Erythromycin Treatment of Brassica campestris Seedlings Impacts the Photosynthetic and Protein Synthesis Pathways

Erythromycin (Ery) is a commonly used veterinary drug that prevents infections and promotes the growth of farm animals. Ery is often detected in agricultural fields due to the effects of manure application in the ecosystem. However, there is a lack of information on Ery toxicity in crops. In this study, we performed a comparative proteomic analysis to identify the molecular mechanisms of Ery toxicity during seedling growth based on our observation of a decrease in chlorophyll (Chl) contents using Brassica campestris. A total of 452 differentially abundant proteins (DAPs) were identified including a ribulose-1,5-bisphosphate carboxylase (RuBisCO). The proteomic analysis according to gene ontology (GO) classification revealed that many of these DAPs responding to Ery treatment functioned in a cellular process and a metabolic process. The molecular function analysis showed that DAPs classified within catalytic activity were predominantly changed by Ery, including metabolite interconversion enzyme and protein modifying enzyme. An analysis of functional pathways using MapMan revealed that many photosynthesis components were downregulated, whereas many protein biosynthesis components were upregulated. A good relationship was observed between protein and transcript abundance in a photosynthetic pathway, as determined by qPCR analysis. These combined results suggest that Ery affects plant physiological activity by downregulating protein abundance in the photosynthetic pathway.


Introduction
Antibiotics have been used as chemotherapeutic agents in human and veterinary medicine for many decades, and the use of three biologically active molecules has increased globally [1,2]. These substances affect agricultural environments due to human behaviors such as soil fertilization during composting of sludge or manure [3][4][5]. These contaminants accumulate in the soil and affect the soil ecosystem and microbial communities according to their antibiotic resistance [6,7]. to Ery antibiotics and suggest that environmental antibiotics can adversely affect crop plant growth and development.

Detection of Erythromycin Residue in Brassica campestris Seedlings
To detect Ery in B. campestris seedlings, shoots of 4-day-old seedlings grown in the presence of Ery (0, 5, and 10 mg/L) were harvested and washed three times with 50% methanol. To analyze accumulated Ery in planta, washed seedling tissues were freeze-dried, measured (dry weight), and milled using liquid nitrogen. Samples were prepared for liquid chromatography with tandem mass spectrometry (LC-MS/MS), as described Wang et al. [30].

Total Protein Extraction and One-Dimemnsional Gel Electrophoresis
For total protein extraction, harvested seedlings were ground into fine powder using liquid nitrogen, three volumes of ice-chilled protein extraction buffer [1 × PBS pH 7.4, 0.1% Triton X-100, protease inhibitor cocktail tablets (Complete Mini, Roche, Indianapolis, IN, USA)] were added, and the suspension was mixed well. The tubes were incubated in ice for 15 min and then centrifuged for 10 min at 4 • C. The supernatants (total protein extracts) were transferred to new tubes.

In-Gel Digestion
A one-dimensional SDS-PAGE lane containing all protein bands was excised from top to bottom using a razor blade, and excised gel slices were washed twice with 100 µL of distilled water for 15 min at room temperature. Excised gel bands were destained using acetonitrile. The gel slices were dried completely in a vacuum, and then alkylated by incubating with 55 mM iodoacetamide/0.1 M ammonium bicarbonate for 30 min at room temperature in the dark. After alkylation, the gel slices were dried again, and then the dried gel slices were swollen in digestion buffer (25 mM ammonium bicarbonate, 0.1% n-octyl glucoside, and 50 ng/mL of sequencing grade trypsin (Promega, Madison, MI, USA)) for rehydration. Peptides were extracted from the gel slices using 66% acetonitrile, 33% water, 0.1% trifluoroacetic acid (TFA). Extracted peptides were dried with a speedvac (Hanil, Korea), and stored at −80 • C before analysis [31].

Data Analysis
After MS/MS analysis, data files were processed using UniProt and ProteinPilot 5.0.1 (SCIEX, Redwood City, CA, USA). Based on the combined MS and MS/MS spectra, proteins were successfully identified at ≥95% confidence interval using their scores in the MASCOT v 2.6 search engine (Matrix Science Ltd., London, UK) and the following search parameters: plant database (Brassica campestris subsp. napus var pekinensis MAKINO database), trypsin as the digestion enzyme, single missed cleavage sites, fixed carbamidomethyl (C) modifications and methionine oxidation, ±0.1 Da precursor ion tolerance, and ±0.1 Da MS/MS fragment ion tolerance. The database search results were manually curated to yield the protein identifications using 1% global false discovery rate (FDR) determined by the FDR tool in ProteinPilot software. Scaffold v 4.11.0 (Proteome Software Inc., Portland, OR, USA) was used to validate MS/MS-based peptide and protein identifications. The identified proteins were searched, and information on functional grouping was obtained using PANTHER (http://pantherdb.org) and STRING (http://string-db.org) databases for gene ontology (GO) analysis. MapMan software (v 3.6.0RC1; http://mapman.gabipd.org) was used for the functional pathway analysis [32]. A workflow chart of the proteomic analysis is presented in Figure S3.

RNA Extraction and qRT-PCR
Total RNA was extracted from B. campestris seedlings using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions, treated with DNase I (Thermo Fisher Sci, Waltham, MA, USA), and RNA was purified using a Riboclear Column (GeneAll, Seoul, Korea). First-strand cDNA was synthesized from 2 µg of total RNA using a cDNA synthesis kit (Thermo Fisher Scientific). Then, qRT-PCR was performed on a CFX Connect Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) using the cDNA template, gene-specific primers (Supplementary Table S1), and AccuPower 2× GreenStar qPCR Master Mix (Bioneer, Daejeon, Korea). EF1a or Act7 was used as the internal reference gene for data normalization. Average gene expression levels were determined using the comparative Ct method (2 −∆∆Ct ).

Erythromycin Inhibits Brassia campestris Seedling Growth
The effect of Ery on crop physiology, growth, and development was investigated in B. campestris seedlings ( Figure 1). Seeds were placed on 1.2% agar media containing the indicated Ery concentration (0, 2, 5, and 10 mg/L), and the primary root lengths and chlorophyll contents were examined in cotyledons. High Ery concentration (10 mg/L) reduced primary root growth, whereas primary root growth was similar under low Ery concentrations (2 and 5 mg/L) and control conditions (0 mg/L Ery) ( Figure 1B).
Primary root growth can be affected by radical emergence during seed germination. Therefore, we examined radicle emergence in the absence and presence of Ery (10 mg/L). Seed coat rupture was examined every 6 h after imbibition until radicle emergence. There was no significant difference between 0 and 10 mg/L Ery until 24 h after imbibition ( Figure S1), suggesting that Ery does not affect seed germination. Chlorophyll contents in cotyledons were dramatically reduced by Ery ( Figure 1C); specifically, both chlorophyll a and chlorophyll b were reduced as the total chlorophyll content was reduced ( Figure S2). These combined results indicated that Ery inhibited chlorophyll accumulation, suggesting that Ery may affect photosynthetic metabolic process. Primary root growth can be affected by radical emergence during seed germination. Therefore, we examined radicle emergence in the absence and presence of Ery (10 mg/L). Seed coat rupture was examined every 6 h after imbibition until radicle emergence. There was no significant difference between 0 and 10 mg/L Ery until 24 h after imbibition ( Figure S1), suggesting that Ery does not affect seed germination. Chlorophyll contents in cotyledons were dramatically reduced by Ery ( Figure 1C); specifically, both chlorophyll a and chlorophyll b were reduced as the total chlorophyll content was reduced ( Figure S2). These combined results indicated that Ery inhibited chlorophyll accumulation, suggesting that Ery may affect photosynthetic metabolic process.

Erythromycin Accumulation in Vegetative Tissue
Ery influences leafy tissue development ( Figure 1). We hypothesized that Ery was absorbed through the root and subsequently translocated and accumulated in leaf tissue. To investigate Ery levels in shoots, we collected cotyledons from seedlings grown on agar media in the absence (0 mg/L) and presence of Ery (5 and 10 mg/L) for 4 days and examined Ery levels using LC-MS/MS. Although Ery was applied once by adding to the agar medium, Ery absorption through roots may last until harvesting. The Ery residual levels in seedling aerial parts in the presence of 5 and 10 mg/L Ery were 18.3 ± 1.89 and 34.6 ± 0.99 mg/kg (dried weight), respectively. Three times replicated experiments were exhibited similar value. By contrast, Ery was not detected in seedlings grown in the absence of Ery (0 mg/L). These results revealed that Ery was absorbed from contaminated medium, transported from root to shoot, and accumulated in the shoots. These results suggest that Ery residues accumulate in edible plant tissues and may influence human health.

Differential Proteomic Analysis of Brassica campestris Seedlings
To better understand the effects of Ery during seedling growth, we performed differentially expressed proteomic analysis using B. campestris seedlings grown in the absence (0 mg/L, E0) or presence (5 mg/L, E5) of Ery. First, we separated proteins in a size-dependent manner by subjecting them to 12% SDS-PAGE, observed significantly changed protein abundance between E0 and E5 around 50, 20, and 10 kDa (Figure 2A, arrowhead), and confirmed with four times other biological replicates. As shown by the reduction in chlorophyll contents (Figure 1), those protein bands are predicted as RuBisCO large subunits (rbcL) and RuBisCO small subunits (rbcS) [33]. The most

Erythromycin Accumulation in Vegetative Tissue
Ery influences leafy tissue development ( Figure 1). We hypothesized that Ery was absorbed through the root and subsequently translocated and accumulated in leaf tissue. To investigate Ery levels in shoots, we collected cotyledons from seedlings grown on agar media in the absence (0 mg/L) and presence of Ery (5 and 10 mg/L) for 4 days and examined Ery levels using LC-MS/MS. Although Ery was applied once by adding to the agar medium, Ery absorption through roots may last until harvesting. The Ery residual levels in seedling aerial parts in the presence of 5 and 10 mg/L Ery were 18.3 ± 1.89 and 34.6 ± 0.99 mg/kg (dried weight), respectively. Three times replicated experiments were exhibited similar value. By contrast, Ery was not detected in seedlings grown in the absence of Ery (0 mg/L). These results revealed that Ery was absorbed from contaminated medium, transported from root to shoot, and accumulated in the shoots. These results suggest that Ery residues accumulate in edible plant tissues and may influence human health.

Differential Proteomic Analysis of Brassica campestris Seedlings
To better understand the effects of Ery during seedling growth, we performed differentially expressed proteomic analysis using B. campestris seedlings grown in the absence (0 mg/L, E0) or presence (5 mg/L, E5) of Ery. First, we separated proteins in a size-dependent manner by subjecting them to 12% SDS-PAGE, observed significantly changed protein abundance between E0 and E5 around 50, 20, and 10 kDa (Figure 2A, arrowhead), and confirmed with four times other biological replicates. As shown by the reduction in chlorophyll contents (Figure 1), those protein bands are predicted as RuBisCO large subunits (rbcL) and RuBisCO small subunits (rbcS) [33]. The most abundant protein in B. campestris seedling leaf was RuBisCO, which accounted for 11.56% of total leaf protein [34]. Next, we divided five different fractions (F1 to F5, Figure 2A) and analyzed embedded proteins using LC-MS/MS. In-gel digestion with trypsin and cleanup was performed for loading onto LC-MS/MS and analyzed peptides were quantified and profiled for protein identification ( Figure S3). A total of 662 proteins were identified in both E0 and E5 as shown in a Venn diagram analysis ( Figure 2B); 73 proteins were specifically expressed in E5 compared to E0, and 201 proteins were not detected in E5. Although 388 proteins were commonly detected in both E0 and E5, >2-fold difference was quantified between E0 and E5 for 178 proteins (FDR < 0.01), with 17 downregulated proteins and 161 upregulated proteins ( Figure 2C and Table S2). These results indicated that Ery modulates the levels of many B. campestris seedling proteins.
proteins using LC-MS/MS. In-gel digestion with trypsin and cleanup was performed for loading onto LC-MS/MS and analyzed peptides were quantified and profiled for protein identification ( Figure S3). A total of 662 proteins were identified in both E0 and E5 as shown in a Venn diagram analysis ( Figure  2B); 73 proteins were specifically expressed in E5 compared to E0, and 201 proteins were not detected in E5. Although 388 proteins were commonly detected in both E0 and E5, >2-fold difference was quantified between E0 and E5 for 178 proteins (FDR < 0.01), with 17 downregulated proteins and 161 upregulated proteins ( Figure 2C and Table S2). These results indicated that Ery modulates the levels of many B. campestris seedling proteins.

Gene Ontology Analysis of Differentially Expressed Proteins
To deduce functional biological processes, the differentially abundant proteins (DAPs) in E5 were analyzed using PANTHER (http://pantherdb.org) and STRING (http://string-db.org) databases and categorized using gene ontology (GO) analysis ( Figure S3). GO analysis identified four categories of DAPs: biological process, molecular function, cellular component, and protein class ( Figure 3). In the biological process analysis of DAPs, 37% cellular process and 33% metabolic process were dominant among shown other GO biological categories including cellular component organization or biogenesis, localization, biological regulation, response to stimulus, reproductive process, reproduction, multiorganism process, signaling, developmental process, multicellular organismal process, and growth ( Figure 3A,B). Proteins involved in reproductive process, signaling, or growth were rarely changed in response to Ery, suggesting that Ery affects energy generation or consumption for plant survival rather than plant reproduction or development. Although most DAPs in the cellular component category were cytosolic proteins with 54% cell and cell part, the second major portion was 27% organelle and organelle parts ( Figure 3C,D). The remaining 20% portion was involved in membrane, membrane parts, membrane-enclosed lumen, protein-containing complex, extracellular region, supramolecular complex, cell junction, and plasmodesma. Proteins in the molecular functional level were categorized as follows: 52% catalytic activity, 29% binding, 14% structural molecule activity, and 5% others including transporter and regulator function ( Figure 3E). Comparison of E0 and E5 showed that more proteins displayed increased levels than decreased levels ( Figure 3F). More specifically, we further classified the protein classes ( Figure 3G,H). As expected

Gene Ontology Analysis of Differentially Expressed Proteins
To deduce functional biological processes, the differentially abundant proteins (DAPs) in E5 were analyzed using PANTHER (http://pantherdb.org) and STRING (http://string-db.org) databases and categorized using gene ontology (GO) analysis ( Figure S3). GO analysis identified four categories of DAPs: biological process, molecular function, cellular component, and protein class ( Figure 3). In the biological process analysis of DAPs, 37% cellular process and 33% metabolic process were dominant among shown other GO biological categories including cellular component organization or biogenesis, localization, biological regulation, response to stimulus, reproductive process, reproduction, multiorganism process, signaling, developmental process, multicellular organismal process, and growth ( Figure 3A,B). Proteins involved in reproductive process, signaling, or growth were rarely changed in response to Ery, suggesting that Ery affects energy generation or consumption for plant survival rather than plant reproduction or development. Although most DAPs in the cellular component category were cytosolic proteins with 54% cell and cell part, the second major portion was 27% organelle and organelle parts ( Figure 3C,D). The remaining 20% portion was involved in membrane, membrane parts, membrane-enclosed lumen, protein-containing complex, extracellular region, supramolecular complex, cell junction, and plasmodesma. Proteins in the molecular functional level were categorized as follows: 52% catalytic activity, 29% binding, 14% structural molecule activity, and 5% others including transporter and regulator function ( Figure 3E). Comparison of E0 and E5 showed that more proteins displayed increased levels than decreased levels ( Figure 3F). More specifically, we further classified the protein classes ( Figure 3G,H). As expected from biological process and molecular function analysis, 51% metabolic interconversion enzyme and 23% translational proteins represent the major proportions of these classes, with 28% protein modifying enzyme, transporter, scaffold/adaptor protein, chaperone, cell adhesion molecule, protein-binding activity modulator, transfer/carrier protein, nucleic acid binding protein, and cytoskeletal protein. These combined analyses suggested that Ery induced changes in the expression levels of many proteins that may be involved in the function of metabolic enzymes in organelles.
Life 2020, 10, 311 7 of 18 23% translational proteins represent the major proportions of these classes, with 28% protein modifying enzyme, transporter, scaffold/adaptor protein, chaperone, cell adhesion molecule, protein-binding activity modulator, transfer/carrier protein, nucleic acid binding protein, and cytoskeletal protein. These combined analyses suggested that Ery induced changes in the expression levels of many proteins that may be involved in the function of metabolic enzymes in organelles.

MapMan Pathway Analysis of Differentially Expressed Proteins
Our proteomics analysis revealed that Ery modulated the levels of a number of proteins involved in central metabolism. To further identify Ery-induced changes in the functional metabolic pathway during plant growth, we mapped quantified DEP values to MapMan pathways (MapMan ontology version 3.6.0), which refines plant-sourced protein classification and annotation frameworks [32]. Functional pathways contain 18.3% photosynthesis and photorespiration (66 DAPs), 19.4% proteins synthesis  Figure S4). Although we identified 452 DAPs, 360 DAPs were applied to investigate the protein abundance of individual DAP based on molecular functional category, since 92 DAPs were difficult to use gene accession number, which is required to analyze transcript abundance (Table S2). Then, we analyzed the change in each DAP in the molecular functional pathway to understand the modulation of indicated pathways. Most DAPs in each pathway were distributed with similar numbers of upregulated and downregulated, except for photosynthesis and protein biosynthesis. The photosynthetic pathway and protein biosynthesis pathway had much higher numbers of downregulated DAPs and upregulated DAPs, respectively ( Figure S4). The quantified values of characterized proteins (Table S2) were represented in MapMan pathway, and the relative protein abundances were compared in log2 range ( Figure 4). As expected, most proteins characterized in the photosynthetic pathway displayed decreased expression levels (color gradient green). By contrast, many protein functions in protein biosynthesis, protein modification, and protein homeostasis displayed increased expression levels (color gradient red) (Figure 4). The list of significantly expressed DAPs in these MapMan pathways are presented in Table 1 for photosynthesis and Table 2 for protein synthesis. Given that all ribosomal proteins are estimated as approximately 8.47% of total lead proteins [34], many ribosomal proteins were identified in Table 2. For the protein annotation in Tables 1 and 2, we also used Arabidopsis AGI locus as an identifier (http://plants.ensembl.org/Brassica_rapa/; http://brassicadb.org/) due to the lack of Brassica gene accessions.
Our proteomics analysis revealed that Ery modulated the levels of a number of proteins involved in central metabolism. To further identify Ery-induced changes in the functional metabolic pathway during plant growth, we mapped quantified DEP values to MapMan pathways (MapMan ontology version 3.6.0), which refines plant-sourced protein classification and annotation frameworks [32]. Functional pathways contain 18.3% photosynthesis and photorespiration (66 DAPs), 19.4% proteins synthesis (70 DAPs), 21.2% metabolic pathway including carbohydrate metabolism (80 DAPs), 9.4% protein homeostasis (34 DAPs), 3.6% redox homeostasis (13 DAPs), 5% transport (18 DAPs), 2.5% cell structure (9 DAPs), and 19.4% not assigned (70 DAPs) ( Figure S4). Although we identified 452 DAPs, 360 DAPs were applied to investigate the protein abundance of individual DAP based on molecular functional category, since 92 DAPs were difficult to use gene accession number, which is required to analyze transcript abundance (Table S2). Then, we analyzed the change in each DAP in the molecular functional pathway to understand the modulation of indicated pathways. Most DAPs in each pathway were distributed with similar numbers of upregulated and downregulated, except for photosynthesis and protein biosynthesis. The photosynthetic pathway and protein biosynthesis pathway had much higher numbers of downregulated DAPs and upregulated DAPs, respectively ( Figure S4). The quantified values of characterized proteins (Table S2) were represented in MapMan pathway, and the relative protein abundances were compared in log2 range ( Figure 4). As expected, most proteins characterized in the photosynthetic pathway displayed decreased expression levels (color gradient green). By contrast, many protein functions in protein biosynthesis, protein modification, and protein homeostasis displayed increased expression levels (color gradient red) (Figure 4). The list of significantly expressed DAPs in these MapMan pathways are presented in Table 1 for photosynthesis and Table 2 for protein synthesis. Given that all ribosomal proteins are estimated as approximately 8.47% of total lead proteins [34], many ribosomal proteins were identified in Table 2. For the protein annotation in Tables 1 and 2, we also used Arabidopsis AGI locus as an identifier (http://plants.ensembl.org/Brassica_rapa/; http://brassicadb.org/) due to the lack of Brassica gene accessions.

DAPs in Photosynthesis and qRT-PCR Analysis
Given that Ery downregulated protein function in photosynthesis, we investigated the role of DAPs in the photosynthesis pathway and/or chloroplast function using MapMan ( Figure 5). As shown in Table 1, we found that most DAPs (E5) in the photosynthesis pathway had lower quantitative values than expressed proteins in the control (E0) (color gradient green), indicating that Ery reduced the expression levels of protein functions in photosynthesis, including those corresponding to PSII in the Life 2020, 10, 311 12 of 18 light reaction ( Figure 5A) and RuBisCO in the Calvin cycle ( Figure 5B). This result suggested that Ery reduced photosynthetic activities of these DAPs.
Life 2020, 10, x FOR PEER REVIEW 12 of 18 Ery reduced the expression levels of protein functions in photosynthesis, including those corresponding to PSII in the light reaction ( Figure 5A) and RuBisCO in the Calvin cycle ( Figure 5B). This result suggested that Ery reduced photosynthetic activities of these DAPs. To further investigate the relationship between DAPs and transcription, we performed qRT-PCR. Although the DAPs were identified from seedlings treated with 5 mg/L Ery (E5), the transcript levels of corresponding proteins were evaluated in seedlings grown on 0, 2, 5, and 10 mg/L Ery ( Figure 6). The abundance patterns of targeted Brassica genes encoding functional proteins in a photosynthetic pathway (Table S2) were merged with our results in Table 1 and Figures 4 and 6. Photosyntheticrelated genes included Bra040977, Bra041106, Bra041120, Bra011329, Bra034200, Bra028087, Bra031534, Bra014908, Bra036240, Bra000837, Bra040927, Bra011792, and Bra026951. Some of these genes are encoded in the chloroplastic genome and play roles in the chloroplast according to Arabidopsis identifier information, such as Bra040977, Bra041106, and Bra041120 ( Figure 6A). This result indicated that Ery can influence chloroplastic gene expression and chloroplastic protein abundance. Most tested genes in the photosynthetic pathway were downregulated and their gene expression was significantly reduced (Student's t-test; * p < 0.05, ** p < 0.01, *** p < 0.001) ( Figure 6B). Bra026951, which was annotated as an upregulated DAP, was not induced significantly at the transcriptional level, whereas Bra041120 was highly upregulated at low Ery concentration (2 mg/L) and dramatically declined at high Ery concentrations (5 and 10 mg/L) ( Figure 6A,C). These results suggested that some identified DAP genes were regulated within different temporal points for gene expression or protein expression/stability depending on Ery concentration. We also observed similar consistent results in different pathways; for example, Bra030284 encoded an RNA-binding protein and was categorized in both protein biosynthesis pathway and RNA processing. Bra008784 serves a role as a phosphate solute carrier at the mitochondrial membrane in an unassigned group ( Figure 6D,E). These results demonstrate that our proteomic and transcript analyses results are consistent. To further investigate the relationship between DAPs and transcription, we performed qRT-PCR. Although the DAPs were identified from seedlings treated with 5 mg/L Ery (E5), the transcript levels of corresponding proteins were evaluated in seedlings grown on 0, 2, 5, and 10 mg/L Ery ( Figure 6). The abundance patterns of targeted Brassica genes encoding functional proteins in a photosynthetic pathway (Table S2) were merged with our results in Table 1 and Figures 4 and 6. Photosynthetic-related genes included Bra040977, Bra041106, Bra041120, Bra011329, Bra034200, Bra028087, Bra031534, Bra014908, Bra036240, Bra000837, Bra040927, Bra011792, and Bra026951. Some of these genes are encoded in the chloroplastic genome and play roles in the chloroplast according to Arabidopsis identifier information, such as Bra040977, Bra041106, and Bra041120 ( Figure 6A). This result indicated that Ery can influence chloroplastic gene expression and chloroplastic protein abundance. Most tested genes in the photosynthetic pathway were downregulated and their gene expression was significantly reduced (Student's t-test; * p < 0.05, ** p < 0.01, *** p < 0.001) ( Figure 6B). Bra026951, which was annotated as an upregulated DAP, was not induced significantly at the transcriptional level, whereas Bra041120 was highly upregulated at low Ery concentration (2 mg/L) and dramatically declined at high Ery concentrations (5 and 10 mg/L) ( Figure 6A,C). These results suggested that some identified DAP genes were regulated within different temporal points for gene expression or protein expression/stability depending on Ery concentration. We also observed similar consistent results in different pathways; for example, Bra030284 encoded an RNA-binding protein and was categorized in both protein biosynthesis pathway and RNA processing. Bra008784 serves a role as a phosphate solute carrier at the mitochondrial membrane in an unassigned group ( Figure 6D,E). These results demonstrate that our proteomic and transcript analyses results are consistent. . The experiments were replicated three times with similar results. Asterisks indicate statistically significant differences relative to the control (0 mg/L Ery). Student's t-test; * p < 0.05, ** p < 0.01, *** p < 0.001.

Discussion
Environmental residual chemicals such as heavy metals and antimicrobial drugs originating from use in humans and animals affect plant growth and development. These chemicals can accumulate in crop plants, affect food chains, and impact human health [35]. Erythromycin antibiotics have been detected in agricultural fields used for cultivating crops [36,37]. However, it is not well The relative transcript levels of indicated B. campestris genes were analyzed by qRT-PCR. Total RNA was extracted from seedlings grown in the presence of 0, 2, 5, and 10 mg/L Ery for 4 days. Gene expression was normalized to that of EF1a. Bars represent mean ± SD (n = 12). The experiments were replicated three times with similar results. Asterisks indicate statistically significant differences relative to the control (0 mg/L Ery). Student's t-test; * p < 0.05, ** p < 0.01, *** p < 0.001.

Discussion
Environmental residual chemicals such as heavy metals and antimicrobial drugs originating from use in humans and animals affect plant growth and development. These chemicals can accumulate in crop plants, affect food chains, and impact human health [35]. Erythromycin antibiotics have been detected in agricultural fields used for cultivating crops [36,37]. However, it is not well understood how Ery affects crop physiology and/or productivity. This study investigated the effects of Ery in crop growth and development using B. campestris seedlings. Although the chemical structure of Ery is not easily absorbed and translocated through cellular membranes [38], Ery accumulation and translocation was reported in hydroponically grown poplar plants irrigated with a recirculating Hoagland's nutrient solution containing~0.01 mg/L Ery [15]. We confirmed Ery accumulation and translocation from root to shoot in B. campestris. Poplar plants grown on 1 mg/L Ery exhibited similar health as those grown on 0 mg/L Ery [15], consistent with our observed phenotypes of B. campestris grown on 0 and 5 mg/L Ery (Figure 1). These results indicated that low Ery concentrations do not alter morphological phenotypes and suggested that it would be difficult to identify potential toxicity from low Ery concentrations during crop cultivation. By contrast, we observed a severe chlorophyll-deficient phenotype in seedlings grown on higher concentrations than 2 mg/L Ery (Figure 1 and Figure S1) [39], and growth defects at 10 mg/L Ery, suggesting that higher Ery concentrations affect plant growth and development. These combined results suggest that Ery levels detected in the environment (<0.1 mg/L, [36,37]) may not induce morphological and physiological effects and reduced chlorophyll contents due to Ery absorption may not be distinguishable from senescent leaf development during crop cultivation.
Higher Ery concentrations induced physiological differences in B. campestris seedlings such as chlorophyll-deficient phenotype (Figure 1 and Figure S1). We identified many Ery-induced differentially abundant proteins (DAPs) (Table S2). These DAPs primarily function in cellular processes and are predicted to have catalytic activity in metabolite interconversion enzyme and protein modifying enzyme ( Figure 3). Consistently, studies using a fish report that Ery affects catalytic activities [40,41]. Ery inhibits acetylcholinesterase (AChE) and glutathione S-transferase (GST) activity but enhances superoxide dismutase (SOD) and catalase (CAT) activitys [40][41][42]. Identified DAPs were rarely involved in cell division, cytoskeleton, plant hormone, and vascular tissue development according to classification of biological and molecular functional process, thereby supporting evidence that Ery does not primarily affect morphology.
Ery predominantly inhibits prokaryotic protein synthesis by binding irreversibly to the 50S ribosomal subunit [43,44]. Prokaryotic ribosome is composed of 30S and 50S subunits containing four rRNA species and approximately 55 ribosomal proteins (r-proteins), and its structure and function resemble chloroplast ribosome [45,46]. Functional pathway analysis using MapMan indicated that Ery either directly or indirectly affects photosynthetic protein expression and protein biosynthesis ( Figure S4 and Figure 4). Many DAPs in the photosynthetic pathway were downregulated ( Figure S4 and Figure 4, Table 1), whereas DAPs classified in protein biosynthesis, modification, and homeostasis were predominantly upregulated ( Figure S4 and Figure 4, Table 2). Chloroplast organelles possess an independent protein synthesis pathway [47]. Approximately one-third of chloroplast r-proteins are encoded by the chloroplast genome, and two-thirds are encoded in the nucleus, synthesized in the cytoplasm, and migrate to the chloroplast [48]. As shown in Table 2, Ery contributed to the abundant expression of ribosomal proteins in the chloroplastic genome and nuclear genome compared to the control (E0), suggesting that Ery interrupted the chloroplast translation machinery, including ribosomal structure and organization that results from chloroplastic and nucleic genomes.
Ery affected metabolic process ( Figure 3). As photosynthetic activity, which is an important metabolic process in plants [15,42,49], we observed decreases in photosynthetic pigment, chlorophyll content ( Figure 1 and Figure S1), a photosynthetic pathway of DAPs functional distribution ( Figure 4 and Figure S4), and downregulation of chloroplastic DAPs and/or DAPs functioning in photosynthetic machinery such as rbcL and rbcS (RuBisCO) ( Table 1). Chlorophylls play key roles in all aspects of the photosynthetic light reaction, including light harvesting, energy transfer, and light energy conversion, suggesting that the alteration of chlorophyll fluorescence parameters may reflect photosynthesis change. The photosynthetic activity of Porphyra yezoensis was reduced by Ery treatment, which reduced F v /F m and changed color [49]. Studies using Microcystis flos-aquae and Selenastrum capricornutum showed that Ery inhibits F v /F m and the F v /F 0 ratio by targeting a candidate thylakoid membrane system [42,50]. These Ery-induced effects on chlorophyll fluorescence quenching and F v /F m depended on Ery dosage and exposure time [49,51,52]. In the light reaction during photosynthesis, photochemical energy conversion by charge separation in PSII reaction centers is reduced by changes in the linear electron transport rate to PSI through the cytochrome complex [53,54]. As expected, Ery inhibited protein synthesis of components in the thylakoid, including the PSII reaction center and cytochrome complex, and reduces membrane proteins including ATPase ( Figure 5A) [50]. These results suggested that Ery inhibited primary photochemistry due to PSII reaction center deficiency and/or disrupted electron transport from PSII to PSI due to cytochrome complex deficiency ( Figure 5A) [51]. Studies using eukaryotic green algae that possess chloroplasts showed that levofloxacin and amphotericin B antibiotics significantly inhibit photosynthetic electron transport [55,56].
Ribulose bisphosphate carboxylase is a rate-limiting enzyme in the photosynthetic carbon reduction cycle that catalyzes the first step of the carbon assimilation process ( Figure 5B). Ery inhibited the synthesis of ribulose bisphosphate carboxylase subunits (rbcL and rbcS), reduced their contents (Figure 2A, Table S2 and Table 1) [52], interfered with assembly, and attenuated RuBisCO activity of this enzyme. Transgenic tobacco plants expressing <50% of wild-type RuBisCO activity had reduced starch and sucrose levels [57,58], and leaf development and growth of transgenic plants expressing only 20% of normal RuBisCO activity were altered with reduced photosynthetic capacity [59]. Ery also affected photosynthetic gene transcription ( Figure 6) [60] and protein abundance ( Table 1), suggesting that disrupted electron transport, decreased reducing power of ATP and NADPH levels, and low RuBisCO activity affect carbon fixation and assimilation.
In conclusion, we suggest that environmental Ery acted as a protein synthesis inhibitor, thereby changing protein abundance and modulating plant physiology involved especially in photosynthesis, and crop production. As this is the first proteomics study of the effects of veterinary antibiotics in plants, we also provided a comprehensive overview of proteomics analyses of plant response to macrolide antibiotics. Our study may lead to a broader understanding of molecular and physiological responses to environmental antibiotics in crop plants.

Conflicts of Interest:
The authors declare no conflict of interest.