From Stroma to Scalpel: Celebrating a Mentor in Science and Surgery

It is a tremendous honor to participate in this celebration of Dr Norton’s extraordinary career. Let me first start by expressing my gratitude to Drs Mary Hawn, Electron Kebebew, and George Poultsides for creating a fantastic festschrift event on Stanford University campus 13–14 October 2023. Next, I would like to acknowledge Dr. Norton’s family—his wife Cathy, his children, Tim, John, Meg, and Patrick, and grandchildren, Molly, Jack, and Frank—to thank them for supporting Dr. Norton and sharing him with the surgical community and particularly Stanford Surgery for many years. I am a recent graduate of the General Surgery Residency Program at Stanford and celebrate Dr. Norton as a leader in the Department of Surgery, a clinical and research mentor, a member of my PhD committee, and most fundamentally as a role model surgeon—scientist.

I distinctly recall being a third-year medical student and first hearing about Dr. Norton—thinking, “Wow, if I could find a way to rotate on his service—such a giant in endocrine and surgical oncology—that would be incredible, that would open the door to this big world of surgery.” Not only did I have the opportunity to rotate on his service as a student, but to follow, he graciously provided a letter in support of my residency application. I was fortunate enough to match at Stanford and to rotate on Dr. Norton's service multiple times as a resident.

In terms of mentorship, Dr. Norton is extremely generous. He has such a keen sense for the value and weight his words carry and uses them to propel people forward. He believes in trainees in a way that makes them believe more in themselves. Many surgeons across phases of their careers and across the country, even the world, can attest to this in terms of receiving tangible support—for example unsolicited letters of support. In short, Dr Norton knowns how to use his impact to advance the careers of those around him.

Clinically, Dr. Norton was truly exceptional to train with (Fig. 1). His patients absolutely adored him and as a student or resident on his service, you got to ride this wave. In this regard, I spoke with several recent graduates of Stanford’s training program about the experience of rotating with Dr. Norton—from these conversations and my own experience, I will share a few anecdotes:

Dr. Norton cared not only about the patients, but about everyone involved in the clinical arena. He made a point of rounding solo with the medical student on service between cases, demonstrating through action that each member of the team has a critical role.

Dr Norton was always available for clinical issues and he would usually be standing right beside you as a resident if something went wrong, helping to trouble shoot the situation. He always answered his phone by the second ring in the middle of night. He was also reliably punctual for weekend rounds—this sort of rigor leaves a lasting impression on those working with him.

Dr Norton knew everything about his patients, which was a great example as a chief resident to see how deeply he cared even though he was so senior and had been doing the work for many years.

Dr Norton has an incredible sense of humor, he knows how to keep it light and to see moments of humor even when not everything is funny, which is so often the case in surgery. He also had a certain way of forgetting the “right” stuff, so to speak, of seeing the big picture.

As residents, Dr Norton pushed us to really understand operations, to be independent, he taught us how to “trust our gut,” and to not be afraid to make a “big move.” He showed us the confidence to get out of trouble in the operating room.

Fig. 1

Dr. Norton the Surgeon. Intraoperative image of Dr. Jeffrey Norton operating alongside Dr. E. John Harris, Dr. George Poultsides, and their respective surgical trainees (Drs. Katie Blevins, Patrick Thompson, and Ioana Baiu) at Stanford Hospital, 2021

I have also had the opportunity work with Dr. Norton extensively in the laboratory for over yearly 6 years now (Fig. 2). Transitioning to the lab and slowing down clinically was a significant change for Dr. Norton, but he was totally committed to the endeavor. In his recent years working in Dr. Michael Longaker’s Laboratory in the Hagey Laboratory at Stanford, he brought his extraordinary wealth of knowledge and experience right to the bench where the students, residents, and post-doctoral fellows were working.

Fig. 2

Dr. Norton the Scientist. Photograph in Hagey Laboratory at Stanford University reviewing slides containing human PDAC tissue samples for research evaluation with post-doctoral fellow Dr. Deshka Foster

Early on in our time together in the lab, we started a project to explore cancer associated fibroblasts (CAFs) in pancreatic ductal adenocarcinoma (PDAC), which has since expanded into multiple pieces and is a big focus in the Longaker laboratory.^1,2,3,4,5 Pancreas is notoriously finicky tissue with which to work. Many conventional assays do not run smoothly requiring trouble shooting and this can be experimentally frustrating. Some of my best memories of my PhD are spending late nights myself, Dr Norton, and several of the students rotating in the lab, doing flow cytometry on human pancreatic tumor tissue samples to isolate CAFs for analysis. Dr Norton would have operated all day, but then he would stay late into the evening to help sort the cells—a testament to how patient and dedicated he is.

But Dr Norton is also impatient, he wants experiments to move quickly, he wants results. He is always circling back to the underlying goal of making an impact on the disease and keeping the kind of urgency patients feel in our minds. This serves as a reminder of the fragility of time and of life. I have developed a deep appreciation for this perspective. Another feature of working with Dr. Norton across the board, but perhaps especially in the laboratory, is that nothing is ever “too small” for Dr Norton. He is always willing to help with tissue staining for example, and whenever there is an experiment with long incubation steps, these pauses come with stories from Dr. Norton, which was truly wonderful as a trainee.

One particular piece of our CAF work involves our effort to untangle the heterogeneity and functional roles of CAF subtypes.¹ The solid tumor microenvironment is composed of transformed cancer cells, along with non-transformed stromal cells including CAFs. CAFs secrete the extracellular matrix that makes desmoplastic tumors, such as pancreatic cancers, firm and interact extensively with other cells in the tumor microenvironment (Fig 3). However, CAFs have yet to become a target in mainstream cancer therapy because CAF heterogeneity and their roles in solid tumors remain incompletely understood. We set out to study CAF heterogeneity with the aim of identifying reliable therapeutic targets and we hypothesized that there might be commonalities among CAFs derived from different types of solid tumors that could be leveraged in this regard. We began by conducting a series of single-cell RNA-seq and assay for transposase-accessible chromatin using sequencing (ATAC-seq) experiments, including multiome sequencing, which entails paired assessment of gene expression and chromatin accessibility from the same cell at the single cell level, across multiple types of solid tumors in mice from cells harvested at various time points over the course of tumor progression. From the transcriptionally defined cell populations, we focused our analysis on the CAF subgroups to identify functional characteristics on the basis of gene expression and chromatin accessibility, which pooled the subtypes into three CAF superclusters termed immunomodulatory (IM), steady-state like (SSL), and mechanoresponsive (MR). We wondered how these CAF subtypes were distributed throughout the tumor; to explore this, we applied the 10x Genomics Visium spatial transcriptomics protocol with cell level mapping of our multiome sequencing to the spatial transcriptomics dataset, which revealed spatial delineation of the CAF clusters. Quantification of cell-cell neighbor interactions showed strong interactions between IM CAFs and immune cells, whereas MR CAFs interacted most with epithelial tumor cells and each other. These data support the functional relevance of the delineated CAF subtypes in terms of how the tumor microenvironment is constructed and behaves.

Fig. 3

Studying cancer associated fibroblasts in pancreatic ductal adenocarcinoma. Representative immunofluorescent stained slide of a mouse pancreatic ductal adenocarcinoma (PDAC) tumor sample created using allograft injection of PAN02 cells showing infiltration of cancer associated fibroblasts (CAFs) into the tumor parenchyma and surrounding fibrotic stroma. CAFs and stroma stained with aSMA (a-smooth muscle actin, green), COL1 (collagen 1, yellow), COL3 (collagen 3, red), and DAPI (blue) stained cell nuclei

We then applied the same experimental paradigm to human PDAC and breast cancer specimens and developed an integrated dataset of CAFs from these two tumor types. We found that our mouse-cancer defined CAF clusters were well recapitulated in these human data. We then integrated CAFs from a third human cancer type—basal cell carcinoma. Using this pooled solid tumor dataset, we validated the functional relevance of the CAF subtypes we defined by showing that (1) fibroblast-specific disruption of mechanoresponsive pathway signaling leads to increased SSL CAFs and more aggressive tumor biology, and (2) conversely that immune checkpoint inhibition results in a reduction in SSL CAFs and a proportional increase in IM CAFs in patients with treatment responses to the checkpoint inhibition. These results show a ‘‘push-pull’’ dynamic among CAF superclusters, the balance of which has clear translational implications (Fig. 4).¹

Fig 4

Multiomic analysis reveals conservation of cancer associated fibroblast phenotypes across species and tissue of origin. Graphical abstract, reproduced with permission (Foster DS, Januszyk M, Delitto D et al. Multiomic analysis reveals conservation of cancer-associated fibroblast phenotypes across species and tissue of origin. Cancer Cell 2022; 40 (11): 1392–1406 e1397.)

Using our extensive experience in the Hagey laboratory studying benign fibrosis in wound healing as a fresh lens by which to approach this topic, another project that Dr. Norton has collaborated on now for many years explores abdominal adhesions. We kindly joke that such a project feels somehow appropriate since being such a busy surgical oncologist for so many years, Dr. Norton has probably spent more time creating adhesions than any surgeon in the country! But in all seriousness, abdominal adhesions are a major clinical problem for which there are no proven therapies to prevent or treat. One reason for the dearth of preventative therapies is that up until recently, research into the biology underlying adhesion formation has been limited. Several years ago we set off on the daunting task to try to bridge this gap in knowledge.⁶ We began by developing a mouse model for abdominal adhesion formation, and then analyzed abdominal adhesions in mouse and human tissues at a cellular, transcriptomic, and protein level. First, we wondered about the origin of adhesion forming cells. Using a parabiosis mouse model, we showed that adhesions form largely from local cells. Using the rainbow mouse model with local application of 4-hydroxytamoxifen liposomes, we showed that adhesions proliferate polyclonally, suggesting local heterogeneity. To delineate the tissue source of these cells, we next designed an abdominal wall transplant model and found that the majority of cells in the adhesion interface are derived from the visceral rather than parietal peritoneum—consistent with our clinical observations. We then examined adhesion cells at the single cell level from mouse and human adhesion tissue; we identified that adhesions are formed by a heterogeneous population of fibroblasts. Considering mouse adhesion fibroblasts isolated at multiple timepoints postoperatively, we see upregulation of proinflammatory and profibrotic genes including genes in the JUN pathways—which is activated early and expression is maintained over the course. We validated this timecourse of early JUN activation and downstream pathway expression at the protein level, and further found that the amount of adhesion fibrosis dramatically increases with increased JUN expression using a JUN over-expressing mouse model.

Next, we applied a small molecule JUN inhibitor called T-5224, which dramatically decreased adhesion formation. JUN is believed to have conserved function across species. To validate the translational potential of our results, we found that primary human adhesion fibroblasts cells also express JUN. Using CRISPR Cas9 to knockdown JUN in human adhesion fibroblasts, we see decreased expression of fibrosis genes and decreased fibroblast proliferation. With virally mediated JUN overexpression, collagen expression was significantly elevated compared with control and significantly decreased with Jun knockdown. These results provide us with a comprehensive understanding of abdominal adhesion biology and a pathway towards prevention of this challenging clinical problem (Fig. 5).⁶ Given our successful prevention of adhesion formation in our model using local application of T-5224, we are compelled to move towards clinical translation of this therapeutic pathway, which is currently underway at this time.

Fig. 5

A Intraoperative photographic examples of abdominal adhesions, green arrows indicate adhesive bands tethering bowel to the abdominal wall (left panel) and tethering loops of bowel together (right panel); B surgical schematic showing adhesions forming primarily after open laparotomy (right panel) compared with uninjured abdominal anatomy (left panel); and C schematic showing proposed JUN-related signaling pathways driving the process of abdominal adhesion formation. The left panel shows the acute phase response following tissue injury by which JUN is initially activated. The right panel shows the chronic profibrotic state that is established in adhesion fibroblasts. Blue circle highlights the AP-1 complex of which JUN and FOS are members, red indicates the proposed therapeutic role of JUN inhibition. ECM extracellular matrix. Panels B and C reproduced with attribution (Foster DS, Marshall CD, Gulati GS et al. Elucidating the fundamental fibrotic processes driving abdominal adhesion formation. Nat Commun 2020; 11 (1): 4061. Copyright © 2020, The Author(s)

I must briefly acknowledge individuals who have worked extensively on the projects I discussed in this article, as basic and translational research in particular is very much a “team sport”—primarily Dr. Longaker, in whose laboratory all of these projects were conducted, Drs. Daniel Delitto, Michael Januszyk, Jason Guo, Michelle Griffin, and Charlotte Berry, and many others, as well as our ongoing collaborations with the Appel, Chang, and Wernig laboratories at Stanford University.

In summary, Dr. Norton represents an exemplar surgeon–scientist and is an extraordinary inspiration in this regard, particularly as he remains forever humble, kind, and caring for all those he encounters. I would like to thank Dr. Norton for his mentorship, encouragement, kindness, guidance, and enthusiasm for those of us working our way along the surgeon–scientist path, and to honor this role that he has played for so many students, residents, fellows and post-docs across the country over the years.