Polyphenol oxidase silencing avoids protein cross‐linking and enzymatic browning in <i>Nicotiana benthamiana</i> leaf extracts

Browning of extracts during purification of recombinant proteins from agroinfiltrated leaves is a widely observed phenomenon that is commonly quelled with reducing agents and absorbing materials. In fruits and vegetables, browning results from the oxidation of phenolics into brown quinones which react with themselves and other molecules to form a brown melanin-like polymer (Figure 1a, Sui et al., 2023). The oxidation of phenolics is catalysed by polyphenol oxidase (PPO), which is localized to chloroplasts but oxidizes vacuolar phenolics upon cell disruption. Enzymatic browning in sliced apple and bruised potato has been suppressed by silencing PPO (Carter, 2012; González et al., 2020). To prevent browning of extracts from agroinfiltrated leaves, we depleted PPO transcripts by virus-induced gene silencing (VIGS). The N. benthamiana genome (Ranawaka et al., 2023) contains six PPO genes, of which five are predicted to encode putative functional enzymes (Figure 1b). However, only NbL17g16540 is significantly expressed in leaves, and its expression increased upon agroinfiltration (Figure 1c). We cloned a 300 bp fragment to silence NbL17g16540 and its homeolog (NbL10g18390, Figure S1) into RNA2 of the bipartite genome of tobacco rattle virus (TRV). Young N. benthamiana plants were co-agroinoculated with TRV1 and TRV2 carrying fragments of PPO or GUS (β-glucuronidase fragment, negative control). Three weeks post-infiltration, TRV::PPO plants showed no growth or developmental phenotypes compared to TRV::GUS plants (Figure 1d). Western blot analysis of leaf extracts from these plants confirms that PPO was successfully depleted from TRV::PPO plants (Figure 1e). Importantly, cleared leaf extracts of TRV::PPO plants in Tris-buffered saline (TBS) remained green after 4 h of incubation at 4 °C, in contrast to browning observed in extracts of TRV::GUS plants (Figure 1f). When these incubated extracts were separated on protein gels, TRV::PPO samples revealed much stronger signals at 55 kDa, whereas TRV::GUS samples showed high molecular weight (HMW) signals at >180 kDa (Figure 1g). Western blot analysis revealed that much of the large subunit of ribulose bisphosphate carboxylase large chain (RBCL) runs at HMW in the TRV::GUS sample, unlike the 55 kDa signal found in TRV::PPO samples (Figure 1g). This indicates that PPO catalyses cross-linking of RBCL, possibly fixing RBCL tetramers in the multimeric RBCL complex (Duff et al., 2000). The HMW signal was less prominent at the t = 0 time point (Figure 1h), indicating that cross-linking occurs during incubation. To examine the impact of PPO depletion on protein expression and accumulation, we transiently expressed cytoplasmic GFP in TRV::GUS and TRV::PPO plants. Interestingly, extracts from TRV::PPO plants contained significantly higher levels of GFP protein than the TRV::GUS control plants (Figure 1h). Further, the intensity of the GFP signal is reduced in TRV::GUS extracts upon incubation for 4 h at 4 °C but not in TRV::PPO plants. However, we did not detect HMW complexes containing GFP in TRV::GUS control plants (Figure 1h). Measuring GFP fluorescence directly from leaves revealed a significant, 2.8-fold higher GFP fluorescence from TRV::PPO plants when compared to TRV::GUS plants (Figure 1i), consistent with Western blot analysis and supporting the increased GFP accumulation in TRV::PPO plants. Although the underlying mechanism behind the increased transient expression is unclear, this feature might be a valuable additional advantage of PPO silencing. Thus, PPO silencing avoids enzymatic browning and protein cross-linking, and this may increase the yield and quality of purified proteins and improve routinely performed experiments such as co-immunoprecipitation and metabolic experiments. While PPO silencing does not affect plant growth or development, it may reduce immunity to pests and pathogens (Zhang and Sun, 2021). Alternative ways to deplete PPO activity include genome editing, the use of chemical or protein-based inhibitors or physical methods such as those used in the food industry (Sui et al., 2023). We would like to thank Ursula Pyzio for excellent plant care; Felix Homma for supporting RNA-Seq analysis; Sarah Rodgers, Caroline O'Brian and Patricia Bowman for technical support; and George Lomonossoff for providing pEAQ vectors. This project was financially supported by ERC project 101019324 (CH and RH) and BBSRC projects DDT00230 (EW); BB/W013932/1 (SS, KZ and RH) and BB/Y00969X/1 (KZ and RH). RH conceived the project; CM performed most experiments with the help of EW, KZ and SS; RH wrote the manuscript with the help of all authors. None declared. The data that supports the findings of this study are available in the supplementary material of this article. Data S1 Supplemental materials. Table S1 Used plasmids. Table S2 Oligonucleotides. Figure S1 Alignment of VIGS fragment with 6 PPO genes of N. benthamiana. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.