Mid-gestation Exposure of C57BL/6 Mice to 2,3,7,8-Tetrachlorodibenzo- p -dioxin Causes Postnatal Morphologic Changes in the Spleen and Liver

Abstract

Pregnant C57BL/6 mice were exposed to 5 μg/kg 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) or vehicle by oral gavage between gestation days (GDs) 11 and 13. The thymus, spleen, and liver of the pups were examined histologically, and cell surface antigen expression was assessed on postnatal days (PNDs) 1, 14, 25, and 46. In addition to the expected decrease in thymic weight on PND 1, TCDD caused an increase in splenic weight on PND 14 and in hepatic weight on PNDs 14 and 25. The apoptotic index was increased and the corticomedullary border poorly defined in thymuses of TCDD-exposed mice on PND 1, but not at later endpoints. T lymphocytes were increased and B lymphocytes decreased in spleens of the TCDD-exposed mice on PND 46. TCDD-exposed mice had a nearly significant (p =.051) decrease in the number of splenic germinal centers on PND 46. Foci of extramedullary hematopoiesis (EMH) were increased in number in the livers of TCDD-exposed mice on PND 14, suggesting possible increased production of immune cells of unknown phenotype and function in this organ. These results suggest that late-gestation thymic architectural changes caused by TCDD resolve shortly after birth: however, abnormalities in other immunologically important areas may appear later in postnatal life.

Introduction

Polychorinated dibenzo dioxins (PCDDs) are organic compounds that are highly persistent, both environmentally and within organisms. Most of the environmental PCDD contamination is a result of various industrial processes that produce PCDDs as by-products (Schecter et al. 2006). Of the PCDDs, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) has been the most widely studied and is the most toxic. Sources of exposure to TCDD include ingestion of residue-containing food and water, exposure to contaminated soil, and inhalation of TCDD-containing dust, smoke, and air. Although there continue to be critical gaps in our knowledge of the mechanisms by which TCDD exerts its toxic effects, it is well documented that TCDD binds to the aryl hydrocarbon receptor (AhR), resulting in deleterious biological consequences. The AhR is a cytoplasmic protein that, once bound to a ligand, translocates to the nucleus and forms a heterodimer with the AhR nuclear translocator (Arnt). The AhR/Arnt heterodimer then interacts with the promoter region of Ah-responsive genes to regulate transcription, resulting in the majority of observed toxic effects (Mandal 2005). In addition to the classical mechanism described above, Enan and Matsumura (1995) have demonstrated that ligand-bound AhR activates cytoplasmic protein kinases, resulting in non-transcriptionally mediated effects. Furthermore, AhR cross-talk with hormonal and nonhormonal signaling pathways, as well as interaction with a variety of cellular regulatory and signaling proteins, have been described (Carlson and Perdew 2002).

TCDD has many toxic effects on numerous body systems including, but not limited to, the reproductive, nervous, and immune systems. In addition to the toxic effects caused by adult exposure, developmental exposure to TCDD, even at a dose that has no appreciable toxic effects in the mother, causes an array of systemic derangements in mammalian and non-mammalian species. Both placental and lactational transfer of TCDD are important routes of exposure for fetuses and neonates (Gehrs et al. 1997). In accidental human gestational TCDD exposures, an association with low birth weight, low thyroid hormone level, psychomotor and cognitive deficits, and altered immune function has been uncovered by epidemiologic research (Feeley and Brouwer 2000). Experimentally produced changes in laboratory animals after prenatal TCDD exposure include reduced growth rates, hemorrhage, edema, cleft palate, neurobehavioral defects, hydronephrosis, and lymphoid organ hypoplasia. Additionally, the reproductive competence of both male and female rodents is deleteriously affected by prenatal exposure (Birnbaum 1995).

Among the immune derangements caused by developmental TCDD exposure are a number of functional changes, including suppression of the delayed-type hypersensitivity response that persists into adulthood in rats (Gehrs and Smialowicz 1999). These authors previously detected changes in the relative abundance of immune cell populations that were present throughout the lives of rats following low-level developmental TCDD exposure. In late gestation, these changes included a decrease in the percentage of thymic TCR⁻CD4⁺CD8⁺cells and an increase in the percentage of thymic TCR⁻CD4⁻CD8⁺cells (Gehrs and Smialowicz 1997). During the neonatal period, these rats showed a decrease in the percentage of thymic TCR⁻CD4⁻CD8⁻, TCR⁺CD4⁻CD8⁻, and TCR⁺CD4⁺CD8⁺cells and an increase in the percentage of thymic TCR⁺CD4⁻CD8⁺cells (Gehrs and Smialowicz 1997). Finally, as adults, rats displayed decreased percentages of thymic γδTCR⁺CD4⁻CD8⁺, γδTCR⁺CD4⁻CD8⁻, and MHCI⁻MHCII⁻ cells, along with decreased percentages of splenic γδTCR⁺CD4⁻CD8⁻ and MHCI⁻MHCII⁻ cells. The relative weight of the thymus was decreased in fetal and adult rats, whereas the relative weights of the liver and spleen in adult rats were increased (Gehrs et al. 1997; Gehrs and Smialowicz 1997).

In possible conjunction with the thymocyte phenotypic changes described above, we recently detected marked architectural changes in the thymus of late-gestation mice following maternal exposure to 5 or 10 μg/kg TCDD on gestation days (GDs) 14 (Camacho et al. 2005) and 16 (Besteman et al. 2005). Changes at GD 18 included a disruption of the architecture of the thymic corticomedullary junction, a decrease in thymic cellularity, and an increase in the number of pyknotic thymocytes. These alterations in normal thymic development were accompanied by altered thymocyte phenotypes and a dose-dependant increase in early apoptotic CD4⁺CD8⁺thymocytes, suggesting possible impaired thymocyte selection. Mice dosed with TCDD as above also displayed fetal hepatomegaly, decreased liver c-jun, bcl-2, and pKcα gene expression, and hepatocyte anisocytosis and anisokaryosis (i.e., fetal hepatocytes and their nuclei were too large and were not the same size relative to each other) (Besteman et al. 2007). It is not known if, or for how long, such disrupted thymic and liver development may persist into postnatal life. The purpose of the present study was to assess histopathologic changes in the thymus, spleen, and liver to determine how long, and to what degree, late-gestation changes in organ architecture may persist into the postnatal period. Concurrent flow cytometry was also conducted to provide immune cell maturation data for comparison to organ architecture.

Materials and Methods

Animal Model

Twenty-six 13-week-old female C57BL/6 mice were mated to thirteen 13-week-old male C57BL/6 mice by placing one male in a cage with two females for three days. The male and female mice were descendants of mice obtained from Charles River Laboratories, Inc. (Wilmington, MA, USA). The female mice were weighed prior to breeding and again twelve days after the first day of breeding to determine whether or not they were pregnant and to approximate the GD. The estimated day of mating was defined as GD 0. Those females that were determined by weight change to be pregnant (a weight gain of six or more grams) were arbitrarily assigned to control or treatment groups. Pregnant females were then separated and housed individually for the remainder of the experiment. The mice were supplied with Harlan Teklad 2018 Global 18% Protein Rodent Diet (Harlan Teklad, Madison, WI, USA) and distilled water ad libitum before and during the experiment. A temperature of 66°F–72°F was held constant and a twelve-hour light/dark cycle was maintained in the animal housing facility for the duration of the experiment. To ensure the accuracy of recording date and time of birth, the pregnant females were observed every twelve hours and the approximate time of parturition was recorded. The day of parturition was counted as postnatal day (PND) 0. After birth, the pups were maintained with the dams until PND 21. No pups were added to or removed from the litters before PND 21 other than for use on the specified PND for experiments. Pups were then separated from dams and maintained in single-sex groups of one to five mice per cage.

Data for PND 1 were obtained from six mice of arbitrarily selected gender, each from different litters, on two separate collection days. Data from two mice from the TCDD group were collected on the first day. Data from one mouse from the TCDD group and three mice from the control group were collected on the second day. Data from the two days were pooled for analysis. Data for the PND 14, PND 25, and PND 46 end points were collected from five control mice and four TCDD-exposed mice of arbitrarily selected gender, each again from different dams, and at each endpoint these data were collected in a single day. All procedures were reviewed and approved by the Virginia Tech Institutional Animal Care and Use Committee (IACUC) prior to commencing the experiments. The investigators followed IACUC guidelines for humane treatment and care of animals in research.

Chemical Exposure

A concentration of 0.9 μg/mL of TCDD (AccuStandard, Inc., New Haven, CT, USA) was achieved by dissolving the compound in corn oil (Sigma Aldrich, St. Louis, MO, USA). The prepared TCDD mixture and additional unadulterated corn oil were stored in a lightproof box at 4°C until needed. Dams in the treatment group were dosed with 5 μg/kg of TCDD, based on weight obtained at the time of dosing, in a total volume of 140–180 μL (4.5 mL/kg) by oral gavage. The dose was administered between GD 11 and GD 13, as estimated by breeding date and weight gain. Pups from dams that were dosed six or seven days prior to parturition were used for PND 1 experiments, whereas pups from dams that were dosed seven days prior to parturition were used for experiments at all other end points to ensure accuracy of GD 12 dosing. A comparable volume of corn oil was administered by oral gavage to dams in the control group at the same time that dams in the treatment group received TCDD.

Tissue Collection

PND 1 mice were euthanized by hypothermia followed by severing the spinal cord at the foramen magnum, and PND 14, PND 25, and PND 46 mice were euthanized with carbon dioxide (Airgas, Inc., Radnor, PA, USA). Cadavers were immediately packed in ice after euthanasia. Each pup was weighed on an Ohaus Precision Standard T5120 scale (Ohaus, Pine Brook, NJ, USA) before tissue collection. Additionally, the numeric size of the litter from which each pup was taken was recorded. The thymus, spleen, and liver were removed by dissection from each pup at all four end points. After removal, each organ was immediately weighed. An organ-to-body-weight ratio was calculated for each organ. One lobe of the thymus and a portion of each spleen and liver were separated into a 1.5 mL polypropylene tube (Fisher Scientific, Pittsburgh, PA, USA) containing 1 mL of Bouin’s fixative (Ricca Chemical Co., Arlington, TX, USA) for histopathologic analysis. The remaining portion of each organ was placed in a 5 mL polystyrene tube (Fisher Scientific, Pittsburgh, PA, USA) containing incomplete RPMI-1640 media (MediaTech, Herndon, VA, USA) on ice until mechanical dissociation was performed. Bone marrow was harvested from each pup at all end points except for PND 1 by flushing 1 mL of incomplete RPMI-1640 media through the cavity of each femoral diaphysis with a 27-gauge needle (Becton Dickinson & Co., Franklin Lakes, NJ, USA) and 1 mL syringe (Sherwood Medical, St. Louis, MO, USA) into a polystyrene petri dish (Fisher Scientific, Pittsburgh, PA, USA). The culture medium containing the flushed bone marrow cells was then transferred to a 5 mL polystyrene tube on ice.

Cell Enumeration and Flow Cytometry

Cells from the liver, spleen, and thymus were released mechanically by gentle dissociation with curved forceps over a metallic sieve screen (Sigma Aldrich, St. Louis, MO, USA) into cold incomplete RPMI-1640 media (varying amounts from 1 to 4 mL depending on PND and organ size). Bone marrow was mechanically dissociated only if clumps were obtained from flushing. All samples were centrifuged at 200 x g for seven minutes at 7°C in an Eppendorf 5810 R centrifuge (Eppendorf, Hamburg, Germany). The thymus, liver, marrow, and PND 1 spleens were immediately resuspended in 0.3 to 4 mL phosphate buffered saline (PBS) (MediaTech, Inc., Herndon, VA, USA), with the amount of fluid added depending on pellet size and expected lymphocyte concentration. One mL of lysis buffer containing ammonium chloride, potassium bicarbonate, sodium EDTA, and hydrochloric acid (all obtained from Sigma Aldrich, St. Louis, MO, USA) was added to the pellets of PND 14 spleen cells, whereas 2 mL of ACK lysis buffer was added to the pellets of PND 25 and PND 46 spleens for three minutes to lyse red blood cells. The lysis reaction was stopped by adding cold incomplete RPMI-1640 media applied at a 4:1 ratio of media:lysis buffer by volume. The lysed spleen samples were then centrifuged and resuspended in 2–3 mL PBS.

All samples were then analyzed on a Multisizer 3 Coulter counter (Beckman Coulter, Inc., Fullerton, CA, USA) to enumerate lymphocytes. After enumeration, all samples were diluted or resuspended to achieve a final concentration of 5.0 × 10⁶ lymphocytes/mL. For each sample of liver, thymus, and bone marrow, 100 μL of each sample, containing 0.5 × 10⁶ lymphocytes, was placed into a 5 mL polystyrene tube. For each spleen sample, two such aliquots were made. To each of the tubes containing thymic cells, 100 μL of a solution containing 0.5 μg of anti-mouse CD4 antibodies labeled with phycoerythrin (PE) (eBioscience, San Diego, CA, USA) and 0.5 μg of anti-mouse CD8 antibodies labeled with fluorescein isothiocyanate (FITC) (eBioscience, San Diego, CA, USA) were added. To one of the two spleen aliquots for each sample, 100 μL of a solution containing 0.5 μg of anti-mouse CD45R (B220) antibodies labeled with PE (eBioscience, San Diego, CA, USA) and 0.5 μg of anti-mouse CD90.2 (Thy1.2) antibodies labeled with FITC (BD Pharmingen, San Jose, CA, USA) were added. To the second spleen sample and the bone marrow and liver samples, 100 μL of a solution containing 0.5 μg of anti-mouse CD44 antibodies labeled with FITC (eBioscience, San Diego, CA, USA) were added. All labeled samples were incubated in the dark at 4°C for thirty minutes. The labeling reaction was stopped by adding 2 mL of cold PBS, followed by centrifugation. The cells were then resuspended in 0.5 mL PBS and analyzed on an Epics XL-MCL flow cytometer (Beckman Coulter, Inc., Fullerton, CA, USA).

Histopathology

All tissue samples collected for histopathology were fixed in 1 mL of Bouin’s fixative for twenty-four to twenty-six hours. Samples were then washed in distilled deionized water until the yellow color of the fixative could no longer be visually detected. The samples were then stored in 70% ethanol until tissue processing and paraffin embedding. After embedding, a 5 μm section was cut from each slide, which was subsequently stained with hematoxylin and eosin (H & E). Tissues were then evaluated with a light microscope and by manual and digital analysis of photomicrographs. A numerical identification system was used such that the evaluator was blinded as to mouse treatment group.

Three randomly selected thymus sections from each of the end points for each of the treatment groups were assessed for their apoptotic indices, mitotic indices, and distinctness of the corticomedullary border. Apoptotic and mitotic indices were determined by counting the number of apoptotic and mitotic figures, respectively, in each of five 600X light micrographs taken at arbitrary locations in the thymic cortices. The counts from the five photomicrographs from each of the six thymuses were averaged, and this number was used for comparison of the indices from the two treatment groups. The distinctness of the corticomedullary border was assessed by side-by-side comparison of 40X photomicrographs of each of the six thymuses, which were subjectively scored on a scale from one to five with lower numbers correlating to a less distinct corticomedullary border.

Three randomly selected spleen sections from each of the end points for each of the treatment groups were analyzed. Spleens from PND 25 and PND 46 were analyzed by digital measurement using ImageJ 1.37v (NIH, Bethesda, MD, USA). The total organ area (in mm²) present in a 40X photomicrograph was measured digitally for each spleen. The total number of germinal centers (GCs) and the total number of periarterial lymphoid sheaths (PALS) in each of these 40X photomicrographs were counted manually. The number of GCs per mm² of spleen and the number of PALS per mm² of spleen was calculated by dividing the counted number of each structure by the total organ area present in the photomicrograph. The diameter of each PALS and each GC in each photomicrograph was measured digitally. The diameter measurements for each type of structure were averaged for each spleen. Using ImageJ, the white pulp structures were manually selected and measured to give a total white pulp area for each photomicrograph. The percentage of white pulp for each spleen was calculated by dividing the white pulp area by the total organ area in each photomicrograph. The spleens for PND 1 and PND 14 had not yet developed sufficiently defined boundaries between red and white pulp to facilitate the acquisition of similar measurements. These spleens were analyzed subjectively by side-by-side comparison of 40X, 200X, and 600X photomicrographs.

Three randomly selected liver sections from each of the end points for each of the treatment groups were assessed for the number of inflammatory foci and the number of foci of extramedullary hematopoiesis (EMH) present in each of five 400X photomicrographs from each liver. The average number of inflammatory foci and EMH foci per 400X micrograph for each liver was calculated and used for comparison. The five photomicrographs from each liver were also each given a subjective score of hepatocyte vacuolation on a scale from one to four, with higher numbers correlating to more vacuolation.

Statistical Analysis

In all tests, the dam was the statistical unit. In assessing the effects of gestational TCDD exposure on organ weight, statistical tests were performed only on organ weight/body weight ratios, as this was considered to be the most valid indicator of organ weight change. A one-sided Student’s t test performed in Microsoft Office Excel 2003 was used to analyze the apoptotic and mitotic rates in thymuses, with expected direction of change based on prior results reported by Besteman et al. (2005). A one-sided Monte Carlo permutation test performed on StatXact-7 (Cytel, Inc., Cambridge, MA, USA) was used to analyze the corticomedullary border scores, and a two-sided Monte Carlo permutation test was used to analyze hepatocyte vacuolation scores. A two-sided Student’s t test performed in Microsoft Office Excel 2003 was used to analyze all other data.

Results

Organ and Body Weights

No significant difference was observed in total litter size or body weight at any of the end points. The mean number of pups per litter was 7.8 (SEM = 0.37) for both the TCDD-treated and control groups (data not shown). No spontaneous pup or dam deaths occurred during the experimental period in either treatment group.

The relative thymic weight was significantly (p = 011) reduced in pups born from TCDD-treated dams (Table 1). On average, a reduction of greater than 50% in relative thymic weight was seen in PND 1 mice that were exposed to TCDD gestationally. By PND 14, changes in relative thymic weight were no longer detectable in the mice.

The relative splenic weight differed significantly (p = .050) between treatment and control groups only on PND 14, at which point an approximate 15% increase in relative spleen weight was observed in mice that were exposed to TCDD gestationally.

The relative liver weight was increased in mice that were exposed to TCDD gestationally. The relative liver weight was arithmetically, although not significantly (p =.088), higher in TCDD-exposed mice on PND 1, with a 31% increase in relative weight on average. There was a highly significant (p =.0001) increase in the relative liver weight of PND 14 TCDD-exposed mice, with a 31% increase in relative weight on average. A significant (p =.031) increase in relative liver weight in TCDD-exposed mice was also observed at PND 25, with a 12% increase in relative weight on average.

Cellular Antigen Expression

A near-significant (p =.054) decreased percentage of CD4⁺CD8⁻ thymocytes in mice that were exposed to TCDD during gestation was observed on PND 1 (Table 2). In TCDD mice, the percentage of CD4⁺CD8⁻ thymocytes was reduced to 64% of the control level. Although the mean percentage of CD4⁻CD8⁺and CD4⁻CD8⁻ was more than doubled in TCDD mice on PND 1, a large variability in individual measurements within each group was present, contributing to p values greater than 0.1. A change in thymocyte surface antigen expression was not observed at any of the later end points (data not shown).

Gestational TCDD exposure produced a significant (p =.034) increase in the percentage of cells in the spleen that were negative for CD44 expression on PND 1. An average of 25% more of the cells from the spleens of PND 1 TCDD-exposed mice were CD44⁻ compared to controls (Table 3). On PND 46 a significant (p =.030) decrease in the percentage of CD45R⁺lymphocytes and a significant (p =.009) increase in the percentage of CD90⁺lymphocytes were observed. An average of 6.5% fewer splenic lymphocytes were CD45R⁺, and an average of 5.8% more splenic lymphoctes were CD90⁺in PND 46 TCDD-exposed mice (Table 3).

No significant differences were observed in expression of CD44 in cells collected from PND 1 mouse livers. A near-significant decrease (p =.052) in the percentage of CD44^lo cells was observed in the livers of PND 1 mice. On average, there were only 77% as many CD44^lo cells in the livers of TCDD-exposed mice as there were in control livers (Table 4). PND 14 TCDD-exposed mice showed a near-significant decrease (p =.053) in the percentage of CD44^hi cells in the liver. On average, only 68% as many CD44^hi cells were present in the livers of TCDD-exposed mice as there were in controls (Table 4).

No significant changes were seen in the expression of CD44 in the bone marrow of TCDD-exposed mice at any of the end points (data not shown).

Histopathology

The apoptotic index was significantly (p =.049) increased in thymuses of PND 1 mice exposed to TCDD gestationally, as assessed by an increase in pyknotic nuclei and apoptotic bodies. The apoptotic indices in the thymuses of TCDD-exposed mice were nearly three times greater than the apoptotic indices observed in controls (Table 5, Figures 1a and 1b). Verification of increased apoptosis in similarly treated mice at GD 18 using the marker 7-aminoactinomycin D was performed previously and was not repeated for this series of experiments (Besteman et al. 2005). The corticomedullary border exhibited a significant (p =.049) lack of distinction in the thymuses of PND 1 TCDD-exposed mice when compared to controls (Figures 2a and 2b). Although the mean mitotic index was nearly doubled in TCDD-exposed mice, a large variability between individual measurements within each group contributed to a p value greater than .1. No significant differences were detected in any of the thymic assessment criteria at end points after PND 1.

There were no statistically significant histologic differences between the spleens of TCDD-exposed mice and control mice; however, PND 46 TCDD-exposed mice showed a near-significant decrease (p =.051) in the number of germinal centers per mm² (Table 6). The spleens of TCDD-exposed mice had an average of only 64% as many germinal centers per mm² as spleens of control mice on PND 46. Although the mean diameter of the germinal centers in TCDD-treated mice was 26% greater than that of control mice, a large variability in measurements within each group contributed to a p value greater than .15.

There was a highly significant (p =.005) increase in the number of foci of EMH in the livers of PND 14 TCDD-exposed mice (Figures 3a and 3b). Livers from TCDD-exposed mice had 9.3 ± 0.4 foci of EMH per 400X field (given as mean ± SEM), whereas control livers had 6.6 ± 0.2 foci of EMH per 400X field. No other significant changes were observed in livers at any of the other end points (data not shown).

Discussion

Developmental exposure to TCDD produces permanent postnatal suppression of T-cell–mediated immune function in rodents (Gehrs and Smialowicz 1999; Holladay and Smialowicz 2000). Changes in thymocyte surface antigen expression similarly persist into adulthood (Gehrs et al. 1997). Late-gestation thymic atrophy is severe in these animals; however, this effect is transient and the animal recovers shortly after birth (Fine et al. 1989; Holladay et al. 1991). We recently described noteworthy morphologic differences in the GD 18 mouse thymus after perinatal TCDD exposure, which have not been observed in adult-exposed mice (Besteman et al. 2005). It was not known if these structural changes reflected permanent effects on organ patterning, which may in part explain persistent altered antigen expression and depression of T-cell function by TCDD.

The present mice continued to show indistinct thymic corticomedullary junctions at PND 1, as well as increased pyknotic cells previously identified as apoptotic thymocytes (Besteman et al. 2005). The thymuses of these mice also showed a decrease in basophilic staining cells when viewed at low magnification. All of these thymic histopathologic differences had resolved by PND 14, such that TCDD thymuses were no longer visibly different from control. This observation suggests that perinatal disruption of the thymic corticomedullary border induced by TCDD may be a visual manifestation of severe thymocyte hypocellularity, with potential contribution from increased eosinophilic-staining stromal cell prominence owing to a lower density of basophilic-staining thymocytes. It is unclear why similar corticomedullary effects have not been observed histologically in adult mice after TCDD, given a similar induction of thymocyte hypocellularity. Similarly, it is not known if the transient perinatal alteration in normal organ corticomedullary architecture may affect thymocyte selection, although such may be suggested by altered surface antigen expression and changed rates of apoptosis (Besteman et al., 2005) and increased peripheral T cells expressing Vβ⁺T-cell receptors (Silverstone et al. 1998).

Perinatally the mouse spleen serves as a hematopoietic organ, shifting to a secondary lymphoid center over the first several weeks of postnatal life (Landreth and Dodson 2005). Effects of developmental exposure to TCDD on the early postnatal spleen have not been previously described. At PND 1 the present TCDD spleens showed increased CD44⁻ cells, with trends toward reduced CD44^lo and CD44^hi staining. In other mouse hematopoietic compartments (fetal liver and bone marrow), CD44^hi cells are of granulocyte-lineage and CD44^lo cells are of lymphocyte-lineage, whereas the CD44⁻ cells include stem cells and other progenitors (Holladay and Smith 1994). Thus, the increased CD44⁻ spleen cells suggest an early effect of TCDD on spleen hematopoiesis, similar to fetal liver (Fine et al. 1989), which resolved by PND 14. Relative spleen weight was changed only at PND 14, at which time TCDD spleens were marginally (15%) larger than controls. Gehrs et al. (1997) similarly noted increased spleen weights in fourteen-week-old rats following gestational exposure to TCDD. Reasons for this transient window of increased postnatal spleen weight in both mice and rats remain unclear.

Additional changes occurred in the spleens of TCDD-exposed mice between PND 25 and PND 46. Within this window, the mouse spleen completes its transition from a hematopoietic organ to a secondary lymphoid center (Holladay and Smialowicz 2000). The percentage of splenic B cells (CD45R⁺) was decreased and the percentage of splenic T cells (CD90⁺) increased at PND 46 in mice that received prenatal TCDD. Total numbers of B and T cells (percentages x organ cellularity) could not be calculated in these mice because a portion of each spleen was used for histopathology. However, spleen weights were nearly identical (0.07 ± 0.003 and 0.07 ± 0.002 g in control and TCDD mice, respectively), suggesting the phenotypic difference may translate to changed total numbers of B and T cells. Histologically, there was a near-significant decrease in the number of splenic germinal centers (p =.051) at PND 46 that may support the observed decreased percentage of splenic B cells, although an absolute increase in splenic T cells cannot be ruled out. Interestingly, there was no statistically significant evidence of a change in the size of GCs in the TCDD mice. In what may be a related observation, Inouye et al. (2003) reported an inhibition of splenic GC formation in young adult mice for at least two weeks after TCDD exposure, which they hypothesized was a result of decreased proliferation owing to alteration in the interaction with helper T cells. Because the splenic changes noted in the present study were at the last end point evaluated (PND 46), it is not known whether they resolve or persist longer into adulthood. A similar experiment with later end points is warranted to answer this question.

An increase in relative liver size in late gestation (Besteman et al. 2005) and in the early postnatal period after perinatal TCDD exposure has been previously documented (Gehrs and Smialowicz 1997; Lucier et al. 1975; Lin et al. 2001). This hepatomegaly was correlated to enzyme induction (Lucier et al. 1975) and increased hepatocyte cytoplasmic volume and nuclear size (Besteman et al. 2007). In the present experiments the hepatomegaly was still present at PND 25 and had resolved by PND 46. A highly significant increase in the density of foci of EMH was histologically evident at PND 14 and has not previously been reported. This increase occurred in the presence of the hepatomegaly, which might intuitively be expected to dilute the foci of EMH over a larger area in the liver, causing a decrease in the density of these foci. Using flow cytometry, Silverstone et al. (1998) detected extrathymic T-cell differentiation in enzymatically digested postnatal liver after developmental exposure to TCDD, including increased expression of Vβ⁺T cells that are normally deleted in the thymus. This observation raises questions regarding the increased hematopoietic centers detected histologically at PND 14 in the present TCDD mice, and the nature of the cells in these particular foci. CD44 staining of liver revealed a near-significant trend toward a decrease in the percentage of CD44^lo cells in the liver on PND 1, which may be suggestive of a previously reported decrease in the percentage of prolymphocytes in the liver detected at late gestation (Fine et al. 1990). Further investigation in postnatal fetal livers, using a larger panel of antibodies, will be necessary to determine the identity of the hematopoietic cells that are residing in the postnatal livers after TCDD exposure.

In conclusion, mid-gestation exposure to TCDD caused transient disruptions of thymic architecture and thymocyte apoptosis that resolved by PND 14. It remains unknown if these changes may affect neonatal establishment of self-tolerance. Limited changes were detected in the perinatal spleen after developmental TCDD; however, the adult spleen showed altered T- and B-cell percentages and a trend toward reduced germinal centers. Hepatomegaly was persistent in the TCDD-exposed mice and was still present at PND 25. The liver transitions from a hematopoietic to a metabolic compartment perinatally in mice (Holladay and Smialowicz 2000). Hepatocyte-derived growth factors are important for supporting liver hematopoiesis (Hackney et al. 2002; Wineman et al. 1996), thus alterations in this compartment by TCDD may contribute to altered hematopoiesis. The observation of increased EMH at PND 14 in livers from TCDD mice suggests effects on liver hematopoiesis, including potential increased production of immune cells of unknown phenotype and function, and warrants further investigation.