In addition to the direct effects of thyroid hormone (TH) on peripheral organs, recent work showed metabolic effects of TH on the liver and brown adipose tissue via neural pathways originating in the hypothalamic paraventricular and ventromedial nucleus (PVN and VMH). So far, these experiments focused on short-term administration of TH. The aim of this study is to develop a technique for chronic and nucleus-specific intrahypothalamic administration of the biologically active TH tri-iodothyronine (T3). We used beeswax pellets loaded with an amount of T3 based on in vitro experiments showing stable T3 release (∼5 nmol l−1) for 32 days. Upon stereotactic bilateral implantation, T3 concentrations were increased 90-fold in the PVN region and 50-fold in the VMH region after placing T3-containing pellets in the rat PVN or VMH for 28 days respectively. Increased local T3 concentrations were reflected by selectively increased mRNA expression of the T3-responsive genes Dio3 and Hr in the PVN or in the VMH. After placement of T3-containing pellets in the PVN, Tshb mRNA was significantly decreased in the pituitary, without altered Trh mRNA in the PVN region. Plasma T3 and T4 concentrations decreased without altered plasma TSH. We observed no changes in pituitary Tshb mRNA, plasma TSH, or plasma TH in rats after placement of T3-containing pellets in the VMH. We developed a method to selectively and chronically deliver T3 to specific hypothalamic nuclei. This will enable future studies on the chronic effects of intrahypothalamic T3 on energy metabolism via the PVN or VMH.
Thyroid hormone (TH) is a major determinant of basal metabolic rate as well as glucose and lipid metabolism (Mullur et al. 2014, Sinha et al. 2014). A number of recent studies have shown that in addition to the well-known direct effects of TH at the tissue level, some of its metabolic effects are mediated indirectly through specific neuronal targets within the hypothalamus. In particular, this has been shown in rats for hepatic glucose metabolism, brown adipose tissue (BAT) function, and the central control of blood pressure and heart rate (for reviews see Fliers et al. 2010, Warner & Mittag 2012, Lopez et al. 2013). Studies on rats have shown that thyrotoxicosis increases glucose production in the liver, while reducing hepatic insulin sensitivity. These effects appeared to be modulated by selective hepatic sympathetic and parasympathetic denervation (Klieverik et al. 2008), and follow-up experiments showed that TH stimulates hepatic glucose production via a sympathetic pathway originating in the hypothalamic paraventricular nucleus (PVN) (Klieverik et al. 2009). Similarly, indirect effects of TH on BAT via the hypothalamic ventromedial nucleus (VMH) were reported by several groups. First, mice heterozygous for a mutant TRα1 with low affinity for T3 appeared to be hypermetabolic due to increased BAT activity, resulting in increased thermogenesis and energy expenditure. The metabolic phenotype was blunted after a functional denervation of sympathetic signaling, suggesting that the central nervous system (CNS) controlled the hypermetabolism of these mice through the autonomic nervous system (Sjogren et al. 2007). More recently, López and coworkers showed that activation of the thermogenic program in BAT through the sympathetic nervous system (SNS) depends on T3-mediated activation of de novo lipogenesis in the hypothalamic VMH (Lopez et al. 2010), establishing a role for T3 in the VMH in the regulation of BAT. Together, these studies revealed a distinct role of hypothalamic T3 in the regulation of peripheral organs through the autonomic nervous system.
It should be noted that the above-mentioned metabolic effects of intrahypothalamic T3 were observed in acute experiments, i.e., a few hours after T3 administration, whereas metabolic dysfunction in thyrotoxicosis, including insulin resistance and weight loss, is mostly due to long-term exposure to excessive TH. Thus, the relationship between chronically elevated intrahypothalamic T3 concentrations and metabolic dysfunction, which may be of clinical relevance, has not been reported yet. Before embarking on experiments to explore this relationship, we aimed to develop a method for selective and chronic (4 weeks) administration of T3 in the rat PVN or VMH region.
We implanted T3-containing or control pellets bilaterally into either the PVN or the VMH region of rats for a period of 28 days. Plasma T3, T4, and TSH concentrations were measured at regular time intervals after placement of the pellets. Hypothalamic tissue punches containing either the PVN or the VMH, as well as pituitary glands, were collected at the end of the experiment for the assessment of local T3 tissue concentration and gene expression analysis. We conclude from our experiments that it is feasible to deliver T3 selectively and chronically to distinct hypothalamic nuclei.
Materials and methods
Preparation of pellets
3,3′,5-tri-iodo-l-thyronine (Sigma-Aldrich) was mixed with melted beeswax (Sigma-Aldrich) at a ratio of 1:9 (w/w) in a 70°C water bath. The mixture was packaged into a length of 1mm cylindrical pellet using a 23G stainless cannula. The weight of the pellets was around 76.8 µg, including 10% of T3 amounting to 7.68 µg. Blank beeswax pellets without T3 were used as control pellet. During surgery, the pellet was extruded out of the cannula by a stainless steel wire plunder.
In vitro experiment
T3-containing pellets (T3Ps) or control pellets (cPs) were incubated in 500 μl bovine serum on a 37°C homoeothermic shaker for 28 days. After 1 day and every 3 days thereafter, a sample of 250 μl serum was taken from the system for assessment of T3 concentration and replaced by an equal amount of fresh serum.
Male Wistar rats weighting 280–320 g (total number 92; Charles River Breeding Laboratories, Sulzfeld, Germany) were housed individually in a 12 h light:12 h darkness cycle environment. Chow and water were provided ad libitum. All procedures were approved by the Animal Care Committee of the Royal Netherlands Academy of Arts and Sciences.
After 1 week of acclimatization in the facility, a total number of 84 animals divided into eight groups (see Table 1) were anesthetized by an intramuscular injection of Hypnorm (fentanyl, 0.252 mg/kg BW; fluanisone, 8 mg/kg BW; Janssen, High Wycombe, Buckinghamshire, UK) followed by a subcutaneous Midazolam (1 mg/kg BW; Roche) injection. T3Ps or cPs were implanted bilaterally into the hypothalamic PVN (anteroposterior: −1.8 mm; lateral: 2.1 mm; ventral: −6.9 mm; angle: 10°) or VMH (anteroposterior: −2.3 mm; lateral: 2.2 mm; ventral: −8.1 mm; angle: 10°) region using a standard Kopf stereotaxic apparatus (Paxinos & Watson 2005). Pellet placements were checked in 20 μm cresyl violet-stained brain sections of each animal at the end of the experiment. Blood samples (around 250 μl) were taken by tail incisions using heparin-coated capillary Microvette (Sarstedt AG & Co, Nümbrecht, Germany) 3 days before and 3, 7, 14, 21, and 28 days after pellet placement. The plasma samples were collected after centrifugation at 4000 rpm for 15 min and stored at −20°C. After 28 days, rats were decapitated; trunk blood was taken and stored at −20°C until analysis. Pituitaries were snap frozen in liquid nitrogen and stored in −80°C until further use. Brains were removed and frozen on dry ice. The hypothalamus was cut into 300 μm coronal sections using a cryostat (Leica CM1950). For T3 tissue concentration and gene expression measurements, the PVN and VMH regions were collected by punching the area with a 17G stainless steel needle according to Palkovits (1973). The third ventricle and major fiber tracts served as landmarks for orientation. Tissue was then homogenized either in lysis buffer (Roche Applied Science, Penzberg, Germany) for RNA isolation or in PBS for the assessment of tissue T3 concentration. Validation of the punched blocks was performed by measuring the expression levels of nucleus-specific genes by qPCR.
Design of the experimental groups.
|1||Intact||Untreated||8||For qPCR in hypothalamic punches|
|2||cP_PVN||Control pellets in the PVN||8|
|3||T3P_PVN||T3-containing pellets in the PVN||12|
|4||cP_VMH||Control pellets in the VMH||8|
|5||T3P_VMH||T3-containing pellets in the VMH||12|
|6||cP_PVN||Control pellets in the PVN||11||For T3 brain tissue concentrations|
|7||T3P_PVN||T3-containing pellets in the PVN||11|
|8||cP_VMH||Control pellets in the VMH||11|
|9||T3P_VMH||T3-containing pellets in the VMH||11|
Plasma thyroid hormones
Plasma T3 and T4 concentrations were determined using an in-house RIA as reported before (inter-assay variation T3, 6.2% and T4, 7.3%; intra-assay variability T3, 3.6% and T4, 6.6%) (Wiersinga & Chopra 1982). Plasma TSH was determined by a chemiluminescent immunoassay, using the Immulite 2000 (Siemens). For TSH, the inter-assay variation was ±9% and the intra-assay variation was 3–4%. Within the same experiment, all samples were measured within one run to prevent inter-assay variation.
Hypothalamic tissue T3 concentrations
The PVN and VMH punches were weighed and homogenized in 300 µl methanol using a Magna Lyser (Roche Molecular Biochemicals, Mannheim, Germany). The homogenates were then extracted using a double volume of chloroform. The mixtures were centrifuged at 3184 g at 4°C for 15 min. Supernatants from the extractions were transferred into glass vials and vaporized under nitrogen gas for 30 min. Subsequently, samples were dissolved in 70 μl alkaline T3 RIA assay buffer and 70 μl T3-free plasma. T3 concentrations were measured by T3 RIA as described above.
RNA isolation and real-time PCR (qPCR)
Total RNA from pituitary, the PVN, and VMH punches were isolated using the MagNA Pure LC 2.0 Instrument (Roche Molecular Biochemicals) with Magna pure tissue III total RNA kit (Roche Molecular Biochemicals). RNA yield was determined using the Nanodrop (Nanodrop), and cDNA was synthesized with equal RNA input with the First-Strand cDNA synthesis kit (AMV) for qPCR with oligo-d (T) primers (Roche Molecular Biochemicals). As a control for genomic DNA contamination, a cDNA synthesis reaction without reverse transcriptase was included. Quantitative PCR was performed using the LightCycler 480 (Roche Molecular Biochemicals) and LightCycler 480 SYBR Green I Master mix (Roche Molecular Biochemicals). The primers used for qPCR are listed in Table 2. Quantification was performed using the LinReg software. PCR efficiency was checked individually and samples with a deviation of more than 5% of the mean were excluded from the analysis. Calculated values were related to the geometric mean expression of Gapdh and Hprt, reference genes showing stable expression under the experimental conditions.
Primers used for qPCR.
|Gene||symbol||Forward (5′-3′)||Reverse (5′-3′)||Products length (bp)|
|Hypoxanthine guanine phosphoribosyl transferase||Hprt||GCAGTACAGCCCCAAAATGG||AACAAAGTCTGGCCTGTATCCAA||84|
|Thyrotropin releasing hormone, prepropeptide||Trh||TCTGCAGAGTCTCCACTTCG||AGAGCCAGCAGCAACCAA||59|
|Deiodinase type 3||Dio3||AGCGCAGCAAGAGTACTTCAG||CCATCGTGTCCAGAACCAG||61|
|Steroidogenic factor 1 (NR5A1)||Sf1||CCAGTGTCCACCCTTATCCG||ACCTTGTCACCACACACTGG||117|
|Thyroid-stimulating hormone, β subunit||Tshb||TCGTTCTCTTTTCCGTGCTT||CGGTATTTCCACCGTTCTGT||245|
Data are expressed as mean ± standard error of the mean (s.e.m.). Differences between two groups were analyzed using an independent two-tailed Student’s t-test. When data were not normally distributed, the independent non-parametric Mann–Whitney U test was used. The changes in T3 and T4 concentrations in the plasma were evaluated by two-way ANOVA with repeated measurements followed by Bonferroni post hoc analysis. Statistical significance was defined at a level of p < 0.05.
In vitro experiments
Before the in vivo experiments, we tested the release of T3 from T3P in serum. The in vitro results showed that T3 was constantly released into the medium at a concentration of ∼5 nmol l−1 from the T3P (corresponding to a T3 release of around 0.26 ng per day), but not from cP for a period of 32 days (Fig. 1).
Pellet placement and validation of hypothalamic dissection
T3Ps or cPs were placed either in the PVN or in the VMH. Post-mortem examination revealed that the majority of pellets were placed laterally from the PVN and dorsolaterally in the VMH with an anteroposterior location ranging from −1.0 to −1.7 mm from bregma (Fig. 2). Nine animals were excluded from the analysis because of the misplacement of the pellets. In the animals implanted with cPs, no adverse effects were found in terms of food and water intake or growth compared with an intact, non-operated control group (n = 8). In addition, no differences were found in gene expressions (i.e., Dio3, Hr, and Trh in hypothalamus and Tshb in pituitary) or plasma TSH, T3, and T4 concentrations between these two groups (data not shown).
In order to validate the anatomy of the punches, we assessed oxytocin (Oxt) mRNA, which is PVN-specific (Swanson & Sawchenko 1983) and steroid factor 1 (Sf1) mRNA, which is VMH-specific (Roselli et al. 1997). As shown in Fig. 2C, Oxt mRNA was expressed 1000-fold less in the VMH compared with the PVN. In contrast, Sf1 mRNA expression was expressed over 20-fold higher in the VMH compared with the PVN (Fig. 2D), supporting the selectivity of the anatomical punches.
Tissue T3 concentrations in the PVN and VMH
We then evaluated the local T3 concentration after 28 days of implantation in micro-dissected PVN and VMH punches. As expected, local T3 concentrations were much higher in the T3P-treated group compared with the cP group. T3 concentrations increased 90-fold in the PVN region and 50-fold in the VMH region compared with the cP group when the T3Ps were placed in the PVN or VMH respectively (Fig. 3). When T3Ps were placed in the PVN, we also observed increased T3 in the VMH compared with the cP group (Fig. 3A), although to a much lesser extent than in the PVN. When T3Ps were placed in the VMH, no increased T3 concentrations in the PVN were observed. In fact, T3 concentrations were even significantly lower compared with the cP group (Fig. 3B).
Differential effect of T3 administration in the PVN and VMH on T3-responsive genes
We assessed the effects of local T3 administration according to the mRNA expression of T3-responsive genes in the targeted nuclei. When implanting the T3Ps in the PVN, both Dio3 and Hr mRNA expressions were significantly increased in the PVN region, whereas Dio3 and Hr mRNA expressions in the VMH region were not affected (Fig. 4A and B). Conversely, placement of T3Ps in the VMH induced a significant increase in Dio3 and Hr mRNA in the VMH region, but not in the PVN region (Fig. 4C and D). Gene expression was not different between intact animals and animals implanted with cPs (data not shown).
Effect of T3 administration in the PVN on HPT axis
Placement of T3Ps in the PVN did not change Trh mRNA in the PVN or VMH region, while decreasing pituitary Tshb mRNA expression (p < 0.001) (Fig. 5A and B). There was no significant effect on plasma TSH concentrations (Fig. 5C). Plasma T3 concentrations were significantly decreased in rats with T3Ps in the PVN at day 3, 14, and 21 after implantation, whereas plasma T4 concentrations were decreased at days 3 and 28 (Fig. 5D and E).
Effect of T3 administration in the VMH on HPT axis
Placement of T3Ps in the VMH did not change Trh mRNA in the PVN; however, it decreased Trh mRNA in the VMH region (Fig. 6A). As expected, pituitary Tshb mRNA, plasma TSH (Fig. 6B and C), and plasma TH (Fig. 6D and E) concentrations did not change throughout the time course of this experiment.
This study shows that it is feasible to selectively deliver T3 for 4 weeks to the PVN or the VMH, without leakage of T3 into the systemic circulation. Also within the hypothalamus, diffusion was limited, as T3-containing pellets in the PVN did not affect T3-responsive gene expression in the VMH or vice versa. As expected, chronic T3 administration in the PVN, but not in the VMH, decreased pituitary Tshb mRNA expression as well as plasma T3 and T4 concentrations.
To set up the experimental model, we first performed pilot experiments in vitro. Incubation of T3P in plasma resulted in rather constant T3 concentrations in the physiological range for more than 1 month. Subsequently, we implanted the pellets into specific hypothalamic nuclei, followed by assessment of local tissue T3 concentrations in hypothalamic tissue punches. The anatomical selectivity of the punch technique was confirmed by the selective expression of Oxt and Sf1 in either the PVN or the VMH respectively. Twenty-eight days after placement of T3Ps, T3 concentrations increased 90-fold in the PVN region and 50-fold in the VMH region compared with cP when T3Ps were placed in the PVN or VMH respectively. Although the local T3 concentration was somewhat higher in the VMH after implantation of T3Ps in the PVN, this did not induce T3-responsive gene expression. These results show that T3P released T3 in a rather limited anatomical region for more than a month.
Next, we investigated the effects of T3 administration by assessment of T3-responsive gene expression. Dio3 is a T3-sensitive gene coding for a deiodinating enzyme that is critical in maintaining local TH homeostasis (Dentice & Salvatore 2011). As a main physiological inhibitor of TH action, it protects cells from excessive TH by converting T3 (or T4) into its inactive derivatives, T2 (or T3) (Dentice and Salvatore 2011). In rats, Dio3 mRNA expression was shown to increase in the brain during hyperthyroidism, while it was undetectable during hypothyroidism (Tu et al. 1999, Bianco et al. 2002). In line with these observations, the present experiments showed that Dio3 mRNA was only increased in the PVN but not in the VMH when the T3Ps were placed in the PVN, while Dio3 mRNA was only increased in the VMH but not in the PVN when the T3Ps were placed in the VMH. Of note, Dio3 mRNA expression may be enhanced by neuronal injury (Li et al. 2001). However, Dio3 mRNA expression in non-operated rats did not differ from rats implanted with cPs, which excludes an effect of tissue damage by the pellet itself. In addition to Dio3, expression of Hr, which is another well-known T3-responsive gene (Potter et al. 2002), was significantly increased only in the target nuclei where the T3Ps were placed, again suggesting a nucleus-specific T3 response as a result of locally increased T3 concentrations. A few additional reported T3-responsive genes such as oxytocin (Oxt) (Dellovade et al. 1999) and monocarboxylate transporter 8 (Mct8 (Slc16a2)) (Herwig et al. 2014) did not show significant changes between control and T3-treated groups (data not shown). We did not observe any change in Trh mRNA expression measured by qPCR in a punch of the PVN region 28 days after placement of T3Ps. A previous study using a hypothyroid rat model showed that unilateral T3 crystalline implants within the anterior hypothalamus induced a marked reduction of Trh mRNA only in the medial parvocellular division of PVN as shown by in situ hybridization within 4 days (Dyess et al. 1988). Moreover, a study by Murphy et al. (2012) showed a mild suppression of hypothalamic Trh mRNA assessed by in situ hybridization 6 weeks after T3 implants in the hypothalamus of Siberian hamsters. Therefore, the absence of an effect on Trh mRNA in the PVN after T3 administration to the PVN in our study can probably be explained by the fact that only a subpopulation of TRH neurons in the PVN, i.e., the hypophysiotropic TRH neurons, are responsive to T3 (Segerson et al. 1987, Fekete & Lechan 2014). In our approach, using the whole PVN region for qPCR, selective changes in the subpopulation of hypophysiotropic TRH neurons may have gone unnoticed as all TRH neurons in the micro-punch are included in the PCR assessment.
Pituitary Tshb mRNA was lower after 28 days in the PVN in T3P group compared with rats with cPs, in accordance with negative feedback regulation (Fliers et al. 2014). Plasma TSH concentration, however, was not significantly altered. Of note, a previous study showed that T3 crystalline implants in the hypothalamus did not influence TSH secretion in hypothyroid rats, although IR-pro-TRH was remarkably inhibited (Dyess et al. 1988). The production and secretion of TSH by the pituitary gland are positively stimulated by TRH, while TRH also modulates the glycosylation of TSH, which affects TSH bioactivity (Magner 1990). Part of the explanation for unaltered plasma TSH in the presence of decreased pituitary Tshb mRNA as observed in this study may be a discrepancy between the biological activity and the immunological detectability of TSH (Persani & Bonomi 2014).
In line with negative feedback at the central level and decreased pituitary Tshb mRNA after T3Ps implantation in the PVN, plasma T3 and T4 concentrations were significantly decreased throughout the experiment. In summary, our results indicate that the T3Ps implantation in the PVN delivers T3 locally for at least 28 days. As an autonomic nervous system integration center, the PVN is involved in many aspects of energy metabolism, including hepatic glucose production. The investigation of metabolic effects of chronic intrahypothalamic T3 administration in the PVN will be the focus of our future experiments.
Placement of T3Ps in the VMH did not affect the HPT axis in any way, again supporting a locally restricted T3 administration. Recent experiments have shown that TH in the VMH is a powerful determinant of BAT activity (Lopez et al. 2013). Thus, we expect that the present technique will enable studies on long-term metabolic consequence of central T3 on energy balance in addition to glucose metabolism. Finally, as a recent study showed a novel central role for T3 in cardiovascular regulation by acting on a group of pre-autonomic neurons in the anterior hypothalamus (AH) (Mittag et al. 2013), our model may be helpful to study the long-term effects of T3 in the AH on blood pressure and heart rate as well.
In conclusion, our pilot experiments provide a promising technique to selectively deliver T3 for at least 4 weeks to specific hypothalamic nuclei. The use of these T3-containing pellets holds promise for studies on metabolic effects of nucleus-specific chronic T3 delivery within the hypothalamus.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
This project was funded by a grant from the China Exchange Program (CEP) of the Royal Netherlands Academy of Sciences and the Chinese Academy of Sciences (grant #11CDP001, to E.F.) and by a grant from the Chinese Scholarship Council (CSC) (CSC ID: 201206340004, to Z.Z.).
We would like to thank the staff of the Laboratory of Endocrinology for measuring serum thyroid hormones.
DyessEMSegersonTPLipositsZPaullWKKaplanMMWuPJacksonIMLechanRM1988Triiodothyronine exerts direct cell-specific regulation of thyrotropin-releasing hormone gene expression in the hypothalamic paraventricular nucleus. Endocrinology1232291–2297. (doi:10.1210/endo-123-5-2291)
HerwigACampbellGMayerCDBoelenAAndersonRARossAWMercerJGBarrettP2014A thyroid hormone challenge in hypothyroid rats identifies T3 regulated genes in the hypothalamus and in models with altered energy balance and glucose homeostasis. Thyroid241575–1593. (doi:10.1089/thy.2014.0169)
KlieverikLPJanssenSFvan RielAFoppenEBisschopPHSerlieMJBoelenAAckermansMTSauerweinHPFliersE2009Thyroid hormone modulates glucose production via a sympathetic pathway from the hypothalamic paraventricular nucleus to the liver. PNAS1065966–5971. (doi:10.1073/pnas.0805355106)
KlieverikLPSauerweinHPAckermansMTBoelenAKalsbeekAFliersE2008Effects of thyrotoxicosis and selective hepatic autonomic denervation on hepatic glucose metabolism in rats. American Journal of Physiology: Endocrinology and Metabolism294E513–520. (doi:10.1073/pnas.0805355106)
PersaniLBonomiM2014Chapter 27 – Uncertainties in endocrine substitution therapy for central endocrine insufficiencies: hypothyroidism. In Handbook of Clinical Neurology pp 397–405. Eds MK Eric Fliers & AR Johannes. Philadelphia, PA, USA: Elsevier. (doi:10.1016/B978-0-444-59602-4.00027-7)