CNTF receptor alpha is expressed by magnocellular neurons and expression is upregulated in the rat supraoptic nucleus during axonal sprouting

Abstract

Ciliary neurotrophic factor (CNTF) is expressed by glial cells at multiple levels of the magnocellular neurosecretory system (MNS). CNTF is present in astrocytes in the hypothalamic supraoptic nucleus (SON) as well as in perivascular cells in the neurohypophysis, and a several fold increase in CNTF immunoreactivity occurs in the SON following either axotomy of magnocellular neurons or during axonal sprouting by intact magnocellular neurons. CNTF also promotes survival and stimulates process outgrowth from magnocellular neurons in vitro. While these findings suggest that CNTF may act as a growth factor in support of neuronal plasticity in the MNS, little is known regarding possible expression of receptors for CNTF in the MNS. We have therefore used immunocytochemistry and in situ hybridization to examine the expression of CNTF receptor alpha (CNTFRalpha) in the rat MNS. Robust immunoreactivity for CNTFRalpha was observed associated with oxytocinergic and vasopressinergic neurons distributed throughout the SON. Astrocytes located within the ventral glial lamina (VGL) of the SON were also immunoreactive for CNTFRalpha. Robust hybridization of an anti-sense [(35)S]-cRNA probe to CNTFRalpha mRNA was observed throughout the SON, while binding of a control sense probe was much lower. Grains were found clustered predominantly over neuronal somata, indicative of expression by magnocellular neurons within the SON. We next examined changes in expression of CNTFRalpha mRNA by magnocellular neurons 7 days following unilateral transection of the hypothalamo-neurohypophysial tract. The level of CNTFRalpha mRNA was increased 32% (compared to age-matched intact controls; p<0.05) in magnocellular neurons in the SON contralateral to the lesion, which are undergoing extensive collateral axonal sprouting, but was unchanged in axotomized magnocellular neurons in the SON ipsilateral to the lesion. These findings suggest that CNTF produced by MNS glia and acting via CNTFRalpha may exert neurotrophic effects on magnocellular neurons.

Figure 1

CNTFR-ir was observed in magnocellular neurons and astrocytes throughout the SON. (A–C) Dual fluorescent colocalization of anti-CNTFRα (A), with anti-oxytocin (B), revealed extensive colocalization in oxytocin-ir neurons (C, arrows). Likewise, dual fluorescent colocalization of anti-CNTFRα (D), with anti-vasopressin (E), revealed complete colocalization in all vasopressinergic neuron soma (F, arrows). In addition, CNTFRα-ir (G, arrows) was prevalent in GFAP-ir astrocyte cell bodies within the VGL (H). Close examination revealed CNTFRα-ir astrocyte cell processes originating from the VGL extending into the magnocellular region of the SON (arrowheads). As shown in panels J and K, the observed differences in mRNA levels (see Figure 2) are also reflected in the relative level of immunoreactivity observed when comparing the contralateral intact sprouting SON (J) to the lesioned SON (K) of the same animal at 7 days post lesion. (L) Preabsorption controls revealed that a 10 fold excess of purified rat recombinant CNTFRα completely eliminated immunoreactivity. Negative omission controls revealed the same (not shown). OC = optic chiasm, Magnification bars = 50μM.

Figure 2

In situ hybridization revealed that CNTFRα message was expressed in magnocellular divisions of the PVN and in magnocellular neurons in the SON. (A) This low magnification image of the intact (non-lesion control) PVN shows dense grain distributions located predominantly in the lateral magnocellular division (PVL, circle) and the medial magnocellular division (PVM, arrows) of the PVN. Higher magnification images from the same section shown in (A) further illustrate the distribution of grains in the PVL (B) and PVM regions (C, circle). Note the relatively low grain densities in juxtaposed parvocellular regions of the PVN and the surrounding neuropil of the hypothalamus. (D) The SON of intact non-lesion controls, outlined with arrows, also showed significantly higher grain densities than the surrounding neuropil indicative of high levels of endogenous receptor expression. (E) Sense strand control section from the same animal as in D showing grain densities comparable to background levels. (F) Grain densities within the sprouting SON contralateral to the lesion are significantly higher at 7 days post-lesion ( p< 0.05) than in intact control SON. (G) Sense strand control from the same 7 day post-lesion animal shown in F. OC = optic chiasm. Magnification bars: A = 200 μM, D–G = 100 μM.