In the previous post, we looked at what kinds of chemical preparation and linking processes are needed to bring in free amino acids that had been quietly maintaining a stable state, use them as the building materials for proteins, and connect them one by one into a peptide chain. Then, does the peptide bond formed through all that effort simply end as a link connecting amino acids in a chain? Now let us shift our attention from the process of making the bond to the consequences brought about by the peptide bond once it has been formed. Above all, the resonance structure formed by the peptide bond has a very important influence on protein structure. Let us continue by looking at how this restricts the structures that proteins can adopt, and how a variety of structures can still become possible within those restrictions.
A peptide bond is not simply a chemical bond that connects amino acids. Because of resonance, it has planarity, and this property has a decisive influence on hydrogen bonding and the formation of protein secondary structure. Proteins, however, do not simply exist as a series of flat boards. In this post, I would like to look step by step at how peptide bonds that maintain planarity can nevertheless create a variety of three-dimensional structures through the φ (phi) and ψ (psi) dihedral angles.
Free amino acids exist in a very stable state because of their stable resonance structures, so participating in reactions is neither particularly attractive nor favorable for them. This means that the dehydration condensation reaction between two amino acids that we commonly know does not readily occur spontaneously in the body. In other words, to connect amino acids and form a peptide chain, enzymes must step in to activate the free amino acids and guide them toward bond formation. Against this background, in the previous post we examined in great detail the entire process: how cytosolic aaRS wakes up free amino acids that had been peacefully remaining in a stable state and activates them so that they can participate in reactions, how it loads them onto tRNA and sends them to the ribosome, and how, in the ribosome, peptidyl transferase transfers the peptide chain carried by the peptidyl-tRNA in the P site to the newly arrived tRNA in the A site, thereby creating the conditions for a peptide bond to form. In this way, in the previous post we followed the actual process by which a peptide bond is formed. Now let us look more closely at the peptide bond itself that results from this process. What special properties does this bond, chemically an amide bond, have?
A peptide bond is not formed by a simple dehydration condensation reaction.
If you look up peptide bond formation, it is usually described as a dehydration condensation reaction in which the OH of the carboxyl group (COOH) reacts with an H from the amino group, producing one molecule of H₂O as the bond forms. In living systems, however, two free amino acids do not readily meet and spontaneously join together by eliminating H₂O in this way. Instead, an aaRS first uses ATP to activate the amino acid, and then links the amino acid to the ribose of tRNA through an ester bond, producing aminoacyl-tRNA. An acyl-transfer reaction then takes place in the ribosome, ultimately forming the peptide bond, which is an amide bond. In other words, two free amino acids do not directly join together while a molecule of H₂O is released during the reaction in the way we commonly imagine. This was one of those moments when I felt what Koreans call a kind of “reality check.” The same is true not only for peptide bonds but also for the phosphodiester bonds that make up the backbone of nucleic acids. It makes me wonder whether what we know is wrong. But perhaps the reason there is such a difference between how a chemical reaction is written and how the reaction actually takes place is that they focus on different things: “what change actually occurred and how?” versus “what ultimately changed into what?” From a chemical point of view, what matters is how the atoms present at the beginning are rearranged by the end—in other words, what is ultimately produced. A simplified reaction equation represents the net transformation of a reaction; it does not show the complex and detailed mechanism by which the reaction actually occurs. Reaction equations are therefore useful for understanding the overall chemical change, but how that change is actually carried out in a living system has to be explored at a separate level.
Looking at the background and formation of peptide bonds naturally leads to another question. Why does the human body go through such a complicated process, even using enzymes and spending ATP, just to connect amino acids? Perhaps the answer has something to do with stability. Proteins responsible for the structure and function of the body have to maintain their form and function for a long time in the body, and if the bonds between amino acids broke easily, enzymes, cytoskeletal proteins, transport proteins, and membrane proteins would all be unable to perform their functions properly. Fortunately, the fact that there are separate enzymes specialized in breaking down proteins also seems to suggest that the peptide bonds forming proteins are quite stable and robust even under physiological conditions. Then why can peptide bonds be so stable? To answer this question, we first need to look at the stability of free amino acids.
Free amino acids exist as very stable molecules under physiological pH conditions. In the near-neutral environment of the body, the carboxyl group (COOH) of a free amino acid loses a proton and exists as a carboxylate (COO⁻), and resonance exists between the carbonyl (C=O) carbon and the two oxygen atoms (O, O⁻) of the carboxylate. Because of this resonance structure, free amino acids can exist as very stable molecules. The amide formed by a peptide bond also has resonance involving the carbonyl carbon and the NH group, allowing it to remain in a very stable state. Molecular stability can be understood through resonance structures. So let us review once again what resonance is.
Resonance
Resonance refers to a state in which electrons are not fixed on one particular atom but are distributed over several neighboring atoms. When electrons are crowded into a narrow space, the state is less stable and higher in energy, whereas when they are spread over a wider space so that charge density is lower, the state becomes more stable and lower in energy. It is a little like the difference between being stressed in a room packed with people and feeling more comfortable when everyone can spread out across several rooms. This spreading of electrons over a wider region is called electron delocalization. The electrons involved in resonance are mainly π electrons or lone-pair electrons, and for them to spread across several atoms, p orbitals that contain these electrons and can overlap with one another are required. If I were to draw it, perhaps it would look something like a cloud-like pathway. To understand this, let us bring back the resonance structure of the carboxylate group that we looked at in the previous post.
The Stable Resonance Structure of Carboxylate
The carbon atom of a carboxylate group is sp²-hybridized and lies in the same plane as the two oxygen atoms. In an sp² structure, three hybrid orbitals form single bonds (σ bonds), while one p orbital that does not participate in hybridization remains perpendicular to the plane. The carbon forms sigma (σ) bonds with the oxygen atoms along the internuclear axes, keeping the atoms at a certain distance from one another, and in addition to the orbitals forming these single (= sigma) bonds, there is a perpendicular p orbital capable of forming a pi bond (the additional bond in a double bond). The two oxygen atoms are also sp²-hybridized and likewise have perpendicular p orbitals. The three p orbitals of the carbon and oxygen atoms can align parallel to one another and overlap, forming one broad electron cloud, or π-electron system, in which the electrons are shared across the atoms. This is resonance. How large an area the p orbitals overlap across is related to the degree of stabilization provided by resonance. Within this π-electron system, the negative charge of the carboxylate group is also not confined to one particular oxygen atom but is distributed across both oxygen atoms.
Resonance is usually explained by drawing two resonance structures with a double-headed arrow between them, but the arrow does not mean that the molecule moves back and forth between the two structures. In reality, the molecule cannot be represented clearly by either structure alone; it is more accurate to think of it as existing in an intermediate state in which the two are mixed together. This combined state is called a resonance hybrid, and the electrons are spread throughout this hybrid structure, which stabilizes them. Because the negative charge is not located on either one oxygen atom but is distributed over both, the two C–O bonds have almost identical bond lengths and have properties intermediate between a single bond and a double bond. The key point is that when electrons are spread over a wider region through resonance, the energy of the system decreases and its stability increases, and this is also why carboxylate is stable and relatively unreactive under physiological conditions.
To synthesize proteins, however, this stable carboxylate group has to be made to participate in a reaction, and the problem is that carboxylate is simply too stable. Living organisms therefore use aaRS enzymes and ATP to first raise the amino acid to a higher free-energy state. The free amino acid is converted into aminoacyl-AMP, and in the process the stable resonance state originally possessed by the carboxylate disappears, producing instead an activated intermediate that can more readily transfer its aminoacyl group to another molecule. Aminoacyl-AMP is called a high-energy intermediate, which means not only that energy is stored in it but also that it is unstable and has a strong thermodynamic tendency to return to its original stable state. In general, the more stable a state is, the lower its energy, while the more unstable it is, the higher its energy level. It therefore has a strong tendency to participate in another reaction so that it can quickly return to a lower-energy, more stable state. This is somewhat similar to a radical, which carries an unpaired electron and is extremely unstable, with a very high energy level, and reacts aggressively with surrounding molecules in an attempt to obtain an electron and reach a more stable state, such as a filled outer shell. I previously discussed in detail, in the posts on reactive oxygen species (ROS), why high reactivity can be dangerous. However, whereas ROS can behave like destructive runaway troublemakers that indiscriminately attack cell membranes, DNA, and other molecules and therefore need to be controlled, living organisms can also use enzymes to harness this kind of reactivity to their own advantage. Of course, the required reaction must occur only within the protected environment inside the enzyme, so it has to be controlled by a precise and tightly regulated reaction mechanism, and aaRS is a representative example of such an enzyme.
Interestingly, the human body deliberately makes stable free amino acids unstable and then uses them to form very stable peptide bonds so that proteins can remain stable. (Give the disease and then give the medicine…) The stability of the peptide bond also comes from resonance. Chemically, a peptide bond is an amide bond. The term “peptide bond” is almost like a special chemical nickname given to an amide bond when it forms part of a peptide chain. Let us focus on the carbonyl (C=O) group and the nitrogen of the amide NH in the peptide bond. In carboxylate, the main players in resonance were the carbonyl carbon and the two oxygen atoms, whereas in a peptide bond the main players are the carbonyl carbon, the oxygen, and the nitrogen. The carbon and oxygen of the carbonyl group are sp²-hybridized and each has a p orbital. The nitrogen also adopts a geometry close to sp² in order to maximize resonance, and its lone pair occupies a p orbital. As a result, the three p orbitals overlap in parallel, forming one continuous π-electron system. The nitrogen lone pair participates in this π-electron system and becomes distributed over all three atoms, stabilizing the peptide bond through resonance. The single bond between carbon and nitrogen is shorter than an ordinary single bond but longer than a double bond, giving it an intermediate character that is commonly described as partial double-bond character, and because of this feature, rotation around the carbon–nitrogen bond is strongly restricted. The broad π-electron system formed by overlapping p orbitals across these three atoms is very similar to what occurs in carboxylate. One small difference, however, is that in carboxylate the two oxygen atoms share the negative charge, so the two carbon–oxygen bond lengths are the same and the electrons are distributed symmetrically, whereas in an amide the carbon–nitrogen and carbon–oxygen bond lengths are not the same, so the electron distribution is asymmetric. In the amide resonance system, electron density is also shifted somewhat more toward oxygen because oxygen is much more electronegative than nitrogen.
To explain resonance in an amide bond, two resonance structures are drawn as shown above. The structure on the left is the usual amide structure, in which a double bond (C=O) exists between carbon and oxygen and the nitrogen is shown in a neutral state with a lone pair. The structure on the right is a resonance contributor in which the nitrogen lone pair participates in the π-electron system of the carbonyl group, forming a π bond between carbon and nitrogen. In this case, the nitrogen formally carries a positive charge (N⁺), while the oxygen formally carries a negative charge (O⁻). However, these two structures do not represent different states that the actual molecule adopts, and the double-headed arrow drawn between the two resonance structures does not mean that the molecule rapidly moves back and forth between them. It is simply a chemical notation used to describe the actual electron distribution. The real peptide bond exists as a single resonance hybrid with an intermediate electron distribution in which the two structures are mixed in appropriate proportions. It is a little like drawing the same mountain from the east and from the west: neither the eastern view nor the western view is the actual complete form of the mountain, and neither view alone is sufficient. Only by combining the two can we form a proper picture of the mountain. The important point is that within the π-electron system formed by overlapping p orbitals across O–C–N, the electrons are not confined to either one of the drawn structures but are broadly distributed across all three atoms. This is the essence of resonance, and the peptide bond is stable because the electrons exist in a highly stabilized state within this resonance hybrid.
The reason for first reviewing and trying to understand the resonance structure formed by the peptide bond in such detail is, above all, to understand why proteins do not easily undergo hydrolysis and break down even in the aqueous environment of the body, but instead remain intact and perform their functions stably. Another important reason is to understand how this amide resonance structure influences the formation of the three-dimensional structure of proteins.
Why Does the Peptide-Bond Region Have Planarity?
The O–C–N atoms in the peptide-bond region form a π-electron system and share delocalized electrons, creating resonance. The electrons involved in resonance are lone-pair electrons or π electrons, and because these electrons occupy p orbitals, the p orbitals of the three atoms must overlap side by side while remaining parallel to one another in order to form a π-electron system. For the p orbitals to overlap effectively, the nuclei of the three atoms—the bonding framework—must lie in the same plane, because only then can the remaining p orbitals on each atom be arranged parallel to one another in a direction perpendicular to that plane. This arrangement must be maintained for the electrons to remain efficiently delocalized, thereby maximizing the resonance effect. For this reason, an amide bond adopts a planar structure in which the atoms of the peptide-bond region lie in the same plane in order to maintain resonance.
In the post on hybridization, we looked at single and double bonds. A single (sigma, σ) bond is formed when orbitals overlap head-on in a straight line along the axis connecting the two atomic nuclei and share electrons, whereas a pi (π) bond is formed when p orbitals located on either side of the bond axis overlap side by side rather than along the bond axis itself. A π bond therefore always exists together with a single (σ) bond and is an additional bond that forms part of a double or triple bond. Single bonds can rotate relatively freely, whereas rotation around double and triple bonds is very difficult because forcing them to rotate would break the π bond. It is a little like how it is easy to turn while holding one hand but difficult to rotate while holding both hands. As we saw earlier, the C–N bond has partial double-bond character, so free rotation around the C–N bond is strongly restricted. Rotation would require disrupting resonance, which requires energy, so a rotational barrier exists between the carbonyl group and the nitrogen that participate in the resonance structure, making rotation very difficult. Therefore, when a peptide bond forms, the atoms around the bond are held rigidly in a single plane, forming a planar structure known as the peptide plane.
Peptide Plane
In structural biology, this planar arrangement around the peptide-bond region is called the peptide plane or peptide unit. Unlike the peptide-bond region itself, the peptide plane is a slightly larger structural unit that includes the α-carbons on both sides, represented as Cα–C–N–Cα (the yellow rectangle in the image below). Because a peptide unit behaves as a single planar unit, like a thin board, the peptide chain does not bend freely in every direction but instead resembles a series of boards connected here and there by hinges. Then where does rotation occur in a protein? The flexibility and rotation of the protein backbone occur not within the peptide unit but immediately before and after the peptide bond, and these rotation angles are called φ (phi) and ψ (psi). φ is the rotation around the single bond between the α-carbon and the nitrogen before it (N–Cα), while ψ is the rotation around the single bond between the α-carbon and the carbon after it (Cα–C). The dihedral angle around the C–N bond within the peptide unit is represented by ω (omega), and because most peptide bonds adopt the trans form in which the two Cα atoms lie on opposite sides, ω is close to 180°.
Peptide Resonance, Hydrogen Bonding, and Protein Secondary Structure
What effect, then, do the planarity of the peptide unit and its restricted free rotation have on protein structure? The planarity of the peptide unit allows the spatial orientations of the carbonyl oxygen and the amide NH to remain relatively constant. This provides the structural and fundamental basis for the formation of regular protein secondary structures such as α-helices and β-sheets. The hydrogen bonds that form α-helices and β-sheets are formed between the carbonyl oxygen, which acts as the acceptor, and the amide NH, which acts as the donor, and hydrogen bonds are known to be highly sensitive not only to distance but also to direction. A hydrogen bond is not simply an attractive force between two atoms that happen to be close together; it is a highly directional interaction that becomes strongest when the hydrogen atom and the lone pair of the acceptor atom are arranged in a favorable direction. In general, the strongest and most stable hydrogen bond forms when the donor and acceptor arrangement approaches a straight line (about 180°), because this maximizes the electrostatic interaction between the lone pair and the partially positively charged hydrogen of the donor. When the angle deviates from this arrangement, the overlap of the electron clouds decreases and the hydrogen bond becomes weaker. As shown in the figure below, the strongest hydrogen bond forms when the hydrogen-bond axis between the acceptor and donor aligns in a straight line with the bond axis of the donor–hydrogen single bond, so directionality is a very sensitive factor that determines the stability of a hydrogen bond.
Considering this, we can begin to see what effect the planarity of the peptide unit has. Because planarity keeps the spatial orientations of the carbonyl oxygen, the hydrogen-bond acceptor, and the amide NH, the donor, almost fixed and constant, repeatedly adopting particular φ and ψ rotation angles allows these groups to become regularly arranged and makes the formation of a stable hydrogen-bond network possible. Within the same polypeptide chain, if the carbonyl oxygen of residue i repeatedly forms a hydrogen bond with the amide NH of the residue four positions away (i+4), an α-helix is formed. Because the peptide plane is fixed, this hydrogen-bond pattern can be maintained, and if the rotation angles before and after it repeat appropriately, a helical arrangement is automatically produced. The same applies to β-sheets. In a β-sheet, several β-strand segments are arranged side by side, and because each peptide plane maintains a consistent orientation, the carbonyl oxygens and NH groups of facing β-strands can regularly face one another, allowing repeated hydrogen bonds to form. If the peptide unit were not a rigid planar structure fixed by resonance but instead could rotate freely and twist, the orientations of the carbonyl oxygen and NH would constantly change, making it difficult to maintain this directionality. In that case, the regular and consistent hydrogen-bonding patterns found in α-helices and β-sheets would be difficult to maintain, and as a result it would also be much more difficult to build the elaborate three-dimensional structures based on these secondary structures.
In fact, resonance not only makes the peptide bond planar and thereby helps hydrogen bonds form more stably, but also strengthens the hydrogen bonds themselves. In general, when hydrogen forms a covalent bond with an electronegative atom such as oxygen (O) or nitrogen (N), the shared electron pair is pulled more strongly toward that atom, leaving the hydrogen with lower electron density and a partial positive charge (δ⁺). Conversely, the oxygen or nitrogen pulls more electron density toward itself and acquires partial negative character. Because of this imbalance in charge, an oxygen or nitrogen atom with an electron-rich lone pair electrostatically attracts the partially positively charged hydrogen, producing a hydrogen bond. In peptide bonds, resonance makes this charge imbalance even more pronounced. Resonance shifts more electron density toward the carbonyl (C=O) oxygen, giving the oxygen a stronger partial negative charge, while the amide nitrogen loses electron density and becomes partially positive. The hydrogen attached to that nitrogen also becomes more strongly partially positive. In other words, resonance makes the carbonyl oxygen a stronger hydrogen-bond acceptor and the amide NH a more attractive hydrogen-bond donor, creating conditions that are even more favorable for hydrogen bonding. In addition, as we saw earlier, resonance maintains the planar structure and thereby provides a consistent and stable directionality favorable for hydrogen-bond formation. In this way, resonance generated by the peptide bond provides planarity geometrically and stronger hydrogen-bonding ability electronically, making it one of the major contributors to the formation of protein secondary structures such as α-helices and β-sheets.
Three-Dimensional Structure and the Two Dihedral Angles φ and ψ
We have looked in detail at how the resonance effect of the peptide bond has several important consequences: it keeps the peptide unit as a rigid plane, strengthens the polarity of the carbonyl carbon and amide NH, and helps form more stable hydrogen bonds with consistent directionality. In particular, we examined how the planarity of the peptide unit contributes to the stable formation of regular secondary structures such as α-helices and β-sheets. Now let us go one step further and look at how proteins form three-dimensional structures. As we know, proteins are not simply made of flat boards lined up in a row, and planar structures alone cannot produce the distinctive three-dimensional structures of proteins that we describe as folding.
The peptide plane itself can hardly rotate, but rotation is possible immediately before and after it. The single bonds between peptide planes act somewhat like flexible hinges connecting two boards, and their rotation angles are represented by φ and ψ. For this reason, in protein structural biology, proteins are often described as chains of planar peptide units connected through φ and ψ angles. It is hardly an exaggeration to say that virtually every three-dimensional protein structure is the result of how these planar pieces are connected to one another.
Protein Backbone Dihedral Angles
It is easy to confuse a dihedral angle with a simple angle between two planes, but a dihedral angle is not merely the angle formed by two planes. When four atoms are connected in sequence, the dihedral angle describes how much the two planes on either side are twisted relative to one another around the bond axis connecting the two middle atoms. In the figure, if the B–C bond is used as the central axis and the two planes containing A and D are rotated in opposite directions, the A–B–C plane and the B–C–D plane become twisted and offset relative to each other, and the degree of that twist is represented by the dihedral angle.
In other words, φ and ψ are often called rotation angles, but strictly speaking they should not be regarded simply as values showing how much a single bond has rotated. To define an amount of rotation, one would first have to determine the reference position from which the rotation is measured, but in protein structure what matters is not the amount of rotation itself. What matters is how the structures on the two sides are arranged relative to each other in space as a result of that rotation. For this reason, instead of directly measuring the rotation itself, structural biology measures the dihedral angle formed by the two planes on either side of the rotation axis and calls these angles φ and ψ. Thus, φ and ψ are structural indicators that describe how the protein backbone bends and twists in three-dimensional space; more precisely, they are dihedral angles defined around particular rotation axes.
The φ (Phi) Dihedral Angle
Let us first look at how the φ dihedral angle is defined. Once φ is understood, ψ can be understood in the same way. Because a dihedral angle is an angle between two planes, two planes are required to define it. One of them already exists as the peptide plane formed by the peptide bond (the yellow region in the image above), so the other has to be defined geometrically. In theory, any three points can define a plane, so the three atoms N–Cα–C centered around the α-carbon can be used to define a reference plane (the green region in the figure). The Cα atom in the amino acid backbone is sp³-hybridized and forms four single σ bonds with four substituents: nitrogen (N), the carbonyl carbon (C), hydrogen (H), and the side chain (R), producing an approximately tetrahedral geometry with bond angles of about 109.5°. These bond angles remain almost constant, but because single σ bonds can generally rotate around their bond axes, the N–Cα and Cα–C single bonds can also rotate around their respective bond axes. Therefore, when rotation occurs around the N–Cα bond axis, the relative spatial arrangement of the atoms connected to Cα also changes, and as a result the relative orientation between the N–Cα–C reference plane and the peptide plane changes as well.
An important point here is that the N–Cα bond axis is part of the peptide plane. Therefore, rotation around the N–Cα bond axis does not simply mean that a single atom rotates; rather, the entire peptide plane on the front side rotates as one rigid planar unit. Ultimately, φ is the angle that represents how much the front peptide plane and the N–Cα–C reference plane are twisted relative to each other as a result of this rotation. If we imagine looking directly along the N–Cα axis, the two planes rotate in opposite directions around the central axis, much like wringing laundry with both hands, as indicated by the red and blue arrows, causing their relative orientations to change. In this way, φ represents how much the two planes are twisted relative to each other around the bond axis, which is why φ is also called a torsion angle. Likewise, because the Cα–C bond is also a single bond, the peptide plane on the other side can rotate around the Cα–C axis. The dihedral angle formed between the N–Cα–C reference plane and the peptide plane on that side is ψ.
So far, we have seen that peptide bonds have resonance, and that because of this the peptide unit has a planar structure in which rotation is not possible, while this planarity contributes to the formation of stable secondary structures. At the same time, the N–Cα and Cα–C bonds on either side of the peptide bond can generate a variety of three-dimensional structures through the φ (phi) and ψ (psi) dihedral angles. Then, how many combinations of φ and ψ are actually possible in real proteins? In the next post, beginning with the Ramachandran plot, which showed that the combinations of φ and ψ dihedral angles in the protein backbone are in fact highly restricted, we will look at how the major historical studies and theories developed as researchers tried to understand protein structure and folding.
[Image sources]
Image 22-1. Chemistry of Bridged Lactams: Recent Developments — CC BY 4.0
Image 3
3-1. Protein backbone PhiPsiOmega drawing.svg — CC BY 3.0
Image 4
4-1. OSC Microbio 07 04 secondary.jpg — CC BY 4.0
Image 5
5-1. Dihedral angle.svg — CC BY-SA 4.0
