Acetone in Drinking Water Does Not Modulate Humoral Immunity in Mice as Measured by the Antibody,Plaque-Forming Cell Assay

Abstract

It has been reported that the repeated topical, nonoccluded application of acetone may modulate antibody production in mice, thus producing humoral immunosuppression. However, the evaporative loss expected following nonoccluded dermal application of acetone makes the systemic effect seem unlikely. This study was designed to investigate the immunotoxicity potential of acetone in mice following a more direct systemic route of dosing via drinking water for 28 days. CD-1 male mice consumed average daily acetone doses of 121, 621 or 1144 mg/kg/day. The antibody, plaque-forming cell (AFC) assay was performed to measure the T cell–dependent, anti-sheep red blood cell immunoglobulin M (IgM) response, and hematology and thymus weights were evaluated to provide additional insight into the potential effects to the immune system. Body weights, white blood cell (WBC), numbers, red blood cell (RBC) counts, and hemoglobin and hematocrit levels showed no treatment-related effects at any dose of acetone. Eosinophil percentages were variable but also showed no dose-related trends. Spleen and thymus weights were not statistically different from controls and there were no effects on spleen cellularity or AFC response as a result of acetone administration. The AFC responses ranged from 1088 to 1401 AFCs/10⁶ splenocytes and were not statistically different from controls (1277 AFCs/10⁶ cells). Mice treated with cyclophosphamide (20 mg/kg) on days 25 to 28 demonstrated a 94% reduction in AFC/10⁶ cells. Thus, the direct systemic administration of acetone did not produce evidence for immunotoxicity in CD-1 mice and the no observed adverse effect level (NOAEL) in this study was determined to be 1144 mg/kg/day.

Acetone (CAS no. 67-64-1), a widely used high production volume chemical, is a normal by-product of fatty acid metabolism and present naturally at measurable levels in human tissues and excreta. The physiological concentrations of acetone can increase as energy requirements of the body increase (e.g., during exercise, dieting, or pregnancy) depending on the concentration of blood glucose available and the need for energy generation via the process of gluconeogenesis (Morgott 2001).

The biochemistry and toxicology of acetone has been extensively studied over the last several decades and acetone has been found to have very low toxicity following either acute or subchronic administration, regardless of the route of exposure. For example, acetone had only minimal effects on hematopoiesis and the kidney and/or the liver in rats and mice following subchronic exposure through drinking water (Dietz 1991; Dietz et al. 1991). Other studies indicate that acetone is neither a neurotoxicant (Christoph, Malley, and Stadler 2003; Spencer et al. 1978), nor does it have reproductive or teratogenic potential (Dalgaard et al. 2000; Larsen, Lykkegaard, and Ladefoged 1991; Mast et al. 1988). A previous study examining the immunotoxic potential of tetradecanoylphorbol acetate using SSIN mice suggested that repeated topical application of acetone, as the vehicle-control group, may modulate antibody production, thus producing humoral immunosuppression when compared to untreated controls (Singh et al. 1996). Acetone administered at 50 to 300 μl/treatment, once or twice weekly for four or eight applications, altered a few of the antibody responses towards sheep red blood cells (SRBCs) but lacked a clear dose-response or a schedule-dependency trend (Singh et al. 1996). Mice treated with acetone volumes of 50 or 100 μl demonstrated lowered antibody, plaque-forming cell (AFC) responses after a single application but not after four or eight treatments. A single application of 200 μl did not reduce the antibody response whereas four applications reduced the AFC count, but eight applications had no effect. Lastly, topical applications using 300 μl reduced the antibody response to SRBCs after one, four, or eight treatments. Acetone did not demonstrate consistent effects for most of the parameters that were measured by Singh and coworkers (1996) including the SRBC antibody response, the lymphoproliferative response to an alloantigen (mixed lymphocyte response) and splenocyte cytometry (B cells and CD4/CD8 T cells). In particular, the observation of a statistically significant increase in the mitogen-induced lymphoproliferative response appears contradictory to the investigators’ conclusions of immunosuppression.

The absence of an immunosuppressive effect for acetone is consistent with its standard use as a diluent in the mouse local lymph node assay (LLNA), a test widely used to determine the contact sensitization potential for many substances (OECD 2002; USEPA 2003). If acetone acted as an immunosuppressant, it would be expected to attenuate the sensitivity of the LLNA to detect sensitizers. However, this assay has been validated against a range of chemical sensitizers having varying potencies. The present study was designed to investigate the immunotoxic potential of acetone by employing methodologies consistent with contemporary regulatory guidelines.

Purpose and Design Summary

The study was designed and conducted in accordance with the United States Environmental Protection Agency’s Immunotoxicology Guideline (OPPTS 870.7800). To evaluate the functional responsiveness of the immune system towards the T cell–dependent antigen SRBCs, CD-1 male mice were exposed to acetone or positive- and negative-control substances for 28 days. The animals were immunized by intravenous injection of SR-BCs 4 days before the end of the exposure period. The antibody, plaque-forming cell (AFC or PFC) assay (Holsapple 1995) was carried out 24 h after the last exposure to determine the effects of the test substance on the splenic anti-SRBC immunoglobulin M (IgM) response. In addition to the anti-SRBC antibody response (which included spleen weights and cell counts), hematology and thymus weights were assessed to further assess adverse effects to the immune system.

MATERIALS AND METHODS

Animals

Male CD-1 IGS [Crl:CD1(ICR)] mice, 9 weeks of age at study start, were obtained from Charles River Laboratories, Portage, MI. Animals were maintained individually using plastic caging under environmental conditions (40% to 70% relative humidity, 22°C ± 1°C room temperature, 12-h light/dark cycle, and room air changes approximately 12 to 15 times/h) and food and water were provided ad libitum. All procedures and animal care were approved through the Animal Care and Use Committee of The Dow Chemical Company. Twice each day a cage-side examination was conducted and to the extent possible the following parameters were evaluated: skin, fur, mucous membranes, respiration, nervous system function (including tremors and convulsions), animal behavior, moribundity, mortality, and the availability of feed and water. All mice were weighed during the pre-exposure period, at least weekly during the dosing period, and on the day of their scheduled sacrifice.

CD-1 mice were selected based upon their effective use with SRBC antibody response assays to evaluate immunotoxicity potential of other xenobiotic chemicals (Kauffmann et al. 1982; Loveless et al. 2002, 2003; Shopp et al. 1984, 1985; White et al. 1985). Male mice were used primarily based upon the evidence of slightly greater sensitivity to acetone toxicity as reflected in heightened incidence and severity of hepatocellular hypertrophy at 20,000 ppm (5000 mg/kg) in male versus female B6C3F1 mice following 13 weeks of acetone exposure via drinking water (Dietz 1991; Dietz et al. 1991). In addition, there are no data that support female mice having a greater sensitivity for immunotoxicity; therefore, female mice were not evaluated in this design.

Test Materials

Acetone and cyclophosphamide monohydrate (CAS no. 6055-19-2) were obtained from Sigma Chemical, St. Louis, MO, and were demonstrated by the manufacturer to be greater than 99% pure.

Rationale for Dose and Route Selection

Inhalation and dermal exposure represent the most likely routes for both occupational and consumer exposure to acetone. Pharmacokinetic studies have shown that acetone is completely absorbed following its oral administration (Plaa et al. 1982) and readily absorbed via inhalation by both rats and mice, although at different uptake efficiencies (Morris 1991; Morris and Cavanagh 1986). Specifically, the uptake efficiencies for inhaled acetone were 21% in Sprague-Dawley rats and 14% in B6C3F1 mice. A recently developed, and fully validated, physiologically based pharmacokinetics (PBPK) model for acetone allows for the extrapolation across routes of administration and species (rats to humans) to determine acetone bioavailability and internal tissue dose at various exposures (Gentry et al. 2003). This model was used to calculate the area under the blood acetone concentration–time curves (AUC) for the acetone drinking water studies of Dietz et al. (Dietz 1991; Dietz et al. 1991). The daily AUC was 10,440 mg-h/L for the drinking water dose of 10,000 ppm (900 mg/kg/day). For a developmental toxicity study where acetone was administered via inhalation (Mast et al. 1988), the daily AUC for exposures to 2200 ppm acetone for 6 h/day was 266 mg-h/L. This inhalation exposure for 6 h per day at 2200 ppm is more than twice the inhalation limit dose of 2 mg/L (≈844 ppm) specified in multiple Environmental Protection Agency (EPA) testing guidelines, including those for Developmental and Immunotoxicology testing. Although these comparisons are for the rat, mice would have a lower internal dose of acetone via inhalation based on the studies of Morris et al. (Morris 1991; Morris and Cavanagh 1986). Hence, the oral administration of acetone in drinking water at the limit dose of 1000 mg/kg/day provided a more rigorous examination of the potential of acetone to produce immunotoxicity than would inhalation exposures at the limit concentration of 2 mg/L.

The permissible exposure limit (PEL) established by the U.S. Occupational Health and Safety Administration for acetone is 1000 ppm [8 h, time-weighted average (TWA)]. Wigaeus, Holm and Astrand (1981) examined the relative uptake and excretion of acetone in human subjects and found that the uptake of acetone remained relatively constant at about 44% of the exposure concentration, regardless of the vapor exposure regimen. Based on this, a 70-kg individual exposed to acetone at the PEL for 8 h would absorb an internal dose of 149 mg/kg. Thus, the highest dose for this study (1000 mg/kg) is approximately 7 times higher than this estimate of the dose received by a permissible occupational exposure.

In the study of the immunotoxic potential of acetone by Singh et al. (1996), the highest nonoccluded topical dose applied was 300 μl/mouse. The absorbed dose in this study was unknown, but the vapor pressure of acetone is high (185 mm Hg at 20°C) (Morgott 2001) and acetone is hydrophilic in nature (log P _ow = −0.24) (Wilkinson and Williams 2001). Therefore, a large fraction of the applied dose likely evaporated from the skin within minutes of application. Assuming that the dermal absorption was as high as 10% of the applied dose, the internal dose for a 20-g mouse would have been 1182 mg/kg. However, based upon in vitro dermal absorption studies using human skin (less than 1% absorbed over 2 h) (Wilkinson and Williams 2001), the actual dose of acetone absorbed dermally in mice was likely much lower than 10% of the applied dose.

In this study, the doses selected were 100, 500, and 1000 mg acetone/kg/day (the latter being the maximum limit dose based on EPA Immunotoxicology Guideline). These doses covered the range of the potentially absorbed doses estimated from the study of Singh et al. (1996).

Acetone Administration

Acetone solutions used as drinking water were prepared using municipal water and were offered to the mice ad libitum using glass water bottles fitted with stainless steel sipper tubes. Stability of acetone in drinking water for up to 7 days was confirmed (data not shown) and concentrations were confirmed using high-performance liquid chromatography (HPLC) with ultarviolet (UV) detection prior to day 1. Both the negative and positive controls were offered municipal water ad libitum. Routine analysis of municipal water determined that the levels of selected contaminants were below those that may compromise or confound the study results.

The test material was administered to the animals (n = 8 per group) via the drinking water continuously for 28 days. Drinking water solutions were prepared once every 7 days and water consumption was determined pre-exposure and once weekly for all animals by weighing the water bottles at the start and end of a measurement cycle. Test material intake was subsequently calculated.

Immunotoxicology Endpoints

On days 25 to 28, the positive-control group was injected intraperitoneally with cyclophosphamide (CYP) at 20 mg/kg in saline; this concentration of CYP typically results in at least 90% reduction in the anti-SRBC AFC response. Four days prior to sacrifice, all of the mice were immunized with a single intravenous injection of 1 × 10⁸ SRBCs via the lateral tail vein. On day 29 the weight of the spleen and thymus was recorded for each animal. Spleens were subsequently processed for the AFC plaque assay similarly to that previously described (Holsapple 1995). The calculations for the SRBC AFC assay included the following: AFCs/ml (original suspension), AFCs/spleen, AFCs/10⁶ spleen cells (as determined using a Z1 Coulter Counter, Beckman Coulter, Miami, FL). The following hematologic parameters were assayed using a Technicon H-1E Hematology Analyzer (Bayer Corporation, Tarrytown, NY): hematocrit percentage, hemoglobin concentration, red blood cell (RBC) count, total white blood cell (WBC) count, platelet count, differential WBC evaluations (eosinophils, basophils, lymphocytes, neutrophils, and monocytes) and RBC indices (mean corpuscular hemoglobin and mean corpuscular volume).

Statistics

Means and standard deviations for body and organ weights, antibody responses, spleen cell number, WBC and RBC counts, and hemoglobin, hematocrit, and platelets concentrations were analyzed using a Bartlett’s test followed by a one-way analysis of variance (ANOVA) (Zar 1999). If analyses indicated a significant difference by the ANOVA, the treatment groups were compared to the drinking water control using a Dunnett’s test (Zar 1999). WBC differentials and RBC indices were not statistically compared. Outliers in all data were identified using Grubb’s test but were not excluded from statistical analysis unless there was sufficient documentation to justify exclusion. SAS, version 6.12, was used to perform all statistical analyses.

RESULTS

Test Material Intake (TMI) and In-Life Observations

Verification of acetone concentrations in drinking water prior to day 1 using HPLC/UV indicated solutions at each dose level were within 5% of target concentrations, thus average TMI was determined using target concentrations of 0.6, 3, and 6 mg/ml, respectively. Acetone treatment groups consumed average daily doses of 121, 621, or 1144 mg/kg/day, as based on the drinking water consumption measured weekly. Water consumption volumes in the acetone-treated groups were not statistically different from controls (data not shown). During the first week, mice provided water containing approximately 6 mg acetone/ml water drank an average of 1.1 g less water than did mice from control groups. However, this level of water consumption was not appreciably less than that measured the week prior to initiation of treatment for these same animals, and did not reduce the average acetone exposure which was targeted at 1000 mg/kg/day (average acetone consumption for days 1 to 8 was 1179 mg/kg/day). Body weights of acetone-treated mice were not statistically different from controls throughout the 28 days of acetone administration (data not shown). No mortalities or overt signs of toxicity occurred in any group.

Hematology

Hematology parameters and WBC differentials were unaffected by acetone consumption (Tables 1 and 2). Eosinophil (EOS) percentages appeared to be lower than controls for the mice treated with the 100 mg/kg/day dose of acetone. However, there was no apparent dose-related effect on this parameter and all the group means were within the range of historical control data (range = 1.5%–4.1%). Further, there was an EOS value that was identified as an outlier in the negative control group (9.6%) that was not excluded from the calculation of the mean, which raised the mean from 2.97% to 3.8%. These data do not support a treatment-related effect of acetone on EOS.

CYP administered as the positive control demonstrated effects on hematology with reduced values for RBC counts, WBC numbers, and hemoglobin and hematocrit levels (Table 1), as well as slight alterations to lymphocyte and neutrophil differentials (Table 2). Although the mean value for EOS in the CYP-treated group appeared lower than controls at 1.3%, CYP at higher doses typically induces eosinophilia, therefore this value was not considered to be treatment related.

Lymphoid Organs

Acetone treatment for 28 consecutive days did not result in statistical differences in spleen or thymus parameters between control and acetone-treated animals. The relative thymus weights of mice administered acetone were not statistically different from control thymus weights (Figure 1). Mice administered municipal water and treated with CYP demonstrated a greater than 50% reduction in thymus weights (p < .001). Likewise, spleen weight and spleen cellularity were unaffected by 28 days of acetone treatment (Table 3).

No effects were noted for the SRBC IgM antibody response following acetone exposures. Neither AFCs per spleen (data not shown) nor AFCs per 10⁶ splenocytes were statistically different between controls (1277 AFCs/10⁶ splenocytes) and mice administered acetone (1088 to 1401 AFCs/10⁶ splenocytes) (Figure 2). CYP treatment resulted in a 94% decrease in AFCs per 10⁶ splenocytes.

DISCUSSION

Because acetone is an endogenous metabolite of intermediary metabolism and its production rate in humans can readily increase as result of fasting, exercise, or other physiologic changes such as obesity or diet, it is thought unlikely that it would express activity as an immunosuppressive agent at physiologic concentrations. Exogenous exposure to acetone from numerous sources do not appear to be consistently as high as that measured in expired air due to endogenous production from normal biological mechanisms (reviewed in Morgott 2001). Exposure to airborne acetone can occur from cigarette smoking, furnishings, construction materials, and chemical manufacturing, but the most significant contributors include vegetative releases, forest fires, and photooxidation of propane and similar molecules (reviewed in Morgott 2001 and Singh et al. 1994).

A previous study by Singh et al. (1996) reported immune suppression due to acetone in a study of the immune modulation potential of tetradecanoylphorbol acetate using SSIN mice. Although the SSIN strain has not typically been used in immunotoxicity studies and there was a somewhat inconsistent pattern of response in the parameters measured, it suggested that a reexamination of the immunotoxicity potential of acetone was needed. In this study we chose to use the CD-1 mouse as this strain is known to be responsive in the AFC assay and is generally considered acceptable for routine immunotoxicity testing (Kauffmann et al. 1982; Loveless et al. 2002, 2003; Shopp et al. 1984, 1985; White et al. 1985). However, the observation of immune stimulation potential simultaneously reported by Singh et al. (1996) using a mitogen-induced (T-) lymphocyte proliferation assay was not addressed by this study design. The impact of repeated acetone administration on a cell mediated endpoint might alternatively be addressed by combining a 28-day repeat exposure protocol with a LLNA using CBA mice (a measure of cell-mediated, dermal sensitization potential) in a manner similar to that recently reported (van den Berg et al. 2005).

The no observed adverse effect level (NOAEL) in this study was determined to be 1144 mg/kg/day as acetone administration to mice in the drinking water for 28 consecutive days did not alter spleen and thymus weights, hematological parameters (including circulating numbers of neutrophils, lymphocytes, and monocytes), or the T-dependent antibody response. Administration of acetone via drinking water at concentrations near the advised limit doses (1000 mg/kg/day) did not suppress the primary IgM antibody response towards SRBCs (Figure 2). The primary antibody response to a T-dependent antigen requires a coordinated response from T, B, and antigen-presenting cells to generate a specific antibody response, and is generally considered to be a primary indication of immunotoxicity potential. Hence acetone did not show evidence of an effect using these robust measures of immunotoxicity potential.

The results of this study are in agreement with another report which evaluated acetone as a vehicle during a humoral immune response assessment to the pesticide carbaryl by the dermal and inhalation routes (Ladics et al. 1994). Nose-only exposures to the highest concentration of carbaryl were estimated to have also resulted in an exposure of 17 g/m³ of acetone/day. Additional CD rats were administered acetone alone as a vehicle control for 5 days per week over 2 weeks using approximately 17 g/m³ (assuming a minute volume of 0.133 L/min and 50% absorption, the internal dose would approximate 1500 mg/kg/day). These animals did not demonstrate a reduced IgM AFC response to SRBCs compared to air only controls. In parallel studies evaluating the dermal immunotoxicity potential of carbaryl, Ladics et al. (1994) dermally administered acetone (2 ml) for 5 days per week over 2 weeks and occluded the site of dosing. Dermal exposures to acetone for 2 weeks at approximately 7000 mg/kg/day did not reveal effects on the SRBC primary antibody response.

The primary antibody response to a T-dependent antigen demonstrates concordance above 90% for known human immunotoxicants when combined with an endpoint such as thymus weight (Luster et al. 1992). In this study, acetone treatment did not produce any alterations in thymus or spleen weights that were statistically different from controls. Any suggestion of a treatment-related effect on thymus weights was not supported by a treatment-related effect for WBC counts or lymphocyte differentials (Tables 1 and 2). This is consistent with other studies that have also shown no effects on the thymus following acetone administration. Ladics et al. (1994) did not report thymus effects after 2 weeks of either nose-only (internal dose ∼1500 mg/kg/day) or occluded topical (∼7000 mg/kg/day) administrations to CD rats. In addition, a drinking water study that supplied acetone for either 2 weeks or 13 weeks at doses greater than 10000 mg/kg did not demonstrate effects on thymic pathology or alterations in thymus weights using male and female mice (Dietz et al. 1991). Dietz et al. (1991) also reported a lack of effect on WBC differentials following these same acetone administrations in drinking water for up to 13 weeks. Although the eosinophil differentials in the low-dose acetone group in this study appeared to be lower than controls, this observation is not consistent with the literature and was not deemed to be treatment related on the absence of any dose-response and the historical control data for this parameter.

Lastly, the absence of an immunosuppressive effect for acetone is consistent with its recommended use as a vehicle in the mouse local lymph node assay (LLNA), a test which has been validated using a variety of chemical sensitizers and is accepted as an accurate measure of dermal sensitization potential when compared to human and guinea pig data (Basketter et al. 2002; Haneke et al. 2001). If acetone had immunosuppressive potential, it would be expected that its use in this assay would reduce the sensitivity of the LLNA. However, comparative analyses for both weak and strong dermal sensitizers have determined that the LLNA is suitable as a method to identify some weak dermal sensitizers. The topical dose of acetone in an LLNA might range from 300 mg/kg (80% v/v solution) to 1500 mg/kg (less than 1% v/v solution) over 3 consecutive days. Although the LLNA typically evaluates sensitization potential on the 3rd day after these exposures, van Och et al. demonstrated a similar potency for DNCB (using an acetone vehicle) after a 60-day LLNA exposure period (8 weekly applications) as compared to the results of a standard 6-day assay (van Och et al. 2003).