Effects of temperature and flow rate on frontal and elution chromatography of aggregating systems

Phenomena of insulin peak fronting in size exclusion chromatography and strategies to reduce fronting

Insulin peak fronting in size exclusion chromatography (SEC) results in more than 10% yield loss in the production of insulin. The goal of this study is to understand the mechanisms of peak fronting and to develop strategies to reduce fronting and increase insulin yield. Chromatography experiments ruled out pressure surge, viscous fingering, and adsorption as the cause for peak fronting. Theoretical analysis based on a general rate model indicated that reversible dimerization is the major cause for peak fronting and reducing the dimerization equilibrium constant is the most effective method for reducing fronting. Two strategies were developed and tested to reduce the degree of dimer formation. The first strategy was to use 0.1N acetic acid as the presaturant and eluent. The second strategy was to use 0.8 or 2.8N acetic acid in 20vol.% denatured ethanol as the mobile phase. The experimental results showed that both strategies can reduce insulin peak fronting in SEC, maintain desired product purity, and increase insulin yield to higher than 98%.

Theoretical analysis of the effects of reversible dimerization in size exclusion chromatography

Reversible dimer formation in size exclusion chromatography (SEC) can cause peak splitting, merging, tailing, and fronting. Such behavior can be predicted by the association rate and the dissociation rate relative to the convection rate. Slow association and dissociation result in separated monomer and dimer peaks. Fast association and slow dissociation result in one single dimer peak. Slow association and fast dissociation result in one single monomer peak. Intermediate association and dissociation result in a merged, broad peak with either fronting when monomers dominate or tailing when dimers dominate. A diagram based on the two relative rates is generated to predict general peak shape and retention behavior in SEC.

In situ probing of insulin aggregation in chromatography effluents with spectroturbidimetry

Probing protein aggregation in situ is quite important for analyzing and developing chromatographic protein purification processes. A spectroturbidimetry method with a photodiode array detector is developed and tested for probing insulin aggregation in solution and determining the aggregation number, n(m). All aggregates examined are in the Rayleigh light scattering regime, where the turbidity between 400 and 350 nm is proportional to lambda(-4). Insulin at 25 degrees C in 3.5 N acetic acid is mainly monomeric (non-aggregated). At 25 degrees C and lower acetic acid concentrations, from 0.1 to 1 N, the average insulin aggregation number n(m) ranges from 2.9 to 1.6. Aggregates, with n(m) = 2-3, are found in 2.6 N acetic acid with 20 vol% acetonitrile. In 0.8 N acetic acid with 20 vol% denatured ethanol, n(m) = 1.2. At 4 degrees C, as acetic acid concentration decreases from 3.5 to 0.1 N, n(m) decreases from 2.4 to 1.8. In 2.8 N acetic acid with 20 vol% denatured ethanol at 4 degrees C, insulin exists mainly in monomer form. In situ probing of size exclusion chromatography, SEC, effluents in 3.5 N acetic acid at 4 degrees C shows n(m) = 1.6 at the fronting portion (a mixture of monomers and dimers or other oligomers) and n(m) = 1.1 (mostly monomers) at the tailing portion of the main peak. In another example, for LysPro-insulin in reversed phase chromatography at 4 degrees C, complex elution patterns and broad peaks are due to substantial aggregation. For a linear gradient of acetonitrile from 10 to 60 vol% at 4 degrees C, n(m) ranges from 2.2 to 12, in order of elution. For a linear gradient of ethanol from 30 to 50 vol% at 4 degrees C, n(m) ranges from 14 to 27, in order of elution. Analytical HPLC results at 25 degrees C imply that the aggregates are reversible.