PERSpECTIVES

into new peptide therapeutics continues at

Trends in peptide drug discovery a steady pace, with more than 150 peptides
in clinical development and another
400–600 peptides undergoing preclinical
Markus Muttenthaler   , Glenn F. King   , David J. Adams    studies3,4. Considering the increase in
and Paul F. Alewood    investment and research efforts in the
peptide field, the maturity of peptide
Abstract | Since the introduction of insulin almost a century ago, more than synthesis technology, the success of biologics
80 peptide drugs have reached the market for a wide range of diseases, including and the pressure for the pharmaceutical
diabetes, cancer, osteoporosis, multiple sclerosis, HIV infection and chronic pain. industry to maintain new drug approval
rates, we anticipate that the peptide
In this Perspective, we summarize key trends in peptide drug discovery and
therapeutics market (Fig. 2; Table 1) will
development, covering the early efforts focused on human hormones, elegant continue its momentum and expand over
medicinal chemistry and rational design strategies, peptide drugs derived from the coming years.
nature, and major breakthroughs in molecular biology and peptide chemistry that In this Perspective, we first outline how
continue to advance the field. We emphasize lessons from earlier approaches peptide drug discovery and development
that are still relevant today as well as emerging strategies such as integrated strategies have advanced, analysing the
chemical, structural and mechanistic
venomics and peptide-​display libraries that create new avenues for peptide drug features of successful peptide drugs, as well
discovery. We also discuss the pharmaceutical landscape in which peptide drugs as the key approaches for transforming
could be particularly valuable and analyse the challenges that need to be peptide leads into drugs. The sections on
addressed for them to reach their full potential. trends below are presented approximately
historically (although there are overlaps),
from the era of peptide hormones through
The field of peptide therapeutics started in of ~20 orally available new drugs per year to the latest peptide-​display technologies
1922 with the first medical use of insulin — was the norm. Moreover, large-​scale peptide and chemical modifications, with each
extracted from animal pancreases — which manufacturing was considered prohibitively section progressing from the origins through
revolutionized the treatment of type 1 expensive, such that only peptide hormone to the current status and opportunities for
diabetes (Fig. 1). Four decades passed before agonists that were effective at low doses were further advances. All approved peptide
synthetically produced peptide hormones, commercially viable. This resulted in limited drugs and diagnostic agents are listed in
namely oxytocin and vasopressin, entered interest in peptides from the pharmaceutical Supplementary Tables 1–5, which include
the clinic. Industrial groups such as those industry and stagnation of peptide drug further references. We also highlight
of Robert Schwyzer at Ciba and Charles development. selected examples of serendipity in Box 1.
Huguenin at Sandoz entered the field and Nevertheless, the use of peptides as We focus mainly on peptide therapeutics
increased commercial interest in peptides subtype-​selective probes for receptor in the range of 4–100 amino acid residues
as therapeutics. However, at that time, studies and their continued presence as and exclude shorter peptidomimetics such
synthesis of peptides by solution-​phase lead compounds provided the basis for a as angiotensin-​converting enzyme (ACE)
chemistry required months to years second wave of peptide drug development inhibitors (see refs5–9 for reviews), which
of work, and it took the invention of beginning in the late 1980s, this time are often similar to typical small-​molecule
solid-​phase peptide synthesis (SPPS) backed by venture funds and biotechnology drugs, as well as bacterium-​derived
in 1963 (ref.1), in combination with the companies. The commercial success of macrocyclic peptide antibiotics such as
development of purification methods such human insulin produced using recombinant vancomycin and daptomycin, for which
as high-​performance liquid chromatography, technology (approved in 1982) as well as the there are other discovery trends and
to attract significant attention from the synthetic gonadotropin-​releasing hormones development challenges (see refs10,11 for
pharmaceutical industry. leuprolide and goserelin (approved in reviews). We conclude this Perspective
Soon, the importance of peptides as 1985 and 1989, respectively) confirmed by discussing overarching issues and
key biological mediators, along with their the viability of the peptide drug market, opportunities for future peptide drugs.
remarkable potency, selectivity and low and advances in drug delivery technology,
toxicity, was established. At the same time, formulation and synthesis encouraged The era of peptide hormones
however, their limitations, including low increased investment and research. Peptide research at the beginning of the
oral bioavailability, low plasma stability and Consequently, the number of peptides twentieth century focused largely on
short circulation time, were recognized. entering clinical trials from 2000 to 2010 the role of human signalling hormones.
These developments occurred during the was nearly twice that in the 1990s2. Insulin is the classic example of the use of
golden age (1970 to 1980s) of small- At present, there are around 80 peptide an endogenous hormone as a therapeutic;
molecule pharmaceuticals, when approval drugs on the global market, and research it was the first peptide drug to be used

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 309
Perspectives

clinically, and it remains by far the most were used to increase their stability as More systematic truncations, as
commercially successful, with the market for well as improve other properties, such as well as scans to assess the effects of
insulin therapies showing a 10% compound potency, selectivity, pharmacokinetics and replacing particular residues with alanine,
annual growth rate owing to the rise in the pharmacodynamics13–15. Some of the earliest N-​methylated versions or d-​amino
prevalence of obesity12. From a chemical attempts were established with oxytocin acids, were performed with the peptide
point of view, however, it was short peptides and vasopressin16, for which chemical hormone somatostatin, which comprises
such as oxytocin, vasopressin, somatostatin modifications such as d-​amino acids17,18, 14 residues and a single disulfide bond32–35.
and gonadotropin-​releasing hormone unnatural amino acids19, amino-​terminal Somatostatin, which inhibits the secretion
(GnRH) that initiated the field of peptide (N-​terminal) capping, deamination20–22, of growth hormone, insulin, glucagon,
drug development, and many analogues extensions of N termini or carboxy gastrin, secretin and thyroid-​stimulating
of these hormones are still in use today termini (C termini)23–25 and disulfide bond hormone, was of little therapeutic value
(Supplementary Table 1). mimetics26–29 were explored to increase the itself due to its short in vivo half-​life of
It became clear early on that the stability, potency and selectivity of these less than 3 min. These studies identified
short metabolic half-​lives (minutes) endogenous ligands, eventually resulting its β-​turn pharmacophore, which became
of endogenous human hormones are in successful drugs such as desmopressin, the lead sequence in the quest for more
problematic for therapeutic development, terlipressin, carbetocin and atosiban30,31 potent analogues with increased metabolic
and various medicinal chemistry approaches (Supplementary Table 6). stability35,36 (Fig. 3a). This led to the

Venomics

New chemical
approaches

Flexizyme technology

Overcoming renal clearance

1922 1954 1963 Phage display

First medical Oxytocin Merrifield
use of insulin synthesis SPPS
Recombinant technology

1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020

1950 1978 1980 1996 2002 2010

Corticotropin Calcitonin Oxytocin Glatiramer Teriparatide Liraglutide

1978 1982 Insulin 1998 2003 2012

Desmopressin (recombinant) Eptifibatide Enfuvirtide Linaclotide

1983 1999 2004 2014

Cyclosporine Ganirelix Ziconotide Dulaglutide

1985 2005 Insulin 2017

Leuprolide detemir Semaglutide

Human hormones 1988 2005 2017

Nature derived Octreotide Exenatide Etelcalcetide
Medicinal chemistry/rational design 2008 2018
Serendipitous discovery Romiplostim Lutetium Lu 177
DOTA-TATE
Overcoming renal clearance
2009
Display technologies Ecallantide

Fig. 1 | Historical timeline of key milestones, developments and drug increase the molecular weight of peptides through conjugation to lipids,
approvals in the peptide therapeutics field. The first peptide drug was larger proteins and polyethylene glycol helped overcome the problem of
insulin, extracted from bovine and porcine pancreas. Polypeptides were renal clearance and increased plasma circulation times. Display technolo-
first chemically synthesized in 1954, when Vincent du Vigneaud’s group gies such as phage display now allow target-​oriented discovery of peptides
published the total synthesis of oxytocin and vasopressin (recognized with with more drug-​like properties from vast libraries. Flexizyme technology
the Nobel Prize in Chemistry in 1955)226,227. Another leap forward was Bruce allows the incorporation of non-​proteinogenic amino acids into display
Merrifield’s visionary idea to automate peptide synthesis by assembling libraries. Natural peptide discovery, particularly peptides from venoms,
amino acids on a solid phase, leading to the invention in 1963 of solid-​ and new chemical approaches are also advancing the field. Selected pep-
phase peptide synthesis (SPPS)1 (recognized with the Nobel Prize in tide drugs based on advances in the areas discussed in this Perspective and
Chemistry in 1984). The advent of recombinant technology in the 1980s their initial regulatory approval dates are highlighted in the lower part
enabled clean production of larger peptides. Subsequently, strategies to of the figure.

310 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

approval of octreotide for the treatment of a Global pharmaceutical market (2019) b Peptide drug approvals
acromegaly, adenomas and pancreatic, breast 52
and prostate tumours. N- to C-​terminal 20% Biologics
cyclization was used in the development of
pasireotide (Fig. 3a), contributing to its long 29
75% $1.2 5% Peptides
half-​life of ~12 h, a vast increase compared trillion
with somatostatin.
Small-
The observation that many tumours molecule 5
overexpress at least one of the five drugs
somatostatin receptor subtypes37 not only 1960–1979 1980–1999 2000–2019
drove development of subtype-​selective
analogues such as lanreotide, vapreotide c Distribution of function d Therapeutic indications
and pasireotide (Fig. 3a) but also introduced Antagonists Cardiovascular
peptide scintigraphy and the field of peptide 8% 11% indications
Inhibitors Oncology 17%
receptor radionuclide therapy, where 9% Reproductive
Decoy 8% medicine
radioactive elements such as 111In, 90Y, 68Ga Surfactant
or 99mTc were attached via chelating groups to 6% Gastroenterology
2%
selective somatostatin analogues38. This has 1% 17% 4% Osteology
resulted in approved radiopharmaceutical Metabolic
indications Miscellaneous
agents such as indium In 111 pentetreotide, 80% Agonists 20%
17% (CNS, pain,
technetium Tc 99m depreotide, gallium Ga Endocrinology infections, etc.)
68 DOTA-​TATE, gallium Ga 68 DOTA-​TOC
and copper Cu 64 DOTA-​TATE to detect
tumours. Furthermore, the recent approval Fig. 2 | The peptide drug market. Peptide drugs occupy a distinct pharmaceutical space in between
small-​molecule drugs and biologics. a | They account for 5% of the global pharmaceutical market, with
of lutetium Lu 177 DOTA-​TATE for the
global sales exceeding US$50 billion in 2019. b | Approvals have steadily increased over the last six
treatment of somatostatin receptor-​positive
decades, with an average growth rate of 7.7% for the global peptide therapeutics market228. Insulin
gastroenteropancreatic neuroendocrine and analogues are responsible for ~50% of peptide drug revenue ($25 billion), followed by the
tumours has revitalized interest in the glucagon-​like peptide 1 (GLP1) receptor agonist dulaglutide, marketed as Trulicity for the treatment
application of peptides in radionuclide of diabetes ($4.4 billion), the GLP1 receptor agonist liraglutide, marketed as Victoza and Saxenda for
therapies39–41 (Supplementary Table 2). the treatment of diabetes and obesity, respectively ($4.1 billion), and the synthetic gonadotropin-
The strategy of developing superagonists releasing hormone analogue leuprolide, marketed as Lupron and Eligard for the treatment of cancer
to desensitize and downregulate receptors to ($2 billion) (Table 1). c,d | Of the currently approved peptide therapeutics, most are agonists (part c),
produce a therapeutic response similar and the most commonly targeted indications are related to endocrinology, metabolism and oncology
to treatment with an antagonist was (part d). CNS, central nervous system. Sources: 2019 company financial reports and global sales
analysis reports.
first successfully translated in the drug
development programmes based on
GnRH. Blockade of GnRH hormonal poly(d,l-​lactic-​co-​glycolic acid)/poly(lactic not ideal as it often induced allergic reactions.
function is applied clinically in cancer acid) microparticle depot formulations able Therefore, selective expression of endogenous
treatment, delay of puberty, management to deliver stable drug concentrations with human peptides and proteins in cell culture
of oestrogen-​dependent female disorders, once-​weekly or less frequent dosing (for systems was highly desirable, and the advent
sex reassignment and in vitro fertilization example, once in 6 months for Lupron Depot of recombinant technology was a milestone
therapy42. Some of these early superagonists for the treatment of prostate cancer)45. There in peptide drug development (Fig. 1).
(Supplementary Table 7) are still very are currently eight peptide drugs on the Somatostatin was the first human peptide to
successful; for example, leuprolide acetate market that use a poly(d,l-​lactic-​co-​glycolic be produced recombinantly47, and in 1982,
and goserelin acetate had annual sales of acid) or poly(lactic acid) extended release insulin, a complex heterodimer comprising
US$2 billion and $813 million, respectively, delivery system (Supplementary Table 5)46. a 21-​residue A chain, a 30-​residue B chain
in 2019 (based on company financial Further details on individual and three disulfide bonds (Fig. 3b), was the
reports). The development of GnRH drug candidate development of the first licensed drug to be produced by this
antagonists that directly arrest GnRH action aforementioned drugs and additional technology (developed by Genentech and
followed soon after, and advances in SPPS supporting examples, such as pramlintide, Eli Lilly)48. Advances in genetic engineering
allowed more extensive use of unnatural an analogue of amylin (Supplementary Fig. 1), soon permitted single amino acid alterations
amino acids in antagonist design, ultimately icatibant, a bradykinin receptor antagonist to modulate the absorption, distribution,
producing successful drugs such as cetrorelix (Supplementary Table 8), and afamelanotide, metabolism and excretion characteristics of
and ganirelix43 (Supplementary Table 7). bremelanotide and setmelanotide, analogues peptides49,50. This was particularly applicable
Different peptide delivery systems of α-​melanocyte-​stimulating hormone, can to insulin, as it naturally forms inactive
were subsequently developed, including be found in Supplementary information. hexamers during long-​term storage in the
slow-​release subcutaneous or intramuscular body, which is therapeutically undesirable.
injections, as well as intranasal delivery. The advent of recombinant technology Genetic engineering allowed development of
The most common delivery form used was Despite the success of early hormone fast-​acting and slow-​release insulin analogues
that of hydrophobic depots, which increased analogues, production of longer peptides was for different treatment options51 (Fig. 3b),
the duration of action44. This eventually limited by the synthetic methods available, which now constitute a multibillion-​dollar
led to the approval of biodegradable and insulin derived from pigs or cows was market (Fig. 2).

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 311
Perspectives

Table 1 | Top peptide drugs by sales

Generic name Most common brand names Companies Indication First Sales in 2019
approval (US$ million)
Insulin and Ademelog, Apidra, Humulin, Humalog, Novo Nordisk, Eli Lilly, Sanofi Diabetes 1982 25,000
analogues Insuman, NovoMix, NovoRapid, Lantus,
Levemir, Ryzodeg, Tresiba, Toujeo
Dulaglutide Trulicity Eli Lilly, Dainippon Sumitomo Diabetes 2014 4,394
Liraglutide Victoza, Saxenda Novo Nordisk Diabetes, obesity 2010 4,142
Leuprolide Lupron, Eligard AbbVie, Astellas, Takeda Cancer 1985 2,022
Semaglutide Ozempic, Rybelsus Novo Nordisk Diabetes, obesity 2017 , 2019 1,694
Octreotide Sandostatin Novartis Cancer 1988 1,585
Glatiramer Copaxone, Glatopa Teva, Sandoz MS 1996 1,531
Teriparatide Forteo Eli Lilly Osteoporosis 2002 1,405
Cyclosporine Restasis Allergan Immune diseases, 1983 1,189
organ transplants
Lanreotide Somatuline Ipsen Acromegaly 2007 1,124
Carfilzomib Kyprolis Amgen Multiple myeloma 2012 1,044
ACTH Acthar Mallinckrodt IS, MS 1950 953
Linaclotide Linzess, Constella Allergan, Astellas Pharma IBS-​C 2012 877
Romiplostim Nplate, Romiplate Amgen, Kyowa Kirin Chronic ITP 2008 841
Goserelin Zoladex AstraZeneca Cancer 1989 813
Etelcalcetide Parsabiv Amgen, Ono Pharma Hyperparathyroidism 2017 693
Exenatide Byetta, Bydureon AstraZeneca Diabetes 2005 659
Teduglutide Gattex, Revestive Takeda Short bowel syndrome 2012 555
Vasopressin Vasostrict Endo Pharma CDI NR 532
Sources: 2019 company financial reports, https://www.fda.gov, and global sales analysis reports. ACTH, adrenocorticotropic hormone; CDI, central diabetes
insipidus; IBS-​C, irritable bowel syndrome with constipation; IS, infantile spasms; ITP, immune thrombocytopenia; MS, multiple sclerosis; NR, not recorded.

Two further key examples of peptide teriparatide, a recombinant form of the first In recent years, there has been an
hormones for which recombinant 34 residues of human PTH (the bioactive increased focus on technologies for
approaches have been harnessed are portion of the hormone) that was approved expanding the genetic code to produce
calcitonin and parathyroid hormone for the treatment of osteoporosis in 2002. peptides and proteins containing non-
(PTH). Calcitonin, a 32-​residue peptide Teriparatide was the first agent that promotes canonical amino acids58–60. Furthermore, the
hormone, was discovered in 1961 as a bone formation (rather than inhibiting bone combination of genetic code expansion with
substance that lowers blood calcium levels resorption) to be approved for the treatment display technologies (discussed further later)
in dogs52–54. It was subsequently shown of osteoporosis. More recently, a recombinant enables incorporation of Nα-​methylated
to potently inhibit bone resorption, and full-​length PTH therapy to control low amino acids61,62, d-​amino acids63–65, β-​amino
salmon calcitonin, which differs from human blood calcium concentration in patients acids66,67 and thioamides68 into peptides
calcitonin by 16 residues and is 40–50-​fold with hypoparathyroidism56 was approved in and the creation of peptoids69 and various
more potent, was approved for the treatment 2015, and abaloparatide, a synthetic analogue macrocycles that can be screened in
of hypercalcaemia and postmenopausal of PTH-​related protein (PTHrP) with 41% large display libraries against therapeutic
osteoporosis in 1978. Salmon calcitonin identity to teriparatide57, was approved for targets70–72. Such technologies are likely to
was initially extracted from the thyroid-​like the treatment of osteoporosis in 2017. be increasingly used for the discovery and
glands of salmon, but a synthetic version Recombinant technology also provided production of larger and more complex
was approved in 1978 and a recombinant a reliable and facile alternative to synthetic peptide drugs.
version was approved in 2005. PTH, an peptide production, which further
84-​residue hormone discovered in 1925 accelerated peptide research and led to Peptide drugs based on natural products
(ref.55), opposes the effects of calcitonin. drug approvals for recombinantly produced Over the years, increasing numbers of
It acts on PTH 1 and PTH 2 receptors and glucagon, carperitide, lepirudin, nesiritide, peptides from bacteria, fungi, plants and
increases blood calcium levels largely by mecasermin, desirudin and romiplostim, animals have been characterized that often
enhancing the release of calcium from the with the majority being used for the have better therapeutic properties than
reservoir in bones. Prolonged elevation of treatment of cardiovascular conditions and their human counterparts. These include
PTH levels will deplete bone stores, but haematological disorders (Supplementary higher selectivity, potency and in vivo
intermittent exposure activates osteoblasts Table 1). Details of another example of a stability, which are especially important
(the cells responsible for creating bone) recombinant peptide drug, teduglutide, for therapeutic development. Since many
rather than osteoclasts. The first therapeutic a glucagon-​like peptide 2 (GLP2) analogue, receptors in the animal kingdom are very
based on PTH to reach the market was are provided in Supplementary information. similar to their human counterparts, these

312 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

vast natural sources provide novel bioactive that featured higher selectivity, potency the treatment of type 2 diabetes in 2005.
peptides that can be mined for therapeutic and in vivo stability due to a more rigid GLP1 belongs to a group of gastrointestinal
development. Supplementary Table 3 lists secondary structure78–85. Evolutionarily hormones called ‘incretins’ that cause an
approved peptide therapeutics derived from optimized venoms have become particularly increase in insulin secretion after food
natural products. Two early key examples are attractive for such discovery efforts, consumption, even before blood glucose
the immunosuppressive drug cyclosporine providing millions of unique and highly levels become elevated91. Its biologically
isolated from fungi and the thrombin potent bioactive venom peptides that target active form GLP1(7–37) also inhibits
inhibitor bivalirudin isolated from saliva a wide range of proteins, including ion glucagon secretion, appetite and food intake.
of the medicinal leech. channels, enzymes, G-​protein-​coupled Considerable efforts were made to develop
Cyclosporines were discovered in 1970 receptors (GPCRs) and transporters86–90 GLP1(7–37) into a drug after it was found
at Sandoz during a project that attempted to (see Box 2). Two examples of approved that intravenous infusion has dramatic
identify new antifungal agents73,74. Although venom-​derived peptide drugs are exenatide, effects on insulin secretion and blood
the crude extracts of two new strains of a GLP1 receptor agonist from the venom of glucose in patients with type 2 diabetes92.
fungi imperfecti had only low antifungal the Gila monster (Heloderma suspectum), However, GLP1(7–37) is rapidly broken
activity, their toxicity was unusually a venomous lizard, and ziconotide, a down by dipeptidyl peptidase 4 (DPP4)
low, which prompted further screening. peptidic inhibitor of voltage-​gated calcium and undergoes renal clearance within
Observation of immunosuppressive effects (CaV) channels from the venom of a 1–2 min (ref.91). Although replacement of
ultimately led to the isolation and structural predatory marine cone snail. Both drugs Ala at position 2 with other short-​side-​chain
characterization of cyclosporine, a neutral, are identical to the native venom peptide. amino acids protected the molecule against
hydrophobic, cyclic 11-​residue peptide that Exenatide became the first of a highly enzymatic degradation, fast renal clearance
has some peculiar features. Its high degree successful class of drugs that act as agonists still rendered it unsuitable for therapy91.
of N-​methylation and its cyclic structure of the GLP1 receptor with its approval for In the early 1990s, a 39-​residue peptide
make it resistant to proteolytic degradation,
and its hydrophobicity and conformational
flexibility due to its intramolecular Box 1 | Peptide drugs based on serendipitous discoveries
hydrogen bonding network render it orally While medicinal chemistry and structure–activity relationship studies were the core concepts in
bioavailable75 (Fig. 4a). Cyclosporine was early peptide drug development, it was serendipity coupled with outstanding cross-​disciplinary
approved as an immunosuppressive drug in science that played a major role in development of the peptide drug blockbuster glatiramer, which
1983, and it remains an inspiring example introduced the important concept of decoys for immunomodulators. This concept also found an
of the value of natural product discovery as important application in development of the HIV-​entry inhibitor enfuvirtide. The discovery of both
well as a reminder that peptides and peptide of these peptide drugs is described here.
Glatiramer (formerly known as copolymer 1 or Cop-1) is not a single peptide but is a synthetic
mimetics can be orally bioavailable.
polymer of four amino acids, Ala, Lys, Glu and Tyr, in a molar ratio of 4.2, 3.4, 1.1 and 1.0, respec-
Medicinal leeches were a focus of tively229,230. Given that the sequences can differ from one peptide to another, the composition and
research in the late eighteenth century due to molar ratios are fixed. The copolymers range from 40 to 90 residues in length and the molecular
the anticoagulant properties of their saliva, mass ranges from 4.7 to 10 kDa (ref.230). Glatiramer was initially synthesized to resemble the struc-
but it took nearly a century until hirudin, ture of myelin basic protein. The purpose was to study the interaction of myelin proteins with
the peptide responsible for this action, was lipids capable of inducing experimental autoimmune encephalomyelitis, which has many parallels
isolated and its structure was determined76. to the human disease multiple sclerosis230. However, rather than inducing the disease, it protected
Hirudin, a 65-​residue peptide with a against acute and chronic relapsing experimental autoimmune encephalomyelitis230. This discovery
compact N-​terminal domain stabilized prompted further study and eventually led to its market approval in 1996 for the treatment of
by three disulfide bonds and a flexible multiple sclerosis231. The exact mechanism is still not completely understood, but its resemblance
to myelin basic protein suggests that it acts by inducing a decoy response, diverting the auto­
C-​terminal domain, is a potent thrombin
immune response away from myelin232,233. Although sales declined from US$3.8 billion in 2017 to
inhibitor. A recombinant form, lepirudin, $1.5 billion in 2019 due to the approval of generic versions, glatiramer is still one of the most
with two modifications (replacement of successful peptide drugs.
Leu with Ile at position 1 and removal of Tyr Another example of serendipity in peptide drug discovery that had important implications for
sulfonation at position 63) was approved in the field was the development of enfuvirtide for the treatment of HIV infection. The discovery
1998 as an anticoagulant. It was followed of the HIV life cycle in the 1980s led to the identification of a variety of HIV treatment targets234.
in 2002 by the first ‘hirulogue’, bivalirudin, A set of peptides was derived from HIV glycoprotein gp41 for an epitope mapping experiment to
a 20-​residue fragment of hirudin that find an effective HIV vaccine. However, these peptides turned out to exhibit antiviral effects when
contains the active site of thrombin inhibitor incubated with human T cells. Additional studies of the fusion process and understanding of how
d-​Phe-​Pro-​Arg linked via a Pro-(Gly)4 the envelope glycoproteins interact showed that these peptides act as decoys, inhibiting HIV-1
entry into cells. Enfuvirtide (formerly known as T-20 or DP-178) was finally selected out of a pool of
linker to a dodecapeptide analogue of the
similar peptides due to its above-​average antiviral efficacy235. It is a linear 36-​mer that corresponds
C terminus of hirudin77. Thrombin can to residues 643–678 of the α-​helical carboxy-​terminal region of gp160; it consists of only l-​amino
slowly cleave the N-​terminal inhibitor acids with an acetylated amino terminus and a carboxy-​terminal amide235. With a serum half-​life
off at the Arg-​Pro site of the linker, which of 3.8 h, it is administered by subcutaneous injection twice daily. This requirement for frequent
leads, in contrast to hirudin, to beneficial administration resulted in an unprecedented scale of peptide production194,236. Large-​scale
reactivation of thrombin’s active site and production of enfuvirtide involves 106 steps (in comparison to an average of 8–12 steps for a
subsequent haemostasis (the half-​life of small-​molecule drug), and enfuvirtide is to date the most complex synthetic peptide manufactured
bivalirudin is 20–30 min). at such a scale (more than 3 metric tons per year in 2004). Achieving such complexity and scale at
Continuous advances in synthetic good manufacturing practice level involved innovative engineering, method development and
methodology permitted the synthesis optimizations, processes that have been beneficial for improving the capabilities and increasing
the efficiency of large-​scale production of peptide therapeutics.
of larger and more protein-​like peptides

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 313
Perspectives

a Somatostatin drug and diagnostic agent development Fig. 3 | Selected examples of therapies based
Compound Drug name N terminus Sequence Half-life Year
on peptide hormones. a | Somatostatin drug
and diagnostic agent development. Systematic
Somatostatin AGCKNFFWKTFTSC 3 min 1973
structure–activity relationship studies identified
Vale compound CFwKTC ND 1979 somatostatin’s pharmacophore (FWKT, blue) and
Vapreotide Sanvar fCYwKVCW 0.5 h 1987 the shortest active mimetic (Vale compound),
Octreotide Sandostatin fCFwKTCT-ol 2h 1982 which became the lead sequence in the
Indium In 111 pentetreotide Octreoscan 111
In–DTPA– fCFwKTCT-ol 2h 1991 quest for more potent and stable analogues.
Lanreotide Somatuline D-Naph-CYwKVCT 1h 1988 Amino-terminal (N-​terminal) D-​Phe (f) and a
Pasireotide Signifor See structure below 12 h 2012 carboxy-terminal (C-​terminal) amino alcohol
Lutetium Lu 177 DOTA-TATE Lutathera 177
Lu–DOTA– fCYwKTCT 3.5 h 2013 (T-ol) extension were introduced to mask enzy-
Gallium Ga 68 DOTA-TATE Netspot 68
Ga–DOTA– fCYwKTCT 68 min 2016 matic recognition sites, increasing activity and
stability (shown in salmon pink and green, respec-
Gallium Ga 68 DOTA-TOC SomaKit TOC 68
Ga–DOTA– fCYwKTCT-ol 68 min 2019
tively). This led to the approval of octreotide for
Copper Cu 64 DOTA-TATE Detectnet 64
Cu–DOTA– fCYwKTCT 12.7 h 2020
the treatment of acromegaly and various cancers.
H
N Further subtype-​selective analogues, such as
O lanreotide and vapreotide, have been developed
H2N
O H H for oncology applications, including peptides
N with radioactive elements such as 111In, 90Y, 68Ga,
N
O
64
Cu or 99mTc attached via chelating groups to the
O O NH N terminus for peptide scintigraphy, targeted
HN O O NH radiotherapy and positron-emission tomography
O
of tumours. N- to C-terminal cyclization increased
NH
N peptide half-life significantly and led to the
H
NH2 approval of pasireotide for the treatment of
Cushing disease. Cys residues that form a disul­
fide bond are shown in orange. The half-​lives
Pasireotide
listed are in vivo/elimination half-​lives, except
O
for gallium Ga 68 DOTA-​TATE, gallium Ga 68
DOTA-​TOC and copper Cu 64 DOTA-​TATE, for
b Insulin drug development which the radionuclide half-​lives are given.
A chain b | Insulin drug development. A schematic struc-
G S N ture of insulin highlighting the chemical modifi-
I V S C T
E Q E N Y K cations used in drug analogues is shown at the
C C T L P
S I C S L Y Q S
G
T A/TRR/ΔT top. The purple residues and the table highlight
B chain Y P/K/E point modifications in insulin drug analogues,
S A G V F K/D
F V S S F with fast-​acting analogues on a blue background
N Q R G C14/16 fatty acid and long-​acting analogues on an olive back-
H L
C G S H L C G E
V E A L Y L V ground. Green and underlined residues are cru-
K cial for receptor binding, demonstrating why it is
difficult to develop a small-​molecule drug that
Analogue A chain B chain Approval can directly activate the insulin receptor. The first
Porcine T30A 1966 drug to use single amino acid alterations was
Bovine T8A, I10V T30A 1966 fast-​acting insulin lispro, in which the C-​terminal
Lys and Pro residues of the B chain were swapped,
Sheep T8A, S9G, I10V T30A NA
preventing formation of inactive hexamers with-
Human 1982
out affecting activity. Other fast-​acting ana-
Lispro P28K, K29P 1996 logues followed: insulin aspart in 2000 and insulin
Aspart P28D 2000 glulisine in 2004. Long-​acting analogues such as
Glargine N21G +31R, +32R 2000 insulin glargine were produced by adding two
Glulisine N3K, K29E 2004 positively charged Arg residues to the C terminus
Detemir K29K (C14 fatty acid) 2005 of the B chain, which increased the isoelectric
Degludec ΔT30, K29E (C16 fatty acid) 2015 point. This rendered the analogue less soluble at
physiological pH, resulting in slower release from
the injection site. Addition of fatty acids pro-
called ‘exendin 4’ with 53% identity to the treatment of this disease, including duced further long-​acting analogues — insulin
GLP1(7–37) was isolated from venom of liraglutide, albiglutide, dulaglutide, detemir and insulin degludec — where the fatty
the Gila monster. It was found to be a full lixisenatide and semaglutide, with some also acid delays the release from the injection site
agonist of the GLP1 receptor93,94 (Fig. 4b) gaining approval for the treatment of obesity. and reduces renal clearance due to binding to
serum albumin. NA, not applicable; ND, not
with stability against DPP4 degradation Advances in this major peptide drug class
determined.
and lower renal clearance in humans also illustrate other trends that are discussed
(5–7 h)91,93,95, supporting its development as further below.
a drug (named ‘exenatide’ and administered Initial studies on ziconotide date back to disulfide-rich peptides, generally 10–40
twice daily by subcutaneous injection) for the 1980s, when the potential therapeutic residues in length, that target various ion
type 2 diabetes. Several other GLP1 agonists applications of the rigid and protein-​like channels, GPCRs and transporters with
(as well as a longer-​acting formulation of peptides from cone snail venom were first high potency and selectivity. Ziconotide is a
exenatide96) have since been approved for explored90. These conotoxins are small 25-residue peptide derived from Conus magus

314 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

with three disulfide bonds that stabilize and (Supplementary Fig. 2) — are provided in Although the primary application of these
orient the short β-​sheets in a unique three- Supplementary information. Looking to the peptides is treatment of infections, for which
dimensional arrangement97. Its defined future, genome mining of microorganisms104 several candidates are in clinical trials, their
topology and compact structure enable this as well as other organisms105–109 and new immunomodulatory functions are also
conotoxin to selectively inhibit CaV2.2 chan- methods to grow previously uncultured being pursued for the treatment of chronic
nels. Although hundreds of analogues were bacteria or promote their production inflammatory disorders and cancer, as
studied for their pharmacological properties, of novel secondary metabolites, such as adjuvants for vaccine formulations and for
it was the naturally occurring peptide that cultivation in their natural environment wound healing113.
was finally approved by the FDA in 2004 or the use of specific growth factors, will
for the treatment of severe chronic pain. lead to the discovery of novel therapeutic Strategies to overcome renal clearance
Notably, this compound is about 1,000 times peptide leads, as recently exemplified by Unmodified peptides are rapidly cleared
more potent than morphine without the the discovery of the antimicrobial peptide (within minutes) from plasma, and
problems related to addiction. A major teixobactin110–112. Cationic host defence it was realized early in the history of
downside, however, is that its administration peptides are another interesting class of peptide drug development that enzymatic
requires an implanted pump for infusion natural products that encompass not just protection alone was not the solution.
into the cerebrospinal fluid because the tar- direct microbicidal properties but also On the basis of the understanding of
get population of CaV2.2 channels is located interactions with the host immune system the size-​dependent mechanism of renal
in central spinal cord neurons. to fend off infections113. This peptide class clearance114–116, strategies to increase the
There are millions of neurologically is produced by vertebrates, invertebrates, molecular weight of peptides emerged,
active venom peptides remaining to be plants and fungi, mostly as the first line such as lipidation, conjugation to larger
explored, with spiders, scorpions and cone of defence against microbial pathogens. proteins and pegylation, which substantially
snails providing some of the richest chemical
diversity. These venom libraries remain
largely untapped, in part because of the slow a Cyclosporine
pace of the traditional discovery approach O
H
based on bioactivity-​guided fractionation, O N 3 4 N
N
in which crude venom fractions are screened 2
5
O
against known targets, followed by further HN
O N
6
purification and sequencing to elucidate O O
the sequence of a hit. However, various 1
HN
N
technological advances are now forming 7
OH
the basis of a powerful approach termed O O HN O
‘integrated venomics’, which combines 11
N N
collated sequence data from venom 10
N 8
gland or venom duct transcriptomes with O 9 O

proteomic data obtained from crude venom

via liquid chromatography–tandem mass
spectrometry in a rapid and cost-​effective
b
manner98,99 (Fig. 5). With this method,
it is possible to obtain thousands of novel
sequences that can be produced by SPPS
DPP4
or recombinant methods in bacteria, Helical regions
yeast and insect cells100. Whereas most of GLP1 HAEGTFTSDVSSYLEGOAAKEFIAWLVKGRG
these native disulfide-​rich peptides fold
into their bioactive isomer, new synthetic
approaches use selenocysteine-​directed
folding to facilitate synthetic library design Exenatide HGEGTFTSDLSKQMEEEAVRLFIEWLKNGGPSSGAPPPS
and efficient folding of modified peptides
for structure–activity relationship (SAR)
studies101,102. Numerous high-​throughput
screening platforms (such as FLIPR, Cell
Lux, FDSS 6000, AlphaScreen and Tango) Fig. 4 | Selected examples of therapies based on natural product peptides. a | Chemical structure
can be used to screen these peptides against of the immunosuppressant cyclosporine, initially isolated from a fungus. Cyclosporine can be classified
a large panel of validated targets for therapy, as a peptide mimetic, considering that 7 of the 11 peptide bonds are N-​methylated (orange); it con-
with ion channels being a target class for tains two unusual amino acids (3-​hydroxy-4-​methyl-2-​methylamino-6-​octonoic acid at position 1 and
which venom peptides could be particularly aminobutyric acid at position 2, blue) and one D-​amino acid (D-​Ala at position 8, blue). These modifi-
cations translate into outstanding stability and oral bioavailability. b | Sequence alignment and
promising, given that identifying modulators
three-​dimensional structures of human glucagon-​like peptide 1 (GLP1) and the natural product
with very high selectivity is often needed in exenatide (53% sequence identity) isolated from the saliva of the Gila monster (Heloderma suspectum).
ion channel drug discovery98,103 (Box 3). GLP1 is rapidly broken down by dipeptidyl peptidase 4 (DPP4) and undergoes renal clearance within
Further examples of approved peptides 1–2 min. By contrast, exenatide is stable to DPP4 degradation and remains in circulation for 5–7 h. The
derived from natural products — eptifibatide, alteration from G16 in GLP1 to E16 in exendin 4 reduces flexibility and further stabilizes its helical
romidepsin, carfilzomib and linaclotide conformation, which might be one of the reasons for its increased stability.

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 315
Perspectives

increased plasma circulation times117,118. for glomerular filtration, thereby prolonging phase I trials for the treatment of obesity130.
Steric hindrance by these size-​increasing the duration of action. This strategy was Oxyntomodulin is an agonist of both GLP1
moieties also protects against proteolytic used for dulaglutide and albiglutide to and glucagon receptors and is a natural
degradation117,118. extend the half-​life and thereby allow appetite suppressant.
Lipid conjugation moieties that have been once-​weekly injections117,123,124. Dulaglutide
used in approved peptide therapies include consists of two identical, truncated and Display technologies
a C14/16/18 fatty acid in insulin detemir, modified GLP1 analogues, each covalently Display systems — which include phage dis-
insulin degludec (Fig. 3b), liraglutide and linked by a flexible peptide to a modified play, yeast display, mRNA display, ribosome
semaglutide, resulting in longer-​acting human immunoglobulin G4 heavy chain display and DNA display — play an impor-
analogues that allow administration fragment (Fc), which are joined through tant role in today’s state-​of-​the-​art peptide
once daily, or once weekly in the case of two disulfide bonds125. Albiglutide consists drug discovery (Fig. 5). They create a link
semaglutide. The covalently attached fatty of two tandem copies of a truncated and between phenotype (peptide) and genotype
acid binds to serum albumin, leading to modified GLP1 analogue fused to human (DNA or RNA coding sequence) to enable
reduced degradation and elimination of albumin125,126. The GLP1 point modifications affinity (or more rarely activity) selection at
the peptides, significantly improving their and protein linkages also protect the the peptide level, with sub­sequent recovery
pharmacokinetics117,119,120. The use of fatty GLP1(7–37) segments from DPP4 and amplification of the genetic material
acids has additional benefits, including degradation. (through PCR or infection of host cells by
delayed release from the injection site and Finally, in the past two decades, phage virions). Phage display, the oldest
reduction of immunogenic responses. pegylation, in which polyethylene glycol technique131, can produce libraries com-
A similar approach was applied in the (PEG) chains are attached to peptides prising up to 1010 unique peptides. In vitro
development of tesamorelin, an analogue of or proteins, has gained momentum, display systems (in particular mRNA display
human growth hormone-​releasing hormone with multiple FDA-​approved pegylated and ribosome display) have the advantage
(GHRH) that stimulates the synthesis and proteins (such as PEG–bovine adenosine that there is no need for host cell transfection,
release of endogenous growth hormone. deaminase and PEG–α-​interferon) and allowing the creation of more diverse libraries
Tesamorelin was approved in 2010 to some pegylated peptides in clinical (1013–1015 peptides)70,132.
treat HIV-​related lipodystrophy121. This trials118,126,127. For example, the complement The first approved peptide therapy
synthetically produced 44-​residue sequence C3 inhibitor pegcetacoplan, a pegylated discovered using display technology was
of GHRH has a hexenoyl moiety at the compstatin analogue, is in clinical trials peginesatide, an erythropoiesis-​stimulating
N-​terminal Tyr residue, which protects for the treatment of paroxysmal nocturnal agent133–135. Phage display libraries were
against DPP4 degradation122. haemoglobinuria, geographic atrophy, screened to isolate small peptides that
Selective conjugation to serum albumin and C3 glomerulopathy128,129. A pegylated bind to and activate the receptor for the
or immunoglobulin — two plasma proteins form of exenatide is in phase II trials for cytokine erythropoietin136, leading to
with an unusually long circulation time the treatment of Parkinson disease and the identification of a synthetic dimeric
(weeks instead of days) — is another strategy type 2 diabetes, while a pegylated analogue peptide comprising two identical 21-​residue
used to exceed the molecular weight cut-​off of the gut hormone oxyntomodulin is in chains each with a single disulfide bond137.
The peptides have no homology to
erythropoietin, yet their effects appear
Box 2 | Animal venoms as a rich source for peptide drug leads to be identical to those induced by the
natural ligand136,138. Covalent linkage of
Venom acquisition transforms predator–prey interactions from physical encounters to chemical the dimeric peptide (~5 kDa) to a single
battles, thereby enabling venomous animals to both prey on and defend themselves against much Lys-​branched bis(methoxypoly(ethylene
larger animals. Venom acquisition is such a transformative event in animal evolution that about
glycol)) PEG chain (~40 kDa) to improve
15% of all extant species are venomous88.
Although there are some exceptions (for example, reptiles and some ant subfamilies), the
its pharmacokinetic characteristics resulted
dominant components of most venoms, such as those from centipedes, cone snails, sea anemones, in peginesatide137, which was approved
scorpions, spiders and most venomous insects, are disulfide-​rich peptides. The disulfide bonds by the FDA in 2012 for the treatment of
provide a rigid scaffold upon which pharmacophores can be displayed within the intercystine anaemia associated with chronic kidney
loops. These loops are typically highly permissive to mutation, so during millions of years of disease in adult patients receiving dialysis.
evolution there has been massive pharmacological diversification within each venom-​peptide However, in 2013, postmarketing reports of
family. This permissiveness to mutation makes these disulfide-​rich peptides ideal for engineering serious hypersensitivity reactions, including
new functions237. potentially life-​threatening anaphylaxis135,
The most widely recruited disulfide scaffold in animal venoms is the three-​disulfide inhibitor led to peginesatide being withdrawn from
cystine knot (ICK or knottin), in which two of the disulfide bridges and the intervening sections
the market.
of the peptide backbone form a closed loop that is pierced by the third disulfide to form a
pseudoknot238. The ICK may be the most ubiquitous disulfide-​rich scaffold in the natural world as
Phage display also led to other approved
it is found in viruses, fungi and animals, and cyclized versions (cyclotides) are present in plants239. peptide drugs (that is, romiplostim139
Although cystine knots are not true knots in a mathematical sense, they nevertheless provide and ecallantide140, see Supplementary
knottins with exceptional stability and resistance to proteases, making them excellent drug leads. information), but it has generally found
The analgesic drug ziconotide is an ICK peptide derived from a venomous cone snail97. Amgen240, greater applications in the development of
Janssen241, MedImmune242 and Merck243 have all developed ICK peptides from spider venom as monoclonal antibody therapeutics than
leads for analgesics that target the NaV1.7 voltage-​gated sodium channel. Venom-​derived ICK peptide drugs141, in part because peptides
peptides are also being developed as drug leads for the treatment of epilepsy, hypokalaemic from phage display often have poor
periodic paralysis and stroke, and they are providing templates for engineering of novel diagnostic pharmaceutical properties. More drug-​like
agents and therapeutics237.
properties, such as peptidase resistance,

316 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

cell membrane permeation ability and

oral bioavailability, can be achieved by
macrocyclization and the introduction of
non-​proteinogenic amino acids (see earlier).
Venomous animals Display technologies
A natural example of a macrocyclic peptide
is cyclosporine, but recent technological
advances now allow the synthesis and
high-​throughput screening of such peptides.
Venom mRNA DNA Library design
For example, split-​intein circular ligation of
Crude venom proteome Venom-gland Venomous animal • DNA
peptides and proteins (SICLOPPS) can be transcriptome genome • mRNA
used to generate macrocyclic peptides inside
cells and phenotypically screen cells for
them; however, it is limited to proteinogenic Assay-guided fractionation Integrated venomics Biopanning
amino acids and has a maximum library size
of ~106 (ref.142). Phage display can also be
Target identification Identification and selection of peptide sequences
used to identify macrocyclic peptides;
for example, the Bicycle screening
platform is a relatively new approach Synthesis, structure-activity relationship studies, medicinal chemistry, preclinical studies
that identifies bicyclic peptides using Increase or improvement of:
phage display143,144. Bicyclic peptides are • Potency and selectivity • Pharmacokinetics and pharmacodynamics
• Stability and bioavailability • In vivo efficacy and biodistribution
conformationally more constrained than
monocyclic peptides and therefore bind
with higher affinity and are typically Therapeutic candidate selection
more resistant to proteolytic degradation,
especially if this property is selected for145. Fig. 5 | Peptide drug discovery strategies. Integrated venomics and display technologies are two key
Using this platform, a bicyclic peptide technologies for therapeutic lead discovery. Integrated venomics utilizes bioinformatics to analyse
inhibitor of coagulation factor XII has been genomic and transcriptomic data of venomous animals along with proteomic data obtained from
developed146, and a modified version of the crude venom samples. This approach can identify a large number of venom peptide sequences, which
peptide has shown promise in preclinical can then be produced synthetically or recombinantly and can be screened against therapeutic targets.
Display technologies can produce vast peptide libraries (1010–1015) that are panned against thera­
work as a thrombosis treatment147. However,
peutic targets. This process usually yields high-​affinity target binders after a few rounds of selection.
as for SICLOPPS, the Bicycle platform is Medicinal chemistry strategies are then used to improve the drug-​like properties of these leads.
generally limited to proteinogenic amino
acids. An elegant strategy that could overcome
this is mirror-​image phage display to identify entered late-​stage clinical trials, one peptide sites, including N-​terminal acetylation,
l-​peptides that bind potently to chemically (zilucoplan), developed using similar tech- N-​methylation, the use of d-​amino acids
synthesized all-​d drug targets148–151. This, in nology (Ra Pharmaceuticals mRNA display or unnatural amino acids and amide bond
turn, delivers blueprints for metabolically platform)153, is currently in phase III trials mimetics (for example, thioamides, peptoids
ultrastable all-​d-​peptide drugs that — for the treatment of generalized myasthenia and β-​amino acids) (Fig. 6). N- to C-​terminal
combined with chemical modifications to gravis (a neuromuscular disease). Zilucoplan cyclization and disulfide-​bond mimetics
evade renal clearance or formulations that was identified from a library of macrocyclic (particularly thioether bonds) will be
improve oral uptake — have the potential peptides using complement component C5 increasingly used in the design of peptide
to become a game-​changer for peptide drug as a target. It inhibits cleavage of C5 into C5a drugs to tune metabolic stability and
development. and C5b, thereby preventing formation of the bioavailability155–158.
Techniques using mRNA display can also terminal complement complex, which can Multiple highly stable cyclic cystine-​knot
be used to screen large libraries of macro- damage and destroy the postsynaptic mem- scaffolds have been identified that allow
cyclic peptides (more than 1012) including brane, disrupt ionic channel conductance grafting of therapeutically relevant pharma­
non-​proteinogenic amino acids (as well as and impair neuromuscular transmission154. cophores into their loops, and it will be
d-​amino acids), due to the use of in vitro Overall, display technologies along with interesting to see how such grafted cyclic
translation71. For example, RaPID (random natural product discoveries are likely to be peptides fare in the clinic159. Advances in
non-​standard peptide integrated discovery) a major source of next-​generation peptide bioinformatics will facilitate the design of
uses ‘flexizymes’ (catalytic RNAs) that can drugs (Fig. 5). such grafts as well as the design of peptido-
charge almost any tRNA with a wide variety mimetics that replicate secondary structure
of non-​proteinogenic amino acids72,152. Such New chemical approaches elements of peptides and proteins. Current
techniques may be useful for targeting pro- Medicinal chemistry strategies in peptide in silico design, however, still faces chal-
teins that have proved intractable to targeting drug development have traditionally been lenges, including the sheer complexity of
with small molecules and are increasingly relatively simple (yet effective) and largely folding, the chirality of peptides, the lack
being used in the pharmaceutical industry. focused on increasing the metabolic and of high-​resolution structures of receptor–
For example, the company PeptiDream circulation half-​life of already very good peptide complexes, subtle conformational
has commercialized this technology and (in terms of potency and selectivity) changes involved in ligand binding and
provided multiple therapeutic leads to the therapeutic peptide leads. We expect the presence of large surface areas instead
pharmaceutical industry. Although no this trend to continue with very specific of binding pockets. More complex chemis­
peptides developed by PeptiDream have yet modifications at identified cleavage tries could eventually appear in clinical

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 317
Perspectives

Box 3 | Ion channels as promising targets for peptide drugs peptide lead is another concept that is
increasingly pursued in peptide drug devel-
Ion channels are a diverse class of pore-​forming membrane proteins that gate ions across cell opment, ranging from pegylation or protein
membranes to initiate action potentials and signalling cascades. They regulate many physiological conjugation as half-​life extension strategies
functions, and their accessible location in the cell membrane and diverse biological roles have made
and cytotoxic payloads for cancer therapy to
them the third most common drug target after GPCRs and kinases244. They comprise numerous
multivalent display, dual pharmacophores,
structural classes245 that are divided into multiple subtypes (see Supplementary Fig. 3): for example,
there are nine subtypes of voltage-​gated sodium (NaV) channels (NaV1.1–NaV1.9) that have distinct targeted delivery vectors or the use of
anatomical locations and functions246,247. Individual subtypes often have distinct (patho)physiological imaging tags for future theranostics. Recent
functions (for example, NaV1.7 is an analgesic target, whereas NaV1.4 regulates skeletal muscle advances in enzymatic ligation of chemically
contractility), and therefore subtype selectivity is crucial for therapeutic targeting of this class synthesized peptide fragments (chemo­
of proteins. enzymatic peptide synthesis) facilitate this
Given their critical role in the nervous system of all animals, it is not surprising that voltage-​gated trend and allow cost-​effective synthesis of
ion channels are a major target class for venomous animals, and that venom peptides modulate the complex peptides and peptide conjugates.
activity of these channels via diverse molecular mechanisms. These channels consist of a central Chemoenzymatic peptide synthesis uses
pore surrounded by four voltage-​sensor domains (VSDs) that enable the channel to respond to
ligases such as sortase, butelase, peptiligase
changes in membrane potential. NaV channels comprise four homologous but non-​identical domains
or omniligase-1 for efficient tandem ligation,
denoted DI–DIV (coloured orange, green, magenta and blue, respectively, in the figure). Some
venom peptides are simple pore blockers (antagonists) that sterically prevent ion flow, such as the peptide modifications and cyclization167.
16-​residue μ-​conotoxin KIIIA from a venomous cone snail that occupies the extracellular vestibule Their cost-​effectiveness and large-​scale
of the human NaV1.2 channel and occludes the channel pore 248 (see the figure, part a). In contrast, compatibility have been demonstrated with
venom peptides known as gating modifiers modulate channel gating (that is, cycling between the gram-​scale exenatide production, resulting
open, closed and inactivated states) by interacting with one or more of the VSDs. An example is in a cost reduction of 50% compared with
the 57-​residue spider venom peptide Dc1a, which extensively interacts with both VSDII and the pore full SPPS168. The use of fully automated and
domain to stabilize an open conformation of the insect NaVPaS channel, thereby acting as a channel scalable flow peptide synthesis is another
agonist249 (see the figure, part b). Both pore-​blocker and gating-​modifier peptides make extensive, technological development that shows
multifocal contacts with ion channels, imparting high affinity and selectivity compared with promise for accelerated synthesis and
small-​molecule modulators. Venom peptides can also indirectly modulate ion channels, as seen
SAR studies169.
with GABAB receptor-​mediated inhibition of CaV2.2/Cav2.3 and potentiation of the channels GIRK1
and GIRK2 with conotoxin Vc1.1 from cone snail venom (see Supplementary information and
Supplementary Fig. 4). Advances in delivery
Most peptide drugs are delivered by injec-
a Human NaV1.2 channel b Cockroach NaVPaS channel tion, which comes with several drawbacks,
including poor patient adherence, accidental
Cone snail venom Spider venom peptide Selectivity injuries, risk of infection, improper use and
peptide μ-conotoxin
Turret
Dc1a (57 residues) filter and pore biohazardous needle waste. The development
KIIIA (16 residues)
region Turret of alternatives to needle-​based injection has
region therefore been a high priority. Here we out-
Selectivity line some of the most recent developments
DII filter and DIV and refer the reader to a review that discusses
VSD pore VSD these strategies in more detail44.
Various pump technologies offer
S3 S5 S6 alternatives to injection. Implantable pumps
Na DI for delivering insulin have been around
S2 S1
VSD since the 1980s, but investment in this
S4
technology dwindled in the first decade of
the twenty-​first century. There are some
DII VSD signs of a comeback; for example, with the
development of microtechnology-​based
implantable pumps, which have more
drug candidates, driven mainly by new cell penetration and target affinity, while precise control over delivery170. However,
platform technologies (for example, from simultaneously decreasing proteolytic implantation of pumps is invasive and they
display/automated assembly/combinatorial/ degradation. So far, two stapled peptides, require refilling.
bacterial expression technologies developed both developed by Aileron Therapeutics, Alternative approaches to delivering
to improve barrier crossing and increase have entered clinical trials. ALRN-6924 drugs across the skin without injection
metabolic stability)70–72,132,142,144,145,148,152,160,161. is designed to reactivate p53 in tumours have included needle-​free injection; for
Another approach that has received by inhibiting its major negative regulators example, liquid jet injectors171. However,
much attention and financial support is the MDM2 and MDMX, and it is currently in pain and bleeding at the site of injection
concept of stapled peptides162–164. In this phase II trials165,166. ALRN-5281 is designed can occur, and there is variability in the
approach, the binding motifs of therapeutic to activate the cell-​surface receptor GHRH to amount of drug delivered, in part due to
proteins are stabilized within short peptides increase growth hormone release in people variation in skin properties among patients.
through rational positioning of covalent with rare endocrine disorders and has Other approaches include those that aim to
crosslinks (staples). In particular, α-​helical completed phase I trials. temporarily disrupt the skin structure, such
stabilization has been pursued by multiple As highlighted earlier, conjugation to as the use of skin-​penetrating peptides172,
biotechnology companies to increase improve the therapeutic properties of a whereas others use ultrasound or electric

318 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

fields to increase skin permeability173. human and in vivo animal studies184,185. the treatment of sudden sensorineural
Microneedle technologies for the delivery of Although these developments are important hearing loss (phase III) and intraocular
insulin174 or abaloparatide175 are in phase II and promising, we are still far from achieving inflammation causing pain in patients
and phase III trials, respectively. true oral bioavailability, which is generally undergoing cataract surgery189. Another
The pulmonary route is another potential defined as more than 20% of the orally intriguing construct is nerinetide,
delivery pathway, but it is accompanied by administered compound observed in which has recently completed phase III
safety issues and limited delivery efficiency. systemic circulation156,186. trials for the treatment of acute ischaemic
Inhalable insulin (Pfizer’s Exubera) received Finally, given the many intracellular stroke190. Nerinetide is composed of Tat
FDA approval in 2006 but was discontinued protein drug targets that may require a linked to the C terminus of NR2B9c,
in 2007 owing in part to safety concerns and substantial surface area for binding or a nine-​residue inhibitor of the intracellular
high costs. With the development of Afrezza inhibition of protein–protein interactions, interaction between postsynaptic
(approved by the FDA in 2014), MannKind it is not surprising that chemical strategies to density protein 95 (PSD95) and NMDA
hopes to revive interest in inhalable insulin, deliver peptides into cells have been actively (N-​methyl-​d-​aspartate) receptors in brain
but safety concerns remain, with side effects pursued for many years. Although most neurons. In this instance, Tat is used
including cough or throat irritation176. peptides have low membrane permeation to deliver intravenously administered
Nasal delivery is another potential route, ability, several peptides with translocation nerinetide across the blood–brain barrier
with desmopressin available as a nasal capacities, known as cell-​penetrating and into neurons. There are now numerous
spray. However, as with the pulmonary peptides (CPPs), have been discovered. CPP-​conjugated therapeutics in preclinical
route, the impact of long-​term delivery These peptides, typically comprising and clinical development191, and we expect
on sensitive mucosa remains a concern. 5–30 residues, can cross membranes multiple CPP constructs to be approved in
Therefore, these routes may be more suited via energy-​dependent endocytosis or the coming years.
for occasional use; for example, glucagon is energy-​independent direct penetration,
available as a dry nasal spray to treat severe although the mechanisms have not been Outlook for peptide drug development
hypoglycaemia reactions177. fully elucidated187,188. CPPs can be used The preceding examples illustrate some
Much effort has been invested in to transport various cargoes, including of the directions in which the peptide
developing strategies to enable oral delivery other peptides, proteins, nanoparticles, therapeutics field has evolved over the last
of peptide drugs, a route that is limited DNAs, small RNAs and small drugs, into six decades, and they provide insight into
by enzymatic digestion of peptides in the cells. One of the most clinically advanced viable strategies and future directions. It is
gastrointestinal tract and their limited and intriguing CPP–drug constructs is clear that the field has come a long way
permeation across the intestinal epithelium AM-111/D-​JNKI-1/XG-102, which is a in overcoming key challenges, yet there
(see refs178,179 for recent reviews). One 31-​residue-​long d-​retro-​inverso construct remains much room for improvement and
strategy to overcome these barriers for of the cell-​penetrating Tat moiety linked to new developments.
which there has been notable clinical a JUN N-​terminal kinase peptide inhibitor. Advances in synthetic methodology,
success is co-​formulation with permeation This construct has been investigated for recombinant production and automation
enhancers. In 2019, semaglutide co‐
formulated with the permeation enhancer
sodium N‐[8‐(2‐hydroxybenzoyl)amino]
Pegylation
caprylate (SNAC) for once‐daily oral S S

administration was approved for the

treatment of type 2 diabetes on the basis
N N
of a phase III programme involving eight H H
O S
clinical trials and several thousand patients Lipidation
β-Amino acid Thioamide
(reviewed in ref.178). SNAC is a small fatty
acid derivative that promotes absorption S–S
of semaglutide across the gastric mucosa S S mimetics
via effects on transcellular pathways180,181.
An orally administered octreotide product O
in which the peptide is encapsulated with N N N
H H H N
medium-​chain free fatty acids and the O O O H
permeation enhancer sodium caprylate is in D-Amino acid L-Amino acid Mimetic Acetylation
phase III trials182, and an orally administered
insulin candidate (ORMD-0801) including S Stapling
a permeation enhancer, soybean trypsin N
O
inhibitor and a chelator is in phase II trials183. O
N
Other strategies that continue to be explored
O S
include pH modulation, mucus-​penetrating
agents, enzyme inhibition, hydrogels, N-methylation Peptoid
intestinal patches, enteric-​coated capsules Fig. 6 | Medicinal chemistry strategies for peptide drugs. Overview of well-​validated chemical
and devices178. For example, devices that modifications used in peptide drug development to increase metabolic stability and bioavailability.
use either drug-​loaded microneedles or Approaches include replacement of selected amino acids by D-​amino acids, β-​amino acids, thioamide
milliposts to deliver peptides into intestinal or side chain mimetics, N-​methylation and N-​terminal acetylation, peptoid design, disulfide bond
tissue recently showed promise in ex vivo mimetics, pegylation, lipidation and stapling to stabilize α-​helices.

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 319
Perspectives

will continue to reduce manufacturing build a strong case for peptides targeting implanted pump) but production costs can
costs of structurally complex peptides. central nervous system (CNS) disorders pose problems for clinical translation. Even
Although synthetic peptide drugs might considering that most peptides struggle to though the multiton synthetic production
never be cheaper to manufacture than their pass the gut barrier. This is unfortunate, of the 36-​mer enfuvirtide was an important
small-​molecule counterparts, it is likely that as peptides would be ideal candidates to step towards commercial viability of peptide
the overall research and development costs treat CNS disorders due to their superior therapeutics, the large-​scale synthetic
of peptide drugs will be lower than those of selectivity and endogenous role as central production costs for a 5,000-​Da peptide
their small-​molecule counterparts due neurotransmitters. Although exceptions will still exceed those of a typical 500-​Da
to their intrinsic synthetic feasibility (which exist, such as chlorotoxin (which targets small-​molecule drug by 10–100-​fold194,195.
enables rapid and extensive SAR studies glioblastomas192) and nerinetide (which Nevertheless, it is important to realize that
for lead optimization) and lower likelihood targets brain neurons in ischaemic stroke190), production costs make up only ~3–5% of
of off-​target effects (which could lead to this barrier will persist until new, robust the total research and development
higher success rates in clinical trials). Recent and broadly applicable peptide technologies and distribution costs for a new drug, and
peptide blockbusters such as the GLP1 emerge that overcome this delivery problem. technological advances in recombinant
receptor agonists liraglutide and dulaglutide Unfulfilled revenue expectation production of peptide therapeutics that
have contributed to the growing industry is another reason why peptide drug also include unnatural amino acids and
interest in this drug class and its acceptance development is often not prioritized. For modifications could play an important role
by physicians and patients. Advances in example, ziconotide, a first-​in-​class drug in addressing this limitation.
peptide drug delivery technologies and for the treatment of severe neuropathic Despite these well-​known barriers, there
peptide receptor radionuclide therapy have pain, which is an area of high unmet need, is a growing market for peptide drugs,
further fuelled this momentum. had unexpected side effects (dizziness, and the question drug developers need
One aspect that needs to be clearly nausea, confusion and nystagmus)193, and to ask is as follows: in which therapeutic
understood, however, is that peptides occupy the US rights were sold to Azur Pharma space might peptides hold advantages over
a specific biophysical and pharmacological for only $14.6 million. Enfuvirtide for the small-​molecule drugs? Insulin illustrates
space, and concepts that work for small treatment of HIV infection (Box 1) had a this concept perfectly, highlighting how a
molecules or biologics might not apply decline in sales due to a combination of peptide drug can successfully fill a particular
to peptide drug development. Despite resistance development and cheaper and niche in the drug market — despite some
advances in delivery technology, formulation orally available small-​molecule drugs of its disadvantages, all attempts to find
techniques and medicinal chemistry, 90% of receiving market approval. This raises an small-​molecule replacements have failed.
all peptide drugs are delivered by injection, important point: not only is the lack of oral In the remainder of this section, we discuss
and lack of oral bioavailability remains bioavailability an obstacle (enfuvirtide needs the characteristics of a good peptide drug
the major limiting barrier in peptide drug to be injected twice daily and ziconotide and the types of targets and indications for
development. Furthermore, it is difficult to is administered intrathecally via an which they are most likely to be successful;
a summary is provided in Box 4.
Box 4 | Characteristics of an ideal peptide drug and peptide drug target
Characteristics of a good peptide drug.
Ideal peptide drug One key advantage of peptides over small
• Potent (effective concentration for half-​maximum response/half-​maximum inhibitory molecules is their larger surface area and
concentration in the low nanomolar range or lower) and selective (more than 50-fold) over key greater chiral and structural complexity.
off-​targets. These attributes can be exploited for drug
• Agonist. Only low receptor occupancy (5–20%) is necessary for receptor activation, whereas targets that require interaction at multiple
antagonists generally must occupy more than 50% of receptors to be effective. Furthermore, and distant sites for target activation. This is
antagonism can be achieved by allosteric receptor interactions, leaving ample room for the case for insulin and is a major reason for
competing orally available small-​molecule drugs. its success: SAR studies illustrate that the
• Antagonist/inhibitor for targets where a large surface area provides an advantage over N terminus and C terminus of insulin’s
small-​molecule drugs (for example, protein–protein interactions and ion channels). A chain and individual residues from the
• Stable due to rigid three-​dimensional structure with secondary structural motifs stabilized by B chain (~14 residues/16 Å apart) are crucial
disulfide bonds, thioethers or staples; N/C termini cyclized, capped or truncated; incorporation for receptor activation196 (Fig. 3b).
of unnatural amino acids to remove metabolic cleavage sites. Agonism versus antagonism is another
• Replacement of methionine residues (for example, with norleucine) to avoid oxidative shelf-​life consideration when peptide drugs are being
problems. developed. Peptide antagonists were slower
• Pharmacokinetics and dynamics matched with biological action, dose regime and safety. to reach the market; this is related to the
• Incorporation of fatty acids (C14/16), pegylation or protein conjugation to evade renal clearance, high production costs in the early days and
if required. Rapid clearance can be beneficial in some cases as it leads to a desirable side effect the fact that only small quantities of agonists
profile. (5–20% receptor occupancy) are necessary
• Compatible with a delivery route/formulation strategy that maximizes patient adherence. for receptor activation, whereas antagonists
Ideal peptide drug target have to compete with the native ligand and
• Extracellular and peripheral to bypass delivery challenges.
must occupy more than 50% of the receptor
population to be effective197,198. More
• Multiple receptor subtypes to leverage the selectivity advantages of peptides.
important, however, is that antagonism
• Targets that require a large surface area for a therapeutic response (for example, insulin receptor,
can be achieved by allosteric receptor
ion channels and protein–protein interactions).
interactions, where the larger surface area

320 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

of peptides may provide little advantage and selectivity that peptides are known disorders. How much oxytocin actually
over competing small-​molecule drugs. for, and sometimes even metabolic stability. gets across the blood–brain barrier and the
An exception is antagonists developed exact mechanism of action are still under
for targets that require drugs with a large Preferred therapeutic targets and indications debate207,208. A potential downside, however,
surface area and structural complexity for peptides. Most peptide drugs are used to is the uptake variability of intranasal
to achieve subtype selectivity and avoid treat endocrine, metabolic or cardio­vascular administration, which could make it
off-​target effects (for example, ion channels; disorders, or cancer, with metabolic disorders difficult for clinical trials to find dosing that
Box 3). Of note, receptor inactivation by (particularly diabetes) and cancer account- is both effective and safe. Other emerging
agonist-​mediated receptor desensitization ing for the largest revenue (Fig. 2; Table 1). approaches and delivery technologies (for
leading to receptor downregulation has been Most peptide drugs modulate peripheral example, blood–brain barrier shuttles and
successfully exploited by the class of GnRH extracellular targets, which is no surprise the gut–brain axis) have been covered in
superagonist peptide drugs. considering their difficulty in crossing cell recent reviews209–214.
Administration by injection and rapid membranes. We predict this trend to con- The fundamental role of peptides as
clearance are often-​cited weaknesses of tinue, with the ideal targets for peptide drugs molecular probes and diagnostic tools in
peptide drugs; however, in many cases, being extracellular receptors with multiple pharmacological and neurological studies
this is beneficial as it allows fast onset of functional subtypes201. For such receptors, has led to several diagnostic agents and
action and leads to a favourable side effect it is difficult to develop small-​molecule devices reaching the commercial market
profile due to rapid clearance. Particularly drugs, and therefore the better selectivity (Supplementary Table 2). Their main
in the treatment of diabetes, fast-​acting profile of peptides (driven by their larger therapeutic application lies in oncology,
drugs are desirable so that drug intake does surface area and chiral complexity) can be where these peptides — often equipped with
not have to precede mealtimes by hours, fully exploited. One example of such a com- radiopharmaceuticals — are used for tumour
giving people with diabetes more flexibility plex receptor subtype system is the various detection and targeted radiotherapy215.
in their daily schedules. At the same time, ion channel subtypes involved in pain signal­ Cancer, in general, has been a steady target
insulin is cleared rapidly with a half-​life of ling, illustrated in Box 3. An additional class for peptide drugs, since an average of one
4–6 min, providing little chance for side of productive targets for peptide drugs is in five peptides entering the clinic may have
effects. Oxytocin, which is clinically used protein–protein interactions that are often application in cancer treatment. This is for
to induce labour, is another example where classified as ‘undruggable’ due to the lack a variety of reasons beyond the substantial
a short half-​life is beneficial due to the of binding pockets in either of the binding unmet medical needs in the oncology area
ubiquitous presence of oxytocin and closely partners202,203. A surface area of 1,500–3,000 Å2 and the supportive regulatory pathways
related vasopressin receptors that can cause must be covered for specific recognition, for new drug development to meet them.
complications during birth. which is a viable task for peptides202 but less First, many peptide hormone receptors
The three-​dimensional structure of so for small molecules. are upregulated in tumour cells, thereby
a peptide often displays evolutionarily Reaching intracellular targets, on forming ideal targets to differentiate
conserved motifs, such as α-​helices, β-​sheets, the other hand, is still a daunting task them from healthy cells, and peptides
β-​turns and γ-​turns that are crucial for for peptide drugs, although recent (including imaging tags or drug cargo) can
receptor recognition, potency, selectivity developments in cell-​penetrating187,191 and enter tumour cells via receptor-​mediated
and proteolytic stability95,199. Insulin, for stapled162–164 peptides are promising. Similar internalization. Second, peptides can have
example, possesses three helical regions limitations hold for CNS targets, even extraordinary receptor subtype selectivity
stabilized by three disulfide bonds that align though the commercial potential is attractive similar to that of monoclonal antibodies,
the residues important for activity196. Such and the selectivity advantage of peptides is but with superior tumour tissue penetration
structural and chiral complexity is often more pronounced considering the density and less immunogenicity. In addition to
lost in small-​molecule drugs, in part due and subtype diversity of receptors in the their selectivity, peptides are also rapidly
to the controversial decision to abandon brain (many modulated by neuropeptides). cleared, thereby further reducing the
natural product discovery and move to We do not predict many new peptide chance of off-​target effects and toxicity.
combinatorial libraries in the 1980s200. drug approvals in this space until the Third, some peptide hormones have the
One of the main reasons for this decision emergence and clinical validation of a new ability to reduce tumour growth and slow
was the synthetic challenges of accessing delivery technology or a different route of cancer progression. Finally, injection is
the steric complexity of natural products in administration that applies to a broad range widely accepted by patients with cancer.
a time-​effective and cost-​effective manner. of peptides. One potential delivery route that We therefore anticipate that this area will
This is, however, not the case in peptide is being explored is intranasal delivery204, continue to expand, including in the field
synthesis, where assembly of the intrinsically which is strongly driven by the behavioural of peptide-​based cancer vaccines, which is
chiral building blocks through SPPS effects observed in human test participants receiving considerable attention216–219.
produces chirally and structurally complex given oxytocin intranasally205,206. Intranasally Another promising anatomical target that
peptides. The same synthetic method can administered oxytocin had been marketed to has received little attention (due to stability
be used to rapidly advance SAR and drug facilitate lactation for breastfeeding mothers concerns) is the gastrointestinal tract, which
optimization studies, a task that is not so but it was withdrawn for commercial is the most highly innervated organ after
simple with complex chiral small molecules. reasons. Initial approval, however, has the brain. There are many therapeutically
It is this chiral complexity, large surface enabled multiple human studies validating relevant gastrointestinal receptors as well
area and the protein-​like structural features CNS effects induced by intranasally as numerous peptides that are expressed
(often stabilized by disulfide bonds) that is administered oxytocin, which is now and function in the hostile gastrointestinal
the true power of this drug class, as these being investigated as a potential treatment environment (for example, defensins and
features result in the remarkable potency option for autism, migraine and bipolar guanylin). Additionally, disulfide-​rich or

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 321
Perspectives

cyclic peptides (for example, cyclotides Peptide therapeutics occupy a 14. Adessi, C. & Soto, C. Converting a peptide into a drug:
strategies to improve stability and bioavailability.
and sunflower trypsin inhibitor) can be well-​defined space in the pharmaceutical Curr. Med. Chem. 9, 963–978 (2002).
grafted to display target specificity as well landscape, in which they can outperform 15. Gentilucci, L., De Marco, R. & Cerisoli, L. Chemical
modifications designed to improve peptide stability:
as prolonged stability in the gut159. The small molecules and larger biologics. incorporation of non-​natural amino acids, pseudo-​
feasi­bility of this approach has been vali- Understanding this space is crucial for peptide bonds, and cyclization. Curr. Pharm. Des. 16,
3185–3203 (2010).
dated with orally administered linaclotide long-​term success, and decisions need to 16. Jost, K., Lebl, M. & Brtnik, F. CRC Handbook of
and plecanatide for the treatment of irritable be made at an early stage as to whether Neurohypophyseal Hormone Analogs. Volumes I & II
(eds Jost, K., Lebl, M. & Brtnik, F.). (CRC Press, 1987).
bowel syndrome with constipation220. This a peptide lead should be moved forward 17. Zaoral, M., Kolc, J. & Sorm, F. Amino acids and
approach should be of particular interest into clinical trials or not, depending on peptides. LXXI. Synthesis of 1-deamino-8-D-​gamma-
aminobutyrine vasopressin, 1-deamino-8-D-​lysine
to the pharmaceutical community as it the therapeutic target, its mechanism of vasopressin, and 1-deamino-8-D-​arginine vasopressin.
presents a shift to orally administered pep- action and how difficult it is to develop a Collect. Czech. Chem. Commun. 32, 1250–1257
(1967).
tide drugs with a high therapeutic window small-​molecule drug for the same target. 18. Dimson, S. B. Desmopressin as a treatment for
(due to non-​systemic exposure) without Incorporation of internal/external peptide enuresis. Lancet 1, 1260 (1977).
19. Melin, P., Trojnar, J., Johansson, B., Vilhardt, H. &
having to overcome long-​standing oral discovery and development programmes Aakerlund, M. Synthetic antagonists of the myometrial
bioavailability issues. alongside existing small-​molecule pipelines, response to vasopressin and oxytocin. J. Endocrinol.
111, 125–131 (1986).
Finally, topical applications, particularly which is currently not the norm, could be 20. Du Vigneaud, V., Winestock, G., Murti, V. V.,
for antimicrobial, antiviral and anti-​ageing beneficial to companies, given the important Hope, D. B. & Kimbrough, R. D. Jr. Synthesis of
1-beta-​mercantopropionic acid oxytocin (desamino-​
peptides, are also expected to grow in the role peptides play as molecular probes for oxytocin), a highly potent analogue of oxytocin. J. Biol.
future221. Indeed, this may be the most bioassays as well as leads to complement and Chem. 235, PC64–PC66 (1960).
21. Hope, D. B., Murti, V. V. S. & du Vigneaud, V. A highly
promising route for antimicrobial peptides, guide small-​molecule drug development. potent analog of oxytocin, deaminooxytocin. J. Biol.
which often fail in preclinical testing or Chem. 237, 1563–1566 (1962).
in the discovery phase due to proteolytic Markus Muttenthaler   1,2 ✉, Glenn F. King   1, 22. Manning, M., Balaspiri, L., Acosta, M. & Sawyer, W. H.
Solid phase synthesis of [1-deamino,4-valine]-8-
David J. Adams   3 and Paul F. Alewood   1 ✉
degradation or toxicity in vivo222. Peptides D-​arginine-vasopressin (DVDAVP), a highly potent
1
Institute for Molecular Bioscience, The University of and specific antidiuretic agent possessing protracted
are also becoming increasingly well known effects. J. Med. Chem. 16, 975–978 (1973).
Queensland, Brisbane, QC, Australia.
in the cosmetics industry, where they are 23. Kyncl, J. & Rudinger, J. Excretion of antidiuretic activity
2
Institute of Biological Chemistry, Faculty of Chemistry, in the urine of cats and rats after administration of the
marketed as anti-​ageing agents. For example, University of Vienna, Vienna, Austria. synthetic hormonogen, Nα-glycyl-glycyl-​glycyl-[8-lysine]-
a synthetic hexapeptide marketed under the 3
Illawarra Health and Medical Research Institute,
vasopressin (triglycylvasopressin). J. Endocrinol. 48,
157–165 (1970).
name Argireline acts similarly to botulinum University of Wollongong, Wollongong, NSW, Australia. 24. Kruszynski, M. et al. [1-(β-​mercapto-β,
neurotoxins to reduce wrinkles223. Palmitoyl ✉e-​mail: [email protected]; β-cyclopentamethylenepropionic acid),2-(O-​methyl)
tyrosine]arginine-​vasopressin and [1-(β-​mercapto-β,
pentapeptide 4 (Matrixyl), launched in [email protected] β-cyclopentamethylenepropionic acid)]arginine-​
2000, is a fragment of procollagen I and is https://doi.org/10.1038/s41573-020-00135-8 vasopressin, two highly potent antagonists of the
vasopressor response to arginine-​vasopressin. J. Med.
thought to reduce wrinkles by stimulating Published online 3 February 2021 Chem. 23, 364–368 (1980).
extracellular matrix renewal in fibroblasts224; 25. Meraldi, J. P., Hruby, V. J. & Brewster, A. I. R.
Relative conformational rigidity in oxytocin and
the palmitoyl chain improves delivery across 1. Merrifield, R. B. Solid phase peptide synthesis. I.
[1-penicillamine]oxytocin: a proposal for the relation
The synthesis of a tetrapeptide. J. Am. Chem. Soc. 85,
skin. Copper peptide GHK-​Cu is a naturally 2149–2154 (1963). of conformational flexibility to peptide hormone
agonism and antagonism. Proc. Natl Acad. Sci. USA
occurring tripeptide with high affinity for 2. Reichert, J. Development trends for peptide
74, 1373–1377 (1977).
therapeutics (Peptide Therapeutics Foundation,
copper ions that has anti-​inflammatory 2010). 26. Walter, R. & du Vigneaud, V. 1-Deamino-1,6-
properties and stimulates collagen and 3. Fosgerau, K. & Hoffmann, T. Peptide therapeutics: L-selenocystineoxytocin; a highly potent isolog of
1-deaminooxytocin. J. Am. Chem. Soc. 88, 1331–1332
current status and future directions. Drug Discov.
glycosaminoglycan synthesis in skin Today 20, 122–128 (2014). (1966).
fibroblasts225. 4. Lau, J. L. & Dunn, M. K. Therapeutic peptides: 27. Walter, R. & du Vigneaud, V. 6-Hemi-​L-selenocystine-​
historical perspectives, current development trends, oxytocin and 1-deamino-6-hemi-​L-selenocystine-​
and future directions. Bioorg. Med. Chem. 26, oxytocin, highly potent isologs of oxytocin and
Conclusions 2700–2707 (2018). 1-deamino-​oxytocin. J. Am. Chem. Soc. 87,
5. Matchar, D. B. et al. Systematic review: comparative 4192–4193 (1965).
The last 60 years has seen a steady approval effectiveness of angiotensin-​converting enzyme 28. Yamanaka, T. et al. Crystalline deamino-​dicarba-
of peptide therapeutics, a trend that we inhibitors and angiotensin II receptor blockers for oxytocin. Preparation and some pharmacological
treating essential hypertension. Ann. Intern. Med. properties. Mol. Pharmacol. 6, 474–480 (1970).
expect to accelerate, since biologics have 148, 16–29 (2008). 29. Sweeney, G. et al. Pharmacokinetics of carbetocin,
removed the traditional assumption that 6. Izzo, J. L. Jr. & Weir, M. R. Angiotensin-​converting a long-​acting oxytocin analog, in nonpregnant women.
enzyme inhibitors. J. Clin. Hypertens. 13, 667–675 Curr. Ther. Res. 47, 528–540 (1990).
a drug needs to be orally available to be (2011). 30. Manning, M. et al. Oxytocin and vasopressin agonists
successful. The peptide field has reached 7. Regulska, K., Stanisz, B., Regulski, M. & Murias, M. and antagonists as research tools and potential
How to design a potent, specific, and stable angiotensin- therapeutics. J. Neuroendocrinol. 24, 609–628
maturity in many aspects, including converting enzyme inhibitor. Drug Discov. Today 19, (2012).
platforms that can reliably and rapidly 1731–1743 (2014). 31. Manning, M. et al. Peptide and non-​peptide agonists
8. Acharya, K. R., Sturrock, E. D., Rirodan, J. F. & and antagonists for the vasopressin and oxytocin V1a,
produce structurally complex peptides of up Ehlers, M. R. W. ACE revisited: a new target for V1b, V2 and OT receptors: research tools and
to ~100 amino acids on a large scale, which structure-​based drug design. Nat. Rev. Drug Discov. potential therapeutic agents. Prog. Brain Res. 170,
2, 891–902 (2003). 473–512 (2008).
enables efficient and cost-​effective SAR 9. Regulski, M. et al. Chemistry and pharmacology of 32. Ling, N., Burgus, R., Rivier, J., Vale, W. & Brazeau, P.
and lead optimization studies. Emerging angiotensin-​converting enzyme inhibitors. Curr. Pharm. Use of mass spectrometry in deducing the sequence of
Des. 21, 1764–1775 (2015). somatostatin, a hypothalamic polypeptide that inhibits
peptide drug discovery technologies provide 10. Luther, A., Bisang, C. & Obrecht, D. Advances in the secretion of growth hormone. Biochem. Biophys.
a rational and powerful approach to develop macrocyclic peptide-​based antibiotics. Bioorg. Med. Res. Commun. 50, 127–133 (1973).
Chem. 26, 2850–2858 (2018). 33. Theodoropoulou, M. & Stalla, G. K. Somatostatin
potent and selective lead compounds on 11. Brown, E. D. & Wright, G. D. Antibacterial drug receptors: from signaling to clinical practice. Front.
short timescales, and the integration of new discovery in the resistance era. Nature 529, 336–343 Neuroendocrinol. 34, 228–252 (2013).
(2016). 34. Biron, E. et al. Improving oral bioavailability of
and optimized delivery technologies with 12. Infoholic Research LLP. Global Human Insulin Market peptides by multiple N-​methylation: somatostatin
innovative chemical strategies will continue 2018–2024. Research and Markets, ID: 4470733 analogues. Angew. Chem. Int. Ed. 47, 2595–2599
(2018). (2008).
to contribute to the future growth of the 13. Nestor, J. J. The medicinal chemistry of peptides. 35. Janecka, A., Zubrzycka, M. & Janecki, T. Somatostatin
peptide field. Curr. Med. Chem. 16, 4399–4418 (2009). analogs. J. Pept. Res. 58, 91–107 (2001).

322 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

36. Vale, W., Brown, M., Rivier, C., Perrin, M. & Rivier, J. 65. Achenbach, J. et al. Outwitting EF-​Tu and the 93. Nielsen, L. L., Young, A. A. & Parkes, D. G.
Development and applications of analogs of LRF and ribosome: translation with D-​amino acids. Nucleic Pharmacology of exenatide (synthetic exendin-4):
somatostatin. in Brain Peptides: A New Endocrinology, Acids Res. 43, 5687–5698 (2015). a potential therapeutic for improved glycemic control
71–88 (Elsevier/North-​Holland Biomedical Press, 66. Fujino, T., Goto, Y., Suga, H. & Murakami, H. of type 2 diabetes. Regul. Pept. 117, 77–88 (2004).
1979). Ribosomal synthesis of peptides with multiple β-​amino 94. Eng, J., Kleinman, W. A., Singh, L., Singh, G. &
37. Susini, C. & Buscail, L. Rationale for the use of acids. J. Am. Chem. Soc. 138, 1962–1969 (2016). Raufman, J. P. Isolation and characterization of
somatostatin analogs as antitumor agents. Ann. 67. Katoh, T. & Suga, H. Ribosomal incorporation of exendin-4, an exendin-3 analogue, from Heloderma
Oncol. 17, 1733–1742 (2006). consecutive β-​amino acids. J. Am. Chem. Soc. 140, suspectum venom. Further evidence for an exendin
38. De Jong, M., Breeman, W. A. P., Kwekkeboom, D. J., 12159–12167 (2018). receptor on dispersed acini from guinea pig pancreas.
Valkema, R. & Krenning, E. P. Tumor imaging and 68. Maini, R. et al. Ribosomal formation of thioamide J. Biol. Chem. 267, 7402–7405 (1992).
therapy using radiolabeled somatostatin analogues. bonds in polypeptide synthesis. J. Am. Chem. Soc. 95. Ruiz-​Gomez, G., Tyndall, J. D., Pfeiffer, B., Abbenante, G.
Acc. Chem. Res. 42, 873–880 (2009). 141, 20004–20008 (2019). & Fairlie, D. P. Update 1 of: over one hundred peptide-​
39. Kwekkeboom, D. J. et al. [177Lu-​DOTA0Tyr3]octreotate: 69. Kawakami, T., Murakami, H. & Suga, H. Ribosomal activated G protein-​coupled receptors recognize
comparison with [111In-​DTPA0]octreotide in patients. synthesis of polypeptoids and peptoid-​peptide hybrids. ligands with turn structure. Chem. Rev. 110,
Eur. J. Nucl. Med. 28, 1319–1325 (2001). J. Am. Chem. Soc. 130, 16861–16863 (2008). PR1–PR41 (2010).
40. Brabander, T. et al. Long-​term efficacy, survival, and 70. Huang, Y., Wiedmann, M. M. & Suga, H. RNA display 96. DeYoung, M. B., MacConell, L., Sarin, V., Trautmann, M.
safety of [177Lu-​DOTA0,Tyr3]octreotate in patients with methods for the discovery of bioactive macrocycles. & Herbert, P. Encapsulation of exenatide in poly-
gastroenteropancreatic and bronchial neuroendocrine Chem. Rev. 119, 10360–10391 (2018). (D,L-​lactide-Co-​glycolide) microspheres produced an
tumors. Clin. Cancer Res. 23, 4617–4624 (2017). 71. Taylor, R. D., Rey-​Carrizo, M., Passioura, T. & Suga, H. investigational long-​acting once-​weekly formulation
41. Strosberg, J. et al. Phase 3 trial of 177Lu-​Dotatate for Identification of nonstandard macrocyclic peptide for type 2 diabetes. Diabetes Technol. Ther. 13,
midgut neuroendocrine tumors. N. Engl. J. Med. 376, ligands through display screening. Drug Discov. Today 1145–1154 (2011).
125–135 (2017). Technol. 26, 17–23 (2017). 97. Miljanich, G. P. Ziconotide: neuronal calcium channel
42. Millar, R. P. & Newton, C. L. Current and future 72. Passioura, T. & Suga, H. A RaPID way to discover blocker for treating severe chronic pain. Curr. Med.
applications of GnRH, kisspeptin and neurokinin B nonstandard macrocyclic peptide modulators of drug Chem. 11, 3029–3040 (2004).
analogues. Nat. Rev. Endocrinol. 9, 451–466 (2013). targets. Chem. Commun. 53, 1931–1940 (2017). 98. Vetter, I. et al. Venomics: a new paradigm for natural
43. Tan, O. & Bukulmez, O. Biochemistry, molecular 73. Borel, J. F., Feurer, C., Gubler, H. U. & Staehelin, H. products-​based drug discovery. Amino Acids 40,
biology and cell biology of gonadotropin-​releasing Biological effects of cyclosporin A: a new 15–28 (2011).
hormone antagonists. Curr. Opin. Obstet. Gynecol. antilymphocytic agent. Agents Actions 6, 468–475 99. Dutertre, S. et al. in Venoms to Drugs: Venom as a
23, 238–244 (2011). (1976). Source for the Development of Human Therapeutics
44. Mitragotri, S., Burke, P. A. & Langer, R. Overcoming 74. Saehelin, H. F. The history of cyclosporin A (ed. King, G. F.) 80–96 (Royal Society of Chemistry,
the challenges in administering biopharmaceuticals: (Sandimmune) revisited: another point of view. 2015).
formulation and delivery strategies. Nat. Rev. Drug Experientia 52, 5–13 (1996). 100. Klint, J. K. et al. Production of recombinant disulfide-​
Discov. 13, 655–672 (2014). 75. Lipinski, C. A., Lombardo, F., Dominy, B. W. & rich venom peptides for structural and functional
45. Zhang, J., Desale, S. S. & Bronich, T. K. Polymer-​based Feeney, P. J. Experimental and computational analysis via expression in the periplasm of E. coli.
vehicles for therapeutic peptide delivery. Ther. Deliv. 6, approaches to estimate solubility and permeability in PLoS ONE 8, e63865 (2013).
1279–1296 (2015). drug discovery and development settings. Adv. Drug 101. Muttenthaler, M. et al. Solving the α-​conotoxin folding
46. Wang, Y., Qu, W. & Choi, S. H. FDA’s regulatory Deliv. Rev. 46, 3–26 (2001). problem: efficient selenium-​directed on-​resin
science program for generic PLA/PLGA-​based drug 76. Rydel, T. J. et al. The structure of a complex of generation of more potent and stable nicotinic
products. Am. Pharm. Rev. 20, 52–55 (2017). recombinant hirudin and human α-​thrombin. Science acetylcholine receptor antagonists. J. Am. Chem. Soc.
47. Itakura, K. et al. Expression in Escherichia coli of a 249, 277–280 (1990). 132, 3514–3522 (2010).
chemically synthesized gene for the hormone 77. Warkentin, T. E. & Koster, A. Bivalirudin: a review. 102. Muttenthaler, M. & Alewood, P. F. Selenopeptide
somatostatin. Science 198, 1056–1063 (1977). Expert Opin. Pharmacother. 6, 1349–1371 (2005). chemistry. J. Pept. Sci. 14, 1223–1239 (2008).
48. Johnson, I. S. Human insulin from recombinant DNA 78. Behrendt, R., White, P. & Offer, J. Advances in Fmoc 103. Vetter, I., Hodgson, W. C., Adams, D. J. & McIntyre, P.
technology. Science 219, 632–637 (1983). solid-​phase peptide synthesis. J. Pept. Sci. 22, 4–27 in Venoms to drugs: venom as a source for the
49. Zaykov, A. N., Mayer, J. P. & DiMarchi, R. D. Pursuit of (2016). development of human therapeutics (ed. King, G. F.)
a perfect insulin. Nat. Rev. Drug Discov. 15, 425–439 79. Coin, I., Beyermann, M. & Bienert, M. Solid-​phase 97–128 (Royal Society of Chemistry, 2015).
(2016). peptide synthesis: from standard procedures to the 104. Ziemert, N., Alanjary, M. & Weber, T. The evolution of
50. Goeddel, D. V. et al. Expression in Escherichia coli synthesis of difficult sequences. Nat. Protoc. 2, genome mining in microbes - a review. Nat. Prod. Rep.
of chemically synthesized genes for human insulin. 3247–3256 (2007). 33, 988–1005 (2016).
Proc. Natl Acad. Sci. USA 76, 106–110 (1979). 80. Paradis-​Bas, M., Tulla-​Puche, J. & Albericio, F. 105. Makarewich, C. A. & Olson, E. N. Mining for
51. Hirsch, I. B. Insulin analogues. N. Engl. J. Med. 352, The road to the synthesis of “difficult peptides”. Chem. micropeptides. Trends Cell Biol. 27, 685–696 (2017).
174–183 (2005). Soc. Rev. 45, 631–654 (2016). 106. Hetrick, K. J. & van der Donk, W. A. Ribosomally
52. Inzerillo, A. M., Zaidi, M. & Huang, C. L. H. Calcitonin: 81. Schnölzer, M., Alewood, P. F., Jones, A., Alewood, D. synthesized and post-​translationally modified peptide
physiological actions and clinical applications. & Kent, S. B. H. In situ neutralization in Boc-​chemistry natural product discovery in the genomic era. Curr.
J. Pediatr. Endocrinol. Metab. 17, 931–940 (2004). solid phase peptide synthesis. Rapid, high yield Opin. Chem. Biol. 38, 36–44 (2017).
53. Copp, D. H. & Cheney, B. Calcitonin-​a hormone from assembly of difficult sequences. Int. J. Pept. Protein 107. Tietz, J. I. et al. A new genome-​mining tool redefines
the parathyroid which lowers the calcium level of the Res. 40, 180–193 (1992). the lasso peptide biosynthetic landscape. Nat. Chem.
blood. Nature 193, 381–382 (1962). 82. Dawson, P. E., Muir, T. W., Clark-​Lewis, I. & Kent, S. B. Biol. 13, 470–478 (2017).
54. Copp, D. H. & Cameron, E. C. Demonstration of a Synthesis of proteins by native chemical ligation. 108. Mendel, H. C., Kaas, Q. & Muttenthaler, M.
hypocalcemic factor (calcitonin) in commercial Science 266, 776–779 (1994). Neuropeptide signalling systems - an underexplored
parathyroid extract. Science 134, 2038 (1961). 83. Miranda, L. P. & Alewood, P. F. Accelerated chemical target for venom drug discovery. Biochem. Pharmacol.
55. Collip, J. B. The extraction of a parathyroid hormone synthesis of peptides and small proteins. Proc. Natl 181, 114129 (2020).
which will prevent or control parathyroid tetany and Acad. Sci. USA 96, 1181–1186 (1999). 109. Gruber, C. W. & Muttenthaler, M. Discovery of defense-
which regulates the level of blood calcium. J. Biol. 84. Kent, S. B. H. Total chemical synthesis of proteins. and neuropeptides in social ants by genome-​mining.
Chem. 63, 395–438 (1925). Chem. Soc. Rev. 38, 338–351 (2009). PLoS ONE 7, e32559 (2012).
56. Kim, E. S. & Keating, G. M. Recombinant human 85. Kent, S. Chemical protein synthesis: inventing 110. Ling, L. L. et al. A new antibiotic kills pathogens
parathyroid hormone (1–84): a review in synthetic methods to decipher how proteins work. without detectable resistance. Nature 517, 455–459
hypoparathyroidism. Drugs 75, 1293–1303 2015). Bioorg. Med. Chem. 25, 4926–4937 (2017). (2015).
57. Haas, A. V. & LeBoff, M. S. Osteoanabolic agents for 86. King, G. F. Venoms as a platform for human drugs: 111. Karas, J. A. et al. Synthesis and structure-​activity
osteoporosis. J. Endocr. Soc. 2, 922–932 (2018). translating toxins into therapeutics. Expert Opin. Biol. relationships of teixobactin. Ann. N. Y. Acad. Sci.
58. Huang, Y. & Liu, T. Therapeutic applications of genetic Ther. 11, 1469–1484 (2011). 1459, 86–105 (2020).
code expansion. Synth. Syst. Biotechnol. 3, 150–158 87. Robinson, S. D., Undheim, E. A. B., Ueberheide, B. & 112. Gunjal, V. B., Thakare, R., Chopra, S. & Reddy, D. S.
(2018). King, G. F. Venom peptides as therapeutics: advances, Teixobactin: a paving stone toward a new class of
59. Young, D. D. & Schultz, P. G. Playing with the challenges and the future of venom-​peptide discovery. antibiotics? J. Med. Chem. 63, 12171–12195
molecules of life. ACS Chem. Biol. 13, 854–870 Expert Rev. Proteom. 14, 931–939 (2017). (2020).
(2018). 88. Holford, M., Daly, M., King, G. F. & Norton, R. S. 113. Mookherjee, N., Anderson, M. A., Haagsman, H. P. &
60. Arranz-​Gibert, P., Vanderschuren, K. & Isaacs, F. J. Venoms to the rescue: insights into the evolutionary Davidson, D. J. Antimicrobial host defence peptides:
Next-​generation genetic code expansion. Curr. Opin. biology of venoms are leading to therapeutic functions and clinical potential. Nat. Rev. Drug Discov.
Chem. Biol. 46, 203–211 (2018). advances. Science 361, 842–844 (2018). 19, 311–332 (2020).
61. Subtelny, A. O., Hartman, M. C. T. & Szostak, J. W. 89. Jin, A.-H. et al. Conotoxins: chemistry and biology. 114. Johnson, V. & Maack, T. Renal extraction, filtration,
Ribosomal synthesis of N-​methyl peptides. J. Am. Chem. Rev. 119, 11510–11549 (2019). absorption, and catabolism of growth hormone. Am. J.
Chem. Soc. 130, 6131–6136 (2008). 90. Akondi, K. B. et al. Discovery, synthesis, and structure-​ Physiol. 233, F185–F196 (1977).
62. Kawakami, T., Murakami, H. & Suga, H. Messenger activity relationships of conotoxins. Chem. Rev. 114, 115. Maack, T., Johnson, V., Kau, S. T., Figueiredo, J. &
RNA-​programmed incorporation of multiple N-​methyl- 5815–5847 (2014). Sigulem, D. Renal filtration, transport, and metabolism
amino acids into linear and cyclic peptides. Chem. Biol. 91. Drucker, D. J. & Nauck, M. A. The incretin system: of low-​molecular-weight proteins: a review. Kidney Int.
15, 32–42 (2008). glucagon-​like peptide-1 receptor agonists and 16, 251–270 (1979).
63. Goto, Y., Murakami, H. & Suga, H. Initiating translation dipeptidyl peptidase-4 inhibitors in type 2 diabetes. 116. Katz, A. I. & Emmanouel, D. S. Metabolism of
with D-​amino acids. RNA 14, 1390–1398 (2008). Lancet 368, 1696–1705 (2006). polypeptide hormones by the normal kidney and in
64. Fujino, T., Goto, Y., Suga, H. & Murakami, H. 92. Elahi, D. et al. The insulinotropic actions of glucose-​ uremia. Nephron 22, 61–72 (1978).
Reevaluation of the D-​amino acid compatibility with dependent insulinotropic polypeptide (GIP) and 117. Pollaro, L. & Heinis, C. Strategies to prolong the
the elongation event in translation. J. Am. Chem. Soc. glucagon-​like peptide-1 (7–37) in normal and diabetic plasma residence time of peptide drugs. Med. Chem.
135, 1830–1837 (2013). subjects. Regul. Pept. 51, 63–75 (1994). Commun. 1, 319–324 (2010).

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 323
Perspectives

118. Kolate, A. et al. PEG - a versatile conjugating ligand for identification to clinical practice. MAbs 6, 73–85 168. Nuijens, T. et al. Engineering a diverse ligase toolbox
drugs and drug delivery systems. J. Control. Release (2014). for peptide segment condensation. Adv. Synth. Catal.
192, 67–81 (2014). 142. Tavassoli, A. SICLOPPS cyclic peptide libraries in drug 358, 4041–4048 (2016).
119. Kurtzhals, P. et al. Albumin binding of insulins acylated discovery. Curr. Opin. Chem. Biol. 38, 30–35 (2017). 169. Mijalis, A. J. et al. A fully automated flow-​based
with fatty acids: characterization of the ligand-​protein 143. Rentero Rebollo, I. & Heinis, C. Phage selection of approach for accelerated peptide synthesis. Nat.
interaction and correlation between binding affinity bicyclic peptides. Methods 60, 46–54 (2013). Chem. Biol. 13, 464–466 (2017).
and timing of the insulin effect in vivo. Biochem. J. 144. Deyle, K., Kong, X.-D. & Heinis, C. Phage selection of 170. Farra, R. et al. First-​in-human testing of a wirelessly
312, 725–731 (1995). cyclic peptides for application in research and drug controlled drug delivery microchip. Sci. Transl. Med. 4,
120. Elbrond, B. et al. Pharmacokinetics, pharmacodynamics, development. Acc. Chem. Res. 50, 1866–1874 122ra21 (2012).
safety, and tolerability of a single-​dose of NN2211, a (2017). 171. Hogan, N. C., Taberner, A. J., Jones, L. A. & Hunter, I. W.
long-​acting glucagon-​like peptide 1 derivative, in healthy 145. Kong, X.-D. et al. De novo development of Needle-​free delivery of macromolecules through the
male subjects. Diabetes Care 25, 1398–1404 (2002). proteolytically resistant therapeutic peptides for oral skin using controllable jet injectors. Expert Opin. Drug
121. Falutz, J. et al. Metabolic effects of a growth hormone-​ administration. Nat. Biomed. Eng. 4, 560–571 (2020). Deliv. 12, 1637–1648 (2015).
releasing factor in patients with HIV. N. Engl. J. Med. 146. Baeriswyl, V. et al. A synthetic factor XIIa inhibitor 172. Kumar, S. et al. Peptides as skin penetration
357, 2359–2370 (2007). blocks selectively intrinsic coagulation initiation. enhancers: mechanisms of action. J. Control. Release
122. Ferdinandi, E. S. et al. Non-​clinical pharmacology and ACS Chem. Biol. 10, 1861–1870 (2015). 199, 168–178 (2015).
safety evaluation of TH9507, a human growth 147. Middendorp, S. J. et al. Peptide macrocycle inhibitor of 173. Zhang, Y. et al. Advances in transdermal insulin
hormone-​releasing factor analogue. Basic Clin. coagulation factor XII with subnanomolar affinity and delivery. Adv. Drug Deliv. Rev. 139, 51–70 (2019).
Pharmacol. Toxicol. 100, 49–58 (2007). high target selectivity. J. Med. Chem. 60, 1151–1158 174. Kochba, E., Levin, Y., Raz, I. & Cahn, A. Improved
123. Baggio, L. L., Huang, Q., Brown, T. J. & Drucker, D. J. (2017). insulin pharmacokinetics using a novel microneedle
A recombinant human glucagon-​like peptide (GLP)- 148. Zhao, L. & Lu, W. Mirror image proteins. Curr. Opin. device for intradermal delivery in patients with type 2
1-albumin protein (Albugon) mimics peptidergic Chem. Biol. 22, 56–61 (2014). diabetes. Diabetes Technol. Ther. 18, 525–531
activation of GLP-1 receptor-​dependent pathways 149. Zhou, X. et al. A novel D-​peptide identified by (2016).
coupled with satiety, gastrointestinal motility, and mirror-image phage display blocks TIGIT/PVR for 175. Daddona, P. E., Matriano, J. A., Mandema, J.
glucose homeostasis. Diabetes 53, 2492–2500 cancer immunotherapy. Angew. Chem. Int. Ed. 59, & Maa, Y.-F. Parathyroid hormone (1-34)-coated
(2004). 15114–15118 (2020). microneedle patch system: clinical pharmacokinetics
124. Matthews, J. E. et al. Pharmacodynamics, 150. Diaz-​Perlas, C. et al. Protein chemical synthesis and pharmacodynamics for treatment of osteoporosis.
pharmacokinetics, safety, and tolerability of combined with mirror-​image phage display yields Pharm. Res. 28, 159–165 (2011).
albiglutide, a long-​acting glucagon-​like peptide-1 D-peptide EGF ligands that block the EGF-​EGFR 176. Kim, E. S. & Plosker, G. L. AFREZZA® (insulin human)
mimetic, in patients with type 2 diabetes. J. Clin. interaction. ChemBioChem 20, 2079–2084 (2019). inhalation powder: a review in diabetes mellitus.
Endocrinol. Metab. 93, 4810–4817 (2008). 151. Rudolph, S. et al. Competitive mirror image phage Drugs 75, 1679–1686 (2015).
125. Glaesner, W. et al. Engineering and characterization display derived peptide modulates amyloid beta 177. Sherr, J. L. et al. Glucagon nasal powder: a promising
of the long-​acting glucagon-​like peptide-1 analogue aggregation and toxicity. PLoS ONE 11, e0147470 alternative to intramuscular glucagon in youth with
LY2189265, an Fc fusion protein. Diabetes Metab. (2016). type 1 diabetes. Diabetes Care 39, 555–562 (2016).
Res. Rev. 26, 287–296 (2010). 152. Tsiamantas, C., Otero-​Ramirez Manuel, E. & Suga, H. 178. Drucker, D. J. Advances in oral peptide therapeutics.
126. D’Souza, A. A. & Shegokar, R. Polyethylene glycol Discovery of functional macrocyclic peptides by means Nat. Rev. Drug Discov. 19, 277–289 (2020).
(PEG): a versatile polymer for pharmaceutical of the RaPID system. Methods Mol. Biol. 2001, 179. Brayden, D. J., Hill, T. A., Fairlie, D. P., Maher, S. &
applications. Expert Opin. Drug Deliv. 13, 1257–1275 299–315 (2019). Mrsny, R. J. Systemic delivery of peptides by the oral
(2016). 153. Guillen Schlippe, Y. V., Hartman, M. C. T., Josephson, K. route: formulation and medicinal chemistry approaches.
127. Park, E. J., Choi, J., Lee, K. C. & Na, D. H. Emerging & Szostak, J. W. In vitro selection of highly modified Adv. Drug Deliv. Rev. 157, 2–36 (2020).
PEGylated non-​biologic drugs. Expert Opin. Emerg. cyclic peptides that act as tight binding inhibitors. 180. Granhall, C., Soendergaard, F. L., Thomsen, M.
Drugs 24, 107–119 (2019). J. Am. Chem. Soc. 134, 10469–10477 (2012). & Anderson, T. W. Pharmacokinetics, safety and
128. Sahu, A., Kay, B. K. & Lambris, J. D. Inhibition of 154. Howard, J. F. et al. Clinical effects of the self-​ tolerability of oral semaglutide in subjects with renal
human complement by a C3-binding peptide isolated administered subcutaneous complement inhibitor impairment. Clin. Pharmacokinet. 57, 1571–1580
from a phage-​displayed random peptide library. zilucoplan in patients with moderate to severe (2018).
J. Immunol. 157, 884–891 (1996). generalized myasthenia gravis: results of a phase 2 181. Ahnfelt-​Roenne, J. et al. Transcellular stomach
129. Liao, D. S. et al. Complement C3 inhibitor randomized, double-​blind, placebo-​controlled, absorption of a derivatized glucagon-​like peptide-1
pegcetacoplan for geographic atrophy secondary to multicenter clinical trial. JAMA Neurol. 77, 582–592 receptor agonist. Sci. Transl. Med. 10, eaar7047
age-​related macular degeneration: a randomized (2020). (2018).
phase 2 trial. Ophthalmology 127, 186–195 (2020). 155. Craik, D. J., Fairlie, D. P., Liras, S. & Price, D. The 182. Biermasz, N. R. New medical therapies on the horizon:
130. Bianchi, E. et al. A PEGylated analog of the gut hormone future of peptide-​based drugs. Chem. Biol. Drug Des. oral octreotide. Pituitary 20, 149–153 (2017).
oxyntomodulin with long-​lasting antihyperglycemic, 81, 136–147 (2013). 183. Eldor, R., Arbit, E., Corcos, A. & Kidron, M. Glucose-​
insulinotropic and anorexigenic activity. Bioorg. Med. 156. Nielsen, D. S. et al. Orally absorbed cyclic peptides. reducing effect of the ORMD-0801 oral insulin
Chem. 21, 7064–7073 (2013). Chem. Rev. 117, 8094–8128 (2017). preparation in patients with uncontrolled type 1
131. Smith, G. P. Filamentous fusion phage: novel 157. Muttenthaler, M. et al. Modulating oxytocin activity diabetes: a pilot study. PLoS ONE 8, e59524 (2013).
expression vectors that display cloned antigens on the and plasma stability by disulfide bond engineering. 184. Abramson, A. et al. A luminal unfolding microneedle
virion surface. Science 228, 1315–1317 (1985). J. Med. Chem. 53, 8585–8596 (2010). injector for oral delivery of macromolecules. Nat. Med.
132. Davis, A. M., Plowright, A. T. & Valeur, E. Directing 158. Erak, M., Bellmann-​Sickert, K., Els-​Heindl, S. 25, 1512–1518 (2019).
evolution: the next revolution in drug discovery? & Beck-​Sickinger, A. G. Peptide chemistry toolbox - 185. Abramson, A. et al. An ingestible self-​orienting system
Nat. Rev. Drug Discov. 16, 681–698 (2017). Transforming natural peptides into peptide for oral delivery of macromolecules. Science 363,
133. Schmid, H. Peginesatide for the treatment of renal therapeutics. Bioorg. Med. Chem. 26, 2759–2765 611–615 (2019).
disease-​induced anemia. Expert Opin. Pharmacother. (2018). 186. Moroz, E., Matoori, S. & Leroux, J.-C. Oral delivery
14, 937–948 (2013). 159. Northfield, S. E. et al. Disulfide-​rich macrocyclic of macromolecular drugs: where we are after almost
134. MacDougall, I. C. et al. Peginesatide for anemia in peptides as templates in drug design. Eur. J. Med. 100 years of attempts. Adv. Drug Deliv. Rev. 101,
patients with chronic kidney disease not receiving Chem. 77, 248–257 (2014). 108–121 (2016).
dialysis. N. Engl. J. Med. 368, 320–332 (2013). 160. Liu, R., Li, X. & Lam, K. S. Combinatorial chemistry in 187. Copolovici, D. M., Langel, K., Eriste, E. & Langel, U.
135. Hermanson, T., Bennett, C. L. & MacDougall, I. C. drug discovery. Curr. Opin. Chem. Biol. 38, 117–126 Cell-​penetrating peptides: design, synthesis, and
Peginesatide for the treatment of anemia due to (2017). applications. ACS Nano 8, 1972–1994 (2014).
chronic kidney disease – an unfulfilled promise. Expert 161. Huo, L. et al. Heterologous expression of bacterial 188. Shi, N.-Q., Qi, X.-R., Xiang, B. & Zhang, Y. A survey
Opin. Drug Saf. 15, 1421–1426 (2016). natural product biosynthetic pathways. Nat. Prod. on “Trojan Horse” peptides: opportunities, issues and
136. Wrighton, N. C. et al. Small peptides as potent Rep. 36, 1412–1436 (2019). controlled entry to “Troy”. J. Control. Rel. 194, 53–70
mimetics of the protein hormone erythropoietin. 162. Walensky, L. D. & Bird, G. H. Hydrocarbon-​stapled (2014).
Science 273, 458–463 (1996). peptides: principles, practice, and progress. J. Med. 189. Staecker, H. et al. Efficacy and safety of AM-111 in the
137. Wrighton, N. et al. Increased potency of an Chem. 57, 6275–6288 (2014). treatment of acute unilateral sudden deafness - a
erythropoietin peptide mimetic through covalent 163. Verdine, G. L. & Hilinski, G. J. Stapled peptides for double-​blind, randomized, placebo-​controlled phase 3
dimerization. Nat. Biotechnol. 15, 1261–1265 intracellular drug targets. Methods Enzymol. 503, study. Otol. Neurotol. 40, 584–594 (2019).
(1997). 3–33 (2012). 190. Hill, M. D. et al. Efficacy and safety of nerinetide for
138. Fan, Q. et al. Preclinical evaluation of Hematide, 164. Cromm, P. M., Spiegel, J. & Grossmann, T. N. the treatment of acute ischaemic stroke (ESCAPE-​NA1):
a novel erythropoiesis stimulating agent, for the Hydrocarbon stapled peptides as modulators of a multicentre, double-​blind, randomised controlled
treatment of anemia. Exp. Hematol. 34, 1303–1311 biological function. ACS Chem. Biol. 10, 1362–1375 trial. Lancet 395, 878–887 (2020).
(2006). (2015). 191. Guidotti, G., Brambilla, L. & Rossi, D. Cell-​penetrating
139. Molineux, G. & Newland, A. Development of 165. Chang, Y. S. et al. Stapled α-​helical peptide drug peptides: from basic research to clinics. Trends
romiplostim for the treatment of patients with chronic development: A potent dual inhibitor of MDM2 and Pharmacol. Sci. 38, 406–424 (2017).
immune thrombocytopenia: from bench to bedside. MDMX for p53-dependent cancer therapy. Proc. Natl 192. Cohen-​Inbar, O. & Zaaroor, M. Glioblastoma
Br. J. Haematol. 150, 9–20 (2010). Acad. Sci. USA 1-10, 10 (2013). multiforme targeted therapy: the chlorotoxin story.
140. Lehmann, A. Ecallantide (DX-88), a plasma kallikrein 166. Carvajal, L. A. et al. Dual inhibition of MDMX and J. Clin. Neurosci. 33, 52–58 (2016).
inhibitor for the treatment of hereditary angioedema MDM2 as a therapeutic strategy in leukemia. 193. Williams, J. A., Day, M. & Heavner, J. E. Ziconotide:
and the prevention of blood loss in on-​pump Sci. Transl. Med. 10, eaao3003 (2018). an update and review. Expert. Opin. Pharmacother. 9,
cardiothoracic surgery. Expert Opin. Biol. Ther. 8, 167. Schmidt, M., Toplak, A., Quaedflieg, P. J. L. M. & 1575–1583 (2008).
1187–1199 (2008). Nuijens, T. Enzyme-​mediated ligation technologies for 194. Bray, B. L. Large-​scale manufacture of peptide
141. Nixon, A. E., Sexton, D. J. & Ladner, R. C. Drugs peptides and proteins. Curr. Opin. Chem. Biol. 38, therapeutics by chemical synthesis. Nat. Rev. Drug
derived from phage display: from candidate 1–7 (2017). Discov. 2, 587–593 (2003).

324 | April 2021 | volume 20 www.nature.com/nrd

 Perspectives

195. Zompra, A. A., Galanis, A. S., Werbitzky, O. & recent advances and future perspectives. Theranostics 238. Pallaghy, P. K., Nielsen, K. J., Craik, D. J. & Norton, R.
Albericio, F. Manufacturing peptides as active 10, 6011–6023 (2020). A common structural motif incorporating a cystine
pharmaceutical ingredients. Future Med. Chem. 1, 218. Skwarczynski, M. & Toth, I. Peptide-​based synthetic knot and a triple-​stranded β-​sheet in toxic and
361–377 (2009). vaccines. Chem. Sci. 7, 842–854 (2016). inhibitory polypeptides. Protein Sci. 3, 1833–1836
196. Mayer, J. P., Zhang, F. & DiMarchi, R. D. Insulin 219. Malonis, R. J., Lai, J. R. & Vergnolle, O. Peptide-​based (1994).
structure and function. Biopolymers 88, 687–713 vaccines: current progress and future challenges. 239. Undheim, E. A. B., Mobli, M. & King, G. F. Toxin
(2007). Chem. Rev. 120, 3210–3229 (2020). structures as evolutionary tools: using conserved
197. Lien, S. & Lowman, H. B. Therapeutic peptides. Trends 220. Busby, R. W. et al. Pharmacologic properties, 3D folds to study the evolution of rapidly evolving
Biotechnol. 21, 556–562 (2003). metabolism, and disposition of linaclotide, a novel peptides. Bioessays 38, 539–548 (2016).
198. Pangalos, M. N., Schechter, L. E. & Hurko, O. therapeutic peptide approved for the treatment of 240. Murray, J. K. et al. Engineering potent and selective
Drug development for CNS disorders: strategies for irritable bowel syndrome with constipation and analogues of GpTx-1, a tarantula venom peptide
balancing risk and reducing attrition. Nat. Rev. chronic idiopathic constipation. J. Pharmacol. antagonist of the Nav1.7 sodium channel. J. Med.
Drug Discov. 6, 521–532 (2007). Exp. Ther. 344, 196–206 (2013). Chem. 58, 2299–2314 (2015).
199. Gruber, C. W., Muttenthaler, M. & Freissmuth, M. 221. Hancock, R. E. W. & Sahl, H.-G. Antimicrobial and 241. Flinspach, M. et al. Insensitivity to pain induced by a
Ligand-​based peptide design and combinatorial host-defense peptides as new anti-​infective therapeutic potent selective closed-​state Nav1.7 inhibitor. Sci. Rep.
peptide libraries to target G protein-​coupled receptors. strategies. Nat. Biotechnol. 24, 1551–1557 (2006). 7, 39662 (2017).
Curr. Pharm. Des. 16, 3071–3088 (2010). 222. Kang, H.-K., Kim, C., Seo, C. H. & Park, Y. The 242. Revell, J. D. et al. Potency optimization of
200. Cragg, G. M. & Newman, D. J. Natural products: therapeutic applications of antimicrobial peptides Huwentoxin-​IV on hNav1.7: a neurotoxin TTX-​S
a continuing source of novel drug leads. Biochim. (AMPs): a patent review. J. Microbiol. 55, 1–12 sodium-​channel antagonist from the venom of the
Biophys. Acta Gen. Subj. 1830, 3670–3695 (2013). (2017). Chinese bird-​eating spider Selenocosmia huwena.
201. Davenport, A. P., Scully, C. C. G., de Graaf, C., 223. Blanes-​Mira, C. et al. A synthetic hexapeptide Peptides 44, 40–46 (2013).
Brown, A. J. H. & Maguire, J. J. Advances in therapeutic (Argireline) with antiwrinkle activity. Int. J. Cosmet. Sci. 243. Schmalhofer, W. A. et al. ProTx-​II, a selective inhibitor
peptides targeting G protein-​coupled receptors. 24, 303–310 (2002). of Nav1.7 sodium channels, blocks action potential
Nat. Rev. Drug Discov. 19, 389–413 (2020). 224. Robinson, L. R. et al. Topical palmitoyl pentapeptide propagation in nociceptors. Mol. Pharmacol. 74,
202. Tsomaia, N. Peptide therapeutics: targeting the provides improvement in photoaged human facial skin. 1476–1484 (2008).
undruggable space. Eur. J. Med. Chem. 94, 459–470 Int. J. Cosmet. Sci. 27, 155–160 (2005). 244. Santos, R. et al. A comprehensive map of molecular
(2015). 225. Pickart, L. The human tri-​peptide glycine-​histidine-lysine drug targets. Nat. Rev. Drug Discov. 16, 19–34
203. Milroy, L.-G., Grossmann, T. N., Hennig, S., Brunsveld, L. and tissue remodeling. J. Biomater. Sci. Polym. Ed. 19, (2017).
& Ottmann, C. Modulators of protein-​protein 969–988 (2008). 245. Yu, F. H., Yarov-​Yarovoy, V., Gutman, G. A. &
interactions. Chem. Rev. 114, 4695–4748 (2014). 226. Du Vigneaud, V., Ressler, C., Swan, J. M., Roberts, C. W. Catterall, W. A. Overview of molecular relationships in
204. Lochhead, J. J. & Thorne, R. G. Intranasal delivery of & Katsoyannis, P. G. The synthesis of oxytocin. J. Am. the voltage-​gated ion channel superfamily. Pharmacol.
biologics to the central nervous system. Adv. Drug Chem. Soc. 76, 3115–3121 (1954). Rev. 57, 387–395 (2005).
Deliv. Rev. 64, 614–628 (2012). 227. Du Vigneaud, V., Ressler, C. & Trippett, S. The sequence 246. Catterall, W. A. Voltage-​gated sodium channels at 60:
205. Kosfeld, M., Heinrichs, M., Zak, P. J., Fischbacher, U. of amino acids in oxytocin, with a proposal for the structure, function and pathophysiology. J. Physiol.
& Fehr, E. Oxytocin increases trust in humans. Nature structure of oxytocin. J. Biol. Chem. 205, 949–957 590, 2577–2589 (2012).
435, 673–676 (2005). (1953). 247. Ahern, C. A., Payandeh, J., Bosmans, F. & Chanda, B.
206. Veening, J. G. & Olivier, B. Intranasal administration 228. Global Information Inc. Global Peptide Therapeutics The hitchhiker’s guide to the voltage-​gated sodium
of oxytocin: behavioral and clinical effects, a review. Sales Market Report 2018. QYResearch, 387893 channel galaxy. J. Gen. Physiol. 147, 1–24 (2016).
Neurosci. Biobehav. Rev. 37, 1445–1465 (2013). (2018). 248. Pan, X. et al. Molecular basis for pore blockade of
207. Leng, G. & Ludwig, M. Intranasal oxytocin: myths and 229. Weinstock-​Guttman, B., Nair, K. V., Glajch, J. L., human Na+ channel Nav1.2 by the μ-​conotoxin KIIIA.
delusions. Biol. Psychiatry 79, 243–250 (2016). Ganguly, T. C. & Kantor, D. Two decades of glatiramer Science 363, 1309–1313 (2019).
208. Walum, H., Waldman, I. D. & Young, L. J. Statistical acetate: from initial discovery to the current 249. Shen, H. et al. Structural basis for the modulation of
and methodological considerations for the development of generics. J. Neurol. Sci. 376, 255–259 voltage-​gated sodium channels by animal toxins.
interpretation of intranasal oxytocin studies. Biol. (2017). Science 362, eaau2596 (2018).
Psychiatry 79, 251–257 (2016). 230. Teitelbaum, D., Meshorer, A., Hirshfeld, T., Arnon, R.
209. Oller-​Salvia, B., Sanchez-​Navarro, M., Giralt, E. & & Sela, M. Suppression of experimental allergic Acknowledgements
Teixido, M. Blood-​brain barrier shuttle peptides: an encephalomyelitis by a synthetic polypeptide. Eur. J. The authors thank K. Woolcock for help with editing the man-
emerging paradigm for brain delivery. Chem. Soc. Rev. Immunol. 1, 242–248 (1971). uscript. M.M. is supported by the European Research Council
45, 4690–4707 (2016). 231. Johnson, K. P. et al. Copolymer 1 reduces relapse under the European Union’s Horizon 2020 research and inno-
210. Chen, Y. & Liu, L. Modern methods for delivery of rate and improves disability in relapsing-​remitting vation programme (714366), by the Australian Research
drugs across the blood-​brain barrier. Adv. Drug Deliv. multiple sclerosis: results of a phase III multicenter, Council (DE150100784 and DP190101667) and by the
Rev. 64, 640–665 (2012). double-​blind placebo-​controlled trial. Neurology 45, Vienna Science and Technology Fund (WWTF; LS18-053).
211. Dockray, G. J. Gastrointestinal hormones and the 1268–1276 (1995). P.F.A., G.F.K. and D.J.A. were supported by Program Grant
dialogue between gut and brain. J. Physiol. 592, 232. Aharoni, R. The mechanism of action of glatiramer APP1072113 from the Australian National Health & Medical
2927–2941 (2014). acetate in multiple sclerosis and beyond. Autoimmun. Research Council (NHMRC) and NHMRC Principal Research
212. Lalatsa, A., Schatzlein, A. G. & Uchegbu, I. F. Rev. 12, 543–553 (2013). Fe l l ows h i p s to G . F. K . ( A P P 11 3 6 8 8 9 ) a n d P. F. A .
Strategies to deliver peptide drugs to the brain. 233. Lalive, P. H. et al. Glatiramer acetate in the treatment (APP1080593).
Mol. Pharm. 11, 1081–1093 (2014). of multiple sclerosis: emerging concepts regarding its
213. Lajoie, J. M. & Shusta, E. V. Targeting receptor-​ mechanism of action. CNS Drugs 25, 401–414 Competing interests
mediated transport for delivery of biologics across the (2011). The authors declare no competing interests.
blood-​brain barrier. Annu. Rev. Pharmacol. Toxicol. 234. Matthews, T. et al. Enfuvirtide: the first therapy to
55, 613–631 (2015). inhibit the entry of HIV-1 into host CD4 lymphocytes. Peer review information
214. Acar, H., Ting, J. M., Srivastava, S., La Belle, J. L. Nat. Rev. Drug Discov. 3, 215–225 (2004). Nature Reviews Drug Discovery thanks J. Mayer and the
& Tirrell, M. V. Molecular engineering solutions for 235. Wild, C., Greenwell, T. & Matthews, T. A synthetic other, anonymous, reviewers for their contribution to the peer
therapeutic peptide delivery. Chem. Soc. Rev. 46, peptide from HIV-1 gp41 is a potent inhibitor of virus-​ review of this work.
6553–6569 (2017). mediated cell-​cell fusion. AIDS Res. Hum. Retroviruses
215. Fani, M., Maecke, H. R. & Okarvi, S. M. Radiolabeled 9, 1051–1053 (1993). Publisher’s note
peptides: valuable tools for the detection and 236. Bruckdorfer, T., Marder, O. & Albericio, F. From Springer Nature remains neutral with regard to jurisdictional
treatment of cancer. Theranostics 2, 481–501 production of peptides in milligram amounts for claims in published maps and institutional affiliations.
(2012). research to multi-​tons quantities for drugs of the
216. Hirayama, M. & Nishimura, Y. The present status future. Curr. Pharm. Biotechnol. 5, 29–43 (2004). Supplementary information
and future prospects of peptide-​based cancer vaccines. 237. Kintzing, J. R. & Cochran, J. R. Engineered knottin Supplementary information is available for this paper at
Int. Immunol. 28, 319–328 (2016). peptides as diagnostics, therapeutics, and drug https://doi.org/10.1038/s41573-020-00135-8.
217. Chen, X., Yang, J., Wang, L. & Liu, B. Personalized delivery vehicles. Curr. Opin. Chem. Biol. 34, 143–150
neoantigen vaccination with synthetic long peptides: (2016). © Springer Nature Limited 2021

NATure RevieWS | DrUG DISCovery volume 20 | April 2021 | 325