Next, we have Jacqueline Boyle, who's going to discuss the activation of the canonical EGFR pathway in Cushing's disease. Good afternoon. I'm Jackie Boyle. I am a third-year medical student currently working at the NIH in the Chivalina Lab, and today I'm going to be talking about the role of EGFR in Cushing's disease. So Cushing's disease is caused by macroadenoma one centimeter in size or smaller that exists in the pituitary. These tumors secrete excess ACTH, which goes to the adrenal glands, causing a state of hypercortisol anemia related to a high morbidity in mortality in patients. This is why the gold standard is currently the surgical resection of these adenomas. During surgical resection, we are often able to collect normal tissue as well that can act as a syngenetic comparator to our adenoma tissues for a variety of our experiments. In addition, many of these tumors are not able to be completely resected and often recur, which is why we need an additional type of treatment and other research. So it's well established that EGFR is overexpressed in pituitary microadenomas, which is unique to these benign tumors compared to other pituitary adenomas, non-secretory and secretory. In 2015, it was established that approximately 30 percent of these corticotropanomas carry a USP8 mutation. The USP8 mutation is a mutation of the deubiquitinase of the EGF receptor. This causes increased deubiquitination of the receptor, increased recycling to the cell membrane, and therefore an activation of the ERG pathway and ACTH synthesis. Despite this knowledge, there is a subselect of corticotropanomas that have EGFR amplification but lack a USP8 mutation, and this is where our research is most interested in investigating. So to do this, we first started with samples of adenoma with adjacent normal from our patient samples looking at a multiplex immunohistochemistry for EGFR as well as activated phosphorylated tyrosine 992 EGFR. We noted that in the adenoma and the normal, there was abundant expression as well as increased expression in the adenoma compared to normal in the activated form. To confirm, in order to understand how this EGFR signaling was occurring, we first needed to look at two different pathways. EGFR signaling occurs through the primary canonical pathway. It's a ligand-activated pathway either through heparin-binding EGF, EGF, or a variety of others causing dimerization, phosphorylation, and primarily activation of the GRB2-RAS-RAF pathway and the AKT pathway. In addition, there is a pathway that occurs without ligand activation. This is the non-canonical pathway through autophosphorylation and dimerization and targets a variety of different genes, primarily interferon genes such as IFIT1 and IFI27. To understand which pathway was happening in our tumors, we took human tumor samples from surgery, both the normal and adenomatous tissue, and sent them for a gene expression microarray of approximately 29,000 probes. Our findings first found statistically significant increased levels of POMC in our adenoma compared to normal, which is unsurprising as POMC is the precursor to ACTH. Interestingly, though, there were no changes found in the EGFR receptor at the gene level. But upon further investigation, we looked at the two different pathways, canonical ligand-activated and non-canonical. Finding amphoregulin and noregulin-1, which are both ligands of EGFR, were upregulated in our adenoma compared to normal. RAS pathway, a pathway of the canonical activation, was also activated. However, non-canonical EGFR genes such as TRAIL and IFI27 were not found to be statistically different between our adenoma and tumor. To confirm these findings, we looked at protein expression of the two ligands thought to be activating the pathway, noregulin-1 and amphoregulin. Noregulin-1 was found to be abundantly expressed in both USP8 mutated and USP8 wild type. So regardless of their mutational status, they had this noregulin present. In addition, we found amphoregulin to be upregulated in the tumor, be greater expressed there, as well as the cytoplasm of these tumors, which is where we'd expect it to be activated. Another focus of our lab is understanding the metabolic reprogramming that happens in these pituitary adenomas. Interestingly, despite their benign status, they have similar patterns to cancer cells. They have increased levels of glute. They have activation of glute-1, as well as recently findings of the lactate dehydrogenase A enzyme, which converts pyruvate to lactate and is an important driver of glycolysis. We found increased levels of LDHA in our adenoma compared to normal in USP8 wild type tissues. So the question was, is EGFR a potential driver of these glycolytic changes that we see in tumors? In order to do this, we first needed to have a baseline of our tumors. So we took normal and adenoma samples collected from patients during surgery and did a cell culture with these, then processed them through the XF96 system with Seahorse to determine their oxygen consumption rate, which is a surrogate for mitochondrial respiration, as well as their extracellular acidification rate, which is a surrogate for glycolytic flux. Finding little differences in mitochondrial respiration, but significant differences in the glycolytic flux between the adenoma and the normal tissues. So understanding this, we needed to bring EGFR into the picture. We decided to transfect EGFR into Chinese hamster ovary cells since these are a very benign and otherwise normal cell. Upon EGFR transfection, we found abundant level increases of glycolysis at both the baseline and when the mitochondrial were stressed and glycolysis was forced to create upregulation. So this proof of concept demonstrating that EGFR can create a glycolytic shift similar seen in our adenoma cells was the first step in connecting these two. Our future studies will be focused on connecting this EGFR transfection, our EGFR expression to our USP8 wild type tumors and understanding their role in shifting the tumors into a glycolytic type state. So in conclusion, aberrant EGFR signaling plays a significant role in corticotropanomas regardless of their USP8 mutation and the EGFR signaling to the cell membrane. EGFR expression appears to be occurring through a canonical pathway, ligand activated, potentially with amphoregulant and noregulant upregulation. And aberrant EGFR signaling may account for the metabolic changes seen in corticotropanomas and lead to a targeted treatment strategy. And I would like to thank Dr. Chittigoina and the rest of the surgical neurology branch at the NIH. Thank you.

EGFR activation

Cushing's disease

pituitary adenomas

USP8 mutation

gene expression microarray analysis