Environ. Sci. Technol. 2001, 35, 270-277
Olfactory Bulb Uptake and Determination of Biotransfer Factors in the California Ground Squirrel (Spermophilus beecheyi) Exposed to Manganese and Cadmium in Environmental Habitats G R A H A M B E N C H , * ,† T I N A M . C A R L S E N , † PATRICK G. GRANT,† JIM S. WOLLETT, JR.,† ROGER E. MARTINELLI,† JOHNNYE L. LEWIS,‡ AND KEVIN K. DIVINE§ Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, California 94550, Center for Population Health, University of New Mexico, Albuquerque, New Mexico 87131, and Lovelace Respiratory Research Institute, P.O. Box 5890, Albuquerque, New Mexico 87185
Manganese (Mn) and cadmium (Cd) profiles in olfactory bulbs of California ground squirrels (Spermophilus beecheyi) trapped at Lawrence Livermore National Laboratory’s Site 300 facility in California were determined with proton induced X-ray emission (PIXE). As a reference, Mn profiles in olfactory bulbs from laboratory rats exposed via nose-only inhalation to 0.53 mg/m3 Mn in the form of MnCl2 were also determined with PIXE. Atomic absorption spectrophotometry was used to measure soil Mn and Cd contents from the trapping sites and Mn and Cd contents in ground squirrel liver and leg muscle tissues. The data from laboratory rats revealed that Mn uptake into the olfactory bulb occurs via inhalation exposure. Data from ground squirrels and knowledge of the collection sites indicate that although several routes of exposure may occur, fossorial rodent olfactory uptake affords a significant exposure route to Mn and Cd in soils. Measured biotransfer factors (ratio of leg muscle tissue metal content to soil metal content) for Cd in ground squirrels were 103-fold greater than exposure modeling estimates based on oral Cd uptake data from livestock. The measurements for ground squirrel tissues show that when conducting ecological risk assessments for natural habitats considerable care should be taken in selecting transfer factors. Specifically, transfer factors derived from data pertaining to comparable exposure pathways and ecological setting should be used wherever possible.
Introduction Ecological risk assessments increasingly are used to evaluate the potential impact contaminants may have on a site’s ecosystem. The United States Environmental Protection Agency has produced guidance documents describing various assessment methods (1-5). A frequently used assessment method is the hazard quotient (HQ) and related techniques (6-9). The HQ technique involves estimating contaminant exposure to individual members of a species potentially found 270
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
at the site. Typically, the species selected for HQ analysis represent site-specific trophic levels and are those most likely to be exposed to contaminants. Exposure pathways are identified and exposure estimated. The estimates are compared to a critical endpoint (often from the literature) to determine the potential ecological hazard. Although it has been recognized that estimating impacts to individuals of a species may not be meaningful when the ecological unit of concern is the population, community, or ecosystem (2, 3, 10-11), the HQ technique remains useful in estimating impacts to threatened or endangered species (2, 3). The HQ method can also be illustrative when initially evaluating a contaminated site. While many of the guidance documents discuss techniques for estimating exposure, selecting endpoints, and determining the HQ, there is no specific guidance on selecting routes of exposure. The HQ technique has been most commonly applied to oral exposure pathways and has been useful in estimating exposures to compounds which biomagnify and/or bioaccumulate in the food chain (6, 1214). Many terrestrial ecosystems contain fossorial (burrowing) vertebrates. Such species are in close contact with contaminated soil compared to nonfossorial species and so have, in addition to a potential risk for contaminant ingestion, a risk from contaminant inhalation and/or dermal absorption. Inhalation exposures generally have concentrated on measurement of lung burdens of inhaled substances and subsequent systemic distribution and toxic effects (15). However, deposition in the olfactory epithelium via the nasal airways is a frequently overlooked pathway. In animal models, toxicants (including metals) can enter the central nervous system (CNS) directly via the olfactory epithelium (16). Olfactory neurons are in contact with both the nasal lumen and the olfactory bulbs, making them nearly the first tissue accessible to inhaled toxicants and potentially providing a direct single-cell pathway to the CNS. Studies have shown that this can occur even with inhalation of low, environmentally relevant doses of toxicants (17). This study had two major objectives. The first was to demonstrate that, for fossorial mammals, olfactory uptake affords a significant potential exposure route to manganese (Mn) and cadmium (Cd) in soil. Uptake of these metals via this route is of concern since Mn is a neurotoxicant (18) and Cd is a putative neurotoxicant (19). The second objective was to determine the accuracy of previously modeled biotransfer factors for Cd used in the HQ analysis conducted on ground squirrels at Lawrence Livermore National Laboratory’s (LLNL) Site 300 facility (8).
Experimental Section Field Site Description. Site 300 is 2711 ha. in area and is located in the Altamont hills of the Diablo range about 100 km southeast of San Francisco, California. It is primarily comprised of California annual type-grassland. Contaminants, including volatile organic compounds, high explosive compounds, metals (including Cd and Mn), and radioisotopes, have been detected in soils and surface water in isolated areas about the site (20); soils exhibit some of the highest * Corresponding author phone: (925)423-5155; fax: (925)423-7884; e-mail:
[email protected]. Contact information: Center for Accelerator Mass Spectrometry, L-397, Lawrence Livermore National Laboratory, Livermore, CA 94550. † Lawrence Livermore National Laboratory. ‡ University of New Mexico. § Lovelace Respiratory Research Institute. 10.1021/es0014180 CCC: $20.00
2001 American Chemical Society Published on Web 12/01/2000
FIGURE 1. Aerial photograph showing the two ground squirrel collection areas at Site 300. Elevation contours are at 50 foot intervals. Black regions designate the areal extent of the two collection sites. contaminant concentrations of any of the environmental media (8). Several of these areas are large enough to pose a potential threat to resident California ground squirrels (Spermophilus beecheyi). California Ground Squirrels. California ground squirrels are fossorial mammals common to the xeric grasslandshrubland habitats of the San Joaquin valley and adjacent hills, including the Diablo range. Ground squirrel colonies are abundant at Site 300. They den underground in burrows connected by a system of tunnels. At Site 300 ground squirrels spend 80% of their time in burrows (8). Typically the squirrels do not stray far from burrow entrances; their home ranges average 0.16 ha. for males and 0.24 ha. for females (21-23). Due to their fossorial nature ground squirrels can be expected to come into frequent contact with contaminants when living in areas containing contaminated soil. They are omnivorous, but primarily consume vegetation and can ingest soil incidentally (8). California ground squirrels rarely live longer than 4 years (22). Juveniles and yearlings respectively represent about 50% and 30% of the population (23). Ground squirrels begin breeding the year following birth and produce one litter annually. The breeding season occurs between March and June. Gestation takes 30 days and at birth pups weigh ∼10 g (22). Growth is rapid, young are weaned by 55 days, and by 18 weeks average mass is 485 g. Average adult male mass is ∼620-660 g (23, 24). Selection of Collection Areas. Two areas in the northwest corner of Site 300 in hilly terrain were selected for collection of ground squirrels during the winter of 1998-1999 (Figure 1). One collection area was located in a vale west of the B850 explosives firing table and will be referred to as the B850 collection area. Previous studies revealed that this ∼2 ha. area contained elevated levels of soil Mn and Cd. The other collection area was located in a vale west of the west observation point (WOP). Previous studies revealed that this ∼3 ha. area had low soil Mn and Cd concentrations. Each
collection area had at least 11 active ground squirrel burrows in a colony structure. Collection and Processing of Soil Samples. Before trapping ground squirrels, soil was sampled at three locations within each collection area to verify previously measured Mn and Cd soil levels. At both sites, sampling locations were separated by ∼25 m, were near the center of the collection area, and were surrounded by active ground squirrel colonies. Each sample consisted of ∼100 cm2 of surface soil to a depth of 3 cm. Visible organic matter was removed, and each sample was dried for 12 h at 70 °C and then for 48 h at 103 °C. Each sample was successively passed through a series of three sieves with respective mesh spacings of 5 mm, 3 mm, and 1 mm to isolate the e1 mm soil fraction. For each sample, approximately 10 g of the e1 mm fraction was digested at 80 °C for 48 h in 70% nitric acid, centrifuged, and the supernatant filtered (0.45 µm pore size). The soil pellet was then rinsed with deionized water and centrifuged. The aqueous fraction was then filtered and added to the prior supernatant for analysis of Mn and Cd content by atomic absorption spectrophotometry (AAS). The chemical form of Mn and Cd in the soils and the size distribution of the Mnand Cd-bearing particles were not measured. Ground Squirrel Trapping. Ground squirrel trapping and euthanization protocols were reviewed and approved by the LLNL institutional animal care and use committee. Ground squirrels were collected using live traps, covered with a tent. The traps were baited with oats and sunflower seeds. The traps were opened at dawn, checked periodically until 3 P.M., and left closed overnight. Although adult California ground squirrels can be torpid during winter months, juveniles are active throughout the first winter after birth (23). Activity peaks during sunlight hours. Because the goal was to obtain yearling males, trapping occurred during January and February 1999 (winter months). Yearling males captured during this time period had an average age of 9 months. A roughly square grid system of trapping was employed at each VOL. 35, NO. 2, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
271
TABLE 1: Attributes of Yearling Male California Ground Squirrels Trapped in This Study mean air intakeb mean age of mean lifetime nasal wta (kg) (m3/day) trapped animalsa (days) air intakeb (m3) 0.60
0.40
270
a Measured or determined in this study. equations (25).
110 b
nasal passage air intake (m3/day) ) 0.5913µB0.78 (1) where BM is in kg. The validity of eq 1 for rodents was verified by calculating daily air intake in rats and comparing the result with rat respiratory data (26). To determine lifetime nasal air intake of the euthanized ground squirrels eq 1 was calculated using a 0.6 kg body mass, and the result was multiplied by the average age of the trapped males. Equation 1 scales with body mass so this procedure will probably overestimate squirrel lifetime nasal air intake. However, as growth of pups is rapid the estimate is presumed to be accurate to within a factor of 2. The allometric equations also indicate that brain and many other organ masses increase nearly linearly with body mass in rodent species. Consequently, olfactory bulb mass was assumed to scale linearly with body mass for rats and ground squirrels. Table 1 shows attributes assumed as typical for the ground squirrels. Nose Only Inhalation Exposure of Laboratory Rats. Ideally one would like to perform inhalation exposures to ground squirrels under controlled laboratory conditions. Unfortunately, no commercial supply of ground squirrels was available, and import of feral animals into the animal care facilities at LLNL was prohibited. Consequently, laboratory exposures were performed on rats. Although rats and ground squirrels may have differences in anatomy and physiology, olfactory bulb metal uptake and subsequent clearance are unlikely to differ substantially between the two species as they are similar sized fossorial rodents with similar nasal pathologies. Laboratory rat exposure protocols were reviewed and approved by both the Lovelace Respiratory Research Institute (LRRI) and LLNL institutional animal care and use committees. Laboratory exposures were nose only exposures. Nose only exposure minimizes systemic accumulation from either skin absorption or from grooming, swallowing, and subsequent gastric absorption. Although it would have been preferable to perform exposures with both Cd and Mn, sufficient funds were only available for exposures with one 9
mean wta (kg)
air intakeb (m3/day)
0.20
0.17
a
Measured in this study.
estimated Mn dose to nasal airways in Mn-exposed ratsa (mg/animal) 0.36 ( 0.07 b
Derived using allometric equations (25).
Derived using allometric
collection area with individual traps set ∼30 m apart over an area of ∼150 × 150 m. Topographic departures from the square grid system were necessary due to the terrain and the desire to place traps adjacent to active burrows. Trapping continued at each area until six males were collected. Visual identification of sex prior to handling was made, and females were released back into the environment. Males were collected in canvas bags and euthanized within 2 h of collection. Before euthanization ground squirrels were weighed. Each had a mass between 550 and 670 g, averaging ∼600 g, for both collection areas. Examination of the pelts, teeth, and claws of the squirrels after euthanization indicated that all but one were yearlings. Ground Squirrel Nasal Air Intake. Little data are available on ground squirrel respiration and olfactory bulbs. However, lung tidal volumes and respiratory rates can be calculated from animal body mass (BM) using allometric equations reported by Peters (25). Assuming nasal air intake is equivalent to the product of lung tidal volume and respiratory rate yields (25)
272
TABLE 2: Attributes of Laboratory Rats Used in This Study
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
compound. MnCl2 was chosen as the exposure compound as Mn is neurotoxic following inhalation in occupational environments (18). Nose only inhalation exposure methodology has been described elsewhere (27). The exposure apparatus consisted of a self-contained closed system on a multiport chamber. MnCl2 aerosols were generated by a venturi disperser (28) and monitored with a light scattering mass monitor. The atmosphere was confirmed by measuring the flow rate, aerosol concentration, and size distribution of particles. For the data reported here five male Harlan Sprague Dawley, Fisher 344 adult rats were exposed 8 h/day for 12 days to soluble MnCl2 (0.53 ( 0.01 mg/m3 of Mn, mass mean aerodynamic diameter (MMAD) 2.1 ( 0.5 µm) via nose-only exposure. The NIOSH threshold limit value for Mn inhalation exposure in the form of MnCl2 is 1 mg/m3 Mn, and a 2 µm particle MMAD ensures a high deposition efficiency in the olfactory region of the nose. A vehicle control group of five rats was concurrently exposed to filtered room air on a different multiport exposure chamber. In each exposure group rats had a mass between 170 and 240 g (mean ∼ 200 g). Mn nasal inhalation in Mn exposed rats was estimated using eq 1 to be 0.030 ( 0.006 mg/day. Table 2 shows attributes assumed as typical for the laboratory rats. Animal Euthanization and Tissue Preparation. Animals were anesthetized by carbon dioxide inhalation and exsanguinated by cardiac puncture and intracardial saline perfusion. During animal autopsy previously described efforts were made to reduce the possibility of contamination (29). This included removing the skin from the autopsied areas so that the hair side never touched the flesh. Olfactory tissue section preparation for PIXE analysis has been reported elsewhere (29) and is summarized here. Following perfusion, the dorsal surface of the skull was removed, the dura cut, olfactory nerve severed, and the brain and olfactory bulbs removed. Tissues were bisected down the midline, and each half was separately frozen in liquid nitrogen and stored at -80 °C until cryosectioning. Tissues were cryosectioned sagitally into 10 µm sections, which were mounted on nylon foils, freeze-dried, and stored in a dry environment until PIXE analysis (29). Nasal cavities were destroyed in order to extract the brain and olfactory bulbs from ground squirrel skulls, and, as a result, viable olfactory tissues from nasal airways were not collected. The liver and left hind leg thigh muscle were also collected from euthanized squirrels. As the ground squirrels are likely exposed to soil contents via ingestion and dermal as well as inhalation routes, measurement of metals in the gut and blood are unlikely to provide data that solely reflect systemic Mn and Cd resulting from inhalation. However, as metal contents in liver are frequently among the highest found in body tissues measurement of liver metal contents provides a useful measure of systemic metal exposure. Samples were immediately frozen in dry ice and subsequently stored at -80 °C. Prior to digestion samples were allowed to thaw and then oven dried for 48 h at 103° C. Samples were digested at 80 °C for 6 h in Aqua Regia (50:50 HCl:HNO3) for analysis by AAS. AAS Analysis of Digested Samples. All AAS measurements were performed at LLNL using a Perkin-Elmer 5100 series atomic absorption spectrophotometer. Flame mode was used
TABLE 3: Mean Mn and Cd Concentrations (N ) 3) and the Range of Mn and Cd Concentrations in the e1 mm Soil Fractions at the B850 and WOP Vale Collection Areas
area
Mn (mg/kg)
B850 340 ( 5 WOP vale 239 ( 23a
FIGURE 2. Total X-ray image (1-15 keV) from PIXE analysis of an olfactory bulb tissue section of a rat. The gray scale ranges from black (low) to white (high). The white bordered box corresponds to a central transect (1 mm wide) through the olfactory bulb to illustrate the orientation of the profiles. The tip of the olfactory bulb is toward the lower left. to determine Mn and high-temperature Zeeman corrected graphite furnace mode was used to determine Cd. Measured concentrations were multiplied by the final sample volume and divided by the sample mass used for digestion to calculate mg/kg concentrations. PIXE Analysis of Olfactory Bulbs. PIXE (30) is an X-ray fluorescence technique that can use focused MeV energy proton beams to interrogate specimens and has been previously used to study olfactory uptake of inhaled metals (29, 31). It provides accurate quantitation and simultaneous multielement detection and is capable of micron scale spatial resolution whilst maintaining down to 0.1 µg/g elemental sensitivity. Two tissue sections from each rat and ground squirrel olfactory bulb (two per animal) were analyzed by PIXE (32). A beam spot size of 10 µm were scanned over the olfactory bulb in each tissue section to produce images of X-ray distribution (29). From these images profiles of Mn and Cd concentration along the central transect through each olfactory bulb tissue section (see Figure 2) were determined (29). To improve X-ray counting statistics each data point in a profile corresponded to the concentration in a spatial bin of 0.2 × 1.0 mm2 within the transect. For each olfactory bulb, respective Mn and Cd profiles obtained from the two tissue sections were quantitatively similar and were averaged. Average Mn and Cd contents in the olfactory bulb portion of each tissue section were also calculated by analyzing the total X-ray spectrum produced from the PIXE analysis of the olfactory bulb as previously described (29). Respective Mn and Cd contents in the two tissue sections analyzed for each olfactory bulb were quantitatively similar and were averaged. Data Analysis and Presentation. When relevant, mean Mn and Cd concentrations and standard deviations in concentrations were calculated from the AAS and PIXE analyses. These are shown in Tables 3-6. Differences in the Mn and Cd contents of related samples were assessed by unpaired two-tailed students t-tests. A significance level of 0.05 when compared to B850 Mn leg muscle contents by Students t-test. d P < 0.05 when compared to B850 Cd leg muscle contents by Students t-test.
FIGURE 3. Average (mean ( standard deviation) Mn concentration profiles in rat olfactory bulbs. Dosed uptake corresponds to olfactory bulbs (N ) 7) from Mn exposed animals that showed measurable uptake of Mn while dosed no-uptake corresponds to bulbs (N ) 3) from Mn exposed animals that showed no measurable uptake of Mn when compared to controls. Control corresponds to olfactory bulbs (N ) 10) from vehicle controls. Regression analyses produced curves of the form log (Mn (mg/kg)) ) 1.45-0.30 × distance (mm), R 2 ) 0.8 for the dosed uptake data, log (Mn (mg/kg)) ) 0.45-0.15 × distance (mm), R 2 ) 0.8 for the dosed no-uptake data, and log (Mn (mg/kg)) ) 0.41-0.13 × distance (mm), R 2 ) 0.7 for the control data.
Results Table 3 shows the results of soil analyses, which are consistent with previous analyses of soils from the collection areas (30). The
