUC of the IgG responses for any of the HAs. In addition, there were no significant differences in IgG responses between the control and irradiated groups after the animals had received only the prime dose (Fig E3). Thus, irradiation did not affect the amount of anti-HA IgG produced in response to vaccine strain HA proteins.
Figure 5Two years after total body irradiation (TBI), irradiated animals show normal antibody responses to seasonal influenza vaccination. One hundred eighteen and 122 weeks after irradiation, macaques were vaccinated with a tetravalent seasonal influenza vaccine. Sera were assayed for (A, B) hemagglutinin-binding immunoglobulin G (IgG) and (C,D) neutralizing antibodies against the influenza strains included in the vaccine. (A) Baseline-normalized hemagglutinin (HA)-binding IgG for each animal and HA were plotted as shown and area under the curve (AUC, shaded) calculated. (B) Heatmap shows log(AUC + 1) for the above-baseline response to each HA over time for all animals and HAs. (C) Neutralization titers for each influenza strain and animal were plotted as shown and the AUC calculated. (D) Heatmap shows the AUC for the neutralization response over time for all animals and influenza strains.
Sera were then tested for their ability to neutralize the 4 influenza virus strains represented in the vaccine. Neutralizing titers for each virus were plotted against time and AUC calculated as shown in Figure 5C. The AUCs for all 4 viruses and 16 animals are shown in Figure 5D. Baseline normalization was not required as there were no detectable neutralizing responses before the boost.
The responses to the 2 influenza B vaccine strains were weak in all animals, with no animals generating neutralizing antibodies to B/Colorado/6/2017 and only 2 generating neutralizing antibodies to B/Phuket/3073/2013. Influenza B does not naturally infect NHPs.
This was borne out by the neutralizing responses to A/Michigan/45/2015 and A/Singapore/INFIMH-16-0019/2016. There were no significant differences in influenza A responses between the irradiated and sham-irradiated animals (Fig 5D).
The humoral response to 23vP was quantified by measuring serum IgG for each of the 23 pneumococcal polysaccharides using a multiplex bead array. IgG titers for each polysaccharide were normalized, plotted against time, and then the above-baseline AUC was calculated. An example plot for 1 animal and serotype is shown in Figure 6A, and the AUCs for all serotypes and animals are shown in Figure 6B. Some serotypes, for example, 3, 2, 10A, and 9V, elicited similar IgG responses in control and TBI animals. However, the irradiated animals showed significantly weaker IgG responses to 10 of the 23 serotypes in the vaccine. The difference in response to serotype 19F was particularly striking, with 100% (n = 6 of 6) of control animals exhibiting a response and only 10% (n = 1 of 10) of the TBI animals responding (P < .00001). In addition, while all 6 of the control animals showed some IgG response to a minimum of 17 serotypes, 3 of the 10 irradiated animals showed responses to 12 or fewer serotypes. Therefore, despite the total WBC counts and frequencies of DC, B, and T cell populations all returning to their expected setpoints, irradiated animals mounted significantly weaker humoral responses to select bacterial polysaccharide antigens over 2 years after radiation exposure.
Figure 6Selectively deficient immunoglobulin G (IgG) responses to 23-valent polysaccharide vaccine 2 years after irradiation. One hundred eighteen weeks after irradiation, macaques were vaccinated with a 23-valent pneumococcus polysaccharide vaccine and then serotype-specific IgG antibodies were assayed. (A) Data were plotted as shown and area under the curve (AUC, shaded) calculated. (B) Heatmap shows log(AUC + 1) for the above-baseline response to each serotype over time. ∗P < .05, †P < .01, ‡P < .001, §P < .0001 by t test.
Most previous studies of the long-term effect of radiation on immunity have been observational, retrospective analyses years after the initial radiation exposure or have been performed in human patients with underlying diseases. Here, we report a multiyear prospective study of immunologic recovery in rhesus macaques after 4 Gy TBI. We found that after a period of neutro- and lympho-penia, the numbers and ratios of the key cellular mediators of immune responses returned to those of age-matched sham-irradiated controls. However, radiation exposure led to subtle, long-term differences in the ability of animals to mount a humoral response to polysaccharide antigens in the setting of vaccination.
The animals in the current study recovered from lymphopenia approximately 3 months after TBI, but both CD4+ and CD8+ T cells were initially skewed toward memory populations, and the memory:naïve balance gradually recovered over time. This suggests T cell repopulation was initially driven by peripheral expansion of surviving memory cells. This is consistent with prior observations from lymphopenic mice, humans, and irradiated NHPs.
we did not observe a significant change in TCR sequence entropy once the memory:naïve cell balance returned to baseline, suggesting any TCR skewing was transient.
Although we did not observe a statistically significant difference in TREC numbers 4 to 9 months after radiation exposure, T cells from irradiated animals had greater variability in CD8 TREC numbers in comparison to the nonirradiated controls. This implies that thymic output contributed to the peripheral repopulation of CD8 cells but that the degree or kinetics of thymopoietic recovery varied between animals. Notably, this phenomenon has also been reported in other NHP TBI studies.
The transient decrease in the CD4/8 T cell ratio following irradiation indicated that CD8 T cells recovered faster than CD4 cells, as previously observed in patients recovering from lymphopenia.
vaccinated 5.1- to 6.4-year-old macaques with 23vP 2 years after irradiation at 7.5 to 8.5 Gy using linear accelerator-derived photons. Comparison of IgG responses to 11 polysaccharide antigens at baseline and 16 weeks after vaccination suggested irradiated animals may be unable to generate a sustained response to select polysaccharides 2 years after TBI. The current prospective study reinforces this finding, as the IgG response to 10 of the 23 pneumococcal polysaccharide antigens tested was either weakened or completely ablated 2 years after TBI. This indicates that despite a quantitative recovery of a TBI-ablated immune system, there are qualitative deficiencies in adaptive immunity 2 years postirradiation. Interestingly, the responses to specific serotypes differed between the 2 studies, with Hale and colleagues observing compromised responses to serotype 5, a serotype for which we found no difference in response postirradiation. Conversely, Hale et al reported only a nonsignificant reduction in the number of animals responding to serotype 19F (75% of controls vs 44% of TBI animals). These distinctions may reflect radiation dose-dependent differences in response, the use of a binary classifier for response in Hale et al. (above-baseline IgG titers at 16 weeks after vaccination) versus a quantitative metric (area under the time-response curve from 0-16 weeks) in the current study, or the preclearance of nonspecific homologous antibodies from sera in the current work but not in the assays reported by Hale et al.
Although irradiated animals had deficiencies in their ability to generate IgG responses to several polysaccharide antigens, we did not observe any defect in IgG responses to any of the 4 influenza protein antigens administered in parallel to the polysaccharides. We tested more polysaccharides than proteins. However, our previous study of NHPs irradiated at a mixture of ages and doses also found no significant difference in the ability of the irradiated animals to mount an IgG response to tetanus toxoid protein.
Further, our study of A-bomb survivors indicated that irradiation did not cause a deficit in the ability of people to mount an antibody response to seasonal trivalent influenza vaccines.
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. Oct 29 2018;36(45):6650-6659.
Our cumulative findings in multiple cohorts and species therefore suggest that high-dose TBI causes long-term defects in the ability to mount an adaptive immune response to polysaccharide but not protein antigens.
Although humoral responses to protein vaccines arise from follicular B cell differentiation to memory B cells, a process that requires T follicular helper cell help (T-dependent), responses to plain polysaccharide vaccines such as 23vP are T-independent and don’t generate memory B cells. Instead, polysaccharide vaccines likely act by cross-linking B cell antigen receptors on splenic marginal zone B (MZ B) cells to drive an extrafollicular plasma cell response. Notably, children under 2 years old have few MZ B cells and do not mount effective responses to 23vP but can create humoral immunity in response to protein vaccines.
Although the effect of radiation on splenic MZ B cells in NHP has not been reported, we previously observed that TBI of NHP causes an acute contraction of splenic B cell counts and disrupts the splenic architecture.
The poor response to polysaccharide antigens observed in irradiated animals may, therefore, be a consequence of MZ B cell loss or malfunction.
Regardless of the mechanisms, our finding is of importance as it suggests irradiated individuals may be less able to mount a response to some infections or vaccines even after the WBC stabilizes. This is of particular concern with regard to S. pneumoniae as pneumonia is the most common serious infection found in patients with cancer.
Further, human infections by S. pneumoniae serotypes 19F, 11A, and 9N, the 3 serotypes with the most significant radiation-induced deficiencies in IgG response, have each been associated with increased mortality.
Collectively, the study reported here adds to the growing body of mouse, NHP, and human data indicating that in addition to the dramatic acute effect of radiation exposure, TBI leads to subtle long-term immunologic defects.
Acknowledgments
The authors would like to thank Jean Gardin, John Olson, Matthew Dwyer, Russell O’Donnell, James Devore, and Melissa Samo 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, NC, under the direction of Dr. Derek Cain. Influenza neutralization assays were performed in the Virology Unit of the Duke RBL under the management of Dr. Thomas H. Oguin, III. TCR sequencing data were acquired in the Duke Human Vaccine Institute Viral Genetic Analysis Facility, Durham NC, under the direction of Dr. Wilton Williams.
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.
[The epidemiological analysis of monitoring of the immune status in liquidators of consequences of the Chernobyl accident for early identification of risk groups and diagnostics of oncological diseases. Report 1].
T cells of atomic bomb survivors respond poorly to stimulation by Staphylococcus aureus toxins in vitro: Does this stem from their peripheral lymphocyte populations having a diminished naive CD4 T-cell content?.
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. Oct 29 2018;36(45):6650-6659.
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. Aug 2019;192(2):121-134.
Sources of support: This research was supported by Department of Defense grant W81XWH-15-1-0574 (J.M.C., G.D., G.D.S., A.N.M., B.R.S.) and the NIH/NIAID Radiation Countermeasures Center of Research Excellence grant U19 AI67798 (J.M.C., G.D.S., A.N.M., B.R.S.). The Duke Regional Biocontainment Laboratory (RBL) received partial support for construction from the National Institutes of Health, National Institute of Allergy and Infectious Diseases (UC6-AI058607: GDS).
Disclosures: The authors have no conflicts of interest to declare beyond the funding statement.
Data sharing statement: Research data are not available at this time.
00035-X&id=gr1.jpg){kind=link}
00035-X&id=gr2.jpg){kind=link}
00035-X&id=gr3.jpg){kind=link}
00035-X&id=gr4.jpg){kind=link}
00035-X&id=gr5.jpg){kind=link}
00035-X&id=gr6.jpg){kind=link}