Influenza A virus following radiotherapy amplifies lung injury and monocyte-derived macrophage responses

Abstract

Improvements in radiation therapy (RT) for thoracic cancers have increased survival, thus preventing radiotoxicity in normal lung tissue becomes even more important. Respiratory infection is a lung stressor that increases the risk of RT toxicity. However, this risk factor remains under-studied with no effective treatment approaches. While RT is toxic to tissue-resident alveolar macrophages, recruited monocyte-derived macrophages (MDMs) drive fibrogenesis. We therefore investigated how these macrophage populations are impacted by a respiratory infection subsequent to lung RT. Mice received whole thorax RT (5–12.5 Gy) then influenza A virus (IAV) 1 or 20 weeks later. Chronic lung injury and acute and chronic macrophage responses were evaluated. RT plus IAV was lethal at doses that were well-tolerated when either were administered singly. IAV potentiated chronic pathology from even a benign RT dose of 5 Gy, even when IAV was delayed for 20 weeks. Macrophage dynamics shifted towards more predominant pro-inflammatory, pro-fibrotic MDM responses. Acutely, RT plus IAV amplified loss of tissue-resident alveolar macrophages but increased inflammatory MDMs. Expression of maturation receptors and antigen presentation factors by inflammatory MDMs decreased while pro-fibrotic factors increased. These novel findings warrant further investigation of the risks of respiratory infection for those receiving thoracic radiation.

Introduction:

Radiotherapy (RT) remains an indispensable treatment for thoracic cancers. However, despite significant technical advances, radiation-induced lung injury continues to affect a significant number of RT patients (1–3). Advances in cancer treatment have increased survivorship of thoracic cancers such as lung and breast cancer (4, 5). However, as survival has improved, and more patients live with cancer and stable disease, preventing radiation-related toxicities to normal tissues has become more important to ensure quality of life (6). Radiation-induced lung injury consists of pneumonitis and pulmonary fibrosis (7). Radiation pneumonitis is a common but serious side effect occurring in 17–47% of patients receiving thoracic RT (8, 9). These patients are at high risk of developing radiation-induced pulmonary fibrosis as a late toxicity (7, 10). Progression of radiation-induced pulmonary fibrosis to a severe condition is not well understood, but is dictated by additional stressors including age, interstitial lung disease (11) or respiratory infection (9). Pulmonary fibrosis is debilitating, irreversible, is often life threatening, and has limited treatment options (12).

There is some clinical evidence that a respiratory infection occurring post-thoracic RT increases the risk of developing RT-related toxicities (9). Our preclinical studies demonstrated that a respiratory infection by influenza A virus (IAV) after RT has significant negative effects that intensify both rodent mortality and radiation-induced lung injury (13–16). Despite this risk to patients receiving thoracic RT, the impact of a subsequent respiratory infection on developing treatment related toxicities remains ill-defined, and no biological targets are available for direct treatment approaches.

Previous studies demonstrated that monocyte-derived macrophages (MDMs) drive the development of radiation-induced pulmonary fibrosis (17). This subset is functionally distinct from tissue-resident alveolar macrophages (TR-AMs) and vary in their capacity to carry out inflammatory versus homeostatic functions (18–20). While TR-AMs are generally immunosuppressive and function to maintain or reestablish lung homeostasis (21–23), MDMs are effector cells that are known to promote inflammation and fibrosis in multiple lung injury models (24–26). Because macrophages are necessary for maintenance of lung homeostasis, to resolve infection and promote repair, when TR-AMs are lost during injury and infection, it is essential to recruit monocytes into the tissue where they differentiate into MDMs, to different extents, in response to tissue-specific signals in their microenvironment (27–29). TR-AMs can be depleted by sufficient radiation doses, since they are present in the lung at the time of exposure. Highly differentiated MDM-derived alveolar macrophages (AMs) repopulate the depleted alveolar niche, becoming more similar to TR-AMs over time. However, these AMs remain distinct from TR-AMs by retaining a functional and transcriptional profile that is similar to the less differentiated macrophages from which they originate (30). As such, replacement of depleted TR-AMs by MDM-derived AMs alters macrophage responses to subsequent challenges, having long term consequences that shift the balance of inflammatory regulation within the lung (31, 32). Therefore, we hypothesized that IAV infection secondary to RT promotes the retention and pro-fibrotic activation of inflammatory MDMs.

Methods:

Detailed methods are provided in supplemental methods.

Animals.

C57BL/6J mice (female, 8 weeks of age) were obtained from Jackson Laboratory. All animal protocols conform to NIH guidelines and were approved by The University Committee on Animal Resources at the University of Rochester.

Irradiation and IAV Inoculation.

Mice were irradiated as previously described (17). Mice received whole thorax irradiation at doses of 5Gy, 10Gy, or 12.5Gy. Controls received identical handling without irradiation. IAV (PR8; strain A/Puerto Rico/8/1934 H1N1) was prepared and 0.01 HAU or sham (PBS only) administered as previously described (33). Body weight and survival was monitored until the end of the experiment. Mice losing >25% of their initial weight were euthanized.

Sample Collection, Magnetic Cell Sorting, and Flow Cytometry.

Lungs were collected 2 weeks or 26 weeks from the start of the experiment for assessment of acute or late effects, respectively. CD45+ immune cell enriched lung digests were generated as previously described (17, 32). For flow cytometric analysis cells were stained for CD11b (BD Pharmingen), Ly6G, CD206, Ly6C, CD45 (Biolegend), CD11c, F4/80 (eBioscience) and LIVE/DEAD cell stain (Life Technologies). Data were analyzed using FCS Express (De Novo Software).

Single Cell Sequencing.

Lung digests enriched for CD45+ cells were stained with anti-CD45 (Biolegend) and DAPI. Lung cells from n=3 mice per condition were pooled to equal n=1 sample per condition. CD45+, DAPI- cells were selected using a FACSAria cell sorter (BD Biosciences). Samples were processed and analyzed using 10X technology and URMC Genomics Core quality control metrics as described in detailed methods. Cell type identification was performed with ScType, contrasting the Immune system and Lung mouse packages. The dataset underwent further refinement to focus on macrophages, visualized using UMAP and clustered using the Louvain method. Data were deposited on Gene Expression Omnibus (GEO) under the accession number GSE270113.

Histology and Immunohistochemistry.

Lungs were inflation-fixed and paraffin embedded. For immunohistochemistry, antibody binding was performed using anti-Arginase 1 (Proteintech) and anti-CD64 (Invitrogen). For histology, slides were stained with Gomori’s Trichrome (Leica Biosystems).

Image Quantification.

For each lung, 10–12 100x fields of view (regions of interest; ROI’s) were imaged and blinded for each trichrome, arginase-1 and CD64 assay for use in quantification. Trichrome-stained lungs were analyzed by semiquantitative scoring for lung injury, as outlined in detailed supplemental methods. For quantification of Arginase-1, stained cells were enumerated using FIJI.

RT-qPCR.

Quantitative RT-PCR was performed on whole lung tissue or CD45+ immune cells. RNA was isolated using Trizol with phenol/chloroform separation, as previously described (33). cDNA was prepared using the iScript kit (BioRad). Quantitative RT-PCR was performed using PerfeCTa SYBR Green FastMix Low ROX (Quantabio). CT values were normalized to the average of three housekeeping genes (Gapdh, B2m, and Tbp). RT-qPCR primers are provided in detailed supplemental methods.

Statistical Analysis.

Data were analyzed with GraphPad Prism 10 (GraphPad Software, LLC). Two-way ANOVA with Tukey’s multiple comparisons test determined statistical significance at P<0.05. Data are presented as mean ± SEM.

Results:

Well-tolerated low doses of RT are sufficient to cause mortality and lung pathology when combined with a secondary IAV infection.

To Investigate how radiation-induced lung injury is affected by a subsequent IAV infection, pneumonitis and fibrosis prone C57BL/6 mice were exposed to whole thorax radiation then infected with IAV (Figure 1A). The radiation doses selected were based on well-studied outcomes for this strain when administered without subsequent IAV, as established in our prior studies and in the literature. (34–37). These responses ranged from those known to cause no observable pathology (5 Gy), to induce pneumonitis only (10 Gy), and to cause both pneumonitis and pulmonary fibrosis (12.5 Gy). As such, all the RT doses were sub-lethal when administered without IAV exposure, except for a small number of mice that acquired late developing radiation injury at the highest doses. We similarly selected a dose of IAV that was not lethal when administered alone without RT exposure. However, when RT was followed 1 week later with an IAV infection (RT + IAV exposure), we observed an RT dose-dependent increase in lethality. This suggests that even low doses of RT are sufficient to increase susceptibility to future insults. Body weight loss, an indicator of viral infection and clearance, was also assessed. Weight loss and recovery after IAV infection was similar between irradiated and non-irradiated mice, except that irradiated mice continued to lose weight for an additional 2–3 days before recovering (Supplementary Figure 1). All surviving IAV-treated mice recovered to their pre-infection weights and achieved weights similar to the non-infected control group by week 9.

Figure 1.

Combined exposure to RT + IAV decreases survival and intensifies lung injury compared to either RT or IAV alone. A) C57BL/6J, 8-week-old, female mice were exposed to either thoracic RT alone (5, 10, or 12.5 Gy), IAV alone (0.1HAU PR8), or 5 – 12.5 Gy RT followed one week later by inoculation with PR8 (RT + IAV), and survival was followed for 26 weeks. B) Experimental timeline of early and late RT + IAV exposures. Mice were exposed to either 5 Gy RT alone, IAV alone, or 5 Gy RT followed by IAV either 1 week later (early IAV) or 20 weeks later (late IAV) and harvested 26 weeks from the start of the experiment. C) Representative images of lungs from each condition showing regions of pathology where present. Semi-quantitative analysis of lung pathology by D) average pathology score per mouse and by E) the frequency of each score per condition. Bars represent mean ± SEM. Symbols (circles) represent each mouse. n=3–5 mice per condition. Significantly different comparisons are as indicated by brackets. *P<0.05, **P<0.01, ***P<0.001.

For assessment of lung pathology, the least toxic radiation dose, 5 Gy, was selected since it resulted in no mortality when administered alone and the least mortality when given as a combined RT + IAV exposure. Lungs were assessed 26 weeks after RT. This time point, which has been well characterized in our previous studies, allows sufficient time for the late developing radiation sequalae, pneumonitis and pulmonary fibrosis, to occur (32, 35, 36) and is long after viral clearance in IAV exposed mice (38, 39). Also, to determine whether the underlying effects of 5 Gy RT on the secondary IAV infection are transient or persistent, in addition to administering IAV 1 week after RT (early IAV), in a separate cohort IAV inoculation was delayed for 20 weeks following RT (late IAV) (Figure 1B).

When given alone, 5 Gy of RT did not cause overt lung pathology, as these lungs were indistinguishable from control lungs upon histological examination (Figure 1C). In contrast, major lung damage was observed after combined RT + early IAV exposure. Pathology included focal areas of congestion, immune cell accumulation and collagen deposition, reflective of a mixed pneumonitis/fibrotic outcome. Furthermore, this response was exaggerated compared to IAV alone. Quantification of lung injury supported these findings (Figure 1D). Importantly, a significant enhancement of lung injury was also observed after RT + late IAV, despite a 20-week period between the two insults, demonstrating that even benign doses of RT cause persisting and long-term susceptibility to a secondary injury. IAV only, in contrast, caused lung injury that resolved, since scores were only significantly elevated after late IAV alone (lungs were collected 6 weeks post IAV) but not after early IAV alone (lungs were collected 25 weeks post IAV). Because both RT and IAV cause heterogenous, patchy pathology, multiple regions of interest (ROI’s) spanning across each lung were assessed and the distribution of the ROI scores for each condition was evaluated (Figure 1E). This revealed that only exposure to RT + IAV produced areas of pathology that yielded the highest scores, reflecting the most severe damage.

Aberrantly activated macrophages are localized specifically to regions of pathology after RT + IAV.

Lung analysis of the pan-macrophage marker CD64 at the 26-week time point revealed that the patchy areas of immune cell congestion within areas of pathology were densely populated with accumulated macrophages (Figure 2A). Within these pathologic regions cells also positively stained for Arginase-1 (Arg-1), a marker for pro-fibrotic aberrantly activated macrophages (40, 41) (Figure 2B; corresponding images to panel A, via serial section). Quantification revealed that RT + IAV significantly increased the numbers of Arg-1 positive cells compared to IAV alone, and that this response occurred even when IAV inoculation was delayed for 20 weeks (Figure 2C). In contrast, Arg-1 positive cells were virtually absent outside of pathological regions, aside from a sparse number of peri-bronchial cells, even though macrophages were present (Figure 2D and Supplementary Figure 2). This was also the case for control and RT only conditions, which both lack regions of inflammation. This specificity provides further evidence of the involvement of aberrantly activated macrophages in late developing radiation-induced lung pathology.

Figure 2.

RT + IAV increases the presence of Arg-1 positive cells, localized to regions of lung pathology. Lungs were collected from mice 26 week following exposure to RT and IAV, either alone or in combination, where early IAV was administered 1 week after RT, and late IAV was administered 20 weeks after RT, or from non-exposed control mice. Immunohistological staining was conducted for A) the pan-macrophage marker CD64, and B) arginase-1 (Arg-1). C). Quantification of Arg-1 positive cells within regions of pathology. D). Quantification of Arg-1 positive cells in regions without pathology. Bars represent mean ± SEM. Symbols (circles) represent each mouse. n=3–5 mice per condition. Significantly different comparisons are as indicated by brackets. *P<0.05, **P<0.01, ***P<0.001. ROI’s were acquired at 100x.

RT causes long-term shifts in lung macrophage populations, where TR-AMs are reduced but MDMs are expanded.

To more comprehensively assess the macrophages that accumulated within the pathology, lung digests were collected at the 26-week time point and distinct macrophage subsets were identified and characterized by flow cytometry. In CD45+ myeloid cells, following exclusion of doublets, dead cells and Ly6G+ granulocytes, F4/80+ macrophages were identified. Resident and recruited macrophage populations were distinguished within this population based on expression of well described surface markers (42–44). This enabled identification of TR-AMs (CD11c+, CD11b^low) and two populations of MDMs: an AM differentiated population (Mo-AMs; CD11c+, CD11b+) and a non-AM differentiated population (Mo-MPs; CD11c-, CD11b+), which also includes interstitial macrophages (45) (Figure 3A). Both RT alone and RT + IAV significantly decreased the proportion of TR-AMs but significantly increased the proportion of Mo-AMs, as a percentage of the total number of macrophages, by 26 weeks (Figure 3 B,C). Additionally, RT + early IAV further increased the proportion of Mo-AMs compared to RT alone. Exposure to IAV alone, however, was not sufficient to alter the proportions of these populations at this time. Conversely, Mo-MPs proportions were not significantly altered by RT alone, and were instead decreased by IAV alone, regardless of when IAV was administered. RT + early IAV further decreased the proportion of this population, to a greater extent than RT + late IAV. These results suggest that at this time RT had a more outsized effect than IAV on the population dynamics of the macrophage subsets, since responses to RT + IAV were more similar to RT alone than to IAV alone.

Figure 3.

RT and IAV alter lung macrophage population size and phenotype. CD45+ immune cells isolated from lung digests, collected 26 week following exposure to RT and IAV, either alone or in combination, or from non-exposed control mice, were analyzed by flow cytometry. Early IAV was administered 1 week after RT, whereas late IAV was administered 20 weeks after RT. A) Following the exclusion of doublets, dead cells, and Ly6G+ cells, macrophages were identified by the positive expression of CD45, F4/80, and, for TR-AMs, autofluorescence. Distinct lung macrophage subsets were identified as tissue-resident alveolar macrophages (TR-AMs; CD11b^low, CD11c+), monocyte-derived alveolar macrophages (Mo-AMs; CD11b+, CD11c+), and monocyte-derived macrophages (Mo-MPs; CD11b+, CD11c-). B) Proportions of TR-AMs, Mo-AMs, and Mo-MPs within the total macrophage population. C) Percentages of TR-AMs, Mo-AMs, and Mo-MPs in each experimental condition. D) Percentages of CD206+ cells within TR-AMs, Mo-AMs, and Mo-MPs. E) Percentages of Ly6C+ cells within Mo-AMs and Mo-MPs. Bars represent mean ± SEM. Symbols (circles) represent each mouse. n=4 mice per condition. Significantly different comparisons are as indicated by brackets. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Phenotype markers were also assessed for each macrophage subset. CD206 is known to be expressed by TR-AMs, but is also a considered to be a pro-fibrotic marker during pathologic conditions (46). As expected, the vast majority of TR-AMs expressed CD206 (Figure 3D). This was not impacted by either RT or IAV exposures, other than a small but significant decrease after RT + late IAV, when compared to IAV alone. The majority of Mo-AMs also expressed CD206, and this was significantly increased after RT + early IAV. While Mo-MPs expressed CD206 at the lowest percentages, this was increased by RT + IAV, regardless of when IAV was administered. Interestingly, neither RT alone nor IAV alone was sufficient to significantly alter CD206 expression in any of the macrophage subsets. The pro-inflammatory marker Ly6C is expressed by monocytes and MDMs (47). In Mo-AMs, Ly6C expression was not significantly altered by RT alone but was increased by late IAV alone (Figure 3E). RT + IAV prevented this increase, as expression was significantly lower than IAV alone, and was not significantly different from the control condition. Mo-MPs showed a greater baseline expression of Ly6C, but in contrast to Mo-AMs, this was not significantly altered by RT or IAV, either alone or in combination. Taken together this demonstrates that RT + IAV caused long-term shifts in the expression of these markers such that MDMs (Mo-AMs and Mo-MPs) adopt a more pro-resolving/pro-fibrotic phenotype.

Acutely, RT and IAV each decrease TR-AMs and increase MDMs, and this is enhanced by combined RT + IAV exposure.

The acute responses of lung macrophages to RT and IAV precede and contribute to the development of the later occurring chronic pathology. Therefore, the initial responses of lung macrophages were assessed at the 1-week time point. Lung macrophage subsets were identified by flow cytometry (Figure 4A), as described for the 26-week results. TR-AMs were significantly decreased, and Mo-AMs were significantly increased by RT alone, IAV alone, and RT + IAV (Figure 4B,C). Furthermore, the TR-AM population was significantly smaller after RT + IAV when compared to either exposure alone. Mo-MPs were only significantly increased after RT + IAV exposure.

Figure 4.

RT and IAV acutely alter lung macrophage population size and phenotype. CD45+ immune cells isolated from lung digests, collected 1 week following exposure to RT + IAV (or 1 week post IAV alone, or 2 weeks post RT alone) were analyzed by flow cytometry. A) Macrophages were identified by the exclusion of doublets, dead cells, and Ly6G+ cells, and the positive expression of CD45, F4/80, and, for TR-AMs, autofluorescence. Distinct lung macrophage subsets were identified as tissue-resident alveolar macrophages (TR-AMs; CD11b^low, CD11c+), monocyte-derived alveolar macrophages (Mo-AMs; CD11b+, CD11c+), and monocyte-derived macrophages (Mo-MPs; CD11b+, CD11c-). B) Proportions of TR-AMs, Mo-AMs, and Mo-MPs within the total macrophage population. C) Percentages of TR-AMs, Mo-AMs, and Mo-MPs in each experimental condition. D) Percentages of CD206+ cells within TR-AMs Mo-AMs, and Mo-MPs. E) Percentages of Ly6C+ cells within Mo-AMs and Mo-MPs. Bars represent mean ± SEM. Symbols (circles) represent each mouse. n=4 mice per condition. Significantly different comparisons are as indicated by brackets. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

The phenotype of each macrophage subset was also analyzed. Both RT and RT + IAV decreased TR-AM CD206 expression, which didn’t occur from IAV alone (Figure 4D). In contrast, in MDMS (Mo-AMs and Mo-MPs) RT alone did not alter CD206 expression, and was instead increased by IAV, either alone or from RT + IAV. For Mo-MPs, this increase was significantly greater after RT + IAV than after IAV alone. The proportion of MDMs expressing Ly6C was increased by either IAV alone or RT + IAV (Figure 4E). In Mo-AMs this was significantly greater after RT + IAV when compared to IAV alone. A small percentage of TR-AMs also expressed Ly6C in conditions with IAV (Supplementary Figure 3 and supplemental text).

An inflammatory MDM population is present after IAV and is further expanded by RT + IAV.

Single cell RNA-sequencing was also performed on the immune cells isolated from the lung at the one-week time point. The major immune cell types typically contained within the lung were identified, including several distinct monocyte and macrophage populations, reflecting macrophage heterogeneity within the lung during quiescence and inflammation (Supplementary Figure 4). These monocyte/macrophage clusters were then subclustered for further analysis (Figure 5A). This consisted of a distinct alveolar macrophage population (C5), a macrophage population expressing genes related to iron metabolism (C8), a macrophage population expressing M2 related genes (C3), clusters of M1 and M2 activated monocytes (C4 and C2, respectively) and 2 clusters expressing markers of both monocytes and macrophages (C0 and C1). Differential expression of known markers for monocytes, immature macrophages, and differentiated macrophages were assessed for each cluster, and confirmed their designation (Figure 5B). To identify MDMs, coexpression of monocyte and macrophage markers was assessed (Figure 5C). Coexpression of the monocyte marker, Ccr2, with the pan-macrophage marker, Cd68, varied across the clusters. Alveolar Macrophages (C5) and M2 Macrophages (C3) expressed the macrophage marker Cd68 but not Ccr2. Likewise, the monocyte marker Ccr2 predominated in M1 and M2 Monocyte clusters (C4 and C2). However, these markers were coexpressed within the M1 Monocyte/Macrophage clusters (C0 and C1), suggesting a transition from monocytes to MDMs within these clusters. Further analysis of monocyte and macrophage signatures also confirmed that M1 Monocyte/Macrophage clusters shared expression of several macrophage and monocyte markers (Figure 5D). A trajectory between the most recently arrived infiltrating monocyte populations and more differentiated MDM populations was noted (Figure 5E). As such, markers associated with monocyte cell adhesion and homing to tissues such as F13a1 (Factor XIIIA) and Sell (CD62L/L-Selectin) were most highly expressed in the M1 and M2 Monocyte clusters, more so than the M1 Monocyte/Macrophage clusters. Markers related to differentiation of monocytes to macrophages, such as Prtn3 and Cd274 (PD-L1) were expressed by both monocytes and MDMs. In contrast, markers of resident alveolar macrophages, such as Marco and Car4 were only present in the Alveolar Macrophage cluster.

Figure 5.

RT and IAV alter the composition of macrophage populations in the lung. CD45+ cells were isolated from lung digests collected 1 week following exposure to RT + IAV (or 1 week post IAV alone, or 2 weeks post RT alone) and analyzed by single cell sequencing. Immune cell identification was performed, and populations identified as monocytes/macrophages were subclustered for further analysis. A) Seven unique subsets were identified. B) Expression of macrophage and monocyte markers within each subset. C) Coexpression of Cd68, a macrophage marker, with Ccr2, a monocyte marker, varied across the multiple subsets. D) Monocyte and macrophage marker expression corresponds with monocyte and macrophage designation, respectively, with mixed expression in the M1 Monocyte/Macrophage subset. E) Expression of markers of monocyte trafficking and differentiation do not overlap with markers of resident alveolar macrophages. F) The size and composition of the monocyte and macrophage subsets present varied between each condition. G) The proportion of cells in M1 Monocyte/Macrophage 0 or 1 subsets varied between IAV and RT + IAV. H) Gene coexpression of Itgax (CD11c) with Itgam (CD11b) in monocyte and macrophage clusters

The number and proportion of cells in each of these clusters varied between conditions (Figure 5F). The Alveolar Macrophage cluster (C5) was present in the control condition but was virtually absent in mice receiving RT and IAV, either alone or in combination, mirroring flow cytometry findings. The control condition also contained M2 Monocyte and M2 Macrophage clusters (C2 and C3, respectively), both of which persisted after RT exposure. While the M2 Macrophage cluster (C3) also remained following either IAV alone or RT + IAV, the M2 Monocyte cluster (C2) did not. The M1 Monocyte cluster (C4) and the M1 Monocyte/Macrophage clusters (C0 and C1) were only present in the IAV alone and RT + IAV conditions. The proportion of cells in M1 Monocyte/Macrophage 1 cluster (C1) was more expanded following RT + IAV compared to IAV alone (Figure 5G). Recapitulating flow cytometry findings, Itgax (CD11c) expressing Alveolar Macrophages (C5) were lost after RT alone, IAV alone, and RT + IAV, while Itgam (CD11b) expressing cells in the M1 Monocyte/Macrophage clusters (C0 and C1) were increased following IAV alone and RT + IAV (Figure 5H).

RT + IAV acutely alters monocyte and macrophage differentiation signatures, increases expression of pro-fibrotic factors and decreases antigen presentation genes.

To assess factors related to monocyte and macrophage differentiation and maturation, genes responsive to the colony stimulating factors (CSF) were investigated (48). Establishment of TR-AMs in the neonatal period as well as maintenance of mature terminally differentiated AMs into adulthood is known to rely on Csf2 expression (49–51). Whereas Csf1 expression is important broadly for differentiation of monocytes into macrophages and for maintaining IM populations (52) and also supports polarization to an M2 phenotype (53). We therefore queried the expression of gene signatures related to either Csf2 or Csf1-induced signaling, or lack of Csf2 (54), in each condition (Figure 6A). The Csf2 signature genes were prominent only in the control condition. In contrast, the Csf1 signature was enhanced following IAV alone and RT + IAV, and for many genes the greatest expression occurred in the RT + IAV condition. The CSF receptor genes Csf1r and Csf2ra were expressed by cells in each cluster present in each condition (Supplementary Figure 5), while expression of the ligands, Csf1 and Csf2, was low across all clusters in all conditions (data not shown).

Figure 6.

RT and IAV alter gene expression within macrophage populations. A) Expression of gene signatures related the CSF monocyte/macrophage maturation factors in each condition. B) Monocyte and macrophage subsets present in IAV alone and RT + IAV conditions. Expression of genes related to C) fibrotic processes and D) antigen presentation in subsets present after IAV alone or RT + IAV.

Since MDMs were present after IAV alone and after RT + IAV, we sought to understand how their responses differed between these conditions by assessing gene expression in the present clusters (C0, C1, C3 and C4) (Figure 6B). Interestingly, although M1 polarization is typically associated with pro-inflammatory phenotypes, genes related to fibrotic processes appeared most prominently in the M1 Monocyte/Macrophage 0 and 1 clusters in both conditions (Figure 6C). Furthermore, these genes were more highly expressed after RT + IAV compared to IAV alone, an effect also present in the M1 monocyte and M2 macrophage clusters. Genes for major histocompatibility complex 2 (MHC-II), for antigen presentation, were expressed by cells in clusters present following IAV alone, with greatest expression in C0 and C1 (Figure 6D). Across all the clusters, expression of these genes decreased following RT + IAV compared to IAV alone (see also, Supplementary Figure 6A). Pathway analysis was performed on the Alveolar Macrophage cluster (C5) since it was reduced by RT and IAV, both alone and in combination, (Supplementary Figure 6B). This revealed a reduction in several signaling processes related to the homeostatic functions of macrophages. These included the proteosome, antigen processing, Fc-gamma mediated phagocytosis, and endocytosis, as well as cytokine-cytokine receptor interaction and toll-like receptor signaling.

RT + IAV elevates pro-inflammatory factors in both the lung and in immune cells.

Gene expression of several factors involved in immune cell recruitment and acute pro-inflammatory responses were assayed in both whole lungs and in a portion of the isolated CD45+ immune cells that were also used for flow cytometry analysis at the 1-week time point. In whole lungs, both IAV alone and RT + IAV significantly increased each of the inflammatory factors analyzed (Figure 7A). This did not occur after RT alone, however. Additionally, RT + IAV exposure did not significantly further alter expression from IAV alone. In contrast to whole lung samples, in isolated immune cells, many of these factors were only elevated by RT + IAV exposure, but not by IAV alone, other than the chemokines Ccl2 and Cxcl2, which were increased to a lesser extent. Additionally, Il6 was increased after RT alone (Figure 7B).

Figure 7.

RT and IAV alter inflammatory factors in the lung and in lung immune cells. Relative mRNA expression of inflammatory chemokines and cytokines, assayed in A) whole lung digests and in B) CD45+ immune cells isolated from lung digests, collected 1 week following exposure to RT + IAV (1 week post IAV alone, 2 weeks post RT alone). Bars represent mean ± SD. Symbols (circles) represent each mouse. n=4 mice per condition. Significantly different comparisons are vs. control condition or as indicated by brackets. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Discussion:

While radiation damage to the lung is often latent, it is serious once it occurs. This is because it progresses over relatively long periods of time, increasing in severity, and can become irreversible (1, 2). This contrasts with the often acutely severe, but then resolving, pathogen response to a respiratory infection. Though it remains understudied clinically (9), our findings demonstrate that an IAV infection subsequent to thoracic RT intensifies the severity of the later developing radiation-induced lung injury. When combined with IAV, even the lowest dose of radiation (5 Gy) produced pneumonitis and fibrosis, characteristic of exposure to a much higher dose (i.e., 12.5 Gy) (17, 36, 55–57). Importantly, this occurred even when IAV exposure was delayed 20 weeks, a substantial portion of a mouse’s lifespan. This suggests that even RT doses considered benign (5 Gy) have underlying long-term effects that can exacerbate harm from a secondary injury, pushing the lung microenvironment past its threshold for recovery back to a quiescent state. This supports our previously published works using the IAV strain HKx31 following lung or total body irradiation and with IAV preceding irradiation (13, 14, 33).

Our findings suggest that the mortality from RT + IAV was not related to active infection but to impaired recovery from the lung injury. Lethality occurred 2–8 weeks following infection, a time frame that is post viral clearance. Additionally, recovery of body weight lost from IAV was similar between irradiated and non-irradiated mice. Our prior studies also found that virus clearance and virus-specific immune responses are not impaired in lung irradiated mice (14). This also supports that the observed pathology from RT + IAV is a radiation outcome. Because the clinical sequalae of radiation-induced lung injury are chronic and late developing, they become apparent by 26 weeks in our mouse models (17, 32, 55). This is long after viral clearance and epithelial cell recovery from infection (14, 38, 39, 58), as well as recovery of body weight lost from IAV infection.

A cellular mechanism for the enhanced response to RT + IAV has not been determined. Previously we established that radiation induced pulmonary fibrosis is dependent on the replacement of depleted lung-resident macrophages with recruited bone marrow originating macrophages (17). Additionally, we found that a maladaptive lung microenvironment develops within the irradiated field that supports the accumulation of aberrantly activated macrophages (59). The present study indicates that the pathology observed in the lungs of RT + IAV exposed mice contains a predominance of MDMs with a profibrotic phenotype. The specificity by which Arg-1 positive macrophages localized to the lung lesions of RT + IAV exposed mice reflects the pro-fibrotic processes occurring in this region and signifies the involvement of aberrantly activated macrophages in the worsened outcomes. Arg-1, a pro-fibrotic macrophage marker, is associated with fibrotic remodeling in both patients and mouse models (60–62). Indeed, others have found that interaction of Arg-1 expressing MDMs with fibroblasts drives pro-fibrotic processes (63). Furthermore, at this late time point, RT + IAV shifted macrophage dynamics towards a greater prevalence of Mo-AMs at the expense of the TR-AM population, and decreased pro-inflammatory Ly6C+ cells but increased CD206+ cells, signifying a pro-fibrotic phenotype (64–66). The observation that RT + IAV altered macrophage dynamics more similarly to RT alone than IAV alone further supports that the pathologic processes were RT-related and exacerbated by IAV rather than prolonged IAV responses.

Radiation-induced lung injury, and pulmonary fibrosis in general, is progressive. The response to an initiating injury, i.e. radiation, produces a series of downstream consequences, such as immune cell dysregulation, that culminate in the later developing pathology (67–69). Therefore, examining differences in the early responses of lung macrophages to RT and IAV exposure provides insight to the initiation of the disease progression. Acutely, RT + IAV exacerbated the depletion of TR-AMs and expansion of Mo-AMs that occurred from either insult alone. Furthermore, at this early time point RT appeared to intensify an IAV predominant M1-like proinflammatory response, which contrasts with the more RT-driven M2-like pro-fibrotic outcomes observed at the 26-week time point. As such, RT + IAV expanded the clusters with an M1 Monocyte/Macrophage designation to a greater extent than after IAV alone, mirroring the increased proportions of MDMs captured by flow cytometry. Also, the enhancement of the phenotypic markers CD206 and Ly6C in MDMs occurred in conditions with IAV, but not RT alone. Interestingly, CD206 decreased in TR-AMs in conditions with RT, possibly reflecting their direct exposure to radiation.

The appearance of the M1 Monocyte/Macrophage clusters after IAV and RT + IAV is likely reflective of MDMs, since it contained a transition in the expression of monocyte and macrophage markers. Furthermore, compared to IAV alone, RT + IAV appeared to shift the proportions of these recruited macrophages towards more immature subsets. The M1 Monocyte/Macrophage cluster 1, which possessed higher expression of monocyte markers and Itgam (CD11b) was more greatly expanded after RT + IAV. Also, loss of macrophage expression of the signature for the macrophage differentiation and survival factor GM-CSF (Csf2) likely reflects the loss of mature terminally differentiated TR-AMs from these insults, as this factor is required to maintain the lung-specific identity and functionality of AMs throughout adulthood (50, 70). Furthermore, the increased expression of the Csf1 signature suggests that these macrophages may instead be M-CSF (Csf1) driven at this time point. Modulation by M-CSF, as well as a lack of GM-CSF, is known to be associated with chronic inflammatory conditions in the lung as well as in pulmonary fibrosis (54, 71, 72). This is in line with our findings of enhanced expression of a profibrotic gene signature in the inflammatory monocyte and macrophage clusters with an M1 designation after RT + IAV. Although the M1 phenotype is typically associated with pro-inflammatory activation, this intensified expression of genes related to fibrotic processes could reflect a more dysregulated inflammatory response and the early development of the later emerging fibrotic state. MDM accumulation and activation is associated with pulmonary fibrosis in a growing body of work in both patients and preclinical models (73). Furthermore, CD45+ immune cells required both RT and IAV insults to increase expression of several genes in our panel of inflammatory factors, suggesting that pro-inflammatory activation of immune cells by IAV was enhanced in irradiated mice. Additionally, the reduced MHCII gene expression in the monocyte/macrophage clusters after RT + IAV suggests that RT may disrupt antigen presentation to T-cells, possibly affecting adaptive immunity (74).

The molecular mechanisms by which MDMs mediate fibrosis is multifactorial and involve their polarization to an M2-like phenotype. These cells produce pro-fibrotic factors including TGF-β and matrix metalloproteinases that promote the activation, differentiation and survival of fibroblasts. Our findings reveal that potentiation of radiation induced lung injury by IAV is accompanied by a more outsized MDM response in the lung. Specifically, the accumulation of pro-fibrotic macrophages during late occurring pathology is preceded by a greater loss of tissue-resident lung macrophages and expansion of recruited inflammatory MDMs in the acute phase. Their upregulated pro-fibrotic gene signatures would further enhance the development of a fibrotic tissue microenvironment. Future studies using MDM-depleted or MDM-tagged transgenic mice to perform lineage tracing of this population throughout the progression of the RT response will be valuable for further defining the precise mechanism mediating the enhanced pathology. The identification of novel mechanisms through which a secondary insult, such as a respiratory infection, potentiates and exacerbates radiation induced lung injury will open new clinical approaches for more effective treatment.