Senescence and defense pathways contribute to heterosis

in the of ABSTRACT 37 Hybrids are used yield underpinning Recent evidence has suggested that a decrease in basal defense 40 response gene expression regulated by reduced levels of salicylic acid (SA) may be 41 important for vigor in certain hybrid combinations. Decreasing levels of SA in the 42 Arabidopsis thaliana accession C24 through the introduction of the SA catabolic enzyme 43 salicylate 1 hydroxylase (NahG) increases plant size, phenocopying the large sized 44 C24/Landsberg erecta (L er ) F1 hybrids. C24♀ x L er ♂ F1 hybrids and C24 NahG lines 45 shared differentially expressed genes and pathways associated with plant defense and leaf 46 senescence including decreased expression of SA biosynthetic genes and SA response 47 genes. The expression of TL1 BINDING TRANSCRIPTION FACTOR 1 ( TBF1 ) , a key 48 regulator in resource allocation between growth and defense, was decreased in both the F1 49 hybrid and the C24 NahG lines, which may promote growth. Both C24 NahG lines and the 50 F1 hybrids showed decreased expression of the key senescence-associated transcription 51 factors WRKY53 , NAC-CONTAINING PROTEIN 29 ( NAP) , and ORESARA1 (ORE1) with a 52 delayed onset of senescence compared to C24 plants. The delay in senescence resulted in 53 an extension of the photosynthetic period in the leaves of F1 hybrids compared to the 54 parental lines, potentially allowing each leaf to contribute more resources towards growth.


INTRODUCTION
Some hybrid systems show a decrease in basal expression of defense response genes in 83 the seedling, which in the absence of pathogens may increase resource allocation to plant 84 growth potentially contributing to hybrid vigor (Groszmann et al, 2015;Miller et al, 2015;85 Yang et al, 2017). Conversely, altering the growth-defense balance towards defense can 86 result in hybrid weakness, where hybrids with hyper-activated defense pathways have 87 reduced growth (Todesco et al, 2010). Hybrids with a decrease in basal defense gene 88 expression may have an associated decrease in the level of salicylic acid (SA; Groszmann 89 et al, 2015). Reducing levels of SA in the Arabidopsis accession C24 using the bacterial 90 degradative enzyme salicylate 1 hydroxylase (NahG), phenocopies hybrid vigor including the 91 increased plant size, suggesting that SA-regulated pathways may have an important role in 92 certain hybrid combinations (Groszmann et al, 2015). 93 We determined whether the size increases of F1 hybrids and C24 NahG lines result from 94 changes to the expression level of genes in the same pathways. In both genotypes the 95 decrease in SA is associated with reduced expression of basal defense response genes and 96 a delay in senescence, both of which may promote increased growth. 97 98 99 RESULTS 100

C24 NahG lines show increased growth at late stages of seedling development 101
Reducing SA levels in C24 by introducing the bacterial NahG gene into its genome results in 102 an increased plant size similar to that of F1 hybrids between C24 and Landsberg erecta (Ler;103 Groszmann et al, 2015). From 0 -21 days after sowing (DAS), three T3 C24 NahG lines 104 were the same size as C24 (Figure 1 & Supplemental Figure S1). After 21 DAS, all three 105 C24 NahG lines began to outgrow C24 resulting in an 18-36% increase in rosette diameter 106 and a 27-45% increase in fresh weight relative to C24 ( Figure 1B,1C, Supplemental Figure  107 S1 & Supplemental Table S1). Ler has low levels of SA which are only slightly reduced in 108 the Ler NahG line and there is only a small increase in plant size ( Figure 1D & Supplemental 109 Figure S1C). Unlike C24 NahG, the F1 hybrids are larger than the parental accessions at all 110 developmental stages suggesting that gene expression changes not related to SA also 111 contribute to the hybrid vigor phenotype (Supplemental Figure S1C & Supplemental Table  112 S1). 113

NahG alters gene expression patterns in C24 but only minimally in Ler 114
Transcriptomes of C24, C24 NahG (T2 independent lines 2 and 3), Ler, Ler NahG and the 115 F1 hybrid were analyzed at 21 DAS to determine which altered pathways were associated 116 with the size increases in C24 NahG and the F1 hybrids. The NahG transgene was 117 expressed in the NahG lines (Supplemental Figure S1D). 6816 genes (35% of the expressed 118 genes) were differentially expressed between the two parental lines (C24 and Ler) at the 21  Table S2; p ≤ 0.01; 120 FDR ≤ 0.01; fold change ≥ ± 1.2). In the F1 hybrid, 3371 genes were differentially expressed 121 compared to the average expression of the two parents (Mid Parent Value (MPV)), with 122 ~60% of these downregulated ( Table S2). Only genes altered in both C24 NahG lines were used 126 for subsequent comparisons. Most of the downregulated C24 NahG DEGs were also 127 downregulated in the F1 hybrid (75%; Figure 2B). The F1 hybrid contained many more 128 down-regulated DEGs than C24 NahG which may reflect changes in expression of both C24 129 and Ler alleles in the F1 hybrid. There was only a small overlap in upregulated DEGs 130 between C24 NahG and the F1 hybrid indicating that if they do share a common mechanism 131 for increased size it is likely due to the shared downregulated genes ( Figure 2B). 132 In Ler NahG, only 225 genes were differentially expressed compared to Ler, with 87% (154) 133 of the downregulated DEGs overlapping with those of C24 NahG ( Table S2). The difference in the number of DEGs between Ler 135 NahG and C24 NahG compared to their parents is likely to be a consequence of the higher 136 initial level of SA in C24 where there is a greater scope for downregulation ( Figure 1D) Figure 3A). The decrease of expression of these SA 146 biosynthesis genes is associated with decreased expression of downstream genes such as 147 WRKY transcription factors, PATHOGEN RELATED GENES (PR1-5) and ACCELERATED 148 of SA biosynthetic genes and downstream targets of SA demonstrates that the decreased 155 expression of these genes in the F1 hybrid is through a reduction in expression of the C24 156 alleles with little change in the expression of the Ler alleles (Supplemental Figure S3B; 157 Supplemental Table S3).
Several other defense-related pathways including stress 158 response, biotic stimuli, and programmed cell death are also downregulated ( Figure 3C; 159 Supplemental Table S4). The downregulation of all these pathways in the F1 hybrid and C24 160 NahG compared to the C24 parent suggests that changes to these pathways may contribute 161 to the increased growth of F1 hybrids and C24 NahG. 162 While many downregulated genes and pathways were shared between the F1 hybrid and 163 C24 NahG, those with upregulated expression levels differed between the two lines (   Table S4). While not statistically significant, there 168 was a trend for upregulation of genes in the carbon cycle and chlorophyll biosynthesis in 169 C24 NahG suggesting that these pathways could influence plant size in C24 NahG (fold 170 change ≥ 1.5; Supplemental Figure S4). Hybrids but not NahG lines had increased 171 expression of PIF4 and PIF5, genes which have been implicated in hybrid vigor through 172 effects on auxin ( Figure 4A; Wang et al, 2017). This difference could result in the larger size 173 of F1 hybrids compared to C24 NahG at early stages of development when these genes are 174 highly expressed (Wang et al, 2017). Another difference was that in the C24 NahG lines 175 there was an altered level of jasmonic acid (JA) responsive genes and glucosinolate 176 biosynthesis genes which was not present in the F1 hybrid nor in Ler NahG lines. While JA 177 responsive genes showed altered expression in C24 NahG no changes were observed in JA 178 biosynthetic gene expression or JA hormone levels (Supplemental Figure S5). 179 TBF1, a master regulator of defense and growth, is downregulated in F1 hybrids and 180

C24 NahG 181
The vigor observed in both F1 hybrids and C24 NahG plants is associated with an increase 182 in leaf cell size (Groszmann et al, 2015). Several genes which increase cell size including 183 In the F1 hybrid the decrease in SA and in expression of related defense pathway genes 192 could release more resources to be allocated to plant growth.

TL1 BINDING 193
TRANSCRIPTION FACTOR 1 (TBF1) is a transcription factor which is induced by SA and 194 regulates resource allocation between defense and growth (Pajerowska-Mukhtar et al, 195 2012). TBF1 represses genes associated with plant growth, including chloroplast proteins 196 and enhances expression of genes involved in plant defense (Pajerowska-Mukhtar et al, 197 2012). In F1 hybrids we found a > 3-fold reduction in TBF1 expression relative to C24, with 198 levels below both parents ( Figure Figure S7C). The regulation of TBF1 occurs both at 208 transcriptional and translational levels (Pajerowska-Mukhtar et al, 2012). A network analysis 209 of the genes regulated by TBF1 suggests that TBF1 protein activity is downregulated in the 210 F1 hybrid and C24 NahG (Supplemental Figure S8A & B). This did not occur in Ler NahG 211 where genes regulated by TBF1 showed only small changes in gene expression 212 (Supplemental Figure S8C). 213

Decreased expression of SA associated genes is consistent with delayed senescence 214 in F1 hybrids and NahG lines 215
Senescence is a tightly regulated process that allows a plant to repurpose nutrients from 216 older leaves into the development of new leaves or to reproductive structures. Senescence  Figure 6A). 230 We analyzed the expression pattern of 3211 genes known to be upregulated during 231 developmental senescence and 2496 genes known to be downregulated during 232 developmental senescence (Supplemental Table S5; Allu et al, 2014). Relative to C24, C24 233 NahG, Ler, Ler NahG and the F1 hybrid displayed a pattern of expression consistent with a 234 delay in senescence ( Figure 6B). Approximately 30% of the genes downregulated in hybrids 235 are genes known to be upregulated during senescence, while 40% of the F1 hybrid's 236 upregulated genes are known to be downregulated during senescence (Supplemental 237 Figure S9). A similar pattern was observed in C24 NahG but not Ler NahG where there was 238 no difference in the overlap between genes upregulated or downregulated during 239 senescence (Supplemental Figure S9C). These patterns of expression are consistent with 240 the F1 hybrid and C24 NahG having delayed senescence. Hormone levels were also 241 altered in a way consistent with a delay in senescence. Along with the decreased level of 242 SA we observed a statistically significant decrease in levels of abscisic acid (ABA) at 21 243 DAS which is also known to promote senescence ( Figure 6C; reviewed in Kim et al, 2017). 244 The combination of gene expression patterns and hormone levels suggested that both the 245 F1 hybrid and the C24 NahG plants have delayed senescence. 246 Under long day conditions both C24 NahG and the F1 hybrid had delayed senescence of 247 leaf 7 compared to C24 (35 DAS; Supplemental Figure S10). While Ler NahG also had 248 delayed senescence compared to Ler the delay was less. The delay in senescence was 249 observed across the whole rosette with the F1 hybrid having fewer senescing leaves 250 compared to either parent (Supplemental Figure S10B & Supplemental Figure S11). 251 Flowering is a strong inducer of senescence and could influence the senescence patterns. (Ler), 10.7±0.5% (Ler NahG), 9.5±0.5% (C24 NahG) and 8.3±0.3% (F1 hybrid; Figure 7B). 270 These values demonstrate that altering levels of SA has a stronger impact on senescence in 271 the C24 genotype than in the Ler genotype (6% versus 1.9% difference in chlorosis rate 272 between the NahG line and wild type). To take into account differences in leaf size between 273 genotypes, we combined absolute chlorophyll area (whole leaf) with PSII efficiency (F v /F m ) to 274 obtain photosynthetic capacity of the leaf. The F1 hybrids had greater photosynthetic 275 capacity than the other genotypes potentially contributing more resources to growth ( Figure  276 7A). 277 We examined the impact that delayed senescence had on leaf longevity in intact rosettes 278 under short day conditions. We identified 3 stages of leaf development, "early" when leaves 279 are increasing in size, a "plateau" stage where leaves have reached maximum size and have 280 the maximum photosynthetic potential and "senescence" where we begin to see chlorosis 281 Figure S12). Leaves 5 and 6 attained their maximum size in all 282 genotypes between 40 -44 DAS (within 5% of maximum size). C24 stopped leaf growth at 283 40 days while Ler, Ler NahG and the F1 hybrid increased in size for 2 more days (42 DAS). 284 C24 NahG showed the largest extension time for growth reaching maximum size by 44 DAS. 285 C24 retained maximum photosynthetic potential for 2-3 days before initiating senescence 286 Figure S12). The F1 hybrid and Ler both plateaued for 7 days 287 but once senescence began the rate of chlorosis was slower in the F1 hybrid ( Figure 7C; 288 Supplemental Figure S12B). The addition of NahG to C24 resulted in the extension of 289 maximum photosynthetic potential by 14 DAS while in Ler, maximum photosynthetic 290 potential was only extended by 4 DAS ( Figure 7C). The extended period of maximum 291 photosynthetic potential could produce more resources for growth in F1 hybrids and NahG  C24/Ler F1 hybrids and C24 NahG both have increased size compared to the C24 302 genotype. We asked whether the size increase of both F1 hybrids and C24 NahG resulted 303 from similar changes in gene expression. The two lines had similar changes to a number of 304 developmental processes which could influence growth, however they also showed changes 305 unique to each genotype. 306 Both F1 hybrids and C24 NahG had reduced SA concentrations and low expression of basal 307 defense response genes. A reduction in expression of defense response genes at 21 DAS 308 may allocate more resources to growth than to defense (Brown 2002;Tian et al, 2003; defense pathway genes. While not significant C24 NahG did display a trend for upregulation 319 of photosynthetic genes suggesting that both F1 hybrids and C24 NahG, through a SA 320 driven change in TBF1, may have altered resource allocation towards plant growth. 321 The decrease in basal defense response genes may impact the F1 hybrid's ability to 322 respond to biotic and abiotic conditions. However this does not seem to be the case in C24 323 derived hybrids that respond normally to either biotic or abiotic stress conditions suggesting 324 that the SA-related changes are reversible (Rohde et al, 2004;Groszmann et al, 2015;Yang 325 et al, 2015). The similarity between F1 hybrids and C24 NahG did not extend to Ler NahG. 326 Ler NahG lines did have decreased expression of a number of SA-related defense response 327 genes however the reduction in expression compared to Ler was much smaller than that 328 observed for the C24 NahG system. Ler NahG also lacked changes in expression of SA 329 biosynthetic genes and TBF1 suggesting that unlike C24 NahG, Ler NahG does not have 330 any change in its resource allocation. These differences may reflect why C24 NahG shows 331 F1-like vigor while Ler NahG shows only a slight increase in biomass compared to Ler. 332 A difference between the F1 hybrids and the C24 NahG lines was the developmental timing 333 at which vigor was observed. The increased longevity of the leaves in F1 hybrids resulted in an extension in the period of 357 photosynthesis. This change in photosynthesis along with the observed change in gene 358 expression related to resource allocation may explain the increased size of F1 hybrids and 359 C24 NahG. Ler NahG also showed a delay in senescence, however the difference between 360 Ler and Ler NahG in senescence was much smaller than that for C24 and C24 NahG. This with an increase in seed weight and grain yield (Spano et al, 2003). 373 In a Chinese hybrid wheat variety, the high yield of the hybrid was related to higher levels of 374 CO 2 assimilation and PSII activity compared to the parents in late stages of leaf 375 development (Yang et al, 2007). In super high-yield hybrid rice, a delay in senescence 376 correlates with higher levels of photosynthetic activity during mid to late development and 377 increased seed yield compared to parents (Zhang et al, 2007, Chang et al, 2016. Ear 378 leaves of some maize hybrids show a delay in senescence compared to the parents with an 379 increase in photosynthetic functions during mid to late stages of growth which correlates with 380 increased leaf size and yield (Song et al, 2016). 381

Conclusion 382
In the hybrid nucleus interactions between the two parental genomes and epigenomes alter 383 the expression of some genes, contributing to hybrid vigor. A reduction in the level of SA in 384 transgenic C24 NahG leads to increased growth and decreased expression levels of basal 385 defense response genes. Parallel changes occur in C24/Ler hybrids, which have decreased 386 levels of SA. In both systems the decrease in SA and associated changes in defense 387 response gene expression may affect resource allocation through the action of TBF1 which 388 facilitates resource distribution between growth and defense. The decrease in SA is 389 correlated with a delay in senescence which extends the period of photosynthesis per leaf 390 increasing energy resources available for continued growth. The change in resource 391 allocation and the delayed onset of senescence may be important contributors to the 392 development of heterosis. 393

Plant material 395
Wild-type Arabidopsis thaliana C24 and Ler, two independent T2 C24 NahG lines, C24 396 NahG 2 and C24 NahG 3, a T2 Ler NahG line and C24♀ x Ler♂ F1 hybrids were used for 397 transcriptome sequencing. For senescence and hormone experiments three T3 C24 NahG 398 lines C24 NahG 2-5 (derived from C24 NahG 2), C24 NahG 3-3 and C24 NahG 3-4 (derived 399 from C24 NahG 3) were used to compare to parents and F1 hybrids. Hybrid seed was 400 background. REViGO was then used to account for redundancies with settings previously 440 described (Groszmann et al, 2015). For heat maps present in the paper gene expression 441 changes greater than 3 were maximized to 3. 442 Chloroplast associated genes repressed by SA through a TBF1 dependent manner were 443 obtained from Supplemental Table S1 of Pajerowska-Mukhtar et al, 2012. The gene list was 444 run through agriGO using TAIR9 as a background reference. 445

Single nucleotide polymorphism (SNP) analysis 446
SNP analysis was carried out as previously described (Zhu et al, 2016). The number of read 447 counts for each allele was defined by samtools mpileup (samtools 1.3.1). Replicates were 448 combined for this analysis. Only positions of known SNPs between C24 and Col or between 449 Ler and Col were reported using the -l option. SNPs were obtained from the 1001 genome 450 project (www.1001genomes.org). If the same SNP was present in both C24 and Ler it was 451 excluded from the analysis. Bedtools intersect function was used to place SNPs within 452 genes followed by the bedtools groupby function to combine reads numbers of all SNPs 453 within a gene. The resulting read numbers were normalized to the highest sample. Only 454 genes with at least 3 SNPs present were used for the analysis. Allele ratios were then 455 compared between the parents and the F1 hybrid. 456

Senescence Associated Gene analysis 457
Senescence associated genes were obtained from Allu et al, 2014. Only genes where one 458 sample had at least 30 read fragments was used (Supplemental Table S5

TBF1 protein activity 464
The regulatory networks were reverse engineered by ARACNe from multiple RNA-seq data 465 sets in seedling Arabidopsis: ARACNe was run with 100 bootstrap iterations using all probe 466 clusters (Margolin et al, 2006;Groszmann et al, 2015, Wang et al, 2015, Wang et al, 2017. 467 Parameters were set to 0.2 DPI (data processing inequality) tolerance.
Function 468 aracne2regulon was used to generate regulon objects from networks reverse engineered 469 with the ARACNe algorithm. This step took two arguments as input: the ARACNe output file, 470 and the expression data-set used by ARACNe to reverse engineer the network. Gene 471 expression signatures were identified by filtering for differentially expressed genes using the 472 function rowTtest which is included in the viper package that efficiently performs Student's t-473 test for each row of a dataset. The rowTtest function took an 'ExpressionSet' object as 474 argument and produces a list object containing the Student's t-test's p-value that by default 475 is estimated by a 2-tail test. The msVIPER analysis is performed by the msVIPER function 476 (Alvarez et al, 2016). It requires a GES, regulon object and null model as arguments, and 477 produces an object of class 'msVIPER', containing the GES, regulon and estimated 478 enrichment, including the Normalized Enrichment Score (NES) and p-value, as output. 479

Reverse transcription quantitative PCR (RT-qPCR) 480
For RT-qPCR, aerial tissue of 21 DAS were snap frozen in liquid nitrogen and ground in a 481 mortar and pestle. Total RNA was extracted using the Sigma spectrum plant total RNA kit 482 (Sigma; STRN250). Genomic DNA was digested from 2 µg of total RNA using DNase I 483 following the manufacturer's instructions (Invitrogen 18068015).
cDNA was then 484 synthesized from the RNA using Superscript III reverse transcriptase following the 485 manufacturer's instructions (Invitrogen; 18080044). Then, 20 µl of cDNA was diluted with 486 380 µl of water with 5 µl used per reaction. Quantitative PCR was performed using SYBR 487 green and Platinum Taq Polymerase (Invitrogen; 10966026). For qPCR 2-4 technical 488 replicates were used with at least 3 biological replicates. Reactions were carried out on the 489 Corbett thermocycler using the following conditions: 94°C for 10 min; 95°C for 20 sec, 58°C 490 for 20 sec, 72°C for 20 sec, and 45 cycles. Gene transcripts were normalized to the control 491 gene At4g24610. Primer sequences can be found in Supplemental Table S6. 492

Detached leaf senescence 493
Leaves were collected at either 5 weeks (35-38 DAS) or after 7 weeks (51-53 DAS) from 494 plants grown under short day conditions. For 5-week-old plants all the leaves from a rosette 495 were detached and placed in a 350 mm x 250 mm square petri dish containing 2 pieces of 496 Whatman paper. 35 ml of sterilized water was added to each petri dish. Detached leaves 497 were placed in the petri dish and wrapped in foil and kept at room temperature for several 498 days. Photos and trayscan images were taken each day at the same time until 12 days of 499 dark treatment. For 7-week-old plants leaves 13 and 14 were collected and placed in dishes 500 as described above. 501

Developmental senescence 502
Plants were grown on Gamborgs media plates (see above) until 14 DAS when plants were Assigned classifications of 'partial senescence' were considered when yellowing was 507 present over 20%-50% of the leaf, 'full senescence' was deemed when the whole leaf was 508 yellow. For analyzing the differences in senescence and leaf longevity of leaf 5/6 of intact 509 rosettes, plants were propagated as above. Once in soil, plastic tags were positioned under 510 leaves 5/6 to avoid premature senescence triggers due to contact with the soil. To avoid 511 shading induced senescence triggers from overlapping leaves and to expose leaves 5/6 for 512 over-head imaging, later developing leaves were kept aside using wooden toothpicks. 513

Fluorescence imaging 514
Chlorophyll distribution and F v /F m ratio, reflecting the maximal quantum yield of PSII 515 photochemistry, were measured using the Photon System Instruments (PSI; www.psi.cz) 516 PlantScreen system. Measured basal fluorescence (F 0 ) and maximum fluorescence (F m ) 517 and calculated F v /F m ((F m -F 0 )/F m ) values from images were derived using PSI Flurocam 7 518 (version 1.5.0.46). Images of leaves from intact rosettes were taken 1 hr pre-dawn to ensure 519 leaves were fully dark adapted. Leaf growth was defined as the chlorophyll area produced 520 using the PSI Flurocam 7. Healthy pre-senescing leaves had F v /F m values of ≥0.83 521 indicating setting for the light pulses were correctly calibrated. All leaves were imaged from 522 their adaxial side. False color images of F m and F v /F m allowed the visualization of the spatio-523 temporal progression of senescence in the leaves by observing the reduction in fluorescent 524 area. A 24 color card (www.cameratrax.com) was present in each imaging batch and used 525 to color standardize all RGB images using an in-house script. 526 Hormone Analysis 527 21 DAS plants grown under long day conditions were collected and frozen in liquid nitrogen. 528 7 replicates were prepared for C24, C24 NahG 3-3, Ler, Ler NahG and C24♀ x Ler♂. Each 529 replicate contained two plants. Material was ground in liquid nitrogen and 100 mg of the 530 ground tissue was used for the subsequent analysis. Hormone samples were extracted as 531 previously described (Xu et al, 2016). Hormone samples and standards (5 μl) were injected 532 onto an Agilent Zorbax Eclipse 1.8 μm XDB-C18 2.1 × 50 mm column. Solvent A consisted 533 of 0.1% aqueous formic acid (v/v) and solvent B, methanol with 0.1% formic acid (v/v). The 534 plant hormones were eluted with a linear gradient from 10% to 50% solvent B over 8 min, 535 50% to 70% solvent B from 8 min to 12 min (then held at 70% from 12 min to 20 min) at a 536 flow rate of 200 μl/min. Solvents were LC-MS grade from Fisher Chemical. The eluted 537 hormones from the column were introduced to the mass spectrometer via a heated 538 electrospray ionization (HESI-II) probe and analyzed using Q-Exactive Plus (Thermo 539 Scientific, Waltham, MA, USA). The HESI negative ion polarity parameters were as follow: 540 The electrospray voltage was 2.5kV, and the ion transfer tube temperature was 250°C. The 541 vaporize temperature and the S-lens RF level were 300 o C and 50 V, respectively. The 542 sheath gas flow was 45 litres min -1 of nitrogen, 10 litres min -1 of auxiliary gas and 2 litres min -543 1 of sweep gas, respectively. Targeted parallel reaction monitoring (PRM) was acquired in 544 the quadrupole-Orbitrap mass spectrometer over the mass range m/z 100 -1500 with a 545 mass resolution of 17500 at 1.0 microscan. Supplemental Table S7 shows the tandem mass 546 spectrometry acquisition parameters. The AGC (Automatic Gain Control) target value was 547 set at 1.0E+05 counts, maximum accumulation time was 50 ms and the isolation window 548 was set at m/z 4.0. Data were acquired and analyzed using the Thermo Scientific Xcalibur 549 4.0 software. 550