Lung Deposition and Clearance of Inhaled Vanadium Pentoxide in Chronically Exposed F344 Rats and B6C3F1 Mice

Jeffrey A. Dill*,1, Kyeonghee M. Lee*, Kathleen H. Mellinger*, Derrick J. Bates*, Leo T. Burka{dagger} and Joseph H. Roycroft{dagger}

* Battelle, Toxicology Northwest, Richland, Washington 99352, and {dagger} National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709

Received August 6, 2003; accepted September 25, 2003


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Female F344 rats and B6C3F1 mice were exposed to vanadium pentoxide (V2O5) at concentrations of 0, 0.5, 1, or 2 mg/m3 (rats) and 0, 1, 2, or 4 mg/m3 (mice) for 6 h/day, 5 days/week (for up to 18 months), by whole-body inhalation. Lung weights and lung burdens of vanadium were determined for exposed animals after 1, 5, and 12 days and after 1, 2, 6, 12, and 18 months of V2O5 exposure. Blood vanadium concentrations were determined at 1, 2, 6, 12, and 18 months for all animals including controls. A model that assumed a first-order deposition rate and a first-order elimination rate for vanadium was employed to fit the lung burden data. Comparisons between exposed groups indicated a progressive increase in lung weight with exposure concentration and time on exposure for both species. The vanadium lung burdens appeared to reach steady state in the lowest exposure groups (0.5 and 1 mg/m3 for rats and mice, respectively) but showed a decline in the higher exposure groups. This deposition pattern was similar between rats and mice but the maximum lung burdens were observed at different times (1 or 2 months in mice vs. 6 months in rats). The vanadium deposition rate decreased faster in mice, while the elimination half-lives of vanadium lung burdens were about six- to nine-fold shorter in mice than in rats at 1 and 2 mg/m3. Thus, the retention of vanadium in the lungs at 18 months was lower in mice (~2% retained) compared with rats (13–15% retained) at the common exposure concentrations of 1 and 2 mg/m3. The lung burden data were approximately proportional to the exposure concentration in both species, likely due to concomitant decreases in deposition and elimination to a similar extent with increasing exposure. The area under the lung burden versus time curves and the area under the blood concentration (control-normalized) versus time curves were also proportional to exposure concentration. The progression of pathological changes in the lung with exposure and time is thought to affect the pattern and/or extent of vanadium deposition in the lungs following repeated exposures to V2O5.

Key Words: vanadium; vanadium pentoxide (V2O5); lung deposition; clearance; chronic inhalation; toxicokinetics.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Vanadium is a group Vb transition element commonly found in various types of ores, coals, and oils. Vanadium pentoxide (V2O5) is the most common commercial form of vanadium and is the primary form found in industrial exposure situations. Vanadium is used as an alloy to increase strength, suppress aging effects, and improve mechanical properties in steel, and V2O5 is the principal starting material for production of vanadium compounds. V2O5 is also used for manufacturing yellow glass, as a depolarizer, and as developer in photography (HSDB, 2003Go; NTP, 2002Go). Many metallurgical processes and boiler-cleaning operations generate dust containing V2O5 and vanadium trioxide compounds.

Atmospheric emissions of vanadium from natural sources have been estimated at 8.4 metric tons/year globally (ranging from 1.5 to 49.2 metric tons). By far the most important source of environmental contamination with vanadium is the combustion of oil and coal; about 90% of the ~64,000 metric tons of vanadium that are emitted to the atmosphere each year from both natural and anthropogenic sources comes from oil combustion (WHO, 2001Go). The permissible exposure limit set by the Occupational Safety and Health Administration (OSHA) is 0.5 mg/m3 (dust; NIOSH, 1997Go).

A 16-day inhalation study of V2O5 was previously performed to characterize the acute toxicity, rates of cell proliferation in the lungs, and lung deposition and clearance of vanadium (NTP, 2002Go). Female F344 rats and B6C3F1 mice were exposed by whole-body inhalation to 0, 1, 2, or 4 mg/m3 (rats) and 0, 2, 4, or 8 mg/m3 (mice) V2O5 for 6 h/day, 5 days/week for up to 12 exposures over 16 days. The severity of lung lesions increased with exposure concentration and time of exposure in both species. There was a significant concentration-related increase in bronchiolar epithelium bromodeoxyuridine-labeled nuclei in rats exposed to 2 or 4 mg/m3 V2O5 for 6 days, correlating with the histopathology results. Half-lives for vanadium clearance from the lung were 4 to 5 days in rats exposed to 1 and 2 mg/m3 V2O5 and 2 to 3 days in mice exposed to 2 and 4 mg/m3 V2O5, respectively. Vanadium lung burdens during the exposure increased in proportion to exposure concentration for each species.

There are other data available in the literature concerning the elimination of vanadium from the lungs in animals exposed to vanadium compounds. Most studies employed a short-term single exposure and the reported elimination of vanadium varies greatly, depending on the chemical and physical form of vanadium, the dose, and the route of administration. In rats, Conklin et al. (1982)Go reported that 40% of the intratracheally instilled V2O5 dose (40 µg V/animal or 0.3 mg V/kg) was eliminated from the lungs within 1 h after dosing. By 3 days after dosing, 90% of the administered dose had been cleared and no further decline in vanadium lung burden was observed. Rhoads and Sanders (1985)Go intratracheally instilled female rats with suspensions of V2O5 (40 µg V/animal) and, subsequently, observed a biphasic elimination profile. The rapid-phase elimination half-life was 11 min and the terminal-phase elimination half-life was 1.8 days. Other investigators (Sharma et al., 1987Go) intratracheally instilled male rats with soluble sodium orthovanadate or less soluble V2O5 at a higher dose (200 µg V/animal or 0.91 mg/kg). When administered as V2O5, clearance of vanadium from the lung followed a simple, first-order elimination pattern. On days 1, 7, and 28, the vanadium lung burden represented >29, 22, and 13% of the administered dose, respectively, suggesting a relatively long elimination phase. However, when administered as the more soluble sodium orthovanadate, vanadium was eliminated more rapidly. Vanadium concentrations fell to 11, 3, and 1% of the administered dose at 1, 7, and 28 days after dosing, respectively, but did not appear to follow a straight, first-order profile.

The National Toxicology Program recently completed toxicity and carcinogenicity studies of inhaled V2O5 in male and female rats and mice (NTP, 2002Go). The studies were conducted by whole-body inhalation exposure of F344 rats and B6C3F1 mice to 0, 0.5, 1, or 2 mg/m3 (rats) and 0, 1, 2, or 4 mg/m3 (mice) V2O5, respectively, for up to 2 years; the chronic toxicity and carcinogenic potential of V2O5 were reported in Ress et al. (2003)Go. To evaluate the vanadium lung deposition and clearance for up to 18 months of V2O5 exposure, additional female animals were included at all exposure concentrations in conjunction with the 2-year studies (NTP, 2002Go). Toxicokinetic findings from the lung burden studies are presented in this paper.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Test chemical.
V2O5 (purity >=99.0%) was obtained from Shieldalloy Metallurgical Corp. (Newfield, NJ) and was stored at room temperature. Analysis by inductively coupled argon plasma (ICP)/atomic emission spectrometry (AES) for trace elements (aluminum, boron, barium, beryllium, calcium, cadmium, chromium, copper, iron, magnesium, manganese, molybdenum, sodium, potassium, nickel, lead, strontium, and zinc) indicated that each of these elements was present at a concentration <=151 µg/g (0.015%). Identity as V2O5 was confirmed using X-ray diffraction (XRD) analysis. Purity and identification of bulk test material were also confirmed by ICP/AES. Purity assays were repeated periodically during the course of the 2-year study, in which ICP/AES was used to determine the percent by weight of vanadium as described above. Relative purity for vanadium in the test material ranged from 98.4 to 101% and was acceptable throughout the study. Filter samples were collected from the occupied 0.5 (rat) and 4 mg/m3 (mouse) chambers, the distribution line, the bulk test material, and the generator reservoir and analyses by ICP/AES and XRD indicated that V2O5 was stable in the exposure system.

Generation and monitoring of V2O5 exposure.
The aerosol generation and delivery systems included a linear dust feeder, particle attrition chamber, and an aerosol distribution system. The flow-rate through the distribution line was controlled by Air-Vac pumps (Air-Vac Engineering, Milford, CT) and the pressure was monitored by photohelic differential pressure gauges (Dwyer Instruments, Inc., Michigan City, IN). At the exposure chamber inlet duct, the V2O5 particles were diluted with conditioned air to achieve the desired exposure concentration.

The V2O5 aerosol concentration in the exposure chambers was monitored using real-time aerosol monitors (RAMs; model RAM-1; MIE, Inc., Bedford, MA). The mean concentrations for all exposure chambers were 100–101% of the targets, with overall relative standard deviations (RSDs) of 7–9%. Particle size distribution of V2O5 aerosols was determined throughout the study with a Mercer-style, seven-stage impactor (In-Tox Products, Albuquerque, NM). The mean mass median aerodynamic diameters (µm ± GSD) were ~1.2 ± 1.9 and ~1.3 ± 1.9 for rat and mouse chambers, respectively.

Animals.
Battelle’s animal care and use program is fully accredited by the AAALAC, and the study protocol was approved by the IACUC before the study started. Female F344/N rats and B6C3F1 mice (~4 weeks old at receipt) obtained from Taconic Farms, Inc. (Germantown, NY) were quarantined and acclimated in a humidity- (RH: 55 ± 15%) and temperature- (T: 75 ± 3°F) controlled environment for ~2 weeks prior to the study. Animals were housed individually in stainless steel, Hazelton 2000, inhalation-exposure chambers (Harford System Division of Lab Products, Inc., Aberdeen, MD). Environmental conditions in the chambers were maintained the same as for the quarantine period, with air flow through the chambers at 15 ± 2 air changes/h. Water and food (NTP-2000; Zeigler Bros., Inc., Gardners, PA) were available ad libitum, except food was withheld during exposures. A 12-h light:dark cycle (light started at 0600 h) was maintained throughout the studies.

Experimental design.
Female rats and mice (~6 weeks old at exposure start) were exposed to V2O5 by whole-body inhalation at three concentrations [0, 0.5, 1, or 2 mg/m3 (rats) and 0, 1, 2, or 4 mg/m3 (mice)], plus controls, for 6 h/day, 5 days/week (for up to 18 months and exclusive of weekends and holidays). Following the daily exposure on days 1, 5, and 12, and at 1, 2, 6, 12, and 18 months of study, lungs were collected, weighed, and vanadium concentration was determined from up to five animals per exposure group. Also, at 1, 2, 6, 12, and 18 months, blood was collected from up to five animals per exposure and control group for blood vanadium determination. The blood sampling started about 30 mins postexposure and all exposure groups (in the order of low, middle, and high exposure groups) were bled in ~1 hour.

Sample collections.
During scheduled blood sampling, control animals were anesthetized by 70% CO2, bled via retro-orbital puncture in the exposure room, and returned to the control chambers after bleeding. Each V2O5-exposed animal was wiped with a damp cloth to remove most of the test chemical from the fur before removal from the exposure room. At necropsy, exposed animals were anesthetized by ip pentobarbital and weighed. Two sets of surgical instruments were used during lung removal: one set was used to expose the viscera and a second set was used to remove lungs to reduce the risk of contamination of the lung with the test chemical. Under anesthesia, the ventral side of the animal was wiped with 70% alcohol and the thoracic cavity was opened and blood was collected from the heart. All blood samples were placed in a blood collection tube containing EDTA (Becton Dickinson and Co., Franklin Lakes, NJ) and stored on ice until transferred to a centrifuge tube (Corning, Corning, NY) and frozen at about -70°C. All animals were bled only once with the exception of the samples taken at 12 and 18 months in the control groups, for which the same animals were used at 1 and 2 months, respectively.

The lungs and associated tissues were removed and the mainstem bronchi, heart, and mediastinal tissues were trimmed. The right and left lungs were weighed separately, after which the right lungs were stored at about -70°C for vanadium analysis. The left lung was processed for histopathological examination (NTP, 2002Go).

Sample analyses.
Samples of whole blood and lung were weighed into Teflon®-lined microwave digestion vessels (Teflon PFA liners; CEM Corp., Matthews, NC), digested in nitric acid, diluted to volume, and the digests were analyzed for vanadium using ICP/AES. Whole blood digests were analyzed using an ultrasonic nebulizer (CETAC Technologies, Inc., Omaha, NE). Lung digests were analyzed using a concentric glass nebulizer (Gilson Medical Electronics, Middleton, MI).

Vanadium lung burdens determined from the right lung were corrected to the amount of vanadium in the whole lung (µg V) by multiplying the concentration of vanadium in the right lung (µg V/g) by the total wet lung weight (g). In a preliminary study (data not shown), it was determined that the lung vanadium concentrations from the right lung were not different from the lung vanadium concentrations of the whole lung for both species and that lung burdens were not dependent on the portion of lung analyzed.

Toxicokinetics.
The toxicokinetic data analyses were performed using SAS 6.12 (SAS Institute, Inc., Cary, NC). All statistical tests were conducted at p = 0.05.

Vanadium blood concentrations.
The vanadium blood data were analyzed in two ways. First, they were analyzed as a two-factor analysis of variance (ANOVA) where the two variables, days of exposure and V2O5 exposure concentration, and their possible interaction were investigated. For comparison of the means for each exposure group to the control group, we used Dunnett’s test; we used Duncan’s test to compare days of exposure. Secondly, the blood data were analyzed as one-factor ANOVA on V2O5 exposure concentration within each exposure period. Comparison of the means for the one-factor ANOVAs was done by using defined contrasts involving different exposure groups (i.e., low vs. middle, middle vs. high, and low vs. high). The area under the blood concentration versus time curve (AUC) was calculated using the trapezoidal rule, integrating from the first to last time point.

Lung weights.
The lung weight data were analyzed as a two-factor ANOVA where the two main variables, days of exposure and V2O5 exposure concentration, and their possible interaction were investigated. Then the lung weight data were analyzed as one-factor ANOVA on V2O5 exposure concentration within each exposure period. Comparison of the means for the one-factor ANOVAs was done by using defined contrasts involving different exposure groups.

Vanadium lung burden.
Lung deposition and clearance parameters were calculated from vanadium lung burden data. A kinetic model that assumed a constant (zero-order) deposition rate and first-order elimination rate was tried initially, but as discussed in the Results this model did not provide a satisfactory fit to the data. Instead, a model that was based on linear differential equations to account for proportional (first-order) deposition and proportional (first-order) elimination rates of vanadium in the lung provided a better fit to the lung burden data over the 18-month period. The model used is described by Equation 1Go.


(1)

where V is the mass of vanadium (µg) in the lung at time t; t is time (days of study); kd is the initial lung deposition rate (µg/day); kr is the rate constant for lung deposition rate change (day-1); and ke is the lung elimination rate constant (day-1). As with many compartmental toxicokinetic models, the rates reported are composite rates that reflect various mechanisms leading to distribution or elimination, and it is not implied that any of these are dominated by a specific mechanism. For example, elimination rate for vanadium (Ke) could be a combination of macrophage-mediated phagocytosis and dissolution and absorption in the lung, and we have made no attempt to distinguish these individual rates.

The solution of this differential equation, with the boundary condition of V = 0 at t = 0, is described in Equation 2Go.


(2)

Estimates and uncertainties of toxicokinetic parameters (kd, kr, and ke) were estimated using SAS PROC NLIN (SAS Institute, Inc.) by fitting the model to the lung burden data at each exposure concentration. Based on these parameters, the half-life for lung deposition rate change (tr1/2, day) and the lung elimination half-life (te1/2, day) were calculated from Equations 3Go and 4Go, respectively.


(3)


(4)

Values of kd and kr were used to estimate the overall deposition rate in the lungs over time, and the absolute vanadium lung dose received at each exposure concentration was calculated by integrating these deposition rate curves over 18 months.

It was observed in fitting Equation 2Go to the lung burden data that numerical values for kr and ke could be interchanged and an equally good fit to the data obtained. This presented a problem in assigning the appropriate numerical values to kr and ke. This problem was solved by noting that if numerical values were assigned such that kr > ke, an unlikely scenario resulted in which the overall lung deposition rate fell rapidly to zero within just a few months and lung elimination half-lives were unrealistically long. However, if numerical values were assigned such that ke > kr, the overall deposition rate decreased very slowly and elimination half-lives were much closer to values typically observed for rats and mice. Accordingly, for modeling using Equation 2Go, values of kr and ke were constrained such that ke > kr.

The toxicokinetic parameters and uncertainties were also estimated using a "unified" model that took into account the exposure rate and allowed modeling of vanadium lung burden data from all exposure groups at the same time. Conceptually, this is accomplished by modeling the data simultaneously from all exposure concentrations and ensuring that there is a term in the model to directly account for the exposure rate. When fitting the unified model to the data, both sides of Equation 2Go were first normalized by a term representing the exposure concentration. This indicates that the lung burden data from each exposure group are divided by the exposure concentration for that group before fitting, which provided equal weighting to all data to account for unequal variances among the data. Then the unified model equation was fit to all of the normalized data from all exposure groups at once. When plotting the resulting unified model against non-normalized data from each exposure group, the right side of the unified model equation was first multiplied by the appropriate exposure concentration for that group. Use of exposure concentration as applied in this particular form of the unified model (i.e., a simple multiplication) assumed that lung burdens were proportional to exposure concentrations. Estimates and uncertainties of kr, tr1/2, ke, te1/2, and normalized kd (kd/exposure concentration) were estimated from the unified model.

The goodness-of-fit of the model to the data was evaluated according to Landaw and DiStefano (1984)Go, first by plotting the observed individual data with the model predictions. These plots showed very good agreement between the model and the data with most of the disagreement due to scatter in the replicate values. Possible different weighting schemes were evaluated during the model-building phase. Based on the results, an unweighted analysis was chosen when fitting the individual models, while a form of weighting (division by the exposure concentration) was used during fitting of the unified model. Evaluation of nonrandomness of errors was not applicable to the case at hand, given the multiple observations at each time point and no established run order. A goodness-of-fit test could not be performed since this is only available when weighting of 1/{sigma}2 is used, and this weighting was not used in any of our models. Values of standard error, 95% confidence interval, and coefficient of variation are reported with model-estimated toxicokinetic parameters, and these were all indicative of a reasonably good fit of the model to the data.

The model-independent empirical areas under the vanadium lung burden curves (area under trapezoids [AUTT]; T = time at termination [day] at 18 months on study) were calculated at each exposure concentration using the trapezoidal rule. The theoretical model-derived area under the vanadium lung burden curve at termination (AUCT) was also calculated based on kd and ke (error propagation was not evaluated because of the complexity). Additionally, values for kd, AUTT, and AUCT were normalized to exposure concentration to determine whether the values were dependent on exposure concentration.

The vanadium lung burden data for each group and time point were also normalized by dividing the lung burdens by the exposure concentration. These exposure-normalized data were tested for dose proportionality using PROC GLM in SAS. The analysis initially involved two-way ANOVA of lung burden data on exposure concentration and time on study. The exposure effect was significant as well as some of the interaction terms. Therefore, a final one-way ANOVA was run for each time point.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Blood Data
F344 rats.
Rats exposed to V2O5 had higher blood vanadium concentrations than controls (Table 1Go), although the differences in concentration between exposed and control rats were generally small. Blood concentrations tended to be higher in exposed rats early in the study and then declined slightly up to approximately day 173. Subsequently, blood concentrations in all exposed and control rats remained constant until the end of the study.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Blood Vanadium Concentrations in Female Rats and Mice
 
Several factors made it difficult to judge whether blood vanadium concentrations were proportional to exposure concentration. These factors included the small differences in blood concentration between controls and exposed animals and the relatively high uncertainty in the data. To eliminate the contribution of endogenous vanadium in the blood, mean blood concentrations in the exposed groups were corrected by subtracting the mean concentration in the control group at each time point. The areas under the blood vanadium concentration versus time curves (AUCs) calculated from these corrected blood concentrations were 24 ± 8, 34 ± 7, and 61 ± 9 µg • days/g in rats exposed to 0.5, 1, and 2 mg/m3, respectively (95% confidence interval). The AUCs normalized to exposure concentration were 48 ± 16, 34 ± 7, and 31 ± 5 µgdays/g/mg/m3, respectively. Given the high degree of uncertainty, differences between these values were not considered significant and blood concentrations integrated over time were roughly proportional to exposure concentrations.

B6C3F1 mice.
Mice exposed to V2O5 had higher blood vanadium concentrations than the controls (Table 1Go); except for mice exposed to 4 mg/m3 V2O5, which had the highest blood concentration at day 26, blood concentrations were highest in all mice at day 54.

Similar to rats, small differences in blood concentration between controls and exposed animals, especially for mice exposed to 1 and 2 mg/m3, and the relatively high uncertainty in the data made it difficult to judge whether blood vanadium concentrations were proportional to exposure concentration. When blood concentrations of exposed mice were corrected by subtracting the concentration of vanadium in controls, the AUCs for these corrected blood concentrations were 53 ± 15, 99 ± 13, and 210 ± 20 µgdays/g for mice exposed to 1, 2, and 4 mg/m3, respectively (95% confidence interval). The corresponding AUCs normalized to exposure concentration were 53 ± 15, 50 ± 7, and 53 ± 5 µgdays/g/mg/m3. There were no significant differences among normalized AUCs; therefore, blood concentrations integrated over time increased in proportion to exposure concentration.

Lung Data
F344 rats.
Lung weights among the exposed rats were compared over time to determine whether repeated exposure induced changes in lung weight (Table 2Go). This comparison was complicated somewhat by the fact that there were no lung weights available from control animals. Therefore, normal changes in lung weight with animal growth could not be taken into account to judge the significance of observed changes in exposed animals. However, analysis of variance did show significant increases in the lung weights from exposed rats as exposure concentration increased. Mean lung weights from the 1 mg/m3 group were significantly higher than those from the 0.5 mg/m3 group after days 12, 26, and 542 (p < 0.05). Lung weights in the 2 mg/m3 group were significantly higher than those in the 1 mg/m3 group at all times after day 54 (p < 0.01). Lung weights in the 2 mg/m3 group were significantly higher than those in the 0.5 mg/m3 group at all time points after day 12 (p < 0.05).


View this table:
[in this window]
[in a new window]
 
TABLE 2 Lung Weight and Lung Burden in Female Rats
 
Lung vanadium concentrations and burdens are shown in Table 2Go. Because lung weights increased with time on exposure and exposure concentration, the behavior of vanadium lung deposition and clearance as indicated by lung concentrations was confounded by simultaneous changes in lung weights. Accordingly, vanadium lung burdens (µg V/total lung), rather than lung concentrations, were used to evaluate vanadium deposition and clearance in the lungs.

Various tests were run to determine whether vanadium lung burdens were proportional to exposure concentration. Although exposure-normalized lung burdens (Table 2Go) indicated that lung burdens did not appear to increase in proportion to exposure concentration at many time points, there were several time points where normalized lung burden data increased proportionally with exposure concentration, especially later in the study. Conversely, area under the lung burden versus time curves clearly increased in proportion with exposure concentration (AUTT and AUCT; Table 3Go), and area under the lung burden versus time curves normalized to exposure concentration were equal (Table 3Go). Thus, if integrated over all time points, lung burden tended to increase proportionally with increasing exposure concentration. Visual inspection of the mean lung burden data also supported the finding that lung burdens increased roughly in proportion to exposure concentration, and the departures from proportional behavior were small. Integrated blood concentrations generally increased in proportion to exposure concentration (based on AUC values), although this behavior was made less certain by the variability in the data. Taken together, these observations indicate that lung burden increased proportionally with exposure concentration.


View this table:
[in this window]
[in a new window]
 
TABLE 3 Lung Deposition and Clearance Parameters with Error Estimates and One-Half 95% Confidence Intervals for Female Rats
 
B6C3F1 mice.
The lung weights of exposed mice increased rapidly up to about day 26 and continued to increase thereafter, though at a slower rate (Table 4Go). When lung weights were compared over time, ANOVA showed significant differences in the lung weights among different exposure groups. There were no significant exposure-related effects on lung weight at days 1 or 5. Significant differences in lung weights between the 1 and 2 mg/m3 groups were seen only at day 12 (p < 0.01), while differences in lung weights between the 2 and 4 mg/m3 groups were significant at days 12, 26, and 171 (p < 0.02). The difference in lung weights between the 1 and 4 mg/m3 groups was significant at all times (p < 0.05) from days 12 through 362. As with rats, analysis of variance on the mouse lung weight data showed a progression of increased lung weights with increasing exposure concentration and time of exposure. Therefore, vanadium lung burden rather than lung concentration was used to evaluate vanadium lung deposition and clearance.


View this table:
[in this window]
[in a new window]
 
TABLE 4 Lung Weight and Lung Burden in Female Mice
 
Several tests were applied to determine whether lung burdens increased in proportion to exposure concentration. Lung burdens normalized to exposure concentration (Table 4Go) increased in proportion to exposure concentration for the majority of time points at which lung burdens were determined. The maximum lung burdens observed at days 26 or 54 (Table 4Go) were 3.0, 5.9, and 11.3 µg for 1, 2, and 4 mg/m3, respectively, which are approximately proportional to exposure concentration. Similarly, areas under the lung burden versus time curve (AUTT and AUCT; Table 5Go) increased proportionally with exposure concentration. Area under the lung burden versus time curves normalized to exposure concentration were essentially identical (Table 5Go), which further demonstrates proportionality of lung burden with exposure concentration when integrated over time. Visual inspection of the normalized mean lung burden data indicated that lung burdens increased roughly in proportion to exposure concentration. Integrated blood concentrations increased in proportion to exposure concentration (based on AUC values). Taken together, these observations indicate that lung burden in exposed mice increased in proportion to exposure concentration.


View this table:
[in this window]
[in a new window]
 
TABLE 5 Lung Deposition and Clearance Parameters with Error Estimates and One-Half 95% Confidence Intervals for Female Mice
 
Lung Deposition and Clearance Calculations Using First-Order Deposition Models
The initial model investigated total vanadium lung burden based on a linear differential equation with a constant deposition rate (zero-order) and a proportional elimination rate (first-order) of vanadium from the lung. This model generally predicts asymptotic steady-state lung burdens after sufficient exposure time, regardless of exposure concentration. The model was adequate for lung burden data from the 0.5 and 1 mg/m3 groups for rats and mice, respectively, which showed an apparent steady state; but, a steady state was not seen for the groups at higher concentrations. Differences between the model simulation (figures not shown) and the data appeared to be due to a real difference in lung burden behavior rather than random variability, suggesting that the model assumption of zero-order deposition might not be adequate to fit the lung burden data (data for zero-order deposition not included). The final model employed a nonconstant (first-order) deposition rate to describe the decreasing deposition of vanadium in the lungs with time.

F344 rats.
The model fit the lung burden data from each exposure concentration well, including those from the latter phase where the lung burden was decreasing (Fig. 1Go). Toxicokinetic parameter estimates obtained from the model are shown in Table 3Go. The vanadium elimination rate constant (ke; Table 3Go) decreased and the elimination half-life (te1/2; Table 3Go) increased between 0.5 and 1 mg/m3, although the difference in elimination rate between 1 and 2 mg/m3 was not significant. The rate constant for the deposition rate change (kr; Table 3Go) was 0 for the 0.5 mg/m3 group, meaning that over the course of the study there was no change in the deposition rate and the lung burden achieved a steady state. Accordingly, the theoretical half-life for deposition rate change (tr1/2) could not be calculated. However, kr had a finite value that was not significantly different at exposure concentrations of 1 and 2 mg/m3, indicating a slow exponential decline in deposition rate over time at these exposure concentrations. The normalized initial deposition rate (kd/exposure concentration; Table 3Go) was not constant so the initial deposition rate of vanadium increased in a less than proportional manner as exposure concentration increased. Therefore, due to a reduction in both the deposition rate and the elimination rate with increasing exposure concentration, the retained lung burdens were roughly proportional to exposure concentration.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 1. Vanadium lung burdens in female rats. Curve represents the fit of the lung deposition and clearance model to the data.

 
The overall simulated deposition rate in the lungs (Fig. 2Go) was essentially constant for the 0.5 mg/m3 group but decreased over time for the two higher concentration groups. After 18 months of exposure, the deposition rate for the 1 mg/m3 exposure group, originally about twice that of the 0.5 mg/m3 group, decreased to a deposition rate similar to that of the 0.5 mg/m3 group. Integrating the deposition rate curves (Fig. 2Go) over 18 months gives the absolute vanadium dose that the lungs of animals received at each exposure concentration (Table 3Go). The dose was marginally different for rats exposed to 0.5 or 1 mg/m3, while the lung dose for rats exposed to 2 mg/m3 was approximately twice that seen at the lower exposure concentrations. The normalized lung vanadium dose (Table 3Go) decreased with increasing exposure concentration due to the fact that overall deposition rates decreased over time at the higher exposure concentrations. Based on the last measured lung burdens (Table 2Go), rats retained ~10–15% of the estimated lung doses at 18 months on study.



View larger version (12K):
[in this window]
[in a new window]
 
FIG. 2. Model of simulated vanadium lung deposition rates in female rats.

 
Figure 3Go shows the unified model developed to estimate the vanadium lung burden for various exposure concentrations. The model resulted in good fits to all three exposure groups simultaneously. This suggests that the unified model is capable of estimating vanadium lung burden at exposure concentrations ranging from 0.5 to 2 mg/m3.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 3. Vanadium lung burdens in female rats—unified model. Curve represents the fit of the lung deposition and clearance model to the data.

 
B6C3F1 mice.
As with rats, the model fit the vanadium lung burden data from each exposure concentration well, including the data from the latter phase when the lung burden was decreasing (Fig. 4Go). Toxicokinetic parameter estimates obtained from the model for mice are shown in Table 5Go. The ke decreased slightly as exposure concentrations increased, with te1/2 (Table 5Go) more than doubling between the 1 and 4 mg/m3 groups. The estimate of kr for the 1 mg/m3 group was very small with a relatively high standard error (0.00031 ± 0.00023 day-1). Accordingly, the theoretical tr1/2 (Table 5Go) was unrealistically long (2210 ± 1650 days). This suggested that, over the course of the study, there was little change in the deposition rate and the lung burden achieved a steady state at 1 mg/m3. However, kr had higher values that were not significantly different at exposure concentrations of 2 and 4 mg/m3, indicating a slow exponential decline in deposition rate over time at these exposure concentrations. The normalized initial deposition rate (kd/exposure concentration; Table 5Go) was not constant, decreasing as exposure concentration increased. As with rats, the reduction in both deposition and elimination of vanadium with increasing exposure produced retained vanadium lung burdens that were approximately proportional to exposure concentration.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 4. Vanadium lung burdens in female mice. Curve represents the fit of the lung deposition and clearance model to the data.

 
The overall deposition rate for mice (Fig. 5Go) exposed to 1 mg/m3 was essentially constant, whereas overall deposition rates for the two higher concentrations decreased over time. By 18 months of exposure, the lung deposition rate for mice exposed to 1 mg/m3 was nearly equal to that of mice exposed to 4 mg/m3, but greater than that for mice exposed to 2 mg/m3. The total lung doses received from the 1 and 2 mg/m3 exposure were nearly identical (Table 5Go), whereas the total lung dose for the 4 mg/m3 group was slightly greater than that from the 2 mg/m3 group at 18 months. Based on the last measured lung burdens (Table 4Go), mice retained about 2 to 3% of the estimated total lung doses at 18 months of study.



View larger version (13K):
[in this window]
[in a new window]
 
FIG. 5. Model of simulated vanadium lung deposition rates in female mice.

 
Figure 6Go shows the unified model developed to estimate the vanadium lung burden for all exposure concentrations. The model resulted in fairly good fits, particularly for the two lower exposure groups. This indirectly supports the finding that lung burden was approximately proportional to exposure concentration, since proportionality is a fundamental assumption of the unified model.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 6. Vanadium lung burdens in female mice—unified model. Curve represents the fit of the lung deposition and clearance model to the data.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Daily exposure of female F344 rats and B6C3F1 mice to V2O5 at concentrations of 0, 0.5, 1, or 2 mg/m3 (rats) and 0, 1, 2, or 4 mg/m3 (mice) for up to 18 months caused a progressive increase in lung weight with exposure concentration and time of exposure. These increases in lung weights were accompanied by increases in incidence and severity of lung lesions that progressed throughout the study (Ress et al., 2003Go). Increases in chronic inflammation, interstitial fibrosis, and alveolar and bronchiolar hyperplasia/metaplasia and squamous metaplasia were observed in the lungs of exposed male and female rats. Alveolar/bronchiolar (A/B) neoplasms occurred in male rats exposed to 0.5 and 2 mg/m3 V2O5. A marginal increase in A/B neoplasms was also observed in female rats exposed to 0.5 mg/m3 V2O5. As with rats, increases in chronic inflammation, interstitial fibrosis, and alveolar and bronchiolar epithelial hyperplasia were observed in lungs of mice exposed to V2O5. A/B neoplasms were significantly increased in all groups of exposed mice.

Despite the progressive pathological changes in the lungs during the 2-year study (Ress et al., 2003Go), the lung burden data in female rats and mice were approximately proportional to the exposure concentration. The area under the lung burden vs. time curves were also proportional to the exposure concentration. Additionally, the areas under the blood vanadium concentration versus time curve were proportional to exposure concentration, provided the vanadium blood concentrations were first corrected for the concentrations in control animals.

The approximate proportionality of lung and blood data to exposure concentration may be explained by a comparable decrease in both the lung deposition and elimination rates of vanadium as exposure concentrations increased. A possible mechanism for decreased deposition rates for the higher exposure groups in this study is a detrimental change in pulmonary function brought about by V2O5-induced lung lesions. During a 13-week study of V2O5 (NTP, 2002Go), lung lesions were observed in the rats that included fibrosis, epithelial hyperplasia, perivascular inflammation, and bronchiolar exudate. In the same animals, there were increases in respiration rate, decreases in tidal volume, shortened times for inspiration and exhalation, and decreases in dynamic lung compliance. Exposure-related changes in minute volume, esophageal pressure, peak inspiratory and expiratory flows, and expiratory resistance were also observed, although the significance of these varied depending on exposure concentration. These data collectively suggested a restrictive-type lesion in the 4 and 8 mg/m3 groups, whereas the data from the 16 mg/m3 group were indicative of both restrictive and obstructive diseases. While fibrosis alone could have accounted for a restrictive lesion, epithelial hyperplasia in small airways could also contribute to this functional deficit. Lung lesions observed microscopically correlated well with the restrictive functional deficit (NTP, 2002Go). During the 2-year study, the incidence of abnormal breathing increased with vanadium exposure in mice. Although pulmonary function of exposed animals was not measured in the 2-year study, the progression of pathological changes in the lungs of rats and mice with exposure and time could possibly lead to a deficit in respiratory function, hence affecting the pattern and/or extent of vanadium deposition in the lungs following repeated exposure.

A significant outcome of modeling the vanadium lung burden data from this study is that, for both species, the model indicated a time-dependent decline in the deposition rates in the middle and high exposure groups but not in the low exposure group. The model-estimated decline in the deposition rate was more apparent in the mouse data: between day 1 and 18 months of study, the deposition rate for mice exposed to 4 mg/m3 declined from 0.62 to 0.27 µg/day, while that for mice exposed to 2 mg/m3 declined from 0.41 to 0.22 µg/day. The deposition rate for mice exposed to 2 mg/m3 actually fell below the deposition rate for mice exposed to 1 mg/m3 after about 300 days of exposure. These changes give rise to total lung doses that were not proportional to exposure concentrations. For each species, by the end of 18 months, the overall lung doses in the lower two exposure concentrations were almost identical.

The decline in deposition rate with time of exposure was demonstrated using a model that assumed that deposition is not constant and is allowed to vary with time. With few exceptions, however, other investigators modeling lung burden data for various particulates have used a model that assumes a constant deposition rate regardless of exposure concentration. The model used by other investigators also assumes that the elimination rate changes in proportion to the amount of chemical in the lung at any time (Raabe, 1982Go; Stober, 1999Go; Wolff et al., 1987Go). Using this constant deposition model, lung burdens during exposure initially increase at an almost linear rate that depends on the difference in the constant deposition rate and the elimination rate. After an exposure time equal to about five times the elimination half-life, retained lung burdens reach a steady-state value as a quasi-equilibrium between deposition and elimination is attained. However, in this study, lung burdens in the two higher exposure groups for both rats and mice did not appear to reach a steady-state value; rather, they declined after 2 (mice) or 6 (rats) months until 18 months of study. Thus, attempts to fit the commonly used model (constant deposition rate and proportional elimination rate) produced less than satisfactory results, especially for the higher exposure groups beyond 2 or 6 months. Instead, a model that allowed the deposition rate to decrease with time of study provided a better fit to these data.

Modeling the data as discontinuous processes was considered as an alternative to the chosen model. One possibility was modeling the early points (while the lung burden increases) using a zero-order deposition rate regardless of exposure concentrations. Subsequent time points could be modeled using another first-order deposition process with a slower deposition rate, once potential toxic effects of vanadium exposure altered the deposition. However, this approach proved impractical since it was not possible to determine when the onset of nonlinear effects became significant. Moreover, some of the phenomena that could lead to decreased deposition or clearance, as discussed below, could occur from the beginning of exposure. Both deposition and clearance changes must be taken into consideration simultaneously to understand the net lung burden. Therefore, the lung burden data were best handled by a single model that assumed continuous change, rather than using more than one discrete model in an attempt to explain temporal effects.

Although many lung burden models reported in the literature assume a constant deposition rate, it is well recognized that there are many different factors that may control or affect particle deposition in the lungs (Schlesinger, 1995Go). These factors include particle characteristics of size, shape, and electrical charge; respiratory tract geometry; ventilation characteristics of the exposed animal; and other factors such as exposure to irritants and lung disease. For instance, an increase in the respiratory rate increases the velocity of particles in the airways, causing greater deposition in the upper airways and lower deposition in the pulmonary region. Bronchoconstriction caused by inhalation of irritants tends to increase impaction deposition in the upper airways, thereby decreasing the number of particles available for deposition in the pulmonary region of the lung. Structural alterations, such as restriction or obstruction caused by disease, are also expected to reduce deposition or eliminate it altogether in some parts of the lungs (Schlesinger, 1995Go).

Much of the data in the literature concerning the elimination of vanadium from the lungs in animals exposed to vanadium compounds indicates that this process occurs rapidly but varies, depending on the chemical and physical form of vanadium administered, the dose, and the route of administration. Following the 16-day inhalation exposure to V2O5 (NTP, 2002Go), the elimination half-lives of vanadium in the lungs of rats exposed to 1 and 2 mg/m3 were 4 to 5 days, while those in mice exposed to 2 and 4 mg/m3 were 2 to 3 days. Elimination half-lives in rats and mice exposed to V2O5 for 18 months were much longer and increased with exposure concentrations, and elimination half-lives for mice (11–14 days) were considerably shorter than in rats (59–61 days) at the common exposure concentrations of 1 and 2 mg/m3. Therefore, although vanadium appears to be eliminated rapidly from the lungs of animals exposed to V2O5 following low doses or a short-term exposure, elimination is slowed with high doses or repeated exposure.

One possible reason for the variance in clearance half-lives of vanadium from the lung is that V2O5 and other vanadium compounds can alter alveolar macrophage integrity and function, thereby adversely affecting a vital clearance mechanism in the respiratory tract (Fisher et al., 1986Go; Schiff and Graham, 1984Go; Waters et al., 1974Go). Using decreased phagocytosis as the measure of functional macrophage impairment, Fisher et al. (1986)Go demonstrated a relatively steep dependency of this impairment on vanadium concentration. Accordingly, in vivo lung clearance studies using a variety of doses and durations of exposure may logically be expected to show considerable variance, as seen in the different studies reported in the literature. It is worth emphasizing that the deposition or elimination rates reported here are composite rates that collectively reflect various disposition mechanisms. Therefore, a specific toxicokinetic study designed to elucidate various mechanisms of vanadium disposition would be necessary to confirm the above hypotheses.

Blood concentrations were several-fold lower than lung concentrations, and, at the common exposure concentrations of 1 and 2 mg/m3, were one- to two-fold greater in mice than in rats. Although V2O5 is relatively soluble, low vanadium concentrations in the blood are most likely due to the fact that vanadium absorbed from the lungs is rapidly distributed to the liver, kidney, bone, and spleen, and is rapidly eliminated from all organs except bone (Conklin et al., 1982Go; Oberg et al., 1978Go; Sharma et al., 1987Go).

In conclusion, exposure of rats and mice to V2O5 for up to 18 months resulted in a concentration- and time-dependent increase in lung weights and lung burdens that was consistent with the pathological changes in the lung. The kinetic analysis of lung burden data suggested a reduction in lung deposition over time with the reduction being greater in the mouse. Lung elimination rates for vanadium were also longer at the higher exposure concentrations. These changes in deposition and clearance rates resulted in a lower than expected dose to the lung at the higher exposure concentrations for each species. The change in deposition rate may be attributed to the pathological changes in the lung and possibly to changes in pulmonary function, while the reduction in the elimination rate may be attributed to adverse effects of vanadium on lung macrophage integrity and function.


    ACKNOWLEDGMENTS
 
Although this research has been funded by the National Institute of Environmental Health Sciences (NIEHS, Research Triangle Park, NC), it does not necessarily reflect the views of the Institute, and no official endorsement should be inferred.


    NOTES
 
1 To whom correspondence should be addressed at Battelle, Toxicology Northwest, MS K4-16, PO Box 999, Richland, WA 99352. Fax: (509) 375-3649. E-mail: dillj{at}battelle.org. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Conklin, A. W., Skinner, C. S., Felten, T. L., and Sanders, C. L. (1982). Clearance and distribution of intratracheally instilled 48vanadium compounds in the rat. Toxicol. Lett. 11, 199–203.[CrossRef][ISI][Medline]

Fisher, G. L., McNeill, K. L., and Democko, C. J. (1986). Trace element interactions affecting pulmonary macrophage cytotoxicity. Environ. Res. 39, 164–171.[ISI][Medline]

Hazardous Substances Data Bank (HSDB) (2003). Vanadium Pentoxide (updated November 8, 2002). Maintained by the National Library of Medicine. Retrieved on April 6, 2003 from http://toxnet.nlm.nih.gov/cgi-bin/sis/search/f?./temp/~AAAUDaqNO:1.

Landaw, E. M., and DiStefano, J. J. III (1984). Multiexponential, multicompartmental, and noncompartmental modeling. II. Data analysis and statistical considerations. Am. J. Physiol. 246, R665–R677.[ISI][Medline]

National Institute for Occupational Safety and Health (NIOSH) (1997). NIOSH Pocket Guide to Chemical Hazards. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, Washington, DC.

National Toxicology Program (NTP) (2002). Toxicology and carcinogenesis studies of vanadium pentoxide (Case 1314–62–1) in F-344/N rats and B6C3F1 mice (inhalation studies), NTP Technical Rep. No. 507. NIH Publication No. 03-4441, U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, Research Triangle Park, NC.

Oberg, S. G., Parker, R. D. R., and Sharma, R. P. (1978). Distribution and elimination of an intratracheally administered vanadium compound in the rat. Toxicology 11, 315–323.[ISI][Medline]

Raabe, O. G. (1982). Deposition and clearance of inhaled aerosols. In Mechanisms in Respiratory Toxicology (H. Witschi and P. Nettesheim, Eds.) Vol. 1, pp. 27–76. CRC Press, Boca Raton, FL.

Ress, N. B., Chou, B. J., Renne, R. A., Dill, J. A., Miller, R. A., Roycroft, J. H., Hailey, J. R., Haseman, J. K., and Bucher, J. R. (2003). Carcinogenicity of inhaled vanadium pentoxide in F344/N rats and B6C3F1 mice. Toxicol. Sci. 74, 287–296.[Abstract/Free Full Text]

Rhoads, K., and Sanders, C. L. (1985). Lung clearance, translocation, and acute toxicity of arsenic, beryllium, cadmium, cobalt, lead, selenium, vanadium, and ytterbium oxides following deposition in rat lung. Environ. Res. 36, 359–378.[ISI][Medline]

Schiff, L. J., and Graham, J. A. (1984). Cytotoxic effect of vanadium and oil-fired fly ash on hamster tracheal epithelium. Environ. Res. 34, 390–402.[ISI][Medline]

Schlesinger, R. B. (1995). Deposition and clearance of inhaled particles. In Concepts in Inhalation Toxicology (R.O. McClellan and R. F. Henderson, Eds.), pp 191–224. Taylor and Francis, Washington, DC.

Sharma, R. P., Flora, S. J., Drown, D. B., and Oberg, S. G. (1987). Persistence of vanadium compounds in lungs after intratracheal instillation in rats. Toxicol. Ind. Health 3(3), 321–329.[ISI][Medline]

Stober, W. (1999). POCK model simulations of pulmonary quartz dust retention data in extended inhalation exposures of rats. Inhal. Toxicol. 11(4), 269–292.[CrossRef][ISI][Medline]

Waters, M. D., Gardner, D. E., and Coffin, D. L. (1974). Cytotoxic effects of vanadium on rabbit alveolar macrophages in vitro. Toxicol. Appl. Pharmacol. 28, 253–263.[ISI][Medline]

Wolff, R. K., Henderson, R. F., Snipes, M. B., Griffith, W. C., Mauderly, J. L., Cuddihy, R. G., and McClellan, R. O. (1987). Alterations in particle accumulation and clearance in lungs of rats chronically exposed to diesel exhaust. Fundam. Appl. Toxicol. 9, 154–166.[ISI][Medline]

World Health Organization (WHO) (2001). Vanadium pentoxide and other inorganic vanadium compounds. Concise International Chemical Assessment Documents, Document 29.





This Article
Abstract
FREE Full Text (PDF)
All Versions of this Article:
77/1/6    most recent
kfh005v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Disclaimer
Request Permissions
Google Scholar
Articles by Dill, J. A.
Articles by Roycroft, J. H.
PubMed
PubMed Citation
Articles by Dill, J. A.
Articles by Roycroft, J. H.