TOXICOLOGICAL ScrENCES 86(2), 239-247 (2005)doi: 10.1093/toxsci/kfi 196Advance Access publication May 11, 2005

f

Rapid Uptake, Metabolism, and Elimination of Inhaled SulfurylFluoride Fumigant by Rats

A. L. Mendrala,*,l D. A. Markham,* and D. L. Eisenbrandtt

*Toxicology & Environmental Research and Consulting, The Dow Chemical Company, Midland, Michigan 48674; tDow AgroSciences LLC,Indianapolis, Indiana 46268

Received March 17,2005; accepted May 9, 2005

n

).

,fdI.

c1-

{,

h

rrc

i.

J.

dd

~.

c

).

,.

Ie

IC

1-

Sulfuryl fluoride (S02F2) is a structural fumigant gas used tocontrol drywood termites and wood-boring beetles. The pharma-cokinetics and metabolism of inhaled S02F2 were evaluated)nmale Fischer-344 rats exposed to 30 or 300 ppm 35S-labeled~()2F2for 4 h. Blood, urine and feces were collected during and after theexposures and analyzed for radioactivity, 35S-labeled flllOrosulfateand sulfate, and fluoride (urine and feces only). Selected tissueswere collected 7 days post-exposure and analyzed for radioactiv-ity.During and after unlabeled S02F2 exposures, blood, brain, andkidney were collected and analyzed for fluoride ion. S02F2 wasrapidly absorbed, achieving maximum concentrations of radioac-tivity in both plasma and red blood cells (RBC) near the end of the4-h exposure period. Radioactivity was rapidly excreted, mostlyvia the urine. Seven days post-exposure, small amounts of radio-activity were distributed among several tissues, with the highestconcentration detected in respiratory tissues. Radioactivity asso-ciated with the RBC remained elevated 7 days post-exposure, andhighly perfused tissues had higher levels of radioactivity thanother non-respiratory tissues. Radioactivity cleared from plasmaand RBC with initial half-lives of 2.5 h after 30 ppm and 1-2.5 hafter 300 ppm exposures. The terminal half-life of radioactivitywas 2.5-fold longer in RBC than plasma. Based on the radio-chemical profiles, there was no evidence of parent 35S02F2 inblood. Identification of fluorosulfate and sulfate in blood and urinesuggests that S02F2 is hydrolyzed to fluorosulfate, with release offluoride, followed by further hydrolysis to sulfate and release ofthe remaining fluoride.

Key Words: sulfuryl fluoride; fluorosulfate; fluoride; inhalation;rat.

INTRODUCTION

Sulfuryl fluoride (S02F2, VIKANE* gas fumigant) is a color-less, odorless gas that has been used to fumigate buildings

The studies reported irr this article were presented at the 42rrd annual meetingof the Society of Toxicology, March 2003, Salt Lake City, UT (The Toxicologist72,713, March 2003).

] To whom conespondence should be addressed at Toxicology & Environmen-tal Research and Consulting, 1803 Building, The Dow Chemical Company,Midland, Michigan48674. Fax: (989) 638-9863. E-mail: ALMendrala@dow.com.

* VIKANE is a trademark of the Dow AcroSciences LLC.

since 1961 to control pests, inclllding drywood termites andw()()d-boring beetles. The American Conference of Govern-n'lental Industrial Hygienists (ACGIH) threshold limit value andOSHA permissible exposul-~ ..level for S02F2 are 5 ppm TWAand 10 ppmSTEL (ACYItf,/1998). Extensive toxicologicalstudies have been conducted \',lith S02F2 over the past 30 years,and the toxicity is well charactel'ized (re;riewed in Nitschke andEisenbrandt, 2001). S02F2 is acutely toxic to mammals wheninhaled, with a LCso (4 h) for rats of about 1000 ppm. Ratsexposed to S02F2 at concentrations of 1500 ppm or 2000 ppmshowed central nervous system depression (sedation) within 20to 40 min, and continued exposure resulted in convulsions andadverse effects on the respiratory tract, liver, and kidneys,followed by death within 3-4 h (unpublished data of TheDow Chemical Company). Exposure of rats to 4000, 10,000,20,000, or 40,000 ppm S02F2 decreased the time to incapaci-tation and death as the vapor concentration increased (Nitschkeet at., 1986),

Fluoride ion may playa role in the mechanism of action ofS02F2 in insects and mammals. Many of the observations inrodents overexposed to S02F2 are typical of acute fluoridepoisoning (Drill, 1954; Goodman et at., 1980; GreemvC)od,1940; Whitford, 1996). Eisenbrandt and Nitschke (1989)studied the toxicity of S02F2 in rats and rabbits exposed to 0,30, 100, or 300 ppm for 6 h/day, 5 days/week for 13 weeks. Astatistically significant and dose-related increase in ionicfluoride concentration was reported in the serum of rabbitsexposed to S02F2 for 13 weeks, although serum levels of ionicfluoride were not significantly elevated in rats. Inhalationexposure of rats to 300 ppm S02F2 resulted in decreased bodyweights; histopathological changes in the respiratory tract,brain, and kidneys; as well as mottled teeth and decreasedspecific gravity of the urine. At 100 ppm, the tooth enamelchanges were the only exposure-related changes in the rats. Notreatment-related effects were observed at 30 ppm. Neuropath-ological changes in the rats consisted of minimal vacuolation inthe area of the caudate-putamen nuclei that was moreprominent in the white fiber tracts of the internal capsule thanin the adjacent neuropil. Alteration of neurological function inthe rats exposed to 300 ppm S02F2 for 13 weeks was

© The Author 2005. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved.For Permissions, please ernail: joumals.permissions@oupjournals.org

mailto:ALMendrala@dow.com.mailto:joumals.permissions@oupjournals.org

240 MENDRALA, MARKHAM, AND EISENBRANDT

demonstrated by a decreased flicker fusion threshold, anda slowing of flash, auditory, and somatosensory evokedpotentials (Mattsson et ai., 1988). The functional responsesas well as brain histology were within normal limits at 2 monthspost -exposure.

Mottled incisor teeth in rats exposed to 100 or 300 ppmS02F2 for 13 weeks were consistent with an increase insystemic fluoride as produced by several other inorganic andorganic fluoride compounds (Eagers, 1969; Pattison, 1959).The current metabolism and disposition study was undertakento better understand the mammalian toxicity of S02F2 and todetermine the systemic availability of fluoride following S02F2exposure.

MATERIALS AND METHODS

Test material. Sulfuryl fluoride, radiolabeled with 35sulfur, was obtainedfrom Moravek Biochemicals, Inc., Brea, California with a specific activity of10 mCi/mmo!. A radiochemical purity of 100% 35S02F2 was established usinggas chromatography with flow radiogas monitoring and identification byretention time confirmation using an authentic standard of S02F2 and thermalconductivity detection. Structural confirmation was accomplished by gas chro-matography/electron impact ionization/mass spectrometry. Non-radiolabeledS02F2 was obtained from Dow AgroSciences LLC (Indianapolis, IN). Analysisof this test material prior to the study revealed the sample to be 99.8% S02F2'All other chemicals and solvents were reagent grade or better.

Animals. Male Fischer-344 rats (197-242 g) were used because they werethe same strain used in previous inhalation toxicity studies of SOlF2(Eisenbrandt and Nitschke, 1989; Mattsson et al., 1988; Nitschke et al.,1986). Upon arrival at the laboratory, which is fully accredited by theAssociation for Assessment and Accreditation of Laboratory Animal CareInternational (AAALAC International), each animal was evaluated by a labo-ratory veterinarian and found to be in good health and acceptable for studypurposes. The rooms in which the animals were housed had the relativehumidity maintained within a range of 40-70% and the room temperaturemaintained at 22° ± 3°C. An approximate 12 h light/dark photocycle wasmaintained for all animal rooms. Room air was exchanged 12-15 times/h. Oneach of 2 days prior to exposure, the animals were stage-acclimated to the nose-only chamber (1 h on day 1, 2 h on day 2), with the nose cones connected to theexposure chamber and airflow comparable to what it was during exposure.

Rats cannulated in the jugular vein by the supplier (Hilltop Lab Animals,Inc. Scottsdale, PA) were used for the 35S time-course experiments. To increasethe likelihood that the cannulae remained patent, acclimation of the rats to thelaboratory environment was 1imited to 4 days prior to 35S02F2 exposure. Theywere returned to their cages for -24 h before 35S02F2 exposure, and during thisperiod the rats underwent exteriorization of the cannulae under light isofluraneanesthesia, approximately 18 h prior to 35S02F2 exposure.

Non-cannulated rats used for the remainder of the study were obtained fromCharles River Laboratories (Raleigh, NC). The animals used for 35S dispositionwere acclimated to metabolism cages for at least 2 days. They were returned tothe metabolism cages for -24 h before 35S02F2 exposure. Non-cannulatedanimals used for fluoride analysis were housed one per cage in stainless-steelcages. Non-cannulated animals were selected from those available and assignedto exposure groups using a computer-driven randomization procedure.

Following exposure to 35S02F2, non-cannulated animals were housed oneper cage in glass metabolism cages. Air was drawn through the metabolismcages at -500 ml/min. Cannulated animals were housed in plastic tubs withcorncob bedding for the duration of blood collection. Animals exposed tonon-radiolabeled S02F2 were housed one per cage in stainless-steel cages

containing a feed container and a pressure activated nipple-type wateringsystem. Cages were suspended above catch pans, and feed and water wereprovided ad libirwn except during the exposures. Animals were providedLabDiet Certified Rodent Diet #5002 (PMI Nutrition International, St. Louis,MO) in pelleted form. To minimize the animals' exposure to environmentalfluoride, the animals were given Milli-Q de-ionized water (Millipore Co.,Bedford, MA) after their alTival at the laboratory. No attempt was made toremovelreduce fluoride ion from nonaqueous sources, i.e., feed (13 ppmfluoride ion, Nutritional Analyses of Certified Rodent Diet #5002 by PMINutrition International).

Exposure system. The 35S01Fl exposures were conducted in a 2-1iterPlexiglas nose-only exposure chamber, stationed in a closed, vented, Plexiglasbox, which was located inside a secondary vented area. The rats were held inplace by use of a tubular restraint device constructed from #3 mesh galvanizedhardware cloth (7.26 mm openings) attached to a polycarbonate nose cone. Anexcreta separator-collector was attached to the bottom of the restraint device,and the collectors were placed in dry-ice. Chambers used for the non-radiolabeled exposures consisted of a 42-liter Dow-modified ADG nose-onlychamber (ADG Developments, Ltd., UK) stationed in a secondary vented area.All exposure chamber air was exhausted through a series of charcoal traps. Thechamber atmosphere was not recycled or returned to the chamber. Sham-exposed control animals were placed in a Dow-modified ADG nose-onlychamber of similar design and were provided air of the same qnality as exposedrats but not containing S02F2'

Atmosphere preparation. The 35S02F2 was received from the supplier inflame-sealed glass flasks with two side-arms and ground glass fittings. To dilutethe radiolabeled material with unlabeled SOzFz, as-liter Ted1ar bag (SKC, Inc.,Eighty Four, PA) containing unlabeled S02F2 was connected to one arm of theflask with a 20 gauge X 30 em hypodermic needle that reached through a break-seal to the bottom of the flask. A 1.5-liter gas-tight syringe was connected to theother arm of the flask. The plunger on the syringe was pulled to create a vacuumin the syringe, drawing the unlabeled S02F2 through the flask, mixing with the3SS02F2, and into the syringe. The entire contents of the syringe were thentrans felTed to an empty Tedlar bag that was clamped, sealed, and detached fromthe syringe. This bag, capped with a rubber septum to allow sampling witha syringe or needle attached to a pump, was the stock gas bag used to generatethe desired chamber concentration. Liquid scintillation spectrometry (LSS) ofseveral samples of the 3sS-labeled atmosphere taken from the stock gas bag wasused to confirm the concentration of radioactivity and the homogeneity of the3SS02Fl in the atmosphere.

Chamber concentrations of 30 ppm and 300 ppm were generated by meteringat a control1ed rate from the stock gas bag into the chamber air stream, where thegas was mixed and diluted to the desired concentration before the air streamentered the chamber. Exposures were extended the appropriate length of time toaccount for chamber equilibration time (9 min for the 2-liter chamber and 7 minfor the 42-liter chamber). The chamber concentrations were monitored with anAgilent Gas Chromatograph 6890A with electron capture detection. The airsample was drawn from the breathing zone of the chamber through a closed gas-sampling loop of 0.53 mm deactivated fused silica capillary line in 300 IIIaliquots. Chamber concentrations were measured at least three times duringeach exposure, in duplicate. The area counts were determined using integrationby TurboChrom (PE Nelson, PerkinElmer, Inc, San Jose, CA). A standard curvewas generated using known S02Fl gas standards prior to and after eachexposure period. The average integrated area count at each sampling time wasdetermined, the standard curve was applied, and the interpolated values wereused to calculate a time-weighted average (TWA) S02FZ chamber concentra-tion. Non-radiolabled exposures were conducted in a similar manner, except thestock gas bag contained only non-radiolabeled SOZF1'

Blood collectioll. The jugular vein-cannulated animals were used onlyduring the radiolabeled portion of the study for blood collection during andafter exposure. All other specimens were collected from non-cannulatedanimals. Certain tissues, noted below, and feces were solubilized with Soluene350 tissue solubilizer (PerkinElmer, Inc.) according to the methods described

SULFURYL FLUORIDE METABOLISM 241

by the manufacturer. For the radio labeled portion of the study, four jugularvein-cannulated rats and four non-cannulated rats were exposed to each35S02F2 concentration. Venous blood samples (-0.15 ml) were collected via theindwelling jugular cannulae at 0.25, 0.5, 1, 2, 3, and 4 h during the inhalationexposure and at 0.5, 1,2,4,6, 12,24,36,48,72,96, 120, 144, and 168 h post-exposure. Blood sampling times are designated in the text and figures as - 3. 75,-3.5, ·-3, -2, ..-I, and 0 h for the 0.25, 0.5, 1,2,3, and 4 h during-exposuresamples, respectively. Blood was obtained at sacrifice via cardiac punCture. Theblood was centrifuged to separate plasma and red blood cells (RBC). Plasmaand RBC were analyzed for radioactivity from which plasma and RBC 35Sconcentration-time courses were constructed. Aliquots of plasma were counteddirectly by LSS while RBC were solubilized and then analyzed by LSS. Inaddition, approximately equal-volume aliquots of blood were pooled (persample time) and stored at -80°C until chemical analysis.

During the non-radiolabeled portion of the study, groups of three rats weresacrificed at -2 h (during exposure), at 0 h (end of exposure), and at 2, 4,8, and20 h post-exposure. Sham-exposed control animals in groups of two weresacrificed at -4, 0, 4, 8 and 20 h. Blood was obtained at sacrifice via cardiacpuncture. The blood was centrifuged to separate plasma and RBC, and theplasma was analyzed by fluoride ion-specific probe (see below) for fluoride ioncontent, and a fluoride ion concentration-time course was constructed. Plasmawas not pooled for fluoride analysis.

Urine specimens. For non-cannulated rats exposed to 35S02F2' all urinevoided during the study was collected in dry-ice-cooled traps. The urine trapswere changed at the end of exposure (0 h) and at 6, 12,24,48,72,96, 120, 144,and 168 h post-exposure. Each urine specimen and rinse was analyzed forradioactivity by LSS as described below. Equal volume aliquots of urinesamples from the 0-6 hand 6-12 h post-exposure collection intervals werepooled and stored at -80°C prior to radiochemical profile analysis.

Feces specimens. For non-cannulated rats exposed to 35S02F2, all feceswere collected in dry-ice-chilled containers through the exposure period (0 h)and at 24, 48, 72, 96,120, 144, and 168 h post-exposure. Aqueous homogenates(-25% w/w) were prepared, and weighed aliquots of these homogenates weresolubilized and quantitated for radioactivity by LSS. In addition, equal volumealiquots of fecal homogenates were taken from the 0-24 h post-exposurecollection interval and pooled. These pooled samples were stored at -80GCprior to radiochemical profile analysis.

Expired air. Air was drawn through the metabolism cage at approximately500 m1lmin. Where expired air exited the cage, volatiles in it were trapped oncharcoal by passage of the expired air through a charcoal trap. The charcoal wasthen desorbed with toluene and aliquots were analyzed for radioactivity byLSS. No radioactivity was detected in the 0-24 h post-dosing charcoal trapsafter the 300 ppm exposure. Therefore, expired air was not passed throughcharcoal for the remainder of the 300 ppm sample collection intervals nor forany of the sample collection intervals for the 30 ppm exposure.

Terminal sacrifice. At 168 h post-exposure, the animals were anesthetizedwith CO2 and killed by exsanguination via cardiac puncture. The metabolismcages were then washed, and the cage wash was analyzed for radioactivity.

Tissues. The following tissues were collected at sacrifice: terminal blood,brain, gastrointestinal (GI) tract, kidney, liver, lung, perirenal fat, residualcarcass, respiratory turbinate, olfactory turbinate, skin, and spleen. Therespiratory and olfactory turbinates were surgically separated and removedduring dissection. The brain, GI tract plus contents, kidney, liver, lung, carcass,and spleen were collected, homogenized individually (-33% homogenate), anda weighed aliquot was solubilized and analyzed for radioactivity by LSS. Bloodwas obtained at sacrifice via cardiac puncture, and an aliquot was solubilizedand analyzed for radioactivity by LSS. The skin was removed from the carcassand a representative skin sample was solubilized and analyzed for radioactivityby LSS. Perirenal fat and the nasal turbinates, both olfactory and respiratory,were directly solubilized without homogenization.

For the non-radiolabeled fluoride ion portion of the study, aqueoushomogenates of brain and kidney from groups of three rats sacrificed at ·..2 h

(during exposure), 0 h (end of exposure), 2 h, and 4 h post-exposure wereanalyzed for fluoride ion by !Juoride ion-specific probe as described below.

35S-Analysis. Radioactivity was quantified in a liqnid scintillation spec-trometer (Beckman LS-180!, LS-3801, or LS-6000 or Packard Tri Carb2900TR). Samples were counted within 24 h after collection; therefore nocorrection for radioactive decay was performed (half-life 35S = 88 days).

Chemical analysis. Whole blood samples submitted for radiochemicalprofiling were collected directly into tared glass vials containing 0.5 ml ofacetonitrile extraction solvent. These vials were sealed, vortex mixed -30 s, andcentrifuged at 3000 rpm for IO min, after which the liquid was transfen-ed to] .5-ml autosampler vials for high pressure liquid chromatography/radioactivitymonitor (HPLC/RAM) analysis. Because of the low levels of radioactivity inthe samples, 20-s fractions were collected and assayed by LSS to improve themethod detection limit.

Urine samples for radiochemical analysis were selected based upon theamount of radioactivity present in the samples. Pooled urine samples from the-4 to 0 h exposure period interval and the 0 to 6 h, 6 to 12 h, and 12 to 24 h(300 ppm only) post-exposure intervals were analyzed. The urine samples werepooled by combining equal 250 "tl aliquots from each of the four animals percollection interval. The vials were centrifuged at 3000 rpm for IS min, and theliquid was transferred to 1.5-ml autos ampler vials for HPLC/RAM analysis.Because of the low levels of radioactivity in the samples, fractions werecollected and assayed by LSS to improve the method detection limit.

The LSS data were used to generate reconstructed radiochromatograms, andthese radiochromatograms were integrated to determine peak area response ofthe radioiabeled peaks. The concentrations of sulfate and fluorosulfate in thesesamples were estimated from the calculated specific activity of sulfate andfluorosulfate and the total radioactivity in the radiolabeled peaks. Theconcentration of fluoride ion in the urine of rats exposed to both concentrationsof 35S01F2 was determined by ion selective electrode (ISE; Orion Research,Cambridge, MA). Also, ISE determination of the fluoride ion concentration inthe plasma, as well as the brain and kidney tissue homogenates, of rats exposedto both concentrations of non-radio labeled S02F2 was conducted.

Data analysis. Radioactivity data are presented as mean ± SD of fouranimals, and ISE tissue fluoride data are presented as mean ± SD of threeanimals. All other data are single measurements of samples pooled perexposure concentration and time. Radioactivity was converted to S02F2concentration by dividing the measured radioactivity value by the specificactivity of the S02F2 gas in the chamber, and the results are expressed asmicrogram or micro mole equivalents (Ilg-Eq or "lmol-Eq) S02F2. Pharmaco-kinetic analyses were performed on individual plasma and RBC concentration-time data to calculate the area under the concentration-time curves (AUC) andhalf-life of elimination of 35S using PK Solutions phannacokinetic modelingprogram (Summit Research Services, Montrose, CO). Fluoride levels in theplasma, brain, and kidney were evaluated by analysis of variance (ANOVA).The analyses were can-ied out at alpha = 0.05. Plasma fluoride statisticalanalysis was conducted with 0, 4, 8, and 20 h samples. Brain and kidneyfluoride statistical analysis was conducted with 0 and 4 h samples.

RESULTS

Sulfuryl Fluoride Exposure

The nominal exposure concentration of 300 ppm 3SS02F2resulted in an actual time-weighted average (TWA) of274 ppmthrough the 4-h exposure period and a 3SS concentration ofapproximately 2.8 /lCi/l of atmosphere (specific activity0.25 mCi/mmol). The nominal non-radio labeled 300 ppmS02F2 exposure resulted in a TWA of 312 ppm. The nominalexposure concentration of 30 ppm 3SS02F2 resulted in an actual

A /', /', /', /',ojU~ A10 A :;j 10ojp:; A-'?"-'?" A A ~'~ A 06' i:! 0\: A Vi iiVi 0

0 0 00"' 06' J A '" 0'" A DnDo ! =L 0=L I •••• HIHI

HI HI

0.1i0 0.1 io24 48 72 96 120 144 168 24 48 72 96 120 144 168Exposure period Time (h) Exposure period Time (h)

242

100

MENDRALA, MARKHAM, AND EISENBRANDT

100

FIG. 1. Plasma concentration-time course oesS02Frderived radioactivityduring and after exposure of rats to atmospheres containing 30 (l1li) or 300 (AI.)ppm 35S02F2. Data are mean ± SO (n = 4).

TWA of 28.4 ppm through the 4-h exposure period and a 35Sconcentration of approximately 0.26 flCi/l of atmosphere(specific activity 0.22 mCi/mmol). The nominal non-radio labeled30 ppm S02F2 exposure resulted in a TWA of 31.2 ppm.S02F2 was not detected in any of the analyses of the controlchamber atmosphere at a level exceeding the lowest levelquantified [-1 ppm].

All animals survived the 4-h S02F2 exposure. Those animalsexposed to 30 or 300 ppm 35S02F2 were maintained for 1 weekpost-exposure prior to sacrifice. Animals exposed to non-radio labeled S02F2 were sacrificed at the specified times pre-,during, and post-exposure.

Plasma Radioactivity

Radioactivity derived from 35S02F2 achieved quantifiableplasma levels by 15 to 30 min after initiation of the exposure(Fig. 1). Plasma levels of radioactivity continuously increasedthroughout the exposure period, ultimately achieving peakmean plasma concentrations of 5.2 and 38 flg-Eq S02F2/g atthe end of the 30 and 300 ppm exposures, respectively. Aftertermination of 35S02F2 exposure the plasma radioactivitydecreased rapidly during an initial Ci phase with a half-life of-2.5 h for both exposure concentrations. A secondary ~ phasebeginning about 24 h post-exposure was observed with a half-lives of -83 and -56 h for the 30 and 300 ppm exposures,respectively. The plasma AUCc_4 h to 00) for the 30 and 300 ppmexposures were calculated as 96.7 and 756 ~Lg-h ml-I.

RBC Radioactivity

The kinetics of RBC-associated radioactivity were similar tothat of plasma radioactivity in uptake and initial distributionphases, but a longer terminal half-life for the elimination ofRBC radioactivity was observed (Fig. 2). Peak RBC radioac-tivity was measured at the end of the 4-h exposure period forboth 30 and 300 ppm 35S02Fl exposures. Immediately

FIG. 2. Red blood cell concentration-time course of 35S02Frderivedradioactivity during and after exposure of rats to atmospheres containing 30(D) or 300 (6) ppm 35S02F2' Data are mean ± SO (/1 = 4).

following the 35S02Fl exposures, mean peak RBC radioactiv-ity reached 4.7 and 40.3 flg-Eq S02Fl/g RBC for the 30 and300 ppm 35S0lF2 exposure concentrations, respectively. Afterthe peak RBC radioactivity levels at termination of exposure,an initial Ci elimination phase half-life of 2.5 and 1.1 h wascalculated for the 30 and 300 ppm exposure levels, respec-tively. After the 30 ppm exposure the ~ phase was estimated tobe 222 h and following the 300 ppm exposure the ~ phase wasestimated at 139 h. These longer terminal half-lives may bea result of either (1) nonspecific incorporation of the 35Sradiolabel as the radiolabel becomes available through normalsulfate pool metabolism, transformed into amino acids, andincorporated into tissues or (2) a result of binding of someunknown 35S-containing component to RBC. Radioactivityremained at measurable, albeit small, concentrations until thetime of terminal sacrifice, 168 h post-exposure. The AUCc-4 h_oo)values for RBC, 863 and 5492 flg-h ml-1 for 30 and 300 ppmexposure concentrations, respectively, are 9-fold to 7-foldlarger than the AUC obtained for plasma.

Excretion

Once absorbed, the 35S02Frderived radioactivity rapidlyappeared in the urine and feces (Fig. 3). Urine contained 88.9%and 85.6% of the total excreted radioactivity through 7 dayspost-exposure for the 30 and 300 ppm S02Fl exposures,respectively. The urine samples excreted during the exposureperiod contained 273 and 2766 flg-Eq S02Fl, or 42% and 51 %of the total urinary radioactivity, for the 30 and 300 ppm S02F2exposures, respectively. The 0 to 6 h interval urine samplescollected immediately following the end of the exposurescontained 167 and 936 flg-Eq S02Fl, or an additional 26% and17% of the total urinary radioactivity, for the 30 and 300 ppmS02F2 exposures, respectively. Initial urinary elimination ratesof 68 and 691 flg-Eq h-1 for 30 and 300 ppm exposures,respectively, rapidly decreased through subsequent collection

SULFURYL FLUORIDE METABOLISM

FIG. 3. Excretion of 35S02FTderived radioactivity during and after expo-sure of rats to atmospheres containing 30 (ft urine; + feces) or 300 (0 urine;o feces) ppm 35S02F2. Data are mean ± SD (11 = 4).

243

to9

o

4-3-4

10

0.1

0.01if>

-'"oE::L 0.001

-2 -1 0

Collection Time (h)

FIG. 6. Whole blood levels of sulfate (~,O), fluorosulfate (+ ,0), andplasma fluoride ion (l1li,0) during and after exposure of rats to atmospherescontaining 30 (closed symbols) or 300 (open symbols) ppm 35S02F2. Plasmafluoride ion in control plasma is indicated by x. Sulfate and fluorosulfate dataare single measurements of blood pooled from four animals per exposureconcentration and time. Fluoride ion data are mean ± SO (n = 3).

OJ

~'" oj.5 "-~ (;if; "D

00

~ ::0u

" Eoj"D i3"" 0...D.~

0.5

245

6

h_

n

+ ---+

__________ nn ----~-------------- X

o 2 4Collection Time (h)

-4 -2Exposure period

0.1

0.0

0.4'"~~Dn 0.3...'"0..

r:;:;en 0.2~o

~

the time of peak levels of fluoride. Figure 6 presents the bloodsulfate plus fluorosulfate data and the plasma fluoride datatogether in one figure for each exposure concentration. Theoverall shapes of the curves during and after the S02F2exposures are quite similar, indicating that the kinetics offormation and elimination from plasma of these three metab-olites are similar or are interrelated.

Fluoride ion analysis of urine from non-exposed rats and fromrats exposed to S02F2 was conducted with selected urinesamples. Figure 7 presents the urine sulfate plus fluorosulfatedata and the fluoride data together in one figure for each exposureconcentration. As observed with plasma, the overall shapes ofthe curves during and after the S02F2 exposures are quitesimilar, indicating that the kinetics of formation and eliminationof these three metabolites are similar or are interrelated.

Elevated levels of fluoride ion were detected in urine duringand after the S02F2 exposures. Non-exposed control rats hadconcentrations of fluoride ion of approximately 0.12-0.13~unol/ml urine. By the end of the 4-h 30 ppm exposure, theurine levels of fluoride ion reached a maximum concentration of-O.S ~lmol/ml urine. This concentration was maintained through6 h post-exposure. By 12 h post-exposure the concentration haddiminished to near background levels, 0.14 f--lmollmlurine. Ina similar fashion, by the end of the 4-h 300 ppill exposure theurine levels of fluoride ion reached a maximum concentration of4.0 f--lmol/mlurine, about 8-fold higher than that obtained afterthe 30 ppm exposure. However, this concentration diminishedthrough 6 h post-exposure to 1.7 ~lmollml urine and by 24 hpost-exposure to -0.3 ~lmollml urine.

Fluoride levels in brain and kidney tissue during and afterexposure to 30 and 300 ppm S02F2 are presented in Figure 8.Kidney fluoride levels were roughly 2- to 2.5-fold higher at allcollection times than those measured in control rats. Controlrats had mean fluoride levels of 0.12 f--lillol/gkidney tissue,

FIG. 8. Brain (8,0) and kidney (II1II,0) concentration-time course offluoride ion during and after exposure to atmospheres containing 30 (closedsymbols) or 300 (open symbols) ppm S02F2' Control tissues are indicated by +(brain) and x (kidney). Data are mean ± SO (11 = 3) except for control tissues,for which the data are the average of two animals.

SULFURYL FLUORIDE METABOLISM

Fluoride Analysis

The ISE technique used in this study measures free fluorideonly. Three types of fluoride make up the "total" fluoride levelin tissues, ionic (free), acid labile, and nonionic (bound) (fora review, see Singer and Ophaug, 1982). Acid labile fluorideaccounts for a very small amount of the total fluoride (0.01 ppm).However, the contribution of nonionic (bound) fluoride to thetotal fluoride level can be as high as 50% of the total in plasma.The determination of nonionic fluoride involves a very labor-intensive combustion method that provides highly variableresults (ibid.), and therefore was not included in this study.

Ion selective electrode analysis of plasma and tissues wasconducted with rats exposed to non-radiolabeled S02F2 andwith control rats exposed to clean air. Levels of plasma fluorideion in control animals ranged from 0.023 to 0.070 ,..unol/mlplasma through a 24-h cycle. These measurements are withinthe control range of serum fluoride concentration reported byEisenbrandt and Nitschke (1989).

An apparent slight elevation of plasma fluoride from controllevels was observed during the 30 ppm and 300 ppm exposures,achieving concentrations of 0.046 and 0.132 f--lmollmlplasma,respectively. But, plasma fluoride levels rapidly returned tocontrol levels (-0.03-0.05 f--lmol/mlplasma) by about 2 h afterexposure was terminated (Fig. 6). Plasma fluoride was 1.6- andSA-fold higher than control levels at the end of exposure for the30 and 300 ppm S02F2 exposures, respectively. The maximumconcentration of plasma fluoride measured at the termination ofthe 4-h 30 ppm S02F2 exposure was 0.04 ~lmollml and issimilar to that reported by Eisenbrandt and Nitschke (1989)following 6-h/day, 5 days/week for 13 weeks exposure to30 ppm S02F2, although considerable variation was reported.The plasma concentration reported here after the 300 ppmexposure was 0.13 f--lmollmland is nearly twice as large as thatreported by Eisenbrandt and Nitschke (1989). However, ourdata indicate a rapid clearance of fluoride from the plasma anda return to background levels within -2 h following terminationof exposure. Because serum was collected from the rats in the13-week study at necropsy the day after the final exposure,measurement of serum fluoride was conducted subsequent to

fluorosulfate. The total ~unoles of sulfate plus fluorosulfaterecovered in the urine as determined by HPLC/RAM (3.8 and43 flIl1oles; 30 and 300 ppm, respectively) compares well withthe ~lmol-Eq S02F2 in the urine + rinse as determined from theradiolabeled portion of this study (5.7 and 45 ~lInoles; 30 and300 ppm, respectively).

Conversion of sulfate and fluorosulfate urine concentrationsto rate estimatcs allowed calculation of half-lives for theelimination of sulfate and fluorosulfate in urine. Sulfate waseliminated in the urine with a half-life of 2.2 and 3.8 h for the30 and 300 ppm exposure groups, respectively. Fluorosulfatewas eliminated slightly faster, with a half-life estimate of 1.2and 2.4 h for the 30 and 300 ppm exposure groups, respectively.

MENDRALA, MARKHAM, AND EISENBRANDT2:46

while both 30 and 300 ppm exposure levels resulted in meanfluoride concentrations of about 0.26 jlmol/g kidney tissue.These levels were measured by the second hour of exposureand were maintained through 4 h post-exposure.

A slight 1.5-fold elevation in fluoride levels in brain tissuerelative to control rats was observed during and after exposureto 30 ppm S02F2' The mean control brain tissue fluoride ionconcentrations were roughly 0.03 ~Llnol/g tissue at all sacrificetimes. Fluoride levels increased to 0.04 ~lmollg brain after the30 ppm S02F2 exposure, whereas after the 300 ppm S02F2exposure, concentrations as high as 0.12 ~Llnollg tissue weremeasured, -4-fold higher than controls, at the end of theexposure period. The concentrations of fluoride in brain tissueapproached the control levels by 4 h post-exposure.

DISCUSSION

Radiolabeled S02F2 was rapidly absorbed via inhalationexposure, achieving maximum concentrations of 3SS02Frderived radioactivity in both plasma and red blood cells nearthe end of the 4-h exposure periods. Saturation of absorptionwas not achieved during these 4-h exposures as evidenced bythe following findings: (1) the measured levels of radioactivitywere roughly proportional to the exposure concentrations, (2)the level of plasma/RBC radioactivity was increasing at the endof exposure, (3) steady-state 3SS conditions were not achieved,and (4) the initial half-lives for clearance of radioactivity fromplasma and RBC were the same at both exposure concen-trations. Plasma levels of radioactivity during the 4-h exposurenever reached steady-state conditions, most likely because ofrapid elimination of radioactivity from the plasma. The similarpeak concentrations of radioactivity measured with plasma andRBC suggests the 3SS initially was evenly distributed betweenthese two compartments. Once absorbed, the 3SS was rapidlyexcreted, primarily via the urine. A large portion of absorbed3SS was excreted in the urine, even during the 4-h nose-onlyexposure.

Seven days post-exposure, small amounts of the radio labelwere evenly distributed among the tissues, suggesting in-corporation into the tissues of 3SS that had been transformedinto amino acids; however, the chemical composition was notdetermined. Radioactivity was recovered mainly in respiratorytissues, which are the site of first exposure to the gas. The lungshad the highest levels of radioactivity 7 days post-exposure,and the nasal turbinates also had detectable radioactivity.Radioactivity associated with the RBC remained elevated 7 dayspost-exposure. Highly perfused tissues such as spleen andkidney also had higher levels of radioactivity than non-respiratory tissues, a finding attributed to the radioactivity inthe blood. Radioactivity was rapidly cleared from plasma andRBC with initial half-lives of -2.5 h after the 30 ppm exposureand 1-2.5 h after the 300 ppm exposure. However, the terminalhalf-life of radioactivity was -2.S-fold longer in RBC than in

plasma. The identification of ftuorosulfate and sulfate in bloodand urine suggests that S02F2 is first hydrolyzed to fluorosul-fate, with release of fluoride, followed by further hydrolysis tosulfate and release of the remaining fluoride. This apparentmetabolism is supported by the increases in fluoride detected inthe blood and urine following exposure of rats to S02F2'

Fluoride metabolism resulting from inhalation exposure toS02F2 would be expected to be consistent with the metabolismof fluoride from other sources. Whitford (1996) providesa comprehensive review of the metabolism of fluoride. Onceabsorbed, plasma functions as the "central compartment" forfluoride. Fluoride rapidly establishes a steady-state distributionbetween extracellular and intracellular fluids, and thus fluoridelevels in these two compartments parallel each other.

The major route of fluoride elimination from the body isexcretion in the urine. The kidneys are very efficient inremoving fluoride from the body. The concentration of ionicfluoride in the glomerular filtrate is very similar to that ofplasma, and there is no evidence for the tubular secretion offluoride, although 62-78% of the filtered fluoride is reabsorbed.Approximately 50% of an absorbed amount of fluoride isexcreted in the urine within 24 h, whereas most of theremainder becomes associated with calcified tissue.

Fluoride associated with calcified tissues is not irreversiblybound, and this is especially true of recently acquired fluorideon the surfaces of bone crystallites. If the intake of fluoridewere to increase or decrease on a chronic basis, the concen-trations in the calcified tissues would eventually reflect thechange. Fluoride accumulation by bone is inversely related toage. The ability of the developing skeleton to clear fluoridefrom plasma more rapidly than the mature skeleton is largelya function of surface area (mature bone is compact, whereas thecrystallites of developing bone provide extensive surface arefor reactions involving fluoride). However, data in ratsindicate that the uptake of fluoride by the skeleton continuesthroughout the first 18 months of life (75-85% of life span)and accounts for one half the plasma clearance during thisperiod.

A likely cause of S02F2 toxicity is the metabolic release offluoride ions as postulated by Nitschke, et al. (1986) andNitschke and Eisenbrandt (2001). The data presented heresupport the hypothesis that S02F2 toxicity is the result ofmetabolic release of fluoride ions. Inhaled S02F2 is rapidlyabsorbed and hydrolyzed to fluorosulfate and ionic fluoride,followed by further hydrolysis to sulfate and an additionalfluoride ion. Therefore, these data suggest that the toxicityelicited by S02F2 may be due to the release of fluoride ions,rather than a direct toxic action of S02F2'

ACKNOWLEDGMENTS

The authors thank 1'. Card, A. J. Clark, J. Y. Domoradzki, M. J. Filary, J. A.Hotchkiss, C. E. Houtman, M. A. KnoelT, S. M. Krieger, G. 1'. Marty, L. G.

SULFURYL FLUORIDE METABOLISM

McFadden, E. S. Mullen, D. L. Rick, S. Saghir, and C. M. Thornton for theirexcellent technical assistance. Conflict of interest: The Dow ChemicalCompany, Dow AgroSciences LLC.

REFERENCES

ACGIH. (American Conference of Governmental Industrial Hygienists). (1998)."TLVs and other Occupational Exposure Values-1998." Documentation ofthe Threshold Limit Values and Biological Exposure Indices, AmericanConference of Governmental Industrial Hygienists, Cincinnati, OR

Cady, G. H., and Misra, S. (1974). Hydrolysis of sulfuryl fluoride. Inorg. Chem.13, 837-84l

Drill, V. A. (1954). Pharmacology in Medicine. McGraw-Hill, New York.

Goodman, A. G., Goodman, L. S., and Gilman, A. (1980). The Phannacolog-ical Basis of Therapeutics, 6th edition. Macmillan, New York.

Greenwood, D. A. (1940). Fluoride intoxication. Physio!. Rev. 20,582-616.

Eagcrs, R. Y. (1969). Toxic Properties of Inorganic Fluorine Compounds.Elsevier Publishing Company Ltd., New York.

247

Eisenbrandt, D. L., and Nitschke, K. D. (1989). Inhalation toxicity of sulfurylfluoride in rats and rabbits. Fundam. App!. Toxico!. 12, 540-557.

Mattsson, J. L., Albee, R. R., Eisenbrandt, D. L., and Chang, L. W. (l988).Subchronic neurotoxicity in rats of the structural fumigant, sulfuryl fluoride.Neumtox. Terato!. 10, 127-133.

Nitschke, K. D., and Eisenbrandt, D. L. (2001). Sulfuryl fluoride. In Handbookor Pesticide Toxicology, Agents, 2nd Edition, Chap. 88, Vol 2. (RobertKrieger, ed.), Academic Press, San Diego, CA.

Nitschke, K. D., Albee, R. R., Mattsson, J. L., and Miller, R. R. (1986).Incapacitation and treatment of rats exposed to a lethal dose of sulfurylfluoride. Fwzdam. App!. Toxico!. 7,664-670.

Pattison, F. L. M. (1959). Toxic Aliphatic Fluorine Compounds (ElsevierMonographs, Ethel Browning, ed.), Elsevier Publishing Company Ltd.,New York.

Singer, L., and Ophaug, R. (1982). Ionic and non ionic fluoride in plasma(or serum). Crit. Rev. Clin. Lab. Sci. 18, 111-140.

Strandberg, L. G. (1964). S02 absorption in the respiratory tract. Arch. Envimn.Health 9, 160-166.

Whitford, G. M. (1996). The metabolism and toxicity of fluoride. Monogr. OralSci. 16 Rev. 2, 1-153.

page1titlesRapid Uptake, Metabolism, and Elimination of Inhaled Sulfuryl ,.

imagesimage1image2image3image4

page2titles240 MATERIALS AND METHODS

imagesimage1

page3titles241 RESULTS Sulfuryl Fluoride Exposure

imagesimage1

page4imagesimage1image2

tablestable1

page5titlesto o 4

imagesimage1image2

tablestable1

page6titles-'" I ~ 244 o Chemical Analysis-Urine

imagesimage1image2image3image4image5image6image7image8

tablestable1table2

page7titles~ ~ ... ~

imagesimage1image2image3image4

page8imagesimage1

page9titlesREFERENCES 247

imagesimage1