Transient proteolysis reduction of Nicotiana benthamiana-produced CAP256 broadly neutralizing antibodies using CRISPR/Cas9

The hypersensitive response is elicited by Agrobacterium infiltration of Nicotiana benthamiana, including the induction and accumulation of pathogenesis-related proteins, such as proteases. This includes the induction of the expression of several cysteine proteases from the C1 (papain-like cysteine protease) and C13 (legumain-like cysteine protease) families. This study demonstrates the role of cysteine proteases: NbVPE-1a, NbVPE-1b, and NbCysP6 in the proteolytic degradation of Nicotiana benthamiana (glycosylation mutant ΔXTFT)-produced anti-human immunodeficiency virus broadly neutralizing antibody, CAP256-VRC26.25. Three putative cysteine protease cleavage sites were identified in the fragment crystallizable region. We further demonstrate the transient coexpression of CAP256-VRC26.25 with CRISPR/Cas9-mediated genome editing vectors targeting the NbVPE-1a, NbVPE-1b, and NbCysP6 genes which resulted in a decrease in CAP256-VRC26.25 degradation. No differences in structural features were observed between the human embryonic kidney 293 (HEK293)-produced and ΔXTFT broadly neutralizing antibodies produced with and without the coexpression of genome-editing vectors. Furthermore, despite the presence of proteolytically degraded fragments of plant-produced CAP256-VRC26.25 without the coexpression of genome editing vectors, no influence on the in vitro functional activity was detected. Collectively, we demonstrate an innovative in planta strategy for improving the quality of the CAP256 antibodies through the transient expression of the CRISPR/Cas9 vectors.

The hypersensitive response is elicited by Agrobacterium infiltration of Nicotiana benthamiana, including the induction and accumulation of pathogenesis-related proteins, such as proteases. This includes the induction of the expression of several cysteine proteases from the C (papain-like cysteine protease) and C (legumain-like cysteine protease) families. This study demonstrates the role of cysteine proteases: NbVPE-a, NbVPE-b, and NbCysP in the proteolytic degradation of Nicotiana benthamiana (glycosylation mutant XTFT)-produced anti-human immunodeficiency virus broadly neutralizing antibody, CAP -VRC . . Three putative cysteine protease cleavage sites were identified in the fragment crystallizable region. We further demonstrate the transient coexpression of CAP -VRC . with CRISPR/Cas -mediated genome editing vectors targeting the NbVPE-a, NbVPE-b, and NbCysP genes which resulted in a decrease in CAP -VRC . degradation. No di erences in structural features were observed between the human embryonic kidney (HEK )-produced and XTFT broadly neutralizing antibodies produced with and without the coexpression of genome-editing vectors. Furthermore, despite the presence of proteolytically degraded fragments of plant-produced CAP -VRC . without the coexpression of genome editing vectors, no influence on the in vitro functional activity was detected. Collectively, we demonstrate an innovative in planta strategy for improving the quality of the CAP antibodies through the transient expression of the CRISPR/Cas vectors. KEYWORDS plant biotechnology, CRISPR/Cas , genome editing, Nicotiana benthamiana, proteases, immunoglobulin G, human immunodeficiency virus Introduction Broadly, neutralizing antibodies (bNAbs) are antibodies with the ability to neutralize multiple HIV-1 strains by targeting conserved epitopes of the virus (Burton and Hangartner, 2016). Antibody neutralization is exclusively targeted to the Env, due to it being the only surface-exposed virus-encoded protein (Davenport et al., 2011). One such antibody lineage is the CAP256-VRC26 antibodies, which are highly potent bNAb against HIV-1 subtypes A and C strains Doria-Rose et al., 2015). The CAP256-VRC26 bNAbs lineage, like PG9, targets the trimeric variable regions 1 and 2 (V1V2) region of the HIV-1 gp120 envelope glycoprotein (McLellan et al., 2011;Doria-Rose et al., 2014;Bhiman et al., 2015). Transient plant-based production of the CAP256-VRC26.08 and CAP256-VRC26.09, along with numerous other anti-HIV antibodies (Abs), have been demonstrated in Nicotiana benthamiana (N. benthamiana) (Loos et al., 2015;Castilho et al., 2018;Stelter et al., 2019;Singh et al., 2020). Nicotiana species allow for the modulation of posttranslational modifications (PTMs), such as glycosylation and sulfation, which are critical to the functioning of these antibodies (Strasser et al., 2008(Strasser et al., , 2009Loos et al., 2015;Singh et al., 2020).
The inclusion of PTMs on Abs requires transition through the secretory pathway to either the cytosol, vacuole, or apoplast; of which all these environments are enriched with either LLCPs or PLCPs, or both (Hara-Nishimura et al., 2005;Goulet et al., 2012;Hatsugai et al., 2015). Apart from the physiological roles of these LLCPs and PLCPs, these protease families have also been implicated in the degradation of plant-produced anti-HIV antibodies (Niemer et al., 2014;Hehle et al., 2015;Paireder et al., 2017). Proteolytic degradation is not limited to in planta degradation during protein expression, ex planta is also possible during extraction and downstream processing (Rivard et al., 2006;Benchabane et al., 2008;Mandal et al., 2014;Niemer et al., 2014;Hehle et al., 2015). The reduction of protease activity could potentially improve both the quality and yield of the produced recombinant protein (Paireder et al., 2017). It had been demonstrated that the downregulation of the PLCP, CysP6, in Nicotiana tabacum (N. tabacum) had a beneficial effect on the accumulation of recombinantly produced protein (Duwadi et al., 2015). VPEs are intricately linked with the level of activity of PLCPs as some VPEs are responsible for the activation of some PLCPs (Okamoto and Minamikawa, 1999;Roberts et al., 2012;Cilliers et al., 2018).

In silico protease cleavage site analysis
Proteolytic cleavage analysis was conducted as previously described by Pillay et al. (2016). Briefly, proteolytic cleavage analysis for cysteine protease families, such as C1, C13, and C14, was conducted in silico using CLC Main Workbench 7.7 (http://www.clcbio.com) based on the various protease substrate specificities (Mathieu et al., 2002;Choe et al., 2006) and on the Schechter and Berger subsite nomenclature Berger, 1967, 1968).

T endonuclease I based confirmation of genome editing
Confirmation of CRISPR/Cas9-mediated disruption of the targeted cysteine protease genes was determined using the T7 endonuclease I assay. The assay was adapted from the experimental procedure outlined by Guschin et al. (2010). T7 endonuclease I recognizes and cleaves at Holliday junctions in an annealed PCR product (Vouillot et al., 2015). Samples from CAP256-VRC26.25 and hTPST1 expression, combined with or without the disruption of NbCysP6 and/or NbVPE-1a and NbVPE-1b (Supplementary Table 3), were assessed using the T7 endonuclease I assay. Plant genomic DNA was extracted from 200 mg of leaf discs using a DNeasy R plant mini kit (Qiagen, DE), as per the manufacturer's guidelines. Polymerase chain reaction (PCR) primers (Supplementary Table 4) were designed to amplify fragments of the NbCysP6, NbVPE-1a and NbVPE-1b genes, with 250 base pairs (bps) spanning each side of the CRISPR/Cas9 disruption site of the protease genes, which results in a ∼520 bp amplicon. The primer sets were used in separate PCR amplifications of the respective gene targets in the uninfiltrated samples and samples where both sgRNA:NbCysP6 and sgRNA:NbVPE1a/b were utilized. The PCR was set up for the amplification of regions within the NbCysP6, NbVPE-1a, and NbVPE-1b genes using Q5 R Hot Start High-Fidelity 2X master mix (New England Biolabs, USA), as per the manufacturer's instruction. PCR amplicons were purified using the GeneJET PCR purification kit (Thermo Fisher Scientific, USA), as per the manufacturer's guidelines. The T7 endonuclease I (New England Biolabs, USA) was used as outlined by the manufacturers. Post-hybridization step and T7 endonuclease I digestion, DNA fragments were purified using the GeneJET PCR purification kit, as per the manufacturer's guidelines. Digested DNA fragments were separated on a 5% tris(hydroxymethyl)aminomethane/boric acid/ethylenediaminetetraacetic acid (TBE) acrylamide gel, and visualized using GelRed (Biotium, USA).
RT-qPCR analysis of candidate cysteine protease mRNA fluctuations mRNA transcripts for three proteolytic candidate proteases (NbCysP6, NbVPE1a, and NbVPE1b) were assayed on a CFX96 real-time PCR detection system (Bio-Rad, USA) by Inqaba Biotechnical Industries (Pty) Ltd (ZA). Total RNA was extracted from 150 mg of leaf discs using the Zymo Quick RNA Plant kit (Zymo Research, USA), as per the manufacturer's instructions. RNA concentration was determined using the NanoDrop TM One Microvolume UV-Vis Spectrophotometer (Thermo Fisher Scientific, USA). cDNA synthesis was done using the LunaScript R RT SuperMix kit (New England Biolabs, USA), according to the manufacturer's instructions, except it was done in a total volume of 20 µl containing 500 ng total RNA. Quantitative RT-PCR was then performed in 96well-pplates with the Luna R Universal qPCR Master Mix (New England Biolabs, USA) using a dye-based qPCR assay. Each reaction contained 1 µl of cDNA template, 0.25 µm of forward and reverse primers (Supplementary Table 5), and 1X Luna R Universal qPCR Master mix. Reactions were analyzed using a standard two-step PCR program, as described in the Luna Universal qPCR Master Mix manual. Each cDNA was analyzed in triplicates. Amplification of different input templates was evaluated based on the quantification cycle (Cq) value. Fold changes in gene expression were determined using the Livak method (Livak and Schmittgen, 2001). For comparative purposes, relative gene expression levels of the candidate proteases were defined with the value of 1 set for control uninfiltrated plants and normalized against relevant reference genes, as previously described by Pillay et al. (2016).

Enzyme-linked immunosorbent assay
Abs produced in the presence and absence of genome editing vectors were quantified by sandwich enzyme-linked immunosorbent assay (ELISA) over a period of 11 dpi. A total of 96-well-plates were coated with either 5 µg/ml of goat antihuman lambda LC (bound and free) antibody (L1645, Sigma-Aldrich, USA) or goat anti-human IgG (Fc specific) antibody (I2136, Sigma-Aldrich, USA). All wells containing captureantibody were blocked for 16 h at 4 • C with 5% fat-free milk in PBS, pH 7.4. After washing the wells with PBS, pH 7.4 containing 0.1% (w/v) Tween R -20, a standard concentration range was prepared using purified IgG1 (I5029, Sigma-Aldrich, St. Louis, USA), and clarified leaf extracts containing Abs produced in the presence and absence of genome editing vectors were incubated with the respective capture antibody for 16 h at 4 • C. This was followed by wells being washed and bound IgGs being detected using either goat anti-Human IgG (Fc specific)-peroxidase antibody (A0170, Sigma-Aldrich, USA) or goat anti-Human Lambda light chain (bound and free)peroxidase antibody (A5175, Sigma-Aldrich, USA), which were incubated with sample wells for 2 h at 37 • C, followed by a wash. The peroxidase substrate, 3,3' ,5,5'-Tetramethylbenzidine (TMB) substrate solution (Sigma-Aldrich, USA) was used, with peroxidase-TMB reactions being stopped with 1 M H 2 SO 4 , and readings taken using a Hidex Sense Microplate Reader fluorometer (Hidex, EE) at 450 nm.

Host total soluble protein quantification
Quantification of host total soluble protein (TSP) was done, as previously described by Buyel et al. (2016). Briefly, TSP was extracted from 200 mg of leaf discs for each sample and quantified with the Bradford method (Bradford, 1976), using Quick Start TM Bradford 1x dye reagent (Biorad, USA). A 5-µl sample was added to 250 µl of Quick Start TM Bradford 1x dye reagent, of which 200 µl was transferred into a well of a 96well-flat-bottom plate (Nunc, DK). All samples were analyzed in triplicates. TSP concentration was determined against a bovine serum albumin (BSA) standard curve. The BSA standards were prepared by 1:1 serial dilution (10-0.625 mg/ml) of a 20-mg/ml BSA stock (New England Biolabs, USA) using PBS, and pH 7.4. TSP concentration was calculated against a BSA standard curve. A t-tests (p < 0.05 and <0.001) were performed to determine statistical differences between samples.

Ab purification
Briefly, infiltrated 7 dpi whole-leaf samples were homogenized using a Matstone 6-in-one juice extractor (Matstone, PH), protein were extracted using PBS, pH 7.4 containing cOmplete TM ULTRA tablets mini, and EASYpack protease inhibitor cocktail (Roche, CH). The supernatant was clarified by a series of centrifugation and filtration steps. The bNAbs were then purified by protein A affinity chromatography using a HiTrap R Protein A High-Performance column (Cytiva, USA), coupled to a Äkta Avant 150 (Cytiva, USA), as per the manufacturer's instructions.

Oligomeric and degradation analysis of the CAP -VRC Abs
The use of size-exclusion chromatography (SEC) separated molecules according to their size and, therefore, enabling the detection of fully assembled, monomeric, cleaved, and aggregated CAP256-VRC26.25 produced in N. benthamiana ( XTFT), with and without the disruption of NbCysP6 and/or Frontiers in Plant Science frontiersin.org . /fpls. .

NbVPE-1a
and NbVPE-1b. CAP256-VRC26.25 produced in HEK293 cells, was used as a positive control. SEC was performed on an Agilent 1,100 series high-pressure liquid chromatography (HPLC) (Agilent Technologies Inc., DE), which consisted of a binary pump, degasser, well-plate autosampler (WPALS), autosampler thermostat (ALS Thermo), column thermostat compartment, and a diode-array detector (DAD). The analytical column used was an AdvanceBio SEC 300 Å 2.7 um, 7.8 x 300 mm column (Agilent Technologies Inc., DE). All data analysis was performed using Chemstation chromatography data management software, revision B.04.02 (Agilent Technologies Inc., DE). The mobile phase contained 150 mM of Na 2 HPO 4 (Merck KGaA, DE), pH 7. The column was conditioned for 1 h with mobile phase before the injection of samples. CAP256-VRC26.25 bNAbs were eluted isocratically at a flow rate of.35 ml/min for 20 min, and the column effluent was monitored at 220 nm. All data were collected at 30 • C.

Secondary structure analysis
Far-UV Circular dichroism (CD) spectrum (260-180 nm) measurements for 4 µM of purified bNAb samples in 10 mM of Na 2 HPO 4 (Merck, DE), pH 7.4 were taken on an Applied Photophysics Chirascan CD spectrometer (UK) using a 1-mm path length at 20 • C. Averaged ellipticity values were converted to mean residue ellipticity (MRE) and corrected for the buffer blank baseline.

HIV-neutralization assay
HIV-1 neutralization potency for each bNAb sample was assessed using the TZM-bl neutralization assay, as described previously (Montefiori, 2005). Each Ab sample was diluted in a threefold dilution series using Dulbecco's Modified Eagle's medium (DMEM) (Sigma-Aldrich, USA), with 10% Fetal Bovine Serum (FBS) (Sigma-Aldrich, USA) (growth media). Diluted Ab was added in duplicate with 100 µl/ well of a 96-well-plate. About 50 µl/well of 200 TCID 50 of HIV-1 pseudovirus was, thereafter, introduced and incubated at 37 • C for 1 h. This was followed by the addition of TZM-bl cells at the concentration of 1 × 10 4 cells/100 µl of growth medium containing 37.5 µg/ml of DEAE dextran to each well, followed by culture at 37 • C for 48 h. HIV-1 neutralization was determined by measuring the luminescence emitted by the cells using the Infinite F500 plate reader (Tecan, AT). Titers were calculated as the concentration that caused 50% reduction (IC 50 ) of relative light unit (RLU) compared to the virus control (wells with no bNAb) after the subtraction of the background (wells without both the virus and the bNAb).

Results
Transient expression of in planta CAP -VRC . bNAbs The glycoengineered N. benthamiana ( XTFT) (Strasser et al., 2008(Strasser et al., , 2009 host, which lacks glycan residues with core β1,2-xylose and α1,3-fucose moieties, was used for the transient expression of CAP256-VRC26.25. CAP256-VRC26.25 was coexpressed with hTPST1 (Loos et al., 2015;Singh et al., 2020) to enable tyrosine sulfation of the complementarity-determining region (CDR) H3. Expression levels were qualitatively assessed in clarified leaf extracts from 1 to 11 days post-infiltration (dpi) by SDS-PAGE ( Figure 1A) and Western blot ( Figure 1B). Under reducing conditions, detectable production of ∼25 kDa band representing a LC by Western blot analysis was observed from day four, whereas production of ∼50 kDa band representing an HC was observed from day five ( Figure 1B). In addittion a ∼10 kDa and ∼40 kDa degradation species was observed from 5 to 10 dpi ( Figure 1B) despite the addition of protease inhibitor cocktail to reduce ex planta degradation.
Protease cleavage site analysis of the CAP -VRC . bNAbs Putative cleavage sites for the C1, C13, and C14 protease families in the CAP256-VRC26.25 amino acid sequence were investigated in silico (Supplementary Table 1). Amino acids were designated by the single-letter code and cleavage sites were designated by arrows (↓), with the numbers appearing after amino acids referring to the position of potential cleavage. Highly abundant cleavage sites are highlighted in yellow, second highly abundant cleavage sites are highlighted in green, and highly scarce cleavage sites are highlighted in blue. The CAP256-VRC26.25 Ab consists of four subunits, two heavy and two light chains. In silico analysis revealed that the HC has 319 putative papain-like cleavage sites, 73 putative cathepsin L cleavage sites, and two putative legumain cleavage sites, whereas the LC has 105 putative papain-like cleavage sites, 27 putative cathepsin L cleavage sites, and two putative legumain cleavage sites. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) confirmed that the intact IgG1 chains and degradation Frontiers in Plant Science frontiersin.org . /fpls. .

Transient reduction of NbCysP , NbVPE-a and NbVPE-b expression
Three strategies were taken: (i) targeting the NbCysP6 alone, (ii) both NbVPE-1a and NbVPE-1b, and (iii) a combination of all three genes (NbCysP6, NbVPE-1a and NbVPE-1b) to reduce the proteolytic degradation of CAP256-VRC26.25. A nuclease assay approach, as described by Nekrasov et al. (2013), and qRT-PCR were used to determine successful transient genome editing. The T7 endonuclease cleavage assay was utilized to determine the success of genome editing, with the T7 endonuclease I recognizing and cleaving Holliday junctions. Successful mutagenesis of the NbCysP6, NbVPE-1a and NbVPE-1b genes were carried out using Cas9, sgRNA:NbCysP6 and sgRNA:NbVPE1a/b in varying combinations. The success of the mutagenesis of  Figure 4B) depicts the mutagenesis of NbVPE-1a and NbVPE-1b. Despite the concentration of DNA being normalized across all T7 endonuclease I reactions, varying degrees of cleavage efficiency were observed between all reactions. In general, we observed an increase of editing over time from three dpi to seven dpi for NbCysP6, NbVPE-1a, and NbVPE-1b. The fragment size for both NbVPE1a and NbVPE1b was the same, making NbVPE1a and NbVPE1b fragment differentiation impossible.
The disruption of the expression levels of CysP6 by means of RNAi was previously assessed by RT-qPCR (Duwadi et al., 2015). The gene expression levels of the investigated three cysteine proteases under uninfiltrated conditions were set to one. The upregulation of NbCysP6, NbVPE-1a, and NbVPE-1b gene expression was measured at two key intervals, three and seven dpi, throughout the transient production period of CAP256-VRC26.25. Agroinfiltration of untransformed A. tumefaciens LBA4404 highly upregulated all three candidate protease genes, with NbCysP6 being upregulated to a lesser extent than NbVPE-1a and NbVPE-1b (Figure 2). The expression of all three candidate protease genes showed a decrease from three to seven dpi, with only NbCysP6 dropping below the gene expression levels under uninfiltrated conditions. Disruption of NbCysP6 and/or NbVPE-1a and NbVPE-1b without CAP256-VRC26.25 production resulted in a similar change in the pattern of gene expression (Figure 2). CAP256-VRC26.25 expression without genome editing vectors resulted in high upregulation of all three cysteine proteases, in particular, NbVPE-1b, with a fold change in gene expression from 13.05 to 23.20 from three dpi to seven dpi (Figure 2). Samples producing CAP256-VRC26.25, combined with the disruption of NbCysP6, as well as samples targeting NbVPE1a and NbVPE1b, showed an initial decrease in gene expression of all three cysteine proteases at three dpi. These samples showed an increase in the gene expression levels of NbVPE1b on seven dpi; with the samples producing CAP256-VRC26.25 combined with NbVPE1a and NbVPE1b disruption, showing an increase in NbCysP6 and NbVPE1a. Interestingly, when all three candidate protease genes were simultaneously disrupted with concomitant production of CAP256-VRC26.25, there was an overall decrease in gene expression levels of all three cysteine proteases to levels below that of the uninfiltrated control from three dpi to seven dpi ( Figure 2).

Transient reduction in proteolytic degradation of the CAP -VRC . bNAbs
The coexpression of the individual sgRNAs targeting either NbCysP6 or NbVPE-1a and NbVPE-1b showed a decrease in protease gene expression; however, qualitative analysis indicates that these strategies did not result in the reduction/elimination of the CAP256-VRC26.25 degradation species (Supplementary Figure 5). Interestingly, with simultaneous targeting of all three candidate protease genes and concomitant production of CAP256-VRC26.25, there was an overall decrease in gene expression levels of all three cysteine proteases to levels below that of the uninfiltrated control from three dpi to seven dpi ( Figure 2) and a complete disappearance of the ∼10 kDa and ∼40 kDa degradation species (Supplementary Figure 5B). Qualitative and quantitative analysis of CAP256-VRC26.25 with the simultaneous targeting of all three genes was conducted from one to eleven dpi using SDS-PAGE ( Figure 3A), western blot ( Figure 3B), and ELISA ( Figure 4). Under reducing conditions, detectable production of the LC by Western blot analysis was observed from day five, whereas detectable production of HC was observed from day six ( Figure 3B). Some proteolytic degradation products are present on day nine, but this tapers off on day 10 and 11 ( Figure 3B). Intact CAP256-VRC26.25 was detected through ELISA as early as two dpi, without the coexpression of genome editing vectors, however, with the coexpression of genome editing vectors, bNAbs was detected at three dpi. Without the coexpression of genome editing vectors, the concentration of produced bNAbs peaked at 6 dpi, with a concentration of 199.48 ± 5.75 mg/kg (Figure 4). With the coexpression of genome editing vectors, the . /fpls. . concentration of produced intact bNAbs peaked at 6 dpi with a concentration of 192.09 ± 9.04 mg/kg (Figure 4).
E ects of the disruption of candidate cysteine proteases on in planta cysteine protease activity The influence of the change in gene expression of the three candidate genes on global cathepsin L-like ( Figure 5) and legumain ( Figure 6) activity of endogenous proteases within N. benthamiana ( XTFT) plants was also investigated. Statistically significant differences between samples, with or without CAP256-VRC26.25 production, combined without or with the disruption of NbCysP6 and/or NbVPE-1a and NbVPE-1b, were determined by single-factor ANOVA with replication (p < 0.001) and t-tests (p < 0.05). Significant and highly significant differences were represented by * and * * , respectively, for cathepsin L-like (Supplementary Tables 6, 7) and legumain (Supplementary Tables 8, 9) protease activities. Samples without CAP256-VRC26.25 production, combined with the disruption of NbCysP6 and/or NbVPE-1a and NbVPE-1b, showed an increase in cathepsin L-like activity from three to seven dpi ( Figure 5). Untransformed A. tumefaciens LBA4404 had no influence on the legumain activity relative to the negative control, however, a huge spike in cathepsin L-like activity was observed from three to seven dpi. Disrupting NbVPE-1a and NbVPE-1b without CAP256-VRC26.25 production resultd in similar legumain activity to the uninfiltrated control. Interestingly, disrupting NbCysP6 resulted in decreased legumain activity, suggestive of the involvement of NbCysP6 in the regulation of proteases, which display legumain activity. CAP256-VRC26.25 production without editing vectors significantly increased cathepsin L-like activity relative to the uninfiltrated leaf sample, with an increase in cathepsin L-like activity from three to seven dpi.
CAP256-VRC26.25 production, combined with the disruption of NbCysP6 on three and seven dpi, had significantly less cathepsin L-like activity compared to the three dpi samples of CAP256-VRC26.25 produced without genome editing vectors. CAP256-VRC26.25 production, combined with the disruption of NbVPE1a and NbVPE1b on seven dpi, had significantly higher cathepsin L-like activity. Cathepsin L-like activity was significantly lower for samples expressing CAP256-VRC26.25, combined with targeted disruption of NbCysP6 on three and seven dpi, as well as samples producing CAP256-VRC26.25 combined with the disruption of all 3 candidate protease genes on three dpi relative to the seven-dpi sample of CAP256-VRC26.25 expression without genome editing vectors ( Figure 5). Supplementary Tables 6, 7 show statistical differences between samples.
. /fpls. .   The legumain protease activity ( Figure 6) of samples, with or without CAP256-VRC26.25 production, combined with or without the disruption of NbCysP6 and/or NbVPE-1a and NbVPE-1b, was also measured. Disrupting either NbCysP6 or all three candidate proteases without CAP256-VRC26.25 production showed a decrease in legumain activity from three to seven dpi, whereas disrupting NbVPE1a and NbVPE1b without CAP256-VRC26.25 production led to no Frontiers in Plant Science frontiersin.org . /fpls. . notable change in legumain activity between three and seven dpi. Legumain activities on three and seven dpi of the leaf samples producing CAP256-VRC26.25 without genome editing vectors remain relatively unchanged. CAP256-VRC26.25 production in combination with genome editing vectors led to an increase in legumain activity ( Figure 6), including samples in which reduction of NbVPE-1a and NbVPE-1b gene expression, was observed by RT-qPCR. There is an increase in legumain activity from three to seven dpi in CAP256-VRC26.25 samples. Supplementary Tables 8, 9 show statistical differences between samples.

E ects of the disruption of candidate proteases on the host total soluble protein
Changes in the levels of total soluble host proteins were used to determine the effects the disruption of the threecandidate protease had on host's cellular protein. The mean TSP concentrations in samples, with or without CAP256-VRC26.25 production, combined with or without the disruption of NbCysP6 and/or NbVPE-1a and NbVPE-1b, were normalized to the uninfiltrated negative control set at 3.4 ± 0.068 µg/µL (Figure 7). Statistically significant differences between samples with and without the production of CAP256-VRC26.25 were determined by single-factor ANOVA with replication (p < 0.001) and t-tests (p < 0.05). Significant and highly significant differences were denoted by * and * * , respectively, (Supplementary Tables 10, 11). Disrupting NbCysP6, both NbVPE-1a and NbVPE-1b, and a combination of NbCysP6, NbVPE-1a and NbVPE-1b without CAP256-VRC26.25 production, resulted in a decrease in TSP from three to seven dpi, with all the samples presenting lower TSP levels than the uninfiltrated baseline. Producing CAP256-VRC26.25 (three and seven dpi) without genome-editing vectors led to significantly lower TSP than the negative control base line (Figure 7). Producing CAP256-VRC26.25, combined with either the disruption of both NbVPE1a and NbVPE1a or all 3 candidate protease genes, resulted in a decrease in TSP, with the lowest observed when all 3 candidate protease genes were targeted. Interestingly, the production of CAP256-VRC26.25 combined .
/fpls. . with the disruption of NbCysP6 resulted in a significant increase in TSP three and seven dpi (Figure 7). This contrasts with NbCysP6 disruption without CAP256-VRC26.25 production, which showed a reduction in host cell protein.
Purification of CAP -VRC . produced in the presence and absence of genome editing vectors in N. benthamiana ( XTFT) Protein A based purification method is a common method for the purification of bNAbs, which results in >95% purity of the purified bNAb (Kim et al., 2018). The purification efficiency of CAP256-VRC26.25 produced without (Figures 8A,B) and with ( Figures 8C,D) genome editing vectors was analyzed post-Protein A affinity purification by SDS-PAGE and Western blot analysis. The ∼40 kDa proteolytic degradation product ( Figure 8A) is shown to comprise two discrete products that has closely migrated. Both are likely the same polypeptide sequence; however, one species is post-translationally modified through glycosylation, while the other is not. Dectectable cleavage products were observed in the eluents of CAP256-VRC26.25 produced without genome editing vectors (Figures 8A,B). No detectable cleavage products by western blot were seen in the clarified extract and flow through of the CAP256-VRC26.25 produced with the simultaneous disruption of NbCysP6 and NbVPE1a/b ( Figure 8D). Seven dpi was selected as a harvest point for the sample with disruption of NbCysP6 and NbVPE1a/b as no degradation product was observed ( Figure 3B). There is, however, detectable amounts of protease degradation products that were found in the eluates ( Figure 8D). This indicated that there was a low level of cleavage occurring and that the use of Protein A affinity purification has an enrichment effect on these proteolytic degradation products. Differences in expression efficiencies of recombinant protein has been previously observed in different leave types of N.
benthamiana (Goulet et al., 2019). The observed low levels of cleavage can be as a result of a variation in the transient genome editing efficiency in small population of the harvested leaf material.
Structural similarity between CAP -VRC . produced with and without protease disruption Size exclusion chromatography was used to characterize aggregates and levels of proteolytic degradation products in the antibody preparations. Comparative of the CAP256-VRC26.25 coexpressed with and without the simultaneous disruption of NbCysP6 and NbVPE1a/b in N. benthamiana ( XTFT) was done relative to the CAP256-VRC26.25 produced in HEK293. The overlaid views of SEC profiles for the Abs are shown in Figure 9. Each chromatogram was characterized by a peak at 13 min, representing single oligomeric state, (2LC, 2HC), of the intact CAP256-VRC26.25 antibody. There was no detectable cleaved fragment peak in all Ab samples,   including the CAP256-VRC26.25 sample, produced in the absence of genome editing vectors; a sample known to contain proteolytic degradation fragments. A shoulder on the peak of the CAP256-VRC26.25 sample produced in the presence of genome editing vectors was noted. The identity of this shoulder is unclear as this shoulder is not observed in the CAP256-VRC26.25 sample, which was produced in the absence of genome editing vectors. Secondary and tertiary structural analysis of the CAP256-VRC26.25 produced with and without the disruption of NbCysP6, NbVPE1a, and NbVPE1b in N. benthamiana ( XTFT) were assessed relative to the human embryonic kidney 293 (HEK293)-produced CAP256-VRC26.25 positive control. CAP256-VRC26.25 Abs' secondary structure was probed by circular dichroism (Figure 10). The CD spectra ( Figure 10) displayed a characteristic minimum at 217 nm, characteristic of protein structures with dominant β-sheet content, such as the native Ab structures (Doi and Jirgensons, 1970). CAP256-VRC26.25 bNAbs produced with and without the disruption of NbCysP6, NbVPE1a and NbVPE1b in N. benthamiana ( XTFT) had a similar secondary structural content to the HEK 293 produced Ab. Fluorescence spectroscopy was used to probe the folds of the CAP256-VRC26.25 Abs (Figures 11, 12). CAP256-VRC26.25 has 30 Trp and 54 Tyr residues, with all Trp and Tyr residues being distributed similarly throughout the Ab. The fluorescence spectral data of CAP256-VRC26.25 produced with and without the disruption of NbCysP6, NbVPE1a, and NbVPE1b in N. benthamiana ( XTFT) relative to the HEK293 produced CAP256-VRC26.25 positive control depicted a λ emmmax of 338 nm (Figures 11, 12).
Equivalent in vitro HIV-neutralization e cacy between CAP -VRC . produced with and without protease disruption CAP256-VRC26.25 is highly potent against HIV-1 subtypes A and C strains, with reduced potency against subtype B  strains Doria-Rose et al., 2015). CAP256-VRC26.25 produced in HEK293 and in N. benthamiana ( XTFT), with and without the disruption of NbCysP6, NbVPE-1a, and NbVPE-1b, was tested against a multi-subtype panel of 15 pseudoviruses using the TZM-bl neutralization assay (Table 1). The IC 50 values of HEK293-produced CAP256-VRC26.25 were similar to those obtained for N. benthamiana ( XTFT)-produced CAP256-VRC26.25, with and without the disruption of NbCysP6, NbVPE-1a, and NbVPE-1b, neutralizing seven of the eight subtype C viruses and zero of the three subtype B viruses (Table 1). A similar pattern was observed for the neutralization of subtype A viruses, where CAP256-VRC26.25 produced in HEK293 and N. benthamiana ( XTFT) with and without the disruption of NbCysP6, NbVPE-1a and NbVPE-1b neutralized three and four viruses, respectively. Importantly, the exceptional potency against some subtype C and A viruses was maintained for all CAP256-VRC26.25 samples, including that which was produced in N. benthamiana ( XTFT) without the disruption of all three proteases (Table 1).

Discussion
Nicotiana benthamiana has been proven to be an adequate host for the production of efficacious V1V2-targeting HIV antibodies (Loos et al., 2015;Rosenberg et al., 2015;Singh et al., 2020). IgG1 antibodies have a characteristic of a ∼ 50 kDa HC and a ∼ 25 kDa LC two-band pattern under reducing conditions. Additional ∼10 kDa and ∼40 kDa bands were present in the N. benthamiana ( XTFT)-produced CAP256-VRC26.25 sample, similar to that which was observed in the N. benthamiana ( XTFT)-produced CAP256-VRC26.08 and CAP256-VRC26.09 samples (Singh et al., 2020). Fragment analysis confirmed four potential cleavage sites within proximity on each HC subunit of the CAP256-VRC26.25 Abs. Two putative papain cleavage sites and two putative cathepsin-L cleavage sites were identified, with no putative legumain cleavage sites predicted. Proteases from the PLCP and LLCPs have been shown to be upregulated in N. benthamiana in response to Agroinfiltration (Pillay et al., 2016;Grosse-Holz et al., 2018). PLCP and LLCPs have previously been implicated in the degradation of anti-HIV antibodies (Niemer et al., 2014;Duwadi et al., 2015;Paireder et al., 2017). More recently, two subtilisin-like serine proteases have more recently been identified for their roles in the degradation of recombinant proteins in N. benthamiana (Puchol Tarazona et al., 2021). In all instances, proteolytic cleavage is usually targeted to sterically exposed segments of the antibodies (Mandal et al., 2014;Niemer et al., 2014;Paireder et al., 2017;Puchol Tarazona et al., 2020). Given the sensitivity of many anti-HIV Abs to the abovementioned LLCPs and PLCPs, it beckons the question as to whether CAP256-VRC26.25 is prone to degradation by these proteases. The vast majority of protease studies involve the knockout or downregulation of proteases through protease Frontiers in Plant Science frontiersin.org . /fpls. .   inhibitor coexpression, or pH control through the coexpression of proton channels (Girard et al., 2007;Pillay et al., 2012;Richau et al., 2012;Jutras et al., 2015). These methods have a broad inhibitory effect on all proteases and are not targeted to a specific protease of interest, and may result in unexpected changes in the physiology of the plant under defined growth conditions (Van der Vyver et al., 2003). The contributions of NbCysP6, NbVPE-1a, and NbVPE-1b to the degradation of CAP256-VRC26.25, were elucidated in this study by the incorporation of a targeted CRISPR/Cas9-mediated genome-editing approach. Protease gene expression was measured at two key intervals, three and seven dpi, throughout the transient production period of the CAP256-VRC26.25 bNAbs. Agroinfiltration of N. benthamiana ( XTFT) with untransformed A. tumefaciens LBA4404 had shown the upregulation of all three candidate cysteine proteases three dpi, with a decrease in gene expression of all three candidate cysteine proteases seven dpi. Apart from A. tumefaciens LBA4404, there are numerous commonly utilized A. tumefaciens strains, such as A. tumefaciens EHA105, A. tumefaciens GV3101, and A. tumefaciens GV3101::pMP90, which are used for the transient protein expression in N.
benthamiana. It is uncertain as to whether each strain will induce the same protease gene expression profile as was observed for the untransformed A. tumefaciens LBA4404 sample. In this study, agroinfiltration of CAP256-VRC26.25 highly upregulated all the three candidate cysteine proteases on three dpi; these were further upregulated in seven dpi, in contrast to the untransformed A. tumefaciens LBA4404 sample. This difference in the gene expression profile can be accounted for by the vector system being utilized for the expression of the CAP256-VRC26.25 protein. This system relies on the co-infection and co-replication of two non-competing viral vectors, such as tobacco mosaic virus (TMV) and potato mosaic virus (PVX), which would further enhance the upregulation of these proteases (Giritch et al., 2006). The infection of tobacco leaves by TMV has been found to induce, and, ultimately, lead to programmed cell death (Hatsugai et al., 2004). A popular alternative is the pEAQ-HT expression system, which is based on the cowpea mosaic virus (CPMV) (Sainsbury et al., 2009;Peyret and Lomonossoff, 2013). This system has also been shown to enhanced the upregulation of these proteases, NbCysP6, NbVPE-1a, and NbVPE-1b, as seen with the expression of the VP1 protein (Pillay et al., 2016). The full extent of the upregulation of these three cysteine proteases is unknown as sampling was done 24 h post-infiltration; as opposed to protease gene upregulation being monitored for the entire production period (Pillay et al., 2016). In our study, the coexpression of genome-editing vectors with CAP256-VRC26.25 showed successful reduction in gene expression of all three candidate cysteine proteases. The independent downregulation of NbCysP6 or the downregulation of both NbVPE-1a and NbVPE-1b does not prevent/reduce the proteolytic degradation of CAP256-VRC26.25, despite a decrease in protease gene expression. Interestingly, it was only through simultaneous targeting of NbCysP6, NbVPE-1a, and .
NbVPE-1b that we achieved a reduction in gene expression levels of all three cysteine proteases to lower the gene expression levels of these proteases under uninfiltrated conditions. This prominent reduction in protease gene expression also resulted in lower proteolytic degradation of CAP256-VRC26.25 bNAbs. ELISA-based quantification of the produced CAP256-VRC26.25 with and without NbCysP6 and NbVPE-1a/b disruption had suggested that the concentration of the oligomeric bNAb were essentially the same. The ELISA detection antibodies were the same as those used for detection in the western blots, which lacked the ability to differentiate between oligomeric bNAbs and degraded bNAbs fragments. This implied that the total concentration of the oligomeric bNAbs for the sample without NbCysP6 and NbVPE-1a/b disruption includes the quantification of degradation production.
Despite the absence of a putative legumain cleavage site in the CAP256-VRC26.25 Fc region, reduction of the proteolytic degradation products was only observed when all three candidate cysteine proteases were simultaneously downregulated, suggesting a synergistic function. Despite playing an important role in the regulation of physiological processes, the interaction between PLCPs and LLCPs and other proteases are hypothesized (Paulus and Van Der Hoorn, 2019). VPEs are autocatalytic proteases under acidic pH conditions and are known to activate PLCPs Kinoshita et al., 1999;Müntz and Shutov, 2002).
The NbCysP6 ortholog in Arabidopsis, RD21, recombinantly produced in insect cells did not self-mature itself, unless it was mixed with leaf extract (Yamada et al., 2001). RD21 is activated via a two-step mechanism involving other proteases that form part of a proteolytic cascade. It was suggested that VPEs plays a role in the activation of RD21 through this proteolytic cascade, however, formation of mature RD21 was shown to be possible in an Arabidopsis VPE quadrupole knockout mutant (Gu et al., 2012;Paulus and Van Der Hoorn, 2019). In contrast to RD21, NbCysP6 is capable of self-activation. Recombinant NbCysP6 was autocatalytic under acidic pH conditions, however, activation is not strictly limited to acidic pH conditions (Paireder et al., 2017). This suggests that the RD21-activation pathway may not be conserved between plant species (Paulus and Van Der Hoorn, 2019). The roles of PLCPs and LLCPs are synergistic as their upregulation is seen in many physiological processes, such as seed germination and hypersensitive response (Okamoto and Minamikawa, 1999;Zakharov et al., 2004). As seen with many proteases, our observations suggest that NbCysP6, NbVPE-1a, and NbVPE-1b function synergistically in independent linear signaling cascades involved in various physiological processes (van der Hoorn and Jones, 2004).
Proteolytic activity, cathepsin L-like and legumain activity, and protease gene expression were analyzed to determine if there was a correlation among NbCysP6, NbVPE-1a, and NbVPE-1b expressions and activities in the degradation of CAP256-VRC26.25. When infiltrated by untransformed A. tumefaciens LBA4404, there was no increase in legumain activity relative to the negative control. There was, however, a huge spike in cathepsin L-like activity from three to seven dpi. It is unlikely that NbCysP6 is solely responsible for this spike in activity as RT-qPCR shows a decrease in the gene expression of NbCysP6 over time. The MEROPS database currently lists a total of 516 putative peptidases and 97 non-peptidase homologs (Rawlings et al., 2018); amongst these, other listed cysteine proteases could have contributed to this overall increase in activity. Apart from NbCysP6, other proteases from the C1 family are upregulated in response to infiltration with untransformed A. tumefaciens LBA4404, of which 41 members form part of the C1 family (Pillay et al., 2016;Grosse-Holz et al., 2018). The downregulation of NbVPE-1a and NbVPE-1b in the control samples resulted in similar legumain activities to the negative control. Interestingly, the downregulation of NbCysP6 results in a decrease in legumain activity, suggesting the possibility of NbCysP6 involvement in the regulation of proteases, which display legumain activity. Infiltration of CAP256-VRC26.25 without genome-editing vectors did not influence legumain activity. There was an increase in cathepsin L-like activity from three to seven dpi, however, this was as high as those observed for plants infiltrated with A. tumefaciens LBA4404 alone. Samples where CAP256-VRC26.25 was coexpressed with genome-editing vectors showed an increase in legumain activity, including samples, in which reduction of NbVPE-1a and NbVPE-1b gene expression was observed by RT-qPCR. There is an increase in legumain activity from three to seven dpi in CAP256-VRC26.25 infiltrated samples, the upregulation of metacaspase-1 as previously seen with the agroinfiltration of VP1 could account for this observation (Pillay et al., 2016). Despite the knowledge of the protease suite in N. benthamiana being quite comprehensive, deducing a definitive correlation from the proteolytic activity data should be done with caution because proteases function in cascades, which may cause an upregulation of other proteases to compensate for the disruption of a particular protease (Grosse-Holz et al., 2018;Kourelis et al., 2019).
Changes in total soluble host protein levels were used to determine the effects of downregulating NbCysP6, NbVPE-1a, and NbVPE-1b on host cellular protein. Agroinfiltration triggers senescence, of which the first organelle to be disorganized is the chloroplast, which holds 75% of the leaf nitrogen, in the form of Rubisco. Degradation of the chloroplastic protein results in reduced photosynthetic capacity of the leaf (Hörtensteiner and Feller, 2002). The agroinfiltration of genome-editing vectors in the absence of the CAP256-VRC26.25 constructs had shown a decrease in host cell protein, with the extent of the decrease in host cell protein increasing seven dpi. Three samples deviated from this pattern, the agroinfiltration of untransformed A. tumefaciens LBA4404 on three dpi, and the downregulation of NbCysP6 with CAP256-VRC26.25 expression on three and Overall, the technique of agroinfiltration induces the hypersensitive response, which upregulates a suite of proteases in addition to NbCysP6, NbVPE-1a, and NbVPE-1b, which are involved PCD. This might explain the overall decrease in host cellular proteins in most samples prepared in this study. Rubisco can be hydrolyzed in intact chloroplasts or in chloroplast lysates, and it has been suggested the initial stages of senescence degradation of the plastidial proteins occurs inside the organelle (Feller et al., 2008;Kato and Sakamoto, 2010). The chloroplastic environment is rich in serine and metalloproteases, suggesting these to be the prime candidates for initial Rubisco degradation (van der Hoorn, 2008). An upregulation of metalloproteases from two families (M38 and M67) were previously seen in response to agroinfiltration, thereby degrading host cell protein (Pillay et al., 2016). The involvement of cysteine proteases in Rubisco degradation was investigated with the use of E64, a commonly used cysteine protease inhibitor. Rubisco degradation was reduced in the presence of this protease inhibitor (Carrión et al., 2013).
Arabidopsis has three different genes, which encode the cysteine protease Cathepsin B, which are upregulated in developmental leaf senescence. Mutation of all three Cathepsin B genes delayed senescence, when monitored by observed loss of chlorophyll. Cathepsin B may act upstream of other proteases involved in senescence, of which the expression is reduced in the triple mutant Cathepsin B plant. This places Cathepsins in a pathway regulating senescence; however, the targets of Cathepsin B are unknown (Pružinská et al., 2017).
Characterization and quantification of the cleaved states and aggregates is of particular importance for biopharmaceuticals as these species could potentially affect the efficacy of the bNAbs. A single oligomeric state was detected, with no observed peak, which could be indicative of the presence of proteolytic degradation products. This is suggestive, that the proteolytic degradation fragment is still associated with the Ab structure. Apart from the disulfide bridging between the LCs and HCs, strong non-covalent associations exist between the two paired HCs in the CH3 regions (Dall'Acqua et al., 1998;van der Neut Kolfschoten et al., 2007). These strong non-covalent forces are responsible for the non-dissociating cleaved fragment in this previous study (Brezski et al., 2009). Non-denaturing SEC studies on a intact IgG1s and a single-cleaved IgG1s (IgG which is cleaved at a particular site on one HC chain) were indistinguishable (Brezski et al., 2009). This is consistent with observations made under non-reducing conditions of the N. benthamiana ( XTFT)-produced CAP256-VRC26.08 and CAP256-VRC26.09, in which a single intact band was observed (Singh et al., 2020). However, it is unclear in this current study as to whether CAP256-VRC26.25 is cleaved on a single HC, or both, contributing to this observation or whether this is due to the low abundance of the cleaved fragment. The structural similarity between the CAP256-VRC26.25 bNAbs produced in HEK293, and in N. benthamiana ( XTFT) with and without the simultaneous disruption of NbCysP6 and NbVPE-1a/b. Despite the presence of protein degradation products in N. benthamiana ( XTFT) CAP256-VRC26.25 samples produced without protease disruption, there were no detectable secondary, tertiary, and quaternary structural differences when compared to the HEK293 produced bNAbs and with simultaneous disruption of NbCysP6 and NbVPE1a/b.
As seen with CAP256-VRC26.08 and CAP256-VRC26.09, the CAP256-VRC26.25 bNAbs produced in N. benthamiana ( XTFT) without protease disruption, had the same in vitro neutralization potency as the HEK293-produced bNAbs. CAP256-VRC26.25 produced in HEK293 and N. benthamiana ( XTFT) with and without the simultaneous disruption of NbCysP6 and NbVPE-1a/b all displayed high potency against HIV-1 subtype A and C strains, with reduced potency against subtype B strains, consistent with previous work Doria-Rose et al., 2015). The Fab region of the CAP256 antibodies recognizes antigens and activates (or suppress) immune cells through interactions with and between the Fc region and the cell bearing Fc receptors (FcR). The in vitro TZM-bl assay used in this study assays the neutralization of the Fab region of the CAP256 antibodies to the V1V2 region of the HIV-1 gp120 envelope glycoprotein. The effects of the cleaved Fc region of the CAP256-VRC26.25 bNAbs Frontiers in Plant Science frontiersin.org . /fpls. . produced in N. benthamiana ( XTFT) without genome-editing vectors were not detectable by the TZM-bl assay. Binding of the Fc region to the FcRs activates antibody-dependent cellular cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP), or complement-dependent cytotoxicity (CDC). The serum half-life of IgGs is mediated by the pH-dependent binding to the neonatal Fc receptor (FcRn) (Datta-Mannan et al., 2007). Residues, which are responsible for FcRn binding are located between the CH2 and CH3 regions of the HC (Kim et al., 1999;Roopenian and Akilesh, 2007). One of these residues is N297, the glycosylation which influences both the structure and function of the IgG. Potential cleavage sites were identified between amino acid positions 397 and 401, indicative of the likely retention of glycosylation even in proteolytically degraded bNAbs. From this study it is uncertain as to whether a single or both CAP256-VRC26.25 HCs are cleaved. It was previously observed that IgG1s with single cleavage in the hinge region maintain affinity to the FcRn; however, they showed a loss of affinity to the Fcγ family of receptors (Brezski et al., 2009;Brezski and Jordan, 2010). The effects of truncated bNAbs on in vivo potency and humoral immunity are yet to be determined.

Conclusion
From this study and other studies, endogenous proteolytic degradation is a bottleneck in the production of antibodies in planta. This study has shown that this bottleneck can be reduced/avoided through the transient's downregulation of proteases like NbCysP6, NbVPE-1a, and NbVPE-1b, thus, making it possible to produce intact highly potent anti-HIV and other Abs. Though the functional activity of these bNAbs has not been demonstrated in vivo, it is still crucial to consider proper downstream processing to improve the overall quality of the produced protein to avoid any aberrant immune responses, which may arise from the presence of these truncated antibodies and proteolytic fragments. The generation of a stable mutant with the NbCysP6, NbVPE-1a, and NbVPE-1b gene knocked out will eliminate the low levels of cleavage observed through Protein A enrichment during the purification process. For in planta-produced Abs to gain further wide-spread acceptance and adoption as an industry alternative to produce biopharmaceuticals, the findings of this study may be key considerations for the biopharma sector.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions
RC, AS, PP, and TT conceived the study. RC, PP, TT, and JV supervised the study. AS and PP conducted all antibody expressions and genome-editing experimentations. KA provided technical input on neutralization assays, with neutralization assays being conducted by KM. LM and HS designed and optimized hTPST1 construct. HS provided the N. benthamiana ( XTFT) seeds used in this study. AS, PP, PN, and KA performed the experiments. AS, PP, PN, KA, and TT analyzed the data. AS, PP, and TT designed the experiments and wrote the manuscript with input from LyM, JB, LM, and HS. All authors contributed to the article and approved the submitted version.