The two macrocyclic trichothecene-producing strains of S. chartarum used in this study were IBT 9460 (strain A) and IBT 7711 (strain B). Strain A produced sixfold higher concentrations of verrucarol and fragmented to a greater extent (<0.5–2 μm in aerodynamic diameter) compared with strain B (Figure 1). Fungal test articles were aerosolized using a computer-controlled acoustical generator system as previously documented (28). See Data Supplement E1 for further details.

Female B6C3F1/N mice, 5–6 weeks old, were acquired from the National Toxicology Program mouse colony housed at Taconic Biosciences, Inc. Mice were housed in the Assessment and Accreditation of Laboratory Animal Care International–accredited animal facility at the National Institute for Occupational Safety and Health (NIOSH) and were provided an NTP-2000 diet (Harlan Laboratories) and water ad libitum. The mice appeared healthy throughout the study (Figure E2 in the data supplement). All animal procedures were performed under a NIOSH Animal Care and Use Committee–approved protocol.

Mice were randomly separated into six exposure groups with 15 mice per group: Groups 1 and 2 were S. chartarum strain A, viable or heat-inactivated conidia (HIC); groups 3 and 4 were S. chartarum strain B, viable or HIC; and groups 5 and 6 were control groups of high-efficiency particulate absolute–filtered air only. The S. chartarum–exposed mice inhaled 1 × 10⁴ particles (estimated pulmonary deposition/exposure), which was the highest aerosolized dose that reproducibly induced a lymphoproliferative response. Mice were exposed twice per week for 4 or 13 weeks as previously documented (28). Samples were collected 24 hours and 48 hours after final exposure to assess temporal changes in immune responses. Mice were killed via intraperitoneal injection of 200 mg/kg sodium pentobarbital solution (Fatal-Plus Solution; Vortech Pharmaceuticals, Ltd.) and exsanguinated via cardiac puncture. The exposure scheme and harvesting time points were chosen on the basis of preliminary immune response data (27, 29).

Left lung lobes (n = 3/group) were inflated with fixative and collected. The tissue was then embedded, sectioned, and stained with hematoxylin and eosin for routine histopathological evaluation as previously described (29). For additional details, see Data Supplement E2.

Cells collected from BAL (n = 7/group) were enumerated using a Cellometer Vision (Nexcelom Bioscience) and prepared for flow cytometric analysis as previously described (29). Cells were blocked and stained using fluorochrome-conjugated antibodies (Table E1). See Data Supplement E3 for additional details.

Right lung lobes (n = 3/group) were sent to MS Bioworks for quantitative proteomic analysis as previously described (29). Total RNA was extracted for microRNA (miRNA) (Exiqon Services) and mRNA analyses as previously described (28). Additional analyzed genes are reported in Table E2. T-helper cell type 1 (Th1), Th2, and Th17 gene expression data were normalized to gusb (β-glucuronidase) and additional genes to gapdh (glyceraldehyde 3-phosphate dehydrogenase).

Proteomic and RNA data were analyzed by using Ingenuity Pathway Analysis (IPA) software (Qiagen) for core, comparative, and miRNA target filter analyses (28). For additional details, see Data Supplement E4.

Statistical power tests were conducted to determine the appropriate number of samples needed for each endpoint measurement. Fold changes greater than or equal to 2 or less than or equal to −2 were considered altered. P ⩽ 0.05 was considered statistically significant (28, 29). See Data Supplement E5 for additional details.

Inflammation and arterial remodeling were observed in murine lungs after repeated exposure to viable S. chartarum (Figure 2). Differences in lesion development between the two strains were most evident microscopically after 4 weeks (Table 1). No microscopic changes were observed in lungs exposed to HIC or high-efficiency particulate absolute–filtered air controls (Figure E2).

At 4 weeks, strain A exposure produced a robust peribronchiolar and perivascular inflammatory response characterized by eosinophilic and neutrophilic infiltrates centered at terminal bronchioles with extension into adjacent alveoli (Figure 2A). Similar inflammatory cells were also observed surrounding smaller pulmonary arteries and within the vessel wall. Remodeling of pulmonary arteries near the terminal bronchiole was evident after 4 weeks of exposure to strain A (Figure 2A). Additional histological changes observed after 4-week strain A exposure included alveolar histiocytosis, bronchiolar epithelial cell hyperplasia with mucous metaplasia, and vascular thrombi (Table 1), whereas strain B exposure resulted in minimal airway inflammation (Figure 2B).

After 13 weeks of exposure, histopathology was nearly indistinguishable between the two strains. Bronchiolitis and pulmonary arterial remodeling were observed in the lungs; however, there was a shift in inflammatory cell composition from eosinophils and neutrophils at 4 weeks to lymphocytes and macrophages at 13 weeks (Figures 2A and 2B), indicating the mice were further along in the pulmonary arterial remodeling process.

A summary of the histopathology showing a continuum of microscopic changes ranging from arteritis to pulmonary arterial remodeling is presented in Figure 2C. Early changes in pulmonary arteries were characterized by inflammation, whereas later changes in vessels were more severe and characterized by arterial remodeling. Remodeling in affected vessels was characterized by symmetrical thickening of the artery wall and narrowing of the lumen, affecting the small to medium-sized vessels at or below the level of the terminal bronchiole. With increasing exposure duration, there was a shift in inflammatory cell types from polymorphonuclear cells to mononuclear cells.

The BALF was composed primarily of eosinophils, neutrophils, and macrophages (Figure 3). Compared with air-only controls, strain A exposure showed increased eosinophils and neutrophils 24 hours after 4 and 13 weeks (Figures 3A and 3B). Alveolar macrophages were significantly increased 48 hours after 13-week strain A exposure (Figure 3C). Only after 4-week strain A HIC exposure were eosinophils, neutrophils, and alveolar macrophages initially increased (Figures 4A–4C).

Strain B–exposed mice showed significantly increased eosinophils and alveolar macrophages 48 hours after 4-week exposure (Figures 3D and 3F). After 13-week exposure, eosinophils, neutrophils, and alveolar macrophages were significantly increased (Figures 3D–3F). Strain B HIC exposure did not result in significant cell population changes (Figures 4D–4F).

Proteomic analysis of whole-lung homogenate detected 3,465 proteins in total (Table E3). Table 2 shows differentially expressed proteins involved in tissue remodeling and inflammation. Chil3 and chil4 (chitinase-like proteins 3 and 4, respectively) were upregulated after 4-week and 13-week strain A exposures, but only after 13-week strain B exposure. Only chil3 was decreased after 4-week strain A HIC exposure. Lgals3 (galectin-3) was upregulated after 4-week and 13-week strain A exposures. After 4-week strain B exposure, lgals3 was decreased, but it was increased after 13 weeks. In contrast, lgals3 was downregulated after 4-week strains A and B HIC exposure. Von Willebrand factor was upregulated after 13-week strain A exposure but downregulated after strain B exposure. Strains A and B HIC exposure resulted in a downregulation of von Willebrand factor after 4-week exposure. Strain A–exposed mice showed increased sftpa (pulmonary surfactant protein A) after 4-week and 13-week exposures. In contrast, both sftpa and sftpd were decreased after 4-week strain B exposure and after 4-week strains A and B HIC exposure.

Inflammation-associated proteins retnla (resistin-like α) and mrc1 (macrophage mannose receptor 1) were significantly increased after 4-week and 13-week strain A exposures. After 13-week strain B exposure, retnla was upregulated. Although mrc1 was downregulated at 4 weeks, expression was slightly increased after 13-week strain B exposure. Retnla was unaltered in mice exposed to either HIC strain; however, mrc1 was decreased after 4-week strains A and B HIC exposure. Located in the eosinophil matrix, ear1 and ear2 (eosinophil cationic proteins 1 and 2 receptors, respectively) were significantly increased after 4-week strain A exposure. After 13 weeks, both ear1 and ear2 were increased over 100-fold at 48 hours. With strain B, ear1 and ear2 were only upregulated after 13-week exposure, but not to the same degree as with strain A. For HIC exposures, ear1 and ear2 were decreased after 4-week strain A exposure, and ear1 was decreased after 13-week strain B exposure.

Carbohydrate-binding clec proteins (C-type lectins) are nonspecific pathogen recognition receptors. After strain A exposure, clec7a and clec10a were increased after 4 weeks, and after 13 weeks, clec10a and clec3b were increased. Strain B–exposed mice showed increased clec7a after 4 weeks and increased clec10a after 13 weeks. Unaltered after strain A HIC exposure, clec7a was decreased after 4-week strain B HIC exposure. Clec3b was also decreased after 4-week strains A and B HIC exposure.

Core proteomic analysis conducted using IPA identified associations between altered proteins and canonical pathways, diseases and disorders, and physiological system development and functions, as well as the top cardiotoxicity pathways. These top relationships, mechanisms, functions, and pathways relevant to a dataset are determined by having the highest number of proteins that overlap those included in the specific pathway, disease, or function. As reported in Tables 3 and 4, one of the top canonical pathways identified after exposure to strain A is eIF2 (eukaryotic initiation factor 2) signaling, a component of protein synthesis. After strain A exposure, organismal injury and abnormalities were identified as the top diseases and disorders at all time points, followed by the inflammatory response and neurological disease. The top physiological system developments and functions involving the proteins altered after strain A exposure were the hematological system development and function, as well as immune cell trafficking and tissue morphology. The top cardiotoxicity pathways included cardiac enlargement, dysfunction, and fibrosis.

The core proteomic analysis identified eIF2 signaling as well as mitochondrial dysfunction and cytidine diphosphate (CDP)-diacylglycerol biosynthesis I as the top canonical pathways associated with the proteins altered after exposure to strain B (Tables 3 and 4). CDP-diacylglycerol biosynthesis results in the synthesis of CDP-diacylglycerol, which is an intermediate product in the synthesis of cardiolipin, a phospholipid primarily located in the inner mitochondrial membrane. After 13 weeks of exposure to strain B, G_βγ signaling, a contributor to heart failure and inflammation, was one of the top canonical pathways identified. In addition, organismal injury and abnormalities and inflammatory response were among the top diseases and disorders associated with the proteins altered after exposure to strain B.

The RT² Profiler plates (Qiagen) measured genes involved in the Th1, Th2, or Th17 response pathways. Heat maps illustrate fold changes between viable exposed groups versus the air-only control group for both time points after 4-week and 13-week exposures (Figures 5 and 6). The majority of differentially expressed genes associated with Th1/Th2 immune responses were upregulated after strains A and B exposure (Figure 5), although not to the same extent. Many Th17-associated genes were downregulated after strain A versus strain B exposure (Figure 6). Overall, the Th1/Th2 and Th17 gene expression profiles were altered to a greater extent in lungs from strain A–exposed mice than in those from strain B–exposed mice.

Genes involved in tissue remodeling and the inflammatory response were increased after 4-week and 13-week strain A exposures, such as retnla, chil3, arg1 (arginase 1), and mrc1 (Table 5). These genes were upregulated after 13-week strain B exposure, but not to the extent of strain A. HIC exposure did not alter retnla and chil3 expression by either strain; however, arg1 and mrc1 were slightly upregulated after 4-week strain A HIC exposure. Monocyte chemoattractant protein 1, or ccl2, was also significantly increased after 4-week and 13-week strain A exposures (Figure 6), but it was only significantly increased after 13-week strain B exposure.

Strain A increased il13 at both time points after 4 weeks, together with il4 and il6 at 48 hours. After 13-week strain A exposure, ifng (IFN-γ), il2, and il13 expression was increased at both time points, il4 and il6 expression was increased at 48 hours (Figures 5 and 6). In contrast, strain B–exposed mice showed upregulated il6 expression, together with significantly increased il13 expression, after 4 weeks. After 13 weeks, ifng expression was slightly increased, together with significantly upregulated il13 and il5 at 24 hours and increased il6 at 48 hours (Figures 5 and 6).

Involved with eosinophil recruitment, il33 expression was initially upregulated after 4-week and 13-week strain A exposures compared with only after 13-week strain B exposure (Table 5). Ccl7 (monocyte chemotactic protein 3) and ccr3 were upregulated after 4-week and 13-week strain A exposures (Figure 5). Strain B exposure resulted in increased ccl7 expression 48 hours after 4-week exposure and 24 hours after 13-week exposure (Figure 5). Inflammatory mediator cxcl5 (Chemokine [C-X-C motif] ligand 5) and cxcl1 expression was significantly increased after 4-week and 13-week strain A exposures (Figure 6). In strain B–exposed mice, cxcl1 and cxcl5 were upregulated 48 hours after 4 weeks, but not significantly. After 13 weeks, cxcl1 was significantly upregulated (Figure 6).

miRNA profiling analysis identified 531 (strain A) and 512 (strain B) miRNAs in the lung samples (Table E4). After 4-week strain A exposure, 116 (24 h) and 138 (48 h) miRNAs were significantly altered compared with air-only controls. After 13-week strain A exposure, 15 (24 h) and 49 (48 h) miRNAs were significantly differentially expressed, the majority of which were downregulated (Table E4). In contrast, no miRNAs were altered after 4-week strain B exposure, and only 14 (24 h) and 35 (48 h) miRNAs were differentially expressed after 13 weeks.

Using IPA, the miRNA target filter predicts miRNAs and target genes included in this study. A regulator of il13 and il33, miR-339–5p was decreased after strain A and strain A HIC exposures (Table 6). Downregulation of miR-365 after 4-week strain A viable and HIC exposure was predicted to regulate il6. Expression of il10 was predicted to increase due to decreased miR-144–5p or miR-374b-5p expression after strain A viable and HIC exposure. After strain B viable and HIC exposure, the expression of these miRNAs was unchanged (Table 6). Expression of Th1 cytokines, such as il2 and tnf, may be regulated by miR-181b-5p (il2), miR-21a-5p (tnf), and miR-669d-5p (tnf). miR-181b-5b was decreased after 4-week strain A exposure and at 4-week and 13-week strain A HIC exposure. miR-21a-5p was upregulated only after 13-week strain A exposure, but it was downregulated 24 hours after 13-week strain A HIC exposure. After strain A viable and HIC exposure, miR-669d-5p expression was increased. After strain B viable and HIC exposure, the expression of these miRNAs was unchanged. Critical to eosinophilic recruitment, miR-133a-3p and miR-335–5p were downregulated after strain A viable and HIC exposure, but unchanged after strain B exposure. miR-23b-3p, a regulator of arg1 and mrc1, was downregulated after strain A exposure, but not after HIC exposure. In contrast to downregulation after 13-week strain B exposure, miR-23b-3p was unchanged after HIC exposure.