Increased hypothalamic signal transducer and activator of transcription 3 phosphorylation after hindbrain leptin injection

Abstract

Reduction of food intake and body weight by leptin is attributed largely to its action in the hypothalamus. However, the signaling splice variant of the leptin receptor, LRb, also is expressed in the hindbrain, and leptin injections into the fourth cerebral ventricle or dorsal vagal complex are associated with reductions of feeding and body weight comparable to those induced by forebrain leptin administration. Although these observations suggest direct hindbrain action of leptin on feeding and body weight, the possibility that hindbrain leptin administration also activates the Janus kinase/signal transducer and activator of transcription 3 (STAT3) signaling in the hypothalamus has not been investigated. Confirming earlier work, we found that leptin produced comparable reductions of feeding and body weight when injected into the lateral ventricle or the fourth ventricle. We also found that lateral and fourth ventricle leptin injections produced comparable increases of STAT3 phosphorylation in both the hindbrain and the hypothalamus. Moreover, injection of 50 ng of leptin directly into the nucleus of the solitary tract also increased STAT3 phosphorylation in the hypothalamic arcuate and ventromedial nuclei. Increased hypothalamic STAT3 phosphorylation was not due to elevation of blood leptin concentrations and the pattern of STAT3 phosphorylation did not overlap distribution of the retrograde tracer, fluorogold, injected via the same cannula. Our observations indicate that even small leptin doses administered to the hindbrain can trigger leptin-related signaling in the forebrain, and raise the possibility that STAT3 phosphorylation in the hypothalamus may contribute to behavioral and metabolic changes observed after hindbrain leptin injections.

Figure 1

Cumulative change in body weight (A and C) and 24 h food intake (B and D) on 2 d after daily administration of either leptin (n = 6) or vehicle (n = 5) injected via the LV (A and B) or leptin (n = 6) or vehicle (n = 5) injected via the 4V (C and D). Leptin-injected rats exhibited significantly greater loss of body weight, compared with vehicle-injected rats, on both d 1 and 2 of injection, regardless of whether injections were made via LV or 4V. Similarly, leptin-injected rats exhibited reduced food intake, compared with vehicle-injected rats, on both d 1 and 2 after LV leptin injection, but reduction of intake was significant only for d 1 after 4V leptin injection. The failure of reduced food intake to reach statistical significance on d 2 after 4V leptin probably was due to the fact that 4V vehicle itself produced a slight reduction of food intake. Error bars show sem. *, Significant difference from vehicle-injected controls (P < 0.001).

Figure 2

Change in STAT3 phosphorylation (pSTAT3) in the hypothalamus and dorsal hindbrain, after daily LV (A and B) or 4V (C and D) injections of leptin (n = 6) or vehicle (n = 5) over 3 d. Results represent averages from six LV and six 4V leptin-injected rats, five LV and five 4V-vehicle-injected rats. Rats were euthanized 90 min after the final ventricular injection on d 3, at which time brain tissue was collected for Western blot analysis. The top portions of the panels show actual Western blots of pSTAT3 and total STAT3 immunoreactivity for individual animals that received either leptin (L) or vehicle (V). Changes in pSTAT3 immunoreactivity, graphed in the lower portions of the panels, are the average of the ratios of integrated pSTAT3 band density to integrated total STAT3 band density for each treatment group. The Western blot analyses compared the effects of leptin or vehicle treatment within each brain area, but are not suitable for comparisons between brain areas. Error bars show sem. Both LV and 4V leptin injections significantly increased pSTAT3 in both hypothalamus and hindbrain, compared with vehicle injected controls. *, Significant difference from vehicle-injected controls (P < 0.001).

Figure 3

Cumulative change in body weight (A) and daily 24 h food intake (B) for 2 d, after daily administration of either leptin (n = 6) (50 ng in 50 nl) or vehicle (n = 7) (50 nl), injected directly into the nucleus of the solitary tract of the dorsal medulla. Leptin-injected rats lost significantly more body weight than vehicle-injected rats on both d 1 and d 2 of the experiment. Similarly, rats the received leptin ate significantly less than vehicle-injected rats, but the reduction of was statistically significant only for d 1. Error bars show sem. *, Significant difference in reduction of body weight compared with vehicle-injected controls (P < 0.01), and significant difference in reduction of food intake compared with vehicle-injected controls (P < 0.05).

Figure 4

Immunohistochemical staining for phosphorylated STAT3 immunoreactivity in the basomedial hypothalami of rats 90 min after a single 50-nl injection of vehicle (D–F) or leptin (50 ng in 50 nl) (G–I) directly into the nucleus of the solitary tract. A, B, and C illustrate boundaries examined and counted, as illustrated in representative sections D, E, G, and H. To reduce clutter, the boarder between the ArcM and ArcL is not depicted on the immunohistochemical images (D,G, E, and H). The area bounded by the dashed rectangle in C defines the area of the hypothalamus depicted in F and I. ArcL, ARCl nucleus; ArcM, ARCm nucleus; ArcMP, medial posterior ARC nucleus; VMHC, central VMH nucleus; VMHL, lateral VMH nucleus.

Figure 5

Average number of pSTAT3 immunoreactive nuclei in the hypothalamus of rats 90 min after a single 50-nl injection of vehicle or leptin directly into the nucleus of the solitary tract. Error bars show sem. Leptin was associated with significantly increased numbers of pSTAT3 immunoreactive nuclei in the VMH and in all areas of the ARC nucleus (ArcL, ArcM, ArcMP), except the ARCd. There were trends toward increased pSTAT3 in the lateral and perifornical hypothalamus (LH/PeF) and DMH, but these trends did not achieve statistical significance. *, Significantly different from vehicle, P < 0.03.

Figure 6

A–F, Digital images of 30-μm coronal hypothalamic sections illustrating the distribution of fluorogold (yellow) and pSTAT3 immunoreactivity (black nuclear staining). Fluorogold was injected into the hindbrain 24 h before injection of either leptin or vehicle, and brains were collected for immunohistochemistry 90 min after injection of leptin or vehicle. Although fluorogold could be retrogradely transported to neuronal cell bodies with projections to the NTS, the 24 h allowed between FG injection and brain harvesting precludes this route for FG arrival in the hypothalamus. Hence, FG labeling is taken to indicate arrival of the tracer via diffusion in cerebrospinal fluid. A and C, Low- and high-power images from the same section from a vehicle-injected rat. The solid square in panel A encloses the area covered in panel C. Dotted lines indicate the border of the ARC nucleus, including the ArcD, ArcM, and ArcL subnuclei. Note the distribution of fluorogold and the virtual absence of pSTAT3 immunoreactivity in the ARC of this rat that received a vehicle injection (300 μl), but no leptin, 24 h after a 300-μl injection of 2% FG via the same 4V cannula. B and D, Images of comparable fields illustrating the distribution of fluorogold and pSTAT3 immunoreactivity in the ARC when fluorogold was injected into the hindbrain 24 h before injection of leptin via the same 4V cannula. The square in panel B indicates the area covered at higher power in panel D. Note the abundance of pSTAT3 immunoreactive nuclei (solid arrowheads) as well as FG labeling, which exhibits little or no overlap with pSTAT3 immunoreactivity. E and F, Images from the hypothalamus of an animal that received leptin (50 ng in 50 nl) injected into the NTS 24 h after a 50-nl injection of 2% fluorogold injected via the same cannula. Note the very limited fluorgold labeling of hypothalamus adjacent to pial surface and the presence of pSTAT3 immunoreactivity in the VMH (E) and ARC (F). In no case were pSTAT3 immunoreactive cells found to be labeled by fluorogold. Fluorogold distribution was largely nonoverlapping with areas of pSTAT3 immunoreactivity, after 4V injections, and fluorogold labeling of the pial surface could be found in just one of four rats that received 50 ng in 50 nl of fluorogold. G, Dual-label immunohistochemical preparation from a rat injected with 50 ng of leptin in 50 nl, directly in to the dorsal vagal complex. pSTAT3 immunoreactivity was clearly present in both α-MSH immunoreactive neuron (brown, closed arrowhead) and α-MSH-negative neurons (open arrowhead). However, all detectable MSH immunoreactive neurons also were immunoreactive for pSTAT3. α-MSH neurons are directly sensitive to leptin. Thus, colocalization of pSTAT3 immunoreactivity with α-MSH immunoreactivity in the arcuate area is compatible with direct activation of these neurons by hindbrain-injected leptin.