Effects of acrylamide exposure on serum hormones, gene expression, cell proliferation, and histopathology in male reproductive tissues of Fischer 344 rats

Abstract

Acrylamide (AA) is a reactive monomer used in many technological applications, but it is the incidental formation during cooking of common starchy foods that leads to pervasive human exposure, typically in the range of 1 μg/kg body weight (bw)/day (d). AA is carcinogenic in multiple organs from both sexes of several rodent models and a consistent body of evidence points to a genotoxic mechanism based on metabolism to a DNA-reactive epoxide, glycidamide (GA). In F344 rats, tumorigenesis occurs in several hormonally regulated tissues (thyroid, mammary gland, and peri-testicular mesothelium), which has prompted speculation about endocrine dysregulation as a possible mechanism. The present study evaluated the effects of a 14 d exposure to AA administered through the drinking water on reproductive tissues and the hypothalamic-pituitary-testes (HPG) axis in male F344 rats. The doses selected encompass a range from approximately 2.5 mg/kg bw/d, which is carcinogenic after lifetime exposure, to 50 mg/kg bw/d, a maximally tolerable dose that causes hind limb paralysis. AA caused significant changes in serum hormones, histopathology, testicular gene expression, and cell proliferation, especially at the highest dose. Despite strong evidence for activation of the HPG axis subsequent to decreases in testosterone levels, and histopathological changes associated with significant effects on Leydig and germ cells, with concomitant mRNA expression changes, the precise mechanism(s) for AA-induced testicular toxicity remains unclear; however, the absence of evidence for increased proliferation of the peri-testicular mesothelium (Ki-67 immunoreactivity) does not support hormonal dysregulation as a contributing factor to the predisposition of this tissue to the carcinogenic effects of AA.

1. Introduction

Acrylamide (AA) is the monomeric precursor for many industrially important polymeric products (e.g., water flocculents, thickening agents); however, the incidental formation of AA during the cooking process of many common starchy foods (e.g., potato and bakery products) is the primary source of human exposure in the range of 1 μg/kg bw/d (reviewed in Doerge et al., 2008). The α,β-unsaturated amide functional group of AA confers chemical reactivity with protein nucleophiles (LoPachin and Barber, 2006), and metabolism by hepatic CYP 2E1 produces an epoxide, glycidamide (GA), which is reactive toward protein and nucleic acid nucleophiles (Ghanayem et al., 2005). AA is carcinogenic in multiple organs of both sexes in several rodent models including A/J mice (Bull et al., 1984a), Swiss-ICR mice (Bull et al., 1984b), B6C3F₁ mice (NTP, 2011; Von Tungeln et al., in press), and F344 rats (Johnson et al., 1986; Friedman et al., 1995; NTP, 2011). A consistent body of experimental evidence in vivo has been developed that supports a genotoxic mechanism for AA carcinogenicity involving GA–DNA adduct formation (Gamboa da Costa et al., 2003) and somatic cell mutations (Manjanatha et al., 2006; Mei et al., 2010; Wang et al., 2010; Von Tungeln et al., 2009). Several major international bodies, including the International Agency for Research on Cancer (IARC, 1994), the U.S. EPA (EPA, 2008), and the World Health Organization/Food and Agriculture Organization (JECFA, 2010) have evaluated the carcinogenicity of AA and all classified it as a likely human health hazard, based on the rodent data. Epidemiological studies have examined the relationship between dietary exposure to AA and cancers, which are negative at many but not all sites (reviewed in Pelucchi et al., 2011). However, the reliability of these findings is not clear given the significant limitations in statistical power, the potential for non-differential misclassification due to the widespread nature of AA in common foods, and the difficulty in distinguishing high and low consumers, even when food frequency questionnaires and exposure biomarkers are used (Doerge et al., 2008). In addition, it is difficult to predict human tumor sites based on those observed in chronic rodent bioassays (Rice, 2005).

Testicular toxicity has been reported in rodents exposed to AA (Sakamoto et al., 1988; Yang et al., 2005; Wang et al., 2010; Hamdy et al., 2011; Ma et al., 2011). Changes in gonadal and pituitary hormones have been associated with histopathological and gene expression changes but no clear mechanism for testicular toxicity has been delineated.

Three chronic bioassays of AA in F344 rats consistently reported increased tumor incidences in the thyroid (males and females), mammary gland (females only), and the peri-testicular mesothelium (Johnson et al., 1986; Friedman et al., 1995; NTP, 2011). Increased mutant frequencies (MFs) were observed in several tissues (spleen lymphocyte Hprt and thyroid cII, but not mammary gland, whole testes, or liver cII) of Big Blue rats treated with doses of AA similar to those used in chronic studies (Mei et al., 2010). However, the extensive distribution of AA and GA throughout total cellular water leads to similar levels of GA–DNA adducts in all tissues, making additional tissue-specific factors essential to tumor induction. A previous evaluation of the hypothalamic-pituitary-thyroid axis in male F344 rats exposed to AA showed no evidence for its activation that could enhance thyroid carcinogenesis (Bowyer et al., 2008a). The present study examines the same group of rats to evaluate the effect of AA on the hypothalamic-pituitary-testicular (HPG) axis and assess its possible influence on the formation of peri-testicular mesotheliomas (tunica vaginalis). Serum and pituitary hormone levels and testicular histopathology were measured in male F344 rats treated for 14 d with doses of AA in drinking water from 2.5 mg/kg bw/d, a dose causing significantly increased tumor incidences after lifetime exposure (Johnson et al., 1986; Friedman et al., 1995; NTP, 2011), to 50 mg/kg bw/d, a maximally tolerated dose producing hind limb paralysis (Crofton et al., 1996; Barber et al., 2001; Bowyer et al., 2008b; NTP, 2011). In a parallel study, gene expression was evaluated in testes from Big Blue rats treated with AA in drinking water (4 and 8 mg/kg bw/d) for 60 d, the same animals used in another previously reported study (Mei et al., 2010). These endpoints were compared to determine the relative sensitivity of testicular gene expression biomarkers versus changes in hormone levels and histopathology for evaluation of testicular effects of AA and alteration of the HPG axis.

2. Methods

2.1. Reagents

Sigma Chemical Co. (St. Louis, MO) supplied the AA (>99.9% purity, CAS 79-06-1) and all biochemical reagents.

2.2. Animal handling procedures

All procedures involving use of rats were reviewed and approved by the NCTR Laboratory Animal Care and Use Committee. For the 14 d exposure study, male Fischer 344 (F344) rats were obtained from the NCTR colony at weaning [postnatal day (PND) 21], housed two per cage, and maintained on NIH41IRR meal diet (Purina Mills, Brentwood, MO). Starting at PND 70, AA was administered 7 d/wk, for 14 d, in the drinking water at concentrations of 0, 25, 100, and 500 mg/L, concentrations chosen to administer daily doses of approximately 0, 2.5, 10, and 50 mg/kg bw/d, as previously described (Bowyer et al., 2008a,b). Based on twice weekly measurements of cage-average body weights and drinking water consumption, the average doses delivered were 93–100% for the 2.5 mg/kg bw/d dose, 99–100% for the 10 mg/kg bw/d dose, and 85–88% for the 50 mg/kg bw/d dose. Rats (10 per dose group) had access to the drinking water containing AA continuously, until removal at termination as previously described (Bowyer et al., 2008a,b).

For the gene expression study, the testicular tissues were dissected from the same animals used by Mei et al. (2010). Briefly, male Big Blue transgenic rats, derived originally from the F344 strain, were obtained from Taconic Farms (Germantown, NY) as weanlings and housed two per cage. Rats were provided NIH41IRR meal diet (Purina Mills) ad libitum. At ~50 d of age, rats (n = 7–8 per dose group) were exposed to 0, 0.7, and 1.4 mM AA, dissolved in the drinking water, 7 d/wk for 60 d. This resulted in AA doses of approximately 0, 4, and 8 mg/kg bw/d, respectively (Mei et al., 2010). The rats were anesthetized with carbon dioxide and exsanguinated by cardiac puncture within 24 h of the last treatment.

2.3. Histopathology

The right testis and epididymis of the rats exposed for 14 d were removed at necropsy, weighed, and fixed for 24 h in modified Davidson’s fixative, comprised of 30% of a 37–40% solution of formaldehyde, 15% ethanol, 5% glacial acetic acid, and 53% distilled H₂O (Latendresse et al., 2002). Both tissues were trimmed, processed and embedded in infiltrating media (Formula R^®), before being sectioned at approximately 4–5 μm. The epididymal sections were stained with hematoxylin and eosin (H&E) and the testicular sections were stained with periodic acid Schiff. A detailed qualitative examination of the testis and epididymis was conducted, taking into account the tubular stages of the spermatogenic cycle, in order to identify treatment-related effects such as missing germ cell layers or types, retained spermatids, multinucleate or apoptotic germ cells, and sloughing of spermatogenic cells into the lumens of seminiferous tubules. The severity of lesions were scored using numerical values, where 1 = minimal, 2 = mild, 3 = moderate, 4 = severe. Pituitaries from rats exposed for 14 d were removed at necropsy and fixed in 10% neutral buffered formalin (NBF) for 48 h. Pituitaries were then weighed and transferred to 70% ethanol before trimming and embedding using Formula R infiltrating media.

2.4. Leydig cell counts, size, and proliferation

A subset of the right testes specimens was weighed after fixation to confirm minimal to no shrinkage. Specimens were then embedded in paraffin and sectioned at 4–5 μm. From each testis, sections were taken from the midsection and poles. For detection of Leydig cells, the sections of testes were stained immunohistochemically for 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1). Sections were de-paraffinized and rehydrated. Endogenous peroxidase was inhibited by incubation with freshly prepared 3% hydrogen peroxide with 0.1% sodium azide for 10 min at room temperature (RT). After incubation with Tris-buffered saline containing 0.1% Tween 20 (pH 7.4), nonspecific staining was blocked with 10% normal goat serum (Jackson Immunoresearch, West Grove, PA) for 20 min at RT. The sections were then incubated with rabbit polyclonal anti-11β-HSD1 antibody (Cayman Chemical, Ann Arbor, MI) at a dilution of 1:400 (1 μg/ml) for 1 h at 37 °C. After incubation with primary antibody, tissue sections were incubated with biotinylated goat antirabbit IgG (ExtrAvidin Kit, Sigma, St. Louis, MO) at a dilution of 1:30 for 30 min at RT, followed by incubation with streptavidin-conjugated horseradish peroxidase (ExtrAvidin Kit, Sigma) at a dilution of 1:30 for 30 min at RT. Staining was developed with 3,3-diaminobenzidine (DAB; Sigma, St. Louis, MO) substrate for 5 min at RT and the sections were counterstained with hematoxylin and mounted with Permount (Fisher Scientific, Pittsburgh, PA). For the negative control, 1 μg/ml rabbit IgG (Jackson Immunoresearch) or phosphate-buffered saline (PBS) replaced the primary antibody.

Leydig cell counts were determined from digital images of 11β-HSD1-immunostained sections of rat testes captured using an Aperio Scanscope System (Aperio Technologies, Inc., Vista, CA). In these images, the proportion of immunostained area, which corresponds to the percentage volume of Leydig cell cytoplasm (V_p), was evaluated using the Positive Pixel Count Algorithm (Aperio Technologies, Inc., Vista, CA). For each animal, V_p was determined in 3 images (midsection and poles) and the mean value was calculated. Values of V_p were converted to absolute volumes (V_abs) per testis by reference to the testis volume. Testes volumes were obtained directly from weight measurements since the specific gravity of testis does not considerably differ from 1.0 (Mori and Christensen, 1980; Marshall and Plant, 1996). Testis weights before and after fixation were comparable in each animal, indicative that shrinkage was minimal.

$$V_{\text{abs}}=\frac{V_{\text{p}}\times \text{weight}}{100}$$

The mean Leydig cell volume (V_Lc) was determined using the nucleator method (Sharpe et al., 2000; Wreford, 1995) modified by the use of digital imaging (Aperio Technologies, Inc., Vista, CA). In the images of 11β-HSD1-immunostained testes sections, 4 separate radii from the nucleolus to the cell membrane (r_1–4) were measured per cell using the Ruler Tool (Aperio Technologies, Inc., Vista, CA). Fifty Leydig cells were analyzed per animal (n = 6 per treatment group), and the mean V_Lc value was calculated using the following equation (Wreford, 1995):

$$V_{\text{Lc}}=\frac{\pi }{3x(r_1^3+r_2^3+r_3^3+r_4^3)}$$

A similar method was used to determine the mean volume of the Leydig cell nuclei (V_nucl) in each treatment group. For each animal, the number of Leydig cells per testis (N_Lc) was calculated using the corresponding values of V_abs, V_Lc, and V_nucl:

$$N_{\text{Lc}}=\frac{V_{\text{abs}}}{V_{\text{Lc}}-V_{\text{nucl}}}$$

For evaluation of the proliferative activity of the Leydig cells, a separate set of testes sections was immunostained for Ki-67 protein. The sections were placed in an antigen retrieval solution (0.01 M citrate buffer, pH 6.0) for 15 min in a microwave oven at 100 °C at 600 W. Endogenous peroxidase was quenched as described above. After blocking in 0.5% casein, rabbit monoclonal anti-Ki-67 (clone SP6, Lab Vision, Fremont, CA) was applied to the sections at the dilution of 1:200 for 1 h at RT. After incubation with the primary antibody, tissue sections were incubated with horseradish peroxidase conjugated goat anti-rabbit IgG (Vector, Burlingame, CA) at a dilution of 1:50 for 30 min at RT. Staining was developed with DAB and the sections were counterstained with hematoxylin, dehydrated and mounted. The number of Ki-67-positive interstitial cells for each animal was evaluated in 10 randomly selected areas of the testes interstitium by the use of the Nuclear Algorithm (Aperio Technologies, Inc., Vista, CA).

2.5. Staining of follicle stimulating hormone (FSH)/luteinizing hormone (LH)-positive pituitary cells

2.5.1. Immunohistochemistry

Formalin-fixed paraffin-embedded rat tissue sections (4–5 μm) were deparaffinized and rehydrated. For immunohistochemical demonstration of LH or FSH, sections of pituitary gland were placed in an antigen retrieval solution (0.01 M citrate buffer, pH 6.0) for 15 min in a microwave oven at 100 °C at 600 W. Endogenous peroxidase was inhibited by incubation with freshly prepared 3% hydrogen peroxide with 0.1% sodium azide for 10 min at RT. Nonspecific staining was blocked with normal goat 10% serum (Sigma, St. Louis, MO) for 20 min at RT. The sections were then incubated with rabbit polyclonal LH or FSH antibodies (DAKO, Carpinteria, CA) at a dilution of 1:700 or 1:500, respectively, for 1 h at RT. After incubation with primary antibody, tissue sections were incubated with horseradish peroxidase conjugated goat anti-rabbit IgG (Vector, Burlingame, CA) at a dilution of 1:50 for 30 min at RT. Staining was developed with DAB (Sigma, St. Louis, MO) substrate for 5 min at RT, sections were counterstained with hematoxylin, and mounted with Permount (Fisher Scientific, Pittsburgh, PA). For the negative control, rabbit IgG (Jackson Immunoresearch Laboratories, West Grove, PA) or PBS replaced the primary antibody.

2.5.2. Image analysis

LH- and FSH-immunostained rat pituitary gland sections were scanned and digital images were obtained by the use of the Aperio Scanscope System. In these images, the proportion of immunostained area and the intensity of staining were evaluated with Positive Pixel Count Algorithm (Aperio Technologies, Inc., Vista, CA).

2.6. Serum hormones

Male F344 rats (n = 10 per dose group) were treated with 0, 2.5, 10 or 50 mg/kg bw/d AA for 14 d, as described above. Blood was collected by cardiac puncture after anesthesia with carbon dioxide asphyxiation. To reduce variation, all blood was collected between 8:00 AM and 12:00 PM. Blood was allowed to clot in serum separator tubes and centrifuged at 1000 × g for 10 min. The serum aliquots were frozen at −80 °C until analyzed. Clinical chemistry analyses were conducted using the following radioimmunoassay techniques: FSH and prolactin (ALPCO Diagnostics, Salem, NH); testosterone, estradiol, and progesterone (Siemens Medical Solutions Diagnostics, Tarrytown, NY), and LH (DSL, Webster, TX). The radioimmunoassays were counted on a COBRA 5005 gamma counter instrument (PerkinElmer, Shelton, CT). Each assay was run in duplicate as a single batch and included a standard curve and two levels of assayed controls. The procedures were performed according to the manufacturer’s protocol. All maintenance, normalization, and calibration with ¹²⁵I of the COBRA 5005 instrument were done according to the manufacturer’s recommendations.

2.7. Measurement of mesothelial proliferative activity

Ten animals per dose group were treated with 0, 2.5, 10 or 50 mg/kg bw/d AA for 14 d, as described above, and anesthetized with an intraperitoneal injection of pentobarbital sodium (150 mg/kg bw, Abbott Laboratories, Deefield, IL), before being exsanguinated and perfused with 10% NBF. The perfused scrotum, testis, and epididymis were removed as a unit, with both testes and epididymides in essentially normal apposition within the scrotal sac. A transverse section of the unit (skin, testis, and epididymis) was taken midway along the longitudinal axis of the testes (Fig. 1A). Paraffin sections of the entire unit were prepared as described above. The paraffin-embedded sections were deparaffinized and rehydrated. Sections were immunostained for Ki-67 protein as described above, scanned, and digital images were captured using an Aperio Scanscope System. Ki-67 immunopositive mesothelial cells along the serosal surfaces of the scrotum, epididymis, and testis (Fig. 1B–D) were point-counted manually and expressed as counts per linear millimeter of mesothelial lining. Separate counts were done for scrotal sac, epididymis, and testis.

Fig. 1.

Measurement of mesothelial proliferative activity. (A) Transverse section of scrotum, epididymis, and testis dissected as a unit and transected midway along the longitudinal axis of the testes, showing general locations where linear measurements of serosal surfaces and point counts of Ki-67 immunopositive mesothelial cell nuclei were done. S, scrotum; T, testis, E, epididymis. Tissue section stained with H&E. (B–D) Ki-67 immunopositive mesothelial cells (arrows) along the serosal surfaces of tunics associated with scrotum (B), epididymis (C), and testis (D). Tissue sections stained with DAB and counterstained with H. Original magnification 200×.

2.8. Gene expression study

The testicular tissues used for the gene expression analysis were dissected, flash frozen in liquid nitrogen, and stored at −80 °C until processed for RNA isolation.

2.9. RNA purification and cDNA synthesis by reverse-transcription

Whole testes from seven animals of the control (vehicle only) group and eight animals from each treated group were used. The frozen whole testes were macerated in liquid nitrogen and 20–30 mg of powder were used to purify total RNA using an RNeasy Mini kit with on-column DNase I digestion, following the manufacturer’s protocol (QIAGEN, Valencia, CA). RNA purity and concentration were assessed using a Nanodrop 1000 spectrophotometer (Thermo Scientific, Wilmington, DE) and the RNA integrity (RIN) was assessed using a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA). All RNAs had an absorption 260 nm/280 nm ratio >2.0 and a RIN >9.0. One microgram of total RNA was reverse transcribed using random hexamer oligonucleotides and SuperScript III First Strand synthesis system, and the RNA was digested with RNase H (Invitrogen, Carlsbad, CA), following the manufacturer’s protocol. The complementary DNA (cDNA) was diluted 1:15 with nuclease-free water and stored at −20 °C.

2.10. Quantitative analysis of gene expression by real-time PCR

The expression levels of selected genes were analyzed by quantitative real-time polymerase chain reaction (qPCR) using SYBR Green and TaqMan chemistries. The SYBR Green-labeled DNA products were amplified with 5′- and 3′-primers designed using the NCBI Primer-BLAST software and iQ^™ SYBR^® Green Super mix (Bio-Rad Laboratories, Hercules, CA), as previously described (Bowyer et al., 2008a,b). SYBR Green PCR cycling conditions were set at 95 °C for 5 min for the first cycle and 30 s at 95 °C followed by 30 s at 60 °C for the remaining 40 cycles. This was followed by 40 repetitive cycles of 10 s starting at 55 °C and incrementing in temperature by 0.5 °C/cycle to determine a melt curve, as a means of validating the PCR products. The qPCRs that used the TaqMan chemistry were performed in an ABI 7900HT instrument (Applied Biosystems, Foster City, CA), using TaqMan assays (Applied Biosystems) and FastStart Universal Probe master mix (Roche, Indianapolis, IN). TaqMan PCR cycling conditions were set at 95 °C for 10 min for the first cycle and 15 s at 95 °C followed by 1 min at 60 °C for the remaining 40 cycles. Tables S1 and S2 list the genes analyzed, the sequences of the oligonucleotides used and the catalog numbers of the TaqMan assays used. Genes with threshold cycle (Ct) values higher than 35 were considered not expressed. The relative expression levels of each gene of interest was expressed as a percentage of the expression level of the endogenous control gene Gapdh (NCBI: NM 017008), by calculating 2^−ΔCt × 100, where ΔCt value is the Ct value of the gene of interest minus that of Gapdh. Gapdh expression levels had been determined in preliminary absolute quantitation experiments to be unchanged with AA treatment (data not shown).

2.11. Statistical analysis of gene expression data

Values are expressed as mean ± standard deviation (SD). Statistical significance between groups was assessed by one-way ANOVA, followed by Dunnett’s test to compare AA-treated groups to the control, or the t-test to compare the high AA treatment group to the control (SigmaStat v3.11). A p-value ≤0.05 was considered statistically significant.

3. Results

3.1. Effects of AA on serum and pituitary hormones

As reported in our previous studies (Bowyer et al., 2008a,b), male rats were exposed to AA in the drinking water and administered doses of approximately 2.5, 10, and 50 mg/kg bw/d, as determined from body weight and water consumption data. While blood levels were not directly measured in these experiments, this exposure regimen is similar to the one we used in previous drinking water exposures to AA conducted in the same animal model using a dose of approximately 1 mg/kg bw/d (Doerge et al., 2005). These and other related data were used to validate a rat physiologically based pharmacokinetic model (Young et al., 2007) that predicted respective average daily serum concentrations in the present study of approximately 1.5, 6, and 30 μM for AA and 1.2, 5, and 25 μM for GA for the three doses tested. Exposure of male F344 rats to AA produced evidence for activation of the HPG axis, subsequent to a decrease in testosterone levels. As shown in Fig. 2, serum levels of testosterone were significantly decreased by 10 and 50 mg/kg bw/d doses of AA. Serum levels of LH (Fig. 2) and the % area of LH-staining in the pituitary (Fig. S1) were significantly elevated by 10 and 50 mg/kg bw/d doses of AA. The level of FSH in serum (Fig. 2) and the intensity of FSH-staining in the pituitary (but not the % area stained) were significantly decreased only at the highest dose of AA (Fig. S2). Serum levels of progesterone, a metabolic precursor to testosterone, were significantly decreased at the highest dose of AA, while levels of estradiol (E2), the aromatase-derived product of testosterone metabolism, and prolactin were not significantly changed in any dose group compared to the control (Fig. 2).

Fig. 2.

Effect of acrylamide on serum hormone levels in male F344 rats exposed to 0–50 mg/kg bw/d acrylamide in the drinking water. Serum hormone levels were determined using immunochemical methods as described in Section 2 and the units used were: luteinizing hormone (LH), units/ml; follicle-stimulating hormone (FSH), ng/ml; prolactin (PRL), ng/ml; total testosterone (T), ng/dl; free estradiol (E2), pg/ml; and progesterone (Prog), ng/ml. Data are expressed as mean ± SD (n = 10 animals per dose group). *Significantly different from control at p ≤ 0.05 using ANOVA with Dunnett’s test for multiple comparisons.

3.2. Testicular and epididymal lesions induced by AA exposure

Exposure to 50 mg/kg bw/d AA for 14 d induced several testicular lesions, including exfoliated germ cells, depletion of germ cells, spermatid retention, and apoptosis, characterized by deeply eosinophilic cytoplasm in pachytene spermatocytes or round spermatids (Table 1 and Fig. 3); Both the incidence and the severity of the testicular lesions was high in the animals dosed with 50 mg/kg bw/d, but decreased to near control levels in the 10 and 2.5 mg/kg bw/d AA dose groups; the only exception was the incidence (but not the severity) of the spermatid retention, which remained elevated in all AA dose groups compared to the control (Table 1). Since the effect of AA doses below 50 mg/kg bw/d on the testicular histopathology was very limited, further histological investigations focused only on changes induced by the high AA dose group versus the control.

Table 1.

Testicular and epididymal histopathological lesions induced by acrylamide exposure.

Tissue	Morphology	AA Dose (mg/kg bw/d)
		0	2.5	10	50
Testis	Depletion, germ cell	1/10 (1.0)a	0/10	0/10 (1.0)	10/10 (2.1)
	Exfoliation, germ cell	1/10 (1.0)	1/10 (2.0)	0/10	9/10 (2.8)
	Retention, spermatid	2/10 (1.0)	9/10 (1.1)	6/10 (1.0)	9/10 (2.6)
	Apoptosis, including pachytene spermatocytes and round spermatids in Stages VII–VIII	1/10 (1.0)	0/10	1/10 (1.0)	7/10 (1.0)
Epididymis	Exfoliation, germ cell	0/10	1/10 (1.0)	1/10 (2.0)	10/10 (2.7)

Note: Stage VII tubules are characterized morphologically by cell layers including preleptotene and pachytene spermatocytes, round spermatids with large acrosome capping nuclear membrane and mature spermatids (spermatozoa) lining tubular lumen prior to release in Stage VIII; Stage VIII tubules contain preleptotene and pachytene spermatocytes, round spermatids with acrosome oriented toward the base of tubule, and possibly a very few “post-release” straggler spermatozoa still present along the margin of the tubular lumen.

^aAverage severity of the lesion, where 1 = minimal, 2 = mild, 3 = moderate, 4 = severe.

Fig. 3.

Typically observed acrylamide-induced testicular lesions in rats exposed to 50 mg/kg bw/d acrylamide in the drinking water: (A) Exfoliation of germ cells in Stage I of spermatogenic cycle with seminiferous tubular lumen containing cellular detritus, mostly degenerated elongated spermatids and an occasional spermatocyte (arrow); (B) Depletion of germ cells exposing wisps of Sertoli cell cytoplasm (arrows) and germ cell exfoliation (arrowhead) in Stage V; (C) Stage XI seminiferous tubule with two sets of maturing elongated spermatids instead of one, (Step 19 spermatids are normally released in Stage VIII); and (D) Apoptosis of pachytene spermatocytes (arrows) in Stage VII, commonly observed with testosterone insufficiency. All photomicrographs stained with PAS. Panels (A, C, D) original magnification 400×; panel (B) original magnification 200×. Note: Stage I tubules are characterized morphologically by germ cell layers including spermatogonia, pachytene spermatocytes, newly formed round spermatids lacking a distinct acrosome, and elongated spermatids; Stage V tubules contain spermatogonia, pachytene spermatocytes, round spermatids with small acrosome capping nuclear membrane and elongated spermatids; Stage VII tubules contain germ cell layers including preleptotene and pachytene spermatocytes, round spermatids with large acrosome capping nuclear membrane and spermatozoa lining tubular lumen prior to release in Stage VIII; Stage XI tubules have germ cell layers including leptotene and pachytene spermatocytes and elongated spermatids.

3.3. Leydig cell volume, counts, and proliferation

Testicular absolute weight was decreased in the 50 mg/kg bw/d AA dose group compared to control (Table 2). Volumetric measurements of Leydig cells showed a decrease in Leydig cell total and cytoplasmic volume, with no appreciable change in their nuclear size. Furthermore, the estimated cell count in the 50 mg/kg bw/d AA dose group, computed from the absolute volume of the Leydig cell compartment as a proportion of total testicular volume, was not significantly different from control (Table 2). The lack of evidence for Leydig cell proliferation (i.e., no increase in Ki-67-immunopositive cell numbers with AA treatment, data not shown) or increased interstitial cell degeneration (apoptosis or necrosis), further supports the finding of no significant changes in the number of Leydig cells in the 50 mg/kg bw/d AA dose group versus control.

Table 2.

Leydig cell volume and number in control and acrylamide-treated rats. Data are expressed as mean ± SEM, from 6 animals per dose group.

Treatment group	Weight of testis (mg)	Vp (%)	Vabs × 109 (μm3)	VLc (μm3)	Vnucl (μm3)	VLc − Vnucl (μm3)	NLc × 106
Control	1.442 ± 24.0	3.4 ± 0.2	49.8 ± 1.9	859.4 ± 52.5	137.2 ± 14.1	722.2	68.6 ± 4.4
AA (50 mg/kg bw/d)	1.326 ± 15.9*	3.0 ± 0.2*	35.7 ± 3.3*	689.3 ± 49.3*	150.5 ± 17.2	538.8*	72.1 ± 5.7

V_p is % volume of Leydig cell cytoplasm per testis; V_abs is absolute volume of Leydig cell cytoplasm per testis; V_Lc is mean volume of Leydig cells; V_nucl is mean volume of Leydig cell nuclei; N_Lc is total number of Leydig cells per testis.

^*Significantly different from control at p ≤ 0.05.

3.4. Mesothelial cell proliferation

The mean Ki-67 immunopositive mesothelial cell counts along the serosal surfaces of the scrotum, testis, and epididymis of rats in either control or 50 mg/kg bw/d AA dose groups are shown in Fig. 4. The number of Ki-67-positive cells was counted in the scrotal, testicular, and epididymal mesothelium (Fig. 1) and the labeling indices (LI) were expressed per mm of mesothelial lining in each region and also summed for all surfaces. Exposure to 50 mg/kg bw/d AA for 14 d significantly decreased the LI of mesothelial cells of testis, epididymis, and along the combined length of the serosal surfaces, but not of the scrotal sac, compared to the control.

Fig. 4.

Ki-67 immunostaining of peri-testicular mesothelium from untreated rats and those administered 50 mg/kg bw/d acrylamide in the drinking water. Data are expressed as mean ± SD, from 5 to 6 animals per dose group. *Significantly different from respective control at p ≤ 0.05 using the t-test.

3.5. Gene expression of biomarkers of testicular function

To evaluate further the effects of AA on testicular function and to evaluate possible mRNA expression biomarkers, a parallel analysis of AA-induced gene expression changes was under-taken. The testes fixation required for histological evaluation precluded the use of the rats treated for 14 d; however, testes were available from another of our AA studies, which determined the mutagenicity of AA and GA in Big Blue rats (Mei et al., 2010). In this study, rats were exposed to vehicle, 4, or 8 mg/kg bw/d AA for 60 d prior to sacrifice and whole testes collected at sacrifice were stored frozen at −80 °C. These doses were selected to bracket tumorigenic doses observed in the three chronic bioassays of AA (Johnson et al., 1986; Friedman et al., 1995; NTP, 2011) in order to evaluate the potential for somatic cell mutations produced from shorter term exposures in target organs. The expression of several genes involved in the functional pathways for oxidative stress, cell proliferation and apoptosis, cytoskeleton, steroidogenesis and related pathways, germ cell survival, and Sertoli cell-germ cell interaction, was evaluated by qPCR. Since many of the statistically significant changes detected were modest in terms of fold-change, qPCRs were independently confirmed using a different set of oligonucleotides or TaqMan assays. The results of the gene expression analysis are summarized in Table 3.

Table 3.

Effects of acrylamide on testicular gene expression. Data are expressed as mean ± SD.

Gene symbol and functional category	Acrylamide dose (mg/kg bw/d)
	0 (n = 7)	4 (n = 8)	8 (n = 8)
Oxidative stress
Sod1a,b	1.00 ± 0.09	0.95 ± 0.10	0.81 ± 0.11*
Gpxa	1.06 ± 0.40	0.95 ± 0.30	1.31 ± 0.23
Gsra	1.01 ± 0.19	0.92 ± 0.06	0.97 ± 0.19
Gssa,b	1.00 ± 0.07	1.12 ± 0.10	0.87 ± 0.08*
Gclca	1.02 ± 0.20	0.89 ± 0.11	0.90 ± 0.20
Cata,b	1.00 ± 0.09	1.04 ± 0.06	1.28 ± 0.17*
Cell proliferation
Pcnab	1.01 ± 0.11	1.04 ± 0.10	0.81 ± 0.10*
Mki67b	1.03 ± 0.24	1.08 ± 0.13	1.03 ± 0.33
Apoptosis
Casp3a,b	1.00 ± 0.09	1.19 ± 0.07*	1.18 ± 0.13*
Fasla,b	1.01 ± 0.15	1.10 ± 0.17	1.26 ± 0.22
Fasb	1.02 ± 0.22	1.23 ± 0.13	1.55 ± 0.28*
Cytoskeleton
Kif2ca	1.01 ± 0.13	0.91 ± 0.07	0.98 ± 0.13
Vima	1.03 ± 0.26	0.81 ± 0.28	0.93 ± 0.34
Related to steroidogenesis
Lhra,b	1.01 ± 0.15	1.19 ± 0.13	1.33 ± 0.17*
Stara,b	1.03 ± 0.27	1.20 ± 0.33	1.67 ± 0.41*
Tpsoa	1.03 ± 0.07	0.98 ± 0.09	1.00 ± 0.07
Cyp11a1a,b	1.04 ± 0.30	1.05 ± 0.18	1.51 ± 0.31*
Cyp17a1b	1.06 ± 0.40	0.99 ± 0.27	1.16 ± 0.32
Hsd3b1b	1.00 ± 0.11	1.01 ± 0.11	1.09 ± 0.12
Testosterone metabolism
Srd5a1b	1.00 ± 0.11	0.90 ± 0.16	0.81 ± 0.08*
Germ cell survival
Ckita,b	1.00 ± 0.10	1.19 ± 0.11*	1.14 ±0.10*
Scfb	1.01 ± 0.14	1.09 ± 0.10	1.19 ± 0.21
Insl3a,b	1.00 ± 0.06	1.15 ± 0.10*	1.33 ± 0.10*
Sertoli cell-germ cell tight junction
Tesa,b	1.05 ± 0.38	1.94 ± 0.42*	3.73 ± 0.95*

^aSYBR Green chemistry.

^bTaqMan chemistry.

When both chemistries were used for a given gene of interest, and since the two methods yielded similar results, only the TaqMan chemistry data are shown.

^*Significantly different from control at p ≤ 0.05.

The expression of several genes encoding enzymes involved in the oxidative stress pathway, including superoxide dismutase (Sod1), glutathione synthetase (Gss), and catalase (Cat) were affected in animals treated for 60 d with 8 mg/kg bw/d AA, but not 4 mg/kg bw/d. No AA-induced gene expression changes were observed for the redox enzymes glutathione peroxidase (Gpx) and glutathione reductase (Gsr).

Eight mg/kg bw/d AA also induced the apoptosis-related genes Casp3 (caspase 3) and Fas, but not its ligand (Fasl). On the other hand, the proliferating cell nuclear antigen gene Pcna was downregulated by the 8 mg/kg bw/d AA dose, although no transcriptional changes were observed in another cell proliferation marker gene, Mki67. Taken together, these data suggest that the apoptosis:cell proliferation ratio in the whole testis is increased by exposure to 8 mg/kg bw/d AA for 60 d.

The gene that encodes the luteinizing hormone receptor (Lhr) was up-regulated in rats treated with 8 mg/kg bw/d AA. Three additional Leydig cell-specific transcripts were modulated by 8 mg/kg bw/d AA. The Star gene, which encodes the steroidogenic acute regulatory protein responsible for the transport of the testosterone-precursor cholesterol to the inner mitochondrial membrane, and the Cyp11a1 gene, which encodes the steroidogenic enzyme cytochrome P-450ssc responsible for the conversion of cholesterol to pregnelenone, the first metabolic step of testosterone biosynthesis, were up-regulated by 8 mg/kg bw/d. In contrast, the Srd5a1 gene, which encodes 5-alpha-reductase, the enzyme responsible for the conversion of testosterone into the more potent androgen dihydrotestosterone, was downregulated by 8 mg/kg bw/d AA. The lower dose of AA tested (4 mg/kg bw/d) had no effect on these testosterone-related genes. On the other hand, the gene encoding the Leydig cell-specific factor insulin-like peptide 3 (Insl3) was up-regulated by both 4 and 8 mg/kg bw/d AA, making it the most sensitive biomarker gene of Leydig cell function in animals exposed to AA.

The gene expression of testin (Tes), a Sertoli cell-germ cell junction integrity biomarker, showed the largest fold-change increases observed in the current study: 4 and 8 mg/kg bw/d AA upregulated Tes gene 2- and 4-fold, respectively. Moreover, even though the Sertoli cell product c-Kit ligand (Scf) was not affected by AA treatment, both 4 and 8 mg/kg bw/d AA increased the gene expression of its receptor, c-Kit.

4. Discussion

A large and consistent body of data supports a genotoxic mechanism for AA-induced rodent carcinogenicity based on metabolism to GA and subsequent formation of GA–DNA adducts (Doerge et al., 2005). The fact that AA induces tumors in F344 rat tissues that are regulated by the endocrine system (thyroid, testis, and mammary gland) has lead some researchers to hypothesize that alternative, endocrine-disrupting mechanisms could be involved (Shipp et al., 2006). The current work is part of a series of mechanistic studies designed to elucidate tissue-specific factors that could increase the susceptibility of some tissues to develop tumors due to exposure to AA. We reported before that 14 d exposure of male F344 rats to up to 50 mg/kg bw/d AA, a dose regimen that results in neurotoxicity, provided no evidence of alterations in the hypothalamic-pituitary-thyroid axis or on forebrain function (Bowyer et al., 2008a,b). The studies presented here focused on the effects of AA on the HPG axis on the same set of animals, to understand better the potential for interference with cellular function that could predispose the peri-testicular mesothelium to the carcinogenic effects of chronic AA exposure. Prominent histopathological alterations were observed in the testes of rats exposed to 50 mg/kg bw/d AA for 14 d, along with modulation of several hormones associated with the reproductive endocrine system, including LH, FSH, testosterone, and progesterone. The intermediate 10 mg/kg/d AA dose produced more modest effects on hormone levels, with significance only for decreased testosterone and increased LH. Testicular germ cell depletion, exfoliation, and apoptosis were observed in animals exposed to 10 or 50 mg/kg bw/d AA, and the incidence and severity of these lesions increased in a dose-related fashion. At the lowest AA dose tested, 2.5 mg/kg bw/d, the only adverse effect observed was an increased incidence of spermatids retention in the seminiferous tubules relative to control.

Previous studies have evaluated toxicity of AA on the testes. The suppression of circulating testosterone following oral exposure to AA has been reported before (Ali et al., 1983; Yang et al., 2005; Hamdy et al., 2011). Our data show that, as the AA dose was increased, not only did the serum testosterone levels decrease, but also that the serum LH levels increased, in a dose-response manner. The concomitant increase in the serum levels of LH coupled with the decreased circulating levels of testosterone suggests that the HPG axis was functional and able to activate the negative feedback response due to the decreased levels of serum testosterone, even in rats exposed to 50 mg/kg bw/d AA. Yang et al. (2005) also observed an AA dose-related decrease of testosterone, both in vivo and in in vitro Leydig cell cultures. Based on their observations that AA decreased the production of testosterone in the in vitro system, these authors suggested that AA directly targeted Leydig cells. Our data further support a direct effect of AA on the testis, rather than on the hypothalamus and/or pituitary. However, the mechanism for testicular effects is unclear but could include inhibition of testosterone synthesis/release and/or interference with gonadotrophic signaling.

Despite the increased levels of serum LH, Leydig cells seemed to be unable to respond appropriately to the gonadotropic signal. The normal response to LH stimulation is either increased number and/or increased size of Leydig cells (Cook et al., 1999). However, there was an apparent decrease in the size of the Leydig cells in the high AA dose group, and no significant changes in the Leydig cell count compared to control. Furthermore, there was no morphological evidence of ongoing cell degeneration (apoptosis or necrosis) or cell proliferation of Leydig cells in the testicular interstitium (as assessed by Ki-67-positive Leydig cell counting). Taken together, these data further indicate that exposure to AA as high as 50 mg/kg bw/d did not result in Leydig cell hyperplasia or depletion. Some potential causes of a defective Leydig cell response to LH stimulation include LH receptors blocked, damaged or depleted; non-functional intracellular cAMP pathway; and inhibition of steroid biosynthesis. In the current study, the expression of the LH receptor gene Lhr was up-regulated by AA treatment, as was the expression of genes encoding enzymes essential for testosterone biosynthesis, such as Star and Cyp11a1. These observations suggest that AA does not decrease the rate of testosterone biosynthesis by inhibition of the expression of cholesterol-transport or steroidogenic enzyme coding genes, but rather that the increased levels of Leydig cells transcripts may be a compensatory response to the decreased levels of testosterone or defects in gonadotrophic signaling.

Evidence has emerged from several studies suggesting that inhibition of some members of the kinesin and dynein cytoskeletal motor protein families may be the common site of action of AA and GA, which could explain both neurotoxicity and male reproductive toxicity observed in animals (Friedman et al., 2008; Sickles et al., 2007; Tyl and Friedman, 2003; Yang et al., 2005). These families of motor proteins function to integrate cellular cytoskeletal elements (microfilaments, intermediate filaments, microtubules) into various structural and functional units that aid in cell–cell adhesions, scaffolding to maintain cell shape and provide support, cell movement or transport, intercellular communication, and translocation of cellular organelles and substrates to facilitate metabolism and biochemical synthesis and secretion (Rogers and Gelfand, 2000; Oshima, 2007). Rat Leydig cells utilize both serum uptake and de novo synthesis of cholesterol from acetate to produce testosterone (Latendresse et al., 1993). It is possible that cytoskeletal inhibition by AA could result in diminished uptake of cholesterol and consequent reduced testosterone levels. De novo synthesis of cholesterol could also be decreased if the mobilization of cholesterol from cytoplasmic lipid droplets and its transport to mitochondria which are thought to involve, in part, organelle transport by the cytoskeletal system (Bilinska et al., 1997), were impaired. In addition, putative AA-induced cytoskeletal dysfunction could inhibit the synthesis and/or transport of LH receptors to the cell membrane of Leydig cells (Laws et al., 1984), thereby indirectly decreasing testosterone biosynthesis.

Defects at the cytoskeleton level could also be related to the seminiferous tubules lesions observed, including spermatid retention and defects in spermatogenesis and Sertoli cell-germ cell tight-junction (Allard et al., 1993; Boekelheide et al., 1989; Fleming et al., 2003a,b; Hall et al., 1992; Neely and Boekelheide, 1988; Russell et al., 1990; Sickles et al., 2007; Sperry and Zhao, 1996). In fact, several of the tubular lesions here described are similar to the ones reported to be induced by 2,5-hexanedione, a toxicant known to disrupt the microtubule-associated motor proteins in rat Sertoli cells (Hall et al., 1995; Boekelheide et al., 2003). Furthermore, Bryant et al. (2008) reported that the most sensitive biomarker of toxicity induced by 2,5-hexanedione in rat testis is spermatid head retention, similar to what we observed upon AA exposure. Hence, it is plausible that the effects of AA on the cytoskeleton and/or associated proteins could be responsible, at least in part, by the testicular lesions observed in our study.

One of the hallmarks of tumorogenesis is the dysregulation of cell proliferation and apoptosis in that an excessive increase in cell proliferation and/or a decrease in apoptosis ultimately may result in tumor formation. Shipp et al. (2006) speculated that the AA-induced testicular mesotheliomas could be associated with Leydig cell tumor progression. In our study, despite the significantly increased levels of LH observed in animals treated with AA 10 and 50 mg/kg bw/d, we did not observe Leydig cells hyperplasia. Rather, the number of Leydig cells seemed to be unaffected by the AA treatment, while the Leydig cell volume decreased. Moreover, our histopathological observations suggest that AA treatment lead to increased germ cell apoptosis. In the case of Stage VII pachytene spermatocytes, this is likely due to the low levels of testosterone rather than a direct effect of AA. The gene expression data indicated an overexpression of the pro-apoptotic genes Casp3 and Fas, and Zhang et al. (2009) and Oshida et al. (2011) have also reported that exposure to AA induces germ cell apoptosis in mice. We further assessed the cell proliferation index at the peri-testicular mesothelium, the direct target tissue of AA carcinogenesis. The LI for Ki-67 immunopositive mesothelial cells lining the scrotum and covering the testis and epididymis in the high AA dose group (50 mg/kg bw/d) showed a significant 2-fold decrease in proliferative activity compared to control (Fig. 4). This finding is consistent with our previously reported data, that demonstrated an AA-induced 5- to 40-fold decrease in the Ki-67 LI of parenchymal epithelia of the thyroid gland, anterior pituitary gland, and liver (Bowyer et al., 2008a,b). Hence, despite the observed AA-induced testicular lesions and reduced levels of circulating testosterone, our data provide no evidence of increased cell proliferation that would be predicted to enhance the carcinogenic process in testicular mesothelium in any of the male reproductive tissues analyzed at doses in the range of those that induce tumorigenesis from chronic exposure, or even at a maximally tolerated dose of AA.

While abnormal functioning of the mitotic spindle would be expected to decrease cell proliferation, it has been proposed that chromosomal aberrations and abnormal mitotic events caused by AA treatment could induce neoplasms in F344 rats (Friedman et al., 2008). Mitotic arrest was reported to occur after exposure of HT1080 cells at 1–10 mM concentrations of AA and inhibition of bovine brain kinesin was observed at concentrations of AA and GA at or above 100 μM (Friedman et al., 2008). However, the observation of tumorigenesis in only a small subset of tissues from AA-treated F344 rats, and at doses of AA that produce maximal serum and tissue concentrations of AA and GA below 2 μM (see Section 3 above), suggest that an anti-mitotic mechanism is unlikely.

5. Summary

The mode of action of AA and GA toxicity on the male reproductive system is complex and remains unclear. We evaluated the effect of orally administered AA on an array of testicular endpoints, including histopathology, cell proliferation, gene expression, and serum hormones. While several of these endpoints were significantly altered by AA treatment, our data support the existence of a functional hypothalamic-pituitary-testicular axis in rats exposed to AA at daily doses as high as 50 mg/kg bw. The AA-induced changes in testicular histopathology, gene expression, and serum hormones were accompanied by a significant decrease in the levels of cellular proliferation in the peri-testicular mesothelium, a target tissue for AA-induced carcinogenesis in male F344 rats. Although the results from the present studies did not identify tissue-specific cellular mechanisms that predispose the peri-testicular mesothelium of F344 rats to the carcinogenic effects of AA, they did not support disruption of endocrine signaling with increased cellular proliferation as contributory factors in the genotoxic effects of AA treatment mediated by GA.

Supplementary Material

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.toxlet.2012.03.007.