Polar Uncharged and Charged Side Chains

Serine protease drug design. Aspirin acetylates a serine hydroxyl in cyclooxygenase (COX-1/COX-2), covalently blocking the enzyme's active site — a direct pharmacological exploitation of serine nucleophilicity. Organophosphate nerve agents and pesticides (sarin, parathion metabolites) covalently phosphorylate the catalytic serine of acetylcholinesterase, and antidotes like pralidoxime work by reactivating that serine through nucleophilic displacement chemistry built on the same hydroxyl reactivity discussed here.

Cysteine-targeted covalent drugs. A growing class of "targeted covalent inhibitors" — including the EGFR inhibitor osimertinib and the KRAS(G12C) inhibitor sotorasib — exploit a specific, accessible cysteine thiolate near an ATP-binding or GTP-binding pocket, forming an irreversible covalent bond that dramatically increases target residence time and potency, directly leveraging cysteine's superior soft nucleophilicity over serine.

Redox biology and disulfide engineering. Cysteine's reversible thiol–disulfide chemistry underlies protein folding quality control in the endoplasmic reticulum (protein disulfide isomerase), the redox-sensing mechanism of transcription factors like NF-κB and Keap1/Nrf2, and industrial biotechnology, where engineered disulfide bonds are introduced to stabilize therapeutic antibodies and enzymes for extended shelf life.

Histidine tags and metal chelation in biotechnology. The imidazole nitrogen's lone pair, the same feature that makes histidine catalytically versatile, is exploited in recombinant protein purification: hexahistidine (His-tag) fusion proteins bind nickel or cobalt affinity resins through direct imidazole–metal coordination, a technology used in essentially every modern structural biology and protein production laboratory.

Arginine-rich cell-penetrating peptides and antiviral design. The guanidinium group's ability to form bidentate hydrogen bonds with phosphate groups is exploited in arginine-rich cell-penetrating peptides (used for intracellular drug delivery) and is central to how the influenza neuraminidase inhibitor zanamivir was designed, mimicking the transition-state charge distribution normally stabilized by an active-site arginine cluster.

Asparagine deamidation as a molecular clock. Asparagine side chains spontaneously deamidate to aspartate (via a succinimide intermediate) over time in long-lived proteins — a slow, non-enzymatic chemical clock exploited in forensic age estimation from eye-lens crystallin proteins and implicated in the aging-related instability of therapeutic antibodies during storage, a direct consequence of the amide's latent electrophilicity at the carbonyl carbon.

Ask the tutor: How does burial of a charged side chain inside a low-dielectric protein core quantitatively shift its pKa, using the Born solvation model?

Every amino acid shares the same backbone — an α-carbon bearing an amino group, a carboxyl group, and a hydrogen — and differs only in the side chain (R group) hanging off that α-carbon. The eleven side chains treated here divide naturally into four chemical families, and within each family the reactivity follows directly from bond polarity and orbital availability.

The hydroxyl family — serine and threonine. Serine's side chain is –CH₂OH, a primary alcohol; threonine's is –CH(OH)CH₃, a secondary alcohol that also introduces a second stereocenter (threonine is one of only two proteinogenic amino acids, with isoleucine, having two chiral centers). The C–O bond is polarized (O is more electronegative, δ⁻ on O, δ⁺ on C), and the O–H bond is similarly polarized, making the hydroxyl oxygen both a hydrogen-bond donor (via O–H) and a hydrogen-bond acceptor (via its lone pairs). The pKa of this alcohol proton, roughly 13–15 in isolation, is far too high for spontaneous deprotonation at physiological pH — but a nearby general base (a histidine imidazole, an aspartate carboxylate, or even a backbone amide) can lower the effective activation barrier for serine to act as a nucleophile, which is precisely how the serine protease catalytic triad works.

The thiol family — cysteine. Cysteine's side chain is –CH₂SH. Sulfur is less electronegative than oxygen but far larger and more polarizable, and the S–H bond is weaker (bond dissociation energy ~87 kcal/mol vs. ~110 kcal/mol for O–H) and more acidic (pKa ~8.3 for free cysteine vs. ~13–15 for serine's alcohol). This combination — a softer, more polarizable, more acidic sulfur — makes the thiolate (–S⁻) a substantially better nucleophile than an alkoxide toward soft electrophilic centers, and it also makes cysteine uniquely capable of reversible two-electron redox chemistry: two thiols oxidize to a disulfide bond (–S–S–), a covalent cross-link central to protein folding and to intracellular redox signaling.

The amide family — asparagine and glutamine. These are the side-chain carboxamides of aspartate and glutamate: –CH₂CONH₂ for asparagine, –CH₂CH₂CONH₂ for glutamine. The carbonyl carbon is sp² and conjugated with the nitrogen lone pair, exactly as in the peptide bond itself — partial C=N double-bond character restricts rotation and makes the amide nitrogen a poor nucleophile (its lone pair is delocalized into the carbonyl) while the carbonyl oxygen remains an excellent hydrogen-bond acceptor and the N–H protons are hydrogen-bond donors. Because this group cannot ionize near physiological pH, Asn/Gln contribute polarity and directional hydrogen bonding to protein structure without ever carrying formal charge.

The acidic family — aspartate and glutamate. These carry a carboxylic acid side chain (–CH₂COOH for Asp, –CH₂CH₂COOH for Glu) with intrinsic pKa values around 3.9 (Asp) and 4.1 (Glu) — dramatically lower than a simple aliphatic carboxylic acid (pKa ~4.8 for acetic acid) because of inductive effects from the nearby backbone and because the resulting carboxylate anion is resonance-stabilized over two equivalent oxygens. At physiological pH (7.4), essentially all Asp/Glu side chains exist as the deprotonated carboxylate, carrying a full negative charge.

The basic family — lysine, arginine, and histidine. Lysine's side chain (–(CH₂)₄NH₂) is a primary aliphatic amine with pKa ~10.5, protonated to –NH₃⁺ at physiological pH. Arginine's side chain (–(CH₂)₃NH–C(=NH₂⁺)–NH₂ when protonated) terminates in a guanidinium group with pKa ~12.5, the most basic of all the side chains, because the cationic charge is delocalized symmetrically over three nitrogens. Histidine's side chain is an imidazole ring with pKa ~6.0 — closest of any side chain to physiological pH — making it uniquely able to act as either acid or base at pH 7.

The unifying physical concept is resonance delocalization of charge. A simple protonated primary amine (like lysine's) has no resonance stabilization at all — the positive charge sits on one nitrogen, period, which is why lysine's pKa (10.5) is close to a generic alkylamine's. Arginine's guanidinium cation, by contrast, delocalizes its positive charge symmetrically over three nitrogens via three resonance structures, an extraordinary stabilization that makes the conjugate acid (protonated form) enormously favored, pushing the pKa up to 12.5. Aspartate and glutamate show the mirror-image effect for anions: the carboxylate's negative charge is shared equally between two oxygens (confirmed by X-ray crystallography showing equal C–O bond lengths, intermediate between single and double bond), stabilizing the deprotonated (basic) form and thus lowering the pKa relative to what an isolated, non-resonance-stabilized anion would show.

The central quantitative relationship governing every ionizable side chain is the Henderson–Hasselbalch equation, which links the observed ratio of protonated to deprotonated forms to the pH and the group's intrinsic pKa.

where:

• pH is the solution hydrogen ion activity, $-\log_{10}[\text{H}^+]$
• p$K_a$ is the negative log of the acid dissociation constant for that specific ionizable group
• $[\text{A}^-]$ is the concentration of the deprotonated (conjugate base) form
• $[\text{HA}]$ is the concentration of the protonated (conjugate acid) form

Reading: at pH = p$K_a$, the group is exactly half-ionized; each pH unit away from p$K_a$ shifts the ratio by a factor of ten. This single equation, applied with the correct group-specific p$K_a$, tells you the ionization state (and hence the reactivity and net charge) of every Asp, Glu, His, Cys, Tyr, Lys, and Arg side chain at any pH — the detailed titration behavior and coupled equilibria are developed fully in Acid-Base Chemistry of Ionizable Side Chains. The second governing relationship connects the microscopic origin of a shifted pKa — resonance delocalization — to the macroscopic equilibrium constant via the free energy of ionization:

where:

• $\Delta G^{\circ}$ is the standard free energy change for the deprotonation reaction $\text{HA} \rightarrow \text{H}^+ + \text{A}^-$
• $R$ is the gas constant (1.987 cal mol⁻¹ K⁻¹)
• $T$ is absolute temperature (298 K at standard conditions)
• $K_a$ is the acid dissociation constant

Reading: resonance stabilization of the conjugate base (carboxylate) or conjugate acid (guanidinium, imidazolium) directly lowers or raises $\Delta G^{\circ}$ for ionization, and this free-energy change is what the measured pKa reports.

Example 1 — Foundational: ionization state of a glutamate side chain at gastric pH. A glutamate residue on the surface of pepsin experiences the strongly acidic environment of gastric fluid, pH 2.0. Glutamate's intrinsic side-chain pKa is 4.1. What fraction of glutamate side chains are deprotonated (carboxylate, catalytically/electrostatically active form) versus protonated (neutral –COOH)? Apply Henderson–Hasselbalch directly:

So the ratio of carboxylate to neutral acid is about 1 : 126. Converting to a fraction deprotonated:

Only 0.79% of glutamate side chains carry a negative charge at pH 2 — essentially all glutamates in pepsin are neutral (protonated) under these conditions, consistent with pepsin's known use of two active-site aspartate/glutamate-type carboxyl groups functioning largely as neutral acids (general acid/general base pair) rather than as ionized nucleophiles at its acidic pH optimum. Limiting case check: as pH → pKa (2.0 → 4.1), the ratio → 1 and $f_{A^-}$ → 50%, exactly the definition of pKa; the asymmetry of this example (pH far below pKa) correctly predicts near-complete protonation, the opposite of what would be found for the same residue at pH 7.4, where $f_{A^-} = 10^{7.4-4.1}/(1+10^{3.3}) \approx 99.95\%$ deprotonated.

Example 2 — Applied: estimating the number of ionized arginine residues in a DNA-binding protein. A zinc-finger transcription factor contains 14 arginine residues, all with an intrinsic side-chain pKa of 12.5 (assume no perturbation from the local environment for this estimate). At physiological pH 7.4, what fraction of these arginines are protonated (positively charged guanidinium, the form that engages phosphate backbone contacts), and how many of the 14 are charged at any instant?

Since Henderson–Hasselbalch as written gives the ratio of base (neutral guanidine) to acid (protonated guanidinium):

Essentially all 14 arginines (13.9999 out of 14, on average) remain protonated and positively charged at physiological pH — this is exactly why arginine, not lysine, is so heavily used at protein–DNA and protein–RNA interfaces: its very high pKa guarantees full positive charge across essentially the entire biological pH range, and its planar, three-nitrogen guanidinium group can form a bidentate hydrogen-bond/salt-bridge pair with a single phosphate oxygen (two simultaneous N–H···O contacts), a geometric advantage lysine's tetrahedral ammonium cannot match. Limiting case check: if instead the pKa were unusually low (e.g., an arginine buried in a hydrophobic pocket depressing its pKa toward 7.4 itself), $f_{\text{protonated}}$ would fall to 50%, showing explicitly how environment-dependent pKa shifts (developed quantitatively in Acid-Base Chemistry of Ionizable Side Chains) can, in principle, convert a "always charged" residue into a functionally titratable one — this is rare for Arg but well documented for buried His and occasionally Lys residues in enzyme active sites.

Example 3 — Exam/competition level: coupled equilibrium in the Asp–His–Ser catalytic triad. In chymotrypsin's catalytic triad, His57 (imidazole, intrinsic solution pKa ~6.0) is hydrogen-bonded to Asp102 (carboxylate) and simultaneously positioned to deprotonate Ser195 (alcohol, pKa ~15) during nucleophilic attack on substrate. Spectroscopic and mutagenesis studies show that removing Asp102 (Asp102Asn mutant) lowers catalytic efficiency ($k_{cat}$) by roughly $10^{3}$-to-$10^{4}$-fold, and NMR studies show His57's effective pKa in the native triad is shifted to approximately 7.0 (from a free-imidazole reference of 6.0). Using the free-energy relationship, calculate the free-energy stabilization provided by the Asp–His hydrogen bond that accounts for this pKa shift, and explain mechanistically why this raises catalytic power.

Setting up the free energy shift. The shift in pKa reflects a change in the relative stability of the protonated (imidazolium) versus neutral (imidazole) forms of His57 when Asp102 is present and hydrogen-bonded to the protonated Nδ–H, versus the free-imidazole reference state:

This ~1.4 kcal/mol reflects the extra stabilization the buried, hydrogen-bonded Asp102 carboxylate provides to the protonated (imidazolium) form of His57 — Asp's negative charge electrostatically and via a direct hydrogen bond stabilizes the developing positive charge on the imidazolium nitrogen, raising His57's effective pKa above the free-solution value and ensuring a larger resting population of the neutral, catalytically competent (basic, lone-pair-bearing) tautomer is still available to deprotonate serine at the moment of nucleophilic attack, while the transition state — where His57 has picked up the serine proton and become imidazolium — is stabilized by the adjacent Asp102 carboxylate exactly when the positive charge develops.

Mechanistic consequence. During the nucleophilic attack step, Ser195's oxygen attacks the substrate carbonyl carbon while His57 simultaneously accepts the departing serine O–H proton, becoming transiently imidazolium; Asp102's carboxylate — precisely because it sits adjacent and hydrogen-bonded to the other imidazole nitrogen — stabilizes this developing positive charge through electrostatic interaction, lowering the transition-state free energy. Quantitatively, using the Eyring relationship $\Delta G^{\ddagger} = -RT\ln(k_{cat}h/k_BT)$, a $10^{3}$-to-$10^4$-fold drop in $k_{cat}$ upon Asp102 mutation corresponds to:

This ~4.8 kcal/mol transition-state stabilization is considerably larger than the ~1.4 kcal/mol ground-state pKa shift calculated above, showing that Asp102's dominant catalytic role is transition-state electrostatic stabilization of the developing imidazolium charge (the "charge relay" concept), not merely raising His57's resting-state basicity — a distinction that historically took decades of kinetic and structural work (and considerable early confusion, since a pure pKa-shift model alone could not account for the full rate enhancement) to establish. Limiting case check: if Asp102 were replaced by a residue incapable of any hydrogen bonding or electrostatic stabilization at all (e.g., a fully inert hydrophobic side chain in a hypothetical mutant), $\Delta\Delta G^{\ddagger}$ would approach the full ground-state pKa-shift contribution only, and $k_{cat}$ would be expected to drop toward the lower end of, or even below, the experimentally observed $10^3$-fold range — consistent with the interpretation that both ground-state pKa tuning and transition-state charge stabilization contribute, with the latter dominating.

Common traps when reasoning about these side chains: 1. Treating free-amino-acid pKa values as fixed constants inside proteins. The table values are measured for the isolated amino acid or short model peptide in water; burial in a hydrophobic active site, proximity to other charges, and hydrogen bonding can shift a pKa by 2–5 units or more (a buried Asp in a hydrophobic core can have pKa above 7; an active-site Lys can be depressed below 8). This context-dependence is the entire subject of Acid-Base Chemistry of Ionizable Side Chains. 2. Confusing basicity with nucleophilicity. As derived above, thiolate's superior nucleophilicity toward carbonyls does not mean sulfur is "more basic" than oxygen — the conjugate-acid pKa order (alcohols > thiols) shows the opposite thermodynamic basicity trend. 3. Forgetting that Asn/Gln cannot ionize. Students sometimes lump Asn/Gln with Asp/Glu because of the structural relationship (amide vs. acid), but the amide nitrogen lone pair is resonance-delocalized into the carbonyl and the group carries no titratable proton near physiological pH — it never carries formal charge under normal conditions. 4. Assuming all three "basic" side chains behave identically. Lysine and arginine are essentially always fully protonated at physiological pH (both pKa values are far above 7), while histidine's near-neutral pKa means its charge state is genuinely poised and context-sensitive — this is precisely why histidine, not lysine or arginine, dominates as the general acid/base in enzyme catalysis.

The systematic understanding of amino acid side-chain acid-base chemistry owes an enormous debt to Lawrence Joseph Henderson and Karl Albert Hasselbalch, whose independent physiological-chemistry work on blood buffering in the 1910s produced the equation that bears both their names, and to Linus Pauling, whose The Nature of the Chemical Bond (1939) established the resonance framework — partial double-bond character in carboxylate, amide, and guanidinium groups — that explains why these particular pKa values take the values they do, rather than merely cataloguing them empirically.