Time for a Pep Talk

Peptide drugs have very recently been in the limelight because of the success of the Novo Nordisk blockbuster GLP-1 agonist Semaglutide (Ozempic, Figure 1).¹ GLP-1 agonists primarily stimulate glucose- dependent insulin release from pancreatic islets, underpinning its use in the treatment of diabetes. An add-on, equally important, function is the slowing of gastric emptying- reducing appetite and aiding weight loss. The latter has driven a frenzied demand for Semaglutide – some relating to genuine medical morbidity and others for cosmetic reasons- a dangerous trend that must have been anticipated. A very specific set of criteria for prescribing the drug to aid weight-loss is in place, however there are websites that will happily sell you something that may or may not be Semaglutide. Buyer beware. Even more remarkable is that it’s an injectable, with demand unhampered by the fairly common fear of needles. 2023 sales of Semaglutide were in excess of $18 billion.  Novo Nordisk’s market value has exceeded the size of the entire Danish economy!¹ A chunk of that money will no doubt be used to defend the tidal wave of looming patent challenges.  Lotte Bjerre Knudsen, chief scientific advisor at Novo Nordisk, shared the prestigious Lasker prize with Joel Habener and Svetlana Mojsov (both academics) for their work on GLP-1 biology. A recent article in Nature Reviews Drug Discovery by Asher Mullard nicely describes the meteoric rise of Semaglutide to blockbuster status.²

In Semaglutide, the key amino acid substitutions relative to wild-type human GLP-1 are glycine to 2-aminoisobutyric acid (Aib) replacement at position 8 and attachment of an octadecanoic diacid to the side chain of Lys-26. There is also a single amino acid substitution at position 34 (Figure 2).³ The peptide chain modifications render the native peptide increasingly resistant to the proteolytic enzyme DPP-4, and the presence of the fatty acid moiety results in a high binding affinity to serum albumin increasing the half-life to approximately 7 days in humans.

DPP-4 enzyme inhibition has a well validated mechanism in diabetes treatment, with several small-molecule drugs on the market, including the Merck compound Sitagliptin (Figure 3).⁴

According to the CDC, around 460 million patients worldwide suffer with type 2 diabetes (around 6% of the total population). In comparison the incidence of cancers are 1.3%.There are currently 80 therapeutic peptides on the market, 200 in clinical phases, and 600 in advanced pre-clinical stages.  An important and lucrative area for the pharmaceutical industry.¹

Synthesis (and purification) of peptides remains a challenging issue, despite a long history beginning with the first synthesis by Fisher and Fourneau in 1901. Development of orthogonal protecting group principles and solid phase synthesis have revolutionised the field. The latter culminated in a chemistry Nobel prize for Bruce Merrifield in 1984.⁵

All that said I’m somewhat ashamed to say I don’t know a great deal about the subject. I do occasionally read papers, but I have to say they don’t hold my interest for very long. One paper I have read recently however, and enjoyed reading, is by Hattori and Yamamoto in JACS.⁶  In it they describe peptide chain formation using unprotected amino acids via isolable silacyclic dipeptides and describes the first reported cross-condensation between two unprotected amino acids in a convergent synthesis of oligopeptides.

The advantage of this work is that it partially addresses the prevailing general limitations encountered during the coupling of unprotected amino acids, including self-condensation (giving 6-membered diketopiperazines), sequential inversion, cyclisation, and over-reaction (Figure 4a). In addition, epimerization via oxazolone formation, a major problem in traditional peptide synthesis cannot occur using this approach (Figure 4b).⁷

In earlier enabling work by Hattori and Yamamoto they describe utilising a unusual looking five-membered ring formed from two unprotected amino acids, tethered with silicon or aluminium, and subsequent reaction with amino acids protected as tert-butyl esters (Figure 5a). However, this is only half the story as it were.⁸ The formation and coupling of a silacyclic dipeptides in which two unprotected amino acids are condensed to generate the isolable silacycle intermediates represents a powerful extension of the methodology (Figure 5b).

The silyl intermediates are prepared by first reacting the appropriate unprotected amino acid with bis(1-imidazolyl)dimethylsilane in DCM to generate the non-isolable mono-silacylic acid (1hr rt), followed by addition of the second amino acid pre-mixed with N-trimethylsilylimidazole (TMS-IM, 2 eqv.)  and N,O–bis– (trimethylsilyl)trifluoroacetamide (BSTFA, 2 eqv) and heating at 50°C for 24 hrs (Figure 6). TMS-IM is important for solubility, whereas BSTFA functions as a silylating reagent for unprotected amino acids. Both have a synergistic effect and independently give low yields. The paper contains a huge number of examples, covering a wide range of amino acids, with high yields and regioselectivities.

The silacyclic dipeptides are apparently stable under atmospheric conditions and are easy to handle. They can efficiently be converted to unprotected dipeptides by stirring in methanol (Figure 6, rt, 1hr).

An example of utility to finish: formation of a pentapeptide intermediate that can be used in the synthesis of Thymopentin, a compound shown to have an effect in restoring macrophage function and a possible sepsis treatment (Figure 7).⁹ Tantalum ethoxide enabled coupling of the Silacyclic tripeptide gave a tetramer, which upon ring opening using TBAF and coupling with the required amino acid-Pfp ester gave the pentapeptide.¹⁰ Hexa- and Hepta- peptide syntheses were also demonstrated using this methodology.

This approach has several advantages over existing methods including convergency, reduction in the use of additives, and the ability to carry out both N– and C-terminal elongation. C-terminal elongation in particular occurs without the formation of oxazolone intermediates that are prone to racemisation.

An interesting paper from an area I don’t usually follow.

See you next time.

References:

1. Big peptide drugs in a small molecule world: J. D. Wade et al, Chem. 2023, 11, 1302169
2. How GLP-1 went from being a hard-to-handle hormone to a blockbuster success: A. Mullard, Rev. Drug Disscov. 1^st Nov. 2024, https://www.nature.com/articles/d41573-024-00177-2; The year of Ozempic: an IP take: https://www.reddie.co.uk/2024/08/30/the-year-of-ozempic-an-ip-take/
3. One-weekly ozempic (semaglutide) injection mechanism of action: https://www.novomedlink.com/diabetes/products/treatments/ozempic/about/mechanism-of-action.html
4. Sitagliptin: A review in type 2 diabetes: L. J. Scott, Drugs 2017, 77, 209–224
5. a) Introduction to peptide synthesis: G. B. Fields et al,  https://doi.org/10.1002/0471140864.ps1801s69; b) Peptides, solid-phase synthesis and characterization: tailor-made methodologies: F. Guzman et al, https://doi.org/10.1016/j.ejbt.2023.01.005; c) Sustainability in peptide chemistry: current synthesis and purification technologies and future challenges: W. Cabri et al, Green Chem., 2022, 24, 975-1020; d) ‘100 years of peptide synthesis’: ligation methods for peptide and protein synthesis with applications to β-peptide assemblies: D. Seebach et al, Journal of Peptide Research, 2005, 65, 229-260
6. Peptide bond formation between unprotected amino acids: convergent synthesis of oligopeptides: T. Hattori & H. Yamamoto, Am. Chem. Soc. 2024, 146, 25738−25744
7. Epimerisation in peptide synthesis: S. Duengo et al, Molecules 2023, 28, 8017
8. Trimethylaluminum-mediated one-pot peptide elongation: Hattori & H. Yamamoto, Chem. Sci., 2023, 14, 5795
9. Stabilized analogs of thymopentin 1. 4,5-ketomethylene pseudopeptides: J. DeGraw et al, Med. Chem. 1997, 40, 2386–2397
10. Substrate-directed Lewis-acid catalysis for peptide synthesis: Yamamoto et al, J. Am. Chem. Soc. 2019, 141, 12288–12295