A total of 455 female FVB/N-Tg(MMTVneu)202Mul/J mice (hereon denoted as MMTV-Neu; Stock # 002376, Jackson Laboratories) were bred from the frozen embryo, then purchased at 4 weeks of age and entered the study at 6 weeks of age. Mice were housed in the Specific Pathogen-Free Biological Research Facility (BRF) at CNL. Six mice were housed per cage in individually ventilated Thoren cages with ad-libitum access to food (Charles River Rodent Chow 5075) and reverse osmosis (RO) water. Mice were externally exposed to chronic gamma rays (⁶⁰Co) in the BRF Gamma Beam Hall for total absorbed doses of 0 (UT), 10, 100, or 2,000 mGy over 56 days. The 2,000 mGy was used as a control high dose. All animal husbandry and experimental procedures were approved by the Canadian Nuclear Laboratories Animal Care Committee in accordance with the standards of the Canadian Council on Animal Care (CCAC).

The experimental endpoints were 3.5, 6, and 8 months of age. These endpoints were chosen to examine tissues at multiple stages of mammary tumor development and post-exposure response; before tumor development (3.5 months; immediately following irradiation); during tumor development (6 months; 2.5 months post-exposure); and high tumor burden/metastasis development (8 months; 4.5 months post-exposure), as shown in Figure 1 . Mice were euthanized via exsanguination under Isoflurane gas anesthesia, followed by cervical dislocation. Blood, mammary glands, lung, spleen, and tumor samples were collected from 455 mice for further processing and analysis.

At 1.5 months of age, MMTV-Neu mice were exposed to an open beam ⁶⁰Co source (Gamma Beam 150C; Nordion) in the 30m-long Gamma Beam Irradiation Facility. The ⁶⁰Co source had a total activity of 2.97 Ci (as of May 2022). Due to the disparity in dose rates between the low dose (10 mGy and 100 mGy) and the high dose (2,000 mGy) groups, the mice were irradiated in separate batches. The low-dose cohorts were irradiated with 1.75’ of lead shielding. Both physical shielding and the distance of the animal cages from the primary source allowed the required dose rates of 7.7 µGy/h (for the total absorbed dose of 10 mGy) and 77 µGy/h (100 mGy). The high-dose cohort did not require shielding during irradiations; mice received an unshielded dose of 1.49 mGy/h (absorbed dose of 2,000 mGy). Dosimetry was completed using Exradin^® A8 ion chamber and SuperMax™ electrometer readings to determine average gamma radiation dose rates in the hall. Theoretical Monte Carlo N-Particle Transport Code (MCNP) modelling of the shielded source estimated that 61-64% of the total dose originated from pure ⁶⁰Co where the remaining contributions were from wall scatter (unpublished data; CNL internal technical memorandum). During dosimetry measurements and during the 56 days of exposure, passive dosimeters containing Harshaw thermoluminescent dosimeter TLD-100e LiF: Mg, Ti chips (Harshaw Chemical Co., Solon, OH) were placed in empty mouse cages distributed at the corners and centers of each animal rack to record total absorbed gamma dose. Average dose rates for gamma radiation during the animal exposure period were derived by dividing the total gamma dose, determined by the mean TLD measurement per animal rack, by the total exposure period (adjusted for beam downtime necessary to accommodate animal husbandry activities). While in the gamma beam hall, mice were housed in wooden racks in Thorene cages, at 23-24 °C, 37-40% humidity, and an atmospheric pressure of 754–759 mmHg.

Mice were anesthetized using Isoflurane gas and then peripheral blood was taken via cardiac puncture at the time of euthanasia and kept refrigerated under continuous rocking/shaking for no longer than 1-hour post-collection. Blood samples were analyzed on Zoetis Vetscan MS5 hematology analyzer. To collect plasma, whole blood was centrifuged at 12,000 × g for 10 minutes at 4 °C. Plasma was carefully transferred to 1.5 mL Eppendorf tubes, snap-frozen in liquid nitrogen and subsequently stored at -80 °C.

Blood plasma samples stored at -80 °C were thawed on ice, and aliquots of 50 µL were collected for analyses. The concentration of select cytokines, chemokines, and growth factors in the plasma was analyzed using BioPlex 200 system and Bio-Plex Pro™ Mouse Cytokine 23 plex panel, Group I (Bio-Rad, Hercules, California, USA). Each plate was designed and assembled according to the manufacturer’s instructions. Briefly, antibody coupled magnetic beads were incubated with standards and plasma samples. This was followed by incubation with biotinylated conjugate antibodies, and lastly incubation with streptavidin phycoerythrin (SA-PE) conjugate antibodies. Beads were washed using the Integra Viaflo Assist automated multi-channel pipette (Mandel Scientific, Canada) after each incubation to remove any unbound components. Two different dilutions of standards were run at the customary dilution factor of 4 and at a dilution factor of 40. Each plate was run twice on the BioPlex 200 array reader, on both the low photomultiplier (PMT) and high PMT settings with the appropriate standard selected. The cytokines were detected with a dual laser system: The green “reporter” laser (532 nm) and the red “classify” laser (635 nm) excite the PE dye and dyes inside the magnetic beads, respectively, to allow high throughput detection of all 23 cytokines at once. Data was analyzed using the BioPlex Manager™ 6.1 software. The standard curves were modelled using a 5-parameter logistic (5PL) regression, and cytokine concentration results were presented in pg/mL. Each sample was run in duplicate on different plates and two samples were consistently run on every plate to function as inter-plate controls; these controls allowed for normalization between plates and runs.

Spleens were harvested promptly following euthanasia, and a single-cell suspension was generated using the following process: dissociating tissues using a 3 ml syringe plunger and passing through a 70-μm filter (Bio Basic Canada Inc.), centrifugation at 1200 rpm for 10 min at 4 °C; and finally, washing with RPMI containing 2% fetal bovine serum (FBS) (Gibco™, Canada). The cell pellet was reconstituted in 1 mL of red blood cell lysis buffer (Roche) immediately followed by vortexing. Cells were washed with RPMI containing 2% FBS and filtered using a nylon mesh before counting. To obtain leukocytes from lungs, tumors, and mammary glands, tissues were dissociated into small pieces using dissection scissors and then mixed with extraction buffer (RPMI-10% FBS) containing 25U/mL collagenase VIII (Sigma). Sample homogenization was done at 37 °C using gentleMACS™ Dissociator (Milteney biotech, USA) followed by straining with a 70-μm filter. Lymphocyte isolation was performed using Percoll gradient centrifugation (Percoll^®, Millipore Sigma) according to the manufacturer’s instructions. Cells were filtered through a nylon mesh before counting. All the assays were performed on splenic lymphocytes; other tissues were used only for immune profiling.

For ex vivo NK cell intracellular IFNγ measurements, freshly derived spleen leukocytes were stimulated with either a combination of IL-2 (100U/ml, obtained from NCI Preclinical Repository, USA) and IL-12 (10 ng/ml) (eBioscience™), or with plate coated anti-NKp46 (BioLegend™) for 1 hour and then incubated in RP-10 media containing 5 μg/ml brefeldin A (Invitrogen™) for 4 hours, followed by intracellular staining. For ex vivo T cell intracellular IFNγ measurements, freshly derived spleen leukocytes were stimulated with anti-CD3/28 for 16 hours and then incubated in RP-10 media containing 5 μg/ml brefeldin A (Invitrogen™) for a further 4 hours, followed by intracellular staining.

Single-cell suspensions (1×10⁶ cells) were incubated at 4 °C for 10 min with α-CD16/32 (clone 2.4G2, from Bioexpress (USA)) to reduce non-specific binding. Cells were labelled with various combinations of fluorochrome-conjugated monoclonal antibodies (mAbs) and incubated at 4 °C for 25 min. The following mAbs were used: anti-TCRβ (H57-597), anti-CD8 (53-6.7), anti-CD49b (DX5), anti-IFNγ (XMG1.2), anti-CD11b (M1/70), anti-NKG2D (CX5) from eBioscience™; anti-CD19 (1D3), anti-CD4 (RM4-5), anti-F4/80 (T45-2342), anti-CD69 (H1-2F3), anti-Ly6C (AL-21), anti-Gr1 from BD Biosciences™, anti-CD43 (1B11) activation- Glycoform from BioLegend™ and Live/Dead Fixable Yellow Dead Cell Stain from Invitrogen™. The intracellular staining of IFNγ was performed using Cytofix/Cytoperm protocols (BD Biosciences™). Cells were acquired using Thermofisher Attune NxT flow cytometers, and data was analyzed using Kaluza 1.3 Analysis software (Beckman Coulter) or FlowJo (V10).

Tumors were measured weekly using digital calipers, both noninvasively over the skin of live mice during the study and at the time of post-mortem dissection. Mice with cumulative tumor volumes greater than 2.9 cm³ were euthanized at any point in the study, as per the clinical endpoints determined by the CNL Animal Care Committee.

The statistical analysis was performed using One-way ANOVA with Dunnett’s multiple comparisons test using Prism Version 8 (GraphPad Software).

The goal of this study was to investigate the impact of chronic low-dose gamma radiation exposure on the immune status of mice with respect to the development and progression of mammary cancer after whole body radiation. To achieve this, 1.5 month-old MMTV-Neu mice that spontaneously develop mammary adenocarcinomas were exposed to continuous whole-body ⁶⁰Co gamma-ray radiation over a period of 56 days. As a result, the mice received a total absorbed dose of 10, 100, or 2,000 mGy. Mice were sacrificed at three time points for analysis: 3.5 months of age (24 hours post-irradiation), 6 months of age (2.5 months post-irradiation), and 8 months of age (4.5 months post-irradiation) ( Figure 1 ).

Peripheral blood comprises erythrocytes and mature leukocytes that are relatively constant in mice from a particular genotype under physiological conditions, indicating a precise regulation of hematopoietic lineage differentiation and commitment (47). To assess the effects of LDR exposure on the blood components, we performed a complete peripheral blood analysis. We observed noteworthy differences in hematology metrics of blood from mice exposed to both low and high doses of radiation ( Figure 2A ), indicating that radiation exposure affected blood composition in these mice. LDR significantly, albeit minorly, increased the proportion of lymphocytes at all time points except at 6-month time-point where 100 mGy did not reach statistical significance ( Figure 2B ). On the other hand, HDR exposure showed no effect on lymphocytes at the 3.5 month time point, followed by a slight decrease in proportion at the 6 month time point ( Figure 2B ). Later, at 8 months of age, mice exposed to HDR had slightly increased lymphocyte proportions ( Figure 2B ). Generally, we observed decreased frequencies of both monocytes and neutrophils in the blood of mice following LDR exposure ( Figure 2C ). Mice exposed to HDR displayed lower monocyte frequencies at 3.5 and 8 month time points and a higher proportion of neutrophils at the 6-month time-point ( Figure 2C ). Taken together, these data suggest that radiation induces a significant impact on peripheral blood composition in a mouse model of spontaneous mammary gland tumorigenesis.

Next, we used flow cytometry to compare the proportions of different immune cells isolated from various tissues across radiation conditions as outlined in the gating strategy ( Supplementary Figure 1 ). We observed an increased frequency of NK cells (TCRβ^- DX5⁺NKp46⁺) isolated from the spleens of mice that received LDR compared to untreated controls ( Figure 3A ). Specifically, both 10 mGy and 100 mGy radiation exposures induced elevated proportions of NK cells at the 3.5 and 6 month time points. At the 8-month time point, the NK cell proportions returned to baseline for mice exposed to LDR ( Figure 3A ). In contrast, HDR exposure (2,000 mGy) did not significantly affect the frequency of NK cells in the spleen at the early time points, however, a slight but significant reduction was measured at the 8-month time point ( Figure 3A ).

Following NK cells, we evaluated the frequencies of splenic T-cells (TCRβ⁺ NKp46^-/DX5^-) with respect to CD4 (TCRβ⁺ CD4⁺) and CD8 (TCRβ⁺ CD8⁺) cell subtypes. Radiation exposure did not induce any difference in the proportions of CD4⁺ T-cells and CD8⁺ T-cells at any dose or time point ( Figure 3B ). Similarly, both low and high doses of radiation did not affect the proportion of B-cells (TCRβ^- CD19⁺) in the spleens ( Figure 3C ). Therefore, we measured a unique but subtle modulation of NK cells in response to chronic radiation exposure. Conversely, mice exposed to HDR experienced a late reduction in splenic NK cell proportion, with NK cells maintaining baseline levels following exposure until 4.5 months post-exposure.

Furthermore, we analyzed the proportions of these immune populations in the mammary glands and lungs of irradiated mice. At 3.5 months, we observed a slight but statistically significant increase in the proportion of NK cells in the mammary glands of mice exposed to 100 mGy LDR ( Supplementary Figure 2A ). Otherwise, we measured no statistically significant impact of LDR on NK cell proportions in the mammary gland or lungs ( Supplementary Figure 2A, B ). The impact of HDR was solely significant at 8 months, when we measured a reduction in the proportion of mammary gland NK cells ( Supplementary Figure 2A ). We observed no effect of HDR on NK cell proportions in both the mammary gland and lungs ( Supplementary Figures 2A, B ). Consistent with the spleen findings, radiation did not impact the proportion of T and B cells in other examined organs (data not shown). These results highlight a limited impact of chronic LDR exposure on the frequency of NK cell populations in different analyzed organs. Importantly, our data suggests that even the lowest dose of 10 mGy could induce immunomodulation by increasing NK cell proportion.

With higher proportions of NK cells measured following LDR exposure, we decided to investigate whether these cells exhibited changes in activation status. Using flow cytometry, we evaluated the expression of activation-associated surface markers. First, we checked the expression of CD69, an early marker of lymphocyte activation (48). LDR did not induce the expression of CD69 on NK cells at any time point compared to untreated controls; however, increased expression was noted following HDR exposure at the 3.5-month time-point ( Figure 4A ). Consistently, the expression of CD43, a transmembrane glycoprotein expressed on activated NK cells (49), was unchanged across all groups during all three time points ( Supplementary Figure 3 ). Next, we examined the expression of IL-2α receptor (CD25) that can demarcate activated NK cells (50, 51). We observed a positive trend of LDR-induced expression of CD25 on NK cells at 3.5 and 6-month time points, with 10 mGy and 100 mGy reaching statistical significance at 3.5 and 6-month time points, respectively ( Figure 4B ). This effect was not seen in mice exposed to HDR at any time point ( Figure 4B ). We also measured the expression of NKG2D, one of the activation receptors that plays a critical role in NK cell-mediated immune response to transformed cells (11, 52–55). Similarly, we found an increased expression of NKG2D with LDR at 3.5 and 6-month time points, while HDR-induced expression was not observed at any time point ( Figure 4C ). Taken together, these results indicated that LDR exposure can lead to the modulation of important NK cell activation receptors, but these effects were transient as they were not observed at 8-month time point.

The release of cytokines and cytotoxic granules is central to immune responses. To evaluate IFNγ production, a primary cytokine produced by NK cells, splenic lymphocytes were isolated from mice in each group and were stimulated ex vivo with either anti-NKp46 or IL-2/IL-12 for 4 hours. We also assessed the degranulation activity by measuring lysosomal-associated membrane protein 1(LAMP-1). In order to measure the functional activity of T cells, splenic lymphocytes were stimulated with anti-CD3/28 for 16 hours, followed by intracellular staining to determine IFNγ and LAMP-1 production.

We found that NK cells isolated from LDR-exposed mice at both 3.5 and 6-month time points that were stimulated with anti-NKp46 had increased proportions of NK cells producing IFNγ, compared to untreated controls ( Figure 5A ). This response was strictly observed in LDR-exposed groups, as HDR did not modulate IFNγ release. NK cells from mice exposed to any radiation condition exhibited slightly increased proportions of LAMP1-positive cells at the 3.5-month time-point. At the 6-month time point, only NK cells isolated from mice exposed to 100 mGy demonstrated increased proportions of LAMP1-expressing cells. At the 8-month time point, no effect of radiation was observed on IFNγ release or LAMP1 expression following anti-NKp46 stimulation ( Figure 5A ).

Cytokine stimulation (IL-2 and IL-12) slightly increased IFNγ production at the 6 and 8-month time points following LDR exposure, compared to untreated controls; the 10 mGy was not statistically significant at the 8-month time point ( Figure 5B ). T cells from all conditions responded similarly to anti-CD3/28 treatment, with no differences measured in the proportion of IFNγ⁺ or LAMP1⁺ T-cells ( Supplementary Figure 4 ). Collectively, these results indicate that exposure to LDR can impact NK cell function by increasing IFNγ production and degranulation in response to stimulus. Notably, HDR largely did not induce any functional activity regardless of the mode of NK cell stimulation.

Cytokines and chemokines are generally recognized as important mediators of the inflammatory response. To better understand the functional alterations of the immune system, we next analyzed the plasma levels of different cytokines in radiation-exposed and control mice. We measured a wide range of cytokines involved in anti and pro-inflammatory responses. We did not observe any noticeable difference in the plasma levels for most of the common cytokines ( Figure 6 ); however, among all assessed, plasma levels of IL-13, GCSF, KC, MIP-1B RANTES were found altered in different mice groups in all time-points ( Figure 6 ).

To probe further, we analyzed the proportion of various sub-populations of immune cells involved in inflammation via flow cytometry by staining splenic cells for surface markers that are known to be expressed specifically on the inflammatory compartment of immune cells (56). We noticed an increased proportion of macrophages (CD45⁺F4/80⁺ cells) in the spleens of mice exposed to LDR and not HDR at the 3 and 6-month time-point, compared to untreated controls, although 10 mGy did not reach statistical significance at the 6-month time-point ( Figure 7A ). In contrast, HDR decreased the proportion of macrophages at the 6-month time-point ( Figure 7A ). Both low as well as high-dose radiation reduced the frequency of myeloid-derived suppressor cells (Gr‐1⁺CD11b⁺) at 3.5-month time-point and no differences were observed between any groups at later time points ( Figure 7B ). A similar pattern was observed in the proportion of CD11b⁺ cells, where radiation impact was only observed at 3.5 moth time point ( Figure 7C ), while radiation did not induce any differences in the proportion of Gr-1⁺ cells among all tested groups ( Figure 7D ). Notably, at the 8-month time point, we observed no effect of any radiation condition on these populations. Taken together, these results suggest differential and time-dependent effects of radiation on cytokines and immune populations involved in inflammation.

Finally, we examined the effects of LDR on the process of tumorigenesis. In MMTV-Neu mice, tumors can be easily discerned through the skin, allowing us to estimate tumor incidence and latency in all mice throughout the study (n ≈ 100 per treatment group). Figure 8A shows tumor latency (age at first tumor development) as percent tumor-free mice vs. time. Pairwise statistical analysis of tumor latency curves using the Log-rank Mantel-Cox test (p<0.05) showed no significant differences between any of the treatment groups. However, pairwise analysis using the Wilcoxon test, which puts greater weight on earlier time points, demonstrated a significant decrease in tumor latency in the UT vs. 100 mGy groups ( Figure 8A ). Although statistically insignificant, there was a trend of increased tumor number and volume in the low-dose cohorts (10 and 100 mGy) versus the control mice ( Figures 8B, C ). Surprisingly, mice exposed to HDR behaved close to the control group in terms of tumor burden ( Figures 8B, C ). In conclusion, our findings suggest that chronic low-dose gamma radiation may influence early tumor development, as evidenced by the significant decrease in tumor latency in the 100 mGy group, along with the observed trend towards increased tumor number and volume in low-dose cohorts. However, these effects were not consistently strong or significantly robust across all statistical analyses, indicating that the impact of low-dose radiation on tumor progression is subtle.