Male A/J mice, 4–5 weeks of age, were purchased from Jackson Laboratories (Bar Harbor, ME) and housed in an Association for Assessment and Accreditation of Laboratory Animal Care-Accredited, specific pathogen-free, environmentally controlled facility. All mice were free of endogenous viral pathogens, parasites, mycoplasmas, Helicobacter, and CAR bacillus. Mice were housed in groups of five in ventilated cages and provided high-efficiency particulate filtered air under a controlled light cycle (12 h light/12 h dark) at a standard temperature (22–24 °C) and 30–70% relative humidity. Animals were acclimated to the animal facility for 1 week and allowed access to a conventional diet (6% irradiated NIH-31 Diet, Envigo RMS, Inc., Madison, WI) and tap water ad libitum. All procedures were performed using protocols approved by the National Institute for Occupational Safety and Health (NIOSH) Institutional Animal Care and Use Committee.

A/J mice, 120 in total, were organized into four groups using a block design for randomization (Fig. 1). On day 1, mice aged 5–6 weeks were intraperitoneally (IP) injected with the chemical initiator, 3-methylcholanthrene (MCA) (Sigma, St. Louis, MO) dissolved in com oil (CO) (Sigma, St. Louis, MO) at a dose of 10 μg/g of body weight or CO alone. MCA was chosen as the initiating agent based on the efficient response of the A/J mouse to this carcinogen in our oropharyngeal aspiration study (Zeidler-Erdely et al. 2013). One week post-initiation, mice were exposed in whole-body inhalation chambers with individual steel mesh cages to aerosols generated during GMA-SS welding or air for 4 h/d and 4 d/w for 9 weeks at a target concentration of 40 mg/m³ (actual 32.3 ± 2.8 mg/m³). The welding wire used was 0.045 in. diameter Lincoln Electric Blue Max MIG 308LSI, and the welding parameters were set to 25 V DC, 300 in. per minute wire feed, 30 L/min of 95% Argon-5% CO₂ shielding gas, and a typical welding current of 220 amps.

The design and construction of the welding fume aerosol generator were previously described (Antonini et al. 2006). This automated robotic welder continuously generated welding fumes by welding beads onto ¼ in. thick plates of mild steel. The resulting fume was carried into a whole-body exposure chamber through a ¾ in. flexible tube by maintaining the chamber at a negative pressure (0.70 in. H₂0). Particle concentrations within the exposure chamber were continuously monitored with a Data RAM (DR-40000 Thermo Electron Co, Franklin, MA), and gravimetric determinations (37 mm cassettes with 0.45 μ m pore-size Teflon filters) were used to calibrate and verify the Data RAM readings each day. Gas generation, including carbon monoxide (CO), carbon dioxide (CO₂), oxygen (O₂), and ozone (O₃), was continuously monitored. During the welding exposure, O₂ levels were maintained above the OSHA minimal acceptable level. O₃, CO, and CO₂ were below OSHA permissible exposure limits and NIOSH recommended exposure limits (REL) during the entire exposure duration. In the exposure chamber, CO and O₃ levels were not significantly higher than background. The exposure system setup was slightly modified from that described previously (Antonini et al. 2006) to reduce the travel time of the particulate fume from the welding torch to the exposure chamber. This was done to ensure delivery of fresher fumes to more closely mimic a worker’s inhalation exposure.

All mice were weight-matched prior to the inhalation exposure. Mice were weighed weekly throughout the experimental time course and at the 30-week sacrifice.

Male A/J mice were exposed by inhalation to GMA-SS welding aerosols (40 mg/m³) (n = 25) or filtered air (n = 10) for 4 h. Immediately following exposure, whole lungs were excised, trimmed, and lyophilized. The freeze-dried tissue was weighed then acid digested. Inductively coupled argon plasma, atomic emission spectroscopy at NIOSH-Division of Applied Research and Technology (Cincinnati, OH) was used to determine the amount of Cr, Ni, Cu, Fe, and Mn present in the lung according to the draft NIOSH method 8200 modified to accommodate the sample matrix (NIOSH 2003).

At 30 weeks post-exposure, A/J mice were euthanized with Fatal Plus (100–300 mg/kg IP; 390 mg/ml pentobarbital sodium) (Vortech Pharmaceuticals, Dearborn, MI), weighed, and then the vena cava was cut to exsanguinate the animal. All internal organs were examined for the presence of tumors. The whole lung was then excised. All tumors were clearly defined with no apparent merged tumors. The lungs were inflated and fixed with 10% neutral buffered formalin for 24 h. Tumors were counted and measured 24 h after fixation. Lungs were embedded in paraffin, and then a 5-μm standardized section was cut. Slides were stained with hematoxylin and eosin and interpreted by two separate, contracted, board-certified veterinary pathologists in a blinded fashion for evidence of hyperplasia and neoplasia, inflammation, lymphoid tissue response, and foreign materials by light microscopy. Diagnostic criteria for hyperplastic and neoplastic findings were according to goRENI (http://www.goreni.org/), the standard reference for nomenclature and diagnostic criteria in toxicologic pathology and at the same time the Internet discussion platform for the global initiative “INHAND”—the International Harmonization of Nomenclature and Diagnostic criteria (Keenan et al. 2015; Renne et al. 2009). If abnormal changes were found, severity was scored using the following scale: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked. The final severity score reflects the average of the right and left lung lobe scores and are presented as mean ± standard error. Because bronchiolo-alveolar hyperplasia (BAH) and bronchiolo-alveolar adenomas (BAA) represent a continuum of the proliferative process and there is possible overlap between these diagnoses, the numbers of lesions were combined to compare the tumorigenic potential of each treatment (Renne et al. 2009). However, the gross tumor count at necropsy is more representative of the response because examination of a single histological section per lung underestimates the total number of lesions per lung (Rehm and Ward 1989).

Statistical analyses were performed using JMP version 12. Factorial analysis of variance (ANOVA) was utilized on continuous variables to make comparisons between the treatment groups. For some variables, a log transformation was performed on the data to reduce heterogeneous variance and meet the assumptions of an ANOVA. Gross tumor counts and histopathology counts from sections were analyzed similarly. Tumor incidence (presence or absence of tumors) was analyzed using a Chi-square test in SAS ‘Proc Freq,’ while tumor multiplicity (number of tumors per lung) was analyzed using Poisson regression in SAS ‘Proc Genmod’. In cases where overdispersion existed, a negative binomial regression was performed. Analyses were performed independently on CO and MCA treated animals, and only utilized data from those animals surviving to the 30-week time point. For all analyses, a p value of <0.05 was set as the criteria for significance.

A summary of the characteristics of GMA-SS welding fume and images of the fume are presented in Fig. 2. Particle mass size distribution was measured for this newer configuration using a Micro-Orifice Uniform Deposit Impactor (MOUDI, model 110; MSP corp., Shoreview, Minn.) with additional Nano-MOUDI stages (MSP model 115). The mass median aerodynamic diameter (MMAD) was 350 nm. The count-based particle size distributions of the particles were measured using a Scanning Mobility Particle Sizer (SMPS model 3936, TSI Inc, Shoreview, MN). The SMPS estimated the count median aerodynamic diameter to be 230 nm. Particle imaging was achieved with a scanning electron microscope (SEM; JEOL 6400, JEOL Inc). Inspection of physical characteristics of the particles showed many small nano-spheres (10–50 nm) linked together into long chain-like structures often with several branches. Elemental analysis of the welding particles was previously carried out (Zeidler-Erdely et al. 2013). The overall particulate fume was composed of 57% Fe, 20.2% Cr, 13.8% Mn, 8.8% Ni, 0.2% copper (Cu) and trace amounts of silicon (Si), aluminum (Al), and vanadium (V). Hexavalent chromium (Cr(VI)) levels were also measured and determined to be 2929 ppm. GMA-SS welding fume is largely water insoluble with a soluble/insoluble ratio of 0.006 (Antonini et al. 1999).

Table 1 shows the lung metal deposition in A/J mice after 4 h of inhalation of GMA-SS welding fume. The most abundant metal measured in the lung was Fe followed by Cr, Mn, Ni, and Cu, as predicted from the previous characterization of this fume (Antonini et al. 2006). The metal analysis by wt% of the whole lungs (60% Fe, 17% Cr, 14% Mn, 8% Ni, and 0.8% Cu) agrees with the wt% of collected fume.

The analysis of the metals shows a cumulative increase of ~10.1 μg of total GMA-SS fume deposited in the lung from a single 4-h exposure. The alveolar deposition in the mice was equated to the human by the equations below using the previous threshold limit value-time weighted average (TLV-TWA) of 5 mg/m³ for total welding fume and the PEL of 5 μg/m³ for Cr(VI). Previously, we estimated that 70% of the total dose reached the alveolar space (10.1 μg/d × 0.70 = 7.07 μg/d) (Erdely et al. 2011a; Raabe et al. 1988). The mice were exposed for 36 days (9 w at 4 d/w) for an approximate total alveolar deposition of 254.5 μg.

A timeline of the experimental protocol for the two-stage (initiation-promotion) carcinogenesis model is shown in Fig. 1. Initial body weights at week 0 [mean ± standard error (SE)] were 18.56 ± 0.35, 19.1 ± 0.39, 18.25 ± 0.31, and 19.05 ± 0.36 for the CO/air, CO/GMA-SS, MCA/air, and MCA/GMA-SS groups, respectively. Body weights increased steadily and were not changed due to exposure from week 0 to 30 with 10.26 ± 0.58, 9.46 ± 0.56, 11.38 ± 0.65, 9.78 ± 0.61 for the CO/air, CO/GMA-SS, MCA/air, and MCA/GMA-SS groups, respectively. Morbidity and mortality throughout the study were low (~ 5%), and no abnormalities, such as other tumor types besides lung, were found at the terminal sacrifice at 30 weeks. In total, six mice died during the course of the study and were not included in the final analysis of the data. Necropsy determined that all six mice died from morbidities that included enlarged heart or cause of death otherwise undetermined.

GMA-SS welding fume significantly promoted lung tumors in the A/J mouse 30 weeks after initiation with MCA. The grossly observed tumor multiplicity (average tumor number/mouse lung ± SE) for all groups is shown in Fig. 3. There was no effect of welding fume alone on tumor multiplicity (CO/air, 0.32 ± 0.10; CO/GMA-SS, 0.45 ± 0.13; p = 0.44). In animals initiated with MCA, tumor multiplicity was 7.93 ± 0.82 and 16.11 ± 1.18 for air and GMA-SS, respectively (p < 0.0001). Average tumor incidence (% of tumor-bearing mice) was 29% in CO/air and 38% in CO/GMA-SS-exposed animals. Reports in the literature indicate the background tumor frequency in A/J mice between 43 and 53 weeks to be 31–40% (Curtin et al. 2004; Groch et al. 1997). The mice in this study were 35 or 36 weeks old upon sacrifice, indicating the observed tumor incidence is consistent with the literature. As expected, tumor incidence was >96% in all MCA-initiated groups (n = 29 for MCA/air and n = 28 for MCA/GMA-SS groups) which confirmed successful experimental administration as well as its carcinogenic effectiveness in A/J mice. Total and average tumor number per treatment group across each of the individual lung regions is described in Table 2. MCA/GMA-SS-exposed mice had significantly greater lung tumor multiplicity in every lung region compared to MCA/air (p < 0.009). There was no difference between CO/air and CO/GMA-SS groups (p = 0.44).

Gross lung morphology from a GMA-SS-exposed mouse initiated with MCA is shown in Fig. 4. Welding fume deposition was visible in exposed mouse lungs and appeared black-brown in color. Tumors appeared white in color and opaque on initial gross exam and became more well-defined after fixation which aided enumeration. At 30 weeks, tumors were between −0.5 mm and ~3 mm, with most tumors ~1 mm.

CO/air animals had no lymphoid infiltrate or foreign material (brown-black pigment, i.e., welding fume) and hyperplasia was unremarkable (0.09 ± 0.05). There were 5 total lesions reported in the CO/air group. The CO/GMA-SS-exposed group had minimal, but significant, lymphoid infiltrates (0.22 ± 0.063; p < 0.03) and foreign material (1.86 ± 0.07; p < 0.03) compared to CO/air. There were 2 total hyperplastic lesions reported in this group. Welding fume, indicated by black-brown foreign material presence in the lungs, was found in both GMA-SS-exposed groups (CO and MCA) in every lung section. Significant hyperplasia (1.20 ± 0.11; p < 0.03) and increased total preneoplastic/neoplastic lesions (BAH and BAA) were reported in the MCA/GMA-SS animals compared to MCA/air animals (114 versus 70; p < 0.03). Histopathological assessment of the lungs from a separate, second board-certified veterinary pathologist confirmed these findings (Table 3). The total number of proliferative lesions for MCA/GMA-SS-exposed mice was 153 compared to 90 for MCA/air. Increased perivascular mononuclear infiltrate in the MCA/GMA-SS group consisting of small aggregates or cuffs of lymphocytes, sometimes mixed with a few plasma cells, adjacent to or around multiple scattered vessels was also reported. This infiltrate was not observed in any of the MCA/air lungs. Because there were no significant findings in the CO groups besides evidence of welding fume and minimal lymphoid infiltrates in GMA-SS-exposed mice, a second evaluation was not deemed necessary. Figure 5a demonstrates a BAA and two areas of BAH. Figure 5b shows the adenoma in panel a at 20× magnification adjacent to areas of welding fume deposition. Panels c and d demonstrate a BAA at 40× and BAH at 20×, respectively. The BAA were more compact nodules composed of enlarged type II cells that obscured the normal tissue morphology. These benign neoplasms had a smooth margin that frequently caused compression of surrounding tissue. The cells sometimes had prominent nucleoli and frequently formed radiating proliferations around blood vessels. BAH were round to irregular in shape with increased numbers of Type II cells lining alveolar septae. The cells were of normal size and normal tissue architecture was retained. The margins of these lesions were irregular and did not cause compression of surrounding tissues. In addition, the number of adenomas in the MCA/GMA-SS group was significantly greater than in the MCA/air group (p < 0.05; Table 3) (McConnell et al. 1986).