The Notch effector gene Hes1 regulates migration of hypothalamic neurons, neuropeptide content and axon targeting to the pituitary

Abstract

Proper development of the hypothalamic-pituitary axis requires precise neuronal signaling to establish a network that regulates homeostasis. The developing hypothalamus and pituitary utilize similar signaling pathways for differentiation in embryonic development. The Notch signaling effector gene Hes1 is present in the developing hypothalamus and pituitary and is required for proper formation of the pituitary, which contains axons of arginine vasopressin (AVP) neurons from the hypothalamic paraventricular nucleus (PVN) and supraoptic nucleus (SON). We hypothesized that Hes1 is necessary for the generation, placement and projection of AVP neurons. We found that Hes1 null mice show no significant difference in cell proliferation or death in the developing diencephalon at embryonic day 10.5 (e10.5) or e11.5. By e16.5, AVP cell bodies are formed in the SON and PVN, but are abnormally placed, suggesting that Hes1 may be necessary for the migration of AVP neurons. GAD67 immunoreactivity is ectopically expressed in Hes1 null mice, which may contribute to cell body misplacement. Additionally, at e18.5 Hes1 null mice show continued misplacement of AVP cell bodies in the PVN and SON and additionally exhibit abnormal axonal projection. Using mass spectrometry to characterize peptide content, we found that Hes1 null pituitaries have aberrant somatostatin (SS) peptide, which correlates with abnormal SS cells in the pituitary and misplaced SS axon tracts at e18.5. Our results indicate that Notch signaling facilitates the migration and guidance of hypothalamic neurons, as well as neuropeptide content.

Fig. 1. Loss of Hes1 does not alter progenitor proliferation or cell death in the early developing ventral diencephalon

Midsagittal sections of mouse embryos at e10.5 and e11.5 were immunostained with phosphohistone H3 (PH3) to visualize proliferating progenitors in the developing ventral diencephalon (VD). At e10.5, PH3-positive cells are detected in the ventricular zone of the VD in both controls (A, arrows) and Hes1 null (B, arrows) animals. At e11.5, like e10.5, proliferating progenitors are restricted to the ventricular zone in both control (C) and Hes1 null animals (D). Cell death was detected using TUNEL in e10.5 and e11.5 control and Hes1 null animals. At e10.5, few cells are undergoing cell death in the VD in control (E, arrows) or Hes1 null (F, arrows) animals. At e11.5 few if any dying cells can be visualized in the VD of both control (G, arrows) and Hes1 null (H, arrows) animals. Two sections from three individual embryos were examined for each genotype. Scale bar indicates 50 μm.

Fig. 2. Loss of Hes1 causes increased and aberrant GABAergic neuron expression and faulty placement of AVP neurons in the SON and PVN at e16.5

A. Represents area within red box of A′ and shows no immunohistochemical detection of arginine vasopressin (AVP) cell bodies in coronal sections at the level of the PL in controls. B. Represents area within red box of B′ and shows that at the same level of the PL, Hes1 null mice show AVP-positive cells in a medial cluster as well as a few scattered cells laterally. C. Represents area within the red box of C′ at the level of the medial median eminence (ME), 198μm anterior to the PL, and shows medial AVP-positive cell bodies in a ventromedial cluster as well as a cluster superior to the optic nerve in the supraoptic nucleus (SON) in controls. D. Represents area within the red box of D′ and shows more dorsal AVP-positive cells at the medial median eminence (ME), 126μm anterior to the PL in Hes1 null animals. Control animals show GAD67 immunoreactivity around the region of the SON at all levels, but not within the SON (E, red box). In contrast, Hes1 null animals display an increase of GAD67-positive cells in and around the SON (F, red box denotes SON). Control animals show AVP immunoreactivity in the paraventricular nucleus (PVN), forming a trapezoid pattern at the third ventricle (G), while Hes1 null animals have more diffuse AVP-positive cells in that region (I). GAD67-positive cells mostly surround the PVN in control animals (H), but are found within the PVN in Hes1 null mice. Fifty sections from 6 individual embryos were examined for each genotype. Scale bar indicates 50μm.

Fig. 3. Loss of Hes1 causes disruption of AVP cell body placement in the PVN and SON at e18.5

Immunodetection of arginine vasopressin (AVP) cell bodies in coronal sections (red, co-labeled with nuclei stained with DAPI in blue) shows robust staining within the paraventricular nucleus (PVN; A, dotted box). In contrast, AVP-positive cells in Hes1 null animals extend outside this region (B, dotted box). AVP-positive neurons within the superior optic nucleus (SON) form a tight cluster in control embryos (C), while they sit in a linear and more diffuse pattern in the SON of Hes1 null animals (D). Forty sections from 4 individual embryos were examined for each genotype. Scale bar indicates 50μm.

Fig. 4. Loss of Hes1 results in pituitary hypoplasia, reduction of AVP neurons in the posterior lobe, and AVP axon misguidance at e18.5

Immunodetection of arginine vasopressin (AVP) axon terminals in coronal sections (red, nuclei stained with DAPI in blue) fill the posterior lobe (PL) of control animals (A), while AVP-positive axons are substantially reduced in the PL of Hes1 null animals (B). AVP-positive axons are present in the median eminence (ME) of Hes1+/+ mice (C), but are reduced in Hes1 null animals and are present in an ectopic cluster lateral to the ME (D). Thirty sections from 4 individual embryos were examined for each genotype. Scale bar indicates 50μm.

Fig. 5. Loss of Hes1 causes misplacement of the median eminence and ectopic AVP axon termination at e18.5

Parasagittal sections cut 30° to the horizontal plane (A, inset), show immunodetection of AVP neurons (yellow, nuclei stained in red) in a cluster superior to the optic nerve in the supraoptic nucleus (SON) at the level of the anterior lobe of the pituitary (AL; A, dotted box). At the same level, Hes1 null animals show a smaller AL, AVP axons traveling through the median eminence (ME), as well a cluster of AVP neurons lateral to the ME (B, dotted box). More posteriorly, AVP neurons are still present in the SON of control animals (C, dotted box). In Hes1 null animals, AVP-positive neurons are present superior to the optic nerve in the SON (D, dotted box) and AVP axons can be detected traveling through the ME. At the level of the posterior lobe (PL) in control animals, AVP-positive axons are visualized traveling through the ME to their target, the PL (E). However, Hes1 null animals show only clusters of AVP axons at this level (F, dotted box), as well as AVP axons in a smaller PL. F′. Magnification of the axon processes in F. Thirty sections from 3 individual embryos were examined for each genotype. Scale bar indicates 50μm.

Fig. 6. Schematic representation of changes in AVP cell body placement and AVP axonal trajectory in Hes1 null mice

Control animals show arginine vasopressin (AVP) positive cell bodies (blue circles) within the region of the paraventricular nucleus (PVN) and supraoptic nucleus (SON) and few AVP immunopositive cell bodies at the level of the median eminence dorsally (ME; A). AVP axons (pink lines) travel from AVP cell bodies in the PVN to the SON and project medially though the ME to reach the posterior lobe (PL; A). Hes1 null animals display AVP-positive cell bodies outside of the PVN and SON, as well as AVP positive cell bodies at the level of the ME and PL in a more ventral region compared to controls (B). AVP axons travel from the PVN to the SON and are found ectopically at the level of the ME, which itself is aberrantly located closer to the PL in Hes1 null mice. Fewer AVP axons terminate in the PL of Hes1 null animals, which is frequently smaller compared to controls (B).

Fig. 7. Loss of Hes1 affects peptide content within the pituitary at e18.5

MALDI MS profile comparing relative intensities of peptides by mass/charge (m/z) detected in Hes1 null and control pituitaries using an acidified acetone extraction method. Somatostatin [92–100] and Pro-AVP [151–168] were found only in Hes1 null pituitaries. Control animals showed various forms of pro-opiomelanocortin (POMC) as well as detection of melanocyte stimulating hormone (MSH) not found in Hes1 null pituitaries. Both Hes1 null and control pituitaries contained cholecystokinin (CCK-8) and ProAVP/AVP.

Fig. 8. Loss of Hes1 results in reduced SS-positive cells in the aPV, altered SS tracts and aberrant expression of SS in the posterior lobe at e18.5

Immunodetection of somatostatin (SS) cell bodies (red, co-labeled with nuclei stained with DAPI in blue) in the mid-aPV nucleus shows SS-positive cells surrounding the third ventricle (3V) in control animals (A), while Hes1 null animals show a decrease in SS-positive cells in this region (B). At the level of the medial median eminence (ME), there are no SS-positive tracts lateral to the ME in control animals (C), while Hes1 null animals show SS-positive axons along the ventral border of the brain (D). Within the posterior lobe, control animals display no SS-positive cells in the posterior lobe (PL; 0 cells, n=4) or intermediate lobe (IL; E). However, Hes1 null pituitaries show aberrant expression of SS-positive cells in the PL (arrow; 1.75 cells ±0.25, n=4) and immunoreactivity in the IL (F). Forty sections from 4 individual embryos were examined for each genotype. Scale bar indicates 50μm.