Cost-effective production of seamless cyclic peptides from natural or recombinant proteins

Abstract

A method for producing a cyclic peptide, comprising: (i) providing a denatured cyanylated peptide of the formula H2N–(Ai)–(A1)n–(Ad)–(Ac)–(A2)m–COOH in solution, wherein Ai is a terminal amino acid bearing the terminal NH2 group; (A1)n is a peptide linkage containing n independently selected A1 amino acid residues connected to each other by amide bonds, wherein n is 1-28; Ac is a single cysteine with its thiol (SH) group derivatized as a thiocyanide (SCN) group; Ad is an amino acid directly attached to Ac; and (A2)m is a peptide linkage containing m independently selected A2 amino acid residues connected to each other by amide bonds, wherein m is at least 3; (ii) exposing the denatured cyanylated peptide produced in step (i) to a pH in a range of 8-10 to elicit an intramolecular cyanylation- aminolysis reaction between the terminal NH2 group and Ad to produce the cyclic peptide.

Description

COST-EFFECTIVE PRODUCTION OF SEAMLESS CYCLIC PEPTIDES FROMNATURAL OR RECOMBINANT PROTEINSCROSS REFERENCE TO RELATED APPLICATION

[0001] The present application claims benefit of United States Provisional Application No. 63 / 735,415, filed on December 18, 2024 and United States Provisional Application No. 63 / 842,394 filed July 11, 2025, all of the contents of which are incorporated herein by reference.GOVERNMENT SUPPORT

[0002] This invention was made with government support under GM 141459 awarded by the National Institutes of Health. The government has certain rights in the invention.FIELD OF THE INVENTION

[0003] The present invention generally relates to methods of producing cyclic peptides. The present invention is more particularly directed to the production of cyanylated peptides and the conversion of the cyanylated peptides to cyclic peptides.BACKGROUND

[0004] Cyclic peptides are of great interest in pharmacology and increasingly for other biotechnological applications. Cyclization can be achieved via chemical stapling (typically a disulfide or thioether bond) or by seamlessly cyclizing the peptide backbone. The latter approach has distinct advantages in that it yields highly stable, redox-insensitive, yet fully native, safe, and biodegradable product whose chemical properties and conformational propensities are as predictable as possible. Yet, the primary conventional approach to production of industrial-scale (gram and higher) quantities of cyclic peptide remains solidphase synthesis of the linear form with subsequent chemical cyclization. This approach is quite expensive, costing hundreds of dollars per gram of the linear product, with additional costs and yield reductions for the cyclization step. In addition, solid-phase synthesis suffers from the well-known problem that a new synthetic methodology must be developed every time a change in sequence is desired. The method development is laborious, costly, andtime-consuming. A much more cost-effective approach is to produce cyclic peptides by microbial fermentation, which can yield proteins costing less than $0.1 per gram. The sequences of biologically produced peptides can be controlled simply by editing the DNA sequence that encodes them. However, no simple, economical method exists for producing cyclic peptides via microbial fermentation. Cost-effective industrial scale production is necessary to use cyclic peptides in areas beyond pharmacology, such as in agriculture.

[0005] Much bioengineering effort has been expended to solve the challenge of cost- effective biochemical production of cyclic peptides. Most of it has focused on enzymatic cyclization of linear peptides, using enzymes generated by engineering or bioprospecting, such as omniligase and butelase. However, efficiency of the enzymes remains low, and their recombinant production in sufficient quantities to catalyze gram-scale peptide cyclization is very challenging. Another method that has been used to generate seamless cyclic peptides is SICLOPPS, based on split-intein ligation, but this method is also relatively inefficient, has only been applied to produce short cyclic peptides (typically 6-8 residues), and suffers from the difficulty of purifying the cyclic peptide product that is generated within the bacterial cell. None of these methods are amenable for use on a commercial scale, which leaves solid-phase synthesis as the current gold standard. However, as noted above, solid-phase methods have a number of disadvantages, including being laborious, costly, and timeconsuming. There is thus a need in the art for a method that could produce a number of different types of cyclic peptides by more cost-effective and less laborious and time intensive means.SUMMARY

[0006] The present disclosure is foremost directed to a method capable of producing any of a variety of different types of cyclic peptides by more straightforward means and in a more cost effective manner than conventional methods. In brief, the method entails cyanylating a peptide (typically denatured) and subjecting the cyanylated peptide to an alkaline pH (typically in the absence of an enzyme) to elicit an intramolecular cyanylation-aminolysis reaction between the terminal NH2 group and peptide residue attached to and immediately upstream from (i.e., closer to the NH2 terminal group than) the cyanylated residue to result in production of a cyclic peptide. In the method, the produced peptide is excised from theupstream portion of the cyanylated peptide as a result of the bond breakage caused by the intramolecular cyanylation-aminolysis reaction. For purposes of the method, the cyanylated residue is within the interior of the starting peptide and is not a terminal residue, and the cyanylated residue is not part of the final cyclic peptide product. The cyclic peptide (as produced) typically has no more than 20, 25, or 30 amino acid residues.

[0007] More particularly, the method includes the following steps:(i) providing a compound of formula (I):(Ai)-(A1)n-(Ad)-(Ac)-(A2)m-COOH (I) wherein A1is an amino acid comprising a terminal NH2 group, (A1)]! is n independently selected amino acids, wherein n is 1-28, Acis cysteine with its thiol (SH) group derivatized as a thiocyanide (SCN) group, Adis an amino acid, and (A2)mis m independently selected amino acids, wherein m is at least 3; and(ii) exposing the compound of formula (I) to a pH of 8-10 to form a compound of formula II:

[0008] In more particular embodiments, the method more specifically includes the following steps:(i) providing a denatured cyanylated peptide of the formula H2N-(A1)-(A1)n-(Ad)-(Ac)- (A2)m-COOH (I) in solution, wherein A1is a terminal amino acid bearing the terminal NH2 group; (A1)]! is a peptide linkage containing n independently selected A1amino acid residues connected to each other by amide bonds, wherein n is 1-28; Acis a single cysteine with its thiol (SH) group derivatized as a thiocyanide (SCN) group; Adis an amino acid directly attached to Ac; and (A2)mis a peptide linkage containing m independently selected A2amino acid residues connected to each other by amide bonds, wherein m is at least 3; and(ii) exposing the denatured cyanylated peptide produced in step (i) to a pH in a range of 8-10 to elicit an intramolecular cyanylation-aminolysis reaction between the terminal NH2 group and Ad to result in formation of a cyclic peptide of the formula:z(AYA1- Ad(ii) and a linear peptide of the formula:H2N-(Ai)-(A1)n-(Ad)-COOH (III) along with simultaneous production of Y-C(O)-(A2)m-COOH byproduct (IV), wherein Y is an iminothiazole (or more accurately, a thiazolidine-2-imine) ring resulting from intramolecular cyclization between the SCN group and the nitrogen atom of the amide linkage between Adand Ac. In some embodiments, Acis the only cysteine residue in the cyanylated peptide. Note, although Y is known in the biochemical literature as iminothiazole, it is formally a thiazolidine-2-imine. A chemical formula of Y is provided later in this disclosure to ensure clarity. In particular embodiments, the cyclic peptide(s) of Formula (II) is / are within the class of orbitides as known in the art.

[0009] A key advantage of the cyanylation-aminolysis process described herein is that it requires no enzymes and allows for convenient chemically catalyzed peptide cyclization in two steps using only modest quantities of comparatively inexpensive, readily commercially available cyanylating reagent. The methodology of peptide production as a recombinant fusion construct by fermentation, followed by chemical cyclization via cyanylation- aminolysis can be a highly economical method for many peptide sequences. Since cyclization of biomanufactured linear polypeptide precursors involves two chemical steps, it is crucial to develop cost-effective ways of cyclic peptide production. There is no discernible way to achieve such low costs using conventional solid-state synthesis methods, and even enzyme-based cyclizations may struggle to attain it due to the high cost of the enzymes.BRIEF DESCRIPTION OF THE FIGURES

[0010] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

[0011] FIG. 1 is a schematic of an exemplary cyclization process for producing an exemplary seamless N-to-C cyclic peptide from polypeptide precursors of known or defined sequence. The mechanism of the cyclization process is a cyanylation and intramolecular aminolysis process (also referred to as a cyanylation-aminolysis process).

[0012] FIGS. 2a-2c. Production of linear and circular peptides from proteins or peptide- protein fusions via cyanylation-aminolysis chemistry. FIG. 2a is a schematic showing cyanylation-induced cyclization of the / V-terminus of human yD-crystallin, as an example. HPLC peaks of the products, as shown in FIG. 2b, indicate a mixture of peptides that include the cyclic version of peptide 1-17 (shown in the mass spectrograph in FIG. 2c, right) and a competing linear hydrolysis product (shown in the mass spectrograph in FIG. 2c, left).

[0013] FIG. 3 shows the MS / MS fragmentation of the cyclic product. The m / z = 684.34 peak (as also shown in FIG. 2c) reveals a quasi-random ring-opening, wherein the markers show two of the most prominent opening patterns.

[0014] FIG. 4 shows a design of a recombinant peptide -protein fusion construct produced via conventional recombinant expression in E. coli. The construct, which was engineered to include a cysteine residue to its N-terminus along with a His-tag at its C-terminal domain, is useful as a convenient precursor polypeptide for the cyclic peptide / orbitide that can be produced from the precursor by the method described herein. Particular advantages of this construct include its high stability and solubility and lack of any other Cys or Lys residues in its sequence. However, many other precursor constructs are possible on the same general scheme, where the future cyclic peptide is an N-terminal sequence appended to a well- expressed and convenient protein or peptide used as a carrier and to facilitate solubility and purification.

[0015] FIG. 5 is a Coomassie-stained SDS-PAGE gel showing the effect of aminolysis pH and incubation time on cleavage efficiency. Longer aminolysis slightly improved the yield of cyanylation, as evidenced by subsequent Cys-targeted aminolysis or hydrolysis, which results in the “cleaved” product seen on this image. Cleavage at pH 9 had higher yield than at pH 8, which is consistent with aminolysis being the predominant chemical mechanism of cleavage. Higher pH values were not assayed because they favor hydrolytic cleavage.Nonetheless, hydrolytic cleavage can be useful if the goal is to favor production of the peptide in the linear rather than the cyclic form.

[0016] FIG. 6 is a Coomassie-stained SDS-PAGE gel showing the effect of protein concentration and urea in the reaction. The protein concentration and urea did not change the reaction significantly, and the extent of cyanylation, as evidenced by the “cleaved” product, remained approx. 50%, except in the control reactions without cyanylating agent, where there was no cleaved product, as expected. The 3X corresponds to 27.9 pM and IX corresponds to 9.3 pM of the engineered precursor protein schematized in Fig. 4.

[0017] FIG. 7 is a Coomassie-stained SDS-PAGE gel showing the effect of pH, temperature, and concentration of CDAP on the cyanylation reaction. Cyanylation at room temperature was higher than at 4 °C. Increasing CDAP concentration beyond 2 mM did not improve the reaction rate.

[0018] FIG. 8 is a Coomassie-stained SDS-PAGE gel showing the effect of reducing agent TCEP on cyanylation efficiency in the presence and absence of 6M urea. Higher concentration of TCEP in the presence of urea improved the reaction efficiency by preventing dimer formation. In the absence of urea, concentrations of TCEP of approx. 1-3 molar equivalents to the concentration of protein improved the reaction efficiency.

[0019] FIG. 9 is a Coomassie-stained SDS-PAGE gel showing the effect of low CDAP concentration in the presence and absence of 6M urea. Decreasing CDAP concentration up to 100 pM did not decrease the reaction efficiency, but the efficiency dropped substantially below this concentration.

[0020] FIG. 10 shows a design of recombinant peptide -protein fusion constructs to investigate the effects of varying the charge of the residues near the unique Cys. The N- terminal Met residue is removed during protein expression, leaving Gly as the N-terminal residue. The constructs used were as depicted in FIG. 3. For simplicity, only the first few residues of the carrier domain are shown.

[0021] FIG. 11 shows the effect of the sequence of peptide in the vicinity of the Cys residue on cyanylation efficiency. Shown on the vertical axis is the ratio of cleaved to un-cleaved products when using citrate buffer vs. acetate buffer for decreasing pH during thecyanylation step. Data represent mean ± SD from three independent experiments (n = 3). The t-test significance tests between Cl and the other constructs as shown were carried out only for the acetate approach.

[0022] FIGS. 12a- 12b show the effect of the sequence of the peptide in the vicinity of the Cys residue on the relative amounts of cyclic, linear, and modified (acetylated) products. FIG. 12a shows the ratio of cyclic peptide ion current to total assignable ion current (sum of linear and cyclic, unmodified and acetylated forms). The only statistically significant difference was observed between Cl and Cl.7. FIG. 12b shows the ratio of cyclic peptide ion current to linear peptide ion current (both unmodified). For constructs Cl.4 and Cl.7, linear peptide was detected in only one or two of the three replicates, so no statistics could be carried out. The only statistically significant differences were observed between Cl.3 versus Cl.2 and Cl.6. Data represents SD from three independent experiments (n = 3) for the acetate approach.

[0023] FIG. 13 shows an HPLC chromatogram for the cyclic peptide production reaction.

[0024] FIG. 14 is a summary schematic of the most prominent MS / MS fragmentation patterns of the cyclic peptide product from FIG. 13. All possible sites of ring opening are shown as red bars in the top row, and all MS / MS ions with at least five residues and abovethreshold ion intensities are shown in red. Fragments shown in gray were theoretically possible but not detected here.DETAILED DESCRIPTION

[0025] The present disclosure is foremost directed to a method for producing cyclic (cyclized) peptides. The cyclized peptide typically contains no more than 30 amino acid (aa) residues. The aa residues can be selected from among any of the known amino acids. In some embodiments, the amino acids are selected from only the 22 common alpha-amino acids present in most biological systems, while in other embodiments, one or more less common or unnatural amino acids may (or may not) be present. In some embodiments, the cyclized peptide product does not contain a cysteine or a methionine. In some embodiments, the cyclized peptide product does not contain any other type of amino acid containing a SH or SCH3 group (e.g., homocysteine). In different embodiments, thecyclized peptide contains 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 17, 20, 25, or 30 aa residues or a number of aa residues within a range bounded by any two of the foregoing values, e.g., 3-30, 3-25, 3-20, 3-17, 3-15, 3-12, 3-10, 4-30, 4-25, 4-20, 4-17, 4-15, 4-12, 4-10, 5-30, 5-25, 5-20, 5-17, 5-15, 5-12, 5-10, 6-30, 6-25, 6-20, 6-17, 6-15, 6-12, 6-10, 7-30, 7-25, 7-20, 7-17, 7-15, 7-12, 7-10, 8-30, 8-25, 8-20, 8-17, 8-15, 8-12, 10-30, 10-25, 10-20, 10-17, or 10-15.

[0026] In some embodiments, in a first step of the method, there is provided an aqueous solution containing a denatured cyanylated peptide of the following formula:H2N-(Ai)-(A1)n-(Ad)-(Ac)-(A2)m-COOH (I)

[0027] The term “denatured,” as used herein and as well known in the art, means that the peptide of Formula (I) does not possess its natural three-dimensional tertiary or folded morphology while retaining its sequence of amino acid residues connected to each other by amide linkages (depicted as dashed lines in Formula (I) except for bonds to terminal NH2and COOH groups). In the case of functional peptides, such as enzymes, the denaturation results in a loss of function of the peptide. The peptide may be denatured by any of a variety of means well known in the art, including physical or chemical means. Physical means of denaturation include heating and exposure to ionizing radiation. The precursor peptide is denatured before it is cyanylated. In particular embodiments, the peptide (i.e., a precursor peptide) is denatured by exposing it to a denaturing substance, or more particularly, a chaotropic agent, which disrupts the hydrogen bonding or hydrophobic interactions between residues in the peptide. Some examples of chaotropic denaturing agents include guanidinium chloride, urea, thiourea, ethanol, 2-propanol, n-butanol, phenol, and alkyl sulfate salts (e.g., sodium dodecyl sulfate). Such denaturing agents are often contacted with the peptide in the presence of a mild reducing agent, such as tris (2-carboxyethyl) phosphine hydrochloride (TCEP), beta-mercaptoethanol (BME), and dithiothreitol (DTT). Notably, as the precursor peptide is subjected to denaturing conditions, the precursor peptide is not contained in a protective buffer, such as Tris, since the function of protective buffers is to retain the natural three-dimensional tertiary or folded morphology of the peptide.

[0028] The precursor peptide being denatured may be a natural peptide or protein, a peptide produced by chemical or proteolytic digestion of a naturally occurring protein, or a precursor peptide comprising a defined peptide sequence recombinantly fused to a carrierprotein domain. In some embodiments, the carrier protein domain is derived from a protein lacking both Lys and Cys residues. In some embodiments, the protein lacking both Lys and Cys residues is a Cys-free variant of the C-terminal domain of human lens yD-crystallin. In other embodiments, the carrier protein domain bears a peptide tag to facilitate purification, wherein the peptide tag may be, for example, a hexahistidine tag.

[0029] The main function of the carrier domain is to ensure good expression and high solubility, and a secondary function is to facilitate purification, if the domain contains a peptide tag such as hexahistidine. In principle, one could simply fuse the target peptide directly to a hexahistidine tag (with just a Cys residue in-between) and use that tag as both a solubility enhancer and a purification enhancer simultaneously. In the schematic below, the top construct has a tagged carrier domain, and in the bottom construct the tag itself is used as the carrier domain.[Target Peptide] - [Cys] - [Carrier Domain] - [Tag][Target Peptide] - [Cys] - [Tag]

[0030] The term “cyanylated,” as used herein and as well known in the art, refers to the presence of a thiocyanate (-SCN) group in the peptide, wherein the SCN group is typically bound to the methylene side chain linker of cysteine. As well known in the art, an amino acid residue possessing a thiol (SH) can be reacted with a cyanylating agent, such as 1- cyano-4-(dimethylamino)pyridinium tetrafluoroborate (CDAP), ammonium thiocyanate, 2- nitro-5-thiocyanobenzoate (NTCB), 1- cyano- 4-pyrrolidinopyridinium tetrafluoroborate (CPPT or CPIP), 1 -cyano-imidazole (1-CI), 1 -cyanobenzotriazole (1- CBT), p- nitrophenylcyanate, A-cyanotriethylammonium tetrafluoroborate (CTEA), 2- cyanopyridazine-3(2H)one (2-CPO), or trimethylsilyl cyanide (TMSCN) to convert the SH group to a SCN group. The cyanylated cysteine thus possesses a side chain of the formula -CH2SCN (as derived from the natural -CH2SH side chain of cysteine). Typically, the peptide is cyanylated after the peptide has been denatured. The denaturation facilitates the subsequent cyanylation by permitting the cyanylating agent to make contact with one or more thiol-containing residues that would otherwise be protected by the three-dimensional folded structure. The denaturation (specifically, the mild reducing agent) also counteracts the cyanylating agent’s tendency to convert thiol groups into disulfides. The cyanylation istypically conducted at a pH of between 4 and 6 but may be carried out at pH values as low as 3 or as high as 7.

[0031] The variable A1in Formula (I) is an amino acid bearing the terminal NH2 group. A1may be any of the known common 22 amino acids or a less common or unnatural amino acid. In some embodiments, A1is not a cysteine or a methionine or any other type of amino acid containing an SH or SCH3 group (e.g., homocysteine). In some embodiments, A1is a glycine or alanine.

[0032] The variable (A1)]! in Formula (I) is a peptide linkage containing n independently selected A1amino acid residues connected to each other by amide bonds, wherein n is 1-28. In different embodiments, n is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 18, 20, 25, or 28 or a number within a range bounded by any two of the foregoing values, e.g., 1-28, 1-25, 1-20, 1- 18, 1-15, 1-12, 1-10, 1-8, 2-28, 2-25, 2-20, 2-18, 2-15, 2-12, 2-10, 2-8, 3-28, 3-25, 3-20, 3-17, 3-15, 3-12, 3-10, 3-8, 4-28, 4-25, 4-20, 4-17, 4-15, 4-12, 4-10, 4-8, 5-28, 5-25, 5-20, 5-17, 5-15, 5-12, 5-10, 5-8, 6-28, 6-25, 6-20, 6-17, 6-15, 6-12, 6-10, 7-28, 7-25, 7-20, 7-17, 7-15, 7-12, 8-28, 8-25, 8-20, 8-17, 8-15, or 8-12. In some embodiments, the A1amino acids are selected from only the 22 common alpha-amino acids present in most biological systems, while in other embodiments, one or more less common or unnatural amino acids may (or may not) be present. In some embodiments, the A1amino acids do not include a cysteine or a methionine or any other type of amino acid containing an SH or SCH3 group (e.g., homocysteine). In some embodiments, at least one of the A1amino acid residues is a lysine (K), glycine (G), glutamic acid (E), glutamine (Q), tyrosine (Y), or arginine (R), or (Ax)n contains a combination of any two, three, or more of these (i.e., at least one of each amino acid selected).

[0033] The variable Acin Formula (I) is a single cysteine residue with its thiol (SH) group derivatized as a thiocyanate (SCN) group. In some embodiments, Acis attached to a glutamic acid or aspartic acid immediately upstream (closer to the NH2 terminal group) or downstream (closer to the COOH terminal group), wherein the residue immediately downstream from Accorresponds to the residue Ad, as further discussed below. In some embodiments, Acis the only cysteine (or thiol-containing) residue in the cyanylated peptide.

[0034] The variable Adin Formula (I) is an amino acid directly attached to Acand directly downstream of Ac. Typically, Adis selected from only the 22 common alpha-amino acids present in most biological systems, while in other embodiments, one or more less common or unnatural amino acids may (or may not) be present. Typically, Adis not a cysteine or a methionine or any other type of amino acid containing an SH or SCH3 group (e.g., homocysteine). In some embodiments, Adis glutamic acid or aspartic acid (or equivalently, glutamate or aspartate).

[0035] The variable (A2)mis a peptide moiety containing m independently selected A2amino acid residues connected to each other by amide bonds, wherein m is at least or greater than 3, 4, 5, 6, 10, 20, 30, 40, or 50 residues, or a number of residues within a range bounded by any two of the foregoing values. The variable (A2)mcan be considered a carrier peptide (or protein) since, during the process of producing the cyclic peptide, the cyclized peptide of Formula (I) and any linear peptide of Formula (II) is excised from the (A2)mportion, as further discussed below. In some embodiments, the A2amino acids are selected from only the 22 common alpha-amino acids present in most biological systems, while in other embodiments, one or more less common or unnatural amino acids may (or may not) be present. In some embodiments, the A2amino acids do not include a cysteine or a methionine or any other type of amino acid containing an SH or SCH3 group (e.g., homocysteine).

[0036] In a second step of the method, the compound of Formula (I) (more particularly, a denatured cyanylated peptide produced in step (i)) is exposed to a pH in a range of 8-10 (or a pH of precisely or about 8, 8.5, 9, 9.5, or 10, wherein the term “about” may indicate a deviation of ±5, ±2, or ±1) to form a compound of Formula (II), as shown below. The elevated pH elicits or promotes an intramolecular cyanylation-aminolysis reaction between the terminal NH2 group and Adto result in formation of a cyclic peptide of the formula:wherein A1, A1, Adand n are as defined above. All possibilities and examples provided above for each of these variables applies to Formula (II). In different embodiments, n is 1,2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 18, 20, 25, or 28 or a number within a range bounded by any two of the foregoing values, e.g., 1-28, 1-25, 1-20, 1-18, 1-15, 1-12, 1-10, 1-8, 2-28, 2-25, 2- 20, 2-18, 2-15, 2-12, 2-10, 2-8, 3-28, 3-25, 3-20, 3-17, 3-15, 3-12, 3-10, 3-8, 4-28, 4-25, 4- 20, 4-17, 4-15, 4-12, 4-10, 4-8, 5-28, 5-25, 5-20, 5-17, 5-15, 5-12, 5-10, 5-8, 6-28, 6-25, 6- 20, 6-17, 6-15, 6-12, 6-10, 7-28, 7-25, 7-20, 7-17, 7-15, 7-12, 8-28, 8-25, 8-20, 8-17, 8-15, or 8-12. The pH can be raised by addition of a suitable alkaline agent, such as an alkali hydroxide (e.g., NaOH or KOH), bicarbonate, or borate buffer. In some embodiments, the pH is raised to 8.5-9.5. Typically, the pH is raised while the denatured cyanylated peptide is contained in a buffer free of primary, secondary, or tertiary amine groups, or any other group that can serve as an efficient nucleophile in the relevant pH range.

[0037] Due to a competing hydrolysis action, the second step of the method typically also produces a linear peptide of the formula:H2N-(Ai)-(A1)n-(Ad)-C00H (III), typically in a minor amount compared to the cyclic peptide. However, the ratio of the linear and cyclic products may be adjusted by altering the design of the carrier domain or the reaction conditions, such as by increasing pH to 10 or above, or by carrying out aminolysis in a buffer containing a compound that includes an amine group, if producing the linear peptide is also desirable. In Formula (III), A1, A1, Adand n are as defined above. All possibilities and examples provided above for each of these variables applies to Formula (III). The temperature, pH, type of denaturant, and other conditions may affect the relative molar ratios of cyclized to linear peptides.

[0038] Under the present conditions, the cyanylation-aminolysis reaction results in production of an iminothiazole (or more accurately, a thiazolidine-2-imine) ring. In accordance with this principle, the second step of the method also typically results in simultaneous production of Y-C(O)-(A2)m-COOH byproduct (IV), wherein Y is an iminothiazole ring resulting from intramolecular cyclization between the SCN group on Acand the nitrogen atom of the amide linkage between Adand Ac. The cyanylation-aminolysis reaction is shown schematically in FIG. 1 using an exemplary starting (denatured) peptide.

[0039] As shown in FIG. 1, the iminothiazole ring (corresponding to Y) has the following structure:

[0040] Typically, the cyclic peptide (II) is separated from linear peptide (III) and byproduct (IV), such as by a chromatographic method. The chromatographic method may be, for example, high performance liquid chromatography (HPLC), ion-exchange chromatography, hydrophobic interaction chromatography, or size-exclusion chromatography, all of which are well known in the art.

[0041] In some embodiments, the cyclized peptide produced by the method described herein is a cyclic immunosuppressive peptide, such as a cyclosporin peptide. The method may be applied to the production of any seamlessly N-to-C cyclized peptide, provided it does not contain unprotected Cys residues. Possible areas of application include antimicrobial or other therapeutic peptides that naturally occur or are predicted to occur, or that are inspired by such naturally occurring peptides, such as reported by Hostetler et al., ACS Chem. Biol. 16, 2604-2611 (2021). Other potential areas of application include cyclic peptides for pest control or crop protection, as labeling reagents or probes in cellular biology, as purification or selection agents in biochemistry, as diagnostic agents, etc. In some embodiments, the peptides may include an internal proteolytic cleavage site, and / or a unique chemical handle for site-specific labeling or attachment to other molecules. In some embodiments, libraries of cyclic peptides can be generated for subsequent screening or selection. In some embodiments, the peptides are synthesized ribosomally, whereas in other cases peptide precursors may be produced via non-ribosomal synthesis or derived from naturally occurring proteins or peptides. In some embodiments, the cyclic peptides that are produced are intermediates in the biochemical synthesis of more complex cyclic peptides, including N- methylated, isomerized, and other chemically modified versions. In some embodiments, the peptide may be attached to a small molecule or a larger biomolecule. In some embodiments,the peptides may be designed to be membrane-permeable, as reported by Bhardwaj et al., Cell 185, 3520-3532 (2022).

[0042] In some embodiments, the cyclic peptide(s) produced by the method described herein is / are within the class of orbitides as well known in the art. Known as orbitides or circular bacteriocins, such compounds are ribosomally synthesized, post-translationally modified, and enzymatically cyclized by plants and bacteria. As well known, orbitides have no disulfides, only canonical amino acids, and seamless peptide backbones. Orbitides are described in, for example, M. F. Fisher et al., Journal of Natural Products, 82(8), 2152-2158 (2019), the contents of which are herein incorporated by reference. Orbitides typically contain 5-16 amino acid residues within their cyclic structure, i.e., n in (Ax)nin the cyclic peptide of Formula (I) may be, for example, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, or 14, or n is within a range bounded by any two of the foregoing values, e.g., 3-14, 4-14, 5-14, 6-14, 7- 14, 8-14, 3-10, 4-10, 5-10, 6-10, 7-10, 8-10, 3-8, 4-8, 5-8, 3-6, 4-6, or 3-5.

[0043] In one embodiment, the method includes adding linkers to the terminal groups of the cyanylated peptide. In another embodiment, the method excludes adding linkers to the terminal groups of the cyanylated peptide. In an embodiment, the method includes one or more mutations in the cyanylated peptide, such as to remove cysteine and / or methionine residues, while in another embodiment, the cyanylated peptide does not include such mutations. In an embodiment, the method employs an enzyme to facilitate the cyclization process, while in another embodiment, an enzyme is excluded.

[0044] Examples have been set forth below for the purpose of illustration and to describe certain specific embodiments of the invention. However, the scope of this invention is not to be in any way limited by the examples set forth herein.EXAMPLES

[0045] Biochemical Production of Cyclic Peptides Containing at Least 17 Residues

[0046] The following experiment demonstrates the biochemical production of cyclic peptides of at least 17 residues using the presently described method. As an example, as shown in FIGS. 2a-2c, wild-type human yD-crystallin was denatured, chemically cyanylated on its Cys residues, and cyclized by intramolecular aminolysis. The free aminegroup at the N-terminus attacks the peptide bond upstream of the nearest cyano-Cys, yielding a cyclic peptide, as confirmed by LC / MS / MS, as shown in FIGS. 2b, 2c, and 3). A cyclic peptide may open at theoretically any position during collision-induced dissociation at the MS / MS step; however, as FIG. 3 shows, some opening sites are preferred, including both the newly-formed peptide bond joining E and G and the bond joining K and I. In the latter case, the formerly N-terminal GK sequence becomes C-terminal. In FIGS. 2a-2c and FIG. 3, the sequence of the peptide can be arbitrary, provided it does not contain any other Cys residues besides the one shown. The peptide is attached to the N-terminus of a protein (represented by “ - ” in the figures). This can be either a naturally abundant protein(in the particular experiment shown in FIGS. 2a- 2c and 3, this protein is human lens yD- crystallin) or a recombinant carrier protein domain optimized for high-yield expression and simple purification of the peptide-protein fusion construct. FIGS. 2b, 2c, and 3 demonstrate that only two major products result from the chemical treatment: a linear peptide product and a circular (macrocyclic) peptide product, produced in the same reaction. The relative yield of linear to circular products is likely to be controllable by setting reaction conditions. The two major products are readily separated by chromatography (FIGS. 2b and 2c), and their identity was confirmed by 2-dimensional mass spectrometry (FIG. 3). Significantly, the circular (cyclic) peptides produced by this approach are seamless, i.e., they are held together entirely by peptide bonds all around the ring, without any other chemical linkages or “staples.”

[0047] The following methodology was used to produce the cyclic peptide described above:

[0048] Step 1. The protein was incubated with denaturant as needed (guanidinium chloride, urea, etc.) and a mild reducing agent to eliminate any undesirable tertiary structure and disulfide bonds.

[0049] Step 2. The sample from Step 1 was buffer-exchanged (by size-exclusion or ultrafiltration) into buffer without reducing agent and concentrated to at least 10 pM, preferably 100 pM.

[0050] Step 3. A stock solution of cyanylating agent (such as l-Cyano-4- dimethylaminopyridinium tetrafluoroborate, CDAP) was prepared in non-aqueous solvent such as neat acetonitrile and stored under inert gas such as argon at -20 °C.

[0051] Step 4. Sample from Step 2 was incubated with 1-10 mM cyanylating agent from Step 3 in buffer at pH between 4 and 5.5 for 2 hours.

[0052] Step 5. Increasing the pH to between 8 and 9 in amine-free buffer yields two products in proportions determined by experimental conditions (temperature, pH, denaturant, etc.): an unmodified linear peptide from the A-terminus to the residue immediately upstream of the first Cys residues in the protein’s sequence, and the circular (cyclic) product comprised of the same residues as the linear product but with a seamlessly cyclized peptide backbone. In cases where the protein’s sequence contains more than one Cys residue, multiple linear peptides may be produced, but only the A-terminal one will be unmodified, and any seamless cyclic peptides must include the A-terminus.

[0053] Step 6. The linear and circular products may be conveniently separated by conventional HPLC or other chromatographic method.

[0054] Step 7. Steps 1-5 describe generation of circular peptides from a native protein sequence. Circular peptides of arbitrary sequence may be generated from recombinantly expressed protein constructs consisting of the peptide sequence at the A-terminus followed by a single Cys residue, followed by a solubility tag that can be a stable peptide or protein domain that facilitates purification. Preferably, the tag or domain should not contain any Cys or Lys residues to avoid interfering with the reaction.

[0055] Step 8. A convenient solubility tag for step 7 is the isolated C-terminal domain of human yD-crystallin with both its Cys residues replaced with Ser residues. The peptide sequence to be cyclized is attached to the A-terminus of this domain, and a hexahistidine tag is attached to its C-terminus. The resulting construct has no Lys residues (except any that are in the peptide) and no Cys residues except the one between the peptide and the crystallin domain.

[0056] Production of Orbitides

[0057] The following work describes a simple method for producing orbitides from naturally abundant proteins or from recombinantly expressed precursor polypeptides. (Note that the term “orbitide” can include any peptide whose backbone is seamlessly circular, i.e., without termini.) The reaction proceeds under mild aqueous conditions, without the need for enzymes, and using only one chemical reagent, which is readily available. This work demonstrates production of a 17-mer cyclic peptide from a wild- type human eye lens y- crystallin and of a set of 10-residue cyclic peptides from recombinantly expressed polypeptide precursors. The effects of reaction conditions and sequence changes on reaction efficiency were investigated, along with identifying the products by their complex mass spectrometry fragmentation patterns, and chromatographically separating linear and cyclic peptide forms. The described methodology paves the way to large-scale, cost-effective production of stable yet biodegradable, easily designable cyclic peptides for applications not only in medicine, but in such areas as biotechnology, materials, agriculture, and pest control. The methodology may also enable production of diverse orbitide libraries from arbitrarily chosen natural protein sources.

[0058] Herein is described a simple method for enzyme-free biochemical production of seamlessly N-to-C cyclized peptides from recombinantly generated designed linear polypeptide precursors or even from naturally occurring proteins not evolved for this purpose. Cyclization was achieved by modifying the methodology initially developed for Cys-targeted protein cleavage (Gary R Jacobson et al., 1973; Serebryany et al., 2023; Wu & Watson, 1997) to achieve peptide cyclization instead. The present methodology entails only two chemical steps: (1) cyanylation of a Cys residue between the N-terminal peptide and the rest of protein molecule, and (2) intramolecular aminolysis via nucleophilic attack of the free amine group of the peptide N-terminus on the peptide bond immediately upstream of the cyanylated Cys residue, which serves as the leaving group (FIG. 1). Despite its simplicity, the potential of this type of chemistry has been largely overlooked. Although this methodology has been used successfully to fuse a protein’s termini together (Takahashi et al., Journal of Biological Chemistry, 282(13), 9420-9429, 2007), this methodology has not been used to excise orbitides or other cyclic peptide products from linear proteinprecursors. Producing protein precursors in bacteria and obviating the need for enzymes in the cyclization process maximizes robustness of our method and minimizes cyclic peptide production costs.

[0059] The present work demonstrates the applicability of this facile, enzyme-free approach to generating orbitides either from naturally occurring proteins or from recombinant synthetic constructs. The present work demonstrates production of a 17-mer cyclic peptide from the N-terminus of wild-type human lens yD-crystallin (HyD) with the sequence c(GKITLYEDRGFQGRHYE). The present work further demonstrates production of a set of 10-mer cyclic peptides, sequence c(GKFISREFHX), where X = A, R, or E, from a recombinant fusion construct, in which the desired sequence is fused to the N-terminus of a small protein domain acting as a carrier. The resulting head-to-tail cyclized peptide macrocycles were identified by mass spectrometry, which confirmed cyclization by revealing diverse fragmentation patterns arising from the same cyclized precursor sequence. Finally, this work explores the effects of varying the residues upstream and downstream of the cyclization site on the efficiency of initial cyanylation and final cyclization, with implications for the generalizability and yield optimization of this method.

[0060] Results

[0061] Cyclic peptide production from wild-type HyD crystallin

[0062] To validate the possibility of this approach on a natural protein, wild-type Human yD-crystallin (HyD) was used since it has a cysteine residue at position 18 in the mature form (FIG. 2a). This protein was unfolded, at 1.9 mg / mL, overnight at 37 °C in 4M guanidinium chloride (GdnHCl) pH7 and 1 mM Tris(2-carboxyethyl)phosphine (TCEP) (final concentrations) to ensure a fully unfolded structure and fully reduced Cys residues. The unfolded protein was desalted into 4M GdnHCl pH 3 without TCEP. Cyanylation was done by adding l-Cyano-4-dimethylaminopyridinium (CDAP) (2mM final concentration) and incubating at room temperature for 3.5 hours. After the cyanylation step, the reaction was divided and desalted into three buffers (A: IM Phosphate buffer, pH 7; B: IM Phosphate buffer, pH 8; C: IM Carbonate buffer, pH 9). For each buffer, the aminolysis reaction was done by incubating at 4 °C vs. room temperature overnight. The samples were quenched byadding 0.1% formic acid and were analyzed by mass spectrometry. According to the mass spectrometry analysis, the best cyclization yield was observed when the aminolysis step was done at pH 9, 4 °C overnight, which yielded linear and circular products in approximately 1: 1 ratio based on total ion current (FIG. 2b). The molecular mass of the cyclic product differs from that of the linear product by the mass of one water molecule (FIG. 2c). LC / MS / MS was used to make sure the product was cyclic rather than a dehydrated linear peptide (FIG. 2d). Unlike the linear peptide, which fragments only from the ends, the ring of an orbitide can open at multiple positions, some of which may be preferred due to secondary structure or sequence features. Indeed, inspection of the complex MS / MS pattern of the cyclic product readily revealed two major fragmentation patterns (FIG. 2d). Notably, residues 1-18 form most of a P-hairpin in the native protein structure (Basak et al., Journal of Molecular Biology, 328(5), 1137-1147, 2003), and simulations have shown this hairpin to be exceptionally stable even in the absence of cyclization (Serebryany et al., Journal of Biological Chemistry, 291(36), 19172-19183, 2016). Accordingly, in the MS / MS experiments, the peptide appears to open preferentially on either side of the newly formed peptide bond, which probably strains the hairpin structure.

[0063] Design of a Cys- and Lys-free carrier domain for peptide-protein fusions

[0064] Encouraged by the observation of cyclic peptide production from wild-type HyD crystallin, the present work engineered its native C-terminal domain to attach a short peptide plus a cysteine residue to its N-terminus (FIG. 4). This crystallin domain is a particularly appropriate carrier for the short peptide due to its high stability and absence of Lys residues in its sequence (Mills-Henry et al., Journal of Biological Chemistry, 291(36), 19172-19183. 2019). The present work also genetically removed both its native Cys residues, via the C108S / C110S substitution and appended a His6-tag to the C-terminus of the construct to facilitate affinity purification. The peptide-protein fusion construct was produced via conventional recombinant expression in E. coli. The short peptide’s sequence was chosen arbitrarily. The sequence of a previously reported a-crystallin-derived mini-chaperone peptide FISREFHR (Raju et al., Biochimica et Biophysica Acta (BBA )-General Subjects, 1860(1), 246-251 2016) was chosen, adding GK to its N-terminus to facilitate bacterial expression.

[0065] Biochemical production of a cyclized mini-chaperone peptide

[0066] The purified construct containing the mini-chaperone peptide at the N-terminus was subjected to cyanylation and aminolysis similarly to the WT HyD crystallin above.However, since the recombinantly added peptide sequence was not part of the protein core, it was herein found that protein denaturation was optional in this case. The pH was decreased by adding acetate buffer, CDAP was used as the cyanylating agent, and the cyanylation-aminolysis reaction was performed as described herein. The mixture was then analyzed by SDS-PAGE to confirm and quantify overall reaction yield, and by MS to confirm production of the new orbitide and determine the ratio of linear to cyclic products. Several reaction parameters were systematically varied to optimize the efficiency of both reaction steps (cyanylation and aminolysis).

[0067] Optimization of the cyanylation and aminolysis reactions

[0068] The overall efficiency of the cyanylation-aminolysis reaction can be readily monitored and quantified by SDS-PAGE, comparing the intensity of the cleaved vs. uncleaved product. Cleavage can proceed either by aminolysis or by hydrolysis; however, both require cyanylation (G. R. Jacobson et al., Journal of Biological Chemistry, 248(19), 6583-6591 1973; Wu & Watson, Protein Science, 6(2), 391-398 1997). Therefore, the present work first optimized the conditions for aminolysis, keeping cyanylation conditions constant. In this case, variations in the cleaved / uncleaved ratio indicate variations in cleavage efficiency. Then, the highly efficient cleavage conditions were kept constant, and the conditions of cyanylation were varied. In this case, variations in the cleaved / uncleaved ratio indicate variations in cyanylation efficiency.

[0069] FIG. 5 is a Coomassie-stained SDS-PAGE gel showing the effect of aminolysis pH and incubation time on cleavage efficiency. Longer aminolysis slightly improved the yield of cyanylation, as evidenced by subsequent Cys-targeted aminolysis or hydrolysis, which results in the “cleaved” product seen on this image. Cleavage at pH 9 had higher yield than at pH 8, which is consistent with aminolysis being the predominant chemical mechanism of cleavage. Higher pH values were not assayed because they favor hydrolytic cleavage. Nonetheless, hydrolytic cleavage can be useful if the goal is to favor production of thepeptide in the linear rather than the cyclic form. Initial optimization of aminolysis conditions (FIG. 5) confirmed that the reaction proceeded somewhat better at pH 9, while prolonged incubation had limited effect.

[0070] In next experiments, the conditions of the cyanylation step were varied, keeping the pH 9 overnight aminolysis as a constant. These experiments investigated whether urea could improve the reaction rate by increasing structural accessibility of the unique Cys residue (FIG. 6). Based on cleavage ratios, urea was not necessary for the cyanylation of this recombinant precursor protein, nor did protein concentration have any effect, except that high protein concentration led to aggregation. Next, we optimized the pH, temperature of cyanylation, and concentration of CDAP was optimized (FIG. 7). Incubation at room temperature improved the reaction compared to 4 °C. Increasing CDAP concentration from the initial protocol did not have any effect (all concentrations tested in this figure were much higher than the protein concentration). In fact, very high [CDAP] (50 mM) apparently inhibited the reaction, although protein aggregation due to exposure to such a high concentration of acetonitrile (the solvent for CDAP) cannot be ruled out. However, formation of dimers by SDS-PAGE was observed (FIG. 7). It may be hypothesized that the free thiol group of reduced Cys could attack the cyanylated Cys residue, leading to formation of a disulfide bond and loss of cyanylation. Reducing these dimers may therefore give the Cys residues another chance to be cyanylated. Indeed, low concentrations of the reducing agent TCEP improved reaction efficiency (FIG. 8). Conversely, other experiments sought to determine the minimum concentration of CDAP needed for the reaction (FIG. 9). The minimum concentration without changing the reaction rate was lOOpM, or ~3.5 equivalents.

[0071] Investigating the effect of peptide composition on cyanylation and cyclization efficiency

[0072] To investigate the effect of charge and size of residues in the vicinity of the cysteine on cyanylation and cyclization efficiency, 7 constructs were designed based on the minichaperone sequence (FIG. 10). The charge may induce electrostatic interaction between the first and last amino acids in the peptide, and it may change the reaction rate. The size of amino acids next to cysteine may hinder the attacking N-terminus amino acid. For instance,Cl.1 has neutral amino acids, and Cl.5 has negatively charged amino acids. These seven constructs were cloned and purified and compared with the original construct. The same protocol was applied to all the eight constructs and cyanylation efficiency was measured by SDS-PAGE. The experiments were done in triplicate and bands were quantified (FIG. 11). Interestingly, Cl.5, which has two negatively charged amino acids, had the lowest cyanylation efficiency. This might be due to the efficacy of cyanylation of the cysteine when there are two electron-rich amino acids next to it.

[0073] To better understand the effects of the peptide and carrier sequence on cyclization efficiency, the reaction of the eight variants was analyzed by mass spectrometry and assigned ratios based on MSI ion intensities. The results show relatively modest effects of the sequence on small differences in terms of produced peptides as the result of the sequence (FIGS. 12a and 12b). For instance, Cl.7 had lower cyclic peptide compared to Cl in the first analysis. Furthermore, by using another analysis approach, it was herein found that Cl.3 had lower cyclic peptide products compared to Cl.2 and Cl.6. Finally, the experiments detected citrate-conjugated and acetate-conjugated peptides, respectively, when these buffers were used to change pH during the reaction. However, citrate appeared to interfere with the circularization reaction to a greater extent by blocking the N-terminus.Therefore, acetate buffer was chosen for future scale-up experiments.

[0074] Purification of cyclic peptides using HPLC

[0075] Preparative HPLC was used to separate cyclic peptides from the mixture. A 5 mL reaction using Cl.3 was prepared for this experiment. Peptides were purified using RP-8 HPLC on a C18 column with a linear acetonitrile gradient (FIG. 13) as discussed in the methods. The cyclic peptide eluted at 21 minutes in approx. 40% acetonitrile concentration. The identity of the peak was confirmed by mass spectrometry.

[0076] The HPLC peak labeled “Cyclic” in FIG. 13 was collected and analyzed by LC / MS / MS on a ThermoFisher™ Orbitrap Eclipse instrument in positive mode with HCD fragmentation. As previously observed for the HyD-crystallin derived peptide in FIG. 2, both the 3+ and 2+ precursor ions corresponding to the expected m / z of the cyclic product yielded highly complex fragmentation patterns. These patterns were consistent with ring-opening at multiple possible sites, followed by fragmentation predominantly from the C- terminus, leading to a preponderance of B-ions. The fragmentation patterns were analyzed manually using commercial software, setting empirical thresholds for MS / MS peak intensities and assigning them by inspection to the set of all possible circular permutants of the “GKFISREFHA” sequence. The findings are summarized in FIG. 14. Similarly to FIG. 2, ring-opening appeared to proceed preferentially within two peptide bonds of the Lys residue. Some of the MS / MS ions assigned in FIG. 14 could have been produced instead as Y-ion fragments of other assigned MS / MS ions; however, they were assigned in these series by parsimony and, to some extent, by similarity of ion current intensities. These choices do not affect the basic conclusion of the assignment: the orbitide was successfully produced, and it opened stochastically at multiple places during MS / MS fragmentation.

[0077] Assignments of the fragmentation patterns in FIG. 14 were made primarily from the 2+ charged precursor at m / z = 587.31, which is an exact match for the precursor expected for the unmodified cyclic peptide. Both fragment ions and neutral loss ions were assigned. Some additional assignments were made from the 3+ precursor at m / z = 391.69, which is close but not exact relative to the expected 391.88 (which was nonetheless observed in this precursor’s MS / MS spectra). The experiments also examined MS / MS spectra for the expected precursor ions for linear, linear-acetylated, and cyclic-acetylated products.

[0078] Aside from the 2+ charged precursor at m / z = 587.31, no other exact-match precursors were auto-selected for MS / MS. However, traces of the other reaction products were assigned by examining the next-closest MSI precursor m / z’s. Thus, the 3+ charged precursor at m / z = 397.99 was close to the expected m / z of 397.88 at z = 3 for the linear peptide product. MS / MS spectra of this precursor revealed several fragments assignable to fragmentation of the linear product from the N-terminus, in the pattern GK|F|I|S|REFHA, where | indicates observed bond cleavage. Traces of the linear product were expected due to the close spacing of the two HPLC peaks (FIG. 13) and the high sensitivity of mass spectrometry. Even these MS / MS spectra, however, were dominated by peaks assignable only to the cyclic product, due to multiple circular permutations of the GKFISREFHA sequence. Lys acetylation produces a 42.01 Da. shift in the peptide mass. From the 2+ charged precursor at m / z = 608.13 (vs. 608.32 expected for the acetylated cyclic product),10 fragment ions were identified consistent with four distinct opening sites of the orbitide acetylated at the Lys residue. The four ring opening sites observed for the acetylated product were A|G, I|S, F|H, and H|A. All four were among the six opening sites observed for the unmodified product in FIG. 14. It is noteworthy that the cyclic product can only be acetylated at the Lys, since it lacks a free N-terminal amine. Accordingly, no B-ions were observed from the K|F ring-opening site (which was detected confidently for the unmodified product), as any C-terminal fragmentation of this ring-opening product would lead to loss of the acetyl-Lys.

[0079] Discussion

[0080] The ease and speed of the presently described protocol makes it useful for both small-scale and large-scale production of seamlessly head-to-tail cyclized peptides with only basic equipment and expertise. This method also facilitates quick substitution of amino acids without substantial modification of the protocol. Thus, this permits quick in-house production and engineering of cyclic peptides. Moreover, production of cyclic peptides from recombinant proteins reduces the release of toxic waste compared to solid-phase synthesis. Cyanylation-aminolysis chemistry occurs readily and specifically in water under mild conditions and can therefore be considered an example of “Click” chemistry (Kolb et al., Angewandte Chemie- International Edition, 40(11), 2004-+.2001). In contrast, production of peptides by solid-phase synthesis requires using large amounts of environmentally problematic solvents (Martin et al., RSC advances, 10(69), 42457-42492 2020). Moreover, engineering cyclic peptides from natural amino acids for therapeutic targets is promising since it does not produce toxic byproducts after degradation inside the human body. Additionally, production of peptide from genetically-encoded material will accelerate sharing these peptides between laboratories since the DNA sequences and plasmids can be easily shared.

[0081] Production of cyclic peptides recombinantly limits incorporation of non-natural amino acids into cyclic peptides since non-natural amino acids require their own tRNA (Jin et al., Applied microbiology and biotechnology, 103, 2947-2958 2019). Therefore, it is best suited for production of compounds that mimic naturally occurring ribosomally synthesizedcyclic peptides, known as orbitides, bacteriocins, etc. However, recombinant precursor proteins could also be produced via cell-free expression, which allows for incorporation of a wider range of non-natural amino acids (Gao et al., Frontiers in pharmacology, 10, 611 2019), although it also increases the cost.

[0082] It has herein been demonstrated that seamlessly head-to-tail cyclized peptides (“orbitides”) can be excised from proteins via cyanylation-aminolysis. The objective was only to prove feasibility, so the final purity and yield of the products were not rigorously calculated. However, the overall yield can be estimated from the apparent yields of the two steps of the reaction. Cyanylation efficiency ranged from -20-50% (FIG. 11) in acetate buffer, with potentially higher efficiencies in citrate buffer but at the cost of high variability and adduct formation. The four reaction products were assigned based on MSI precursor ions: linear, cyclic, and acetylated versions of either. Yields of unmodified cyclic products relative to all assignable peaks averaged -20% (FIG. 12a), for an overall estimated yield of 4-10%. This would make the yields from the present approach comparable to those of traditional cyclization reactions used in solid-phase synthesis. It may be an underestimate of the true yield because cyclic peptides lack the N-terminal free amine group and may therefore be less ionizable than linear products of the same sequence. Moreover, the observed ratios of cyclic to linear product ions varied based on the residues in the immediate vicinity of the Cys residue (FIG. 12b). The observed variation suggests that small residues near the Cys promote cyclization, perhaps by ensuring good accessibility of the peptide bond to be transferred. Conversely, negatively charged (Glu) residues, particularly after the Cys, may inhibit cyclization and favor the linear hydrolysis product. These observations open the way to further optimization of the reaction chemistry to boost the overall yields when using recombinant precursors.

[0083] Producing comparable quantities of linear and cyclic peptides is a useful feature at the research stage, since in many applications it is of interest to compare the performance of the linear and cyclic versions of the same peptide sequence. In this case, both are produced in one pot. For optimal scale-up, however, it will be important to further improve cyanylation chemistry, rigorously quantify the overall reaction yields, further optimize the aminolysis conditions, and engineer improved carrier proteins for maximum expression,solubility, and suitability for the reaction. It will also be important to investigate how a peptide’s length and structure affect cyclization yields. The cyclic 17-mer (FIG. 2b) and 10-mers (FIG. 12) were produced with comparable cyclic:linear ratios; however, the 17- mer’s P-hairpin propensity may have facilitated this. Cyclization methods in general may favor hairpin-forming sequences due to greater average proximity of their termini.

[0084] Generation of acetylated side products was a surprise given the reaction composition. Non-enzymatic pathways of Lys acetylation are known (Baldensperger & Glomb, Front Cell Dev Biol, 9, 664553, 2021), but they involve compounds such as acetyl phosphate. This was unlikely to form in the present reaction, since all proteins were desalted out of phosphate buffer by size exclusion. It remains to be explored whether reaction conditions that promote Cys cyanylation also promote Lys acetylation.

[0085] An intriguing future possibility opened by the present approach is to generate large libraries of seamlessly cyclic peptides at very low cost by excising them directly from natural proteomes. Natural proteome-derived linear peptide libraries have already been widely used, primarily for characterizing specificities of novel proteases or other peptide modification enzymes and reagents. In at least one case, a proteome-derived linear peptide library was directly used to screen for binders of protein targets (Maaty & Weis, Journal of the American Chemical Society, 138(4), 1335-13432016). Applying the present strategy to this problem promises highly scalable inexpensive generation of peptide macrocycles of a wide variety of sizes and structures in libraries whose sequence composition is derived from natural proteomes and may therefore be more likely to yield biologically available and active initial hits. The hits may subsequently be optimized using focused, designed libraries. By choosing which protein or proteomic samples to use as starting material and whether and how to digest them beforehand, researchers have a meaningful degree of control over the final composition of the cyclic peptide library thus produced. Initial screening for biological activity may then be analogous to the long-established drug discovery approach of screening plant extracts for desired bioactivity, then fractionating the extracts to isolate the active ingredient. In essence, the present approach provides a way to produce orbitides from arbitrarily chosen mixtures of natural protein feedstocks instead of having to extract them from the sap of trees.

[0086] Methods

[0087] Protein expression and purification:

[0088] WT HyD Protein expression and purification were performed as previously described by Serebryany et al. {Journal of Biological Chemistry, 290(18), 11491-11503 2015).

[0089] The protein-fusion constructs were cloned, expressed and purified as follows. The backbone of pET16b plasmid containing sGFP was linearized by PCR. The gene of sGFP was omitted and synthesized gene containing homology region to the backbone was used for in vivo assembly. The bacterial colonies were plated on Ampicillin plates, and their plasmid were sequenced by GENEWIZ. pET16b plasmids containing the genes of variants were transformed into BL21(DE3) and selected with lOOpg / mL Ampicillin. Single colonies were used to inoculate 10 mL TB (Casein digest peptone 12g / L, Yeast extract 24g / L, Dipotassium Phosphate 9.5g / L, Monopotassium Phosphate 2.2g / L) in 50 mL conical flask which were cultured overnight at 37 C, 220 rpm overnight. Overnight cultures were used to inoculate main cultures (1 to 100). Cultures were grown under 37 °C and 220 rpm until reaching OD ~ 0.6. Then the expression was induced by adding IPTG to the final concentration of ImM. The cultures were induced overnight at 16 °C, and harvested the following day by centrifugation at 4000 g for 40 minutes at 4 °C. Bacterial pellets were resuspended in binding buffer (50 mM sodium phosphate, 300 mM sodium chloride, 10 mM imidazole; pH 7.4) and protease inhibitor (commercial mini tablets, one tablet per 600 mL culture) was added. The bacterial cultures were lysed by 30 minutes incubation with lysozyme (0.1 mg per 600 mL of culture) and sonication (commercial sonicator with 1 / 8 inch probe, 70% amplitude, 2s pulse, 2s rest for 5 minutes total sonication time) on ice. Next, the suspension was centrifugated (40000 g, 30 minutes, 4 °C) to separate bacterial debris from soluble proteins. The supernatant was injected into a HisPur™ Cobalt chromatography (Thermo Fisher) column and was purified by elution buffer (50 mM sodium phosphate, 300 mM sodium chloride, 150 mM imidazole; pH 7.4). The eluted proteins after this step were incubated with 5 mM TCEP for Ih to reduce any possible disulfide bonds. The purified proteins from affinity chromatography were applied to size exclusion chromatography(Superdex™ 75). The proteins were then purified in size-exclusion buffer (10 mM PIPES, 50 mM NaCl, pH 6.7). Purified proteins were then aliquoted and stored at -70°C.

[0090] Cyanylation and aminolysis

[0091] For the cyanylation step, the pH was decreased to 4.5 either by adding citrate buffer (IM, pH 4) or acetate buffer (IM, pH 4). Protein concentration was 27 pM. Then CDAP from the stock (400 mM, dissolved in acetonitrile) was added to the reaction (final concentration of 10 mM) and incubated at room temperature for two hours. The pH was then increased to 9 by adding carbonate buffer (IM, pH 9.2) to the reaction for the aminolysis step. The reaction then was incubated at 4 °C for 18 hours. The reaction was then quenched by adding 0.1% formic acid.

[0092] SDS-PAGE and quantification

[0093] The purity of protein purification was assessed by 12% SDS-PAGE using precast BioRad Criterion™ gels and GelCode™ Coomassie stain (ThermoFisher). The rate of cleavage of the constructs was assessed by running 16% SDS-PAGE and fixing the gels before staining according to the manufacturer’s protocol. Quantification of bands was done by commercial software.

[0094] ID mass spectrometry for measuring cyclization

[0095] The samples after completion of the reaction were quenched by 0.1 % formic acid. About 40 pL of each sample at 20 pM concentration was applied to the column (Acquity™ UPLC protein BEH C4 column 2.1 x 50 mm), peptides were separated using a l l min gradient at 300pL / min by 5%, 6.5 min at 40%, 7 min at 98%, hold for 8.75 min, 9 min at 5%, and hold for 11 min. Samples were injected via an autosampler that was kept at 4°C, and the column was kept at 40°C. Eluting peptides were analyzed by mass spectrometry, positive polarity, and acquired over 110-2000 m / z at 0.2 scans / sec. Lockspray data acquired on LeuEnk peptide every 10 sec and 3 scans averaged to apply real-time correction.

[0096] Preparative HPLC

[0097] The mixture was filtered by 0.2 m membrane and purified using reverse-phase high-performance liquid chromatography (RP-HPLC) on a Cl 8 column (150 mm x 4.6 mm, 5 pm particle size, 100 A pore size). The separation was performed on a Teledyne ISCO ACCQPrep™ HP 150 chromatograph, with UV detector set to 214 nm and 254 nm. Mobile phases were: water with 0.1% formic acid (solvent A) and acetonitrile with 0.1% formic acid (solvent B). Cyclic peptides were purified using a linear gradient from 0% to 100% solvent B for 40 minutes at a flow rate of 1.0 mL / min. The column temperature was maintained at 25 °C during the run. Fractions were collected, concentrated and stored at - 20°C until further analysis.

[0098] 2D mass spectrometry for detecting orbitide fragmentation patterns

[0099] Data in FIG. 3 were collected on a ThermoFisher™ qExactive HF+ mass spectrometer in positive ion mode with a standard Ih acetonitrile gradient on a commercial C 18 column. The sample was extensively desalted using a C 18 desalting cartridge prior to loading to remove GdnHCl. MS / MS data summarized in FIG. 14 were collected on a commercial mass spectrometer, with a commercial C18 column and a Ih acetonitrile gradient, in positive ion mode, HCD fragmentation mode, and 15-second dynamic exclusion window to maximize MS / MS spectral quality for highly abundant precursors. All data were analyzed using commercial software by inspection against a manually constructed database of circularly permuted peptide variants and the masses of their B-ion fragments.

[0100] While there have been shown and described what are at present considered the preferred embodiments of the invention, those skilled in the art may make various changes and modifications which remain within the scope of the invention defined by the appended claims.

Claims

WHAT IS CLAIMED IS:

1. A method for producing a cyclic peptide, comprising:(i) providing a denatured cyanylated peptide of the formula H2N-(A1)-(A1)n-(Ad)-(Ac)- (A2)m-COOH in solution, wherein A1is a terminal amino acid bearing the terminal NH2 group; (A1)„ is a peptide linkage containing n independently selected A1amino acid residues connected to each other by amide bonds, wherein n is 1-28; Acis a single cysteine with its thiol (SH) group derivatized as a thiocyanide (SCN) group; Adis an amino acid directly attached to Ac; and (A2)mis a peptide linkage containing m independently selected A2amino acid residues connected to each other by amide bonds, wherein m is at least 3;(ii) exposing the denatured cyanylated peptide produced in step (i) to a pH in a range of 8-10 to elicit an intramolecular cyanylation-aminolysis reaction between the terminal NH2 group and Adto result in formation of a cyclic peptide of the formula:and a linear peptide of the formula:H2N-(Ai)-(A1)n-(Ad)-C00H (II) along with simultaneous production of Y-C(O)-(A2)m-COOH byproduct (III), wherein Y is an iminothiazole ring resulting from intramolecular cyclization between the SCN group and the nitrogen atom of the amide linkage between Adand Ac.

2. The method of claim 1 , wherein Acis the only cysteine residue in the cyanylated peptide.

3. The method of any one of claims 1-2, wherein n is 2-28.

4. The method of any one of claims 1-2, wherein n is 1-18.

5. The method of any one of claims 1-2, wherein n is 2-18.

6. The method of any one of claims 1-5, wherein said pH is in a range of 8.5-9.5.

7. The method of any one of claims 1-6, wherein the cyclic peptides (I) is separated from linear peptide (II) and byproduct (III) by a chromatographic method.

8. The method of claim 7, wherein the chromatographic method is high performance liquid chromatography (HPLC), ion-exchange chromatography, hydrophobic interaction chromatography, or size-exclusion chromatography.

9. The method of any one of claims 1-8, wherein the denatured cyanylated peptide is prepared by subjecting a precursor peptide to denaturing conditions followed by cyanylation of the precursor peptide with a cyanylating agent.

10. The method of claim 9, wherein the denaturation conditions comprise contacting the precursor peptide with a denaturant molecule and a mild reducing agent to eliminate a tertiary structure or folding in the precursor peptide as well as to counteract the cyanylating agent’s tendency to convert thiol groups into disulfides.

11. The method of claim 9, wherein the denaturant molecule is guanidinium chloride or urea.

12. The method of claim 9, wherein the cyanylating agent is selected from the group consisting of l-cyano-4-dimethylaminopyridinium (CDAP), ammonium thiocyanate, 2- nitro-5-thiocyanobenzoate (NTCB), 1- cyano- 4-pyrrolidinopyridinium tetrafluoroborate (CPPT or CPIP), 1 -cyano-imidazole (1-CI), 1 -cyanobenzotriazole (1- CBT), p- nitrophenylcyanate, A-cyanotriethylammonium tetrafluoroborate (CTEA), 2- cyanopyridazine-3(2H)one (2-CPO), and trimethylsilyl cyanide (TMSCN).

13. The method of claim 9, wherein the precursor peptide is a natural peptide.

14. The method of claim 9, wherein the precursor peptide is a peptide produced by chemical or proteolytic digestion of a naturally occurring protein.

15. The method of claim 9, wherein the precursor peptide comprises a defined peptide sequence recombinantly fused to a carrier protein domain.

16. The method of claim 15, wherein the carrier protein domain is derived from a protein lacking both Lys and Cys residues.

17. The method of claim 16, wherein the protein lacking both Lys and Cys residues is a Cys-free variant of the C-terminal domain of human lens yD-crystallin.

18. The method of claim 15, wherein the carrier protein domain bears a peptide tag to facilitate purification.

19. The method of claim 18, wherein the peptide tag is a hexahistidine tag.

20. The method of claim 15, wherein the cyclic peptide is a cyclic immunosuppressive peptide.

21. The method of claim 20, wherein the cyclic immunosuppressive peptide is a cyclosporin peptide.

22. The method of claim 20, wherein the cyclic peptide of Formula (I) is an orbitide.

23. A method for producing a cyclic peptide, comprising:(i) providing a compound of formula (i):(Ai)-(A1)n-(Ad)-(Ac)-(A2)m-COOH (i) wherein A1is an amino acid comprising a terminal NH2 group, (A1)]! is n independently selected amino acids, wherein n is 1-28, Acis cysteine with its thiol (SH) group derivatized as a thiocyanide (SCN) group, Adis an amino acid, and (A2)mis m independently selected amino acids, wherein m is at least 3 ; and(ii) exposing the compound of formula (I) to a pH of 8-10 to form a cyclic peptide of formula I:

24. The method of claim 23, wherein in step (ii) a compound of formula (III) and a compound of formula (IV) are formed:H2N-(Ai)-(A1)n-(Ad)-COOH (III),Y-C(O)-(A2)m-COOH (IV), wherein Y is a ring resulting from cyclization between Adand Ac.

25. The method of any one of claims 23-24, wherein Acis the only cysteine residue in the compound of formula (I).

26. The method of any one of claims 23-25, wherein n is 2-28.

27. The method of any one of claims 23-25, wherein n is 1-18.

28. The method of any one of claims 23-25, wherein n is 2-18.

29. The method of any one of claims 23-28, wherein said pH is 8.5-9.5.

30. The method of any one of claims 23-29, wherein the compound of formula (i) is prepared by subjecting a precursor peptide to denaturing conditions followed by cyanylation of the precursor peptide with a cyanylating agent.

31. The method of claim 30, wherein the denaturation conditions comprise contacting the precursor peptide with a denaturant molecule and a reducing agent.

32. The method of claim 31, wherein the denaturant molecule is guanidinium chloride or urea.

33. The method of claim 30, wherein the cyanylating agent is selected from the group consisting of l-cyano-4-dimethylaminopyridinium (CDAP), ammonium thiocyanate, 2- nitro-5-thiocyanobenzoate (NTCB), 1- cyano- 4-pyrrolidinopyridinium tetrafluoroborate (CPPT or CPIP), 1 -cyano-imidazole (1-CI), 1 -cyanobenzotriazole (1- CBT), p- nitrophenylcyanate, V-cyanotriethylammonium tetrafluoroborate (CTEA), 2- cyanopyridazine-3(2H)one (2-CPO), and trimethylsilyl cyanide (TMSCN).

34. The method of claim 30, wherein the precursor peptide is a natural peptide.

35. The method of claim 30, wherein the precursor peptide is a peptide produced by chemical or proteolytic digestion of a naturally occurring protein.

36. The method of claim 30, wherein the precursor peptide comprises a defined peptide sequence recombinantly fused to a carrier protein domain.

37. The method of claim 36, wherein the carrier protein domain is derived from a protein lacking both Lys and Cys residues.

38. The method of claim 37, wherein the protein lacking both Lys and Cys residues is a Cys-free variant of the C-terminal domain of human lens yD-crystallin.

39. The method of claim 36, wherein the carrier protein domain comprises a peptide tag.

40. The method of claim 39, wherein the peptide tag is a hexahistidine tag.

41. The method of any one of claims 23-40, wherein the compound of Formula (I) is a cyclic immunosuppressive peptide.

42. The method of claim 41, wherein the cyclic immunosuppressive peptide is a cyclosporin peptide.

43. The method of any one of claims 23-42, wherein Adand (A2)mare other than cysteine.

44. The method of any one of claims 23-43, wherein the cyclic peptide of Formula (I) is an orbitide.

Images