Yeastolate is often used as a media supplement in industrial mammalian cell culture or as a major media component for microbial fermentations. Yeastolate variability can significantly affect process performance, but analysis is technically challenging because of its compositional complexity. However, what may be adequate for manufacturing purposes is a fast, inexpensive screening method to identify molecular variance and provide sufficient information for quality control purposes, without characterizing all the molecular components. Here we used Size Exclusion Chromatography (SEC) and chemometrics as a relatively fast screening method for identifying lot-to-lot variance (with Principal Component Analysis, PCA) and investigated if Partial Least Squares, PLS, predictive models which correlated SEC data with process titer could be obtained. SEC provided a relatively fast measure of gross molecular size hydrolysate variability with minimal sample preparation and relatively simple data analysis. The sample set comprised of 18 samples from 12 unique source lots of an ultra-filtered yeastolate (10 kDa molecular weight cut-off) used in a mammalian cell culture process. SEC showed significant lot-to-lot variation, at 214 and 280 nm detection, with the most significant variation, that correlated with process performance, occurring at a retention time of ∼6 min. PCA and PLS regression correlation models provided fast identification of yeastolate variance and its process impact. The primary drawback is the limited column lifetime (<300 injections) caused by the complex nature of yeastolate and the presence of zinc. This limited long term reproducibility because these age-related, non-linear changes in chromatogram peak positions and shapes were very significant.
Keywords: Cell culture media; Chemometrics; Screening; Size exclusion chromatography; Yeastolate.
Copyright © 2023 The Authors. Published by Elsevier B.V. All rights reserved.