MATERNAL SUBCLINICAL AND CLINICAL HYPOTHYROIDISM EFFECTS ON RAT OFFSPRING: A STORY OF THE SKIN AND ITS DERIVATIVES

Epidermis stem cells have a crucial role through the processes of proliferation and differentiation, to replace cells that are constantly lost during tissue turnover or following injury. On the other hand, thyroid hormones regulate the proliferation and differentiation of epidermal cells and thus signi ﬁ cantly in ﬂ uence the homeostasis of the skin. It is well known that maternal hypothyroidism during pregnancy leads to impaired development of many organ systems in their offspring. However, there is a lack of data about the in ﬂ uence of maternal subclinical hypothyroidism during pregnancy and lactation on the development of the skin and its derivatives in the litter


INTRODUCTION
The skin epidermis protects animals against major environmental stresses and has the remarkable ability to enrich the body surface with structures such as hair follicles, claws, sebaceous and sweat glands in mammals or scales and feathers in lower vertebrates [1].
The skin epidermis and its array of appendages undergo ongoing renewal, which is a key feature of skin tissue homeostasis.Stem cells (SCs) in the epidermis have a crucial role in the maintenance of this important barrier through the processes of proliferation and differentiation.Interestingly, the basic mechanisms and signaling pathways that orchestrate epithelial morphogenesis are reused during adult life to regulate skin homeostasis [2].Proliferating cell nuclear antigen (PCNA) is a frequently used marker for the detection of epidermal cells proliferation [3] and caspase-3 (Casp3) is a critical executioner of apoptosis, as it is either partially or totally responsible for the activation of many key proteins in the apoptosis pathway, via the mechanism of proteolytic cleavage [4].Different SCs reside in specifi c niches such as interfollicular epidermis, hair follicles and sebaceous glands, that provide the appropriate microenvironment and molecular cues to preserve their proliferative and tissue regenerative potential [5].In the hair follicles (HFs), SCs reside in a relatively quiescent state within an anatomically distinct region known as the bulge.Bulge cells express the embryonic stem cell transcription factor NANOG, which helps maintain the capacity of selfrenewal and multipotency in SCs [6,7].In adult rats, bulge SCs can contribute to all three epithelial lineages of the skin [8,9].During normal homeostasis, bulge SCs are periodically activated to fuel postnatal hair cycles.Interfollicular epidermis (IFE) and sebaceous gland (SG) both contain resident populations of unipotent progenitor cells [10].
By binding to their receptors, thyroid hormones regulate the proliferation and differentiation of epidermal cells and thus signifi cantly infl uence the homeostasis of this largest organ [11].Maternal hypothyroidism down regulates thyroid hormone receptors in the offspring's skin [12], thus disabling triiodothyronine (T3) effects, as one of the main regulators of epidermal growth and differentiation [13].Numerous studies have shown that the altered function of the maternal thyroid gland during pregnancy and lactation can affect the development of various fetal organs [14,15].However, scarce research has been conducted on the effect of maternal thyroid hormones on the development of skin and its derivatives during prenatal and early postnatal development.In particular, there is a lack of data about the infl uence of subclinical hypothyroidism of mothers during pregnancy on the development of the skin and its derivatives in the litter.It has been documented that subclinical and clinical hypothyroidism in mothers during pregnancy and lactation leads to offspring subclinical and clinical hypothyroidism [15][16][17].Clinical hypothyroidism in mothers is characterized by rough coat, prolonged pregnancy and small litter size [15,18,19].In the pups, the overt form of hypothyroidism results in reduced body mass, lower thyroid hormone levels and higher thyroid index of activation [15].
The aim of this study was to investigate the effects of maternal thyroid dysfunction on the development of the skin and its derivatives in their offspring in the early postnatal period.

Animals
Ten-week-old female Albino Oxford rats (weight 140-160g), obtained from the Department of Laboratory and Experimental Care and Use of Animals Unit of the Institute of Medical Research, Military Medical Academy (Belgrade, Serbia), and their male pups were used in this experiment.Only male pups were used because of small litter size, higher number of males and the fact that female pups were used for investigating effects of maternal hypothyroidism on ovary development [16,17].The animals were kept under a photoperiodic cycle of 12 h light / 12 h dark in an air-conditioned facility.The mean temperature was 21 ± 0.5°C.Pelleted rat feed (PA 20, Veterinary institute Subotica) and drinking water were available ad libitum.The experiment was approved by the Ethical Committee of the Faculty of Veterinary Medicine University of Belgrade (decision number 01-20/6) and by the Ministry of agriculture, forestry and water management -veterinary administration (decision number 323-07-00364/2017-05/8).

Experimental procedure
Antithyroid substance 6-n-propyl-2-thiouracil (PTU) (Sigma Chemical Co., St. Louis, MO, SAD) dissolved in drinking water, was used for inducing subclinical (low dose 1,5 mg/l) and overt hypothyroidism (high dose 150 mg/l).Water with PTU was administered from the beginning of pregnancy and during lactation.
Female Albino Oxford rats had a 7-day adaptation period, after which vaginal smears were made every day to determine the phase of the sexual cycle.After proestrus or estrus phase was determined, the females were mated.The presence of sperm in vaginal smears after mating or presence of the mating plug were considered as gestational day zero.Does were then separated from males and randomized into three groups, each consisting of six animals: control group (C), low dose PTU group (PTU low ), and high dose PTU group (PTU high ).
From each experimental group, six male pups were euthanized within 24h of delivery (neonatal pups) by decapitation using surgical scissors.Seven days after delivery, six more male pups (early infantile pups) from each group were euthanized by cervical dislocation.Does were euthanized at the same time point by using 100 mg/kg body weight Euthasol Euthanasia Solution (Produlab Pharma Production B.V.Raamsdonksveer, Netherlands).

Tissue collection and processing for light microscopy and histochemistry
To avoid wrinkling of the skin during the fi xation process, the carcasses were placed in 10% neutral buffered formalin for six hours, and thereafter the back skin was taken and placed in formalin for eighteen more hours, after which processing, and paraffi n embedding were done by routine laboratory procedures.Serial sections (5 μm thickness) were made using a rotary microtome (Reichert, Wien, Austria).Sections were stained with haematoxylin/eosin (H/E) (Merck Millipore, Darmstadt, Germany) and Van Gieson staining method, to ensure better visualization of the keratinized layer of the epidermis [20].The sections were then mounted using DPX (phthalate-free) mounting media (Fisher Scientifi c, Loughborough, UK).
For semi-thin sections, skin samples were fi xed in cold 4% glutaraldehyde, followed by fi xation in 1% buffered osmium tetroxide, dehydrated in acetone, embedded in Araldite, and 1 µm thick sections were stained with toluidine blue.Cytological characteristics of the epidermis and hair follicles were examined using a light microscope (Olympus CX31, Münster, Germany).

Immunohistochemistry
All immunohistochemical analysis was done according to the following procedure: Antigen unmasking was performed with 0,1M citrate buffer (pH 6) in a microwave for twenty minutes.Inactivation of endogenous peroxidase was achieved with 3% H 2 O 2 for 10 minutes.Sections were then washed in PBS and blocked using a protein block (5% goat serum) for 10 minutes.After the wash in PBS, sections were incubated with primary antibodies (Table 1) diluted in IHC Diluent (Novocastra, Leica Biosystems, Newcastle, UK) overnight at 4°C.For negative controls, primary antibodies were omitted.Primary antibodies signal amplifi cation and visualization were performed using EnVision FLEX/HRP (RTU, Dako, Santa Clara, CA, USA) followed by Liquid DAB+ Substrate Chromogen System (Dako, Carpinteria, CA, USA).Counterstain was carried out with Mayer's haematoxylin.

Histomorphometry and stereology
Histomorphometry and stereological analyses were done on serial sections using a microscope equipped with a digital camera and adequate software (Olympus CX31 with UC50 Soft Imaging Solutions camera and SensEntry 1.13 software, Münster, Germany).
The thickness of the non-cornifi ed layer of the epidermis was measured using a vertical line drawn from the basal cell layer to the end of the granular layer of the epidermis.The thickness of the cornifi ed layer was measured using a vertical line drawn through the cornifi ed layer.These measurements were performed on three different sections, on 20 fi elds of view per section.
Stereological analysis of the dermis was done using a multipurpose stereological grid M42 [21].The number of view fi elds per section was determined using a formula as reported by Kališnik [22].The volume density of hair follicles (V vdf ) was calculated using the formula: V vdf =P f /P t (P f -number of testing points hitting the hair follicle; P t -total number of testing points in the multipurpose stereological grid M42).V vdf is expressed as a percentage of hair follicles in relation to the entire volume of the dermis.
The number of PCNA-positive (PCNA + ) cells was expressed per 1mm 2 of the epidermis area.The number of PCNA + cells, as well as Casp3-positive (Casp3 + ) was expressed per 1mm 2 of hair follicle area.The number of Casp3 + hair follicles was also determined and expressed per 1mm 2 of dermis area.
The number of NANOG-positive (NANOG + ) hair follicles was determined and expressed per 1mm 2 of dermis area.The count of multipotent SCs which were NANOG + was determined and expressed per 1mm 2 of hair follicle area.
All counting of primary antibody positive cells was performed on fi ve different sections, on 10 fi elds of view per section.

Statistical analysis
Two independent researchers reviewed all slides twice for morphological and stereological analysis and three times for immunohistochemical evaluations.Values above and below the confi dence interval (5-95%) were excluded from further processing of results.To determine statistically signifi cant differences between the examined groups, within the framework of the general linear model, analysis of variance for repeated measurements was used.Results were expressed as mean ± standard deviation.Levels of signifi cance bellow 0.05 were considerd signifi cant.

Histology, morphometry and stereology
The skin sections of neonatal pups (T0) showed a formed epidermis and de rmis with HFs (Fig. 1A-F).Epidermis o f C-T0 had a clear boundary between the epithelium and the dermis, a layer of basophilic basal keratinocytes was formed.Stratum spinosum was also formed and spines with desmosomes were observed between the polygonal cells, and between the polygonal cells and the cells of the basal layer.In the stratum granulosum, the cells became fl attened with many osmium-positive granule s and keratohyalin granules.Upwards to the stratum corneum, as a result of terminal differentiation, the cells became irregularly shaped, and thinner, so that in the stratum corneum they became dehydrated corned coots (Fig. 1G).In comparison with the C-T0 group, the thickness of non-cornifi ed layers of the epidermis was not altered in PTU low -T0 group (Table 2; Fig. 1D, E), while there was a statistically signifi cant increase in the thickness of the stratum corneu m (Table 3; Fig. 1D, E).The thickness of noncornifi ed layers was reduced, while stratum cor neum was t hicker in PTU high -T0 group, compared to the same epi dermal layers of the pups from the contro l group (Table 2 and 3, Fig. 1D, F).Compared to the control, both treated groups, did not have a clear dermoepidermal junction and keratinocytes of the basal and spinous layers were irregular in size and shape (Fig. 1G-I).While a perinuclear halo zone was observed in some cells of the stratum spinosum in the control and PTU low -T0 groups, in the PTU high -T0 groups a perinuclear halo zone was present in the majority of cells in this layer of the epidermis (Fig. 1G-I).Abundance of keratohyalin and osmium-positive granules were present in the cells of the stratum granulosum (Fig. 1G-I).In the dermis of the pups from the PTU low -T0 group, as well as PTU high -T0 there was a reduced number of HF, compared to the C-T0 group (Fig. 1A-C).All HFs of hypothyroid pups from both groups were positioned solely in the papillary layer of the dermis (Fig. 1B, C), whereas HFs of the pups from the control group extended from the surface of the epidermis to the subcutaneous muscle (Fig. 1A).The marked delay of HF morphogenesis was present in both treated groups.Stereological analysis of the dermis of hypothyroid pups from both PTU low -T0 and PTU high -T0 groups, showed a statistically signifi cant reduction of the V vdf of HFs. in comparison with the C-T0 group (Fig. 1J).The skin sections of early infantile pups (T7) also showed a formed epidermis and dermis with HFs that are in the later stages of morphogenesis (Fig. 2A-F).In PTU high -T7 group there was a signifi cant reduction in thickness of non-cornifi ed layers of the epidermis, with an increase in thickness of the stratum corneum in comparison with both C-T7 and PTU low -T7 groups (Table 2 and 3; Fig. 2D-F).The epidermis of C-T7 had normal stratifi ed organization of keratinocytes, like newborn controls (Fig. 2G), while in the PTU low -T7 group keratinocytes in the stratum basale were hypertrophic, irregular in shape, with a halo zone around the nucleus (Fig. 2H).Towards the surface, keratinocytes, with a blebbing membrane and irregular nuclear envelope in stratum spinosum, became fl attened and accumulated keratohyalin and osmium-positive granules in the stratum granulosum (Fig. 2H).Flattening of stratum basale, a marked decrease in the number of layers of the stratum spinosum with altered morphology of keratinocytes and hyperthickening of stratum corneum are characteristics of PTU high -T7 group (Fig. 2I).
In the dermis of the pups from PTU low -T7 group the number of HFs was reduced, and the pups from the PTU high -T7 group had a signifi cantly reduced number of HFs in comparison with both C-T7 and PTU low -T7 groups (Fig. 2A-C).Statistically signifi cant reduction of the V vdf was present in both PTU low -T7 and PTU high -T7 groups in comparison with the C-T7 group.There was also a statistically signifi cant reduction of the V vdf present in the PTU high -T7 group in comparison with the PTU low -T7 group (Fig. 2J).

Immunohistochemistry
The proliferative activity of keratinocytes was determined based on the number of PCNA + cells in the germinative layer of the epidermis (Fig. 3A-F).Proliferative activity in both PTU low -T0 and PTU high -T0 groups was not decreased in comparison with the  C-T0 group (Fig. 3D-F, M).In the T7 pups there was a statistically signifi cant decrease in the number of PCNA + cells in the germinative layer of the epidermis in both PTU low -T7 and PTU high -T7 groups in comparison with the C-T7 group (Fig. 3J-M).Also, the number of PCNA + cells of the epidermis was decreased in the PTU high -T7 group in comparison with the PTU low -T7 group (Fig. 3K-M).
The number of PCNA + cells in the HF of the control group pups was increasing with the pups' age (Fig. 3A, G).There was no signifi cant decrease in the number of PCNA + cells of the HF in both PTU low -T0 and PTU high -T0 groups in comparison with the C-T0 group (Fig3A-C, N).Still, in T7 pups from both treated groups the decrease in the number of PCNA + cells of the HF were statistically signifi cant between those experimental groups, as well as in relation to the C-T7 group (Fig. 3G-I, N).
The expression of Casp3 was not detected in HF of T0 experimental groups (Fig. 4A-C).However, in treated groups PTU low -T7 and PTU high -T7 the number of Casp3 + cell in the external hair sheet or in the associated sebaceus glands and the number of Casp3 + HF per 1mm 2 of the dermis area were signifi cantly increased in comparison to the C-T7 group (Fig. 4D-H).
The overall number of HF in the dermis of the T0 and T7 pups from the PTU low and PTU high groups was decreased.However, the number of the NANOG + HFs was signifi cantly increased in both treated T0 groups compared to the C groups (Fig. 5A-C,  G).The number of NANOG + HFs was also signifi cantly increased in PTU high T0 group, compared to the PTU low T0 group (Fig. 5B, C, G).Similar signifi cant increases in number of NANOG+ HFs were observed in T7 treated groups (Fig. 5D-G).Also, the expression of NANOG + HF bulge SCs signifi cantly increased in treated groups of both age categories (Fig. 5B, C, E-H), when compared to the C group (Fig. 5A, D,  H).

DISCUSSION
This study used the rat as an animal model, previously proved appropriate for hypothyroidism investigation [23] and though histological and stereological approach is a basic way to access the subject, the obtained results might be a good starting point for further understanding the effects of the lack of thyroid hormones on skin cells at the ultrastructural and molecular level.The main fi ndings of our investigation were that the subclinical and clinical form of hypothyroidism in pups lead to serious damage of the epidermis in terms of pronounced hyperkeratosis and reduction of the germinal layer along with reduced number of HFs.
Proliferation and differentiation are essential for normal skin structure and homeostasis and they take place alternately, diffusely and they overlap in some epidermal layers [24,25].Proliferation is more pronounced in the germinal layer, with differentiation in the higher ones, while apoptosis is diffusely present in all layers of the epidermis [24].On the other hand, T3 is proven as a suppressor of Casp3 + activity and an important antiapoptotic factor [26].The lack of T3 results in impaired synthesis of extracellular matrix proteins like laminin [23] and nuclear matrix proteins like PCNA, in charge for cell cycle control [27], both with a key role in connection, growth, proliferation and distribution of epithelial cells like keratinocytes [25].Thus, the changes are understandably more pronounced in the pups from mothers with overt hypothyroidism (T0 and T7) where amplifi ed apoptosis and decreased proliferation result in reduced thickness of the non-cornifi ed layer.Furthermore, our observation of the cornifi ed layer light thickening in the treated groups could be associated with the study of keratinocytes cell culture which indicated the promoting effect of T3 on terminal differentiation and shedding of these cells [28].We presume that the reduced concentration of thyroid hormones thus leads to the decreased desquamation rate and since this is the only study to our knowledge associating the thyroid hormones and cornifi cation, further insight to this subject is needed.No changes regarding epidermal thickness and cell morphology are detected in pups obtained from mothers with subclinical hypothyroidism immediately after birth.Still, it appears that there is a temporal demand for subtle changes like reduced proliferation and differentiation to develop, as confi rmed at T7 pups from the same treatment.This could be of interest to further inspect from the aspect of possible long-term outcomes of maternal hypothyroidism, already observed for metabolic disorders and behavior [29].
A similar trend concerning the cells in the external and internal hair sheet or in the associated sebaceous glands was observed regarding reduced proliferation and increased apoptosis of HF cells.Safer et al. (2001) demonstrated that topical application of T3 stimulated hair growth in mice and rats, suggesting that thyroid hormones directly affect HF cell proliferation [13].Tsujio et al. (2008) concluded that the diminishing growth of hair in hypothyroid animals is a consequence of reduced proliferation and accelerated apoptosis of HF cells, due to the reduced concentration of thyroid hormones, while the reduced number of HFs in hypothyroid animals could be explained by their accelerated atrophy [30].This lead to delay of HF morphogenesis in both groups of treated pups, confi rmed in several previous studies [12,23,31].
Reduced number of HFs and their delayed morphogenesis could be discussed considering the regenerative potential of the skin.The skin is an organ with a compartment containing highly active pluri-and multipotent SCs [32,33].As mentioned above bulge cells can be recognized by the expression of the embryonic stem cell transcription factor NANOG, which helps maintain the capacity of selfrenewal and multipotency in SCs [6,7], promoting their proliferation and delaying senescence [6,34,35].In our study, HF bulge region cells in the tissue samples from T0 and T7 pups originating from mothers with both types of hypothyroidism, showed an increase in the number of NANOG + HF, as well as NANOG + cells in HF bulge itself.It is known that under physiological conditions, apoptosis is suppressed in the epidermis, and that the Casp 3 + cells cannot be observed [36,37], but within the bulge region, the activity of this enzyme is expressed.The increased NANOG immunopositivity seems to be in connection with the increase in the number of Casp 3 + cells in the bulge of both hypothyroid T7 groups.The Caspase family role is involved in regulating a large cohort of cellular processes other than cell death, including regulation of SCs properties.The molecular mechanisms by which these enzymes promote regeneration in multicellular organisms have yet to be determined [38].Our fi nding may indicate that the weaker signaling and slower mobilization of stem cells into the upper compartments of HF or impaired epidermis, lead to their consequent accumulation in the HF bulge [39].Reduced regenerative potential of the skin was observed in hypothyroid individuals [39] and the impaired reepithelialisation is associated with a reduced contribution of bulge SCs [10,40].
Bulge SCs are multipotent and can also differentiate into epidermal and sebaceous gland cells [8,9].Impaired differentiation of bulge cells due to the lack of thyroid hormones and the delayed morphogenesis of HF in treated pups could be partially explained by this fact.The decreased number of HF may be the consequence of the disrupted entry of HF into the anagen phase of growth and the persistence of HF of hypothyroid rats in the telogen phase [39].Bulge cells divide several times during anagen [33,41], which enables and maintains an uninterrupted cycle of hair growth.

CONCLUSION
The results have shown that the intensity of the modifi cations on the skin and its derivates are dependent on the hypothyroidism form, as well as on the pups' age.The more pronounced alterations are detected on seven-day-old pups with overt form of hypothyroidism.In addition, this is the fi rst study to the best of our knowledge that combines subclinical hypothyroidism with the clinical form, and sheds light on changes in epidermal morphology from the aspect of regenerative potential/stem cells and important processes such as proliferation and apoptosis.The study could be improved by examining the effects of maternal hypothyroidism in older (pubertal) offspring and adults and assessing the reversibility of changes in skin and its derivates during growing.The obtained results indicate that there is signifi cant impairment in epidermal morphology and thickness, and reduced hairiness in treated rats, especially in the overt form of hypothyroidism.In the offspring obtained from mothers with subclinical form, the observed changes were less pronounced and delayed in occurrence.The future prospective is to investigate the molecular mechanisms of these phenomena, in order to prevent the occurrence of more serious skin damage and to provide better insight into this controversial topic.

Table 1 .
Primary antibodies used for immunohistochemistry

Table 2 .
The thickness of the non-cornifi ed layer of the epidermis in neonatal and 7-day-old pups

Table 3 .
The thickness of the stratum corneum in neonatal and 7-day-old pups PTU low -pups from does treated with low dose of PTU; PTU high -pups from does treated with high dose of PTU; Т0 -within 24h of delivery; Т7-7-day-old; SD-standard deviation; n -number of animals per group; P-signifi cant difference; *group with signifi cant difference p˂0.05; ***groups with signifi cant difference p˂0.001