Genetic susceptibility to interstitial pulmonary fibrosis in mice induced by vanadium pentoxide (V₂O₅)

Abstract

Interstitial lung diseases (ILDs) are characterized by injury, inflammation, and scarring of alveoli, leading to impaired function. The etiology of idiopathic forms of ILD is not understood, making them particularly difficult to study due to the lack of appropriate animal models. Consequently, few effective therapies have emerged. We developed an inbred mouse model of ILD using vanadium pentoxide (V₂O₅), the most common form of a transition metal found in cigarette smoke, fuel ash, mineral ores, and steel alloys. Pulmonary responses to V₂O₅, including dose-dependent increases in lung permeability, inflammation, collagen content, and dysfunction, were significantly greater in DBA/2J mice compared to C57BL/6J mice. Inflammatory and fibrotic responses persisted for 4 mo in DBA/2J mice, while limited responses in C57BL/6J mice resolved. We investigated the genetic basis for differential responses through genetic mapping of V₂O₅-induced lung collagen content in BXD recombinant inbred (RI) strains and identified significant linkage on chromosome 4 with candidate genes that associate with V₂O₅-induced collagen content across the RI strains. Results suggest that V₂O₅ may induce pulmonary fibrosis through mechanisms distinct from those in other models of pulmonary fibrosis. These findings should further advance our understanding of mechanisms involved in ILD and thereby aid in identification of new therapeutic targets.—Walters, D. M., White, K. M., Patel, U., Davis, M. J., Veluci-Marlow, R. M., Bhupanapadu Sunkesula, S. R., Bonner, J. C., Martin, J. R., Gladwell, W., Kleeberger, S. R. Genetic susceptibility to interstitial pulmonary fibrosis in mice induced by vanadium pentoxide (V₂O₅).

Interstitial lung diseases (ILDs) are characterized by parenchymal injury, inflammation, and scarring of the alveoli, leading to impaired lung function. Many ILDs result in interstitial fibrosis for which there are no effective therapies, but prognosis varies depending on cause and pathological pattern. ILDs include more than 200 lung disorders, caused by a variety of stimuli including occupational and environmental exposures (1). However, a number of ILDs, including the prototypical idiopathic pulmonary fibrosis (IPF), have no known cause, making them difficult to model.

Whether of known or unknown etiology, pulmonary fibrosis is a complex trait thought to involve interactions between genetic and environmental factors. Gene-environment interactions in pulmonary fibrosis are supported by observed susceptibility differences in individuals with similar exposures, familial cases, and differential susceptibility among inbred strains of mice exposed to fibrogenic agents such as bleomycin (2), silica (3, 4), and asbestos (5). While extensive research using animal exposures to these known fibrogenic agents has increased understanding of pathology associated with specific exposures, effective therapies to halt the fibrotic process in idiopathic interstitial pneumonias (IIPs) have not been found and may be partially due to unknown environmental factors. Bleomycin has been employed to test numerous antifibrotic agents, yet none of these compounds have been shown to improve patient survival (6, 7). Therefore, alternative models that reflect human exposures associated with IIPs may provide a more relevant context in which to investigate fibrogenic mechanisms and potential therapies.

A number of ILDs are associated with metal exposures (8, 9), and epidemiological studies have consistently found associations between IPF and exposure to metal dusts [odds ratio (OR) 2.44] and cigarette smoking (OR 1.58) (10). In fact, exposure to metal dusts in a Japanese population was very strongly associated with IPF (OR 9.55; ref. 11). Although specific components of metal dusts or cigarette smoke that lead to ILD have not been identified, transition metals are hypothesized to be likely candidates due to their potential to induce cellular damage through production of reactive oxygen species(ROS). In addition, chronic exposure to high levels of vanadium compounds is associated with pneumonitis (12).

Vanadium poses an inhalation risk at high concentrations in occupational settings (13) but is a ubiquitous transition metal found in air pollution particles and cigarettes (14). Vanadium pentoxide (V₂O₅) is the most common chemical form of vanadium and is used extensively as an alloy in steel for strength at high temperatures and as an industrial catalyst (15). In animals, V₂O₅ exposure induces a variety of pulmonary responses, including inflammation, fibroblast proliferation, interstitial and/or peribronchiolar fibrosis, and epithelial cell hyperplasia, depending on the species and strain of animal (16–18). In humans, acute occupational exposure to vanadium has been predominantly associated with decrements in lung function indicative of obstructive disease and bronchitis-like symptoms (19–23). Although a direct link between vanadium exposure and ILD has not been established in human populations, animal studies and suggestive evidence from human occupational exposures imply that vanadium exposure may contribute to some cases of ILD in susceptible individuals.

The current study describes a novel model of fibrosing alveolitis using V₂O₅ to investigate the gene-environment interactions in pulmonary fibrogenesis. Here we describe V₂O₅-induced lung injury and inflammation with interstitial fibrogenesis in a strain-dependent manner. Furthermore, the strain-dependent differences in susceptibility suggest a genetic contribution to V₂O₅-induced pulmonary fibrosis, which was analyzed using a set of recombinant inbred strains. The pathophysiological characteristics and environmental relevance of this model should be valuable to identify mechanisms involved in fibrogenesis that have not been previously identified in other models.

MATERIALS AND METHODS

Animals

Male C57BL/6J (B6), DBA/2J (D2), and 49 BXD recombinant inbred (RI) strains of mice (5–7 wk) were purchased from Jackson Laboratories (Bar Harbor, ME, USA) and housed in an environmentally controlled animal facility with a 12:12 light:dark cycle and provided food and water ad libitum throughout the experiments. Mice were acclimated for at least 5 d prior to study. All animal procedures were approved by the Institutional Animal Care and Use Committees of East Carolina University and the U.S. National Institute of Environmental Health Sciences.

V₂O₅ exposure

Mice were lightly anesthetized with isoflurane gas, and doses of V₂O₅ particles (Sigma-Aldrich, Milwaukee, WI, USA) suspended in PBS or PBS alone were administered by oropharyngeal aspiration (24) of ∼50 μl on study d 0 and 7. For the dose-response study, 0, 0.1, 0.5, 1, 2, or 4 mg/kg V₂O₅ was aspirated, and mice were euthanized on study d 14 (n=6–8 mice/dose). For kinetic studies, 4 mg/kg V₂O₅ was aspirated and mice were euthanized on study d 1, 2, 4, 6, 8, 10, 12, 14, 21, 28, 56, 84, and 112 (n=4–8/group). BXD RI strains were aspirated with 4 mg/kg V₂O₅ or PBS on study d 0 and 7 and euthanized on study d 14 (n=4–8/group).

Lung function measurements

Baseline lung function measurements were made after V₂O₅ or vehicle using a forced oscillation technique on the flexiVent system (Scireq, Montreal, QC, Canada). Animals were anesthetized (∼400 mg/kg tribromoethanol, i.p.), tracheotomized, and ventilated with room air at 10 ml/kg and 150 breaths/min with positive end expiratory pressure of 3 cmH₂O. The animals were then paralyzed with pancuronium bromide (0.8 mg/kg, i.p.), and respiratory mechanics were measured using a script to ensure consistent timing of perturbations relative to total lung capacity maneuvers to standardize volume history. Measures of airway newtonian resistance (R_N) and tissue elastance (H) were obtained using the constant phase model, and static compliance (C_st) was measured using ramped pressure inflation and deflation.

Bronchoalveolar lavage

Immediately after lung function measurements, mice were exsanguinated, the left bronchus was clamped, and the right lung was lavaged with 4 successive aliquots of 26.25 ml/kg cold HBSS. Recovered bronchoalveolar lavage fluid (BALF) was centrifuged at 500 g for 10 min at 4°C. Supernatant from the first BALF aliquot was used for measurement of total protein. All cells from the same individual were pooled, and total cell counts were made with a hemacytometer. Slides were prepared by cytocentrifugation of 20,000 cells at 600 rpm for 5 min (Shandon Cytospin III; Thermo Fisher Scientific, Waltham, MA, USA) and stained (Richard-Allan, Kalamazoo, MI, USA). BALF cell differential counts were performed on 300 cells/slide using standard morphological criteria.

Quantitation of protein in BALF

Total protein concentration in BALF (an index of lung permeability and lung injury) was measured using a Bradford protein assay (Bio-Rad, Hercules, CA, USA) according to the manufacturer's recommendations and read at 595 nm on a spectrophotometer (Beckman-Coulter, Brea, CA, USA). Concentration was calculated from a standard curve.

Measurement of lung collagen

Lavaged right lung lobes were homogenized in RIPA lysis buffer plus protease inhibitor cocktail and PMSF (Sigma-Aldrich). Lung homogenates were centrifuged at 10,000 g for 10 min at 4°C. Soluble collagen was measured by adding 500 μl Sircol Dye Reagent (Biocolor, Carrickfergus, UK) to 50 μl of lung homogenate supernatant. Samples and standards were agitated for 30 min, then centrifuged at 10,000 g for 10 min to pellet collagen-bound dye. Supernatants were discarded, and pellets were dissolved in 1 ml 0.5 M NaOH. Absorbance was read at 540 nm, and sample concentrations calculated from a standard curve.

Histology

Left lungs were inflated with 10% neutral buffered formalin (Azer Scientific, Morgantown, PA, USA) for 24–72 h before cutting into 3 sections and undergoing standard processing. Cross sections (5 μm) were placed in xylene to remove parrafin, hydrated through an ethanol gradient, and stained with Masson's Trichrome to allow visualization of collagen.

PCR fibrosis array

Total lung RNA was isolated from the lower right lung lobe of 3 mice/group euthanized 10 and 21 d after V₂O₅ (4 mg/kg) or PBS using a RNeasy Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's recommendations. Total RNA (1.5 μg) was reverse-transcribed to cDNA using SABioscience's RT² First Strand Kit (Qiagen). Mouse Fibrosis PCR Arrays (Qiagen) were run on an Applied Biosystems StepOnePlus Real-Time PCR System (ABI, Foster City, CA, USA) following kit instructions. Data were analyzed using SABiosciences' web-based PCR array analysis tool (http://pcrdataanalysis.sabiosciences.com/pcr/arrayanalysis.php).

Quantitative real-time (R-T) PCR

The same total lung RNA isolated for the PCR-array was reverse transcribed using a QuantiTect reverse transcription kit (Qiagen). Several genes found to be differentially expressed, including Ccl11 [chemokine (C-C motif) ligand 11], Il1b (interleukin 1β), Lox (lysyl oxidase), and Mmp3 (matrix metallopeptidase 3) as well as Gapdh (internal control) were verified in B6 and D2 mice by quantitative R-T PCR using QuantiTect primer assays and SYBR green master mix (Qiagen). In addition, mRNA expression for genes found in the significant chromosome 4 quantitative trait locus [QTL; Cdk5rap2 (CDK5 regulatory subunit associated protein 2), Tle1 (transducin-like enhancer of split 1), Frmd3 (FERM domain containing 3), Kdm4c (lysine [K]-specific demethylase 4C), and Ptprd (protein tyrosine phosphatase, receptor type D)] was measured in lungs from B6 and D2 mice after vehicle and V₂O₅ treatment. R-T PCR was used to obtain cycle threshold (C_t) values for target and internal control cDNA levels. Data are expressed as the average ± se ΔC_t of 3 animals/group, where ΔC_t is defined as C_{t target} − C_{t internal control}.

Transforming growth factor β (TGF-β) ELISA

TGF-β protein levels were measured in the same lung homogenates as collagen content using the Human/Mouse TGF-β ELISA Ready-SET-Go! kit (eBioscience; Affymetrix, San Diego, CA, USA) according to the manufacturer's instructions. The assay is specific for activated TGF-β; however, further acid activation to measure total TGF-β levels produced readings below the level of detection suggesting that the detergent in the RIPA lysis buffer had already caused dissociation of the latency-associated peptide. Therefore, data represent total TGF-β protein levels.

Genetic linkage analyses

The mean PBS response for each BXD RI strain was subtracted from individual V₂O₅-induced responses, and the difference was used for V₂O₅-induced cell and collagen phenotypes for genetic mapping. Correlations between cell and collagen phenotypes were evaluated using the correlation matrix function in GeneNetwork (http://www.genenetwork.org/webqtl/main.py). GeneNetwork was also used to identify QTLs for V₂O₅-induced increases in lung collagen content, total BALF cells, and cell differentials. Genome-wide scans for QTLs were done with the entire RI data set and parental phenotypes. More than 7600 informative markers and ∼3800 markers with unique strain distribution patterns exist for the BXD RI set. Details of analysis procedures were published previously (25).

Statistical analyses

All data are presented as means ± se for each group. Two-way ANOVA was used to evaluate dose and strain effects in dose-response studies, while treatment (V₂O₅ or PBS) and strain effects were evaluated at each time point for the kinetics and BXD RI studies. Bonferroni post hoc testing was used to determine statistical differences between groups, and P < 0.05 was accepted as statistically significant. SigmaStat 3.0 software (SPSS Science, Chicago, IL, USA) was used for all statistical analyses.

RESULTS

Dose response

Dose-dependent increases in mean BALF protein were found in B6 and D2 mice 7 d after the second of 2 V₂O₅ doses (Fig. 1A). In both strains, protein concentrations were significantly elevated compared to PBS controls after 0.5–4.0 mg/kg V₂O₅. However, the response in D2 mice was significantly greater than in B6 mice after 2 and 4 mg/kg V₂O₅.

Figure 1.

Responses to increasing doses of V₂O₅ in B6 and D2 mice for protein content in BALF (A), total number of cells recovered in BALF (B), and lung collagen content (C). Bars represent group means ± se; n = 7–8 mice/group. *P < 0.05 vs. 0 mg/kg V₂O₅ within strain; ⁺P < 0.05 vs. B6 within dose; n = 6–8 mice/group.

Similarly, mean numbers of BALF cells increased dose-dependently in both strains; however, significantly more cells were found in D2 mice compared to B6 mice at all V₂O₅ doses (Fig. 1B). Macrophages accounted for ∼90% of BALF cells (range 88.7% – 94.3%) across all groups (Table 1); however, numbers of macrophages were significantly higher in D2 mice at all V₂O₅ doses. Numbers of lymphocytes were also significantly elevated in D2 compared to B6 mice for the highest 3 V₂O₅ doses. No change from control collagen content at any V₂O₅ dose was found in B6 mice, whereas V₂O₅ induced a significant increase in lung collagen at 1, 2, and 4 mg/kg V₂O₅ in D2 mice (Fig. 1C).

Table 1.

Differential cell counts from BALF 14 d after increasing doses of V₂O₅

Strain	V2O5 (mg/kg)	Macrophages	Epithelial cells	Neutrophils	Eosinophils	Lymphocytes	Monocytes
B6	0.0	56,905 ± 2900	3842 ± 321	56 ± 36	0 ± 0	535 ± 81	92 ± 62
D2	0.0	53,433 ± 3112	2978 ± 500	24 ± 24	120 ± 120	589 ± 245	44 ± 29
B6	0.1	65,419 ± 3760	4887 ± 696	212 ± 88	213 ± 93*	950 ± 277	105 ± 70
D2	0.1	80,680 ± 5109*,#	4201 ± 352	206 ± 146	387 ± 152	1330 ± 431	383 ± 103#
B6	0.5	88,599 ± 7747*	5354 ± 915	723 ± 148	398 ± 43*	1268 ± 305	88 ± 57
D2	0.5	127,278 ± 5685*,#	7130 ± 1009*	475 ± 246	840 ± 254*	3568 ± 823*	1022 ± 299#
B6	1.0	107,313 ± 7735*	7024 ± 605*	1048 ± 232	489 ± 136*	1433 ± 252	193 ± 142
D2	1.0	213,687 ± 12929*,#	9657 ± 1769*	568 ± 219	2405 ± 1030*	6238 ± 1842*,#	1730 ± 452*,#
B6	2.0	160,721 ± 8994*	9636 ± 1170*	1568 ± 696*	1477 ± 511*	3314 ± 361	855 ± 321
D2	2.0	342,627 ± 28302*,#	11,626 ± 1120*	1877 ± 1357	7667 ± 2282*	15,601 ± 3655*,#	2744 ± 672*
B6	4.0	214,854 ± 13167*	9339 ± 290*	1219 ± 682	798 ± 283*	11,110 ± 1553*	1967 ± 375*
D2	4.0	494,420 ± 45762*,#	16,158 ± 1673*,#	4007 ± 858*,#	9495 ± 3362*,#	29,884 ± 3665*,#	3535 ± 742*

^*P < 0.05 vs. 0 mg/kg V₂O₅ within strain

^#P < 0.05 vs. B6 within dose.

Dose-dependent increases in airway R_N and tissue H were also found in D2 mice but did not change significantly in B6 mice (Fig. 2A, B). Quasi-static C_st was significantly lower compared to PBS control at the 2 highest doses of V₂O₅ in D2 mice. Significantly lower C_st at baseline and at all doses of V₂O₅ were found in D2 mice compared to B6 mice (Fig. 2C).

Figure 2.

Changes in lung function with increasing doses of V₂O₅ in B6 and D2 mice. A) Newtonian resistance, R_N, B) Tissue elastance, H. C) Static compliance, C_st. Bars represent group means ± se; n = 6–8 mice/group. *P < 0.05 vs. control (0 mg/kg V₂O₅) within strain; ⁺P < 0.05 vs. B6 within dose; n = 6–8 mice/group.

Overall, V₂O₅-treated B6 lung sections were similar to vehicle controls, although at the highest dose of V₂O₅, mild peribronchiolar/perivascular inflammation was found in 2 of 7 lungs (Fig. 3). In D2 mice, patchy areas of alveolar inflammation with increased collagen staining appeared after 0.5 mg/kg, and the area and severity of inflammation and alveolar collagen deposition increased with dose. Subpleural regions appeared to be most affected, and at the highest doses, alveolar epithelial type II cell hyperplasia and infrequent fibrotic nodules were found.

Figure 3.

V₂O₅-induced changes in lung histology in B6 and D2 mice. V₂O₅ had little effect on B6 mice at all doses except 4 mg/kg, where minimal peribronchiolar/perivascular inflammation was found (arrow). In contrast, exposure to 0.5 mg/kg V₂O₅ caused focal inflammation, alveolar wall thickening, and collagen deposition in D2 mice and became more extensive with increasing dose, especially in subpleural regions (arrows). Infrequent fibrotic nodules surrounded by hypertrophic epithelium were seen in D2 mice after 4 mg/kg V₂O₅ (arrowhead). Masson's Trichrome staining of 5 μm paraffin sections. Original magnification: ×100; insets, ×400.

Time course

A small but significant increase in BALF protein was found in B6 mice until 6 d after V₂O₅ (Fig. 4A), and a second V₂O₅ aspiration did not increase BALF protein above controls. Compared to B6 mice, significantly greater BALF protein was found in D2 mice, peaking 2–3 d after the first and second V₂O₅ doses. Protein concentrations remained significantly elevated over controls to 21 d in D2 mice.

Figure 4.

Time course of responses to 4 mg/kg V₂O₅ in B6 and D2 mice. V₂O₅-induced increases in BALF protein levels (A), total number of cells recovered in BALF (B), and lung collagen content (C) to a much greater extent in D2 mice than in B6 mice. Solid rectangles on x axis indicate the days when V₂O₅ was aspirated. *P < 0.05 vs. PBS control within strain; ⁺P < 0.05 vs. B6 mice within same treatment group; n = 4–8 mice/group.

Total numbers of BALF cells increased similarly in B6 and D2 mice 1 d after V₂O₅ (Fig. 4B), and they remained near this level until d 21 in B6 mice; cell numbers continued to rise in D2 and peaked 4–5 d after V₂O₅. Unlike BALF protein, cell numbers were enhanced after a second exposure and did not return to controls at 112 d after V₂O₅ in D2 mice. The majority of BALF cells were macrophages at all times except d 1 (Table 2). The greatest numbers of neutrophils were found 1–2 d after V₂O₅ and remained significantly elevated above controls through d 14 in both strains, although peaks were significantly higher in D2 mice. Numbers of eosinophils were greatest 2 d after V₂O₅ in both strains of mice, and the increase was significant in B6 mice only on d 6. The greatest numbers of lymphocytes were found 5–6 d after V₂O₅ in D2 mice and remained significantly elevated above controls until d 21; numbers in B6 mice were significantly increased only on d 10. Numbers of epithelial cells and monocytes were variable in both strains, and no consistent V₂O₅ or strain effects were found.

Table 2.

Time course of differential cell counts from bronchoalveolar lavage fluid after 4 mg/kg V₂O₅

Strain	Treatment	Day	Macrophages	Epithelial cells	Neutrophils	Eosinophils	Lymphocytes	Monocytes
B6	PBS	1	74,568 ± 6335	2851 ± 923	269 ± 127	175 ± 65	323 ± 180	29 ± 29
	V2O5	1	136,850 ± 17,242*	2591 ± 606	17,795 ± 7363*	3490 ± 1519	796 ± 392	354 ± 122
D2	PBS	1	53,608 ± 8521	1227 ± 438	639 ± 318	232 ± 178	328 ± 263	216 ± 129
	V2O5	1	48,573 ± 9115#	3375 ± 789	87,191 ± 18,880*,#	7806 ± 2917*	1120 ± 409*	4435 ± 2543
B6	PBS	2	94,982 ± 6775	704 ± 235	458 ± 299	290 ± 186	632 ± 442	114 ± 73
	V2O5	2	144,811 ± 20,196	1252 ± 407	26,553 ± 6536*	4434 ± 2681	632 ± 195	391 ± 204
D2	PBS	2	47,055 ± 4249	2175 ± 413	2088 ± 1994	79 ± 52	675 ± 388	116 ± 81
	V2O5	2	120,126 ± 36,488*	4131 ± 2064#	98,993 ± 19,045*,#	7451 ± 1647*	2381 ± 817*,#	3793 ± 1118*,#
B6	PBS	4	87,693 ± 10,628	600 ± 155	43 ± 28	108 ± 60	626 ± 200	94 ± 94
	V2O5	4	157,155 ± 14,987	622 ± 358	9938 ± 2522*	2056 ± 1049	2211 ± 1306	517 ± 292
D2	PBS	4	42,701 ± 1274	1879 ± 483	147 ± 67	0 ± 0	225 ± 63	47 ± 47
	V2O5	4	323,328 ± 19,661*,#	3422 ± 1489	54,751 ± 22,372*,#	5163 ± 1395*,#	3334 ± 964*	2085 ± 883
B6	PBS	6	87,914 ± 10,786	704 ± 286	190 ± 97	110 ± 79	342 ± 165	205 ± 100
	V2O5	6	161,864 ± 11,882*	447 ± 220	3575 ± 1459*	1185 ± 375*	1552 ± 347	1377 ± 1068*
D2	PBS	6	55,915 ± 9388	3205 ± 1117#	21 ± 21	32 ± 32	425 ± 222	90± 74
	V2O5	6	148,895 ± 18,494*	2567 ± 653	3826 ± 1102*	1108 ± 174*,#	4317 ± 1110*,#	1430 ± 276*
B6	PBS	8	76,569 ± 8710	1100 ± 402	0 ± 0	106 ± 106	297 ± 184	261 ± 31
	V2O5	8	174,992 ± 14,137*	1458 ± 301	25,297 ± 12,328*	3167 ± 890	1000 ± 355	2419 ± 1001*
D2	PBS	8	64,423 ± 7863	2654 ± 653	685 ± 427	125 ± 77	238 ± 238	0 ± 0
	V2O5	8	208,504 ± 31,004*	2257 ± 1073	82,521 ± 30,037*,#	5211 ± 3282	3429 ± 1244*,#	2007 ± 865
B6	PBS	10	73,567 ± 6493	838 ± 289	60 ± 60	0 ± 0	89 ± 62	89 ± 89
	V2O5	10	154,389 ± 17,799*	1500 ± 655	26,187 ± 9583*	1574 ± 771	2283 ± 538*	2505 ± 838*
D2	PBS	10	48,852 ± 6635	6939 ± 2551#	285 ± 172	63 ± 42	111 ± 61	0 ± 0
	V2O5	10	278,422 ± 41,807*,#	12,121 ± 10,148*,#	38,392 ± 12,957*	5413 ± 1771*	*3016 ± 1108	763 ± 420#
B6	PBS	12	64,322 ± 7381	519 ± 261	578 ± 578	0 ± 0	1242 ± 576	839 ± 477
	V2O5	12	160,214 ± 11,386	329 ± 221	7497 ± 6491*	467 ± 467	1211 ± 315	699 ± 319
D2	PBS	12	47,679 ± 7410	2319 ± 794	492 ± 389	46 ± 46	46 ± 46	44 ± 44
	V2O5	12	463,124 ± 81,903*,#	2460 ± 972#	23,143 ± 11,122*,#	4302 ± 1360*,#	7199 ± 2540*,#	2987 ± 1111
B6	PBS	14	71,730 ± 9969	539 ± 178	21 ± 21	0 ± 0	350 ± 103	217 ± 115
	V2O5	14	177,479 ± 17,122*	1072 ± 522	2270 ± 1326	1440 ± 883	1740 ± 429	1625 ± 785
D2	PBS	14	47,969 ± 7509	4347 ± 1282#	502 ± 180	128 ± 91	698 ± 599	419 ± 405
	V2O5	14	357,179 ± 58,672*,#	19,975 ± 11,651	16,721 ± 14,213*,#	4451 ± 1638*,#	2819 ± 1474*,#	1668 ± 1028
B6	PBS	21	75,254 ± 8887	381 ± 151	54 ± 54	0 ± 0	713 ± 330	385 ± 149
	V2O5	21	139,441 ± 16,420*	1269 ± 506	189 ± 71	328 ± 199	2217 ± 459	932 ± 189
D2	PBS	21	45,451 ± 2588#	1679 ± 614	81 ± 43	35 ± 35	284 ± 85	281 ± 96
	V2O5	21	187,834 ± 12,863*,#	2599 ± 798	75 ± 75	421 ± 234	3672 ± 1463*,#	1024 ± 309
B6	PBS	28	41,733 ± 9489	1250 ± 270	21 ± 21	0 ± 0	121 ± 95	0 ± 0
	V2O5	28	69,504 ± 13,042	1679 ± 641	0 ± 0	0 ± 0	256 ± 162	435 ± 225
D2	PBS	28	55,419 ± 8060	1173 ± 235	0 ± 0	0 ± 0	223 ± 153	60 ± 60
	V2O5	28	139,519 ± 32,980*,#	5847 ± 1564*,#	0 ± 0	0 ± 0	2472 ± 2472	494 ± 494
B6	PBS	42	38,427 ± 2234	488 ± 488	94 ± 63	133 ± 94	233 ± 233	0 ± 0
	V2O5	42	38,696 ± 3080	2688 ± 1930	94 ± 55	413 ± 413	381 ± 249	229 ± 144
D2	PBS	42	48,808 ± 4006	2977 ± 1163	44 ± 44	46 ± 46	0 ± 0	0 ± 0
	V2O5	42	69,867 ± 14,870#	2867 ± 816	146 ± 94	50 ± 50	48 ± 48	148 ± 94
B6	PBS	56	46,079 ± 5268	6777 ± 4116	83 ± 49	29 ± 29	110 ± 66	46 ± 46
	V2O5	56	46,969 ± 8680	7825 ± 4610	0 ± 0	460 ± 460	335 ± 211	35 ± 35
D2	PBS	56	36,596 ± 3878	3496 ± 925	31 ± 31	0 ± 0	308 ± 120	194 ± 117
	V2O5	56	101,039 ± 16,545*,#	6528 ± 1271	0 ± 0	0 ± 0	200 ± 200	567 ± 186*,#
B6	PBS	84	57,538 ± 3895	6233 ± 168	338 ± 84	96 ± 55	296 ± 51	0 ± 0
	V2O5	84	56,108 ± 9472	5238 ± 1155	256 ± 60	140 ± 55	133 ± 133	0 ± 0
D2	PBS	84	57,629 ± 3505	5929 ± 214	96 ± 55	0 ± 0	96 ± 55	0 ± 0
	V2O5	84	87,571 ± 6865*,#	15,254 ± 3958*,#	1150 ± 197*,#	617 ± 98*,#	408 ± 307	0 ± 0
B6	PBS	112	48,846 ± 1123	5596 ± 200	96 ± 55	375 ± 115	88 ± 51	0 ± 0
	V2O5	112	44,906 ± 7175	7104 ± 747	173 ± 78	431 ± 192	510 ± 210	0 ± 0
D2	PBS	112	52,733 ± 1578	8125 ± 188	308 ± 53	517 ± 48	817 ± 221	0 ± 0
	V2O5	112	73,142 ± 7285*,#	17,126 ± 2099*,#	943 ± 184*,#	858 ± 347	431 ± 250	0 ± 0

^*P < 0.05 vs. 0 mg/kg V₂O₅ within strain;

^#P < 0.05 vs. B6 within dose.

V₂O₅ induced significant increases in lung collagen content, and responses were enhanced after the second exposure in D2 mice and remained elevated above controls until the end of the study (112 d), whereas collagen content in B6 mice was not significantly increased after V₂O₅ except at 6 d (Fig. 4C). Collagen content in D2 mice was greater than V₂O₅-treated B6 mice throughout the study.

Lungs from PBS-exposed mice of both strains were histologically normal at all times (Fig. 5). Mild peribronchiolar/perivascular inflammation was most prominent 1–2 d after V₂O₅ in B6 mice, and small focal areas of alveolar inflammation and atelectasis were found in <10% of the sections until 28 d, after which all V₂O₅-induced abnormalities resolved. Much greater parenchymal effects were found in lungs from V₂O₅-treated D2 mice, including intrapulmonary hemorrhage, vascular congestion, edema, and mixed infiltration of inflammatory cells 1 d after V₂O₅. By d 4, predominantly mononuclear organizing pneumonitis was found, which started to resolve by d 6. However, the second V₂O₅ aspiration caused heterogeneous lesions with thickened alveolar septae, early interstitial fibrosis, and type II alveolar epithelial hypertrophy/hyperplasia, as well as mixed granulocytic/mononuclear inflammation, hemorrhage, and alveolar consolidation. Much of the affected area was subpleural, with some central foci interspersed with normal tissue. Most granulocytic inflammation had resolved after 21 d, but large numbers of mononuclear cells remained in the alveoli with increased collagen staining in these areas. Despite resolution of granulocytic inflammation, gross morphological changes were visible with constriction of the lung surface from 28 to 112 d. Increased cellularity, alveolar consolidation, and fibrosis were present to 84 d. Although inflammation and fibrotic tissue had resolved by 112 d, collapsed alveoli remained.

Figure 5.

Time course of V₂O₅-induced lung histology in B6 and D2 mice. V₂O₅ -induced mild inflammatory foci and alveolar septal thickening, which completely resolved by 56 d in B6 mice. In D2 mice, V₂O₅ induced early lung injury with alveolar hemorrhage and edema followed by inflammation, septal thickening, and interstitial fibrosis, as well as epithelial type II metaplasia. Structurally remodeled alveoli persisted to 4 mo after V₂O₅. Masson's Trichrome staining of 5 μm paraffin sections. Original magnification: ×100 and ×400.

mRNA expression of many conventional fibrosis-associated genes was reduced in D2 mice compared to B6 mice 10 d after V₂O₅, including Ccr2 [chemokine (C-C motif) receptor 2], Eng (endoglin), Il1b, Itga1 (integrin α 1), Jun, Lox, Mmp3, Mmp8, Mmp9, Smad6 (SMAD family 6), and Smad7 (Table 3). Relative to B6 mice, greater expression of profibrotic genes, such as Ccl11 and Ctgf (connective tissue growth factor), as well as Fasl (Fas ligand), Il1a, and Serpina1a [serine (or cysteine) peptidase inhibitor, clade A, member 1A], but less expression of Mmp3, Plau (plasminogen activator, urokinase), and Tgif1, was found in D2 mice 21 d after V₂O₅. V₂O₅-induced changes in gene expression detected in the PCR array were verified by RT-PCR for Ccl11, Il1b, Lox, and Mmp3 (Supplemental Fig. S1). Because Tgfb1 mRNA levels were also lower in D2 mice compared to B6 mice, we measured protein levels of this important profibrotic mediator and found that protein levels were also significantly lower in D2 mice compared to B6 mice. While a treatment effect was not found in either strain, a temporal effect was found in B6 mice with increased levels at 21 d (Fig. 6).

Table 3.

Fold up- or down-regulation of V₂O₅-induced fibrosis-related genes in DBA/2J (D2) compared to C57BL/6J (B6) mice

Gene	Description	d 10	d 21
Acta2	Actin, α 2, smooth muscle	−2.84	1.52
Bcl2	B cell leukemia/lymphoma 2	−3.47	1.49*
Bmp7	Bone morphogenetic protein 7	−2.25	1.19
Cav1	Caveolin 1, caveolae protein	−2.77	1.89*
Ccl11	Chemokine (C-C motif) ligand 11 (eotaxin)	−1.41	3.75*
Ccl12	Chemokine (C-C motif) ligand 12	−2.29	1.49
Ccl3	Chemokine (C-C motif) ligand 3 (Mip1-α)	−2.90	−1.33
Ccr2	Chemokine (C-C motif) receptor 2 (Mcp1 receptor)	−4.35*	−1.32
Col1a2	Collagen, type I, α 2	1.78	1.90*
Ctgf	Connective tissue growth factor	1.12	2.38*
Cxcr4	Chemokine (C-X-C motif) receptor 4	−2.77	1.50*
Dcn	Decorin	−2.26	−1.35
Edn1	Endothelin 1	−3.47	1.88*
Egf	Epidermal growth factor	−2.27	1.19
Eng	Endoglin	−3.49*	1.49*
Fasl	Fas ligand (TNF superfamily, member 6)	−3.74	2.94*
Grem1	Gremlin 1	−2.88	1.15
Hgf	Hepatocyte growth factor	−2.19	−1.33
Il13ra2	Interleukin 13 receptor α 2	−2.25	−1.11
Il1a	Interleukin 1 α	−4.62*	2.36*
Il1b	Interleukin 1 β	−11.06*	1.19
Il4	Interleukin 4	−2.21	1.15
Il5	Interleukin 5	−2.19	−1.09
Inhbe	Inhibin β E	−3.55	−1.69
Itga1	Integrin α 1	−2.86*	−1.35
Itga2	Integrin α 2	−2.91	−1.34
Itgb3	Integrin β 3	−2.22	1.20
Itgb5	Integrin β 5	−3.47	−1.68*
Jun	Jun oncogene	−3.48*	−1.33
Lox	Lysyl oxidase	−3.49*	−1.34
Mmp13	Matrix metallopeptidase 13 (interstitial collagenase)	−2.60	1.20
Mmp14	Matrix metallopeptidase 14 (membrane-inserted)	−2.26	−1.34
Mmp3	Matrix metallopeptidase 3 (stromelysin 1)	−7.12*	−3.41*
Mmp8	Matrix metallopeptidase 8 (neutrophil collagenase)	−14.61*	1.49*
Mmp9	Matrix metallopeptidase 9 (type IV collagenase)	−5.91*	−1.07
Myc	Myelocytomatosis oncogene	−2.31	1.50*
Plat	Plasminogen activator, tissue	−3.65	1.88*
Plau	Plasminogen activator, urokinase	−2.21	−2.11*
Serpina1a	Serine (or cysteine) peptidase inhibitor, clade A, member 1A	−2.19	2.55*
Serpine1	Serine (or cysteine) peptidase inhibitor, clade E, member 1	−2.91	−1.08
Smad2	SMAD family member 2	−2.19	1.50
Smad6	SMAD family member 6	−4.40*	1.51*
Smad7	SMAD family member 7	−5.52*	1.17
Snai1	Snail homolog 1 (Drosophila)	−4.42*	1.50
Tgfb1	Transforming growth factor, β 1	−2.20	−1.34
Tgfb3	Transforming growth factor, β 3	−2.76	−1.06
Tgif1	TGFB-induced factor homeobox 1	−2.26	−2.69*
Thbs2	Thrombospondin 2	−2.18	1.50*
Timp4	Tissue inhibitor of metalloproteinase 4	−1.73	2.38

^*P < 0.05 for V₂O₅-treated D2 mice vs. V₂O₅-treated B6 mice.

Figure 6.

Total TGF-β protein levels in lung tissue homogenates from B6 and D2 mice 10 and 21 d after aspiration of PBS or 4mg/kg V₂O₅. Bars represent means + se of 4 mice/group. ⁺P < 0.001 vs. B6 mice within time point and treatment group; ^‡P < 0.001 vs. d 10 within strain and treatment.

V₂O₅ effects in BXD RI mice

We next investigated the genetic basis for differential inflammatory and fibrotic responses in B6 and D2 mice using 49 BXD RI strains. We found treatment and strain effects (Fig. 7) for BALF cell numbers and lung collagen content between RI strains 14 d after PBS and 4 mg/kg V₂O₅. PBS control BALF cell numbers ranged from 1.9 × 10⁴ in BXD11 to 24.1 × 10⁴ in BXD55, while V₂O₅ responses ranged from 8.6 × 10⁴ in BXD11 to 73.2 × 10⁴ in BXD9. Collagen content in the right lung also varied with strain in controls from 19.5 μg in BXD34 to 60.0 μg in BXD44, and in response to V₂O₅ from 28.5 μg in BXD24 to 117.2 μg in BXD44. Because of the strain-dependent differences in controls, the mean PBS response for each strain was subtracted from individual V₂O₅-induced response, and this difference was used for genetic mapping of V₂O₅-induced phenotypes. Total cells, macrophages, and lymphocytes significantly correlated with each other, but we found no correlation between V₂O₅-induced increases in collagen content and any cell type (Fig. 8A).

Figure 7.

BXD RI strain distribution for PBS and V₂O₅-induced BALF total cell numbers (A) and lung collagen content (B). Each BXD RI strain was exposed to 4 mg/kg V₂O₅ or PBS, and phenotypes were measured at study d 14. Bars represent means ± se of 4–8 mice/strain in each treatment group. *P < 0.05 vs. PBS treatment within strain.

Figure 8.

A) Matrix shows significant correlations between V₂O₅-induced total cells, macrophages, and lymphocytes, but a lack of correlation between collagen and any cell type for 49 BXD RI strains. B) Genome-wide interval map of V₂O₅-induced changes in lung collagen content across 49 BXD RI strains using the WebQTL function of GeneNetwork to identify quantitative trait loci. The x axis represents the length of each chromosome, the left y axis indicates the log of the odds ratio (LOD, blue line), and the right y axis indicates the degree to which either C57BL/6J (red line) or DBA/2J (green line) alleles increase phenotypic values. Numbers along the top of the figure are chromosome numbers. The upper horizontal line indicates the significant LOD threshold (P<0.05), and the lower horizontal line indicates the suggestive LOD threshold. C) Detailed linkage map of chromosome 4 shows the significant quantitative trait locus for V₂O₅-induced changes in lung collagen content in the RI strains.

Genome-wide association mapping identified no significant QTLs for V₂O₅-induced changes in BALF cells, although there was a suggestive QTL on chromosome 12 for macrophages (data not shown). Association mapping for V₂O₅-induced changes in collagen content (Fig. 8B) identified a statistically significant QTL on chromosome 4 (Fig. 8C). Suggestive QTLs were found on chromosomes 2 and 9 (Fig. 8B). We then queried the haplotype structure of the QTL and searched for candidate genes (http://phenome.jax.org/db/q?rtn=snp/ret1). All BXD RIs have one of 2 haplotypes that were found in the chromosome 4 QTL (Fig. 9A): 29 RIs are resistant B6 haplotype and 19 are susceptible D2 haplotype. Relative to the B6 haplotype RI strains, mean V₂O₅-induced lung collagen content was significantly greater (P=0.002) in D2 haplotype RI strains (Fig. 9B). Genes within the QTL include Dbc1 (deleted in breast cancer 1), Cdk5rap2, Megf9 (multiple EGF-like-domains 9), Tle1, Aldoart1 (aldolase 1 A, retrogene 1), Rasef (RAS and EF hand domain containing), Frmd3, Kdm4c, and Ptprd. Two haplotypes and 3 haplotypes were found in the chromosomes 2 and 9 QTLs, respectively (Supplemental Fig. S2). Significantly greater V₂O₅-induced collagen content was found in D2-like strains compared to B6-like strains.

Figure 9.

Haplotype structure for the chromosome 4 quantitative trait locus for susceptibility to V₂O₅-induced changes in lung collagen content in BXD RI strains of mice. A) Haplotype with single-nucleotide polymorphism (SNP) location, gene names, and gene function in RI strains with similar haplotype. Parental B6 (left) and D2 (right) mice flank the RI strains. Cs, synonymous coding SNP; I, intronic SNP. B) Mean lung collagen content in RI strains with the B6 haplotype (n=29 strains) and strains with the D2 haplotype (n=19 strains).

Gene expression levels of 5 candidate genes within the collagen QTL on chromosome 4 were measured using mRNA from lungs of B6 and D2 mice at d 10 and 21 (Supplemental Fig. S3). All genes were expressed at relatively low levels (indicated by high C_t values) and Frmd3 was not detected in any of the samples. In D2 mice, V₂O₅ exposure increased Kdm4c expression at d 10, while Tle1 was increased at d 21 compared to controls. In contrast, Ptprd was decreased in B6 mice at d 21 and was undetectable in V₂O₅-treated D2 mice.

DISCUSSION

Interstitial lung diseases include a large number of respiratory conditions with various etiologies, pathological courses, and prognoses. Idiopathic forms of ILD are particularly difficult to study because, in the absence of a known cause and natural history, it has not been possible to develop an appropriate animal model. Existing models have taken advantage of known fibrogenic agents to study mechanisms of disease pathogenesis. While understanding of a number of fibrogenic pathways has been advanced with these models, few have successfully identified effective therapeutic targets for human disease (6). In the current study, we used vanadium to develop a novel, alternative model of pulmonary fibrosis to explore the genetic basis for fibrosis susceptibility. Differential lung injury, inflammation, and fibrogenic responses between responsive D2 and resistant B6 mice after identical V₂O₅ exposure demonstrated genetic background is an important determinant of susceptibility.

Functional measures indicative of restrictive disease were found in the fibrosis susceptible D2 strain, but significant increases in airway resistance were also evident. Increased airway resistance is consistent with V₂O₅-induced bronchitis-like phenotypes in rodents and primates including airway hyperresponsiveness, inflammation, mucus cell hyperplasia, and airway smooth muscle thickening (16–18, 26). Studies of human exposure to vanadium-rich particles in occupational settings have also reported functional and symptom changes indicative of obstructive disease (19–23). However, restrictive and obstructive lung diseases are not mutually exclusive, as seen in cases of combined pulmonary fibrosis and emphysema (27). Increased airway resistance in D2 mice may reflect airway smooth muscle and/or basement membrane thickening, or airway epithelial hyperplasia/hypertrophy. Although these phenotypes were not quantitatively measured, hypertrophy and hyperplasia of airway epithelial cells were found in D2 mice exposed to V₂O₅. Interestingly, these hyperplastic cells did not stain positive for mucus in the 2 strains we examined (data not shown). This contrasts with studies in other strains of mice and rats, in which V₂O₅ exposure induced goblet cell metaplasia, mucous production, thickening of airway smooth muscle, and peribronchiolar fibrosis (17, 28).

Elevated numbers of BALF macrophages were predominant in both strains of mice after V₂O₅, however they persisted throughout the study only in D2 mice. Alveolar macrophage phenotype (M1, M2) is thought to be essential in regulating inflammatory and repair processes after lung injury (29). Alternatively activated alveolar macrophages (M2) were shown be critical for fibrogenesis in the bleomycin model and are the prevalent phenotype in BALF from patients with IPF (30). However, additional subphenotypes and other mechanisms that determine resolution or continued progression of fibrosis are needed to explain progressive fibrosis in IPF compared to fibrosis resolution in mouse models.

Vanadium compounds affect host defense activities including phagocytic ability, production of ROS, enzyme activity, resistance to infection, and production and response to immunomodulatory cytokines (31–35). Previous reports of reduced phagocytic activity of macrophages and increased ROS and cytokine production after vanadium exposure may have contributed to the V₂O₅-induced injury and inflammation we observed, particularly in D2 mice. Changes in cytokine production after exposure to vanadium varies depending on the vanadium compound and cytokines investigated (36–38). A Th2 profile has been hypothesized to play a role in IPF and in bleomycin-induced lung fibrosis (39–43), although whether a Th2 response is required for fibrogenesis is unclear (44). In our study, Th2 cytokines measured in the fibrosis array were either undetectable (Il10 and Il13) or were down-regulated (Il4, Il5, Il13ra2) in D2 mice after 10 d and were not different from B6 at 21 d. Furthermore, in lung fibroblasts, V₂O₅ stimulates a Type I immune response (45). The lack of a significant Th2-like response suggests the immune response to V₂O₅ may differ from other fibrogenic models. It is also possible that the differences in immune responses may be time dependent, and further investigation is necessary to resolve these differences.

We also found greater numbers of eosinophils and elevated expression of eosinophil-specific chemotactic factor eotaxin (Ccl11; ref. 46) in D2 mice relative to B6 mice. Increased eosinophil numbers have been reported after exposure to metal-containing particles and may play a role in particle-induced exacerbation and/or induction of asthma-like parameters such as airway hyperresponsiveness (47–49). Increased eosinophil and neutrophil numbers in BALF from IPF patients have been reported to correlate with decrements in diffusing capacity and reduced survival time (50–54), suggesting that these cells may have a role in pulmonary fibrosis. In addition, inflammation is a key feature of a number of other ILDs that result in fibrosis.

Interestingly, mRNA expression of most fibrosis-associated genes was decreased in V₂O₅-treated D2 mice at 10 d, a time at which models such as bleomycin switch from inflammatory to fibrotic phenotypes (55). Although a significant increase in collagen levels at 10 d was found in D2 mice, there was also cellular inflammation, likely due to the second V₂O₅ exposure at d 7. The extensive decreased gene expression observed in the V₂O₅ mouse model at 10 d is consistent with in vitro studies of human lung fibroblasts in which V₂O₅ suppressed 3–4 times more genes than were induced (56). The list of genes that were down-regulated 10 d after V₂O₅ in D2 mice compared to B6 mice included a number of genes that are typically up-regulated after bleomycin challenge and thought to be key to pulmonary fibrosis such as Mmps and Tgfb1. TGF-β protein levels were also lower in D2 compared to B6 mice, although V₂O₅ treatment did not affect levels in either strain. Notably, Smad6 and Smad7, known to inhibit Tgfb signaling (57–59), were significantly up-regulated in B6 mice, but down-regulated in D2 mice suggesting that a lack of Tgfb inhibition may contribute to development of fibrosis in the V₂O₅ model.

At 21 d, when cell numbers decreased but collagen levels were still near their peak, only 5 genes were significantly up-regulated more than 2-fold, and 3 were down-regulated. Up-regulated genes included Ccl11, Ctgf, Fasl, Il1a, and Serpina1a. Ccl11 and Il1a are inflammatory genes and would be expected to contribute to the mixed eosinophilic/neutrophilic inflammation seen after V₂O₅ exposure; yet, granulocytic inflammation had largely resolved by 21 d. However, increased levels of Ccl11 have been associated with models of fibrotic conditions of the lung (60–63). Ctgf has previously been reported to be increased in response to V₂O₅ exposure in human lung fibroblasts (56) and is thought to be important in early wound healing (64, 65). Serpina1a, a mouse ortholog of α1 antitrypsin, is a protease inhibitor and its induction, along with continued down-regulation of Mmp3, may contribute to accumulation of extracellular matrix. Finally, the Fas-Fas ligand system has been shown to be up-regulated in IPF lung tissue leading to apoptosis of epithelial cells (66). Epithelial cell apoptosis was observed at early time points, but later apoptotic cells likely represented recruited inflammatory cells, while notable fibroblast apoptosis was not observed at any time (data not shown).

The wide range of values for BALF cells and lung collagen content across 49 BXD RI strains suggested that cellular inflammation and fibrosis are complex traits involving multiple genetic factors. Because the role of inflammation in pulmonary fibrosis is controversial in some ILDs such as IPF, we asked whether inflammatory cell phenotypes were related to lung collagen content. The lack of a correlation between collagen and cell numbers could be a temporal effect as significantly greater cellular inflammation was found in D2 mice at early time points, but it is also possible that inflammation and fibrosis may be controlled by different sets of genes as suggested by the lack of concordance of QTLs between the phenotypes.

The lack of a fibrotic response and associated functional changes in B6 mice was surprising, as they are highly susceptible to pulmonary fibrosis induced by bleomycin, silica, and asbestos (2–5), suggesting that V₂O₅ may work through different mechanisms than other known fibrogenic agents. In our genome-wide linkage analyses to identify the genetic basis for the differential responses between strains, we found a significant QTL for the collagen phenotype on chromosome 4 that was largely driven by D2 alleles. A number of gene candidates were found within the QTL, including Kdm4c (or Jmjd2c, jumonji domain containing protein 2C) which was expressed at higher levels in D2 mice at 21 d. This histone methylase has been associated with cancer proliferation (67). Interestingly, the methylase has also been shown to selectively interact with HIF-1α (68), which has a role in inflammation and fibrosis (69). Another gene candidate in the chromosome 4 QTL is protein-tyrosine phosphatase receptor-type δ (Ptprd). Ptprd has been implicated as a tumor suppressor gene in lung carcinogenesis (70, 71), and its decreased expression by V₂O₅ to undetectable levels in D2 mice at 21 d may contribute to the ongoing epithelial hypertrophy/hyperplasia observed in D2 mice. Polymorphisms in PTPRD have also been associated with pediatric asthma (72) and may have a role in other complex lung diseases. The B6 and D2 haplotypes that contain these genes and others (see above) associated with differential V₂O₅-induced lung fibrosis in the RI strains. However, because these genes are in linkage disequilibrium with each other, additional studies will be necessary to characterize their roles in this fibrosis model. Significant QTLs for the cell phenotypes were not identified, which may suggest that numerous genes may have a relatively small impact on these phenotypes.

In summary, the pulmonary effects of V₂O₅, including lung injury, inflammation, and fibrosis, are dependent on genetic background. This V₂O₅ model is useful for investigating fibrotic mechanisms and therapeutic targets, as it appears to be more persistent than the more commonly used bleomycin model and likely works through different mechanisms. Because vanadium is a component of cigarette smoke and metal dusts, which have been associated with ILDs such as IPF, it also has the advantage of being an environmentally/occupationally relevant agent. Whether continued exposures would produce even more stable or progressive fibrosis in susceptible strains presents an interesting area for further research, while genetic studies to address variant alleles across numerous strains of mice are underway.