Delta‐like protein 1 in the pituitary‐adipose axis in the adult male mouse

With the aim of studying delta‐like protein 1 (DLK1) with respect to the relationship between adipocyte leptin and adenohypophyseal hormones, we carried out an immunohistochemical study analysing the presence of receptors for these hormones in the pituitary and adipose cells of male wild‐type (WT) mice (Dlk1 +/+) compared to knockout (KO) mice (Dlk1 −/−). The mRNA expression of these molecules was also determined using the reverse transcriptase‐polymerase chain reaction. The results obtained showed that, in WT adipose cells, all of the adenohypophyseal hormone receptors were present, with a higher mRNA expression for growth hormone (GH) receptor and thyroid‐stimulating hormone (TSH) receptor. Of the total cells in the anterior pituitary lobe, 17.09±0.9% were leptin receptor (LEPR) immunoreactive (‐IR), mainly in GH‐IR and prolactin (PRL)‐IR cells (41.5±3.8%; 13.5±1.7%, respectively). In Dlk1 −/− mice, adipocyte cells showed a significant increase in the TSH receptor mRNA expression level. Moreover, the percentage of LEPR‐IR GH cells showed a statistically significant increase compared to controls, from 41.5±3.8% to 53.1±4.0%. By contrast, only 3.0±0.6% of LEP‐IR anterior pituitary cells were detected in Dlk1 KO mice, as opposed to 6.8±1.1% observed in WT mice. The results suggest that relationships exist between adipocytes and pituitary GH, PRL and TSH cells, in addition to an influence with respect to the synthesis and release of pituitary leptin, particularly in PRL cells.

more recently, gonadotrophin receptors have been described in chicken adipose tissue. 6 In the pituitary gland, LEP exerts a direct or indirect effect on the secretion and regulation of the adenohypophyseal hormones gowthn hormone (GH), FSH/LH, adrenocorticotrophic hormone (ACTH), thyroidstimulating hormone (TSH) and prolactin (PRL). [7][8][9][10][11][12] Numerous studies show that LEP regulates the four main hypothalamic-pituitary peripheral axes (adrenal, thyroid, gonadal and growth hormone) at different levels. 13 Direct LEP action on the pituitary is exerted by binding to its receptor (LEPR) in endocrine cells of the anterior lobe. Thus, homozygote LepR mutant mice displayed lowered TSH and GH secretion. 14 Moreover, the selective deletion of LEPR expression in hypophyseal GH cells in murine models gave rise to a reduction in both somatotroph cells and serum GH levels, whereas abdominal fat increased in adult mice. 9 In addition, the loss of somatotroph leptin resulted in lower prolactin levels in serum of 6-month-old female mice. 12 The presence of the long form of the LEPR in hypophyseal cells of the anterior lobe has been demonstrated in the rat and mouse, 15,16 although the distribution between the different cell types varies among species. In rats, 97% of GH-producing cells express the leptin receptor, whereas less than 1% of the other cell types do so. 16 Moreover, using dispersed cells in culture, high percentages of adenohypophyseal cells expressing LEPR have been described in rats and mice. 9,17 By contrast, in sheep, approximately 70% of somatotroph cells, as well as 30% of gonadotroph and corticotroph cells, express this receptor. 18 In recent years, numerous studies have shown that LEP is also produced by pituitary cells, 15,16,19,20 suggesting a paracrine or autocrine role for this hormone in the pituitary gland. Although there is no agreement to date regarding the cell types that produce LEP, 15,19,[21][22][23] most studies concur that GH cells are the most important LEP-producing cells in the pituitary gland. This is in agreement with the evidence indicating that LEP is important for GH secretion, as reported in several studies. 24,25 Recently, Odle et al 11  Regarding the regulation of adipocyte LEP production, delta-like protein 1 (DLK1) is a transmembrane protein with regulatory effects on adipocyte differentiation. [26][27][28] Dlk1 mRNA is also widely expressed during embryonic development, although, in the adult, Dlk1 mRNA expression is limited to some endocrine glands and subsets of neurones in the brain, including the hypothalamus. [29][30][31][32] In the pituitary gland of the 129/Svj wild-type (WT) mouse, we previously demonstrated that DLK1 is expressed in all types of cells, particularly in somatotroph cells. 33,34 Furthermore, we used a Dlk1 knockout (KO) mouse of the same strain that displays a smaller pre-and postnatal size but increased white fat mass in the adult. 35,36 The results obtained showed increased serum leptin in these KO mice, as well as a slight increase in GH levels, despite having a smaller number of GH-producing cells. 33 Based on these data, the present study aimed to investigate in situ (ie, maintaining the natural context of the cells), the influence of DLK1 protein on the relationship between adipocyte and adenohypophyseal cells, using male WT and Dlk1 −/− KO mice of the strain 129/SvJ.

| Animals
Eighteen adult male mice (129⁄SvJ) approximately 4 months old were used: nine WT mice (Dlk1 +/+ ) and nine Dlk1 KO mice (Dlk1 −/− ). All animals were supplied by the Animal House Facility of the University of Castilla-La Mancha (Spain). Dlk1-deficient mice were generated as described previously. 36 The experimental procedures were approved by the Ethics Committee for Experimental Animal Welfare of the

| Antisera
Antibodies are listed in the Supporting information (Table S1). A polyclonal antiserum (#1125) against DLK1 raised in rabbit was used, 37 along with polyclonal goat and rabbit anti-DLK1 antibodies purchased from Santa Cruz Biotechnology Inc. (Heidelberg, Germany) (for details, see the Supporting information, Table S1). Polyclonal antisera for identifying the pituitary hormones ACTH, PRL, FSH and TSH were produced in rabbits by Dr G. Tramu (University of Bordeaux 1, Bordeaux, France). The specificity of these antisera was evaluated by means of an absorption test, incubating the antisera overnight with the homologous antigens for 12-24 hours. 38 Other antisera obtained from commercial sources were rabbit

| Immunohistochemistry
Eight male mice, four WT and four KO, were deeply anaesthetised with a mixture of ketamine (100 mg kg -1 ; Parke-Davis, Alcobendas, Spain) and 2% xylazine (10 mg kg -1 ; Dibapa, Barcelona, Spain) and transcardiacally perfused with 0.9% saline followed by fixation with Bouin's solution (0.9% picric acid, 9% formaldehyde, 5% acetic acid). The pituitary glands and abdominal adipose tissues were dissected out and postfixed by immersion in the same fixative for 36 hours. Subsequently, they were dehydrated in graded ethanol, cleared in xylene and embedded in paraffin. Horizontal sections, 3 μm in thickness, were cut on a microtome (Shandon Finesse 325; Thermo Electron Corporation, Waltham, MA, USA) and prepared for immunohistochemistry. Sections were collected consecutively onto slides to determine co-localisations.
For the indirect immunohistochemical procedure, deparaffinised sections of abdominal adipose tissues were rehydrated in 0.05 mol L -1 Tris buffered saline (TBS) (Trizma Base 0.05 mol L -1 , NaCl 0.9%) (pH 7.4), which was also used for all further incubations and washes.
To study the expression of DLK1, LEP and LEPR in relation to each pituitary cell producing hormones of the anterior lobe, an immunofluorescence procedure was carried out. To localise two or three markers on the same section, double-or triple-immunofluorescence was used. Sections were first incubated overnight with a selected primary antibody diluted as described below, and then with the corresponding secondary antibody conjugated with a fluorophore for 1 hour at room temperature. The same procedure was followed with the second and the third primary antibodies.

| Quantitative analysis
To study the different peptides and receptors in the pituitary of To quantitatively analyse LEP, LEPR and the receptors for the adenohypophyseal hormones ACTHR, TSHR, FSHR, LHR, PRLR and GHR, eight abdominal adipose tissue samples were used: four from WT and four from KO mice. Two sections per slide were selected, and 10 fields per section were captured at random.

Normal distribution was not found by Shapiro-Wilk's test for small
samples and statistical analysis was performed using a two-tailed nonparametric Wilcoxon-Mann-Whitney test. P<.05 was considered statistically significant. All values are given as the mean±SEM.

| RNA isolation and quantitative reverse transcriptase polymerase chain reaction (qRT-PCR)
The Statistical data for each target gene were determined in five different animals, each carried out in triplicate. The number of animals required for the studies was calculated using granmo (http://www.imim. cat/ofertadeserveis/software-public/granmo/). Accordingly, accepting an α risk of 0.1 and a β risk of 0.2 in a two-sided test, five subjects were required in each group to recognise a difference ≥1.9 units as being statistically significant. The common SD was assumed 1.2. A drop-out rate of 0% was anticipated. Statistical analysis was performed using a two-tailed nonparametric Wilcoxon-Mann-Whitney test; normal distribution was not found by Shapiro-Wilk's test for small samples. P<.05 was considered statistically significant.

| RESULTS
In the present study, we tested immunohistochemically for the presence of receptors for hormones released by cells of the two tissues in WT 129/Svj mice (WT; Dlk1 +/+ ) compared to KO mice (Dlk1 −/− ) to link the presence of DLK1 protein with the other molecules under consideration. The mRNA expression of these molecules was also determined by qRT-PCR.

| Expression of adenohypophyseal hormone receptors in the adult abdominal adipose tissue of WT male mice
Adipocytes were immunoreactive for all specific antibodies against adenohypophyseal hormone receptors ACTHR, TSHR, FSHR, LHR, PRLR and GHR ( Figure 1B-G). The mRNA expression of these receptors in the abdominal adipose tissue of adult mice was subjected to qRT-PCR, using specific oligonucleotides, and analysed with the method of Pfaffl. 40 The data indicated that the receptors with the highest RNA expression were GHR and TSHR, followed by ACTHR, whereas PRLR, LHR and FSHR had a lower level ( Figure 2). In all cases, high variability among samples was observed. whereas the LEPR mRNA levels were somewhat higher but also low (not shown).

| Expression of the LEPR in the anterior pituitary lobe and abdominal adipose tissue of WT male mice
In LEP-producing adipocytes, immunoreactivity against their LEPR receptor was also observed ( Figure 5). The LEPR mRNA expression was very low and similar between all the adipose tissue samples, whereas LEP expression as determined by qRT-PCR was significantly higher but with an elevated variability between samples (Table 2).
T A B L E 1 Mouse primers used for quantitative polymerase chain reaction amplifications

Gene
Forward primer Reverse primer

| Presence of the specific LEPR in adenohypophyseal cells and abdominal adipose tissue from Dlk1 KO mice
In Dlk1 KO mice, LEPR-IR cells were also observed in the anterior lobe of the adenohypophysis (see Supporting information, Figure S1B). In this case, the percentage of LEPR-IR cells was 27.17±1.5%, which is significantly higher than in WT controls (17.09±0.9%; P≤.0001) ( found between controls and Dlk1 KO mice. Therefore, only the GH-LEPR-IR cell type presented significant differences with respect to WT mice (Table 3).
Few LEP-IR cells were observed in the anterior pituitary lobe of KO mice (see Supporting information, Figure S1D), comprising 3.0±0.6% with respect to the total, as opposed to 6.8±1.1% of LEP-IR cells in WT mice (Table 3) (Table 3). By contrast, all LEP-IR cells were also LEPR-IR, as in controls (Table 3).
Finally, LEP mRNA was highly expressed in all abdominal white adipose samples, whereas LEPR mRNA levels were quite low, although no significant differences between KO and WT mice were found in any case (see Supporting information, Figure S1, bottom).

| DISCUSSION
In the present study, we have analysed the impact of the lack of was detected in cultured adipose cells. 5 Later, FSHR was detected at protein and mRNA level in adipocytes of abdominal adipose tissue in female chickens. 6 As in other vertebrate groups, 14  According to previous studies, detection of LEP in pituitary anterior lobe cells indicated great dissimilarities between the species studied.
In rats, LEP was first located mainly in TSH cells. 15 However, GH cells were recognised as the main cells producing LEP in a study using dis-    context of cell interactions being altered. Although observed differences in cell types may not be a consequence of this, they can affect the dissimilarities in the reported percentages.
The potential role of LEP in GH cells has been studied using KO mice for the adipocyte or pituitary LEP, 11,12 a deletion of exon 17 of LepR in somatotroph cells 9 and LEPR deletion in GH cells. 49  This is adipocyte leptin according to Odle et al. 12 These results would confirm that both LEP sources are necessary for correct GH activity.
All of these observations point to a possible feedback between adipose and GH cells.
Additionally, GH hormone can also block preadipocyte differentiation by induction of Dlk1 expression, 50 which is considered to be a modulator of adipogenesis (see Introduction TRH is not only a releasing factor for TSH, but also for GH and PRL hormones. 55 In our mouse model lacking Dlk1, serum GH levels were slightly higher than in the WT male mouse, whereas PRL cells contained less immunoreactivity. 33 This lower immunoreactivity does not appear to be related to higher PRL release because PRL levels in serum were practically undetectable 33 56 However, in our model, PRL does not appear to be the hormone responsible for augmented LEP levels, at least directly, suggesting an indirect action.
The hormone LEP also has an important action on the reproductive system and "may act as the critical link between adipose tissue and the reproductive system, indicating whether adequate energy reserves are present for normal reproductive function". 7 In the pituitary, mice lacking DLK1 protein express significantly less FSH mRNA without showing changes in LH mRNA. 33 These results suggest a difference in the role of DLK1 in the synthesis of these hormones. Furthermore, abdominal adipose tissue shows a nonsignificant increase in LHR expression and a lower FSHR expression level according to the results of the present study. Indeed, published data on the rat indicate that LEP indirectly regulates hypothalamic GnRH production under normal conditions and directly stimulates the production of gonadal hormones by LEPR, expressed in gonadotroph cells. 57 In sheep and rat pituitaries, gonadotrophs were found to be the main type of adenohypophyseal cells expressing LEPR. 16,18 This suggests that LEP produced by gonadotroph cells acts in a paracrine manner on GH cells in these two spe-  33 However, in Dlk1 −/− mice, there was a nonsignificant LEPR increase in gonadotroph cells, together with a significant reduction in LEP in PRL cells. This suggests that adipose LEP acts on gonadotroph cells.
In summary, although physiological studies are needed, the results of the present study suggest that DLK1 is a factor to be considered not only as a modulator of adipocyte differentiation, but also of the differentiated adipocyte function. It may thus be an important factor in the relationships between adipose tissue and the pituitary hormone producing cells. Therefore, the observed adiposity in Dlk1-deficient mice may be a consequence of DLK1 being involved in the different feedbacks between adipose LEP and the production of pituitary hormones.
Hormones GH, PRL, LH and TSH all act on the adipose cells, stimulating or inhibiting LEP synthesis and release, inhibiting lipase, and increasing triglyceride storage or the number of adipose cells. In addition, DLK1 appears to directly or indirectly modulate pituitary LEP synthesis, particularly in PRL cells. Figure 8 presents tentatively proposed interactions between adenohypophyseal cells and adipocytes based on previous and present results, which require further in vitro physiological tests.