Comments for:

---

Published literature does not support a direct and positive relationship between insulin produced in PVN CRH neurons and pituitary GH secretion.

Submitter: Rhonda Kineman | kineman@uic.edu

Authors: Jose Cordoba-Chacon (UIC), Raul M. Luque and Manuel D. Gahete (University of Cordoba, Spain)

University of Illinois at Chicago (UIC)

Published December 8, 2020

Lee et al., clearly show paraventricular (PVN), corticotropin-releasing hormone (CRH) neurons express Ins2 mRNA and produce proinsulin, which is processed to release C-peptide into the median eminence (ME).  Moreover, they report PVN insulin is suppressed by restraint stress and viral-directed delivery of Ins2 shRNA and this was associated with a reduction in pituitary growth-hormone (GH) mRNA and plasma GH levels. Non-specific overexpression of PVN Ins2 prevented the reduction in GH in restrained mice, but had no effect on GH in unrestrained mice.

The authors conclude “PVN insulin neurons send their neurosecretory nerve terminals to the external zone of the ME, transporting insulin to the anterior pituitary, where it stimulates GH gene expression and secretion...”.  This conclusion is depicted in the graphical abstract.

We respectively disagree with this conclusion based on the following:

1. Although it is clear that C-peptide is present in the ME, no convincing evidence was provided that mature insulin is localized to the ME. Perhaps C-peptide is the bioactive peptide released (1)?
2. The lack of change in the expression of key hypothalamic regulators of GH production, GH-releasing hormone (GHRH) or somatostatin (SST), was used to support a direct action of PVN insulin on GH secretion.  However, lack of transcriptional regulation does not eliminate the possibility that the activity of GHRH or SST neurons could have been altered.  It should be noted that the PVN CRH neurons have recently been shown to be bipolar (2), not unipolar as depicted in the graphical abstract, and therefore have the potential to interact with other neurons, including GHRH neurons (3).
3. The authors state “GH-secreting cells of the pituitary express insulin receptors (41). However, it is still unclear whether a classic insulin signaling pathway mediates GH production.”  However, there is overwhelming evidence that insulin acts directly on the pituitary somatotropes to inhibit GH secretion. The direct inhibitory actions of insulin on GH was first reported in the early 80s using primary rat pituitary cell cultures (4,5), and has since been confirmed in other species, including non-human primates (6).  Of note, reference 41 cited by the authors, used primary pituitary cell cultures of mice and primates, as well as mouse models with somatotrope-specific knockout of the insulin receptor or the insulin-like growth factor I (IGF-I) receptor to demonstrate that insulin acts through its own receptor, not that of IGF1, to directly suppress GH secretion, and this inhibitory effect requires PI3K/Akt signaling (7). 

 

1. Souto SB, Campos JR, Fangueiro SF, et al. Multiple cell signalling pathways of human proinsulin C-peptide in vasculopathy protection. Int J Mol Sci 21(2):645, 2020.
2. Wang Y, Hu P, Shan Q, et al. Visualizing dendritic characteristics of corticotropin-releasing hormone neurons at single-cell resolution in the whole mouse brain. bioRxiv 2020.06.23.168310.
3. Peroski M, Proudan N, Grignol G, et al., Corticotropin-releasing hormone (CRH)-immunoreactive (IR) axon varicosities target a subset of growth hormone-releasing hormone (GHRH)-IR neurons in the human hypothalamus. J Chem Neuroanat 78:119-124, 2016.
4. Melmed S. Insulin suppresses growth hormone secretion by rat pituitary cells. J Clin Invest. 73(5):1425-1433, 1984.
5. Yamashita S, Melmed S. Insulin regulation of rat growth hormone gene transcription. J Clin Invest 78(4):1008-14, 1986.
6.  Luque RM, Gahete MD, Valentine RJ, Kineman RD. Examination of the direct effects of metabolic factors on somatotrope function in a non-human primate model, Papio anubis. J Mol Endocrinol 37(1):25-38, 2006.
7.  Gahete MD, Córdoba-Chacón J, Lin Q, et al., Insulin and IGF-I inhibit GH synthesis and release in vitro and in vivo by separate mechanisms. Endocrinology 154(:2410-20, 2013.

---

Response to Kineman and colleagues

Submitter: Eun-Kyoung Kim | ekkim@dgist.ac.kr

Authors: Eun-Kyoung Kim, Jaemeun Lee, Kyungchan Kim

Department of Brain and Cognitive Sciences, Daegu Gyeongbuk Institute of Science and Technology, Republic of Korea

Published December 8, 2020

We submit the following response to the comments of Kineman and colleagues.

a. Immunohistochemical analysis of brain slices using an antibody against mature insulin is unable to discriminate between brain-derived insulin and pancreatic insulin bound to insulin receptors in the brain. Considering that C-peptide is a proxy for processed insulin, we used the C-peptide antibody instead of the mature insulin antibody. Our data with mice lacking Ins2, a brain-specific form of insulin, show that C-peptide is not detected in the external zone of the ME (median eminence). If C-peptide was the bioactive peptide released from the pancreas or sources other than brain, it would be still detected in Ins2 KO mice. Therefore, our results suggest that C-peptide is from PVN neurons and PVN-derived mature insulin is likely to co-exist there.

b. We agree that a lack of transcriptional regulation does not eliminate the possibility of altering the activity of GHRH or SST neurons, because we did not test the activity of these neurons. Although our Fluorogold data show that PVN insulin neurons have the anatomical properties of parvocellular neurosecretory neurons, we haven't investigated whether these neurons are bipolar rather than unipolar. Further study will be necessary to examine whether PVN insulin neurons interact with other neurons, including GHRH neurons.

c. As discussed in our paper and as you note, we are well aware that insulin can act as a negative regulator of GH production. As such, we expected that pituitary GH production would increase via PVN insulin knockdown. Unexpectedly, however, PVN insulin knockdown decreased GH production under our experimental conditions. Our data of the change in p-AKT level in the pituitary shows the positive correlation between insulin signaling and GH production. Given that the physiological outcomes of insulin signaling could be opposite depending on doses of insulin [1-3], the contrary effects of insulin on GH production could be due to the different local concentrations of insulin produced by PVN insulin neurons compared with much higher concentration of insulin exogenously administrated.

 

References:

1.         Demeyts, P., A.R. Bianco, and J. Roth, Site-Site Interactions among Insulin Receptors - Characterization of Negative Cooperativity. Journal of Biological Chemistry, 1976. 251(7): p. 1877-1888.

2.         de Meyts, P., et al., Insulin interactions with its receptors: experimental evidence for negative cooperativity. Biochem Biophys Res Commun, 1973. 55(1): p. 154-61.

3.         Johnson, J.D. and E.U. Alejandro, Control of pancreatic beta-cell fate by insulin signaling: The sweet spot hypothesis. Cell Cycle, 2008. 7(10): p. 1343-7