Or Sunday Fun with Peptide Cutter

Part of my consulting work has led me to reviewing the use of pancratic enzymes for the treatment of solid tumors. At first it sounded bizarre, if not down right dangerous. This post will go over a review by Aitor González-Titos and coauthors, most of whom are from Granada, Spain.

• Both trypsin and chymotypsin are released by the pancreas as zymogens.
• Enterokinase on duodenal enterocytes cleaves tyrpsinogen to the active enzyme trypsin.
• A fine point of the González-Titos review, not mentioned by other reviews I’ve read, is hat It is trypsin that cleaves chymotrypsinogen to the active enzyme chymotrypsin.
• Chymotrypsin and trypsin have similar structures.
• Chymotrypsin is known for its specificity for aromatic amino acids.
• Trypsin prefers basic amino acids like arginine and lysine.
• Three isoforms of trypsin exist: cationic, anionic, and msesotrypsin. The cationic form can auto activate at acidic pH. [1] This has implications under acidic conditons of the tumor micro environment.
• The anionic form of trypsin may auto autoactivate under neutral or alkaline conditions. [1]
• Brief mention was made of mesotrypsin degrading serum trypsin inhibitors like
• González-Titos discussed the four isoforms of chymotrypsin and that all four may be present in the pancreatic secretions used for enzyme therapy.

I’m not including images from González-Titos because this article is not public access. Instead I tried to recreate the essence of one of their figures using public domain software at rcsb.org. One thing I quickly learned was that crystal structures of these enzymes required use of inhibitor to keep them from autoproteolysis one would assume.

Mention was made of pioneering work of Gonzalez and Isaacs who have used pancreatic enzymes to treat many forms of cancer. Both the González-Titos and the Isaacs review make a big deal of the protease-activated receptor 1 as a therapeutic target. There is an excellent review on this receptor. [3] I decided to take one of their target proteins Protease Activated Receptor 1 described in greater detail in a review. [3]

The signal peptide is removed before PAR1 ever reaches the cell membrane. The pro-peptide is removed in order to activate the receptor. Covic and coauthors discussed in length the down stream signaling of PAR1 in cancer. Like the cannabinoid receptors 1 and 2, PAR1 fires through Gα_i, a subunit that activates adenylate cyclase production of second messenger cyclic AMP. Covic and coauthors mentioned other G α subunits with different functions. The Covic review also mentioned the Gα_i12,13, and Gα_q, that have down stream targets other than adenylate cyclase. [3] They did not go into detail as to the mechanisms cleaving a one amino acid versus another. We are now nearing a deep, dark, rabbit hole of heterotrimeric G protein signalling.

2-arachidonyl glyeryl either is a partial agonist of CB1 meaning it does not activate to the same extent as 2-AG. Does clipping PAR1 one amino acid “too soon” result in a partial versus fully activated G protein? If so, for which Gα,pathway? The we get into the realm of antagonist inhibitors versus inverse agonists that take the signalling below baseline and in the opposite direction of the agonist! The non cleaved peptide of PAR1 would be an antagonist.

PPAR1 post translational modifications and signaling…

Pancreatic enzyme therapy may also include amylases. In this thought experiment we can see that removal of glycosylation sites has the potential to influence the binding of the propeptide to the rest of the receptor. Could removal of these glycosylation sites turn the pro peptide from being an antagonist to being an inverse agonist? How does what is removed from the propeptideinfluence the G protein that couples to the intracellular domain? This post is not meant a review, but rather a thoughtful playing around with

PeptideCutter

The featured image of this most is more or less a screen shot from Peptide Cutter. All the user has to do is paste the single letter amino acid code, select the protease and hit “perform.” Unfortunately not all of the proteases used for cancer therapy are available on this site. The matrix metallo proteases that tumor cells produce are also not available. A final caveate is that this site will not allow us to adjust for possible changes in proteolytic activity in the acidic environment of a tumor versus that of the duodenum.

PAR1

Here are the first 90 amino acids of the human PAR1 sequence

MGPRRLLLVAACFSLCGPLLSARTRARRPESKATNATLDPRSFLLRNPNDKYEPFWEDEEKNESGLTEYRLVSINKSSPLQKQLPAFISE

Let’s remove the signal peptide

ARTRARRPESKATNATLDPRSFLLRNPNDKYEPFWEDEEKNESGLTEYRLVSINKSSPLQKQLPAFISE Now let’s do some in silico digestion at PeptideCutter.

This trypsin and chymotrypsin digestion go far beyond the matrix metalloprotease and thrombin digestion of the Covic review. [3] Does the glycosylation on residue 62 have any influence on proteolytic activation? Does really chewing up things change the activation?

Fibrinogen

The Covic review mentioned thrombin. [3] Let’s move on to thrombin’s clot forming substrate fibrinogen. Fibrinogen activation has been covered in a previous post.

MFSMRIVCLVLSVVGTAWTADSGEGDFLAEGGGVRGPRVVERHQSACKDS The first 50 amino acid of fibrinogen. The thrombin site is highlighted. Let’s cut it up!

Interesting that the trypsin site maps with the thrombin site. What does this mean if we are a cancer patient and getting a proenzyme injected in our bodies? Do we worry about blood clots and having vascular event? If we are injecting only the basic chymotrypsin, perhaps we will only have activation in acidic parts of our bodies? The assumption is that this will be a tumor with lactic acid buildup due to the Warburg Effect. What if we have lactic acid buildup from a good workout?

PD-1 and PDL1

Antibody Nivolumab, aka Optiva, is an antibody against the T cell receptor PD1 that binds to the PD-L1 antigen on the surface of many T and B cells. Binding of PD-1 to PD-L1/2 has the effect of activating programmed death in regulatory T cells in such a way as to attenuate autoimmunity but also promote tumor evasion. Nivolumab is used to treat melanoma, renal cell carcinomas, and lung cancer.

Okay, let’s assume that the tumor micro environment is acidic and this activates our basic form of tyrpsinogen to form the active enzyme trypsin. [1] What does this do the the T cell expressing PD-1 that might wander into the tumor? Better yet, let’s go after the PD-L1 on the tumor first.

PD-L1 is really smallish by transmembrane protein standards. It just has one transmembrane domain and a tiny intracellular domain of just 30 amino acids. Let’s just digest the entire extracellular domain minus the signal sequence.

Okay, back to PD-1 on T cells.

We’ve got a little bit more signal peptide in PD-1. Let’s digest 22-170 with trypsin and high fidelity chymotrypsin.

Digestion of the extracellular domain of PD-1 with enterokinase and thrombin yielded no fragments. This speaks to the lack of specificity of digestive enzymes. What do these proteases have to do with checkpoint inhibitors? Well…. everything and nothing. Everything meaning these proteases digest just about everything. Nothing meaning there is absolutely nothing in terms of a specific target.

Trypsin digests everything, but nothing in particular

In this series of thought experiments, lets look at the facilitated glucose transporter Glut1. Glut1 is the name of the protein. GTR1 is the name of the gene. Unlike PD-1 and PD-L1, Glut1 is a large membrane bound protein with 12 membrane spanning helices. Glut1 is important in melanomas that rely on aerobic glycolysis for ATP. There really is not that much on the surface for trypsin and chymotrypsin to digest.

Let’s digest the starred extracellular domain. I decided to throw in all the endogenous human proteases hat are not pancreatic and stomach enzymes. Pepsin was not included. Caspase are proteases released in programmed cell death. Neutrophil elastase is released by neutrophils in response to infections. While it has broad specificity, its site on our Glut1 extracellular peptide are different from trypsin and chymotrypsin.

These pancreatic enzymes are essentially taking a sledge hammer to a tumor rather than a surgical scalpel. If only we could restrict the activity of injected pancreatic enzymes to the acidic tumor microenvironment and not the acidic microenvironment of fatigued skeletal muscle.

Thoughts after playing with Peptide Cutter

It’s starting to become evident why injecting pancreatic enzymes is illegal in the U.S. [2] These enzymes are really good at what they do. It also sort of makes sense why one would want to start with the inactive, zymogen form rather than the active enzymes. Here are a few questions one might want to ask of the clinic

1. Can you prove to me that you are giving me the basic isoform of trypsin?
2. Are you giving me the inactive forms of the enzyme?
3. Are you giving me the amylases too?
4. Are you going to be monitoring for tumor peptides in my blood?
5. Is there any way I can make sure that basic trypsin does not become activated in parts of my body that are naturally acidic?

The pancreatic enzyme approach seems promising. It would be nice if there were a way to target these enzymes to the cancer.

References

1. González-Titos A, Hernández-Camarero P, Barungi S, Marchal JA, Kenyon J, Perán M. Trypsinogen and chymotrypsinogen: potent anti-tumor agents. Expert Opin Biol Ther. 2021 Dec;21(12):1609-1621.
2. Isaacs LL. Pancreatic Proteolytic Enzymes and Cancer: New Support for an Old Theory. Integr Cancer Ther. 2022 Jan-Dec;21:15347354221096077. PMC free article
3. Covic L, Kuliopulos A. Protease-Activated Receptor 1 as Therapeutic Target in Breast, Lung, and Ovarian Cancer: Pepducin Approach. International Journal of Molecular Sciences. 2018; 19(8):2237. https://doi.org/10.3390/ijms19082237  free article