The research team, working with milk from Xinjiang Bactrian camels, first established a reference standard for CLF using heparin affinity chromatography, confirming its identity through liquid chromatography-tandem mass spectrometry and ultra-high-performance liquid chromatography. With a reliable quantification method in hand, they turned to cation-exchange chromatography (CEX), the workhorse technique already proven for large-scale bovine lactoferrin production. CEX exploits electrostatic interactions: most acidic whey proteins carry negative charges under binding buffer conditions and flow through the column unbound, while lactoferrin, with its high theoretical net positive charge, is captured by the CM-Sepharose Fast Flow resin and later eluted with increasing salt concentrations.

Three elution strategies were compared before optimization began. Linear and multi-step gradients induced severe peak broadening, whereas a two-step approach—washing weakly bound impurities, mainly lactoperoxidase, with 40 percent Buffer B (0.4 M NaCl) before eluting the target with 100 percent Buffer B (1 M NaCl)—produced sharp, high-purity CLF peaks. SDS-PAGE of the target fractions displayed a single band, and LC-MS/MS confirmed camel lactotransferrin as the major constituent. The team then systematically varied three parameters through single-factor experiments: mobile phase flow rate (0.5 to 1.5 mL/min), buffer pH (6.0 to 8.0), and impurity elution salt concentration (0.1 to 0.5 M NaCl).

The single-factor results followed classical ion-exchange theory. Purity peaked at 1.00 mL/min, where residence time balanced adsorption efficiency against throughput; slower rates caused extensive co-adsorption of impurities, while faster rates led to premature target elution. Buffer pH proved critical because it modulates protein net surface charge: under acidic conditions (pH 6.0–6.5), CLF and impurities carried positive charges and co-eluted, while near-neutral pH (7.0–7.5) weakened the binding of acidic impurities, enabling effective separation. At pH 8.0, the buffer approached the theoretical isoelectric point of CLF (8.5–9.0), reducing net charge and weakening matrix binding. Impurity elution concentration showed the strongest effect, with 30 to 40 percent Buffer B sufficient to fully remove contaminants before target elution.

Response surface methodology with a Box-Behnken design then refined the parameters. The resulting quadratic regression model explained 98.49 percent of total purity variation, with a non-significant lack-of-fit test confirming no systematic deviation. Factor influence ranked impurity elution concentration highest, followed by buffer pH, with flow rate exerting the weakest effect. The model predicted optimal conditions of 1.01 mL/min, pH 7.09, and 32.78 percent eluent, yielding a maximum predicted purity of 90.4 percent. Rounded to practical values of 1 mL/min, pH 7.0, and 30 percent eluent (0.3 M NaCl), the final CLF preparation reached 90.13 ± 1.06 percent purity. Validation experiments showed relative standard deviations for precision, repeatability, stability, and spiked recovery all below 5 percent, confirming excellent reproducibility and robustness.

With a reliable purification workflow established, the team compared the thermal stability of CLF and bovine lactoferrin (BLF) under two industrial heating regimes: 72 °C for conventional pasteurization and 95 °C for intensified pasteurization, each paired with holding durations of 15 seconds, 2 minutes, and 10 minutes. Iron saturation measurements showed the purified proteins were closely matched—27.19 percent for BLF and 33.38 percent for CLF—ensuring that observed differences in thermal response reflected species-specific molecular features rather than saturation status. Differential scanning calorimetry then revealed a striking divergence: CLF’s N-lobe melted at 67 °C, lower than BLF’s 72 °C, while CLF’s C-lobe melted at 91 °C, slightly above BLF’s 89 °C. The two proteins thus exhibited opposite thermal trends for their N- and C-lobes, likely stemming from differences in amino acid composition, rigid region distribution, and native tertiary folding.

Intrinsic fluorescence and UV–vis spectroscopy tracked changes in the microenvironment around buried aromatic residues. After heating at 95 °C for 10 minutes, CLF’s apparent fluorescence intensity increased by 212.07 percent, compared with only 36.80 percent for BLF, indicating substantially greater exposure of previously buried tryptophan residues. UV–vis spectra showed a stronger hyperchromic effect around 280 nm for CLF under moderate heating, confirming greater aromatic-residue exposure. However, BLF exhibited stronger long-wavelength light scattering in the 350 to 600 nm region, suggesting the formation of larger intermolecular assemblies. Circular dichroism spectroscopy added secondary-structure detail: under severe heating, BLF’s fitted α-helical fraction dropped from 22.8 percent to 7.6 percent while β-sheet rose to 31.9 percent, and CLF showed a comparable loss of ordered structure with a marked increase in random coil.

Dynamic light scattering resolved the apparent contradiction between CLF’s greater conformational perturbation and its weaker light-scattering response. Native BLF and CLF both showed low hydrodynamic diameters and polydispersity indices. At 95 °C, CLF maintained lower particle sizes across all time points, indicating a lower tendency to form large aggregates, while BLF showed a sharp polydispersity increase at 15 seconds, reached maximum diameter at 2 minutes, and then redistributed. The team proposed a distinctive mechanism: CLF adopts a compact global architecture with a disordered interior under high temperature, and its exposed hydrophobic moieties are shielded through intramolecular interactions—salt bridges and hydrophobic contacts—that suppress intermolecular association. Thermally perturbed BLF, by contrast, exhibited a greater tendency toward intermolecular association and large molecular assemblies.

Molecular dynamics simulations running 100 nanoseconds at 25 °C, 72 °C, and 95 °C supported this hypothesis computationally. At 95 °C, CLF displayed larger residue-level fluctuations than BLF, confirming its greater local conformational vulnerability, yet it retained a persistently lower radius of gyration and declining solvent-accessible surface area, indicating that compensatory noncovalent interactions preserved its compact spatial framework. Secondary-structure analysis during simulation showed CLF always contained a higher fraction of random coil and less α-helix and β-sheet than BLF at identical temperatures, with a more remarkable growth in random coil upon heating. Flexible residues concentrated within the N-lobe for both proteins, consistent with the calorimetric finding that the N-terminal domain is heat-sensitive while the C-lobe is more thermostable.

The study’s authors emphasize that aggregation propensity is not determined solely by the extent of thermally induced structural perturbation but also by the conformational state and intermolecular association behavior of the thermally perturbed proteins. CLF’s unique thermal pattern—compact overall morphology with disordered interior domains—offers experimental evidence for species-specific thermal denaturation pathways. The optimized CEX protocol, with its outstanding repeatability and potential for industrial processing, provides a transferable workflow for large-scale CLF manufacture. Future investigations will address bioactivity retention of thermally treated CLF and its conformational changes in complex milk matrices, guiding the exploitation of camel milk resources in high-value dairy products including infant formula and sports nutrition foods.

Subject of Research: Biochemistry

Article Title: Optimization of purification process and thermal stability of lactoferrin from camel milk

Article References: Liu, Y., Liu, Y., Zhang, H., Huanbieke, M., Ma, L., Ran, C., Cheng, Y., Yang, J., Yue, H., & Qu, Y. (2026). Optimization of purification process and thermal stability of lactoferrin from camel milk. Food Chemistry: X, Article 104616. https://doi.org/10.1016/j.fochx.2026.104616

Image Credits: AI Generated

DOI: 10.1016/j.fochx.2026.104616

Keywords: camel milk, lactoferrin, protein purification, thermal stability, molecular dynamics, food chemistry, Optimization, purification, process, thermal, stability, camel