Therapeutic peptides: current applications and future directions

Therapeutic peptides are a unique class of pharmaceutical agents composed of a series of well-ordered amino acids, usually with molecular weights of 500-5000 Da 1 . Research into therapeutic peptides started with fundamental studies of natural human hormones, including insulin, oxytocin, vasopressin, and gonadotropin-releasing hormone (GnRH), and their specific physiological activities in the human body 2 . Since the synthesis of the first therapeutic peptide, insulin, in 1921, remarkable achievements have been made resulting in the approval of more than 80 peptide drugs worldwide. The development of peptide drugs has thus become one of the hottest topics in pharmaceutical research. The first half of the 20th century witnessed the discovery of several life-saving bioactive peptides, such as insulin and adrenocorticotrophic hormone, which were initially studied and isolated from natural sources. The discovery and development of insulin, a peptide with 51 amino acids, has been considered as one of the monumental scientific achievements in drug discovery. It was first isolated by Frederick Banting in 1921 and further developed by Frederick and Charles Best 3 , 4 , and was already available for patients with diabetes mellitus just a year after its first isolation. In 1923, insulin became the first commercial peptide drug and has since benefited thousands of diabetes patients to date. However, the production of human insulin during the 20th century could not keep up with the high market demand, and animal-derived insulins, such as bovine and porcine insulin, dominated the insulin market for almost 90 years until they were replaced by recombinant insulin 5 , 6 . More peptide hormones and their receptors with therapeutic potential were identified and characterized from the 1950s to the 1990s 7 . Meanwhile, the technologies used for protein purification and synthesis, structure elucidation, and sequencing made substantial progress, thus accelerating the development of peptide drugs, leading to nearly 40 peptide drugs being approved worldwide. Notably, synthetic peptides such as synthetic oxytocin 8 , synthetic vasopressin 9 , and recombinant human insulin 10 , 11 began to be developed in addition to natural peptides. Peptide drug development entered a new era with the advent of the 21st century, since when advances in structural biology, recombinant biologics, and new synthetic and analytic technologies have significantly accelerated the process. A sophisticated system of peptide drug development has been established, including peptide drug discovery, drug design, peptide synthesis, structural modification, and activity evaluation. A total of 33 non-insulin peptide drugs have been approved worldwide since 2000 (Table 1 ). In addition, these peptide drugs are no longer simply hormone mimics or composed simply of natural amino acids. For example, enfuvirtide is a 36-amino acid biomimetic peptide mimicking human immunodeficiency virus (HIV) proteins used in combination therapy for the treatment of HIV-1 12 , 13 ; ziconotide 14 , 15 is a neurotoxic peptide derived from the cone snail Conus magus , which was approved in 2004 and is used to manage severe chronic pain; teduglutide is a glucagon-like peptide 2 (GLP-2) analogue used to treat short bowel syndrome 16 , 17 , and is manufactured using a strain of Escherichia coli modified by recombinant DNA technology; and liraglutide is a chemically synthesized analogue of human glucagon-like peptide 1(GLP-1) 18 , 19 , made by attaching a C-16 fatty acid (palmitic acid) with a glutamic acid spacer on lysine residue (position 26 in the sequence), which acts as a GLP-1 receptor agonist to manage type 2 diabetes mellitus (T2DM). All these peptide drugs have been used in a wide range of therapeutic areas, such as urology, respiratory, pain, oncology, metabolic, cardiovascular, and antimicrobial applications 20 – 24 . To date, more than 170 peptides are in active clinical development (Table 2 ), with many more in preclinical studies 1 , 7 . Table 1 Peptide drugs approved since 2000, with their targets and indications Target name Peptide name First approval Approved indication(s) GLP-1 receptor Exenatide 462 2005 Indicated for Type 2 Diabetes Mellitus Liraglutide 463 2009 Lixisenatide 464 2013 Albiglutide 465 2014 Dulaglutide 466 2014 Semaglutide 467 2017 GLP-2 receptor Teduglutide 468 2012 Treatment of Short bowel syndrome and malabsorption GC-C receptor Linaclotide 469 2012 Treatment of irritable bowel syndrome (IBS) with constipation and chronic idiopathic constipation Calcitonin receptor Pramlintide 470 2005 Treatment of Type 1 and Type 2 Diabetes Mellitus GnRH receptor Abarelix 471 2003 Treatment of advanced prostate cancer Degarelix 472 2008 Binding to active site of the 20S proteasome Carfilzomib 473 2012 Treatment of multiple myeloma NOD2 protein Mifamurtide 474 2009 Treatment of high-grade, resectable, non-metastatic osteosarcoma VIP1 receptor Aviptadil 475 2000 Treatment o

Peptide drug discovery Natural peptides/hormones in the human body The history of peptide drug discovery started by exploiting natural hormones and peptides with well-studied physiological functions for treating diseases caused by hormone deficiencies, such as a lack of insulin required to regulate blood glucose levels in patients with T1DM or T2DM. Diabetes is treated either by insulin injection or by stimulating insulin secretion-related targets such as GLP-1 receptor, to produce insulin 48 . Searching for natural peptides and hormones or replace them by animal homologues, such as insulin, GLP-1, somatostatin, GnRH, 8-Arg-Vasopressin, and oxytocin, were the initial strategies used for peptide drug discovery and development (Fig. 3 ). However, the drawbacks associated with these natural peptides aroused interest in optimizing their natural sequences, leading to a series of natural hormone-mimetic peptide drugs. Fig. 3 Structure of human insulin and human insulin-derived drugs. Structure of human insulin (left, PDB: 1XDA). Modifications on its residues (B-Chain: B3: Asn, B28: Pro, B29: Lys; A-Chain: A21: Asn) resulted in several short- and long-acting insulin drugs (right, see table)

Many bioactive peptides from bacteria, fungi, plants, and animals possess therapeutic properties, such as snake venom, which is considered as a vascular endothelial growth factor (VEGF) analogue, VEGF-F or svVEGF 61 – 63 . They are usually disulfide-rich cyclic peptides of no more than 80 residues, which can induce cytotoxicity by targeting ion channels and other membrane-bound receptors 1 , 64 . Venom peptides from snakes and scorpions have been modified for therapeutic applications. In addition, exenatide (Fig. 5a ), optimized from Gila monster venom 65 is a GLP-1 agonist and ziconotide, a venom peptide derived from Conus magus , has been used to treat chronic neuropathic pain 66 , 67 . Fig. 5 Sequences and structures. Exenatide ( a ) and lugdunin ( b ) Non-ribosomal peptides (NRPs) comprise another class of peptides identified from natural products. The non-standard residues contained in the sequence mean that NRPs are not produced through the traditional biosynthesis pathways via ribosomes 68 , but are produced by non-ribosomal peptide synthetases via a pathway consisting of initiation, elongation, and termination modules 69 , 70 . Compared with peptides synthesized by ribosomes, NRPs are more resistant to hydrolases and show increased stability in vivo. The most-studied NRPs are mainly derived from bacteria and fungi, including vancomycin, cyclosporin, lugdunin 71 , 72 . (Fig. 5b ), and teixobactin with antibacterial activities, and a-amanitin, nanocystin A, and actinomycin with anti-tumor activities 73 , 74 . In addition, cyclodepsipeptides are cyclic peptides that comprise a specific class of NRPs usually identified in plants 75 – 77 , such as enniatin B and emodepside 78 , 79 . These peptide drugs display enhanced plasma stability that enables their oral delivery. However, the synthesis and structure-activity relationships study of NRPs represent one of the most challenging and exciting areas of research for NRPs.

Phage display is a highly effective and robust technology used to identify ligands of biological targets, first reported by Smith in 1985 93 . Phage display uses recombinant technologies to engineer target ligands on the surface of the bacteriophage 94 . Only peptides containing proteinogenic amino acids, rather than NRPs, are produced in the phage. This high-throughput sequencing method can be used to identify drug leads, including antibodies and peptides 95 , 96 . Phage display has been widely used to discover new peptide ligands. Lerner et al. reported the discovery of potent peptide analogues of GLP-1 and other membrane receptor ligands by phage display, including proteins, peptides, and venoms, which mainly act as agonists 97 – 100 . In addition, peptides targeting transforming growth factor (TGF)-β1 101 or epidermal growth factor receptor (EGFR) 102 , and peptide antagonists that disrupt the fibroblast growth factor (FGF)-1-FGFR1 interaction 103 are good examples of peptide drugs discovered by phage display. Recent developments in phage display technology have focused on searching for more efficient screening protocols to simplify ligand selection among enormous amounts of data, such as by reducing phage panning cycles 104 . Heinis et al. used an “on-phage” modification technology to obtain chemically modified peptides from traditional phage display to obtain a bis-thioether cyclic peptide 105 . Another strategy involves developing novel display approaches. For example, Schumacher et al. developed a mirror-image phage display to explore D-chirality peptides 106 , 107 , and Szostak et al. performed mRNA display to discover and select macrocyclic peptides with unnatural amino acids 108 – 110 . Suga et al. used ribosomal display to exploit lead peptides, including bioactive macrocyclic peptides, containing D-amino acids and unnatural amino acids 111 – 113 . These developments have allowed the construction of numerous display libraries for the discovery of new peptide candidates.

As a particular class of therapeutic agents, the biological activity of peptides is intimately related to their chemical structure. Following the synthesis of peptides, they need to be modified using medicinal chemistry techniques to mimic, stabilize, or construct an ideal secondary structure to improve their biological activity and achieve selectivity, stability, and solubility of the peptide drugs 132 . Before modification of the lead peptide drug candidate, it is necessary to identify the minimum active sequence with the desired biological properties. Classical sequence scanning, termed alanine-scan 29 , 133 , 134 , is then commonly used to replace each residue with alanine to produce a series of lead peptide analogues to determine which key residues confer the biological activity of the lead peptide: a decrease in activity suggests that the replaced residue was important, whereas a non-significant reduction of activity suggests that the replaced residue was redundant. Further modifications of the replaceable residues and C- and N-termini of the lead peptide are then carried out to produce the final peptide drug 135 . Backbone modification of peptides One of the main reasons for backbone modification is to improve the proteolytic stability of the peptide. Proteolytic sites in the peptide can be identified by stability studies and metabolite determination 136 . Backbone modification includes the substitution of L-amino acids by D-amino acids 137 , 138 , insertion of methyl-amino acids 137 , 139 , and the incorporation of β-amino acids 140 and peptoids 141 – 143 . Introducing these non-natural amino acids into the peptide sequence, particularly at the proteolysis site, is an effective strategy for extending the plasma half-life of peptide drugs. A successful example is selepressin, which was derived from vasopressin and has similar target selectivity but a longer plasma half-life 144 , 145 .

The weak forces in peptides, such as hydrogen bonds, van der Waals forces, and intramolecular hydrophobic interactions are not adequate for a stable secondary structure conformation. Additional modifications of the backbone, N- or C-termini, or side-chains to mimic the structures of natural products or hot spots in PPI and stabilization of secondary structures are therefore needed to produce promising peptide drug candidates 149 , 150 . Peptide cyclization . Cyclization is a common peptide modification technique that can include various strategies, such as head-to-tail, backbone-to-side chain, and side chain-to-side chain cyclization (Fig. 7 ) 151 – 153 . Peptide cyclization can increase proteolytic stability 154 , 155 , and cell-permeability 156 – 158 , and allows mimicking and stabilization of the peptide secondary structure. Without being connected to other peptides, a single peptide sequence cannot form loop or turn structures, but cyclization facilitates the formation of these secondary structures by pre-organizing intramolecular interactions 159 , 160 . Peptide cyclization is also commonly applied to stabilize other secondary structures, such as α-helixes and β-sheets 161 – 163 . Fig. 7 Strategies of peptide cyclization and stabilization of α-helices, β-sheets, and β-strands. The establishment of intramolecular cross-links can stabilize different secondary structures of peptides. Side chain cross-links between i and i + 4 or/and i + 7 and hydrogen bond surrogate cross-links can stabilize α-helices. Side chain-to-side chain, head-to-tail, and side chain-to-tail cyclization can stabilize turn, loop and β structures (β-sheets and β-strands). The D-Pro-L-Pro scaffold can specifically stabilize antiparallel β-hairpins Peptide mimicking of α-helices and stabilization . Helices are one of the commonest types of protein secondary structures, representing about 30%-40% of all protein structures 164 . The α-helix is formed by intramolecular hydrogen bonds 165 and accounts for 90% of helix structures 165 . Mimicking the α-helix in peptides enables the identification of modulators of PPIs. The α-helix can be stabilized either by building cross-links through side chains or replacing hydrogen bonds by covalent bonds (referred as hydrogen bond surrogates, HBS). In the α-helix structure, the side chains of amino acids at positions i, i + 4, and i +7 are on the same side, and building cross-links through i and i + 4 or i and i +7 effectively approach backbone atoms and help to form hydrogen bonds in helical structures 166 – 168 . Great efforts have been made to investigate different cross-links, such as lactam-based cross-links (Fig. 7 ), with the formation of a lactam bridge through the side chain of glutamic acid or aspartic acid with lysine 169 , the formation of disulfide bonds by replacing residues with cysteine or homocysteine 170 , and biselectrophilic linkers 171 , 172 . Stapled peptides represent a recent new cross-linking approach introduced to stabilize the α-helix structure, using non-natural electrophilic amino acids to replace residues at the i and i + 4 or i and i +7 position, and form ligations with nucleophilic cross-links 154 , 173 , 174 . The HBS modification strategy involves replacing one hydrogen bond of the α-helix peptide with a covalent bond to pre-organize the helical structure. Cabezas and Satterthwait first used hydrazine links to build an HBS peptide to mimic an α-helix 175 . The Arora group has also carried out extensive work on HBS peptides, using alkene linkers to stabilize the α-helix 176 – 179 . They recently started to use the HBS strategy to stabilize β-hairpins 180 – 182 , as well as the biological activities of these modified peptides 181 , 183 – 185 . We also used the HBS peptide modification strategy in our previous work, focusing on designing a full SPPS pathway to simplify the application of HBS in α-helix mimicking and stabilization 186 , 187 . Peptide mimicking of β-strands and β-sheets . β-sheets and β-strands represent another class of protein secondary structures, based on turn mimics. The modification of peptides to stabilize β-sheets is usually achieved by the introduction of D-amino acids, such as D-Pro, to form a turn structure in the sequence. D-Pro-L-Pro templates are a well-known scaffold for stabilizing antiparallel β-hairpins in several successful PPI inhibitors 188 , 189 . Macrocyclization or amyloid beta-sheet mimics have also been applied to create β-sheets and β-strand structures 190 – 193 . Chemical modification is an effective method of producing peptide analogues with the desired structures. The improved stability and activity have resulted in the introduction of several peptide drugs into the clinic, such as selepressin, liraglutide, and semaglutide. However, some modifications cannot improve the proteolytic stability and activity simultaneously. For example, the insertion of D-amino acid can usually help to extend the plasma half-life of the peptide, but

Natural proteins are synthesized from 20 canonical amino acids, and this limited and conservative repertoire of amino acids significantly restricts the diversity and complexity of protein structures and functions. Genetic code expansion was developed two decades ago as a technology to overcome this limitation (Fig. 8 ) 218 , 219 . Genetic code expansion allows for the site-specific incorporation of non-canonical amino acids (ncAAs) with novel chemical and physical properties into a growing polypeptide during protein translation 220 , 221 . Four components are required to achieve this: 1) an ncAA with the desired chemical and physical properties; 2) a unique codon that specifies the ncAA, e.g., an amber stop codon (UAG) or quadruplet codon; 3) an orthogonal tRNA that suppresses the unique codon and does not crosstalk with its endogenous counterparts; and 4) an orthogonal amino-acyl tRNA synthetase that can specifically charge the ncAA onto the orthogonal tRNA and does not crosstalk with the endogenous amino-acyl tRNA synthetase/tRNA pairs 218 , 222 – 224 . Fig. 8 Scheme of genetic code expansion. Genetic code expansion enables the site-specific incorporation of an noncanonical amino acid (shown in green filled circle) into a growing peptide chain by suppressing an unique codon (e.g., amber stop codon) More than 200 ncAAs with diverse functionalities have been genetically encoded into different organisms to date, such as Escherichia coli , yeast, mammalian cells, viruses, and even animals, providing an invaluable toolbox for protein studies and engineering 225 – 232 . This expanded set of building blocks, including bioorthogonal chemical conjugation partners, metal chelators, photo-crosslinkers, proximity-enabled crosslinkers, photocaged amino acids, amino acids with post-translational modifications (phosphorylation, sulfation, acylation, etc.), redox-active amino acids, and infra-red, nuclear magnetic resonance, fluorescent probes, has been widely used in the study, manipulation, and evolution of proteins 233 – 242 . The ability to genetically encode diverse ncAAs allows for the rational optimization and production of chemically modified recombinant proteins with defined structures, functions, and stoichiometries 243 , 244 . Here, we focus on the application of genetic code expansion in the evolution of therapeutic peptides and proteins.

Small molecule covalent drugs have many advantages compared with noncovalent drugs, such as increased biochemical efficiency and potency, improved pharmacokinetics, prolonged duration of action, reduced dosage and dosing frequency, and potent inhibition of intractable targets 264 . Safety concerns about their low selectivity and the potential immunogenicity of covalent drug-protein adducts mean that the development of small molecule covalent drugs has been intentionally avoided 265 . However, the development of activity-based protein profiling and other recent technologies mean that small molecule covalent drugs have regained attention, and several small molecule drugs that act by a covalent binding mechanism have been approved for marketing 266 . Theoretically, covalent protein drugs should offer similar advantages to small molecule drugs. However, due to their inherent inability to form covalent bonds of natural proteins, the therapeutic potential of covalent protein drugs has not been fully explored. Li et al. 267 recently reported on a proximity-enabled reactive therapeutics (PERx) strategy to develop covalent protein drugs. They genetically incorporated the latent bioactive ncAA, fluorosulfate-L-tyrosine (FSY) 268 , into human programmed cell death protein 1 (PD-1) at position 129 and showed that the resulting PD-1(FSY) formed covalent bonds selectively with its natural ligand, PD-L1, in vitro and in vivo. Strikingly, PD-1(FSY) significantly enhanced the bioactivities of human naïve T cells and engineered chimeric antigen receptor T cells, compared with wild-type PD-1. PD-1(FSY) showed more potent inhibition of tumor growth and had equivalent or greater anti-tumor effects than a therapeutic anti-PD-L1 antibody in several immune-humanized mouse models. They then applied PERx to the covalent inhibition of the HER2 receptor by a FSY-modified affibody, illustrating that PERx could provide a general platform for developing covalent protein drugs. Compared with noncovalent protein drugs, PERx drugs can be used in their original form and do not require additional modifications to extend their half-life, because the covalent binding decouples the drug efficacy from its pharmacokinetics. Moreover, PERx allows small-protein biologics such as PD-1 (15.6 kDa) to be used as therapeutics, thus greatly expanding the scope of therapeutic proteins. In addition, PERx can minimize the off-target effect due to the inherent affinity between the protein drug and its target, as well as the proximity-driven crosslinking mechanism of the latent bioactive ncAA. These advantages mean that the PERx strategy has the potential to provide a general platform to develop novel covalent protein drugs. The chemistry behind the PERx strategy and more examples of covalent proteins have been reviewed in detail elsewhere 269 . Lipid and larger proteins are frequently conjugated to improve the pharmacokinetics of covalent peptide drugs 270 – 272 . Approved peptide drugs, such as liraglutide, semaglutide, and insulin degludec, were conjugated with C 14/16/18 fatty acids, which increased their plasma circulation times and reduced their degradation during kidney elimination 270 . Two plasma proteins, serum albumin and immunoglobulin, are also used to prolong the peptide-circulation times by increasing their molecular weight, thereby exceeding the molecular weight cut-off for glomerular filtration. For example, this strategy was used to extend the half-life of dulaglutide and albiglutide, administered by once-weekly injections 273 , 274 .

Therapeutic peptides in the treatment of diabetes mellitus T2DM is caused by an acquired insulin deficiency and is common in middle-aged and older people. T2DM has been successfully treated with peptide drugs, including GLP-1 receptor agonists (GLP-1RAs) and the best-known peptide drug, insulin. GLP-1 is an endogenous growth hormone secreted by L-cells in the ileum. Its receptors are present in pancreatic β-cells, the peripheral and central nervous systems, heart and blood vessels, kidneys, lungs, and gastrointestinal mucosa (Fig. 11 ). GLP-1 interacts with its receptor to stimulate islet β-cells to secrete insulin, inhibit the release of glucagon by islet α-cells, increase satiety, and delay gastric emptying in a glucose-dependent manner 284 . Endogenous GLP-1 is degraded by dipeptidyl peptidase-4 (DPP-4) and is rapidly inactivated. In order to prolong the stimulation time of GLP-1 receptors, synthetic GLP-1RAs are required to prevent its degradation. Since the first GLP-1RA, exenatide, was approved by the US Food and Drug Administration (FDA) in 2005, several GLP-1RAs have entered the clinic, including liraglutide (2009), lixisenatide (2013), dulaglutide (2014), and semaglutide (2017) 285 . After injection, these GLP-1RAs effectively reduce glycosylated hemoglobin and average blood glucose levels and improve fasting blood glucose 286 . Fig. 11 Mechanisms of GLP-1 and GLP-1RA peptide drugs in regulation of T2DM. GLP-1 and GLP-1RA peptide drugs treat T2DM by regulating multiple organs functions, such as reducing gastric emptying and gastric acid secretion, reducing appetite, promoting cardiac glucose utilization, accelerating renal natriuresis and diuresis, minimizing glucose production in the liver and increasing insulin secretion in the pancreas Some GLP-1RAs are also effective or the treatment of some complications of T2DM. Diabetic nephropathy is one of the most dangerous complications of T2DM, leading to severe effects on kidney function in diabetic patients, with clinical manifestations including proteinuria and decreased glomerular filtration rate (GFR). In a study of 35 patients with T2DM, lixisenatide reduced the absolute and partial excretion of magnesium, calcium, and phosphate by inhibiting the proximal tubule sodium-hydrogen antiporter 3 (NHE3) and thus increasing the absolute and partial excretion of sodium, chlorine, and potassium and increasing urine pH values compared with insulin glargine 287 . In addition, in a study of 30 T2DM patients, liraglutide significantly reduced GFR, urinary albumin excretion rate, and partial albumin excretion 288 . GLP-1RAs can reduce GFR by increasing sodium efflux to the macula densa, increasing tubulo-glomerular feedback and vasoconstriction of afferent arterioles, and may also reduce albuminuria by reducing plasma renin activity, reducing renal oxidative stress, and increasing natriuresis 289 . However, the extent to which these effects are mediated by GLP-1R remains to be determined. Recent studies have confirmed that the metabolites of GLP-1 retain important antioxidant and anti-apoptotic activities, which are independent of GLP-1R 290 . Cardiovascular disease remains the leading cause of death in patients with T2DM, and the prevention and treatment of cardiovascular complications should thus be considered when choosing treatments for T2DM. GLP-1RAs have been shown to play a beneficial role in cardiovascular disease. Recent clinical trials found that only liraglutide and semaglutide had advantages in terms of cardiovascular benefits, although the mechanism is still unclear and may be anti-atherosclerotic 48 . The protective effects of other GLP-1RAs on cardiovascular disease are not obvious, but they have no harmful effects on other safety parameters, and the risk-benefit distribution of GLP-1RAs is thus well-balanced 291 . GLP-1RAs also showed therapeutic effects on obesity symptoms in patients with T2DM. Thondam reported that a patient with severe hypothalamic obesity and various obesity-related complications, including T2DM, responded well to exenatide, with significantly improvements in weight and blood glucose control, possibly through a central regulatory mechanism increasing satiety and reducing energy intake 292 . A study of 25 obese patients with T2DM showed that patients treated with metformin and sulfonylurea/DPP-4 inhibitors for 6 months who took GLP-1RA (exenatide19, six cases) had significantly reduced average body weight, glycosylated hemoglobin, and intrahepatic lipids 293 . Body mass index and fat thickness also decreased significantly in 25 patients with T2DM treated with exenatide and liraglutide for 3 months 294 . T2DM can lead to bone brittleness and increase the risk of bone-related complications such as fractures and poor fracture healing. Experimental studies found that GLP-1RAs had a significant positive effect on bone quality and strength, possibly by improving the blood supply to the bone necessary for bone health 295 . In one s

Therapeutic peptides in the treatment of intestinal disease In the human body, gastrointestinal flora constitutes a complex micro-ecosystem. Typically, the gastrointestinal flora in the human body constitutes a complex micro-ecosystem. Typically, the epithelium regulates the composition of the intestinal flora at the intestinal mucosal interface by providing a physical barrier and secreting various antimicrobial factors, including antimicrobial peptides (AMPs). The dominant flora (physiological flora) and weak flora (pathogenic bacteria) maintain a dynamic balance 315 , which is disrupted in various intestinal diseases caused by exotic bacteria, viruses, and parasites, food poisoning, adverse drug reactions, and genetic factors, such as enteritis, constipation, intestinal ulcers, and inflammatory bowel disease (IBD). The extensive use of antibiotics may further reduce the biodiversity of symbiotic bacteria, which is not conducive to treatment and may even aggravate the disease; for example, individuals affected by IBD are more likely to have used antibiotics within 2-5 years before diagnosis 316 . Peptide drugs have attracted much attention in this field because of their specificity, efficacy, and low toxicity. Significant changes in the normal intestinal flora and the destruction of host-microbial symbiosis may be the key to the development of IBD 317 . IBD, including Crohn’s disease and ulcerative colitis, is caused by an intestinal immune response, and the associated inflammation is caused by the interaction between environmental and genetic factors 318 . However, the specific pathogenesis of IBD is still unclear and there is currently no effective cure. Intestinal microbial diversity is significantly reduced in patients with IBD 319 , and the two dominant phyla Firmicutes (Lachnospiraceae) and Bacteroidetes, were significantly decreased while the phylum Proteus was significantly increased 320 . Substantial evidence has indicated a key role for members of the phylum Proteus in IBD 321 . Proline-arginine-39, a small cationic AMP that is naturally secreted by porcine bone marrow and lymphoid tissue, has demonstrated antibacterial, immunomodulatory, and intestinal epithelial repair functions and may provide a safe alternative therapy for IBD 322 . Patients with Crohn’s disease are often treated by bowel resection 323 , leading to short bowel syndrome (SBS). Damage to the small intestine and abnormal shortness of the small intestine at birth may also cause SBS, which is defined as symptoms associated with a persistent length of the residual small intestine of <200 cm 324 . GLP-2 is produced by intestinal endocrine L cells and various neurons in the central nervous system (Fig. 13 ) 325 and has recently received extensive attention for the treatment of SBS. GLP-2 has demonstrated various beneficial effects, including stimulating crypt cell growth, reducing intestinal cell apoptosis, promoting intestinal mucosal dilatation, inhibiting gastric acid secretion and gastric emptying, stimulating intestinal blood flow, strengthening intestinal barrier function, reducing anti-inflammatory injury, and promoting nutrition and liquid absorption 326 – 328 . GLP-2 also regulated the expression of amino acid transporters and directly activated mTORC1 to increase the absorption of amino acids in the intestinal epithelium 326 . Some specific amino acids (including glutamine, glutamate, arginine, glycine, lysine, methionine, and sulfur-containing amino acids) have also been shown to play an important role in maintaining intestinal integrity, including preventing intestinal atrophy, improving intestinal barrier function, and reducing inflammation and apoptosis 329 . Endogenous GLP-2 is easily degraded by DPP-4; however, the GLP-2 analogue teduglutide prolongs the half-life from 7 minutes to about 2-3 hours by substitution of alanine by glycine in the second position of the N-terminal of GLP-2, effectively preventing its degradation by DPP-4 330 . Clinical studies have shown that teduglutide can effectively reduce or eliminate the need for parenteral nutrition and/or intravenous infusion support 331 , while the application of teduglutide in young pigs with distal ileectomy significantly increased the weight per unit weight and protein synthesis of the remnant intestine 332 . Teduglutide was approved by the FDA for clinical use in SBS patients in 2012. Wiśniewski et al. designed a series of GLP-2 analogues, including 2-glycine substitution, 10-norleucine substitution, 11- and/or 16-hydrophobic substitution, many of which were more effective against GLP-2R than natural hormones, showing good receptor selectivity and low systematic clearance. Among these, the peptide ([ 2 Gly, 10 Nle, 11 DPhe, 16 Leu] hGLP-2-(1−33)-NH 2 ) was selected as a candidate for clinical development 333 . GLP-1 from the proglucagon family has similar functions to GLP-2 and has been suggested for the treatment of SBS. In one study, five patients with SBS

Traditional cancer treatments include surgery and radiotherapy, which have limited effects in patients with advanced cancer. The subsequent development of targeted therapy and immunotherapy have significantly improved the survival rate of cancer patients. Targeted therapy takes advantage of the reliance of tumor cells on specific molecules or signaling pathways to kill tumor cells using a “guided missile” approach 375 . Immunotherapy drugs do not attack tumor cells directly but modulate the patient’s own immune system and attack tumor cells by targeting immune checkpoints 376 . PD-1/PD-L1 is a well-known immune checkpoint, and five monoclonal antibodies against PD-1/PD-L1 interaction have been approved by the FDA for cancer treatment. However, antibodies have disadvantages including high cost, poor oral suitability, and high immunogenicity. Peptides have also attracted attention in the field of tumor diagnosis and treatment because of their small size, high affinity, easy modification, and low immunogenicity. Some modified peptides have also demonstrated good stability. For example, Carvajal et al. developed stable α-helical peptides as inhibitors of MDM2 and MDMX for the treatment of p53-dependent cancer 377 – 379 . The short half-life of natural peptides in vivo means that peptides targeting various abnormally expressed receptors in tumor cells are usually modified peptide analogues. There are three main methods for the production of these peptides, each with its own strengths and weaknesses: 1) derivation from natural proteins; 2) chemical synthesis and reasonable engineering based on structure; and 3) screening of peptide libraries 380 . Among these, phage-display technology is a traditional and widely used method, with the advantages of simple operation and effective screening of a large number of different peptides 381 . Peptides can be applied in tumor therapy in four main ways: 1) using radioisotopes, dyes, or other reported molecular-labeled peptides as probes for tumor diagnosis and imaging; 2) using peptide-coupled nanomaterials for tumor therapy; 3) using peptide vaccines to activate the immune system for prevention; and 4) using peptides alone as targeted drugs (Fig. 14 ) 382 – 384 . Fig. 14 Application of peptides in tumor therapy. a Screening and identification of peptide candidates from chemically synthesized peptide library and phage library. b Using radiolabeled, dye-labeled, or other designed peptides as probes for tumor diagnosis and imaging. c Application of peptide-conjugated nanomaterials in tumor therapy. d Using peptide vaccine and targeting peptides in tumor immunotherapy and targeted therapy Peptide-based imaging probes bind to receptors specifically expressed in the tumor. These receptors can either be expressed on the cell surface, such as αvβ3 integrin (RGD peptide), EGF receptor, somatostatin receptor, neurotensin receptor, and transferrin receptor; intracellularly, such as Bcr/Abl, cyclin A, and cyclin kinase; or in extracellular matrices, such as fibronectin, matrix metalloproteinases, and prostate-specific antigen 382 . The locations of probes can be visualized by single-photon emission computed tomography/computed tomography scanning to indicate the tumor distribution. This technique has been applied for early tumor diagnosis and surgical resection. Several probes have been developed, such as octreoscan and depreotide, as radiolabeled conjugates of somatostatin-like peptides, have been approved by the FDA for imaging of tumors, such as neuroendocrine tumors and lung cancer 385 , 386 . Unfortunately, depreotide has been withdrawn. Radiolabeled peptides based on RGD peptides have recently received attention, and a series of probes have been synthesized, including (99m) Tc-3PRGD2, which can be used to detect differentiated thyroid cancer by negative whole-body scan of radioactive iodine 387 . Lutetium 177 dotatate is a radiolabeled somatostatin analog recently approved for the treatment of somatostatin receptor-positive gastroenteropancreatic neuroendocrine tumors 388 . It binds to somatostatin receptors and then releases radioactive Lutetium 177 into the tumor cells, which induces cellular damage via the formation of intracellular free radicals 389 , 390 . Based on the same principle as peptide-based probes, internal radiotherapy can be achieved by labeling β-emitters on peptides; however, this is notably limited by radiation damage to normal target tissues with positive receptors, either near or far from the tumor 384 . A possible alternative to reduce the potential side effects involves delivering the peptides to tumors coupled to anti-cancer drugs, genes, and RNAs (small interfering RNA ([siRNA]/miRNA/mRNA) 391 . AN-152 and AN-207 are luteinizing hormone-releasing hormone analogues coupled to adriamycin, with anticancer activity against luteinizing hormone-releasing hormone receptor-positive cancers. The results of phase I and II clinical studies showed that the drug was

Peptides have become a unique class of therapeutic agents in recent years as a result of their distinct biochemical characteristics and therapeutic potential. Although peptides outperform small molecules and large biologics in some aspects, they often suffer from membrane impermeability and poor stability in vivo, due to the intrinsic limitations of amino acids. Extensive research has been carried out in terms of the discovery, production, and optimization of peptide drugs, in order to overcome these drawbacks. The integration of traditional lead peptide discovery methods with novel technologies, such as rational design and phage display, provides a reliable approach for the development of effective and selective lead peptides in a short period of time. The single or combined use of chemical and biological recombination synthetic approaches allows the efficient and reliable production of synthetic peptides on large scales. These peptides can be further modified in a site-specific manner through chemical synthesis or genetic code expansion to enhance their stability and physiological activity. Although the field of therapeutic peptides started with natural hormones, the discovery and development trends have since shifted from simply mimicking natural hormones or peptides derived from nature to the rational design of peptides with desirable biochemical and physiological activities. Major breakthroughs in molecular biology, peptide chemistry and peptide delivery technologies have allowed significant progress in the fields of peptide drug discovery, peptide production, and their therapeutic applications. More than 80 therapeutic peptides have reached the global market to date, and hundreds of peptides are undergoing preclinical studies and clinical development. These peptide drugs have been applied to a wide range of diseases, such as diabetes mellitus, cardiovascular diseases, gastrointestinal diseases, cancer, infectious diseases, and vaccine development. Considering their huge therapeutic potentials, market prospects, and economic values, we expect therapeutic peptides to continue to attract investment and research efforts and to achieve long-term success.