Morphological and biochemical changes in the Harderian gland of hypothyroid rats

The Harderian gland (HG) is an orbital gland present in most groups of terrestrial vertebrates. It is regulated by many exogenous and endogenous factors that vary according to the animal species (for reviews, see Payne, 1994; Chieffi et al., 1996). Numerous functions have been attributed to this gland, including lubrication of the eye and nictitating membrane, thermoregulation(Thiessen, 1988; Shanas and Terkel, 1996) and photoprotection (Hugo et al.,1987; Spike et al.,1990). Furthermore, it is part of the retinal–pineal axis(Hoffman et al., 1985) as well as a source of either pheromones or growth factors(Seyama et al., 1992; Shanas and Terkel, 1996). In some mammals, the HG activity is influenced by endogenous factors (such as prolactin, thyroid hormones and steroid hormones) and exogenous factors (such as light and temperature) (Hoffman et al.,1989a; Buzzell and Menendez-Pelaez, 1992; Buzzell et al., 1992; Menendez-Pelaez et al., 1993; Buzzell et al.,1994). During the 1940s, the interrelationship between the thyroid gland and the HG was suggested by several researchers(Smelser, 1943; Boas and Bates, 1954; Boas and Scow, 1954) who used glandular mass as their main study end point. These studies also suggested species differences: guinea pigs reacted to thyroidectomy with HG hypertrophy(Smelser, 1943) whereas the HGs of thyroidectomised rats decreased in mass(Boas and Bates, 1954). Decades later, thyroxine (T4) injections were found to hasten the appearance of new porphyrins in the HG of neonatal rats(Wetterberg et al., 1970). Recently, we demonstrated that 3,5,3′-triiodothyronine (T3)administration to adult rats induces HG hypertrophy(Chieffi Baccari et al.,2004). T4 injections lead to an increase in porphyrin content in female hamsters, an effect that is reversed under conditions of thyroid hormone deficiency (Hoffman,1971; Hoffman et al.,1989a; Hoffman et al.,1989b; Hoffman et al.,1990). In addition, hypothyroidism induces a consistent increase in N-acetyl transferase (NAT) activity in male hamster HG(Buzzell et al., 1989) as well as in porphyrin levels (Hoffman et al.,1990). Furthermore, Hoffman and co-workers(Hoffman et al., 1990) have demonstrated that thyroid hormones act directly on the hamster HG rather than indirectly through modification of thyroid-stimulating hormone (TSH) synthesis and/or release. Nuclear receptors for T3 have been described in the HG of male and female golden hamsters (Vilchis and Pérez-Palacios, 1989). The presence of type II thyroxine 5′-deiodinase (5′D), an enzyme that converts thyroxine to triiodothyronine, in both rat and hamster HGs, provided further evidence that this gland is a target for thyroid hormones(Delgado et al., 1988; Osuna et al., 1990). Following thyroidectomy, the activity of 5′D in rat HG increases and exhibits a marked circadian rhythm (Guerrero et al.,1987). However, the effects of hypothyroid status on the morphofunctional characteristics of rat HG remain to be elucidated.

In this study, we investigated the effects of experimentally induced hypothyroidism on the morphology of rat HG by histochemistry and ultrastructural analyses. In addition, since we found that some morphological signs are indicative of cell death by apoptosis, we compared both mitochondrial and cytosolic expression of cytochrome c [the principal initiator molecule released from mitochondria to cytosol to initiate the apoptotic program in cells (Green and Reed, 1998)] in euthyroid and hypothyroid rat HGs.

Male Wistar rats, Rattus norvegicus albinus Berkenhaut 1769,weighing 300–350 g, were kept under regulated conditions of temperature(28°C) and light (12 h:12 h L:D cycles). They received commercial food pellets (Mil-Rat, Morini, Italy) and water ad libitum. To induce hypothyroidism, rats (N=8) were injected intraperitoneally (i.p.)with 6-n-propylthiouracil (PTU) (1 mg 100 g^–1 body mass)daily for 4 consecutive weeks. In addition, once a week, they received an i.p. injection of iopanoic acid (IOPA) (6 mg 100 g^–1 body mass)(Lanni et al., 1996). Control rats (N=8) received saline injections. At the end of the treatment,rats were first anesthetised by an i.p. injection of chloral hydrate (40 mg^–1 body mass) and then decapitated. The trunk blood was collected, and the serum was separated and stored at –20°C for later T3 and TSH determinations. The HGs were dissected out, weighed and rapidly immersed in liquid nitrogen for cytochrome c determination or for cryostat sectioning. Pieces of glands were quickly immersed in fixative for light or electron microscopy, as described below. Total T3 levels were determined in 50 μl samples of serum using reagents and protocols supplied by BD Biosciences (Franklin Lakes, NJ, USA). Serum TSH levels were measured by radioimmunoassay (RIA) using materials and protocols supplied by Amersham(Milan, Italy). Experiments were performed in accordance with local and national guidelines governing animal experiments.

Pieces of HGs were rapidly immersed in Bouin's fluid. Then, paraffin sections (5 μm thick) were stained with trichrome Mallory stain. A histochemical test for proteins was carried out using the mercury–Bromophenol Blue method, and the mucosubstances were detected with Alcian Blue–PAS (AB/PAS). For lipid detection, formol-calcium-fixed frozen sections (5–10 μm) were stained with Sudan Black B.

For electron microscopy, pieces of HGs (3 mm³) were promptly immersed and left for 3 h in Karnovsky's fixative in cacodylate buffer (pH 7.4) and then postfixed for 2 h in cacodylate buffer containing 1% osmium tetroxide. The samples were dehydrated through a graded ethanol series and finally embedded in Epon 812. Ultrathin sections, stained with 4% uranyl acetate and then with 1% lead citrate, were examined using a Philips 301 transmission electron microscope (Philips Electronic Instruments, Rahway,NJ).

The terminal uridine deoxynucleotidyl transferase dUTP nick end labeling(TUNEL) reaction was performed on paraffin sections (5 μm thick) of HG using an in situ cell death detection kit (Roche Applied Science,Mannheim, Germany). Specifically, sections were pretreated with the permeabilisation solution (0.1% Triton X-100, 0.1% sodium citrate). Then, they were rinsed twice in PBS and incubated with a mixture containing terminal deoxynucleotidyl transferase (TdT) (1:20) and fluorescein-labelled dUTP, for 60 min at 37°C in a dark humidified chamber. After washing in PBS, the slides were examined under a fluorescence microscope.

HG mitochondria were isolated after homogenisation in an isolation medium consisting of 220 mmol l^–1 mannitol, 70 mmol l^–1 sucrose, 20 mmol l^–1 Tris–HCl, 1 mmol l^–1 EDTA, 5 mmol l^–1 EGTA, and 5 mmol l^–1 MgCl2, pH 7.4 (all from Sigma-Aldrich Corp., St Louis,MO, USA), supplemented with a protease inhibitor cocktail (Roche Applied Science, Mannheim, Germany). After homogenisation, samples were centrifuged at 700 g for 10 min and supernatants were collected and transferred into new tubes for subsequent centrifugation at 17000 g. The obtained mitochondrial pellet was washed twice and then resuspended in a minimal volume of isolation medium and kept on ice. The supernatant contained the cytosolic fraction(Singh et al., 2003).

Protein concentrations of mitochondrial and cytosolic fractions were estimated using a modified Bradford assay (Bio-Rad, Melville, NY, USA).

Proteins from both mitochondrial and cytosolic fractions (30 μg each)were boiled in Laemmli buffer for 5 min. Afterwards, the samples were subjected to SDS–PAGE (13% polyacrylamide) under reducing conditions. Analysis of mitochondrial and cytosolic samples was performed on two separate gels. After electrophoresis, proteins were transferred to a nitrocellulose membrane. Each membrane was treated for 1 h with blocking solution (TBS/T: 5%non-fat powdered milk in 25 mmol l^–1 Tris, pH 7.4; 200 mmol l^–1 NaCl; 0.5% Triton X-100) and then it was incubated overnight at 4°C with a rabbit anti-human polyclonal antibody against cytochrome c (Cell Signalling Technology, Inc., Danvers MA, USA)diluted 1:1000. After washing with TBS/T and TBS, membranes were incubated with the horseradish-peroxidase-conjugated secondary antibody (1:4000) for 1 h at room temperature. The reactions were detected using a chemiluminescence protein detection method based on the protocol supplied with a commercially available kit (NEN Life Science Products, Boston, MA, USA). In both mitochondrial and cytosolic fractions, cytochrome c levels were determined using two preparations containing three glands from three different rats. The amount of cytochrome c protein was quantified by employing a Bio-Rad Molecular Imager FX using the supplied software (Bio-Rad Laboratories, Milan, Italy). The values obtained were compared by ANOVA followed by an unpaired t-test (for between-group comparison). All data were expressed as the mean ± s.d. The level of significance was set at P<0.01.

The combined effect of PTU and IOPA produced hypothyroid rats with T₃ levels significantly lower (0.13±0.02 nmol l^–1) than those observed in euthyroid rats (0.98±0.05 nmol l^–1). The serum TSH level was much greater in hypothyroid rats (20.1±1.2 mi.u. l^–1) than in euthyroid rats (4.3±0.4 mi.u. l^–1). HG mass was lower in hypothyroid rats (0.15±0.08 g) than in saline-injected rats(0.22±0.03 g).

The rat HG is a tubulo-alveolar gland surrounded by a connective tissue capsule. In the HG of euthyroid rats, the glandular epithelium had pyramidal cells with basal nuclei and large lumina(Fig. 1A). There are no morphological differences between the sexes in rat HG (for reviews, see Payne, 1994; Chieffi et al., 1996). We found that the morphology of the HG of hypothyroid rats(Fig. 1B) differed considerably from that of euthyroid rats (Fig. 1A). Indeed, the HG of hypothyroid rats was characterised by glandular cells that displayed pale secretory granules in the cytoplasm. Furthermore, the acinar lumina were filled with secretory granules mixed with nuclei and cytoplasmic fragments of glandular cells(Fig. 1B).

The acinar cells of euthyroid HG were positive for the Bromophenol Blue reaction (Fig. 1C) and weakly positive for the AB/PAS reaction (Fig. 1E). Large Sudan Black-positive vacuoles and porphyrin accretions were present in glandular cells of euthyroid HG(Fig. 1G). Glandular cells of hypothyroid rats were negative or weakly positive for Bromophenol Blue(Fig. 1D). The basal nuclei often appeared pycnotic. Some acini lost their normal structure and appeared detached from the extracellular matrix(Fig. 1D). Scarce connective tissue among acini was present. Coalescence of acini, as well as holocrine secretion, could be observed (Fig. 1D). Hypothyroid glandular cells were positive for AB/PAS and the lumina were filled with a strongly positive AB/PAS secretion(Fig. 1F). A few small Sudan Black-positive vacuoles were observed outside the acini(Fig. 1H).

As largely reported, two cell types are present in rat HG. Type A cells have large secretory vacuoles and type B cells have a few minute droplets (for reviews, see Payne, 1994; Chieffi et al., 1996). Electron micrographs revealed marked differences in the shape of glandular cells between euthyroid and hypothyroid rats. In euthyroid rats, type A cells(Fig. 2A) had a large number of cytoplasmic vacuoles containing a moderately electron-dense substance and basal nuclei with evident nucleoli. Numerous mitochondria with tubular cristae and glycogen granules were observed (Fig. 2B). In hypothyroid rats, type A cells had heterogeneously shaped cytoplasmic vacuoles (Fig. 3A). They were almost devoid of secretory products. Mitochondria with a few cristae often showed a condensed configuration and vacuolisation(Fig. 3B). A few glycogen granules were seen throughout the cytoplasm(Fig. 4A). Occasionally,bodies, interpreted as autophagic vacuoles, were observed(Fig. 4A). Some granules had a characteristic lamellar substructure (Fig. 4B) and they often formed patterns resembling myelin forms(Fig. 4C).

Type B cells from euthyroid rat HG were characterized by empty cytoplasmic vacuoles and basal nuclei (Fig. 5A). Furthermore, these cells, compared to type A cells(Fig. 2A), had an abundant smooth endoplasmic reticulum (SER) and a larger number of mitochondria(Fig. 5A). In hypothyroid rats,type B cells had dark nuclei with peripheral condensation of chromatin and fewer cytoplasmic vacuoles than did euthyroid B cells(Fig. 5B). Loose SER, as well as numerous mitochondria, could be observed throughout the cytoplasm(Fig. 5B).

As shown in Fig. 6, some cells (types A and B) of the hypothyroid rat HG displayed strong signs of degeneration, including condensation of chromatin and pycnosis, profiles indicative of coalescence of secretory vacuoles(Fig. 6A), amorphous cytoplasm,and alteration of mitochondria (Fig. 6B).

Since we found morphological signs of cell degeneration in the HG from hypothyroid rats, we investigated apoptosis by the TUNEL technique. Whereas some nuclei from hypothyroid rat HG were labelled by fluorescein(Fig. 7A), the euthyroid HG nuclei were not (Fig. 7B).

Cytochrome c expression in mitochondrial and cytosolic fractions of euthyroid and hypothyroid rat HGs was compared. In the euthyroid state, the intensity of the cytochrome c band in the mitochondrial pellet was significantly (P<0.01) higher than that in cytosolic fraction(Fig. 8). Conversely, under hypothyroid conditions, the cytochrome c band was significantly more intense (P<0.01) in the cytosolic fraction than in the mitochondrial pellet (Fig. 8).

The combined treatment with PTU and IOPA produces rats with both systemic and local hypothyroidism. Indeed, PTU blocks thyroidal hormone synthesis, via an inhibition of thyroid peroxidase activity and it is also a strong inhibitor of type I 5′-d-deiodinase activity(Oppenheimer et al., 1972; Leonard and Visser, 1986). IOPA inhibits all three types of deiodinase enzymes; however, whereas its effect is strong on type II and III, it is comparatively weak on type I(Kaplan and Utiger, 1978; St Germain, 1994). In the present study, the most notable histological effect of hypothyroidism in the rat HG was the decrease in cytoplasmic lipoproteic secretory granules and the increase in secretion in the lumina. The release of granules, together with cellular debris and some pycnotic nuclei in the lumina, suggests a holocrine type of secretion. This phenomenon was rarely observed in the HG of euthyroid animals in which merocrine secretion is prevalent(Woodhouse and Rhodin, 1963; Abe et al., 1980; Sakai, 1981; Satoh et al., 1992). Hence,the decreased weight of the HG in the hypothyroid condition could be ascribed to the massive release of glandular secretion. Consistently, we have previously demonstrated that the increased mass of the HG in hyperthyroid rats is due to the activation of the secretory activity induced by T3, which,indeed, provokes an accumulation of lipoproteic secretion into the cells and lumina (Chieffi Baccari et al.,2004).

Interestingly, although the rat HG secretes predominantly lipoproteic substances, histochemical tests revealed the presence of mucosubstances in the acinar lumina of hypothyroid rat HG, suggesting that under such conditions a biochemical modification of HG secretion takes place. An increase in the glycosaminoglycan content has been demonstrated in guinea pig HG after thyrotropin injections (Winand and Kohn,1973). Therefore, mucosubstances accumulation in the HG of hypothyroid rats could be an effect of increased TSH serum levels.

The ultrastructural changes observed in hypothyroid rat HG could be attributed to the alterations in glandular activity elicited by thyroid hormone deficiency, thereby leading to eventual cell death. The electron microscopy revealed that in the final stages of cellular damage some cells had pycnotic nuclei and alterations in the cytoplasmic architecture without any accompanying swelling and rupture of nuclear and cytoplasmic membranes,indicative of necrosis. It is well known that cytochrome c is released from mitochondria in response to apoptotic stimuli(Green and Reed, 1998). We found cytochrome c translocation from mitochondria to the cytosol in hypothyroid rat HG. Hence, the alteration in mitochondrial morphology observed in hypothyroid rat HG may be due to the translocation of apoptogenic molecules. Consistently, as evidenced by TUNEL, DNA fragmentation in the gland of hypothyroid rat indicates the occurrence of apoptosis.

Interestingly, besides the occurrence of apoptosis in the hypothyroid rat HG, we also observed autophagic activity. Indeed, autophagic vacuoles and lipofuscins were detected and interpreted as the end result of autophagy. Our interpretation was based on recent studies describing an alternative pathways for active self-destruction called programmed cell death (PCD) type II or autophagy. Apoptosis, or PCD type I is characterised by condensation of cytoplasm and preservation of organelles, essentially without autophagic degradation. Autophagic cell death exhibits autophagic degradation of Golgi apparatus, polyribosomes, and endoplasmic reticulum, events that normally precede nuclear destruction (for a review, Bursch, 2001). Since apoptotic and autophagic PCD were are not mutually exclusive phenomena they could actually coexist in the hypothyroid rat HG. Autophagy has been recently described in hamster HG as a survival mechanism for fighting against cell damage resulting from physiological oxidative stress(Tomàs-Zapico et al.,2005).

Further studies are still needed to clarify the molecular mechanism underlying programmed cell death in rat HG under conditions of hypothyroidism. Although it is well known that thyroid hormone deficiency leads to extensive apoptosis during cerebellar development, the mechanisms by which it occurs still remain unclear. What is known is that thyroid hormones seem to preserve the mitochondrial architecture by inhibiting the release of apoptogenic molecules that can eventually lead to excess apoptosis(Singh et al., 2003). Recently, a possible involvement of pronerve growth factor–p75 neurotrophin receptor pathway in hypothyroidism-enhanced apoptosis has been demonstrated during rat cortical development(Kumar et al., 2006). In addition, it has been demonstrated that hypothyroidism increases the level of p73, a protein involved in apoptosis, during rat liver regeneration(Alisi et al., 2005).

In conclusion, our findings indicate that the neuroendocrine–thyroid axis plays a pivotal role in preserving the structural integrity of rat HG and, therefore, its secretory activity. Without doubt, further studies are indispensable to clarify the mechanisms responsible for the morphological and biochemical changes in the HG and, consequently, the interplay between this gland and thyroid hormone deficiency.

LIST OF ABBREVIATIONS

 
• AB/PAS

  Alcian Blue–PAS

 
• dUTP

  uridine deoxynucleotidyl transferase

 
• HG

  Harderian gland

 
• IOPA

  iopanoic acid

 
• PCD

  programmed cell death

 
• PTU

  6-n-propylthiouracil

 
• T₃

  3,5,3′-triiodothyronine

 
• TdT

  terminal deoxynucleotidyl transferase

 
• TSH

  thyroid-stimulating hormone

 
• TUNEL

  terminal dUTP nick end labeling

We thank Mr Franco Iamunno for technical assistance and Dr Paola Merolla for the editorial revision of the manuscript.