Exposure of Agrobacterium tumefaciens to agroinfiltration medium demonstrates cellular remodeling and may promote enhanced adaptability for molecular pharming

Agroinfiltration is used to treat plants with modified strains of Agrobacterium for the purpose of transient in planta expression of genes transferred from the bacterium. These genes encode valuable recombinant proteins for therapeutic or industrial applications. Treatment of large quantities of plants for industrial-scale protein production exposes bacteria (harboring genes of interest) to agroinfiltration medium that is devoid of nutrients and carbon sources for prolonged periods of time (possibly upwards of 24 h). Such conditions may negatively influence bacterial viability, infectivity of plant cells, and target protein production. Here, we explored the role of timing in bacterial culture preparation for agroinfiltration using mass spectrometry-based proteomics to define changes in cellular processes. We observed distinct profiles associated with bacterial treatment conditions and exposure timing, including significant changes in proteins involved in pathogenesis, motility, and nutrient acquisition systems as the bacteria adapt to the new environment. These data suggest a progression towards increased cellular remodelling over time. In addition, we described changes in growth- and environment-specific processes over time, underscoring the interconnectivity of pathogenesis and chemotaxis-associated proteins with transport and metabolism. Overall, our results have important implications for the production of transiently-expressed target protein products as prolonged exposure to agroinfiltration medium suggests remodeling of the bacterial proteins towards enhanced infection of plant cells. In this study, we use state-of-the-art mass spectrometry to quantify changes in bacterial protein abundance over a time-course of exposure to agroinfiltration medium. Based on our recent work evaluating bacterial growth conditions (e.g., shake-flask vs. bioreactor) for optimal biomass production, we extend these findings to define processes of cellular remodeling specific to bacterium grown under each condition prior to agroinfiltration medium exposure and observe temporal proteome changes in a nutrient-limited environment (Prudhomme et al. In Revision). Our results identify distinct proteome profiles associated with bacterial growth conditions and exposure timing, as well as a progression towards increased cellular remodelling over time. In addition, we describe unique functional characteristics influencing pathogenesis and response to stimulus, as well as changes in bacterial nutrient acquisition and adaptability. Moreover, we validate our findings with in-depth profiling of specific proteins over time and demonstrate a >80% overlap between applications of our proteomics platform for A. tumefaciens profiling. Overall, our results suggest that prolonged exposure of A. tumefaciens to agroinfiltration medium may increase pathogenesis and motility, leading to enhanced infectivity of plant cells and production of the target drug.

D r a f t vacuum chamber to draw air out of interstitial spaces of submerged leaves that is replaced with Agrobacterium-containing medium when the vacuum is released (Garabagi et al. 2012). Although this process requires investment of machinery, including vacuum pumps, vacuum chambers, and large bacterial culture volumes, this approach provides scalability potential, which is instrumental in the success of plant-based production systems (Mclean 2017).
Recent studies explored the impact A. tumefaciens agroinfiltration on gene expression, plant defense response, and target protein production to optimize the process and improve transformation efficiency (Buyel 2015). For example, transcriptomic and proteomic analyses report dynamic reprogramming of the proteolytic machinery of N. benthamiana upon agroinfiltration (Grosse-Holz et al. 2018). Another study explored the role of chemical additives, heat shock, and co-expression of genes known to suppress plant stress and gene silencing or stimulate cell cycle progression to increase agroinfiltration-based transient gene expression (Norkunas et al. 2018). These studies highlight the downstream effects and importance of optimizing gene expression parameters for plant-based protein production. However, few studies have investigated the protein-level impacts of sample preparation and specifically, exposure and D r a f t limitation (Geddes et al. 2015;Smits and Vermeulen 2016;Sukumaran et al. 2019;Muselius et al. 2020). In this study, we use state-of-the-art mass spectrometry to quantify changes in bacterial protein abundance over a time-course of exposure to agroinfiltration medium.
Based on our recent work evaluating bacterial growth conditions (e.g., shake-flask vs. bioreactor) for optimal biomass production, we extend these findings to define processes of cellular remodeling specific to bacterium grown under each condition prior to agroinfiltration medium exposure and observe temporal proteome changes in a nutrient-limited environment (Prudhomme et al. In Revision). Our results identify distinct proteome profiles associated with bacterial growth conditions and exposure timing, as well as a progression towards increased cellular remodelling over time. In addition, we describe unique functional characteristics influencing pathogenesis and response to stimulus, as well as changes in bacterial nutrient acquisition and adaptability.
Moreover, we validate our findings with in-depth profiling of specific proteins over time and demonstrate a >80% overlap between applications of our proteomics platform for A. tumefaciens profiling. Overall, our results suggest that prolonged exposure of A. tumefaciens to agroinfiltration medium may increase pathogenesis and motility, leading to enhanced infectivity of plant cells and production of the target drug.
D r a f t

Strains
A modified version of the A. tumefaciens strain EHA105 (transformed with the T-DNA vector pPFC0058 for expression of the monoclonal antibody Trastuzumab) was used for the experiments (Hood et al. 1993;Mclean 2017). The bacteria were maintained on Lysogeny Broth (LB) agar plates at 28°C supplemented with kanamycin sulfate solution (50 µg/ml) and rifampicin solution (25 µg/ml).
D r a f t sulphate, pH 5.6) (Garabagi et al. 2012). Samples were subjected to static incubation at room temperature for 0, 2, and 24 h post immersion (hpi) in the dark.

Sample collection
Prior to collection, any settled cell material was resuspended by gentle swirling and two ml of culture was collected by centrifugation at 1,500 x g for 10 min. Supernatants were collected for further processing. Cell pellets were washed twice with cold phosphate buffered saline (PBS) and collected for protein extraction. Samples were stored on ice before processing.

Sample preparation for proteomic analysis
Protein extractions were performed as previously described, with modifications (Prudhomme et al. n.d.;. Briefly, bacterial cell pellets were resuspended in 100 mM Tris-HCl (pH 8.5) containing a protease inhibitor cocktail tablet. Samples were lysed by probe sonication (ThermoFisher Scientific) on ice bath for 3 cycles (30% power, 30 s on/30 s off). Two percent sodium dodecyl sulphate (SDS) and 10 mM dithiothreitol (DTT) was added, followed by incubation at 95°C for 10 min with shaking at 800 rpm. Samples were cooled and 55 mM iodoacetamide (IAA) was added followed by room temperature incubation for 20 min in the dark. Next, 100% ice cold acetone (final concentration of 80%) was added prior to storage at -20°C overnight. Samples were collected by centrifugation at 10,000 x g at 4°C for 10 min, washed twice with 80% acetone, and air dried. Pellets were resolubilized in 8M urea/40 mM HEPES and a bovine serum albumin (BSA) tryptophan assay determined protein concentrations (Wisniewski and Gaugaz 2015). Samples were diluted in 50 mM ammonium bicarbonate and digested overnight with a mixture of LysC and trypsin proteases (Promega, protein:enzyme ratio, 50:1). Digestion was stopped with 10% v/v trifluoroacetic acid (TFA) and 50 µg of the acidified peptides were loaded onto STop And Go Extraction (STAGE) tips (consisting of three layers of D r a f t C18) to desalt and purify according to the standard protocol (Rappsilber et al. 2007). Samples were stored as dried peptides at -20°C until measured on the mass spectrometer.
For secretome analysis, the culture supernatant was filtered through 0.22 µM syringe filters. For each sample, 500 µl of filtered supernatant was treated with DTT, IAA, followed by digestion using LysC and Trypsin. Digested peptides were desalted and purified as described above.

Mass spectrometry
Samples were eluted from STAGE-tips with 50 µl buffer B (80% acetonitrile (ACN) and 0.5% acetic acid), dried, and resuspended in 12 µl buffer A (0.1% TFA). Six µl of each sample was analyzed by nanoflow liquid chromatography on an Ultimate 3000 LC system (ThermoFisher Scientific) online coupled to a Fusion Lumos Tribrid mass spectrometer (ThermoFisher Scientific) through a nanoelectrospray flex-ion source (ThermoFisher Scientific). Samples were loaded onto a 5 mm µ-precolumn (ThermoFisher Scientific) with 300 µm inner diameter filled with 5 µm C18 PepMap100 beads. Peptides were separated on a 15 cm column with 75 µm inner diameter with 2 µm reverse-phase silica beads and directly electrosprayed into the mass spectrometer using a linear elution gradient from 4% to 30% ACN in 0.1% formic acid over 45 min at a constant flow of 300 nl/min. The linear gradient was followed by a washout with up to 95% ACN to clean the column followed by an equilibration stage to prepare the column for the next run. The Fusion Lumos was operated in data-dependent mode, switching automatically between one full scan and subsequent MS/MS scans of the most abundant peaks with a cycle time of 3 s. Full scan MS1s were acquired in the Orbitrap analyzer with a resolution of 120,000, scan range of 400-1600 m/z. The maximum injection time was set to 50 ms with an automatic gain control target of 4e5. The fragment ion scan was done in the Orbitrap using a Quadrupole isolation window of 1.6 m/z and HCD fragmentation D r a f t energy of 30 eV. Orbitrap resolution was set to 30,000 with a maximum ion injection time of 50 ms and an automatic gain control target set to 5e4.

Data analysis
For proteome data analysis .Raw files were analyzed using MaxQuant software (version Further analysis of the MaxQuant-processed data ('proteingroups.txt' file) was performed using Perseus (version 1.6.2.2) (Tyanova et al. 2016). Hits to the reverse database, contaminants, and proteins only identified with modified peptides were eliminated. LFQ intensities were converted to a log scale (log 2 ), and only those proteins present in triplicate within at least one sample set were used for further statistical processing (valid-value filter of three in four replicates D r a f t in at least one group). Missing values were imputed from a normal distribution (downshift of 1.8 standard deviations and a width of 0.3 standard deviations). A Student's t-test identified proteins with significant changes in abundance (p-value ≤ 0.05) with multiple hypothesis testing correction using the Benjamini-Hochberg FDR cut-off at 0.05, s0 = 1. A principal component analysis (PCA) was performed. The PCA plot was performed on the 'category enrichment in components' for five components with a Benjamini-Hochberg FDR = 0.05 and relative enrichment by Protein IDs.
Replicate reproducibility was derived from a Pearson correlation with hierarchical clustering by Euclidean distance. For 1D annotation enrichment, Student's t-test (permutation-based FDR = 0.05; s0 = 1) was performed followed by 1D annotation enrichment function in Perseus (Cox and Mann 2012). This analysis generates a numerical 'score' value, which represents the direction in which the protein LFQ intensities within a given category tend to deviate from the overall distribution of all proteins. Visualization of enrichment by Gene Ontology was performed within the RStudio platform (http://www.R-project.org/) (R Foundation for Statistical Computing. 2018).
For network visualization the STRING database was used (https://string-db.org) (Szklarczyk et al.

2019).
D r a f t

Results.
To assess the impact of agroinfiltration medium and the effect of storage of A. tumefaciens cultures prior to the agroinfiltration process, we performed a quantitative proteomic analysis of the cell pellet (total proteome) and secretome (extracellular environment) (Fig. 1). With this information, we analyzed changes to A. tumefaciens at the protein-level in response to transformation preparation, which may require prolonged periods of bacterial storage in the agroinfiltration medium when large numbers of plants are prepared for industrial treatment.
Moreover, we aimed to suggest optimal processing strategies to improve target drug production as influenced by the agroinfiltration process. Furthermore, we explored the influence of sample growth conditions (e.g., bioreactor growth vs. shake-flasks) on bacterial response to the agroinfiltration process to highlight potential protective measures demonstrated by the bacteria.

Storage in agroinfiltration medium alters the cellular proteome of A. tumefaciens
In total, we identified 2,992 proteins (>55% of open reading frames) in the cellular proteome and secretome samples (before valid value filtering) and pursued further analysis of 2,640 proteins ( Fig. 2A). Proteins identified in the supernatant did not meet the valid value filtering criteria, therefore we focused our analysis on the cellular proteome results. Biological replicate reproducibility was > 97% for all samples (

Duration of exposure to agroinfiltration medium influences bacterial cell remodeling
D r a f t Given the distinction between storage time and bacterial growth conditions, we identified proteins with significant changes in abundance under the different parameters. For shake-flasks, we observed the smallest protein-level changes at early immersion (0 h) compared to the later time points. Specifically, 10 proteins were significantly increased in abundance and 10 proteins showed a significant decrease in abundance at 2 h ( Fig. 3A; Supp. Table 2). Of these, we observed an increase in pathogenesis-related protein Atu3295, methyl-accepting chemotaxis protein McpC, and an outer membrane heme receptor Atu2287. Conversely, we observed a decrease in abundance of ABC transport substrate binding proteins (Atu4123, Atu3253) and a transcriptional regulator We also investigated the presence of common and unique proteins produced at each of the time points and report time-specific responses at each stage of incubation, and to a lesser extent, common responses across the comparisons; however, no proteins were detected as significantly D r a f t different at all three stages (Fig. 3D). Three proteins were significantly different at 0 and 24 h but did not demonstrate a change in abundance at 2h. Although this may be due to biological variability or a fluctuation in cell protein responses over the time course of incubation, taken together, our quantitative proteomic profiling of changes to the cellular proteome of A. tumefaciens grown in shake-flask and exposed to agroinfiltration medium at early (0 h), intermediate (2 h), and late (24 h) stages indicates a progression of cellular remodeling over time.
Next, we explored temporal differences in exposure to agroinfiltration medium of bacterial cultures grown in a bioreactor. Again, we observed the lowest number of altered protein profiles at 0 h with seven proteins displaying significant changes in abundance: five proteins increased, and two proteins decreased ( Fig. 3E; Supp. Table 5). All proteins with increased abundance were uncharacterized (Atu3295, Atu8200, Atu1155, Atu2160, Atu5516) and proteins with decreased abundance included a glucose dehydrogenase (Gcd) and uncharacterized protein Atu3752. Very similar to the observation of common and unique proteins produced at each of the time points during shake-flask growth, time-specific responses were present at each stage of D r a f t immersion, and to a lesser extent, we detected common responses between 0 and 2 h, as well as 2 and 24 h (Fig. 3H). Again, we did not detect any significantly different proteins at all three stages of incubation. With this information, we can begin to define the impact of agroinfiltration medium on bacterial cultures over the time course of bacterial processing and we can distinguish differences in response to exposure, dependent on bacterial growth conditions.

Unique functional alternations upon exposure to agroinfiltration medium
Given our identification of significant changes in protein abundance, suggesting cellular remodeling in the presence of agroinfiltration medium during 24 h exposure, we hypothesized that differences in functional groups of proteins may change over time. To explore this idea, we performed a 1D annotation enrichment analysis based on Gene Ontology Biological Processes (GOBP) (Ashburner et al. 2000;Cox and Mann 2012). For shake-flask samples, we saw a significant positive enrichment (based on GOBP) of proteins associated with multi-organism process (e.g., pathogenesis), response to stimulus (e.g., chemotaxis, locomotion, stress responses), and carbohydrate metabolic process (e.g., β-xylanase) as immersion time in agroinfiltration medium increased (Fig. 4A). Conversely, we saw a negative enrichment of proteins associated with metabolism, biosynthetic processes, and translation over time. To define networks of proteins altered over time in agroinfiltration medium, we used the STRING database to analyse proteins showing significant increases in abundance at 24 h from shake-flask samples (Fig. 4B). This analysis highlights four clusters of proteins: i) flagellin, motility, and chemotaxis, ii) metabolism (enzymatic activity), iii) siderophore biosynthesis, and iv) heme-and iron-association.
We performed a 1D annotation analysis for bioreactor samples by GOBP and observed a similar positive enrichment of response to stimulus proteins over time and a negative enrichment of proteins associated with metabolic and biosynthetic processes, translation, aminoacylation, and D r a f t amino acid activation (Fig. 4C). An investigation into protein networks altered during incubation with agroinfiltration medium from bioreactor samples highlighted increased abundance at 24 h of proteins associated with: i) flagellin, motility, and chemotaxis, ii) metabolism (enzymatic activity), and iii) transport (Fig. 4D). Taken together, our functional analysis of the impact of agroinfiltration medium on A. tumefaciens demonstrated growth-and environment-specific processes over time and underscores the interconnectivity of pathogenesis and chemotaxis-associated proteins with transport and metabolism. Furthermore, we observed a consistent reduction in metabolic and biosynthetic processes, as well as ribosome activity and translation, which suggest remodelling of cellular processes over processing of the bacterial culture.

Changes to pathogenesis-and motility-associated proteins highlights bacterial adaptability
Given our observations of diverse biological categories altered during the time course of sample preparation for agroinfiltration, we hypothesized that specific proteins influence these cellular processes and confer a balance between bacterial protection and activation over time. First, based on a positive enrichment of multi-organism process from shake-flask samples, we profiled LFQ differences in all pathogenesis-associated proteins over time. Although no significant difference in abundance was observed among the group of proteins, we detected a significant increase in production of pathogenesis-associated protein Atu3295 by 24 h in shake-flask samples ( Fig. 5A; Supp. Fig. 1A). Moreover, given our observation of the importance of response to stimulus, chemotaxis, locomotion, and transport at 24 h, we explored changes in protein abundance for all flagellin-associated proteins. We noted a significant increase in flagellar protein abundance over time; Flagellin A (FlaA) showed the greatest increase in abundance ( Fig. 5B; Supp. Fig. 1B).
We observed similar protein response profiles for bioreactor samples over time, despite a reduced D r a f t emphasis on pathogenesis but a similar response on bacterial motility (Fig. 5C, D; Supp. Fig 1C,   D).
Finally, we compared our proteomic profiling from the agroinfiltration experiments performed in this study to our recent investigation into the protein-level impact of bacterial growth conditions (e.g., shake-flask vs bioreactor) (Prudhomme et al. n.d.). Of the 2,743 proteins identified between the studies, 80% were common, supporting consistency in our protein extraction, sample preparation, and mass spectrometry methods (Fig. 5E). Moreover, our approaches enabled detection of protein-level changes distinct to each study. For example, 104 proteins were specific to our investigation of shake-flask vs. bioreactor growth conditions and 456 proteins are unique to this agroinfiltration study. Moreover, this comparison allowed us to tease apart growth-vs. environment-specific bacterial responses. For example, shake-flask growth impacted metabolic and biosynthetic pathways, whereas during analysis of temporal agroinfiltration preparation, we observed broad-spectrum changes to pathogenesis, motility, and nutrient acquisition. For bioreactor samples, growth condition analysis emphasized differences in transport and locomotion, which also drive the observations associated with agroinfiltration. Taken together, our quantitative proteomic profiling provided a reliable analysis of cellular remodelling and changes in biological processes associated with multiple steps in the agroinfiltration-mediated transient expression process and highlighted distinct responses that may influence bacterial survival and adaptability.
D r a f t

Discussion.
By profiling the total proteome of A. tumefaciens in response to prolonged exposure to agroinfiltration medium as a preparatory step for plant transformation, we report time-dependent cellular responses and dynamic remodelling of biological processes. Specifically, we observed significant changes in proteins involved in pathogenesis, motility, and nutrient acquisition systems as the bacteria adapt to the new environment. In general, these responses are specific to the exposure time and were not defined by bacterial growth conditions (e.g., shake-flask vs. bioreactor). As expected, shorter periods of exposure to agroinfiltration medium cause fewer protein-level changes to the bacterial cultures. However, analysis of exposure periods of 2 h, which is an accurate representation of the time lag between medium preparation and plant treatment, shows remodeling of cellular processes have begun and, by 24 h, we observed dynamic remodeling. Overall, these results suggested enhanced bacterial adaptability to environmental conditions during the agroinfiltration process, which may promote efficient bacterial invasion upon transformation.
Previous proteome profiling of A. tumefaciens response to plant tissue (e.g., tomato roots) showed induction of 30 proteins, including known virulence factors, transporters, metal-associated enzymes, as well as biosynthetic and metabolic enzymes, supporting their production during the invasion process (Rosen et al. 2003). We observed similar patterns of protein induction over the time course of exposure to agroinfiltration medium, suggesting the incubation period may serve as a 'primer' to activate the bacteria in preparation for host cell invasion. For example, following shake-flask growth, we observed an increase in abundance of proteins associated with pathogenesis, including uncharacterized protein Atu3295 (BA14-like protein associated with immune reactivity) during the time course of exposure (Chirhart-Gilleland et al. 1998). Previous D r a f t reports demonstrated the induction of virulence-associated genes to improve transformation frequencies. Specifically, the inclusion of acetosyringone, a plant-secreted phenolic that induces Agrobacterium virulence gene expression through induction of a small heat shock protein involved in VirB accumulation and promoting virB/D4-mediated DNA transfer, is often used for cocultivation in agroinfiltration media (Stachel et al. 1986;Hiei et al. 1994;Lai et al. 2006;Wydro et al. 2006;Tsai et al. 2009;Norkunas et al. 2018). Not surprisingly, given the absence of the host, we did not observe VirB in our dataset, which may indicate low protein production with the tested parameters. Taken together, these findings suggest the production of pathogenesis-associated proteins in response to additives or prolonged exposure to agroinfiltration medium may 'prime' the bacteria for improved host invasion efficiency.
The physical attachment of bacteria to the host is critical for infection and horizontal gene transfer to the plant (Thompson et al. 2018). There are several structures that function for surface attachment by A. tumefaciens, including flagella, several different forms of pili, and multiple complex polysaccharides (Gotz et al. 1982). The general function of flagella is to propel bacteria through an environment. Flagellar propulsion can enable bacteria to move towards conditions that are favourable, such as high nutrients, and avoid conditions that inhibit growth or damage cells.
However, in some cases, flagella play an important role in surface attachment during infection (Merritt et al. 2007). Here, we observed an increase in abundance of proteins associated with motility and chemotaxis, specifically, flagellar protein production (FlaA, the primary flagellin, required for motility) during the course of exposure to agroinfiltration medium (Deakin et al. 1997). Based on the role of flagellin in motility, the increased production may be associated with searching for nutrients in a nutrient-limited environment or it may support preparation of the bacteria for host invasion. A role in nutrient sensing is corroborated by our observation of activated D r a f t ABC transporters, specifically, those associated with metal-ion uptake (e.g., iron, heme, zinc) which are commonly produced in response to nutrient-limited environments when encountered by bacteria (Heindl et al. 2016;Tanaka et al. 2018). Conversely, a connection between FlaA and bacterial virulence associated with surface attachment has been explored in the interaction between A. tumefaciens and host, demonstrating an influence of light-induction on gene expression (Oberpichler et al. 2008). Overall, our conserved observation of changes in motility during prolonged exposure to agroinfiltration medium, regardless of bacterial growth conditions (shakeflask vs. bioreactor), supports cell remodeling for purposes of nutrient acquisition and perhaps enhanced surface attachment efficiency upon transformation.
Despite our observations of increased production of bacterial proteins connected with pathogenesis, motility, and transportation during the course of incubation in agroinfiltration medium, suggesting adaptation of the bacteria to the surrounding environment and enhanced preparedness for transformation, we also observed a reduction in metabolism and biosynthetic processes, as well as mRNA translation over time. The absence of nutrients (starvation) can induce metabolic stress, DNA damage, and polymerase pathway activation to induce cell death as an adaptation and survival mechanism (Rodríguez-Vargas and Oliver 2016). Moreover, sufficient nutrient supply is essential for optimal mRNA translation and nutrient deprivation can inhibit global protein synthesis, supporting our observation of reduced translation in agroinfiltration medium (Gameiro and Struhl 2018). Given the reduced cellular response to a nutrient-limited environment (i.e., agroinfiltration medium), a balance between bacterial survivability (e.g., increased production of acquisition-associated transporters) and adaptability (e.g., production of pathogenesis and motility proteins) is needed to promote efficient transformation. Furthermore, exploring the relationship between prolonged periods of exposure to agroinfiltration medium and D r a f t cell viability will highlight important relationships between bacterial priming, survivability, and infectivity.
Notably, proteins detected in the supernatant were low in numbers and inconsistently identified (i.e., present in less than three of four replicates), resulting in their removal from further analysis. An important factor to consider when profiling the supernatant of bacterial cells, is the medium used for initial growth. For example, in these experiments, we culture the bacterium in LB medium, which may vary between lots and for optimal manufacturing procedures, one lot should be used for all experiments. Additionally, the use of alternative medium (e.g., yeast extract beef broth) for initial culturing prior to exposure to infiltration medium, may impact the proteins secreted. Further investigation into precise medium components and their impact on bacterial growth may enable a deeper profiling of the A. tumefaciens secretome. In addition, a lack of proteins identified in the supernatant may be due to low abundance of released or secreted bacterial proteins during the exposure to agroinfiltration medium, supporting low occurrence of cell lysis during the incubation period. Alternatively, the supernatant may be dilute and investigation of proteins in the extracellular environment of a concentrated supernatant or following protein precipitation methodology may provide more protein identifications (Caldwell and Lattemann 2004;Chevallet et al. 2007).
D r a f t

Conclusion.
Our results define dynamic remodeling of A. tumefaciens as changes in the cellular proteome over a time course of exposure to agroinfiltration medium as an adaptation to the nutrient-limited environment. We hypothesize that this remodeling promotes virulence, nutrient acquisition, and motility of the bacterium, which will enhance host infectivity and therefore, the transformation pathway may benefit from prolonged exposure periods. To support this hypothesis, future studies will explore target protein production following transformation of plants with A.
tumefaciens cultures collected at early (0 h), intermediate (2 h), and late (24 h) stages of exposure in agroinfiltration medium, as well as assess cellular viability throughout the exposure time course.
Moreover, we aim to knockout and overexpress the pathogenicity gene (Atu3295) and FlaA and assess bacterial infectivity and potential changes to target protein production. Based on our current findings, and previous reports of A. tumefaciens infectivity, we hypothesize that increased pathogenicity and motility will support hypervirulence, improve infectivity, and increase production of the target protein, which we will explore in future studies. However, increased bacterial virulence may initiate an enhanced plant defense response and therefore, we will monitor changes from both the host and pathogen perspectives simultaneously. Industry-scale production of transiently expressed proteins for pharmaceutical use requires an efficient process strategy for speed and purity. Recent research has explored animal component-free media to meet requirements of good manufacturing processes (GMP), and ready-to-use A. tumefaciens stocks to de-couple bacterial growth and infiltration stages (Houdelet et al. 2017;Spiegel et al. 2019). Here we propose the benefits of long-term exposure to infiltration medium to promote a process for efficient bacterial invasion and transformation.  tumefaciens cultures were grown in shake-flasks or bioreactors, diluted in agroinfiltration medium and exposed at room temperature for 0, 2, and 24 h. Total proteome (cell pellet) and secretome (supernatant) were collected and extracted followed by purification, electrospray ionisation (ESI) and liquid chromatography tandem mass spectrometry (LC-MS/MS). Data processing, analysis, and visualization performed with MaxQuant and Perseus platforms.    Venn diagram of quantitative proteomics results demonstrating overlap of proteins identified in D r a f t this study and proteins identified in recent study (Prudhomme et al. n.d.). The majority of proteins (>80%) are common between the datasets, supporting consistency in the proteomics platforms, while also identifies proteins unique to each condition: purple = 104 proteins only identified in previous study comparing bioreactor vs. shake-flask growth (Prudhomme et al. n.d.); blue = 456 proteins only identified in this study. Student's t-test: *p-value ≤ 0.05; ***p-value ≤ 0.0001. Error bars = standard deviation.