Introduction
PS20 and PS80 are the most commonly used surfactants in
biopharmaceutical protein formulation that can improve protein stability
and protect drug products from aggregation and denaturation during
delivery and transportation (Kiese, Papppenberger, Friess, & Mahler,
2008; Martos et al., 2017) . However, polysorbate is labile to
hydrolysis and the hydrolyzed products can lead to the formation of
undesired particulates in the formulated drug substances(Hall, Sandefur,
Frye, Tuley, & Huang, 2016). The mechanism of hydrolysis of
polysorbates had been widely studied by multiple groups across industry.
In addition to acid and base-induced polysorbate hydrolysis, host cell
derived lipases or esterase co-purified with the final drug substance
have also been shown to degrade polysorbate, with distinct degradation
profiles unique to PS20 or PS80. For example, porcine liver esterase was
reported to be able to specifically hydrolyze PS80 but not PS20,
resulting in the formation of PS85 over time in mAb drug
product(Labrenz, 2014). Group XV lysosomal phospholipase A2 isomer X1
(LPLA2) demonstrated the ability to degrade monoesters in both PS20 and
PS80 at a concentration less than 1 ppm(Cheng et al., 2019; Hall et al.,
2016). Unique degradation pattern in both PS20 or PS80 was found in a
range of carboxylesters, including pseudomonas cepacia lipase on
immobead 150 (PCL), candida antarctica lipase B on immobead 150 (CALB),
thermomyces lanuginosus lipase on immobead 150 (TLL), rabbit liver
esterase (RLE), Candida antarctica lipase B (CALB) and porcine
pancreatic lipase type II (PPL)(McShan, Kei, Ji, Kim, & Wang, 2016).
In 2016, phospholipase B-like 2 (PLBD2) was first reported as the
residual host cell protein present in a sulfatase drug product that can
degrade polysorbate 20 over time(Dixit, Salamat-Miller, Salinas, Taylor,
& Basu, 2016). That conclusion was made based on the experiment were
PS20 degradation can be accelerated upon spiking in the recombinant
human PLBD2 protein. However, as the recombinant PLBD2 used in the study
only had a purity of ~90%, the enzymatic contribution
from impurities could not be ruled out.
In this study, we engineered a PLBD2 knockout CHO host cell line.
Several monoclonal antibodies were expressed from the PLBD2 knockout CHO
and wild type CHO host cell lines. The purified mAbs from the PLBD2
knockout cell line showed no significant difference in lipase activity
compared to the same product produced from the control CHO cell line
with wild type PLBD2. Consistent with this result, the removal of PLBD2
protein in mAbs purified from wild type CHO by immuno depletion with an
anti-PLBD2 antibody did not significantly decrease lipase activity. In
addition, our PLBD2 quantitation analysis of multiple formulated mAb
products showed no correlation between the amount of PLBD2 in the sample
and the extent of PS20 degradation. Taken together, our data suggest
that PLBD2 itself does not contribute significantly to polysorbate
degradation.
Proteomics analysis of the commercially available recombinant PLBD2
identified several additional lipases, providing a likely explanation
for the initial misidentification of PLBD2 as the root course of PS20
degradation.