Acid-Base Chemistry of the Alpha-Amino and Alpha-Carboxyl Groups

Knowing the alpha-amino and alpha-carboxyl pKa values (and knowing that they vanish upon peptide bond formation) underlies practical techniques across biochemistry and biotechnology. Isoelectric focusing, a core step in two-dimensional gel electrophoresis for proteomics, separates proteins by their pI — a quantity built up residue-by-residue from side-chain pKa's but, for small peptides and free amino acids, dominated by the alpha-group pKa's discussed here. Ion-exchange chromatography, used industrially to purify amino acids from fermentation broths (for example large-scale monosodium glutamate or lysine production) and in the lab to resolve amino acid mixtures via post-column ninhydrin detection (the classical amino acid analyzer), relies directly on the pH-dependent net charge computed exactly as in the worked examples above. In peptide synthesis, both solid-phase (Merrifield) and solution-phase methods must protect the alpha-amino group (commonly with Fmoc or Boc protecting groups) and activate the alpha-carboxyl group for each coupling step, precisely because free alpha-amino and alpha-carboxyl reactivity — governed by the pKa's and protonation states described here — would otherwise cause uncontrolled self-condensation or wrong-partner coupling. Mass spectrometry-based proteomics exploits the predictable ionizability of the N-terminal alpha-amino group (readily protonated under electrospray ionization's acidic conditions) to generate the b/y-ion fragmentation series used in peptide sequencing. And in enzymology, active-site catalytic mechanisms that use "generic" amino/carboxyl acid-base chemistry (for instance serine protease catalytic triads, or lysine-mediated Schiff-base formation in aldolases) must reckon with exactly this kind of pKa perturbation — a lysine's side-chain amino group buried in a hydrophobic active-site pocket, or sitting near another charged residue, can have its pKa shifted by several units from the free-amino-acid value, by the same inductive/electrostatic logic worked out here for the free alpha-groups; that generalization is the subject of Acid-Base Chemistry of Ionizable Side Chains.

Ask the tutor: How does adding an ionizable side chain (like aspartate's or lysine's) change the pI formula and the shape of the titration curve compared with a simple diprotic amino acid like glycine?

Consider glycine, H₂N–CH₂–COOH, the simplest amino acid. In the solid state and in water, it does not exist as this neutral-drawn structure at all. Both the amino and carboxyl groups are ionizable, and the pKa of the carboxyl (about 2.34) is far below neutral pH while the pKa of the ammonium (about 9.60) is far above it. At pH 7, essentially all carboxyl groups have given up their proton (existing as -COO⁻) and essentially all amino groups have kept theirs (existing as -NH₃⁺). The dominant species in neutral water is therefore ⁺H₃N–CH₂–COO⁻: a zwitterion, net charge zero, but built from two charged groups rather than two neutral ones. Three protonation states are possible for glycine as pH is swept from strongly acidic to strongly basic. At very low pH (pH ≪ 2.34), both groups are protonated: the fully cationic form ⁺H₃N–CH₂–COOH, net charge +1. Between the two pKa values, the zwitterion ⁺H₃N–CH₂–COO⁻ dominates, net charge 0. At very high pH (pH ≫ 9.60), both groups are deprotonated: H₂N–CH₂–COO⁻, net charge −1. The transition between each pair of adjacent states is governed by Henderson–Hasselbalch, treating each ionizable group independently (an excellent approximation when the two pKa values are well separated, as they are here — more on the limits of this approximation below). Why are these pKa values shifted so far from those of simple analogs? Acetic acid, CH₃–COOH, has pKa ≈ 4.76; ethylamine (or methylamine) has pKa ≈ 10.6–10.7. Glycine's carboxyl is about 1000-fold more acidic (pKa 2.34 vs 4.76, a difference of 2.4 pKa units) and its amino group is about 10-fold more acidic than a simple primary amine (pKa 9.60 vs ~10.6). The culprit in both cases is the same: the other ionizable group sitting on the adjacent carbon, exerting an inductive (through-bond, electron-withdrawing) effect and, once charged, an additional through-space electrostatic effect. Take the carboxyl group first. In neutral glycine's cationic form, the ammonium nitrogen (-NH₃⁺) is strongly electron-withdrawing inductively — nitrogen is more electronegative than carbon and, being positively charged, pulls electron density toward itself through the sigma-bond framework. This inductive pull partially delocalizes negative charge away from the developing carboxylate oxygen when the -COOH proton dissociates, stabilizing the conjugate base (-COO⁻) relative to the neutral acid. A more stable conjugate base means a stronger acid — a lower pKa. The effect is compounded electrostatically: once the carboxylate anion forms, it sits only three bonds from a cationic ammonium group, and the direct Coulombic attraction between the two opposite charges further stabilizes the ionized (deprotonated) form. Both effects point the same way, and together they drop the pKa from 4.76 (acetic acid) to about 2.34. The amino group experiences the mirror-image logic, but the electrostatics reverse sign. Consider the zwitterion ⁺H₃N–CH�2–COO⁻ losing its ammonium proton to give the doubly-anionic H₂N–CH₂–COO⁻. The adjacent carboxylate is now anionic and, being electron-rich, is inductively electron-donating relative to a neutral group — but more importantly, the ammonium cation and carboxylate anion in the zwitterion attract each other electrostatically, and this attraction stabilizes the protonated (ammonium) form rather than the deprotonated one. Removing the proton to form the neutral amine breaks this favorable electrostatic stabilization, destabilizing the product state relative to what a simple amine (with no such neighbor) would experience. That makes the amino group's proton easier to remove — the pKa drops from ~10.6 (simple primary amine) to 9.60. Put more carefully in one unified statement: the -NH₃⁺ group stabilizes the adjacent -COO⁻ (favoring carboxyl deprotonation, lowering its pKa) while the -COO⁻ group destabilizes the adjacent -NH₃⁺'s conjugate base relative to an isolated case in a way that actually makes deprotonation of the ammonium relatively more favorable than it would be next to a neutral group — because in the fully-protonated cation ⁺H₃N-CH₂-COOH, the ammonium doesn't yet have that electrostatic assistance from a caboxylate (COOH is neutral), so its intrinsic pKa in that state is closer to a normal amine; it's only once the adjacent group becomes anionic (in the zwitterion) that its charge-charge stabilization of the ammonium proton is lost upon further deprotonation, tipping the true observed pKa down modestly. The dominant, unambiguous effect for both groups is the simple electron-withdrawing inductive pull of a nearby full or partial positive charge, amplified by direct charge-charge (Coulombic) interactions once ionization has actually occurred.

The starting point, as always for a weak acid HA in equilibrium with its conjugate base A⁻, is:

where: $K_a$ is the acid dissociation constant, $[\text{H}^+]$, $[\text{A}^-]$, $[\text{HA}]$ are equilibrium concentrations. For a diprotic amino acid such as glycine, there are two such equilibria in series, because the molecule carries two independently titratable protons:

where: $K_{a1}$ governs loss of the carboxyl proton (numerically the more acidic group), $K_{a2}$ governs loss of the ammonium proton. The Henderson–Hasselbalch equation applies to each equilibrium separately, since the two pKa values are well separated (ΔpKa ≈ 7 units, so cross-contamination between the two titrations is negligible):

where: the ratio inside each log is [deprotonated form]/[protonated form] for that specific ionizable group. The isoelectric point — the pH at which the average net charge on the molecule is exactly zero — for a simple amino acid with only two ionizable groups is the arithmetic mean of the two flanking pKa values:

where: this formula holds specifically because the zwitterion (net charge 0) is flanked on one side by the +1 cation and on the other by the −1 anion, and pI is defined as the pH exactly halfway (in log-concentration terms) between full population of the cation-to-zwitterion transition and the zwitterion-to-anion transition. Finally, the shift in pKa induced by a neighboring charged or polar group can be estimated from the extra electrostatic work needed to remove a proton in the presence versus the absence of that neighbor, via a simple free-energy bookkeeping:

where: $\Delta\Delta G^\circ$ is the difference in standard free energy of deprotonation with and without the neighboring group, $R = 8.314\ \text{J mol}^{-1}\text{K}^{-1}$, $T$ is temperature in kelvin, and the factor $2.303 = \ln 10$ converts natural log to base-10 pKa units.

Example 1 (foundational) — glycine at pH 1.00. At strongly acidic pH, we expect glycine to be almost entirely in its cationic form. Using the same fractional framework with pKa₁ = 2.34, pKa₂ = 9.60:

Since pH = 1.00 is far below pKa₂, $r_2 = 10^{(1.00-9.60)} = 10^{-8.6} \approx 2.5\times10^{-9}$, negligible. So

At pH 1.00, about 96% of glycine is the fully protonated cation ⁺H₃N-CH₂-COOH, with only 4% still in the zwitterionic form — consistent with pH 1.00 being 1.34 units below pKa₁, so the HH ratio predicts roughly a 22:1 excess of protonated over deprotonated carboxyl, matching the direct calculation. This is the textbook titration-curve region where the carboxyl group is being actively titrated and the amino group is essentially untouched (100% protonated) throughout.

Example 2 (applied) — designing an ion-exchange separation of glycine and glutamate at pH 6.0. Suppose a biochemist wants to separate glycine (pKa's 2.34, 9.60; a simple amino acid with no ionizable side chain) from glutamate (whose side-chain carboxyl has pKa ≈ 4.25, in addition to alpha-COOH pKa ≈ 2.19 and alpha-NH₃⁺ pKa ≈ 9.67) using an anion-exchange resin (a positively charged stationary phase that binds anions) running at pH 6.0. At pH 6.0, glycine's alpha-carboxyl (pKa₁ = 2.34) is fully deprotonated (pH is 3.66 units above pKa₁, so essentially 100% -COO⁻) and its alpha-amino (pKa₂ = 9.60) is fully protonated (pH is 3.60 units below pKa₂, so essentially 100% -NH₃⁺). Net charge on glycine at pH 6.0:

So glycine is essentially 100% zwitterionic, net charge ≈ 0 — it will not bind an anion-exchange resin (no net anionic character) and will elute early/unretained. Glutamate at pH 6.0, however, has three ionizable groups: alpha-COOH (pKa 2.19, fully deprotonated at pH 6.0, since 6.0 − 2.19 = 3.81 pH units above pKa), the side-chain gamma-COOH (pKa 4.25, also essentially fully deprotonated at pH 6.0, since 6.0 − 4.25 = 1.75 units above pKa, giving a deprotonated fraction of $10^{1.75}/(1+10^{1.75}) = 56.2/57.2 = 0.983$, i.e., 98.3% deprotonated), and the alpha-NH₃⁺ (pKa 9.67, fully protonated at pH 6.0). Net charge:

Glutamate carries a net charge near −1 at pH 6.0 (dominated by the ionized side chain), so it binds strongly to the anion-exchange resin while glycine (net charge ≈ 0) flows through unretained — this is exactly the physical basis of amino-acid separation by ion-exchange chromatography, and it hinges entirely on knowing the pKa values of every ionizable group, including — crucially — the alpha-amino and alpha-carboxyl groups that are common to both amino acids and therefore don't differentiate them; the side chain is what does the differentiating here, foreshadowing why Acid-Base Chemistry of Ionizable Side Chains matters just as much as the backbone chemistry covered here.

Example 3 (exam/competition level) — pKa shift of the alpha-amino group upon peptide bond formation, and why internal glycines have no titratable alpha-groups. Consider the tripeptide Ala-Gly-Ser. The question: which alpha-amino and alpha-carboxyl groups in this tripeptide are still titratable weak acids/bases with pKa values close to the free amino acid values, and which have vanished? Structurally, a tripeptide has the backbone

Only the N-terminal alanine retains a free alpha-amino group (⁺H₃N–), and only the C-terminal serine retains a free alpha-carboxyl group (–COO⁻). The internal glycine's alpha-amino nitrogen has become an amide nitrogen (–CO–NH–CH₂–), and its original alpha-carboxyl carbon has become an amide carbonyl carbon (–CH₂–CO–NH–). Neither is titratable in the ordinary sense over the physiological pH range (amide N-H has pKa around 15–17 as an acid and amide nitrogen lone pair is barely basic, pKaH around −1, both essentially non-ionizable near neutral pH) — because the amide resonance

delocalizes the nitrogen lone pair into the carbonyl pi system, removing the basicity that a free primary or secondary amine would have, and simultaneously making the carbonyl far less electrophilic/acidic in the relevant sense than a free carboxylic acid. Quantitatively: the N-terminal alanine's free alpha-amino group typically has pKa ≈ 9.69 (close to, but not identical to, free alanine's 9.87, because the adjacent peptide carbonyl is a weaker electron-withdrawing perturbation than a free carboxyl two bonds further along would be) — but the point of the exercise is that this is the only amino group titrating anywhere near pH 9–10 in the whole tripeptide; the "α-amino pKa" of the internal glycine residue simply does not exist as an experimentally accessible acid-base equilibrium. Likewise the C-terminal serine's free alpha-carboxyl has pKa ≈ 2.2, the only carboxyl titrating anywhere near pH 2 in the peptide. A titration curve of Ala-Gly-Ser therefore shows only two alpha-backbone inflections (near pH 2 and near pH 9-10) rather than three pairs (six inflections) that would appear if each residue's alpha-groups were independently titratable — directly demonstrating, from first principles of amide resonance, why peptide bond formation is acid-base-silencing for every internal residue's backbone groups. (Side-chain ionizations, e.g. serine's hydroxyl at very high pH ≈ 13, are a separate story, covered elsewhere.)

The systematic understanding of amino acid acid-base behavior, and the zwitterion concept in particular, emerged from the meeting of physical organic chemistry and the emerging science of protein chemistry in the early twentieth century. The zwitterion hypothesis for amino acids is generally credited to work following from Niels Bjerrum's 1923 theoretical treatment of dipolar ions, and the careful potentiometric titration curves that established the actual numerical pKa values for each of the 20 standard amino acids were largely the work of Edwin J. Cohn and John T. Edsall, whose 1943 monograph Proteins, Amino Acids and Peptides as Ions and Dipolar Ions became the definitive reference compiling titration data, dipole moments, and thermodynamic parameters for the amphoteric behavior of amino acids — establishing, among other things, the now-standard tabulated pKa values (glycine 2.34/9.60, alanine 2.34/9.87, and so on) still quoted in biochemistry textbooks today.