Peptide - an overview | ScienceDirect Topics

Chapters and Articles

You might find these chapters and articles relevant to this topic.

8.5 Physical Biochemistry of Proteins and Peptides

Proteins and peptides have attracted a great deal of attention among scientists in terms of their potential use in diagnostics, therapeutics, and drug delivery systems. For this, scientists must be familiar with the special structures of peptides, the physical and chemical properties of peptides, and so forth. Chemically, proteins and peptides, both are made up of amino acids only with nominal difference of peptides containing lesser than 100 amino acid residues in their structure.

8.5.1 Peptide and Protein Structure

Peptide chains are the key components of all the proteins. A peptide chain of a mammalian peptide or protein consists of a polyamide backbone made up of carbon atoms interspersed between amide groups and alkyl side chains. The peptide chain is formed by condensation of amino acids that involves formation of peptide bonds between the carboxylic acid C terminal and the amino group. In the C-terminus, one end is carboxylic acid while the other end is a free amino group that is known as the N-terminus. The carbon atoms present at the backbone are asymmetric and have an L-absolute configuration. The presence of these chiral centers is responsible for the conformational arrangement of proteins and thereby their stability. There are in all 22 amino acids in the body of a mammal. The side chains are either reactive or inert, depending upon the presence and type of amino acids. The presence of amino acids like arginine, aspartic acid, glutamic acid, and so on, renders the protein molecules quite reactive while those with alanine, glycine, leucine, and so forth, are very inert. These alkyl side chains are also responsible for hydrophilicity or hydrpophobicity of the protein molecule and for the formation of conjugated proteins. These side chains get covalently linked to sugar, phosphate, lipids, nucleic acids, and so on. These covalent linkages with diverse biomolecules impart the chemical and physical properties that render them biologically active and chemically stable. Amino acids are the building blocks of peptides and proteins; the latter are amphoteric in nature. One of the important physiochemical properties possessed by proteins is their isoelectric point, or pH. The isoelectric point, or pH, is defined as the pH of an aqueous solution at which there is no net charge on the molecule, in other words, the molecule is electrically neutral and does not migrate in an electric field on applying external potential. At this pH, due to the absence of either type of charge, the protein or peptide tends to precipitate out from the aqueous solution.

Peptide chain conformation is deduced by the amino acid sequences held together by means of disulfide linkages and total conformational energy. Conformational energy is the sum total of all energies of interactions. These energies of interactions are mainly composed of electrostatic energy, nonbonded energy, hydrogen bonded energy, and torsional energy.

8.5.1.1 Electrostatic Energy

Electrostatic energy arises from ionic interactions that lead to attractive or repulsive forces. This energy is inversely proportional to the dielectric constant of the medium. Ionic interactions may take place between cationic groups, for example, in between N-terminal amines like arginine and guanidine, and anionic groups like carboxylic acid in aspartic and glutamic acid. As discussed earlier, these ionic interactions, and thereby electrostatic energy, are inversely proportional to the dielectric constant of the medium. Hence, there are poor ionic interactions in proteins in aqueous milieu, leading to changed protein conformation in aqueous microenvironment as compared to the same seen in a lipoidal atmosphere.

8.5.1.2 Hydrophobic Forces

Hydrophobic forces are the energy gained in removing hydrophobic groups from water. This energy provides the catalyst for the hydrophobic clustering that may be seen between the lipophilic side chains of many amino acids. Hydrophobic interactions are key contributors to conformational energy in water.

8.5.1.3 Hydrogen Bonds

Hydrogen bonds are intermediate in terms of strength and length when compared to van der waal's forces and covalent bonds. A hydrogen bond is mainly formed between a donor atoms such as an amine, which acts as a ligand, or a lewis base and acceptor atom like carboxylic acid. The larger the number of amino acids in a given protein molecule, the greater is the probability of the formation of hydrogen bonds. These interactions significantly contribute to conformational energy. The major ways in which the chains fold to bring interacting groups together are primarily of four different types: (1) α-helix, (2) β-structure, (3) reverse turns, and (4) random coils. In the α-helix, the peptide chains get coiled in the form of a helical structure with hydrogen bonds between carbonyl groups of each residue and amide nitrogen of the fourth residue. In the β-sheet arrangement, there is a fully extended, unwound flattened portion of the chain, whereas in random coils, there is no well-defined structure.

8.5.1.4 Conformational Stability

Globular proteins tend to a single specific conformational structure. The overall conformational and topological features of a protein are mainly governed by (1) structural cross links composed of disulfide linkages and covalently bonded prosthetic groups along with noncovalent cross-links like hydrogen bonds, hydrophobic interactions, and miscellaneous forces and (2) continuous refolding to native conformations. Sometimes, the addition of organic solvent alters the conformational flexibility and structure of a protein due to change in the dielectric constant, leading to overall change in the stability of the protein molecule under consideration. The addition of reducing agents and materials like guanidine results in a reduction of disulfide linkages that are responsible for maintaining three-dimensional structures of proteins. Reduction of disulfide linkages results in increased intrinsic viscosity, suggesting that globular proteins are unfolding to loose, expanding random coil chains. This process is called denaturation.

The definition of denaturation is the nonproteolytic modification in a structure of a natural/native protein, leading to sea change in its physical, chemical, and ultimately biological activity. Protein denaturation is confirmed and assessed by a number of techniques, such as determination of intrinsic viscosity, optical rotation ultraviolet differential spectroscopy, and others. Protein denaturation finally results in disruption of disulfide bridges and separation of protein or peptide chains.

8.5.2 Methods of Protein Structure Prediction

It has always been a matter of interest and importance to develop specific and sensitive methods for the prediction of protein structure. This is particularly useful in determining the structure of antigens and epitopes that are responsible for diseases like AIDS. Structural models currently used by scientists facilitate easy structural prediction. Some of the popular and widely adopted methodologies for structural prediction of proteins are discussed here.

The availability of state-of-the-art facilities and techniques like X-ray crystallography enables scientists to determine the conformational arrangements of proteins. These techniques can be applied to new proteins where the secondary structures are not completely elucidated.

8.5.2.1 Prediction of Hydrophobicity and Hydrophilicity

Hydrophobicity is a major catalyst for the arrangement of different sections of peptide sequence in different parts of protein. Hence, prediction of hydrophobicity may act as a handy tool for determining secondary protein structure. Nozaki and Tanford [38] were pioneers in establishing a hydrophobicity scale based on solubility of amino acid in aqueous ethanol and dioxane solutions.

8.5.2.1.1 Applications of Predicted Hydrophobicity and Secondary Structure

Hydrophilicity and hydrophobicity data are instrumental in predicting the antigenic determinants. Despite the ability to use conformational techniques, these techniques have also been widely adopted to predict hydrophobicity and secondary structure, to fold new protein sequences into globular structure. If the novel protein bears a sequence analogous to a protein having known X-ray crystal structure, it may be quite easy to predict the real structure due to the concordance of the secondary structures to the original ones. This method was used for the structural prediction of renin.

Though a lot of emphasis has been given to the native conformation of a globular protein, conservation of native conformation may not be observed with noncyclic peptides. Endogeneous endorphins and enkephalins are a few such examples that display multiple conformations.

8.5.2.1.2 Solubility

Classically, proteins were classified on the basis of their solubility profile. Most nonconjugated globular proteins are soluble in water, dilute acid, and dilute salt solutions, but insoluble in concentrated salt solutions. Peptides with fewer ionic groups are apparently less water soluble. Intermolecular attractions leading to aggregation result in reduced water solubility of proteins. This may be the rationale for higher dissolution rates and enhanced solubility of proteins in denaturing solvents.

It is customary to obtain protein as a solid from an aqueous solution. However, thermolability of proteins puts a major restraint on the selection of evaporation or drying procedures. It is preferable to choose degassing and freezing followed by lyophilization to avoid protein degradation and loss. One of the drawbacks of lyophilized powders of proteins is their high bulk volume and hygroscopicity, making them moisture sensitive. Hence, such lyophilized powders should be stored over a desiccant.

8.5.2.1.3 Chiral Stability

Stereochemically active protein molecules mainly raise the issue of chiral stability. All amino acids other than glycine exhibit stereoisomerism. These stereoisomers are mainly enantiomers—nonsuperimposable on their own mirror images or diastereomers—stereoisomers that are not enantiomers. An optical rotation or change in the optical angle, like conversion of an l-amino acid to a d-amino acid form, leads to altered properties.

8.5.2.1.4 Chemical Stability

Stability of conformation has been found to be crucial for overall chemical stability. For instance, a chemical reaction could lead to lowering of conformational energy of a protein in a highly strained conformation, or vice versa. Addition of chemical reagents may also lead to stability issues. The effect of degradation products on the stability of proteins must not be overlooked and is determined quantitatively by modern techniques like HPLC. The destabilization mechanisms are mainly composed of diketopiperazine formation (seen in case of proline, glycine-containing proteins), imide formation (in asparagines, glutamine resulting in formation of cyclic aspartimide, or glutarimide), transpeptidation, oxidation, and so on.

A paradigm shift in biotechnology has led to exploration of proteins as bioactives with high potency. These proteins are major challenges to chemists from a formulation point of view, owing to their large macroscopic size and structural complexity, which render some critical issues during formulation; as well, there is the challenge of averting degradation. As seen earlier, three-dimensional conformational structures of proteins are largely dependent on the presence of chemicals, processing steps, pH conditions, and so forth. Hence, all the materials, processes, and accessories related to protein production must be thoroughly validated. Apart from this, only limited routes of delivery are available for proteins and peptides due to their lability and susceptibility to degrade in an acidic environment and in the presence of proteolytic enzymes. Two major aspects associated with protein formulations are their safety and efficacy. It should always be borne in mind that bioavailability should not be increased at the risk of raising safety concerns.

8.5.2.1.5 Levels of Protein Structure8.5.2.1.5.1 Primary to Quaternary Structure

The arrangement of amino acids or their sequential arrangement is referred to as the primary structure and is determined genetically by ribonucleic acid (RNA) during the process of transcription, and ultimately on the sequence of nucleotides in DNA. These 20 amino acids will be present in triplet codons in various combinations within a biological system. The secondary structure of a protein delineates the arrangement of individual amino acids along the polypeptide backbone of a protein, resulting in α-helices, β-sheets, reverse turns, and random coils.

Tertiary structure is the three-dimensional arrangement of a protein molecule and depicts the mode of interaction of secondary structural elements.

Quaternary structure is the conformation of a protein when it exists in a solid state or in solution and mainly based on noncovalent interaction or stereochemical arrangement of individual protein subunits (monomers).

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B9780123849649000086

2 Characteristics of Peptides in Nature

Natural peptides have a few key properties that determine the way they function in the body:

•

Relatively large molecular weight

•

Hydrophilicity

•

Sensitivity to proteases

•

Low or no plasma protein binding

2.1 Distribution

Their large molecular weight (compared to small molecules) and hydrophilicity prevent peptides from crossing cell membranes, thus the molecular targets that peptides address are extracellular, either exposed on the extracellular aspect of cell membranes or soluble in plasma or extracellular fluids.

2.2 Clearance and Half-Life

Most peptides function as messengers in the body. After secretion (be it endocrine, paracrine, or as a neurotransmitter) their action needs to be terminated to maintain fine regulation of the biological system they control. Paracrine or neurotransmitter actions can be terminated by diffusion out of the effect compartment and/or proteolysis. Endocrine actions require the elimination of the peptide from the circulation. Because most peptides have no or little plasma protein binding, they are always subject to passive glomerular filtration, which is not a very fast mechanism. The glomerular filtration rate in healthy young individuals is about 1.8 mL/kg/min and it decreases significantly with age. In addition, peptides distribute to extracellular water, with a volume of distribution at steady state of about 0.2 L/kg. Although this is a very small volume compared to small molecules, it contributes significantly to lengthening the half-life of peptides. A typical half-life in humans for a peptide subject only to passive renal filtration is 2–3 h, as in the case of the selective V2 agonist desmopressin (Minirin™) 12 or eptifibatide (Integrilin™) 2.3 Faster elimination of peptides from the circulation occurs by proteolysis by (mostly) tissue and/or plasma proteases, and other unknown mechanisms. In many cases, the half-life is difficult to measure as it is of the order of minutes. Although little is known, small peptides can also be taken up by tissues. In rats, the cyclic somatostatin agonist octreotide (Sandostatin™) 9 (MW 1638) is taken up by hepatocytes and secreted into bile via carrier mediated transport.4

2.3 Potency and Selectivity

The structural variety of the 20 natural amino acids and the relatively large number of amino acids in a peptide chain makes them rich in structural information. In spite of being linear and containing a large number of rotatable bonds, the affinity of peptides for their GPCR receptors is high, with Kd values from pM to nM. In many cases, peptides achieve a certain level of ligand preorganization by cyclization via disulfide bridges between cysteines, for example, in endothelin-1 3 (Fig. 1.a)

Fig. 1. Sequence of endothelin-1 3 and stick model of its crystal structure showing the disulfide bridges in black.

Redrawn with PyMol from Protein Data Bank entry 1EDN, without hydrogens.

It could be expected that nature would use these features to construct ligand-receptor systems that operate independently of each other with superb selectivity. However, this is not always the case as many natural peptides exhibit cross reactivities with several receptors in a family. For example, the cyclic nonapeptide arginine vasopressin (AVP) 4 secreted from the posterior pituitary is an agonist at the V1a, V1b, and V2 receptors to cause vasoconstriction, ACTH release (together with corticotropin-releasing factor, CRF), and urine concentration, respectively. The tridecapeptide α-melanocyte-stimulating hormone, α-MSH, 5 is an agonist at the MC1 (causing pigmentation), MC3, MC4, and MC5 melanocortin receptors, but not at the MC2 receptor.

Nature uses a variety of mechanisms to maintain functional selectivity, for example, distribution (CNS vs periphery), different circulating concentrations, costimulants, dilution (endocrine vs paracrine messaging), compartmentalization (neurotransmission in synapses), etc. An excellent review on the families of peptides, their receptors, and functions can be found in the Handbook of Biologically Active Peptides edited by Abba J. Kastin.5

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/S0065774317300064

5 Peptides- promising therapeutic candidates for neuroprotection

For years numerous peptides with a wide variety of biological activities have been identified. Simultaneously, improvements in the technology for peptide synthesis have made these reagents cheap and easy to produce in large quantities. These achievements have created a new class of research reagents, known as biologically active peptides, and encouraged even more to the design of peptides with improved pharmacokinetic properties, and facilitated the implementation of the most efficient methods for process development. Bioactive peptides can be defined as specific portions of proteins that have desirable biological properties and attributes related to human health (including antioxidant, anti-hypertensive, antithrombotic, anti-adipogenic, antimicrobial and antiinflammatory, anticancer, immunomodulatory effects) and food quality (including better sensory properties and shelf-life increase) [18,102]. Many biologically active peptides occur naturally and function as hormones or neurotransmitters, playing a role in activating or regulating various cellular pathways. Some of them are extracted from precursor proteins from different origins (food protein). Many others have been designed to have specific activities. There are some beneficial peptides, including low molecular weight, high activity, and specificity, easy absorption, high safety profile with low toxicity, that make these molecules of particular interest to the food and pharmaceutical industries [100]. Additionally, at this stage of development, scientists proposed many valuable strategies to improve the bioavailability of peptides and increase their use [6,7,15,133].

Based on the research findings, it seems highly indisputable that peptides are substances with multidirectional biological activity and could be considered not only as a drug but also as functional food or nutraceuticals [76]. Their main direction of biological action mentioned above is antioxidant, anti-hypertensive, antithrombotic, anti-adipogenic, antimicrobial, anticancer, antiinflammatory, and immunomodulatory effects. Here, the author will deal with the evidence proving that peptides may have neuroprotective activity.

5.1 Antioxidant peptides

Recently, many authors highlighted the antioxidant activity that peptides can exert and applied them as potential therapeutic agents to improve the health benefits, especially in the context of neuroprotection [seemorein [27,101,102,123,141]]. A variety of antioxidant peptides are products of hydrolysis of diverse food proteins from plant or animal sources (meat muscles, food byproducts, and various plant materials) that have been discussed in several reviews available in the scientific literature [100,102,123]. Peptides prevent oxidation by the same processes as well-known natural antioxidants, thus two main mechanisms are considered: hydrogen transfer and electron donation [100]. Antioxidant peptides react with the free radicals to terminate a radical reaction. Their functionality has been attributed to donating protons to free radicals, chelate metal ions and/or trap lipid peroxyl radicals [141]. Thus, antioxidant peptides are able to reduce the formation of oxidative products through the previously mentioned mechanisms, along with the stimulation of cellular antioxidant enzymes (SOD, or catalase), responsible for the maintenance of the antioxidant system in cells [92].

Generally, structural characteristics such as molecular weight, size, hydrophobicity, amino acid composition, and peptide sequence are considered determinants of a peptide's antioxidant activity [100,123,141,200]. For instance, low molecular weight peptides (500–1500 Da) are generally associated with enhanced antioxidant activity in contrast to larger peptides (above 1500 Da) [141,200]. In addition to the amino acid structure and composition, the position of particular amino acids that they occupy in the sequence is significant. Saito et al. [140] stated that any changes in each amino acid position in tripeptide chains showed different antioxidant activities. It has been postulated that peptides containing amino acid residues such as Asp, Pro, Trp, Tyr, Met, Cys, Leu, Arg, Ala, Phe, and His show higher antioxidant activities [100]. Generally, hydrophobic amino acids such as Ala, Leu, Met or Val (particularly at the N-termini) and aromatic Phe, Tyr, Trp and/or the imidazole ring-containing His (in the sequence) are able to enhance the radical scavenging abilities of peptides [100,141]. Hydrophobic amino acids present in the structure of the peptide enhance accessibility to hydrophobic targets and improve membrane permeability, as well as increase the solubility of the peptide in lipid, facilitating the contacts with hydrophobic radical species.

In contrast, amino acids with aromatic residues donate protons to electron-deficient radicals [100]. Moreover, the presence of Phe contributes to antioxidant activity due to its easily abstracted by free radical hydrogen in the methyl group [136]. Imidazole ring-containing His in peptide chains has been generally found to affect antioxidant properties because it can act as hydroxy radical scavengers, active oxygen quenchers, metal ion chelators [85]. Nucleophilic sulfur-containing amino acid residues (Cys, Met), due to its direct interaction with radicals, have an independently crucial antioxidant action [81]. Notably, the SH group in Cys as a radical scavenger protects tissue from oxidative stress and improves the glutathione activity [141].

Moreover, carboxyl and amino groups in the side chains of amino acids (Asp, Glu, His) serve as a chelator of metal ions [154] and hydrogen donors [128]. Despite the fact that certain amino acids can influence antioxidant activity, it has been suggested that the peptide bond or its structural conformation plays the same role [141]. Fig. 4 concludes the structural basis for enhanced antioxidant activity of peptides.

Fig. 4. Structural basis for enhanced antioxidant activity of peptides.

Generally, antioxidant peptides owing their activity to the composition and positioning of amino acid in the sequence could serve as agents for preventing, delay, control, and/or management of oxidative stress-related diseases such as neurodegenerative diseases. The search for pharmacologically and biologically active peptides of different origin with antioxidant properties is still ongoing. Among many hydrolysates/peptides derived from natural materials exerted neuroprotective activity by suppressing free radicals and oxidative stress markers [see more in 92], the following sequence obtained by Kim et al. [88] from venison protein is noteworthy: Met-Gln-Ile-Phe-Val-Lys-Thr-Leu-Thr-Gly (APVPH1, antioxidative peptides from venison protein hydrolysates I). The authors reported that peptide inhibited productions of nitric oxide (NO), ROS, malondialdehyde (MDA), and cell death against H2O2-induced neuronal cell damage in pheochromocytorna (PC12) cells, that confirmed its neuroprotective properties. Zhao et al. [197] evaluated the neuroprotective mechanisms on glutamate-induced apoptosis in neuronal PC12 cells of two peptides obtained from anchovy hydrolysate of the following structure: Pro-Ala-Tyr-Cys-Ser (PAYCS) and Cys-Val-Gly-Ser-Tyr (CVGSY). Compounds significantly inhibited ROS generation and MDA production and increased SOD and glutathione peroxidase (GSH-px) activities. It was hypothesized that the antioxidant activity of these peptides due to Tyr and Cys and their free radical-quenching ability [197].

Mitochondria are the main ROS source in a wide spectrum of diseases and pathologies, which makes them promising targets for pharmaceutical intervention. The idea to develop the mitochondria targeting antioxidants (MTAs) as a new class of compounds was proposed by Zhao et al. [198]. They designed MTAs, among them peptides could target mitochondria and exerted remarkably beneficial effects with potential neuroprotective outcomes. These small cell-permeable arginine-containing tetrapeptides targeted mitochondria in a potential-independent manner were very potent at reducing intracellular ROS and preventing cell death caused by the oxidant t-butylhydroperoxide (tBHP) and 3-nitropropionic acid (3NP). The structural motif of so-called Szeto-Schiller (SS) peptides centered on alternating aromatic residues and basic amino acids (aromatic-cationic peptides), this combination of the arginine residue and cationic charge probably made them easy to traverse plasma and mitochondria membranes. Their sequence is as followed: SS-02 (Dmt-d-Arg-Phe-Lys-NH2; Dmt = 2-,6-dimethyltyrosine), SS-20 (Phe-d-Arg-Phe-Lys-NH2), SS-31 (D-Arg-Dmt-Lys-Phe-NH2) [159,198]. An amino acid arrangement allowed them to freely penetrate cells despite carrying a 3+ net charge at physiologic pH and passed through the plasma membrane in both directions [160]. Research has shown that their antioxidant action was attributed to the tyrosine or Dmt residue. By reducing mitochondrial ROS formation, SS peptides inhibit mitochondrial permeability transition and cytochrome c release, thus preventing oxidant-induced cell death. Additionally, SS-02 and SS-31 dose-dependently inhibited lipid peroxidation and scavenged H2O2 [198]. As a mitochondrial-targeted peptide, SS-31 has been evaluated in different models of neurodegenerative diseases and showed remarkable neuroprotective effects. The peptide was effective against myocardial ischemia-reperfusion injury in both ex vivo and in vivo models [198]. Preclinical studies support SS-31 potential use for ALS. Peptide (1–100 μm) protected against cell death induced by hydrogen peroxide in vitro in mouse neuronal (N2a) cells cultures stably transfected with either wild-type or mutant Cu/Zn superoxide dismutase 1 (SOD1) [135]. Daily intraperitoneal injections of SS-31 (5 mg/kg) led to a significant delay in disease onset in transgenic familial ALS mice (SOD1G93A) model, improvement in survival, and motor performance. Moreover, SS-31-treated mice showed decreased cell loss and a decrease in immunostaining for oxidative stress markers in the lumbar spinal cord. SS-31 (2 and 5 mg/kg of body weight, ip) also showed some promise for ischemic stroke victims, normalizing ischemia-induced redox changes, and attenuating CD36 (a class B scavenger receptor, mediates free radical production and tissue injury in cerebral ischemia)-mediated tissue injury [26]. Further research revealed that SS-20 and SS-31 (0.1−10 mg/kg) protected against 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) neurotoxicity in male C57 black mice as experimental model of PD [189]. Peptides (0.1−10 nM) also were effective when applied in vitro using the substantia nigra-derived dopaminergic cell line (SN4741) treated with 1-methyl-4-phenylpyridinium (MPP+), the active metabolite of MPTP. Moreover, the authors showed that SS-20 and SS-31 prevented MPP+-induced inhibition of oxygen consumption and ATP production, and mitochondrial swelling using isolated mouse liver mitochondria. In another study, SS-31 (0.1−1 nM) prevented Aβ toxicity in N2a cells, had a protective effect on the mitochondrial structure and function and neurite outgrowth [105]. A similar effect was observed after incubation of SS-31 in primary neurons from mutant amyloid precursor protein (AβPP) transgenic mice (Tg2576 line), as a model of AD. Soon, SS-31 with its extraordinary potential to protect mitochondria against oxidative stress was proposed as a novel neuroprotective ophthalmic drug for protecting retinal ganglion cells (RGC) in glaucoma [132]. Continuing this research-trend, Wu et al. [177] proved that SS-31 significantly reduced MDA levels and increased SOD2 levels after glaucoma induction in an experimental rat model. Additionally, peptide SS-31 was evaluated as an agent in traumatic brain (TB) injury. SS-31 (5 and 10 mg/kg, ip) was applied in the mild TB injury animal model using male ICR mice and showed neuroprotective potential by scavenging ROS from the mitochondria, thus improving the mitochondrial function [199]. Moreover, the activation of transcriptional coactivator PGC-1α, which regulates the genes involved in energy metabolism, was considered crucial for the induction of the mitochondrial ROS-detoxifying process.

Recently novel cell-penetrating SS-peptide was developed and showed superior mitochondrial uptake and superoxide scavenging abilities [20]. The mtCPP-1 peptide differs from SS-31 by the substitution of the lysine residue with an ornithine residue. Peptide didn’t affect mitochondrial functionality, was non-toxic even at high concentrations (100 μM) to human immortalized (HeLa 705), human primary glioblastoma (U87), and mouse brain (bEnd.3) cell lines, and prevented ROS formation. Additionally, mtCPP‐1 could be potentially conjugable to different cargo molecules.

Most of the SS-peptides are distinguished by their high antioxidant and neuroprotective potential, which could be related to the ability to penetrate the brain-blood barrier and then locate in the mitochondrial inner membrane independent of MMP. The designers of SS-peptides emphasize their importance in finding the best therapies and excellent “druggable “properties, indicating their small size, simplicity of synthesis, solubility, stability, and resistance to enzymatic degradation [159].

At the same time, gramicidin-TEMPO conjugates of two fragments: a molecule with ROS-scavenging activities (4-amino-2,2,6,6-tetramethylpiperidine-N-oxyl known as 4-NH2-TEMPO) and a molecule that promotes selective accumulation within mitochondria (the membrane-active Leu-d-Phe-Pro-Val-Orn fragment of cyclopeptide antibiotic, gramicidin S as the targeting sequence), known as XJB peptides family, seemed reasonable and adequate strategy [49]. As a bifunctional antioxidant, XJB-5−131 reduced apoptosis and enhanced cell survival in the mouse embryonic cells in vitro [176], while in a mouse model of HD had remarkably beneficial effects, reduced oxidative damage to mitochondrial DNA, maintained mitochondrial DNA copy number, suppressed motor decline and weight loss, enhanced neuronal survival, and improved mitochondrial function [182]. Table 1 summarizes the most promising peptides with antioxidant properties.

Table 1. The most promising peptides with antioxidant properties.

STRATEGY	SEQUENCE	FINDINGS	REFERENCE
Suppressing free radicals and oxidative stress markers	Met-Gln-Ile-Phe-Val-Lys-Thr-Leu-Thr-Gly (APVPH1)	-Inhibition of NO, ROS, MDA production	[88]
-Inhibition of H2O2-induced cell death
Pro-Ala-Tyr-Cys-Ser (PAYCS)	-Inhibition of ROS, MDA production	[197]
Cys-Val-Gly-Ser-Tyr (CVGSY)	-Elevation of SOD and GSH-px activities in glutamate induced apoptosis
Mitochondria-targeted antioxidants (MTA)	Dmt-DArg-Phe-Lys-NH2 (SS-02)	-Inhibition of ROS production, lipid peroxidation, mitochondrial permeability transition and cytochrome c release in tBHP- and 3NP- induced cell death	[159,160,198]
Phe-d-Arg-Phe-Lys-NH2 (SS-20)
D-Arg-Dmt-Lys-Phe-NH2 (SS-31)
Phe-d-Arg-Phe-Lys-NH2 (SS-20)	-Inhibition of MPTP-neurotoxicity in animal model of PD	[189]
D-Arg-Dmt-Lys-Phe-NH2 (SS-31)
D-Arg-Dmt-Lys-Phe-NH2 (SS-31)	-Inhibition of H2O2-induced cell death	[135]
-Delay in disease onset in vivo
-Improvement in survival and motor performance in animal model of ALS
-Attenuation of ischemia-induced redox changes and brain injury	[26]
-Inhibition of Aβ-neurotoxicity in cell and animal model of AD	[105]
-Attenuation of RGC injury	[132,177]
-Inhibition of MDA
-Elevation of SOD2 activity in animal model of glaucoma
-Reduction of neuronal death and inhibition of ROS production in animal model of TBI	[199]
D-Arg-Dmt-Orn-Phe-NH2 (mtCPP-1)	-Inhibition of H2O2-induced ROS production	[20]
TEMPO-Leu-d-Phe-Pro-Val-Orn (XJB-5−131)	-Reduction of apoptosis, oxidative damage in animal model of HD	[176,182]

5.2 Mitochondria-targeted peptides

Since mitochondria are the primary sites for ROS production within cells and are recognized as a critical neuroprotective target, it seems reasonable that delivering any therapeutic agent directly to these organelles could be an incredibly effective strategy. Cationic arginine-rich peptides (CARPs) are a broad and relatively novel class of peptides being developed as neuroprotective agents, possessing extraordinary biological properties including the ability to traverse cell membranes and enter the CNS, reduce intracellular calcium influx, target mitochondria, inhibit proteolytic enzymes, induce pro-survival signaling, scavenge toxic or reactive molecules, and reduce oxidative stress or inflammation [40,104,112]. CARPs, with their ability to pass over cell membranes, enter cells and mitochondria, could also be used as carrier molecules for the delivery of other putative neuroprotective agents across the BBB and blood-spinal cord barrier. For this reason, they are also known as cell-penetrating peptides (CPPs). This family of peptides typically range in size from 4 to 30 amino acids and are positively charged due to the presence of cationic arginine, lysine, or histidine residues [104]. According to Meloni et al. [112], CARPs with neuroprotective properties are divided into three main groups: (i) poly-arginine peptides, cationic arginine-rich cell-penetrating peptides (CCPPs) or peptides derived from proteins; (ii) putative neuroprotective peptides fused to CCPPs; and (iii) endogenous peptides (Fig. 5). Characterized and classified above as MTA cyto- and mitoprotective SS-peptides belong to this expanding family of compounds.

Fig. 5. CARPs – a novel class of neuroprotective agents and its classification.

Recently, a few comprehensive reviews have been published, detailing CARPs neuroprotective activities in in vitro neuronal injury models (e.g., excitotoxicity, oxygen-glucose deprivation), in in vivo models of acute CNS injury (e.g., stroke, traumatic brain injury, perinatal hypoxia-ischemia, spinal cord injury, and epilepsy) and models of chronic neurodegenerative disorders (e.g., PD or AD) and neuropathic pain [seemorein [40,104,112]]. In these reviews, among many tested candidates for mitochondrial-targeted therapeutics, the author focused only on selected peptides that showed the most promising pharmacological profile (Table 2). According to the historical time-line proposed by Meloni et al. [112], the first short CARPs (hexapeptides: R6; 2−6 mer of arginine) have been identified in 1998 by Ferrer-Montiel et al. [47] and manifested the ability to block NMDA receptors and protect neurons from excitotoxic events. Further studies revealed that the neuroprotective efficacy of CARPs was increasing with expanding peptide arginine content and positive charge, peaking at R15 to R18 for poly-arginine peptides, and the presence of tryptophan [104,109]. All R6 to R15, R18, and tryptophan-containing poly-arginine peptides (at the dose of 0.5–15 μM) showed high neuroprotective potential following glutamate and oxygen-glucose deprivation (OGD) excitotoxicity [47,109]. Additionally, CARPs provided significant neuroprotection and improved functional outcomes in rat models of permanent and/or transient middle cerebral artery occlusion (p and/or tMCAO), perinatal hypoxia-ischemia (HI), and TB injury [112]. l-isoform poly-arginine peptides, such as R12, R15, and R18 (1000 nmol/kg, iv, 30 min post-MCAO), reduced total infarct volume and brain injury, improved the functional neurological outcomes [116]. Neuroprotective efficacy was also observed in the case of d-amino acid enantiomer of R9, R9D (1000 nmol/kg, iv, 30 min post pMCAO onset) reduced infarct volume, but without any improvement of behavioral assessment [109]. d-isoform poly-arginine peptide R18D, showed a more significant reduction in mean infarct volume following pMCAO compared to R18 [115]. In contrast, both poly-arginine peptides (30, 100, 300, or 1000 nmol/kg) increased neuronal survival in a perinatal HI model [39]. The peptide treatments resulted in significant behavioral outcomes, inhibiting the neuronal calcium influx in an in vitro glutamic acid excitotoxicity model. Moreover, R18D (wide range of doses: 10−1000 nmol/kg, ip, depending on administration time) reduced infarct volume and improved behavioral outcomes after HI [38]. R18 (1 μM) was highly neuroprotective following glutamic acid excitotoxicity, reduced neuronal intracellular calcium influx, and improved functional outcomes after TB injury [25]. In a non-human primate MCAO stroke model, R18 (1000 nmol/kg, iv, post MCAO) once again confirmed its effectiveness in reducing the severity of ischemic brain injury and improving functional outcomes after stroke [110].

Table 2. CARPs with the most promising neuroprotective properties.

PEPTIDE NAME	INJURY MODEL	FINDINGS	REFERENCE
Poly-arginine peptides, CCPPs or peptides derived from proteins
R6 (2−6 mer of arginine);	Glutamate excitotoxicity	-Inhibition of glutamate excitotoxicity, reduction of Ca2+ influx, protection of neurons from excitotoxic and OGD events	[47,109]
R15, R18, and tryptophan containing poly-arginine peptides	OGD ischemia models
R12, R15 and R18;	p and/or tMCAO stroke model	-Reduction of infarct volume, cerebral oedema, improvement of functional outcomes	[109,115,116,117]
R9D, R18D
R18, R18D	Glutamate excitotoxicity	-Reduction of Ca2+ influx, improvement of neuronal survival and functional outcomes, reduction of infarct volume	[38,39]
HI model
R18	Glutamate excitotoxicity	-Reduction of Ca2+ influx and improvement of functional outcomes	[25]
TBI model
R18	Non-human primate MCAO stroke model	-Reduction of infarct volume, improvement of functional outcomes	[110]
TAT	Glutamate and kainite excitotoxicity model	-Inhibition of glutamate/kainite excitotoxicity	[111,169]
Penetratin	Glutamate and kainite excitotoxicity model in vitro OGD ischemia model	-Improvement of neuronal survival, inhibition of glutamate/kainite excitotoxicity	[111]

Neuroprotective peptides fused to CCPPs
TAT-fused peptides	Glutamate and kainite excitotoxicity model	-Inhibition of glutamate/kainite excitotoxicity	[169]
TAT-NR2B9c	tMCAO stroke model	-Reduction of infarct volume, cerebral swelling, improvement of functional outcomes	[117]

Endogenous peptides
Apelin -13, -17, -36	Glutamate, quinic acid and HIV toxicity	-Reduction of apoptosis, ROS generation intracellular Ca2+ accumulation, mitochondria depolarization, cytochrome c release and activation of caspase-3	[23,124,193]
Serum-deprivation model
Apelin-13	H2O2 toxicity	-Inhibition of H2O2-toxicity, elevation of VEGF	[83]
Apelin-13	Cerebral I/R injury	-Reduction of the inflammation	[36,179]
Apelin-13	Corticosterone-induced neuronal damage	-Attenuation of p-Akt, p-ERKs and cleaved caspase-3	[201]
Humanin	-Aβ oligomers toxicity, over-expression of familial AD-related genes	-Reduction of neuronal death/ apoptosis	[66,82,122,192]
-Serum withdrawal model	-Amelioration of cognitive impairments
Humanin	NMDA-mediated excitotoxicity	-Inhibition of NMDA-medaited excitotoxicity, reduction of intracellular Ca2+ accumulation, ROS generation, inhibition of JNK and p38 MAPK activation	[188]
HNG	-Serum withdrawal model,	-Protection of neurons, attenuation of apoptosis and poly(ADP-ribose) polymerase activation	[16,31,66,181]
-NMDA-induced excitotoxicity
-cerebral I/R injury
HNG	HI neonatal model	Amelioration of axonal remyelination, BDNF level, neurological recovery	[22]
CN-105	ICH model	-Reduction of cerebral edema, neuroinflammation	[94,171]
-Improvement of neuronal survival and functional outcomes
CN-105	tMCAO stroke model	-Improvement of neuronal survival and functional outcomes	[166]
CN-105	TBI model	-Rrduction of neuronal injury, microgliosis, inflammation	[91]

Penetratin is one of the first CPPs to be discovered [35], also belongs to CARPs class of future neuroprotective agents. This 16 amino acid peptide sequence, derived from the third helix of the Drosophila Antennapedia homeodomain, shares common attributes of hydrophobicity, cationic nature, and ability to traverse across the cellular membrane. Penetratin (5–10 μM) provided significant neuroprotection in a dose-response manner in glutamate and kainite excitotoxicity model of toxicity. However, in in vitro ischemia (OGD) model, penetratin was less effective, and neuroprotective potency was lost with increasing concentrations of the peptide (≥5 μM) [111].

In the context of the search for neuroprotective compounds, it is worth considering the importance of TAT peptide, firstly characterized as CPPs, that showed the ability to transport various large protein cargos across BBB [51]. To date, over fifty different TAT-fused neuroprotective peptides (e.g., TAT-NR2Bct, TAT-NR2Bcts, JNK-1:TAT, and many more) have been shown to exert positive effects in different in vitro and/or animal CNS injury models [see more in 112], and their neuroprotection presumably is mediated by arginine residues and peptide positive charge. Furthermore, several studies demonstrated the TAT peptide itself exhibited modest neuroprotective activity in in vitro excitotoxicity and in vivo ischemic injury models [106,111,112,169]. A well-characterized and promising neuroprotective agent is the NR2B9c peptide fused to the arginine-rich carrier peptide TAT, which was originally known as TAT-NR2B9c, and renamed NA-1 or nerinetide. To this end, NA-1 was tested in different neuronal injury models (excitotoxicity, stroke, HI encephalopathy, intracerebral hemorrhage, AD, epilepsy, and pain) and could efficiently reduce different neuronal lesions [106,113,117,118]. For instance, NA-1 (30, 100, 300, 1000 nmol/kg, iv, post-tMCAO) reduced infarct volume, cerebral swelling and improved functional outcomes (neurological score, adhesive tape and rota-rod) [117]. This TAT-fused peptide is currently being assessed in two Phase III clinical stroke trials (ESCAPE-NA-1 and FRONTIER). Generally, some other CARPs (e.g., mentioned above R18) displayed promising neuroprotective efficacy, even better than NA-1.

Endogenous peptides with neuroprotective properties are also a distinct group of CARPs. Some hopes and challenges for the treatment of neurodegenerative disorders can be focused on the endogenous apelin peptides (apelin-36 and a family of shorter peptides such as apelin-17, -12, and -13), isolated in 1998 from the bovine stomach and identified as a ligand of the orphan G protein-coupled receptor, apelin receptor (APJ) [23,162]. Both distribution of apelin peptides and their receptor in CNS and the recent studies suggest a significant role in the neuronal signaling pathway and neuroprotective function. Apelin-13 and -36 (10 μmol/L) showed pro-survival activity in hippocampal neurons treated by NMDA-mediated toxicity of quinic acid. Similar results (apelin peptides at the dose of 20 μmol/L) were observed when hippocampal neurotoxicity was induced by human immunodeficiency virus (HIV). In human NT2.N neuronal cultures, apelin-13 and -17 increased the phosphorylation of a serine/threonine-specific protein kinases related to retroviral oncogenes (Raf), a serine/threonine-specific protein kinase (AKT), and extracellular signal-regulated protein kinases (ERK) and inhibited apoptosis induced by HIV [124]. Zeng et al. [193] showed that apelin-13 (1.0–5.0 nM) markedly blocked apoptosis in cultured mouse cortical neurons and presented an inhibitory effect on NMDA-induced intracellular Ca2+ accumulation and excitotoxicity. This neural peptide reduced serum deprivation-withdrawal, ROS generation, mitochondria depolarization, cytochrome c release and activation of caspase-3, and maintained phosphorylation states of survival-associated PI3K/Akt and ERK1/2 pathways. In vitro studies with cultured hippocampal neurons showed that apelin (1−100 μM) alone did not affect cell viability or preserve cells from H2O2-induced cell death. However, at the dose of 100 μM apelin was able to enhance the protective effect of vascular endothelial growth factor (VEGF, 50 ng/mL) against H2O2-toxicity [83]. Furthermore, these results suggested that apelin/APJ system in the spinal cord may be considered as an endogenous neuroprotective factor in the SOD1G93A mouse model of ALS. Xin et al. [179] presented that apelin-13 markedly reduced the inflammation in the rat model of cerebral ischemia/reperfusion (I/R) injury, which was observed by decreased expression of inflammatory cytokines such as IL-1β, TNF-α, and ICAM-1. Apelin-13 exerted a neuroprotective activity against corticosterone-induced neuronal damage in PC12 cells and attenuated toxin-induced down-regulation of p-Akt and p-ERKs and up-regulation of cleaved caspase-3 [201]. These effects were blocked by pretreatment of LY294002 (inhibitor of PI3K) or PD98059 (inhibitor of ERKs), indicating that activation of the PI3K and ERKs signal pathways, via sustained phosphorylation of Akt and ERKs, is required for apelin-13-mediated neuroprotection in PC12 cells. Recent reports confirmed that Apelin-13 protected the brain and PC12 cells against I/R-induced ROS-mediated inflammation and oxidative stress [36]. Changes were mediated by 5′AMP-activated protein kinase (AMPK)-mediated inhibitory phosphorylation of glycogen synthase kinase-3 (GSK-3β) downstream of APJ/G-coupled receptors pathway and further induced nuclear factor erythroid-derived 2-like 2 (Nrf2)-mediated antioxidant protein expressions.

A 24-amino acid peptide humanin (HN) isolated from the brains of patients with AD showed cyto- and neuroprotective activity [67]. Many studies classified HN as a potent pro-survival agent that suppressed the neuronal death caused by all AD-related insults, including Aβ oligomers and over-expression of familial AD-related genes [67,121]. HN is also effective against cell death induced by other cellular stress stimuli under different experimental settings, such as serum withdrawal, NMDA-induced excitotoxicity, stroke, etc. [16,31,82,181,192]. Recently, Yang et al. [188] investigated the neuroprotective mechanism of HN (10–40 μM) in vitro, and revealed that peptide reversed NMDA-induced injury by increasing cell viability and cell survival, decreasing lactate dehydrogenase (LDH) release, reducing the intracellular elevation of Ca2+ levels and ROS generation, inhibiting the c-Jun N-terminal kinase (JNK) and p38 mitogen-activated protein kinase (MAPK) activation. The authors suggested that inhibition of the ROS-dependent JNK/p38 MAPK signaling pathway was responsible for HN neuroprotective efficacy. Moreover, HN and its analogs could help ameliorate cognitive impairments induced by different insults [192]. Generally, HN affects the modulation of tyrosine kinase, ERK1/2, AKT, signal transducer, and activator of transcription 3 (STAT3) and JNK signaling cascades [66,87]. Highly potent HN derivative S14G-HN (Gly14-HN = HNG, 0.1 μg, icv/ ip) showed the ability to protect against cerebral I/R injury in mice. Additionally, HNG improved neurological deficits after I/R injury and attenuated both neuronal apoptosis as well as poly(ADP-ribose) polymerase activation. Peptide exerted its neuroprotection in vivo, at least partly by a receptor-dependent inhibition of ERK activity [181]. Moreover, HNG decreased Aβ-amyloid burden in triple transgenic mice that show the age-dependent development of multiple pathologies relating to AD and ameliorated cognitive function in male mice (after three months of intranasal treatment at the dose of 10 nmol) [122]. Additionally, immunohistochemical analyses showed that Aβ levels in brains were markedly lower in HNG-treated male and female mice than control mice. Chen et al. [22] detected axonal remyelination, neurological recovery, and increased level of BDNF in HI brains after delayed HNG (50−100 μg/kg, ip) treatment.

Dr. Laskowitz’s team [91] developed a small pentapeptide CN-105 (Ac-VSRRR-amide), that easily cross BBB, and has demonstrated neuroprotective potential across a wide variety of preclinical models of brain injury and safety in early-phase clinical trials. The peptide mimics the polar face of the apoE helical domain and its ability to counteract neuroinflammation and neurodegeneration. CN-105 has received orphan drug status, its safety was demonstrated in both single escalating dose and multiple dosing paradigms in the Phase 1 clinical trial (clinical trials identifier: NCT02670824). Other currently early phase clinical trials are conducted, including a pilot first-in-disease state trial for patients with primary intracerebral hemorrhage (ICH) (NCT03168581, NCT031711903, NCT03802396) [63]. CN-105 peptide (0.05 mg/kg, iv) showed improved both short-term and long-term functional outcomes in preclinical murine model of collagenase-induced ICH [94]. Cerebral edema was reduced, neuroinflammation was decreased and hippocampal CA3 neuronal survival was increased, suggesting the therapeutic potential of CN-105 for patients with acute ICH. Additionally, neuroprotective and anti-inflammatory properties of CN-105 (0.1 mg/kg, iv) was evaluated in a microglial cell line and murine model of tMCAO [166]. Studies revealed the reduced infarct volume and microglial activation as well as improved survival, and functional outcomes in ischemic stroke. Observed changes resulted via modulation of neuroinflammatory pathways. Treatment with CN-105 (0.05 mg/kg, iv) was associated with histological and functional benefits in a closed-head TBI murine model [91]. Peptide effectively reduced secondary hippocampal neuronal injury, microgliosis, and inflammatory gene expression. Recently published outcomes showed therapeutic potential of CN-105 in translational ICH models that account for hypertensive comorbidity, sex, species, and age [171].

Taken together, CARPs arouse great interest among scientists because of their hydrophobicity, cationic nature, and capability to deliver neuroprotective cargos to the CNS. Moreover, CARPs themselves could be developed into neuroprotective drugs. One of the disadvantages that more or less appears in the research is their nonspecific cell uptake due to their cationic nature by almost all cells in the body, which narrows targeted therapy.

5.3 Opioid peptides

Opioid peptides and opioid systems, with μ, δ, and κ receptors (MOR, DOR and KOR, respectively) widely expressed in the central and enteric nervous system, as well as in peripheral tissues, were traditionally thought to be primarily involved in modulation in pain signaling but also other physiological processes. They are known as modulators/mediators in cardiovascular, immune, gastrointestinal, and neural systems [48]. Moreover, they contribute to stress responses and cell survival, demonstrating antioxidant, antiinflammatory, antiapoptotic properties [75]. New outcomes consistently reported beneficial effects of opioids, describing their molecular mechanisms, including stabilization of ionic homeostasis, stimulation of neurotrophic factor release and transduction of pro-survival signals, reduction of oxidative damage [168]. Therefore opioids are proposed as unique neuroprotectors. Opioid peptides, as well as alkaloid opioid agonists, are able to promote neuronal survival during nervous system development and affect neuronal and glial proliferation, neuronal migration, dendritic growth, and spine formation [75]. Additionally, they are considered neurotrophic factors.

The diverse biological effects of opioids are manifested through specific membrane-bound heptahelical opioid receptors, which are a group of G protein-coupled receptors (GPCRs). It has been shown that, in addition to trophic factors, some GPCRs family are able to stimulate an ERK, a member of the MAPKs, implicated in cell survival and an intracellular /AKT (PI3K/AKT) signal transduction pathway that promotes metabolism, proliferation, cell survival and growth in response to extracellular signals and may be beneficial in neuroprotection. [75].

Preliminary research conducted by Meriney et al. [114] in 1985, indicating the involvement of the opioid system in neuroprotection, concerned morphine, an alkaloid opioid agonist known as an effective drug for diarrhea and pain, that preferentially acts by activation of the μ-opioid receptors. Morphine reduced naturally occurring neuronal death in the ciliary ganglion of the chick embryo. This evidence supports the idea that the opioid system may play a role in regulating neuronal survival and cell proliferation during nervous system development. Iglesias et al. [75] have shown the survival-promoting activity of morphine (0.01−10 μM) on serum-deprived neuroblastoma (SH-SY5Y) cell line. This effect (at the dose of 1 μM) was abolished in the presence of β-FNA (irreversible and μ-selective opioid antagonist, 10 μM), indicating the role of the μ- opioid receptor in neuroprotection. Additionally, morphine (50−200 μM) prevented glutamate-induced toxicity of primary rat astrocytes and C6 glial cells [93] and 6-hydroxydopamine (6−OHDA)- induced injury of human derived neuroblastoma (SH-SY5Y) cell lines [41,44], respectively. Morphine is protected against intracellular amyloid-β (iAβ) toxicity in human and rat primary neuronal cultures, which was mediated through μ-opioid receptor and estrogen receptor activation [32]. Wang et al. [174] showed that morphine (50 μM) preserved SH-SY5Y cells against 6−OHDA-induced impairment, e.g., morphine attenuated apoptosis and ER stress possibly by activation of autophagy, improved mitochondria dysfunction, and suppressed the accumulation of intracellular ROS. Neuroprotective effects of low-dose morphine were also observed after oral administration, concerned with the improvement of midbrain tyrosine hydroxylase (TH) expression, a decline of apomorphine-evoked rotation, and attenuation of pain hypersensitivity in a 6−OHDA-induced PD rat model. Meanwhile, morphine (0.01−0.25 mg/mL) prolonged the lifespan and ameliorate motor function in the transgenic PD Drosophila model. A cross-tolerance between morphine and nicotine's protective effects occurred in 6−OHDA-induced SH-SY5Y cell toxicity, while morphine had no protective effects in chronic nicotine-incubated cells [43].

Since morphine is a ligand of opioid receptors exerted a neuroprotective effect, similar results could be expected from opioid peptides (Table 3). Two naturally occurring endogenous opioid peptides with high affinity and selectivity towards μ-opioid receptor, endomorphin-1 (EM-1, Tyr-Pro-Trp-Phe-NH2) and endomorphin-2 (EM-2, Tyr-Pro-Phe-Phe-NH2) [48,77] were considered as cytoprotective and pro-survival agents. Endomorphins (EMs) have been shown to be present in cells and tissues of the immune system and to modulate a variety of immune parameters or even damages related to inflammatory diseases of the brain [79]. Additionally, they scavenge radicals, inhibit lipid peroxidation, DNA and protein oxidative damage [97], and prevent oxidative damage of LDL in in vitro model of oxidative damages generated by 2,2′-azobis (2-amidinopropane hydrocholoride) (AAPH), H2O2 or copper ions [96]. These findings were consistent with the neuroprotective potential of EM-2, suggesting its role as a lead compound for drug development in AD treatment. EM-2 (1–200 μM) reduced the neurotoxicity of amyloid proteins (Aβ1–42) in the cell-viability assay using SH-SY5Y cell line, protected against the hyperexcitability caused by Aβ1–42 in vivo, preventing against the field excitatory postsynaptic potential attenuation in vivo [158]. Tetrapeptide obviously was able to bind to Aβ1–42, but didn’t arrest fibril formation or disassemble preformed aggregates. Moreover, EMs protected against iAβ toxicity in human and rat primary neuronal cultures and in rat brains in vivo [32]. In neuronal cultures, both peptides at the dose of 1μM reduced iAβ-cell death to ∼25 %, while cAβ1−42 constructs-cell death generated by Adeno-5 virus infection was decreased to ∼50 %. Additionally, EMs and morphine protected mitochondrial function in the I/R- injured brain in the mouse [45]. Using in vitro model of isolated brain mitochondria, the authors demonstrated improvement of mitochondrial respiratory activity and attenuation of oxidative stress-induced changes in mitochondrial membrane fluidity, lipid peroxidation, and cardiolipin release. EMs also blocked the induced release of cytochrome c, which results in the inhibition of cell apoptosis.

Table 3. Opioid peptides tested for neuroprotection.

OPIOID RECEPTOR AGONIST	INJURY MODEL	FINDINGS	REFERENCE
Tyr-Pro-Trp-Phe-NH2 (EM-1), Tyr-Pro-Phe-Phe-NH2 (EM-2)	In vitro model of oxidative damages generated by 2,2′-azobis(2-amidinopropane hydrocholoride) (AAPH), H2O2 or copper ions	EMs provided antioxidant defense in the brain against oxidant- and free radical-induced damage	[96,97]
In vitro/in vivo model of AD	EM-2 reduced the neurotoxicity of Aβ1–42 but didn’t arrest fibril formation or disassemble preformed aggregates	[158]
In vitro/in vivo model of AD	EMs protected against iAβ toxicity	[32]
In vitro mouse brain mitochondria anoxia-reoxygenation model	EMs protected mitochondria against IR-induced oxidative	[45]
Tyr-d-Ala-Gly-N-MePhe-Gly-ol (DAMGO)	In vitro model of serum-deprived apoptosis	Role of opioid receptors in cell survival signals and activation of PI3K/Akt signal transduction pathway	[75]
In vitro model of serum-deprived apoptosis	Role of CDK5 in neuroprotective activity	[175]
In vitro model of PD	DAMGO prevented 6-OHDA-induced toxicity, suppressed ROS elevation	[41]
Tyr-d-Pen-Gly-Phe-d-Pen (DPDPE)	In vivo model of global brain ischemia	Pre-ischemic administration of DPDPE produced increase in CA1 neuron survival	[21]
Tyr-d-Ala-Gly-Phe-d-Leu (DADLE)	In vitro model of glutamate-induced cytotoxicity	Role of δ-opioid receptors in cell survival signals and reduction of glutamate-induced cytotoxicity	[195]
In vitro model of hypoxia-induced injury	Role of δ-opioid receptors in neuroprotection in normoxic and hypoxic environments, pre-hypoxic DADLE treatment abolished the neuronal injury	[194]
In vivo model of cerebral ischemia	Pre-stroke administration of DADLE reduced neuronal apoptosis, attenuates motor asymmetrical behaviors, reduces cerebral infarction	[11]
In vivo model of forebrain ischemia	Pre-ischemic administration of DADLE had protective effect against CA1 neuronal loss	[152,153]
In vivo model of spinal cord I/R injury caused by aortic occlusion	Protected the spinal cord against I/R injury	[99]
In vivo model of cerebral I/R injury	Protected against cerebral I/R injury by decreasing nerve cells apoptosis, via ctivation of PI3K-Akt pathway	[103]
In vitro and in vivo model of dopaminergic terminal damage induced by METH	Protected and reverse neuronal damage even at the genomic level, provided antioxidant defense	[68,69,164,165]
In vivo model of MCAO	Protected neuronal damage, provided antioxidant defense	[187]
In vitro model of serum-deprived apoptosis	Role of MEK-ERK pathway and δ2-opioid receptors	[70]
In vitro model of OGD injury	Reduced neuronal injury through the MAPK pathway	[155]
In vitro and in vivo model of OGD/R injury	Role of AMPK/mTOR/ULK1 signaling pathway	[90]
Tyr-d-Ser-Gly-Phe-Leu-Thr (DSLET)	In vitro model of serum deprived apoptosis	Inhibited apoptosis and prevented the intracellular elevation of the Bcl-2 and Bak proteins	[34]
(Tyr-d-Ala-Gly-Phe-NH–)2 (Biphalin)	In vitro OGD models and in vivo MCAO model	Reduced brain edema possibly through decreased glutamate release and attenuated NKCC activity	[183,184]
In vitro model of NMDA-induced damage	Reduced neuronal damage, provided antioxidant defense	[84]

Impressive results were obtained for the synthetic enkephalin derivative with high μ-opioid receptor affinity and selectivity, DAMGO (Tyr-d-Ala-Gly-N-MePhe-Gly-ol). This compound's potential protective effects on apoptotic cell death were tested using non-transfected fully differentiated SH-SY5Y cells (exposed to retinoic acid for five days) and primary culture of mouse cortical neurons [75]. In differentiated SH-SY5Y cells, DAMGO (0.1–2 μM) promoted cell survival after serum deprivation without inducing cell proliferation. In mouse primary cortical neurons, DAMGO (1 μM) decreased the percentage of apoptotic cells after serum withdrawal. In both models, blockade of μ-opioid receptor response by naloxone (50 μM) and inhibition of PI3K by specific LY294002 (10 or 20 μM), completely terminated DAMGO effects, indicating the role of opioid receptors in cell survival signals. Moreover, DAMGO stimulated AKT phosphorylation via PI3K-dependent signaling cascade, which may be a key downstream kinase involved in this anti-apoptotic effect. It has also been reported that DAMGO (0.1 μM) partially attenuated cell death caused by cyclin-dependent kinase 5 (CDK5) inhibitor in differentiated SH-SY5Y cells as well as in primary cultured hippocampal neurons after serum deprivation [175]. CDK5 is an essential molecule that participates in neuronal migration, neurite outgrowth, survival, and development of the cerebral cortex, while its overactivation is toxic to cells. Wang et al. [175] evaluated the first time activation of CDK5 in response to μ-opioid receptor agonist. In this case, the PI3K pathway was neither stimulated nor responsible for the observed increase in CDK5 activity and up-regulation of a neuron-specific activator (p35), and phosphorylation of STAT3 was involved in neuroprotection. Moreover, DAMGO used in in vitro model of PD, prevented 6−OHDA-induced SH-SY5Y cells toxicity [41]. The neurotoxin-induced ROS and intracellular calcium level elevation was significantly suppressed by this peptide (0.1−1 μM). Additionally, DAMGO attenuated the decrease in mitochondrial membrane potential in injured cells. Immunoblot technique showed that DAMGO could significantly inhibit the release of cytochrome-c and caspase-3 as biochemical markers of apoptosis induction.

Interestingly, the neuroprotective role of DORs and their ligands is based on the reports according to which DOR density is much higher in the turtle brain that makes it highly tolerant to hypoxic/ischemic stress in contrast to the rat brain [144,178]. Thus, DOR agonists have been studied extensively and shown maximum neuroprotective activity among other OR agonists, suggesting that DOR system could be a potential target against neurological impairment [61,150,168]. Data of the late 1980s and the early 2000s provided strong evidence demonstrating the importance of DORs and delta opioid peptides in ischemic and hypoxic conditions [seemorein 61,144]. For instance, Mayfield et al. [108] showed increased survival during hypoxia, when animals were pretreated with a synthetic δ-opioid receptor-ligand DPDPE ([D-Pen2,D-Pen5]-enkephalin, 100 mg/kg, sc). Observed survival improvement was inhibited by DOR antagonists, which confirmed the opioid-dependent mechanism. Using a rat model of global brain ischemia, Charron et al. [21] supported the role of DORs in attenuation of ischemia-induced hippocampal damage and cognitive impairments by pre-ischemic treatment with DPDPE (5 μg, icv). Peptide produced a 59 % increase in hippocampal CA1 neuron survival seven days post-ischemia and had no significant impact on locomotor activity.

Since we have known that synthetic δ-opioid receptor ligand DADLE ([D-Ala2, d-Leu5]-Enkephalin) was highly influential in inducing hibernation [125,150,152], thus could be considered as a model of hypoxia refers to a state in which oxygen supply is insufficient, this opioid peptide became a target object in the search for potential pro-survival agents. Soon, Zhang et al. [195] firstly demonstrated the decreased glutamate-induced injury in neuronal cultures from rat neocortex, after administration of DADLE (0.01−10 μM). Simultaneously, activation of MOR and KOR did not elicit any significant pro-survival effects, suggesting that only DORs are responsible for the observed neuroprotective effects. DADLE diminished the number of injured cortical neurons and significantly reduced glutamate-induced LDH increase. Similar results were obtained using hypoxia-induced injury (1% oxygen), DADLE protected cortical neurons against O2 deprivation, while DORs played a crucial role in this pro-survival mechanism [194]. The authors observed attenuation of neuronal injury after 24-h DADLE treatments in either normoxic or hypoxic conditions, but any substantial benefit was detected after prolonged administration. The importance of DORs in ischemia was confirmed in vivo when mice were subjected to MCAO [11]. Delta binding sites decreased prior to reductions in κ- or μ- binding sites, concomitant with infarct core extension [150], which insinuated that KORs and MORs are less sensitive to brain injury then DORs and stimulation of DORs might display a protective effect [55]. Moreover, Borlongan et al. [9,10] described DADLE as a neuroprotective agent against ischemic stroke, able to induce a hibernation-like state. The study results showed that pre-stroke administration of DADLE (4 mg/kg, ip) significantly ameliorated infarct area, reduced neuronal apoptosis, and prevented behavioral abnormalities. Additionally, peptide caused the increase of glial cell line-derived neurotrophic factor (GDNF) expression levels, but not NGF, in the striatum and cortex, which are the MCA occlusion/reperfusion or stroke target brain areas. Using a rat model of global forebrain ischemia, DADLE (0.2−20 mM, icv) administrated 45 min before ischemia preserved CA1 neuronal loss and improved neurological behavior [153]. Additionally, DADLE (0.05 mg/kg) protected spinal cord I/R injury induced by infrarenal aortic occlusion, improved hind limb motor function, narrowed incidences of paraplegia, and neurologic dysfunction [99]. In a rat MCAO model where cerebral I/R injury was initiated, DADLE (5 mL/kg) decreased nerve cells apoptosis by activation of PI3K-AKT pathway [103].

Taking into account other neurodegenerative impairments, Tsao et al. [164,165] proved the protective property of DADLE (2 and 4 mg/kg, i.p.) against methamphetamine (METH)- induced injury. METH is a neurotoxin and psychostimulant, causes destruction of DA terminal in the CNS, generates ROS [165], as well as involves the activation of immediate early genes and certain transcription factors [152]. Generally, DADLE (18 mg/kg, ip) reversed the DA terminal damage induced by METH and caused a significant, transient restoration of dopamine transporter level in the striatum [164]. Obtained results showed that the action of DADLE might involve opioid receptors as well as the free radical uptake. Therefore, DADLE acts as a free radical scavenger in sequestering the formation of superoxide anions and hydroxyl free radicals, also prevents lipid peroxidation in the synaptosomal preparation [152,165]. As it was mentioned before, METH is able to exert genomic changes and elevate the brain mRNA levels of an immediate-early gene c-fos and the gene of a tumor necrosis factor p53. Results obtained by Hayashi et al. [68,69] indicated that DADLE reversed the effect of METH injury even at the genomic level.

It has been elucidated that DOR agonists protect the brain from oxidative damage, particularly enkephalins and their analogs such as DADLE exhibit free radical scavenging activities and can reduce ROS-induced lipid peroxidation [28,50]. Yang et al. [187] also revealed in a rat MCAO model that DADLE (1 mg/kg, ip) protected neurons not only by inhibition caspase activity but also by attenuation oxidative injury. Moreover, DORs played a crucial role in increasing the action of the antioxidant enzymes, glutathione peroxidase and superoxide dismutase, and decreasing levels of the free radicals, malondialdehyde, and nitric oxide, in the brain.

In another study, DADLE enhanced the survival of serum-deprived PC12 cells [70], but its antiapoptotic action was observed at very low concentrations (pM-nM). The suggested mechanism involved the MAPK-ERK pathway, perhaps through δ2-opioid receptors. Moreover, it was found that high concentrations of DADLE (mM) were cytotoxic in PC12 cells and might cause cell death by activating Fas-ligands (FasL), at least in part, through a mechanism involving μ-opioid receptors. DADLE also had a protective effect against neuronal injury induced by OGD in cortical neurons, and this effect was entirely blocked by naltrindole, delta opioid antagonist [155]. DADLE elevated phosphorylation of ERK and prevented OGD-induced p38 phosphorylation. Neither DADLE nor naltrindole had any appreciable effect on phosphorylation of JNK. A recently published study confirmed the neuroprotective potential of DADLE against ischemia in an oxygen-glucose deprivation/reperfusion (OGD/R) model [90]. DADLE increased DOR levels after ischemia in vivo and in vitro, which enhanced neuronal autophagy after ischemia by triggering the AMP-activated protein kinase, mammalian target of rapamycin (mTOR), and unc-51-like kinase 1/2 (Ulk1/2), three interconnected major junctions within the autophagy regulating signaling network (AMPK/mTOR/ULK1 signaling pathway).

However, in 2000 Dermitzaki et al. [34] examined the effect of various synthetic opioids, including DADLE and DSLET ([D-Ser2]-Leu-enkephalin-Thr, a selective δ2 agonist) on PC12 cell apoptosis, both under normal growing conditions and under serum deprivation. Promising results were obtained for DSLET, while DADLE did not have any effect on PC12 apoptosis. DSLET inhibited apoptosis to 62.4 ± 4% (mean ± SEM) of the parallel controls and prevented the intracellular elevation of the Bcl-2 and Bak proteins, suggesting that opioids might protect, transiently, PC12 cells from entering a state of induced apoptosis following acute fluctuations in the supply of growth factors [34].

A synthetic biphalin [(Tyr-d-Ala-Gly-Phe-NH–)2], a dimeric opioid peptide with an affinity for three types of opioid receptors, showed a more significant effect in decreasing water content in OGD-exposed hippocampal slices, compared with other selective opioid agonists (at the dose of 0.01−10 μM) [186]. Moreover, peptide (5.7 μmol/kg) reduced brain edema and infarct ratios and also improved behavioral outcomes in a 12 h permanent MCAO model of focal ischemia. Opioid receptor-dependent neuroprotection of biphalin in stroke could be initiated by the inhibition of Na+, K+, 2Cl− cotransporter (NKCC) function via the regulation of PKC-dependent cell signaling [185]. Moreover, biphalin (0.025−0.1 μM) evoked a neuroprotective effect in organotypic hippocampal cultures (OHC) exposed to NMDA-induced damage [84]. Other results confirmed pro-survival activity, biphalin reduced neuronal damage induced by glutamate and hypoxia-reperfusion, decreased free radical damage in neurons challenged with hypoxia-reperfusion and significantly reduces brain stroke damage, and improves behavioral endpoints [183].

Generally, for several years the opioid system was investigated as a therapeutic target for neuroprotection. Notably, DOR agonists showed maximum neuroprotective effects, among other OR ligands. Their activity was mediated by the maintenance of ionic homeostasis, alleviation of glutamate excitotoxicity, inhibition of Bax-related apoptosis, increasing the transduction of pro-survival signals, reducing oxidative injury, regulating DOR expression, and their ability to promote neuronal survival involved multiple molecular signaling pathways, such as protein kinase C (PKC) and MAPK/ERK, PI3K/Akt and BDNF/TrkB pathways [seemorein [61,90,150,168]].

5.4 Other peptides

Based on the hypothesis that amyloid formation could be inhibited by peptides homologous to Aβ, Tjernberg et al. [163] characterized peptide sequence Lys-Leu-Val-Phe-Phe (KLVFF; Aβ16−20) that prevented Aβ assembly into amyloid fibrils. The introduction of so-called beta-sheet-breaker (BSB) peptides showed some promise to the design of rational and putative protective peptides because BSBs were able to arrest Aβ fibrillogenesis and/or amyloid neurotoxicity [147,149]. Other peptides with neuroprotective potential are summarized in Table 4.

Table 4. Other peptides with neuroprotective potential.

Peptides	FINDINGS	REFERENCE
BSB
Lys-Leu-Val-Phe-Phe (Aβ16−20)	Prevented Aβ assembly into amyloid fibrils	[163]
Leu-Pro-Phe-Phe-Asp (iAβ5)	Inhibited amyloid formation, neurotoxicity in vitro, reduced amyloid load and cerebral damage in vivo	[134,149]
Leu-Pro-Tyr-Phe-Asp-NH2	Protected neurons from the modulatory effect of Aβ1–42 by binding to the Aβ fibrils, reduced amyloid load, and cerebral damage in transgenic animals	[157]
Arg-Ile-Ile-Gly-Leu-NH2

NAP
Asn-Ala-Pro-Val-Ser-Ile-Pro-Gln (davunetide)	Protected neurons against toxic events in vitro and in vivo	[60,95]
AL-108	Ameliorated the memory function in patients with amnestic mild cognitive impairment	[59]

Fibroblast growth factor family
FK18	Protected neurons in OGD and retinal I/R injury models	[180]
CopA3
CopA3	Protected neurons against 6-OHDA /okadaic acid/ toxin A	[119,191]

Pituitary adenylate cyclase-activating polypeptide
PACAP	Protected neurons against salsolinol/ β-amyloid/ 6-OHDA/glutamate/ oxidative stress/ MPTP/ rotenone;	[13,126,138,139,145,170,172,173]
Empty Cell	Attenuated diabetic retinal injuries	[156]

Proline-rich oligopeptides
BPP-10C	Reduced ROS generation and oxidative stress markers, dependent of AsS activity	[131]
Bj-PRO-5a and Bj-PRO-7a	Reduced ROS generation and oxidative stress markers, independent of AsS activity	[130]

One such compound, protective pentapeptides Leu-Pro-Phe-Phe-Asp (LPFFD, iAβ5) that were synthesized based on the N-terminal domain of Aβ, successfully inhibited amyloid formation and neurotoxicity in vitro (1.5 μg/μl) [149] and reduced amyloid load and cerebral damage in two different transgenic animal models of AD (1 mg in 250 μl saline) [134]. Additionally, the end‐protected version of iAβ5, acetylated at the N‐terminus and amidated at the C‐terminus (iAβ5p), was very stable in human plasma and cerebral spinal fluid (CSF) in contrast to not modified peptide. Szegedi et al. [157] characterized series of pentapeptides using in vitro and in vivo electrophysiological assays and concluded that Leu-Pro-Tyr-Phe-Asp and Arg-Ile-Ile-Gly-Leu amides protect neurons from the modulatory effect of Aβ1–42 by binding to the Aβ fibrils, successfully reduced amyloid load, and cerebral damage in transgenic animals thus could be considered as lead compounds for designing novel drugs for AD, while Phe-Arg-His-Asp-Ser amide and Gly-Arg-His-Asp-Ser amide probably act rather as integrin-type receptor-binding ligands

A synthetic octapeptide presents promising neuroprotective activities and minimal effects on vital organ functions in animals derived from activity-dependent neuroprotective protein (ADNP), NAP peptide (NAPVSIPQ = Asn-Ala-Pro-Val-Ser-Ile-Pro-Gln, known as davunetide) [8,143]. NAP exhibited multifaced neuroprotection at very low concentrations in vitro (femtomolar concentrations) and in vivo (microgram per kilogram doses) [95]. In cell culture, NAP protected neurons against toxic events associated with oxidative stress, glucose deprivation, the β-amyloid peptide, NMDA, (HIV)-1 envelope glycoprotein (gp120), and electric blockade, dopamine, H2O2, nutrient starvation, and zinc overload [summary of in vitro pharmacology is presented in 60]. In animal models, NAP prevented apolipoprotein E and cholinergic deficits, stroke injuries, middle-aged anxiety, cognitive dysfunction, reduced edema formation, mortality, and facilitated recovery in mice model of head-injury [summary of in vivo pharmacology is presented in 60]. The most spectacular results showed that NAP seemed to be a good candidate as a disease-modifying treatment for HI brain injury [89,95], hypoxia-induced seizures [62], and AD [59]. Early human testing indicated that the intranasal formulation of NAP named AL-108 (single dose of 15 mg) and intravenous administration of NAP had a positive impact on memory function in patients with amnestic mild cognitive impairment made NAP peptide a novel tangle-directed drug candidate [57].

Recombinant protein therapeutics strategy involving neurotrophic factors or growth factors became another popular approach for neuroprotection [161,196]. Recently, a novel peptide with some protective potency has been identified and proposed as a prototype for further research. Xiong et al. [180] indicated that 18-amino acid peptide, FK18 from basic fibroblast growth factor (bFGF), has a neuroprotective effect against ischemic injury. Its potential mechanisms were assessed using the in vitro (OGD) model in SH-SY5Y cells and the in vivo retinal I/R injury model to mimic ischemic neuronal injury. Peptide FK18 significantly increased the cell viability, attenuated the cell apoptosis, inhibited the activation of caspase-9, and caspase-3. In the animal model, peptide diminished I/R-induced retinal neuronal apoptosis, damage to RGCs, and morphological and functional damage to the retina. Additionally, FK18 elevated AKT phosphorylation under both normoxic and OGD conditions, reduced mitochondrial translocation of the proapoptotic protein Bad, up-regulated the expression of Bcl-2/Bax. It inhibited the release of cytochrome c from the mitochondria into the cytoplasm [180].

The antibacterial peptide, CopA3 (a d-type disulfide dimer peptide, LLCIALRKK), derived from the amino acid sequence of the α-helical region of the 43- amino-acid insect peptide, Coprisin, originally isolated from the Korean dung beetle, Copris tripartitus [73], has neurotrophic effects on SH-SY5Y cells and neural stem cells isolated from mouse brains [119]. In both cell types, CopA3 (10 μg/mL) time-dependently increased cell proliferation, significantly inhibited the apoptosis and viability losses caused by 6−OHDA and okadaic acid (an AD-mimicking agent). The putative mechanism through which CopA3 exerts neuroprotective and neurotropic effects may be the result of 27Kip1 protein degradation (a negative regulator of cell cycle progression). Recently, CopA3 treatment (10 μg/mL) inhibited Clostridium difficile toxin A-induced neural cell damage, completely inhibited activation of caspase-3 (the proapoptotic factor), and increased proliferation [191]. Moreover, the neuroprotective effects of CopA3 are not associated with well-characterized toxin A pathway through ROS/p38 MAPK/p27Kip1, suggesting new independent signals.

Several studies have revealed great therapeutic potential of pituitary adenylate cyclase activating polypeptide (PACAP, represented by 38 amino acids sequence known as PACAP38, as well as shorter biologically active form with 27 amino acids of the N-terminal, PACAP27) that is widely expressed in the central and peripheral nervous system [see more in 184]. This pleiotropic neuropeptide exhibited a significant neuroprotective and neurotrophic potential through both direct and indirect ways in several in vitro and in vivo neurodegenerative models, especially PACAP inhibited pathological processes in models of AD and PD, and alleviated clinical symptoms. Furthermore, evidence suggested that PACAP might move across the BBB in amounts sufficient to affect the brain functions, which raised its importance in searching for neuroprotective agents. Early studies showed that PACAP attenuated the glutamate-induced neurotoxicity in cultured retinal neurons by increasing the cAMP level, resulting in the activation of protein kinase A (PKA) [145]. Further research proved the ability of PACAP to protect against β-amyloid toxicity in PC-12 cells by inhibition of caspase-3 [126] and primary cortical neurons by activating the AMPK pathway, enhancing Sirtuin 3 expression and improving mitochondrial function [64]. Moreover, neuropeptide protected cerebellar granule neurons against oxidative stress-induced apoptosis [170]. In in vitro model of PD PACAP attenuated rotenone-induced neurotoxicity in PC12 cells through PKA and MAPK pathway [173] and combated inflammatory-mediated toxicity induced by salsolinol in SH-SY5Y cells [13]. Furthermore, PACAP reduced neurons injury and improved behavioral symptoms in a 6−OHDA induced PD rat model [138,139]. The treatment with PACAP declined the memory impairment in MPTP-neurotoxicity animal model of PD [172]. To date, several studies have described the significant neuroprotective potential of PACAP concerning retinal metabolic disorders by regulating cell survival and death [33,53]. For example, intraocular injection of PACAP attenuated diabetic retinal injury. The levels of the antiapoptotic p-Akt, pERK1, p-ERK2, PKC, and Bcl-2 signaling increased, while the levels of the proapoptotic p-p38MAPK pathway decreased [156].

The last peptide presented in this review is bradykinin-potentiating PEPTIDE-10C (BPP-10C) that represented low molecular mass fraction from Bothrops jararaca (Bj) toxins known as argininosuccinate synthetase (AsS) activators, improving l-arginine availability [46]. Generally, bradykinin potentiating peptides (BPPs) are oligopeptides rich in proline that display bradykinin-potentiating activity, including neuroprotective potential [129]. Pre-treatment with BPP-10C (0.1−10 μM) significantly reduced H2O2-induced cell death, ROS generation and lipid peroxidation, iNOS expression, and nitrate levels [131]. Furthermore, BPP-10c protected the mitochondrial membrane against oxidation. Possible neuroprotective mechanisms of BPP-10C was involved with increasing AsS expression in an oxidative stress condition and improving the l-arginine synthesis. At the same time, its metabolites preferentially lead to agmatine formation with a neuroprotective property. A similar study was conducted with other fraction of Bj toxin, also containing proline-rich peptides, and the most promising neuroprotective compounds were Bj-PRO-5a and Bj-PRO-7a [130]. Under the oxidative stress condition, both peptides reduced the ROS production, MDA levels, GSSG formation, and tGSH levels, therefore were able to prevent H2O2 cytotoxicity by inhibiting oxidative products. However, the neuroprotective effects observed in the treatments with these peptides were independent of AsS activity.

Read full article

URL: https://www.sciencedirect.com/science/article/pii/S019697812100036X

1.1 Protein Structure and Reactivity

Amino Acids

Peptides and proteins are composed of amino acids polymerized together through the formation of peptide (amide) bonds. The peptide-bonded polymer that forms the backbone of polypeptide structure is called the α-chain. The peptide bonds of the α-chain are rigid planar units formed by the reaction of the α-amino group of one amino acid with the α-carboxyl group of another (Figure 2.1). The peptide bond possesses no rotational freedom due to the partial double-bond character of the carbonyl–amino amide bond. The bonds around the α-carbon atom, however, are true single bonds with considerable freedom of movement.

Figure 2.1. Rigid peptide bonds link amino acid residues together to form proteins. Other bonds within the polypeptide structure may exhibit considerable freedom of rotation.

The sequence and properties of the amino acid constituents determine protein structure, reactivity, and function. Each amino acid is composed of an amino group and a carboxyl group bound to a central carbon, termed the α-carbon. Also bound to the α-carbon are a hydrogen atom and a side chain unique to each amino acid (Figure 2.2). There are 20 common amino acids found throughout nature, each containing an identifying side chain of particular chemical structure, charge, hydrogen bonding capability, hydrophilicity (or hydrophobicity), and reactivity. The side chains do not participate in polypeptide formation and are thus free to interact and react with their environment.

Figure 2.2. Individual amino acids consist of a primary (α) amine, a carboxylic acid group, and a unique side-chain structure (R). At physiological pH, the amine is protonated and bears a positive charge, while the carboxylate is ionized and possesses a negative charge.

Amino acids may be grouped by type depending on the characteristics of their side chains. There are seven amino acids that contain aliphatic side chains, which are relatively nonpolar and hydrophobic: glycine, alanine, valine, leucine, isoleucine, methionine, and proline (Figure 2.3). Glycine is the simplest amino acid, its side chain consisting of only a hydrogen atom. Alanine is next in line, possessing just a single methyl group for its side chain. Valine, leucine, and isoleucine are slightly more complex with three or four carbon branched-chain constituents. Methionine is unique in that it is the only reactive aliphatic amino acid, containing a thioether group at the terminus of its hydrocarbon chain. Proline is actually the only imino acid. Its side chain forms a pyrrolidine ring structure with its α-amino group. Thus, it is the only amino acid containing a secondary α-amine. Due to its unique structure, proline often causes severe turns in a polypeptide chain. Proteins rich in proline, such as collagen, have tightly formed structures of high density. Collagen also contains a rare derivative of proline, 4-hydroxyproline, found in only a few other proteins. Proline, however, cannot be accommodated in normal α-helical structures, except at the ends where it may create the turning point for the chain. Poly-proline α-helical structures have been formed, but the structural characteristics of these artificial polypeptides are quite different from native protein helices.

Figure 2.3. Common aliphatic amino acids.

Phenylalanine and tryptophan contain aromatic side chains that, like the aliphatic amino acids, are also relatively nonpolar and hydrophobic (Figure 2.4). Phenylalanine is unreactive toward common derivatizing reagents, whereas the indolyl ring of tryptophan is quite reactive, if accessible. The presence of tryptophan in a protein contributes more to its total absorption at 275 to 280 nm on a mole-per-mole basis than any other amino acid. The phenylalanine content, however, adds very little to the overall absorbance in this range.

Figure 2.4. The two aromatic amino acids that have nonpolar and nonionizable side-chain groups.

All of the aliphatic and aromatic hydrophobic residues are often located at the interior of protein molecules or in areas that interact with other nonpolar structures such as lipids. They usually form the hydrophobic core of proteins and are not readily accessible to water or other hydrophilic molecules.

There is another group of amino acids that contains relatively polar constituents and are thus hydrophilic in character. Asparagine, glutamine, threonine, and serine (Figure 2.5) are usually found in hydrophilic regions of a protein molecule, especially at or near the surface where they can be hydrated with the surrounding aqueous environment. Asparagine, threonine, and serine are often found post-translationally modified with carbohydrate in N-glycosidic (asp) and O-glycosidic linkages (thr and ser). Though these side chains are enzymatically derivatized in nature, the hydroxyl and amide portions have relatively the same nucleophilicity as that of water and are therefore difficult to modify with common reagent systems under aqueous conditions.

Figure 2.5. The four amino acids with polar, uncharged side chains. The arrows show the attachment points for carbohydrate that may be present in post-translational modifications on glycoproteins.

The most significant amino acids for modification and conjugation purposes are the ones containing ionizable side chains: aspartic acid, glutamic acid, lysine, arginine, cysteine, histidine, and tyrosine (Figure 2.6). In their unprotonated state, each of these side chains can be potent nucleophiles to engage in addition reactions (see the discussion on nucleophilicity below).

Figure 2.6. The amino acids with ionizable side chain groups possess some of the most important functional groups for bioconjugate applications. The C- and N-terminals of each polypeptide chain also are included in this group.

Both aspartic and glutamic acids contain carboxylate groups that have similar ionization properties to the C-terminal α-carboxylate. The theoretical pKa of the β-carboxyl of aspartic acid (3.7–4.0) and the γ-carboxyl of glutamic acid (4.2–4.5) are somewhat higher than the α-carboxyl groups at the C-terminal of a polypeptide chain (2.1–2.4). At pH values above their pKa, these groups are generally ionized to negatively charged carboxylates. Thus, at physiological pH, they contribute to the overall negative charge contribution of an intact protein (see following section).

Carboxylate groups in proteins may be derivatized through the use of amide bond-forming agents or through active ester or reactive carbonyl intermediates (Figure 2.7). The carboxylate actually becomes the acylating agent to the modifying group. Amine-containing nucleophiles can couple to an activated carboxylate to give amide derivatives. Hydrazide compounds react similarly to amines. While a thiol group is reactive toward an activated carboxylate and results in a thioester linkage, it forms relatively unstable derivatives, which can exchange with other nucleophiles such as amines or hydrolyze in aqueous solutions.

Figure 2.7. Derivatives of carboxylic acids can be prepared through the use of active intermediates that react with target functional groups to give acylated products.

Lysine, arginine, and histidine have ionizable amine-containing side chains that, along with the N-terminal α-amine, contribute to a protein’s overall net positive charge. Lysine contains a straight four-carbon chain terminating in a primary amine group. The ε-amine of lysine differs in pK1 from the primary α-amines by having a slightly higher ionization point (pKa of 9.3–9.5 for lysine versus pKa of 7.6–8.0 for α-amines). At pH values lower than the pKa of these groups, the amines are generally protonated and possess a positive charge. At pH values greater than the pKa, the amines are unprotonated and contribute no net charge. Arginine contains a strongly basic chemical constituent on its side chain called a guanidino group. The ionization point of this residue is so high (pKa>12.0) that it is virtually always protonated and carriers a positive charge. The histidine side chain is an imidazole ring that is potentially protonated at slightly acidic pH values (pKa=6.7–7.1). Thus, at physiological pH, these residues contribute to the overall net positive charge of an intact protein molecule.

The amine-containing side chains in lysine, arginine, and histidine typically are exposed on the surface of proteins and can be derivatized with ease. The most important reactions that can occur with these residues are alkylation and acylation (Figure 2.8). In alkylation, an active alkyl group is transferred to the amine nucleophile with loss of one hydrogen. In acylation, an active carbonyl group undergoes addition to the amine. Alkylating reagents are highly varied and the reaction with an amine nucleophile is difficult to generalize. Acylating reagents, however, usually proceed through a carbonyl addition mechanism as shown in Figure 2.9. The imidazole ring of histidine is also an important reactive species in electrophilic reactions, such as in iodination using radioactive 125I or 131I (Chapter 12, Section 2).

Figure 2.8. Derivatives of amines can be prepared from acylating or alkylating agents to give amide, secondary amine, or tertiary amine bonds.

Figure 2.9. The mechanism of acylation proceeds through the attack of a nucleophile, generating a tetrahedral intermediate, which then goes on to form the product.

Cysteine is the only amino acid containing a sulfhydryl group. At physiological pH, this residue is normally protonated and possesses no charge. Ionization only occurs at high pH (pKa=8.8–9.1) and results in a negatively charged thiolate residue. The most important reaction of cysteine groups in proteins is the formation of disulfide crosslinks with another cysteine molecule. Cysteine disulfides (called cystine residues) are often key points in stabilizing protein structure and conformation. They frequently occur between polypeptide subunits, creating a covalent linkage to hold two chains together. Cysteine and cystine groups are relatively hydrophobic and usually can be found within the core of a protein. For this reason, it is often difficult to fully reduce the disulfides of large proteins without a deforming agent present to open up the inner structure and make them accessible (see Section 4.1).

Cysteine sulfhydryls and cystine disulfides may undergo a variety of reactions, including alkylation to form stable thioether derivatives, acylation to form relatively unstable thioesters, and a number of oxidation and reduction processes (Figure 2.10). Derivatization of the side chain sulfhydryl of cysteine is one of the most important reactions of modification and conjugation techniques for proteins.

Figure 2.10. Sulfhydryl groups may undergo a number of additional reactions, including acylation and alkylation. Thiols also may participate in redox reactions, which generate reversible disulfide linkages.

Tyrosine contains a phenolic side chain with a pKa of about 9.7 to 10.1. Due to its aromatic character, tyrosine is second only to tryptophan in contributing to a protein’s overall absorptivity at 275 to 280 nm. Although the amino acid is only sparingly soluble in water, the ionizable nature of the phenolic group makes it often appear in hydrophilic regions of a protein—usually at or near the surface. Thus, tyrosine derivatization proceeds without much need for deforming agents to further open protein structure.

Tyrosine may be targeted specifically for modification through its phenolate anion by acylation, through electrophilic reactions such as the addition of iodine or diazonium ions, and by Mannich condensation reactions. The electrophilic substitution reactions on tyrosine’s ring all occur at the ortho position to the –OH group (Figure 2.11). Most of these reactions proceed effectively only when tyrosine’s ring is ionized to the phenolate anion form.

Figure 2.11. Tyrosine residues are subject to nucleophilic and electrophilic reactions. The unprotonated phenolate ion may be alkylated or acylated using a variety of bioconjugate reagents. Its aromatic ring also may undergo electrophilic addition using diazonium chemistry or Mannich condensation, or be halogenated with radioactive isotopes such as 125I.

In summary, protein molecules may contain up to nine amino acids that are readily derivatizable at their side chains: aspartic acid, glutamic acid, lysine, arginine, cysteine, histidine, tyrosine, methionine, and tryptophan. These nine residues contain eight principal functionalities with sufficient reactivity for modification reactions: primary amines, carboxylates, sulfhydryls (or disulfides), thioethers, imidazoles, guanidinyl groups, and phenolic and indolyl rings. All of these side-chain functionalities in addition to the N-terminal α-amino and the C-terminal α-carboxylate form the full complement of polypeptide reactivity within proteins (Figure 2.12).

Figure 2.12. The more important polypeptide functional groups are represented by these nine amino acids. Bioconjugate chemistry may occur through the C- and N-terminals of each polypeptide chain, the carboxylate groups of aspartic and glutamic acids, the ε-amine of lysine, the guanidino group of arginine, the sulfhydryl group of cysteine, the phenolate ring of tyrosine, the indol ring of tryptophan, the thioether of methionine, and the imidazole ring of histidine.

Nucleophilic Reactions and the pI of Amino Acid Side Chains

Ionizable groups within proteins can exist in one of two forms: protonated or unprotonated. Carboxylate groups below their pKa values exist in the protonated state and are therefore in the conjugate acid form and carry no charge. However, at pH values above the pKa of the carboxylic group, the acid is ionized and therefore unprotonated to a negative charge. This same relationship is true of the –OH group on the phenol ring of tyrosine. At pH values below its pKa, tyrosine’s side chain is uncharged. Above the pKa, however, the hydrogen ionizes off, leaving a negatively charged phenolate. Conversely, amine nucleophiles below their pKa values are in a protonated state and possess a positive charge. At pH values above the pKa of the amino group, it is then ionized and unprotonated to neutrality.

Each type of ionizable group in proteins will have a unique pKa based upon the theoretical value for the amino acid and modulated from that value by its own surrounding microenvironment. Minute environmental changes will cause amine-containing residues at different structural locations to have different ionization potentials, even if the groups are otherwise chemically identical.

Thus, the actual pKa of each ionizable group within protein molecules may range considerably lower or higher than the theoretical values as the microenvironment of individual groups changes. Identical side chains in differing parts of a protein molecule may have widely varying pKa values depending on the immediate chemical milieu. Such factors as the presence of other amino acid side chains in the vicinity, salts, buffers, temperature, ionic strength, and other effects of the solvent medium all play crucial roles in creating microenvironmental changes that affect the ionization potential of these groups (Tanford and Hauenstein, 1956; Schewale and Brew, 1982).

The Henderson–Hasselbalch equation (1) explains the relationship of pH and pKa to the relative ratios of protonated (acid) and unprotonated (base) forms of an ionizable group. Note that the ionized form of such a group does not have to possess a negative charge, as in the case of unprotonated primary amines. Indeed, in that instance it is the protonated amine that bears a charge of positive one. According to the mathematical implications of this equation, an ionizable group at its pKa value is exactly 50% ionized. This means that aspartic acid side chains placed in a medium with a pH equal to its pKa should have half of its carboxylates ionized to a negative charge and half of them unionized with no charge.

(1)pH=pKa+log{[base]/[acid]}

Further implications of this equation are that at one pH unit below or above the pKa, an ionizable group will be 91% unionized (protonated) or 91% ionized (unprotonated), respectively. Two pH units below or above translate to a 99% unionized or 99% ionized state.

The absolute ratio of protonated : unprotonated forms will change from this theoretical approach based upon the microenvironment each group experiences. The reactivity of amino acid side chains is directly related to them being in an unprotonated or ionized state. Many reactions of modification and conjugation occur efficiently only when the nucleophilic species is in an ionized form. As the unprotonated form increases in concentration, the relative nucleophilicity of the ionizable group increases. Many of the reactive groups commonly used for protein modification will couple in greater yield as the pH of the reaction is raised closer to the pKa of the ionizable target. However, continuing to increase the pH beyond the pKa may not be necessary for increased yield, and may even be detrimental, because many reactive groups will begin to lose activity through hydrolysis at high pHs.

A nucleophile is any atom containing an unshared pair of electrons or an excess of electrons able to participate in covalent bond formation. Nucleophilic attack at an atomic center of electron deficiency or positive charge is the basis for many of the coupling reactions that occur in chemical modification. Thus, an uncharged amine group is a more powerful nucleophile than the protonated form bearing a positive charge. Likewise, a negatively charged carboxylate has greater nucleophilicity than its uncharged, protonated conjugate acid form. In addition, an unprotonated thiolate, bearing a negative charge (RS−), is a much more powerful nucleophile than its protonated, uncharged sulfhydryl form.

According to the theory of nucleophilicity (Edwards and Pearson, 1962; Bunnett, 1963; Pearson et al., 1968), the relative order of nucleophilicity relative to the major groups in biological molecules can be summarized as follows:

R-S−>R-SH

R-NH2>R-NH3+

R-COO−>R-COOH

R-O−>R-OH

R-OH=H-OH

and finally,

R-S−>R-NH2>R-COO−=R-O−

Using these relationships, it is obvious that the strongest nucleophile in protein molecules is the sulfhydryl group of cysteine, particularly in the ionized, thiolate form. Next in line are the amine groups in their uncharged, unprotonated forms, including the α-amines at the N-terminals, the ε-amines of lysine side chains, the secondary amines of histidine imidazolyl groups and tryptophan indole rings, and the guanidino amines of arginine residues. Finally, the least potent nucleophiles are the oxygen-containing ionizable groups including the α-carboxylate at the C-terminal, the β-carboxyl of aspartic acid, the γ-carboxyl of glutamic acid, and the phenolate of tyrosine residues.

According to the theoretical pKa values for the ionizable side chains of amino acids, nucleophilic substitution reactions involving primary amines or sulfhydryl groups on proteins should not be efficient below a pH of about 8.5 (Table 2.1). In practice, however, reactions can be carried out with these groups in high yield at pH values not much higher than neutrality. This discrepancy relates to the changes in pKa due to microenvironmental effects experienced by the residues within the three-dimensional structure of the protein molecule. In reality, the ε-amine groups on lysine side chains within proteins, having theoretical pKas of over 10, nonetheless exist in sufficient quantity in an unprotonated form even at a pH of 7.2 that modification easily occurs.

Table 2.1. pKa of Ionizable Amino Acids

Group Location	Functionality	pKa Range
α-Amine; N-terminus		7.6–8
Lysine’s ε-amine		9.3–9.5
Histidine’s imidazolyl nitrogen		6.7–7.1
Arginine’s guanidinyl group		>12
Tyrosine’s phenolic hydroxyl		9.7–10.1
α-Carboxyl; C-terminus		2.1–2.4
Aspartic acid’s γ-carboxyl		3.7–4
Gutamic acid’s γ-carboxyl		4.2–4.5
Cysteine’s sulfhydryl		8.8–9.1

One important point should be noted, however. The changes that occur in the pKa of ionizable groups in protein molecules due to microenvironmental effects sometimes make it difficult to select certain residues for modification simply by careful modulation of reaction pH. For instance, at least in theory, overlap of the pKa range for sulfhydryls and amine-containing residues would eliminate any chance of directing a reaction toward –SH groups solely by adjusting the pH of the reaction medium. However, because of the microenvironmental changes that occur in complex biomolecules, pH sometimes can be used along with the right reactive group to target thiols without amine modification. Thus, in practice, to effectively site-direct a modification reaction, the proper choice of reactive group and reaction conditions can result in highly discrete conjugation to certain sites within proteins.

Secondary, Tertiary, and Quaternary Structure

Amino acids are linked through peptide bonds to form long polypeptide chains. The primary structure of protein molecules is simply the linear sequence of each residue along the α-chain. Each amino acid in the chain interacts with surrounding groups through various weak, noncovalent interactions and through its unique side-chain functionalities. Noncovalent forces such as hydrogen bonding and ionic and hydrophobic interactions combine to create each protein’s unique organization.

The sequence and types of amino acids and the way that they are folded provide protein molecules with specific structure, activity, and function. Ionic charge, hydrogen bonding capability, and hydrophobicity are the major determinants for the resultant three-dimensional structure of protein molecules. The α-chain is twisted, folded, and formed into globular structures, α-helicies, and β-sheets based upon the side-chain amino acid sequence and weak intramolecular interactions such as hydrogen bonding between different parts of the peptide backbone (Figure 2.13). Major secondary structures of proteins such as α-helicies and β-sheets are held together solely by massive hydrogen bonding created through the carbonyl oxygens of peptide bonds interacting with the hydrogen atoms of other peptide bonds (Figure 2.14).

Figure 2.13. The α-chain structure of alkaline phosphatase illustrates the complex nature of polypeptide structure within proteins (Kim and Wyckoff, 1991).

Figure 2.14. Secondary structures within proteins may be stabilized through hydrogen bonding between adjacent α-chains, forming β-sheet conformations.

In addition, negatively charged residues may become bonded to positively charged groups through ionic interactions. Nonpolar side chains may attract other nonpolar residues and form regions of hydrophobicity to the exclusion of water and other ionic groups. Occasionally, disulfide bonds are also found holding different regions of the polypeptide chain together. All of these forces combine to create the secondary structure of proteins, which is the way the polypeptide chain folds in local areas to form larger, sometimes periodic structures.

On a larger scale, the unique folding and structure of one complete polypeptide chain is termed the tertiary structure of protein molecules. The difference between local secondary structure and complete polypeptide tertiary structure is arbitrary and sometimes of little practical difference.

Larger proteins often contain more than one polypeptide chain. These multi-subunit proteins have a more complex shape, but are still formed from the same forces that twist and fold the local polypeptide. The unique 3-dimensional interaction between different polypeptides in multi-subunit proteins is called the quaternary structure. Subunits may be held together by noncovalent contacts, such as hydrophobic or ionic interactions, or by covalent disulfide bonds formed from the cysteine residue of one polypeptide chain being crosslinked to a cysteine sulfhydryl of another chain (Figure 2.15).

Figure 2.15. Polypeptide chains may be bound together through disulfide linkages occurring between cysteine residues within each subunit.

Thus, aside from the covalently polymerized α-chain itself, the majority of protein structure is determined by weaker, non-covalent interactions that potentially can be disturbed by environmental changes. It is for this reason that protein structure can be easily disrupted or denatured by fluctuations in pH or temperature or by substances that can alter the structure of water, such as detergents or chaotropes.

Not surprisingly, chemical modification to the amino acid constituents of a polypeptide chain may also cause significant disruption in the overall three-dimensional structure of a protein. If amino acid residues critical to folding near functionally important regions are modified with chemical groups that change the charge, hydrophilicity, or hydrogen bonding character of the polypeptide chain, protein structure may be altered and activity may be compromised. This concept will be discussed further in subsequent sections.

Prosthetic Groups, Cofactors, and Post-Translational Modifications

Proteins may contain structures other than polypeptide chains that are important for biological function. Prosthetic groups and cofactors are small organic compounds that are sometimes tightly bound to a protein and aid in forming the active center. A prosthetic group is usually carried within the three-dimensional protein structure in a firm-fitting pocket or even attached through covalent bonds, such as the heme ring associated with cytochrome c molecules which is bonded through thioether linkages with adjacent cysteine residues (Figure 2.16). Cofactors, by contrast, may be bound only transiently to proteins during periods of activity. Enzymes often require cofactors to act as donors or acceptors of chemical groups that are added to or cleaved from a substrate molecule. Some common cofactors are ATP, ascorbic acid, coenzyme A, NAD, NADP, FAD, FMN, and biotin. Sometimes, the enzyme cofactor is also an energy source for the catalytic reaction, as in the case of ATP dependent reactions.

Figure 2.16. The heme ring of cytochrome c is a non-amino-acid, prosthetic group bound to the protein through two cysteine residues.

Frequently, metal ions are associated with the prosthetic group or cofactor. Heme rings usually contain a chelated iron atom. Occasionally, however, these metals are merely bound within folded polypeptide regions with no additional organic constituents required. Many metal ions are known to participate in enzymatic activity. One or more of the ions of Na, K, Ca, Zn, Cu, Mg, and Mn, as well as Co and Mo, are often required by enzymes to maintain activity.

Prosthetic groups and cofactors, whether organic or metallic, may be removed from a protein to create an inactive apo protein or enzyme. Loss of these groups may occur through environmental changes, such as removing metal ions from solution or adding denaturants to unfold protein structure. In many cases, simply re-introducing the needed group into the surrounding medium can restore full activity.

In addition to small organic molecules or metal ions, proteins may have other components tightly associated with them. Nucleoproteins, for instance, contain noncovalently bound DNA or RNA, as in some of the structural proteins of viruses. Lipoproteins contain associated lipids or fatty acids and may also carry cholesterol, as in the high-density and low-density lipoproteins in serum.

During modification or conjugation reactions, prosthetic groups and other associated molecules may be lost or damaged. Metal ions temporarily may be removed by the inclusion of a chelating agent added to maintain sulfhydryl stability during coupling through the –SH groups of a protein. To restore activity after conjugation, it is necessary to remove the chelator and add the required metal salts. Other changes to the prosthetic carriers may not be so easily corrected. For instance, heme-containing molecules are sensitive to the presence of agents that can form a coordination complex with or modify the oxidation state of the chelated metal ion. Some reagent systems may permanently inactivate the heme-containing protein.

Thus, loss of activity can occur not only through changes to the amino acid constituents of a protein, but through prosthetic group or cofactor loss or damage as well. Most of these potential difficulties can be overcome through careful selection of the reaction conditions and through knowledge of the cofactor dependencies that are critical to the activity of the protein being modified.

Post-translational modifications to protein structure are covalent changes that occur as the result of controlled enzymatic reactions or due to chemical reactions not under enzymatic regulation. One of the most common cellular modifications performed on proteins after ribosomal synthesis is glycosylation. Proteins newly synthesized on ribosomes may be transported to the Golgi apparatus, where specific glycosyl transferases catalyze the coupling of carbohydrate residues to the polypeptide chains. Glycoproteins and mucoproteins are formed by the coupling of polysaccharides through O-glycosidic linkages to serine, threonine, or hydroxylysine and through N-glycosidic linkages with the amide side-chain group of asparagine.

The structure of most glycoprotein carbohydrate is branched with the sugars mannose, N-acetylglucosamine (GlcNAc), sialic acid, glactose, and l-fucose being prevalent. Asparagine-linked polysaccharides are well characterized and are known to be constructed of a core unit consisting of three mannose residues and two N-acetylglucosamine residues. The GlcNAc residues are bound to the Asp side-chain amide nitrogen through a β1 linkage (Kornfield and Kornfield, 1985). The three mannose groups then usually form the first branch point in the oligosaccharide chain (Section 2.1.2).

The content by weight of carbohydrate in glycoproteins may vary from only a few percent to over 50% in some proteins in mucous secretions. Although the function of the polysaccharide in most glycoproteins is unknown, in some cases it may provide hydrophilicity, recognition, and points of non-covalent interaction with other proteins through lectin-like affinity binding.

The presence of carbohydrate on protein or peptide molecules can provide important points of attachment for modification or conjugation reactions. Coupling exclusively through polysaccharide chains can often direct the reaction away from active centers or critical points in the polypeptide chain, thus preserving activity. Polysaccharides can be specifically targeted on glycoproteins through mild sodium periodate oxidation. Periodate cleaves adjacent hydroxyl groups in sugar residues to create highly reactive aldehyde functionalities (Chapter 3, Section 4.4). The level of periodate addition can be adjusted to selectively cleave only certain sugars in the polysaccharide chain. For instance, a concentration of 1-mM sodium periodate at temperatures less than 4°C specifically oxidizes sialic acid residues to contain aldehydes, leaving all other monosaccharides untouched. Increasing the concentration to 10-mM and carrying out the reaction at room temperature, however, will cause oxidation of other sugars in the carbohydrate chain, including galactose and mannose. The generated aldehydes can then be used in coupling reactions with amine-or hydrazide-containing molecules to form covalent linkages. Amines can react with formyl groups under reductive amination conditions using a suitable reducing agent such as sodium cyanoborohydride. The result of this reaction is a stable secondary amine linkage (Chapter 3, Section 5.3). Alternatively, hydrazides spontaneously react with aldehydes to form hydrazone linkages, although the addition of a reducing agent increases the efficiency of the reaction (Chapter 3, Section 5.1).

Another form of post-translational modification that may add carbohydrate to a polypeptide is non-enzymatic glycation. This reaction occurs between the reducing ends of sugar molecules and the amino groups of proteins and peptides. See Section 2.1 in this Chapter for further details and the reaction sequence behind this modification.

Protecting the Native Conformation and Activity of Proteins

The goal of most protein modification or conjugation procedures is to create a stable product with good retention of the native state and activity. Ideally, any derivatization should result in a protein that performs exactly as it would in its unmodified form, but with the added functionality imparted by whatever is conjugated to it. Thus, an antibody molecule tagged with a fluorophore should retain its ability to bind to antigen and also have the added functionality of fluorescence.

One of the best ways to ensure retention of activity in protein molecules is to avoid carrying out chemistry at the active center. The active center is that portion of the protein where ligand, antigen, or substrate binding occurs. In simpler terms, the active center (or active site) is that part that has specific interaction with another substance (Means and Feeney, 1971). For the preparation of enzyme derivatives, it is important to protect the site of catalysis where conversion of substrate to product happens. For instance, when working with antibody molecules, it is crucial to stay away from the two antigen binding sites.

The best chemical procedures avoid the active site by selecting functional groups away from that area or by protecting the site through the incorporation of additives. In some cases, the inclusion of substrates, cofactors, ligands, inhibitors, or antigens in the modification reaction will protect the active site. Addition of the appropriate substance can bind the active site and mask it from modification by crosslinking agents. In enzyme derivatization procedures, this is often just a matter of adding a reversible inhibitor or substrate analog. For instance, when working with alkaline phosphatase merely carrying out the reaction in phosphate buffer protects the active center from chemical modification, since phosphate ions bind in the catalytic site. With trypsin, the incorporation of benzamidine similarly masks and protects the active site.

However, protecting the antigen binding sites on an antibody molecule by using this method is often more difficult. Inclusion of antigen to mask the binding sites is effective in blocking these areas, but it may also cause irreversible crosslinking of the antigen to the antibody. This is especially true when the antigen is a peptide or a protein having the same chemical functionalities as the antibody. Any modification reactions that are directed at the antibody may modify the antigen as well. Therefore, only use this method if the antigen is lacking in the chemical targets that are going to be used on the antibody. For instance, if the polysaccharide chains on the antibody are targeted for modification, then using a protein antigen that does not contain carbohydrate to block the antigen binding sites may work well.

An equally effective method of protecting the activity of a protein is by using site-directed reactions that result in modifications away from the active center. In some cases, specific functionalities are known to be present only at restricted sites within the three-dimensional structure of a protein. If these functionalities are not present close to the active site, then using them exclusively for modification reactions should ensure good retention of activity. For instance, sulfhydryl groups or carbohydrate chains are often present in limited quantity and in specific regions on a protein. Selecting reagent systems that target these groups ensures derivatization only at restricted sites within the protein molecule, thus potentially avoiding the active center.

Surprisingly, the goal of some protein crosslinking schemes is to somewhat alter the native presentation of the conjugate. This is especially true in hapten–carrier conjugation as used for immunogen or vaccine preparation. In this case, the main objective is to modify the environment of the hapten to create an immunological response in vivo. A hapten is usually a small molecule that is not able to generate an immune response on its own, but can react with the products of such a response once generated. Most often these products are antibodies having binding specificity for the hapten.

The complexities involved in achieving a successful conjugation strategy are best illustrated in the problems and concerns dealing with hapten–carrier conjugation. In order to produce the initial immune response to a small molecule, the hapten is typically coupled to a larger protein that can generate a response on its own. In simple terms, the larger carrier protein confers immunogenicity to the smaller hapten. The native presentation of the hapten is altered toward the immune system, thus creating the immune response.

The site of attachment of the hapten to the carrier and the nature of the crosslinker are both important to the specificity of the resultant antibodies generated against it. For proper recognition, the hapten must be coupled to the carrier with the appropriate orientation. For an antibody subsequently to recognize the free hapten without the attached carrier, the hapten–carrier conjugate must present the hapten in an exposed and accessible form. Optimal orientation is often achieved by directing the crosslinking reaction to specific sites on the hapten molecule. With peptide haptens, this is typically carried out by attaching a terminal cysteine residue during synthesis. This provides a free thiol group on one end of the peptide for conjugation to the carrier. Crosslinking through this group provides hapten attachment only at one end, therefore ensuring consistent orientation.

In hapten–carrier conjugation, the goal is not to maintain the native state or stability of the carrier, but to present the hapten in the best possible way to the immune system. In reaching this goal, the choice of conjugation chemistry may control the resultant titer, affinity, and specificity of the antibodies generated against the hapten. It may be important in some cases to choose a crosslinking agent containing a spacer arm long enough to present the antigen in an unrestricted fashion. It may also be important to control the density of the hapten on the surface of the carrier. Too little hapten substitution may result in little or no response. A hapten density that is too high may actually cause immunological suppression and decrease the response. In addition, the crosslinker itself may generate an undesired immune response. Fortunately, for the majority of hapten–carrier conjugation problems, a few main crosslinking techniques provide a workable compromise to solving all these concerns and ultimately generating an effective immune response (Chapter 19).

Oxidation of Amino Acids in Proteins and Peptides

The modification of amino acids in proteins and peptides by oxidative processes plays a major role in the development of disease and in aging (Kim et al., 1985; Tabor and Richardson, 1987; Halliwell and Gutteridge, 1989, 1990Halliwell and Gutteridge, 1989Halliwell and Gutteridge, 1990; Stadtman, 1992). Tissue damage through free-radical oxidation is known to cause various cancers, neurological degenerative conditions, pulmonary problems, inflammation, cardiovascular disease, and a host of other problems. Oxidation of protein structures can alter activity, inhibit normal protein interactions, modify amino acid side chains, cleave peptide bonds, and even cause crosslinks to form between proteins.

Due to their abundance in cells relative to other biological molecules, proteins are one of the primary targets of oxidation in vivo. However, sometimes oxidation reactions involving proteins and peptides are thought of solely as the creation of disulfides from thiols on cysteine residues. This is certainly an important form of oxidation that can affect protein structure and function or even cause problems relevant to bioconjugation reactions. The presence of an accessible free thiol on a protein in an aqueous solution can be highly unstable to rapid oxidation unless precautions are taken to prevent disulfide formation. Dissolved oxygen and other potentially catalytic components, such as certain metal salts, can quickly result in disulfides being formed within a protein or between different protein molecules.

From a broader perspective, protein oxidation can result in covalent modification at many sites other than just at cysteine thiols. The earliest reports on protein oxidation date from the first decade of the 20th century, but it took many more years to characterize these reactions and their products (Dakin, 1906).

The significance of protein oxidation became paramount with the advent of recombinant protein biologics used as human therapeutics. Careful characterization of protein stability is essential to maintaining the efficacy of protein pharmaceuticals. If even a single side-chain amino acid residue becomes oxidized, then a protein therapeutic may not have the same activity in vivo as the unmodified protein.

Oxidation of proteins can result from exposure to oxidative species from many sources: reactive oxygen intermediates caused by metabolic reactions within cells (mitochondrial electron transport function and certain enzymes, such as oxidases, peroxidases, and P-450 enzymes), from the byproducts of oxidative stress reactions in cells (Sayre et al., 2001), or through the presence of strongly oxidizing compounds within a solution—all of these can contribute to selective damage or modification to protein structures. Some examples of chemical agents that can oxidatively modify proteins include hydrogen peroxide (H2O2) and other peroxy compounds, such as perborate and peroxycarbonate; hydroperoxyl radical (HO2·); superoxide anion (O2·−); singlet oxygen (1O2); hydroxyl radical (·OH), periodate (IO4−); metal salts in the presence of oxygen species, such as those of iron (Fe3+ and Fe2+) and copper (Cu2+); ozone (O3); peroxynitrite (ONOO−); hypobromous acid (HOBr); hypochlorous acid (HOCl); performic acid (HC(O)OOH); trichloromethylperoxyl radical (CCl3OO·); under the right conditions, metal-chelating compounds, such as porphyrins, texaphyrins, and FeBABE; and gamma radiation and ultraviolet light. For additional information, see Winterbourn and Kettle, 2000; Baynes and Thorpe, 2000; Greenacre and Ischiropoulos, 2001; Halliwell and Gutteridge, 1989, 1990Halliwell and Gutteridge, 1989Halliwell and Gutteridge, 1990; and Stadtman, 1992.

Singlet oxygen (1O2) differs from the predominant oxygen molecule in that O2 is in the ground state or triplet state and its outer two unshared electrons have parallel spins (sometimes designated 3O2), which is nearly unreactive toward other molecules, while singlet oxygen has increased energy and has its outer electrons transformed into an opposite spin orientation, which is highly reactive. Superoxide (O2·−) is different from singlet oxygen in that it is a reduced form of oxygen having an extra unpaired electron, called a radical. The presence of the radical makes superoxide extremely reactive and highly damaging to proteins and other biological molecules.

Singlet oxygen and superoxide, in addition to their modifying effects on proteins, are also important reactive oxygen species in biological applications, as they are intermediates used in some detection methods and in photodynamic therapy (PDT) for the treatment of cancer. One of the more common compounds used in PDT is Photofrin, which is a mixture of oligomers consisting of ether and ester linkages that combine up to eight porphyrin groups (Misawa et al., 2005). The generation of reactive oxygen species takes place by irradiation with 630-nm wavelength laser light, which also penetrates the skin effectively during therapy. Photoactivation of the Photofrin molecule causes radical initiation to form porphyrin-excited states. Transfer of electrons from the porphyrin groups to molecular oxygen then generates the highly reactive singlet oxygen species. Subsequent radical reactions can also form superoxide and hydroxyl radicals, all of which severely damage tissue in the region of the tumor and ultimately cause cancer cell death.

Another compound used to generate reactive oxygen species for PDT is texaphyrin, which contains a metal-chelating ring structure resembling a porphyrin group (Figure 2.17). Typically, a gadolinium atom is chelated in the texaphyrin center and this complex becomes both a photosensitizing agent and an MRI contrast agent to better visualize tumor locations for irradiation therapy (Donnelly et al., 2004).

Figure 2.17. The texaphyrin–gadolinium chelate structure used as a photosensitizer and MRI contrast agent in the detection and treatment of cancer.

The reactive oxygen species involved with protein oxidation can be generally categorized according to their relative reactivity as follows:

HO⋅,HO2⋅>O2⋅2>ROOH,H2O2>1O2,ClO−,BrO−>O2

Thus, radicals are the most reactive and destructive of protein structure, followed by peroxy derivatives, singlet oxygen, and other oxygen compounds. The oxidative reactivity of some of these oxygen species is so high that just contact of the pure compound with paper or cotton fabrics can cause combustion (e.g., superoxide).

In vitro studies of protein oxidation indicate that virtually all proteins and peptides are susceptible to damage by the radicals ·OH and O2·−. Analysis of protein modification products after oxidation indicates the presence of altered molecular weight (either fragmentation or oligomerization), altered net charge, tryptophan destruction, and the formation of tyrosine dimers (Davies, 1987). Even in the presence of very low concentrations of oxidants (nM), SDS polyacrylamide gel electrophoresis of proteins can indicate multiple bands of higher and lower molecular weight due to oxidative damage.

Transition metals in solution can catalyze the formation of reactive oxygen species that are particularly damaging to proteins and other biomolecules. In a series of reactions, reduced transition metals, such as Fe2+ and Cu1+, can be oxidized by oxygen to produce superoxide and ultimately undergo a Fenton reaction to create hydroxyl radicals (Kim et al., 1985). Transition metal-chelating groups can accelerate this reaction, as demonstrated in the process of hydroxyl radical footprinting of protein interactions using EDTA chelates of iron (see discussion on the reagent FeBABE in Chapter 24, Section 4).

Production of superoxide: Fe2++O2→Fe3++O2⋅−

Production of hydrogen peroxide:2O2⋅−+2H+→H2O2+O2

Production of hydroxyl radical:Fe2++H2O2→Fe3++OH⋅+OH−

The potential sites of oxidation within a protein molecule include the peptide backbone and the side-chain amino acid groups. Hydrogen atom abstraction at the alpha carbon of an amino acid chain can occur upon reaction with an oxidative species to form a radical intermediate. Subsequent reaction can result in peptide bond cleavage and fragmentation of the protein structure, often forming carboxylic acids or carbonyls (aldehydes or ketones). This is the basic mechanism of fragmentation caused by the bifunctional chelating reagent FeBABE. When used in the presence of H2O2 and ascorbic acid, polypeptides will fragment in the neighborhood of interacting proteins.

Amino acid side chains can undergo oxidation through hydrogen abstraction, elimination, or by addition reactions. In the presence of oxygen, aliphatic amino acids usually experience oxidation to a peroxy intermediate, which causes either hydrogen atom abstraction or an elimination reaction resulting in the formation of carbonyls, hydroxyls, or other peroxides (Requena et al., 2001) (Figure 2.18). Aromatic amino acids typically undergo addition reactions following exposure to strong oxidants. An example of this type of reaction is the nitrosation of tyrosine groups in the presence of a peroxynitrite (ONOO−) to create o-nitrotyrosine (see Figure 2.19).

Figure 2.18. Reaction of proline, arginine, and lysine residues with hydroxyl radical results in oxidation of side-chain structures that form carbonyls. Both arginine and proline oxidation will result in the same product being formed.

Figure 2.19. Tyrosine and phenylalanine residues can undergo oxidation to modify their phenyl side chain groups. Tyrosine can form covalent dimers that link two side chains together via a radical reaction. Both tyrosine and phenylalanine can be modified by oxidation to add oxygen-containing groups directly to their aromatic ring.

After exposure to an oxidant, the potential types of oxidation products in proteins and peptides can be extensive (Stadtman and Levine, 2000). Cysteine and methionine undergo a variety of sulfur oxidation reactions to yield cysteine disulfides, methionine sulfoxide, methionine sulfone, and sulfonate products (e.g., cysteic acid) (Ghesquiere et al., 2011) (Figure 2.20). Oxidation with performic acid can be used purposely to convert methionine and cysteine in peptides and proteins to more stable products prior to acid hydrolysis and amino acid analysis. Cysteine and methionine are perhaps the most sensitive amino acids to oxidation, and for this reason they are an early indicator of oxidative damage to proteins.

Figure 2.20. Cysteine and methionine are highly susceptible to oxidation reactions. Cysteine thiols can forms disulfide linkages with other cysteine groups or be oxidized to cysteic acid. Methionine is oxidized very easily to the sulfoxide or sulfone products.

Tyrosine is also easily modified through addition reactions due to the ring activating nature of its phenolic group. Using oxidants, tyrosine’s ring can be chlorinated or iodinated, undergo nitrosation or hydroxylation, and even form tyrosine–tyrosine crosslinks. The last product can be formed purposely by use of a peroxidase in the presence of hydrogen peroxide, and this type of reaction has been studied extensively in the manufacture of phenolic polymer resins (Dordick, 1991). In addition, Fancy et al. (1996) as well as Fancy and Kodadek (1997, 1998)Fancy and Kodadek (1997)Fancy and Kodadek (1998) have applied the oxidation of tyrosine to form dityrosine to the study of protein–protein interactions using nickel-chelated 6xHis-tagged fusion proteins in oxidative environments.

Nitrogen-containing side chains in amino acids can be altered by oxidation forming chloramines or even become deaminated. The result is often the formation of carbonyls (e.g., aldehydes) and hydroxyls. Lee et al. (2006) found that Fe-EDTA-mediated oxidation of human serum albumin resulted in extensive aldehyde and ketone formation from modification of lysine, arginine, histidine, proline, threonine, and aspartic and glutamic acids. Some groups will oxidize and convert to another amino acid altogether. For instance, histidine can be converted to asparagine and proline to hydroxyproline, and tyrosine can be changed to dihydroxyphenylalanine (DOPA) through oxidation reactions.

It is obvious that the oxidation of protein molecules can have detrimental effects on protein structure and function. However, there are some unique methods in bioconjugation wherein controlled and purposeful oxidation is carried out to study protein–protein interactions (Chapter 24, Section 4).

Unfortunately, there are no universal methods to detect all types of protein oxidation, because the products formed can be so diverse in nature. However, some forms of protein oxidation can be assayed using chemical modification (Davies et al., 1999; Shacter, 2000). In particular, the formation of carbonyl groups on proteins can be targeted using the reagent 2,4-dinitrophenylhydrazine (DNPH). This compound reacts with aldehydes to form 2,4-dinitrophenylhydrazone derivatives, which create chromogenic modifications that can be detected at high sensitivity in microplate assays or western blot analysis (Buss et al., 1997; Winterbourn et al., 1999).

In addition, a method involving mass spec analysis to determine carbonyl formation as a result of protein oxidation was developed using a novel mass tag. The carbonyl-specific Element-Coded Affinity Mass Tag (O-ECAT) can covalently couple to aldehyde or ketone oxidation sites using an aminoxy group to form an oxime (Lee et al., 2006; see also Chapter 12, Section 3). The ECAT mass tag consists of a bifunctional metal-chelating group that coordinates a lanthanide metal ion of specific mass. Proteins that have been oxidized to contain carbonyls can be labeled with this reagent and the exact sites of modification determined by analyzing the mass spec signature of the labeled peptides after proteolysis.

Solvent Accessibility of Functional Targets in Proteins

Proteins are highly complex, folded polypeptide chains consisting of at least 20 different amino acids that are strung together in unique sequences, which relate to structure and function. Particular amino acids in proteins may be further modified post-translationally to contain a wide variety of covalent modifications normally found in native proteins. The way in which a peptide chain is wrapped and folded governs each amino acid’s relative exposure to the outside environment, but post-translational modifications can also obscure the protein surface from easy access to the solvent environment.

Amino acid side chains are the primary effectors of the three-dimensional structure of a protein, because their properties vary depending on the presence of charged groups, uncharged polar components, aliphatic chains, aromatic rings, and groups able to form hydrogen bonds with other amino acid residues. The relative hydrophilicity or hydrophobicity of an amino acid side chain is a major factor in determining whether the group will be found on the surface of a globular protein or buried within its globular structure.

However, just considering the individual properties of each amino acid type is not enough to determine its accessibility to the surrounding aqueous environment. There have been many attempts at developing analytical models with predictive value for determining buried or surface accessible amino acids in a folded polypeptide chain (Vasicek et al., 2012). These studies have concluded fractional assignments for each residue that relate to its accessible surface area (ASA) or its solvent exposed area (SEA).

In most cases, there are general trends that emerge from theoretical studies in which hydrophilic amino acids are more likely to be found on the surface of a protein and hydrophobic amino acids are more likely to be inside its three-dimensional structure, but we already knew this intuitively so this conclusion is not surprising. However, a real-life study of the positions of amino acids in proteins whose structures are known is more revealing. The data for Figure 2.21 were calculated from 55 proteins in the Brookhaven database by Bordo and Argos (1991), and the graph was derived from the analysis as presented by the Jena Image Library of Biological Macromolecules (http://www.imb-jena.de/IMAGE_AA.html). Although most of these structures were determined using crystallographic means and thus the proteins are “frozen” in a single structural state, the results are revealing as to how often particular amino acids are accessible to the surrounding solvent.

Figure 2.21. Comparison of the solvent exposed surface area of amino acids in proteins. Data are plotted as a percentage of each amino acid in a protein having greater than a 30-Å2 exposure to the aqueous environment. Charged and polar amino acids are seen to have the most solvent exposure, while uncharged, aromatic, or aliphatic amino acids have the least exposure.

Three levels of SEA are presented in the graph for each amino acid, which corresponds to areas in Å2 accessible to the solvent environment: greater than 30 Å2 for highly accessible amino acids, between 10 Å2 and 30 Å2 for medium accessibility, and less than 10 Å2 for those residues that are relatively not accessible to the solvent. Only the SEA for each amino acid of>30 Å2 is shown in the plotted data. The graph shows that the polar amino acids such as serine, threonine, asparagine, glutamine, and tyrosine often have large areas accessible to the solvent, as do the charged amino acids aspartic acid, glutamic acid, lysine, histidine, and arginine. Surprisingly, proline also falls in the highly accessible group, which is not as expected, because it does not carry a charge, nor is it a highly polar amino acid. However, proline does have a unique characteristic that may explain its appearance on the surface of proteins: it cannot freely rotate about its imino group as other amino acids in a peptide chain can do at amide bonds. This effect results in a kink in the polymer (called a beta-turn), and these sharp turns in a peptide backbone probably occur most often near the surface. Thus, proline is found to be frequently accessible to the solvent environment despite its hydrophobic nature.

The nonpolar amino acids glycine, alanine, valine, leucine, isoleucine, methionine, phenylalanine, and cysteine have lower exposure to the solvent environment than charged or polar residues. However, the frequency at which these groups are found to have an SEA of greater than 30 Å2 is much higher than one would expect based solely upon consideration of their hydrophobicity. In fact, nearly 30 to 50% of the time nonpolar amino acids in a protein can be found at the surface.

At the two extremes, lysine is observed as the amino acid most accessible on the surface of proteins while cysteine is the least exposed amino acid. The inaccessibility of cysteine probably stems from the fact that disulfides are typically buried within the polypeptide structure of proteins, whether they are intrachain or in nature, and proteins rarely contain many reduced cysteine thiols.

It is clear from these data that proteins have complex hydrophilic and hydrophobic regions on their surfaces that determine their potential interactions, binding sites, and active centers. For bioconjugation purposes, targeting of an amino acid even with a high SEA for modification or crosslinking may not result in every residue being modified that is theoretically present in a protein based only on knowledge of its amino acid composition. Even when coupling to very polar or charged groups, such as lysine, there are varying degrees of accessibility to a given reagent, because of the complex folding of the polypeptide chains at the protein surface.

Figure 2.22 shows the globular structure of an immunoglobulin (IgG) Fc region to illustrate this point. In this space-filling model, the lysine residues are highlighted in yellow to easily show their locations within the two polypeptides of the heavy chains. Notice that some of the ε-amino groups at the ends of the side chains are protruding far out into the solvent and are therefore highly accessible for modification. Some of these groups, however, are less exposed even though they are still near the surface, and a few lysines are seen to be between the heavy chain regions where it would be difficult to modify them due to crowding.

Figure 2.22. The solvent accessibility of lysine residues in the Fc region of an antibody is illustrated by highlighting the lysine groups in yellow. Some lysine ε-amine groups are extremely accessible to conjugation, while others are only partially exposed, thus making it difficult to modify all of them in bioconjugation reactions.

Figure 2.64 in this Chapter provides data to validate this effect. The reaction of the thiolating reagent SATA with IgG resulted in only a percentage of the available lysines being modified. As the molar ratio of SATA to IgG was increased, the yield of lysine modification actually became lower. This result can be explained by the relative accessibility of each lysine in the immunoglobulin structure. Some residues are easily accessible and they get modified with high yield even with low molar ratios of SATA-to-IgG. As the molar ratio is increased, it gets more difficult to modify those lysines that are less accessible to the solvent environment or are partially obscured by another polypeptide chain. Thus, the solvent accessibility of particular amino acids is a major factor in whether they can be effectively targeted and modified with a given bioconjugate reagent.

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B9780123822390000029

Peptides that act via the carpet mechanism and remain in contact with the acidic phospholipid head groups, should have the following properties: (i) their net charge needs to be highly positive and spread along the peptide chain. (ii) they should bind very weakly, or not bind zwitterionic membranes and as a consequence are not hemolytic. It should be noted however, that highly positively charged antimicrobial peptides can lyse erythrocytes if they form oligomers in solutions (see previous paragraphs). (iii) no preferable structure is required, as long as a certain level of hydrophobicity and number of positive charges are preserved.

Read full article

URL: https://www.sciencedirect.com/science/article/pii/S0196978101004983

5.510.2 Proteins

Proteins, or polypeptides, are ubiquitous and essential components in a functional tissue environment and participate in nearly every cellular process. Therefore, several classes of proteins are targeted as promising candidates for the engineering of bioactive materials and tissue engineering scaffolds. Proteins are polymer structures composed of 20 distinct types of monomer units (amino acids) linked by amide bonds. The properties of complex protein-containing tissues and organs are exclusively determined by the physicochemical properties of these monomers and their sequence. Therefore, these natural biopolymers present unique opportunities for the engineering of a rich complexity of structure and function by controlling sequence, hierarchical organization, and long-range interactions in much the same way as synthetic polymers.1

5.510.2.1 Extracellular Matrix Proteins

A large class of molecules relevant to this type of protein engineering is the extracellular matrix (ECM) proteins. The ECM is a tough, elastomeric network composed of various proteinaceous fibrils and fibers interconnected in a hydrated network of glycosaminoglycan (GAG) chains. These components of the ECM provide mechanical stability and structural integrity to tissues and organs by resisting tensile and compressive stresses. The ECM also contains biochemical cues that promote cellular adhesion, motility, and tissue production, and provides mechanical stimuli to cells through its intimate connections with the cellular cytoskeleton via cell surface receptors (Figure 1). A significant challenge in the fields of biomaterials and tissue engineering is the recreation of this exact microenvironment at the implantation site to promote favorable tissue anchorage and growth. The most promising approaches rely on engineering scaffold materials that mimic the native ECM proteins that contain functional information for promoting cell adhesion, proliferation, and differentiation.2,3

Figure 1. Classes of bioactive molecules that have been engineered for applications in biomaterials, tissue engineering, and regenerative medicine include (1) extracellular matrix proteins on the exterior of the cell, (2) small polypeptides that mimic the epitopes in the extracellular matrix that are involved in receptor binding, (3) soluble or matrix-bound growth factors, (4) plasmid DNA or (5) viral gene carriers that enter the cell and travel to the nucleus to deliver genetic material, (6) small RNAs delivered by polymeric drug carriers into the cytoplasm to disrupt mRNA stability and translation, (7) liposomes that act as drug delivery vehicles, and (8) second messengers that transmit signals from extracellular stimuli to the nucleus to modulate changes in gene expression that are regulated by (9) transcription factors.

Natural ECM proteins perform a wide array of functions in their native context, including structural integrity, cellular adhesion and communication, storage, and catalysis. Their diverse set of functional properties makes them ideal templates for the engineering of bioactive molecules that closely mimic the native system. Collagen is one of the most promising biomaterials in regenerative medicine and tissue engineering as it is the predominant component of extracellular matrices in mammals, providing mechanical support to various connective tissues, such as skin, bone, tendon, ligament, cartilage, and blood vessels. Collagens, characterized by their unique triple-helical structures, not only provide significant structural integrity to tissues and organs, but also regulate a number of biological events, including cell attachment, migration, survival, proliferation, differentiation, and tissue regeneration. Many of these diverse functions are controlled through collagen-targeted cell surface receptors, such as integrins, discoidin domain receptors (DDR1 and DDR2), and annexin V, which promote cell adhesion and specific cellular functions.4–6 The rational design of collagen-based biomaterials and tissue engineering scaffolds often incorporates the binding sequences from the native collagen molecule that are recognized by these various cell membrane receptors. In addition, it has been shown that cell adhesion-mediated signaling pathways are affected not only by the recognition sequence bound by the receptor, but also by the collagen molecule's triple-helical structure and assembly in the three-dimensional matrix.7 For example, studies have demonstrated that fibrillar collagen regulates early integrin signaling differently from monomeric collagen, with the fibrils resulting in the inhibition of smooth muscle cell proliferation.8 Therefore, the rational engineering of bioactive collagen-based materials often incorporates the specific receptor sequences from the native collagen molecule as well as particular forms, structures, and tertiary configurations of the collagen matrix that are most conducive to the design cell function and tissue performance.

Many of collagen's diverse functions are controlled by specific interactions of collagen-binding molecules (cell membrane receptors, soluble factors, and other matrix molecules) with certain structures and/or sequences displayed in the context of the collagen triple helices. Thus, synthetic triple-helical peptides that mimic the structure of native collagens and incorporate receptor-binding sequences have been used to design bioactive surfaces and scaffold substrates. For example, several studies indicate that the α2β1 integrin interaction with type I collagen is a crucial signal for the induction of bone cell (osteoblast) differentiation and matrix mineralization.9–12 The integrin α2β1 recognizes the Gly–Phe–Hyp–Gly–Glu–Arg (GFOGER) motif in residues 502–507 of the α1[I] chain of type I collagen.13 Reyes and colleagues demonstrated that an α2β1 integrin-specific GFOGER-peptide triggers the activation of focal adhesion kinase (FAK) and alkaline phosphatase in osteoblast cells – two proteins that have been implicated in the osteoblastic differentiation pathway.14 In this case, the functional activity of the peptide is entirely dependent on the engineering of the GFOGER recognition sequence in the context of a triple-helical peptide structure. These GFOGER-peptide surfaces also support the expression of multiple osteoblast-specific genes, including osteocalcin and bone sialoprotein, and induce calcification and matrix mineralization in a manner similar to type I collagen, demonstrating that this engineered bioactive peptide is a promising surface modification strategy for the design of collagen-mimetic surfaces that support bone formation. In fact, the GFOGER peptide has been used to dramatically enhance the osseointegration properties of titanium implants in animal models.15

Similarly, Ito and colleagues have rationally engineered collagen-like proteins with characteristics tailored specifically for cartilage engineering.16 They identified the D4 region of human type II collagen as critical for the migration of chondrocytes and genetically engineered a collagen-like protein consisting of tandem repeats of the D4 domain assembled into a triple-helical structure. Scaffolds coated with this protein provided favorable culture conditions for chondrocytes, promoting migration, matrix formation, and maintenance of the chondrocyte phenotype.

Artificial self-polymerizing or self-assembling collagen-like materials have also been designed to mimic collagen suprastructures. Paramonov and colleagues designed collagenous peptide polymers of the form poly(Xaa–Yaa–Gly)n.17 These triple-helical polymers form nanofiber-like structures and can be engineered with various side-chain functional groups for binding. Cys residues on the polymers allow cross linking and additional functional modification. Koide et al. and Kotch and Rain also describe a system for obtaining collagen-like triple-helical supramolecules using a self-assembly process with self-complementary trimers of Pro–Hyp–Gly repeats.18,19 Peptides are tethered by cysteine knots in a staggered arrangement to form elongated triple helices.

In addition to collagen, several other proteins are commonly found in the ECM and are promising templates for the design of bioactive molecules. The abundant ECM protein elastin, found in arteries, lung, and ligament, has unique elastomeric mechanical properties because of alternating hydrophilic and hydrophobic domains and multiple lysine cross-linking sites that allow for significant extensions and subsequent elastic recoil. Elastin-like polypeptides (ELP) have been derived by engineering repeat hydrophobic motifs from mammalian elastin, the most common being (VPGXG)n, where X is any nonproline amino acid.20,21 These polypeptides exhibit mechanical properties similar to those of native elastin, but are particularly suited for tissue engineering applications because they are thermally sensitive and undergo a reversible, inverse temperature phase transition. They have been used as thermally responsive injectable hydrogels that are water soluble below their transition temperature and self-aggregating above their transition temperature to form a gel-like, space-filling scaffold. Design principles from the field of polymer science can be applied to adjust the phase transition temperature to the desired in vivo application.21 In addition, highly controlled genetic engineering strategies may be applied to design the precise amino acid sequence at the genetic level, allowing for the introduction of cross-linking sites that affect network modulus and reactive sites that allow the addition of functional ligands to control cell behavior or enzymatic recognition sites to control degradation.22,23 The combination of polymer design principles and genetic engineering techniques in this system allows for the exquisite control of the physical properties of the scaffold network and the bioactivity and biocompatibility of the material in in vivo applications.

To address tissue engineering for cartilage repair, Betre and colleagues demonstrated that chondrocytes encapsulated in an ELP aggregate gel maintain morphology and phenotype in vitro for up to 15 days.24 These cells grew and secreted the characteristic proteins of cartilage matrix, including type II collagen and sulfated GAG. They also demonstrated that ELP gels support the differentiation of human adult stem cells derived from adipose tissue into chondrocytes, as evidenced by the upregulation of genes that control the specific chondrocyte phenotype.25 Subsequently, these ELP gels have also been used for the controlled release of drugs and other therapeutic agents in the intra-articular26,27 and perineural28 environment.

5.510.2.2 Biomimetic Peptides

In addition to bioactive materials that recapitulate the unique and specific structures of large ECM proteins, such as collagen and elastin, the identification of small oligopeptide adhesive and signaling sequences within such proteins has led to the design of ligand-functionalized materials. These materials capitalize on the mechanical properties of a bulk substrate and the specific receptor-binding activity of an attached peptide ligand that promotes cell adhesion and subsequent cellular activity (Figure 1). Cell adhesion to ECM proteins controls complex biological processes such as development, wound healing, immune response, and tissue function through specific and dynamic regulation of cell behavior.29 Cell attachment to the ECM is primarily mediated by integrins, a widely expressed family of heterodimeric cell surface adhesion receptors consisting of noncovalently associated α- and β-subunits.29 In addition to anchoring cells and providing tissue structure, integrins transmit intracellular signals that direct cell migration, proliferation, cell cycle progression, and differentiation.30–32 Integrins thus function as the principal mediators of the molecular dialog between a cell and its ECM environment by specifically recognizing and binding to short amino acid motifs, or ligands, present in the ECM.

Engineering small bioactive peptides or ligands for immobilization onto a bulk substrate allows for specific tuning of integrin binding and activity by controlling ligand type, density, and spatial distribution (Figure 2). The use of the short peptide over the full-length native ECM protein also results in greater stability during the immobilization or modification process, as well as enhanced steric availability. The RGD (R: arginine; G: glycine; D: aspartic acid) sequence is one of the most effective and widely used peptide motifs for promoting cell adhesion on a synthetic surface. This tripeptide sequence was identified in 1984 by Pierschbacher and Ruoslahti as the minimal essential cell adhesion sequence in the ECM protein fibronectin.33 Since then, RGD sites have been indentified in several other ECM proteins, such as vitronectin, fibrinogen, collagen, laminin, von Willebrand factor, tenascin, as well as in a variety of membrane proteins.34 Nearly half of the 24 known integrins have been shown to bind to ECM proteins in some RGD dependent manner.34 Other peptide sequences, such as Tyr–Ile–Gly–Ser–Arg (YIGSR) and Ile–Lys–Val–Ala–Val (IKVAV) from laminin or Arg–Glu–Asp–Val (REDV) from fibronectin, have also been immobilized onto the surfaces of various biomaterials.35,36

Figure 2. Various design criteria to consider when engineering bioactive molecules into biomaterials for presentation to cells. Each of these parameters can affect the cellular response to the material, including cell differentiation and tissue regeneration.

The recognition sequence itself is not the only factor in designing a bioactive peptide. Many ligand sequences require certain conformational states or neighboring amino acids for optimal activity. Therefore, the design of the bioactive peptide sequence beyond the simple recognition sequence may allow for greater activity or tunable cell behavior. It has long been recognized that the RGD sequence loses affinity and specificity when presented outside the context of the native protein.33,37 In fact, the RGD tripeptide alone shows little to no effect in blocking cell attachment.38 Many efforts have been undertaken to rescue the activity of the RGD ligand by inserting flanking amino acids and adjusting the local conformation of the sequence (Figure 2). The activity of the sequence is improved by blocking the C-terminal carboxyl group on the tripeptide or by adding surrounding amino acids from the native sequence. For example, RGDS has higher activity than RGD and GRGDSP has higher activity that RGDS.38 Furthermore, Petrie and colleagues have recently published a series of studies demonstrating enhanced bioactivity of a larger fragment of fibronectin encompassing RGD and neighboring epitopes also implicated in integrin binding.37 This larger fragment has also been more effective in directing osseointegration of titanium implants.39,40 The conformation and stereochemistry of the RGD sequence has been shown to affect both activity and specificity for particular integrins.41–47 For example, a kinked RGD conformation favors αvβ3 integrin binding and a linear arrangement favors binding to the αIIbβ3 integrin.44 The RGD cyclopeptide shows even greater affinity and cell attachment activity compared to linear peptides, which is most likely due to the fact that RGD exists in an exposed loop conformation in ECM proteins such as fibronectin.48,49 Similarly, the Gly–Phe–Hyp–Gly–Glu–Arg (GFOGER) motif from type I collagen requires a triple-helical conformation to bind to the α2β1 integrin.50 Engineering the flanking environment and conformation of the recognition sequence in a bioactive peptide may allow precise control over a ligand's affinity and selectivity to various integrin subtypes, which in turn determines the cell types that preferentially interact with the implanted surface (Figure 2).

In the engineering of bioactive materials, there are a number of critical design factors that regulate the cellular interactions with the surface at a molecular level through receptor binding (Figure 2). Receptor binding subsequently determines the strength of cell adhesion,51–53 the rates of cell migration and tissue infiltration,54–56 and the extent of cytoskeletal formation and subsequent cell signaling.57,58 These design parameters include immobilized peptide density and spacer length, receptor-ligand affinity, and spatial distribution or concentration of the bioactive peptide in the biomimetic material, all of which have been shown to affect biological responses. Using immobilized bioactive peptides on model substrates, several investigators have demonstrated a clear dependence of cell adhesion and cellular function on peptide density.14,59–61 In fact, cell attachment as a function of ligand concentration shows a sigmoidal increase.50,62,63 Differences in cell adhesion activity using fibroblasts seeded onto RGD-functionalized glass substrates and RGD-functional hydrogels also indicate that the hydrophobic/hydrophilic nature of the substrate affects peptide spacing results.61,64

In the engineering of peptide surfaces, cell adhesion is not the only aspect of cell behavior that should be considered. While high concentrations of bioactive peptide may result in greater cell attachment, they may impede cell migration and proliferation, which would adversely affect tissue growth and infiltration. Palacek et al. describe the biphasic trend of cell migration in response to ligand density, with cell migration increasing and then decreasing with increasing ligand density.56 Similarly, Neff et al. demonstrate that fibroblast proliferation on RGD-modified polystyrene peaks at an intermediate concentration of RGD and then subsequently decreases with higher concentrations.65 Hubbell and coworkers have shown that the amount of in vitro neurite extension in fibrin gels modified with RGD peptides shows biphasic behavior, with an intermediate concentration of peptide leading to maximal values of extension.66,67 A similar biphasic response to adhesion peptides adsorbed onto biodegradable polymer scaffolds has also been observed for bone growth in vivo, with maximal values of growth at intermediate peptide densities.68 These and other similar results showing migration, proliferation, and matrix production peaking at intermediate ligand densities suggest that the concentrations of bioactive peptide should be optimized for the particular cell function of interest rather than simply for cell adhesion.67,69 Additionally, the incorporation of multiple, independent ligands that target unique receptors has been shown to result in synergistic biological activity.70

Integrin-mediated cell adhesion depends not only on receptor occupancy but also on receptor clustering (Figure 2). Therefore, peptide-modified surfaces may stimulate different cellular responses not only because of different peptide surface concentrations, as discussed above, but also because of the presentation of those peptides in a manner that allows or triggers integrin clustering or aggregation.71–74 Maheshwari et al. designed surfaces with controlled RGD peptide density and nanoscale spatial distribution and demonstrated that fibroblasts exhibited greater shear resistance, and well-formed stress fibers and focal adhesions when the RGD peptide was presented in a clustered rather than random distribution.75 Therefore, designing the precise spatial distribution of a bioactive molecule may contribute to greater control over cell attachment and function.

In addition to peptide density and distribution, biological activity of the peptide recognition sequence must be considered while engineering functionalized surfaces (Figure 2). The immobilized peptide must be flexible enough to interact with the target receptor in the proper orientation and must protrude from the bulk surface sufficiently enough to minimize steric hindrance. To achieve these requirements, nonspecific, flexible peptide chains, such as repeating glycines,50,76 or hydrophilic polymer chains, such as poly(ethylene glycol) (PEG),59,77 have been engineered between the substrate or matrix polymer and the bioactive signaling sequence. Hern et al. demonstrated that cells adhere to a greater extent to RGD-functionalized surfaces with a PEG spacer than to those without the spacer, even at high surface densities. Similarly, Shin et al. showed that the length of PEG peptide spacers affected marrow stromal cells attached to modified hydrogels. Beer et al. explained this need for a spacer using an integrin model derived from electron microscopy.78 The RGD peptide must be long enough to reach inside the globular binding pocket of the integrin. The length of the required RGD spacer varies in the literature, most likely because of differences in surface roughness or microconformations. However, cell attachment markedly decreases if that spacer length is too long because of the increased entropic energy of the long chain and the cell's preference for tight binding and rigid surfaces.79,80

Many of the aforementioned studies demonstrate several design principles that should be addressed while engineering bioactive molecules and materials (Figure 2). However, the optimal design parameters, such as peptide concentration and spacer length, vary depending on the cell type of interest, the desired cell function or phenotype, and the intrinsic material properties of the bulk matrix or substrate. The engineer must consider the interplay between all these criteria while designing a specific biomimetic system.

In addition to design factors that affect cell adhesion and functional activity, there are other considerations in the engineering of a bioactive peptide that may improve the ease of production or the long-term stability of the surfaces. In the case of covalent coupling of the ligand, including an accessible N-terminal lysine amino acid or other tethering moiety and avoiding bulky groups that may sterically hinder attachment will enhance coupling yields. Peptide solubility may also be tuned by adjusting the amino acids near the termini, for example, by adding charged side chains.81 The proteolytic resistance of the peptide may be enhanced by using d-amino acids or cyclic conformations.43,82

Finally, the biodegradability of the matrix must be considered (Figure 3). For materials that degrade, surface tethering strategies may be insufficient as the surface can be rapidly removed in vivo. In these cases, it may be advantageous to develop polymers that display the bioactive peptide in the bulk material rather than on the surface. New peptide would be displayed continually as the material degraded. This issue has been investigated using the biodegradable material poly(lactic acid). Langer and coworkers synthesized the copolymer poly(lactic acid-co-lysine) in which the amino groups on the lysyl comonomer provide sites for peptide tethering.83 Therefore, the bioactive peptide is presented on the transient surface through the degradation process. This type of bulk matrix incorporation is also relevant in the case of hydrogel scaffolds that undergo cellular infiltration. Migration-stimulating peptides, such as YIGSR and SIKVAV from laminin, have been impregnated in gels for nerve regeneration applications.84 This allows the infiltrating neuronal cells to encounter the bioactive signal at all times during the infiltration process.

Figure 3. Strategies for incorporating bioactive molecules into biomaterials. (a) The bioactive molecule may be immobilized onto the surface of an implant or device to direct the activity of cells at that interface. This allows for control over the density and geometry of the presentation of the molecule to the cells. (b) The molecule may be incorporated into the bulk of a biodegradable material, such that the molecule is released slowly as the material breaks down and is replaced by host tissue. (c) The advantages of both (a) and (b) can be combined by constructing biodegradable materials that present the bioactive molecules from defined chemical structures.

5.510.2.3 Growth Factors

In addition to the adhesion molecules discussed above, the ECM contains another class of morphogenetic molecules known as growth factors. These polypeptides are secreted from cells and subsequently bind to specific receptors on the surface of target cells and transmit signals that modulate cellular activities.85 In the ECM, growth factors are often bound to proteoglycans, such as heparan sulfate and chondroitin sulfate, which protect them from proteolytic degradation and regulate their bioavailability.86 These growth factors may be bound in the active state or may require enzymatic cleavage of the matrix or growth factor itself to reach maximal activity.87

Biomaterial and tissue engineers have begun to exploit this relationship between the matrix, growth factors, and cells to design systems for tissue repair and remodeling that promote cell growth, tissue production, and cellular invasion. While utilizing growth factors for these applications, it is critical that the functionality and activity of the immobilized growth factor are maintained during the matrix fabrication and release. The simplest approach has been to use an injectable hydrogel scheme in which the growth factor is incorporated into the matrix during the liquid-to-gel transition that occurs under physiological conditions.88 The aqueous environment inside these hydrogels allows for the diffusion of the growth factor through and from the matrix. These injectable hydrogels may be delivered with minimal invasive surgery and can conform to a variety of complex tissue shapes.

Another approach has been to develop hydrogels or solid-phase biopolymeric materials to mimic the growth factor-binding characteristics of the native ECM. As many growth factors, such as vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF), bind to heparan sulfate, several studies have sought to incorporate heparin, a molecule chemically similar to heparan sulfate, onto biomaterial surfaces in order to control the release of growth factors from implants.89 In this case, the dissociation kinetics of the growth factor from the heparin controls the overall rate of release from the implanted matrix.

Alginate gels are one of the more widely used hydrogels in injury and wound management. Several investigators have sought to heparinize alginate gels to allow growth factor retention and controlled release by covalently conjugating heparin molecules90 or encapsulating heparin-sepharose in the matrix.91–93 In one example, sustained release of bFGF from heparinized-alginate pellets resulted in significant improvement of coronary blood flow in a porcine model of myocardial ischemia.92

As previously mentioned, collagen matrices are attractive candidates for tissue engineering scaffolds because of their mechanical properties, and cell adhesion and signaling characteristics. However, these scaffolds often display poor growth factor loading and retention. Several techniques have been employed to improve the incorporation and release kinetics of growth factors from collagen, including covalent attachment of heparin or modification of the network to couple growth factor release to the rate of bulk matrix degradation. For example, Tabata et al. demonstrated that varying the cross-linking density of collagen by using glutaraldehyde slowed matrix degradation rates and prolonged VEGF retention and release.94 The same group showed that cross-linked gelatin (denatured collagen) hydrogels with bFGF incorporation by electrostatic interaction implanted into rabbit cranial defects enhanced bone regeneration.95–97

One of the more interesting engineered matrices for tissue engineering is fibrin, the provisional matrix formed during tissue repair. Because this matrix functions as a temporary scaffold for tissue repair and regeneration, its physiological role is to promote cellular ingrowth; it is gradually replaced by a more stable ECM upon healing. These are precisely the characteristics that are desired in an implantable tissue engineering scaffold. Sakiyama-Elbert and Hubbell have developed a system in which molecules can be linked onto a fibrin scaffold during the network cross-linking process.98 They have used this technique to covalently incorporate a heparin-binding peptide into the fibrin matrix. This peptide binds heparin, which in turn immobilizes growth factors to the scaffold.99,100 With this system, they demonstrated the controlled release of several nerve growth-stimulating growth factors from a fibrin matrix with a high excess of heparin-binding sites. Lee et al. and Taylor et al. subsequently showed that a fibrin matrix with bound neurotrophin-3 (NT3) enhanced regeneration of the sciatic nerve and the spinal cord.101,102

The same investigators also demonstrated that growth factors themselves can be engineered to bind fibrin directly without the heparin intermediate.103,104 In this case, they developed a fusion protein consisting of factor XIIIa that binds fibrin at one terminus, a central domain containing proteolytic cleavage sites, and the bioactive growth factor at the other terminus. These fusion proteins exploit the proteolytic activity of plasmin and matrix metalloproteinases in the wound site to liberate the bioactive growth factor from the matrix by clipping the central linker.

In some cases, covalent tethering of the growth factors imparts a biological advantage compared to the incorporation of the soluble molecule. Griffith and coworkers demonstrated that surface-tethered epidermic growth factor (EGF) promotes both mesenchymal stem cell spreading and survival more strongly than saturating concentrations of soluble EGF. The same group showed that covalent tethering of EGF reduces epidermic growth factor receptor (EGFR) internalization and restricts signaling to the cell surface, enhancing osteogenic differentiation of bone marrow stromal cells.105 This group also demonstrated that EGF conjugated to synthetic polymer surfaces is capable of directing hepatocytes to maintain their liver-specific morphology and function.106

These studies demonstrate that engineering principles may be applied to both the biomaterial matrix and the growth factor of interest to manipulate and optimize activity, concentration, and release kinetics.

5.510.2.4 Transcription Factors

Growth factors and ECM proteins function by inducing cellular signaling pathways that ultimately lead to changes in gene expression in the nucleus (Figure 1). These changes in gene expression are mediated by transcription factors that bind to specific gene sequences and regulate the activation of that locus. Several studies in the area of tissue engineering are now focusing on direct modification of gene regulation by the genetic engineering of cells with transcription factors, rather than by direction of cell activity with extracellular stimuli. For example, García and colleagues have genetically engineered cells to express the transcription factor Runx2 to enhance osteoblastic differentiation and bone formation.107–110 Incorporation of this strategy into an inducible gene expression system led to exogenous control of the magnitude and kinetics of cell differentiation and tissue formation in vitro and in vivo.111 Furthermore, they have discovered novel mutations of Runx2 that mimic natural posttranslational modifications and lead to bone formation by otherwise recalcitrant cell types.112

Transcription factor-based strategies have also been prevalent in approaches for wound healing and therapeutic angiogenesis. In particular, Hif-1α is a transcription factor that regulates genes responsible for angiogenesis, including the growth factors VEGF and platelet-derived growth factor (PDGF). Trentin et al. delivered a stabilized form of Hif-1α, in which a natural degradation domain had been removed, to mouse dermal wounds, using a peptide-based gene delivery vector.113 This led to enhanced angiogenesis relative to direct delivery of VEGF alone, as demonstrated by more stable and mature blood vessels. Similarly, stabilized Hif-1α showed enhanced angiogenesis relative to the VEGF in skeletal muscle.114 Variants of Hif-1α have also been engineered to incorporate the strong transactivation domain from the herpes simplex virus (HSV) VP16 protein to enhance its transactivation properties. These Hif-1α/VP16 engineered molecules have been successful in treating animal models of limb ischemia115 and have subsequently moved into clinical trials.116

In addition to naturally occurring transcription factors, it is now possible to engineer synthetic transcription factors that target-specific genes within mammalian genomes.117 For example, artificial zinc finger proteins have been designed to regulate a variety of genes related to tumor development and sensitivity to cancer therapy.118–121 One particularly successful application of this approach has been in the development of an artificial transcription factor that upregulates the endogenous VEGF gene.122 This molecule was able to protect against experimental diabetic neuropathy and has now moved into clinical trials for a variety of indications.123

5.510.2.5 Methods for Synthesizing Engineered Proteins and Peptides

There are two general strategies for engineering bioactive peptide- and protein-based molecules – recombinant DNA expression and synthetic polymerization. The recombinant DNA approach uses genetically encoded sequences to generate the desired polypeptide. The engineer must first design a recombinant DNA segment that precisely encodes for the therapeutic protein of interest. This segment is then cloned into a suitable DNA plasmid to create a recombinant DNA molecule that can be delivered to cells with the appropriate machinery for expressing the desired protein. This plasmid is transformed into the host cells, which are then grown in culture for a prescribed length of time to allow for protein expression. The protein of interest is then purified from the cells or culture media. A critical step in this process is the generation of the gene of interest. For small polypeptides, short DNA sequences can be directly synthesized using standard solid-phase synthesis. For long proteins with repeating units, including many structural ECM proteins such as elastin, these shorter DNA sequences can be assembled into larger proteins that encode for repetitive structures of the desired molecular weight.20 This can be accomplished by concatenation of oligonucleotides, recursive directional ligation (RDL), or amplification of gene segments by polymerase chain reaction (PCR).20 Concatenation is a statistical method in which a distribution of genes encoding the monomer is ligated and the gene inserts are sorted on the basis of size. The desired insert is then ligated into the vector of choice. In RDL, the short DNA segments are linked using recombinant DNA techniques. The resulting larger genes can then be recursively combined until a gene of a desired length is obtained. This approach is modular and can be used to combine genes encoding different polypeptide sequences.124,125 This method also provides a greater level of control over the size of the insert.

Recombinant protein expression has many advantages over chemical synthesis. It provides well-defined sequences, stereochemistry, and molecular weight based on the engineered genetic template.20 The resultant transformed cells can be stored to provide an ongoing supply of the protein of interest. Finally, and most important for the generation of structurally complex bioactive peptides and proteins, the intrinsic protein folding machinery of the cell has the ability to form physiologically relevant secondary and tertiary structures. For example, integrin α5β1-binding requires both the PHSRN sequence in the ninth type III repeat and the RGD motif in the tenth type III repeat of the ECM protein fibronectin.126 Because of the acute sensitivity of α5β1 to small perturbations in the orientation of these domains, the development of proper structural orientation using synthetic peptides remains a challenge. Recombinant molecular techniques have been successfully applied to recapitulate the native secondary and tertiary structure of this important fibronectin domain.37 The disadvantages of these techniques include the time- and labor-intensive process of assembling the gene and cloning it into the desired vector and then optimizing expression conditions in the host cell line. In addition, these techniques are often limited to the standard 20 amino acids, and the incorporation of unnatural amino acids is a difficult and often inefficient venture. The addition of nonpeptide moieties, such as a synthetic polymer backbone or insert, is also not possible with this approach unless it is done after expression and purification. Finally, in many cases the desired polypeptide is not expressed well in the host cell line, making it difficult to obtain the requisite amounts of purified protein.

Part of the design process of a recombinantly expressed bioactive molecule includes the development of a strategy for protein expression and purification that will optimize the yield and quality of the final product. These conditions will vary according to the sequence, folding structure, and posttranslational modifications of the desired protein. Based on these criteria, different expression hosts are available, including bacteria, insect cells, and mammalian cells. Once a host is selected, the final bulk protein yield can be maximized by optimizing the expression conditions.

Alternatively, the synthetic polymerization approach uses standard techniques such as solid-phase peptide synthesis or α-amino acid N-carboxyanhydride (NCA) polymerization to produce the desired protein from individual amino acid monomers. In solid-phase synthesis, small porous beads are functionalized with linkers upon which peptide chains can be built, amino acid by amino acid. The peptide remains covalently attached to the bead until it is clipped by a reagent such as trifluoroacetic acid. This allows the growing peptide to be retained during several filtration steps, wherein the liquid-phase reagents and reaction by-products are washed away. The general process of solid-phase synthesis includes repeated cycles of coupling and deprotection. The free N-terminal amine of the bead-attached peptide is coupled to a single N-protected amino acid moiety. This amino acid is then deprotected, revealing a new N-terminal amine to which the next amino acid can be linked. The disadvantages of this technique include low yields and an upper size tolerance of around 70–100 amino acids. Beyond this length, this approach is burdened by unavoidable deletions and truncations that result from incomplete deprotection and coupling steps.127

The most advantageous process for synthesizing high molecular weight peptide chains with significant yields in large quantities is the polymerization of NCAs.127 In this technique, a nucleophile or base is used to initiate the polymerization of NCA amino acid monomers. The most common initiators used are primary amines and alkoxide anions.127 The reagents required are simple and inexpensive; thus NCA polymerization is a very economical and expedient way to generate long polypeptides. The disadvantages of conventional NCA polymerizations include the fact that there is no control over the reactivity of the growing polymer chain end during the polymerization reaction.127 After initiation, the resulting primary amine, carbamate, or NC anion end group can undergo a host of undesirable side reactions. New developments in the use of transition metal catalysis, however, can lead to products that, in many ways, rival biologically produced counterparts in terms of complexity and purity.127 Studies have demonstrated the production of block copolypeptides of controlled molecular weight, sequence, composition, and molecular weight distribution with the use of this technique.

Although the recombinant genetic engineering approach may offer greater control over sequence, these synthetic methods allow for the incorporation of artificial or nonstandard amino acids, such as hydroxyproline, and for the creation of alternative stereochemistries, which can prevent protease-mediated degradation. Synthetic techniques also allow for the generation of hybrid molecules that consist of polypeptides fused or merged with other synthetic polymers.

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B9780080552941001641

Peptides from Ant Venoms

Most of the ants only present traces of proteins/peptides in their venoms; however, in those species producing peptides as components of their venoms, these toxins follow the same general pattern already observed for peptides from wasps and bees venoms, that is, short and linear polycationic peptides, presenting a high content of α-helices, which are responsible for cell lysis, hemolysis, histamine release from mast cells, and antimicrobial actions. A few neurotoxins are also known.

“Myr p” Peptides and Pilosulin

The venom of the Australian ant Myrmecia pilosula seems to contain a complex mixture of allergenic peptides, some of them forming heterodimers, apparently maintained by disulfide bridges. From this complex was identified Myr p 1, the major expressed allergenic product being a 56 residue peptide, whereas Myr p 2 is a 27 residue peptide; these peptides may dissociate, or are cleaved to minor fragments, such as those previously identified as pilosulin-1 and pilosulin-2, characterized as the fragments 57–112 from Myr p1 and 49–75 from Myr p2.5 Pilosulin-1 is a α-helical peptide with potent and broad spectrum of antimicrobial activity, both against standard and multidrug-resistant Gram-positive and Gram-negative bacteria, and Candida albicans.29 Other pilosulins have been identified and characterized as hemolysins and histamine release peptides.12,18,25

Poneratoxins

Poneratoxin is a neuropeptide found in the venom of the ant Paraponera clavata. A peptide containing 25 amino acid residues, it affects voltage-dependent sodium channels and blocks synaptic transmission in the insect central nervous system in a concentration-dependent manner. Its NMR structure is in the form of two α-helices connected by a β-turn (Fig. 1C); the helices have quite different characteristics from each other: One of them is apolar, whereas the second contains polar and charged amino acids. This will result in different interactions with cell membranes. The extremely hydrophobic N-terminal helix may interact with uncharged lipid bilayers, while the C-terminal helix, slightly positively charged and terminating with arginine, will be able to attach to negatively charged cell surfaces as previously found for other membrane interacting peptides. Such a toxin can thus use two different complementary modes of interaction to attain its target, cellular membranes.18,23

Ponericins

Ponericins constitute a group of peptides isolated from the venom of the predatory ant Pachycondyla goeldii, exhibiting hemolysis, insecticidal activity against cricket larvae, and antimicrobial action against Gram-positive and Gram-negative bacteria. According to their primary structure similarities, they can be classified into three families: ponericin G, W, and L. Ponericins share high sequence similarities with known peptides: ponericins G with cecropin-like peptides, ponericins W with gaegurins and melittin, and ponericins L with dermaseptins. The comparison of the structural features of ponericins with those of well-studied peptides suggests that the ponericins may adopt an amphipathic α-helical structure in polar environments, such as cell membranes.13

Ectatomin

This peptide is a neurotoxin isolated from the venom of the Ectatomma tuberculatum ant and contains two highly homologous peptide chains (consisting of 37 and 34 amino acid residues) linked to each other by disulfide bonds.25 Each chain consists of two alpha-helices and a hinge region of four residues forming a hairpin structure that is stabilized by disulfide bridges; the hinge regions of the two chains are connected together by a third disulfide bridge (Fig. 1D). Thus, ectatomin forms a four-alpha-helical bundle structure.18 Ectatomin is a potent inhibitor of calcium currents after a latency of a few seconds in rat ventricular myocites.25

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B9780123850959000580

4.08.8 Overview of the Nonribosomal System

Many drugs in use today are derived from natural products. Investigations reveal that many of these produced by various microorganisms16 are of peptide origin and are structurally complex in nature. Such compounds are composed of an oligopeptide skeleton having extensive modifications, such as the incorporation of d-amino acids, hydroxy acids, and other unusual pseudo amino acids, by use of peptide synthetases.

The majority of peptides originate from Gram-positive bacteria and filamentous fungi. Among them Bacilli and Actinomycetes are leading producers. However, bouvardin from the medicinal plant Bouvardia terniflora, lophyrotomin, a benzoylated octapeptide toxin from the sawfly larvae Lophyrotoma interrupta and Arge pullata are promising candidates to prove that some higher organisms contain peptide compounds.94,95 However, lack of genes encoding peptide synthetase within the determined DNA sequence of Haemophilus influenzae has proved that not every organism contains such an enzyme system.95 Most of these enzymatically synthesized peptides102,103 are composed of peptide chains that may be modified by acylation or glycosylation.91

Many Gram-negative bacteria, for example Pseudomonas, produce secondary metabolites in this manner,104,105 Some of the well-known models are syringotoxin from P. syringae,106 phaseolotoxin from P. syringae, var. phaseolica107, the iron siderophores enterobactin108,109 and pyoverdin110–112 from P. aeruginosa. The fresh-water cyanobacterium M. aeruginosa produces mycrocystin,113–115 and Zolarian arboricola produces pneumocandin.116 Table 1 describes other significant peptide products and their producing strains in detail (for further details about peptide products see refs. 65, 88). Nonribosomally produced peptides thus contribute to the remarkable structural diversity of low molecular weight peptide products.117

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B9780080912837000941

4.26.7.1 GALA

Because the peptide GALA is one of the most extensively characterized synthetically engineered fusogenic peptides,251 a more in-depth discussion will be dedicated to this model peptide. The thorough mechanistic understanding of the pH-responsive membrane activity of GALA can be used as a model to lend insight into the biophysical properties of other less well-characterized fusogenic peptides. GALA is a 30-amino acid (AA) sequence (WEAALAEALAEALAEHLAEALAEALEALAA) possessing repeats of glutamic acid-alanine-leucine-alanine (Glu-Ala-Leu-Ala or EALA). The Glu-Ala-Leu-Ala repeating structure was innovatively engineered by Szoka et al. to possess a combination of charged/acidic (Glu), spacer (Ala), and hydrophobic (Leu) AAs that impart pH-dependent structure–function properties that mimic fusogenic viral proteins to form pores in membranes under acidic conditions.252–254 Near neutral pH, Glu residues exist predominantly in an ionized state, making GALA hydrophilic and creating intramolecular charge repulsion that causes the peptide to adopt an extended state. When exposed to the slightly acidic endosomal environment, Glu residues (pKa~6) are protonated, which transforms GALA into a more hydrophobic state with decreased anionic charge density and triggers the peptide to adopt an amphipathic α-helical structure254 (Fig. 9). Within this helix, the Glu residues form a hydrophilic α-helical face, and the Leu residues localize to the opposing, nonpolar face.252 This amphipathic helix is able to interact with lipid bilayers in an orientation parallel to the membrane, and it subsequently self-aggregates and reorients into a pore perpendicular to the membrane surface once a critical number of approximately 10 peptide chains have oligomerized.255–257 These pores, which generate pH-dependent lipid vesicle leakage, are lined on the interior with hydrophilic Glu residues while the more hydrophobic amino acids associate with the alkyl chains that constitute the interior of the lipid bilayer.252,253,255 One potential limitation of GALA is that cholesterol content and some phospholid compositions can reduce pore formation.258,259 However, more recently, new derivatives of GALA have been created that are activated earlier in the endosomal pathway at a higher pH (i.e., pH 6–7 versus pH < 6 for GALA), and these variants have also improved performance in cholesterol-containing membranes.260

Fig. 9. Schematic illustration of GALA pH-dependent activation. (a) The Glu residues in GALA extend the peptide backbone at neutral pH but are protonated at acidic pH, triggering the peptide to adopt an alpha helix structure (b). The GALA alpha helix is amphipathic in nature and (c) initially interacts and aggregates parallel to the lipid bilayer. (d) When a critical number of peptides aggregate (estimated to be 10), they self-assemble into a transmembrane channel with polar Glu residues lining an aqueous pore that allows biomacromolecule diffusion across the bilayer.

GALA has been predominantly applied as a constituent of nonviral vectors for nucleic acid delivery, and Haensler and Szoka were the first to apply GALA for pDNA delivery.147 GALA was conjugated to cationic PAMAM dendrimers (10–13 GALA per dendrimer used in a 1:1 ratio with unmodified dendrimer) to enhance endosomal escape of plasmids electrostatically complexed with PAMAM. Using a luciferase plasmid as a reporter system, the authors found that inclusion of GALA on the dendrimers improved transfection for all seven cell types tested. For some cells (i.e., fibroblast (CV-1), leukemia (K562), and hepatocyte (HepG2) cell lines), luciferase activity increased by four orders of magnitude in GALA-containing dendrimers relative to unmodified PAMAM. Subsequently, Simoes and coauthors used cationic liposomes to electrostatically complex GALA (anionic due to Glu residues) into ternary (lipid/GALA/DNA) and quaternary (lipid/GALA/DNA/transferrin) systems.261–263 They initially developed a ternary system formed by cationic lipids complexed with GALA followed by plasmid condensation to form a net negatively charged carrier, and it was found that this approach produced a fourfold increase in luciferase production relative to controls.261 A follow-up, mechanistic study found that GALA enhancement of transfection was inhibited by bafilomycin, a drug that inhibits endosomal acidification, verifying the key contribution of pH-dependent GALA activity.262 Subsequently, a quaternary system that also included the targeting ligand transferrin was developed and shown to successfully transfect human macrophages, a particularly difficult target for nonviral gene therapy due to their potent lysosomal degradative activity.263 The quaternary complex performed better than two control ternary systems (lipid/DNA/transferrin and lipid/GALA/DNA) and better than an alternative quaternary complex containing the influenza-derived fusogenic peptide, HA-2, as an alternate to GALA. However, a primary shortcoming of using multiple, electrostatically complexed components was exemplified by the significant falloff in activity observed in the presence of serum.

The Harishima laboratory has produced the most recent advances in the use of GALA in liposomal drug delivery.264–269 An early application involved mixing GALA with commercially available cationic liposomes. This resulted in as much as a fivefold increase in transfection efficiency in some conditions tested, further illustrating the efficacy of GALA for enhancing cytoplasmic delivery.265 In other studies, GALA was derivatized with cholesterol (chol-GALA) in order to anchor GALA to the surface of the liposome. Chol-GALA efficiently mediated diffusion of the model molecule rhodamine from the liposomal carrier into the cytosol of treated cells (while GALA encapsulated within the liposomal aqueous interior did not).264 In subsequent iterations of this carrier, transferrin was covalently linked to the liposome surface to target the liposomes for cell uptake. This targeted system effectively delivered nucleic acids (plasmid and siRNA) and the antitumor peptide, mastoparan.267–269 Enhanced in vivo circulation time and production of smaller, more homogeneous particles were also achievable through carrier surface PEGylation. Not surprisingly, it was found that attachment of PEG chains to the liposomal surface inhibited activity of GALA anchored in the membrane.267 However, activity could be restored when GALA functionalization was present both on the lipid membrane and on the free PEG chain ends267 or by using reversible PEG attachments removed in matrix metalloproteinase-rich environments.268

GALA peptides have also been utilized for delivery of therapeutic proteins. One of the first applications was by Anderson and coworkers who covalently conjugated GALA to the Fab fragments of an antitumor antibody with the goal of enhancing cell–surface interactions, tumor retention time, and therapeutic functionality.270 Although their application did not exploit the pH-responsive properties of GALA for cytoplasmic delivery, it did prolong cell-surface antibody retention of the protein in human melanoma cells (although this effect was not as apparent in a similar colon carcinoma application). More recently, Kuehne and Murphy conjugated GALA to OKT9, an antibody against the transferrin receptor, with the ultimate goal of targeted, intracellular delivery of cancer-killing immunotoxins.271 Optimal performance was achieved with 2–3 GALA peptides per therapeutic antibody, although a sevenfold greater concentration of GALA in the OKT9 conjugate form was required to cause 50% leakage compared to free GALA. Inclusion of a flexible PEG spacer between GALA and OKT9 did not rescue pH-dependent activity, although the potential negative effect of the hydrophilic linker on membrane interactions was not addressed. In a recent study, Kobayashi et al. tested the ability of biotinylated GALA to achieve cytoplasmic delivery of avidin (serving here as a model protein) and quantum dots (QDs) surface functionalized with streptavidin.272 Cellular uptake for these systems was enhanced by electrostatic formulation with a commercially available cationic liposome transfection agent. Four GALA peptides per avidin improved cytosolic delivery of the model protein, and GALA functionalization of the QDs decreased their colocalization with lysosomes. However, QD staining remained punctate, and it was not apparent whether decreased presence in lysosomes was due to altered intracellular trafficking or the inability of QDs to escape through small GALA-formed pores in the endolysosomal membranes. Thus, additional work will be needed to further elucidate and develop GALA-induced cytosolic delivery of QD imaging probes.

The most recent developments incorporating GALA functionality have involved its incorporation into genetically engineered proteins. Wang and coauthors recently engineered a biomimetic protein consisting of GALA, a DNA-condensing set of histone H2A repeats, a cathepsin cleavable motif (that separates the targeting ligand and its potential steric hindrances once the full construct reaches the cathepsin-containing endosomes), and a HER2 receptor cancer cell surface targeting motif.273 This novel application showed the importance of GALA activity in achieving gene transfer using this approach and serves as a starting point for development of future biosynthetic vectors with perfectly defined, monodispersed compositions and architectures that may prove to be advantageous to the chemosynthetic approaches utilized previously. Another recent, novel approach employed GALA as a component in an engineered protein designed to produce pH-sensitive ‘nanocages’ that, after further development, may serve as a new type of acid-induced molecular switch.274

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B9780081006917000197

2.1 Introduction

Proteins are polypeptides synthesized intracellularly within ribosomes and are commonly interpreted as the products of cellular expression including tissues that compose enzymes, growth factors, and connective tissues. Since each ribosome is influenced by conformational hindrance and electronic charge, those features regulate the selectivity to a specific amino acid or acids that can be added to a growing peptide chain. As a result, even in a realm where there are many different potential amino acids to choose from, high degrees of replication are commonly coupled with very narrow molecular weight distribution of the polyamide proteins that are produced. The templates used to synthesize these peptides are commonly replicated with such regularity that even with minor differences in sequences between chains of different protein structures yield very modest differences in protein length and molecular weight on the order or hundreds of Daltons with average molecular weights ranging from 3 to 300 kDa. The syntheses used to produce proteins are interesting from a conceptual comparison to chain growth that occurs through a condensation reaction at very modest conditions, and while the growth rates are likely slow, the sheer volume of cells that can have replicate function is so large that slower growth rates can be tolerated for growing replicated tissues. One can also compare protein synthesis to radical polymerization which when controlled usually has a small number of reaction sites and a much wider range of molecular weights distributed within the reaction mixture at the end of the synthesis. In this chapter, a deeper understanding of these distinctions coupled with how chain lengths within polymers and proteins are determined.

Read full chapter

URL: https://www.sciencedirect.com/science/article/pii/B978012809478500002X