pH-dependent causes of N2O emissions and nitrite accumulation in soil

13 Denitrifier community phenotypes often result in transient accumulation of denitrification 14 (NO3 →NO2 →NO→N2O→N2) intermediates. Consequently, anoxic spells drive NO-, N2O15 and possibly HONO-emissions to the atmosphere, affecting both climate and tropospheric 16 chemistry. Soil pH is a key controller of intermediate levels, and while there is a clear 17 negative correlation between pH and emission of N2O, NO2 concentrations instead increase 18 with pH. These divergent trends are probably a combination of direct effects of pH on the 19 expression/activity of denitrification enzymes, and an indirect effect via altered community 20 composition. This was studied by analyzing metagenomics/transcriptomics and phenomics of 21 two soil denitrifier communities, one of pH 3.8 (Soil3.8) and the other 6.8 (Soil6.8). Soil3.8 22 had severely delayed N2O reduction despite early transcription of nosZ, encoding N2O 23 reductase, by diverse denitrifiers, and of several nosZ accessory genes. This lends support to a 24 post-transcriptional, pH-dependent mechanism acting on the NosZ apo-protein or on enzymes 25 involved in its maturation. Metagenome/metatranscriptome reads of nosZ were almost 26 exclusively clade I in Soil3.8 while clade II dominated in Soil6.8. Reads of genes and 27 transcripts for NO2 -reductase were dominated by nirK over nirS in both soils, while qPCR28 based determinations showed the opposite, demonstrating that standard primer pairs only 29 capture a fraction of the nirK community. The -omics results suggested that low NO2 30 concentrations in acidic soils, often ascribed to abiotic degradation, are primarily due to 31 enzymatic activity. The NO reductase gene qnor was strongly expressed in Soil3.8, 32 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint

had severely delayed N2O reduction despite early transcription of nosZ, encoding N2O 23 reductase, by diverse denitrifiers, and of several nosZ accessory genes. This lends support to a 24 post-transcriptional, pH-dependent mechanism acting on the NosZ apo-protein or on enzymes 25 involved in its maturation. Metagenome/metatranscriptome reads of nosZ were almost 26 exclusively clade I in Soil3.8 while clade II dominated in Soil6.8. Reads of genes and 27 transcripts for NO2 --reductase were dominated by nirK over nirS in both soils, while qPCR- 28 based determinations showed the opposite, demonstrating that standard primer pairs only 29 capture a fraction of the nirK community. The -omics results suggested that low NO2 -30 concentrations in acidic soils, often ascribed to abiotic degradation, are primarily due to 31 enzymatic activity. The NO reductase gene qnor was strongly expressed in Soil3.8, 32 Introduction lead to more or less HONO emission than from non-NO2accumulating, acidic soil. Secondly, 126 we investigated if the apparent lack of DNRA activity (dissimilatory nitrite reduction to 127 ammonium) in these soils was due to low abundance of DNRA-related genes and transcripts. 128 Thirdly, we clarified the complex ecophysiology of nosZ carrying bacteria to better 129 understand their hampered N2O reduction under acidic conditions (this study and [38,39]). 130 To do so, we investigated to what extent the two nosZ clades were found in the MGs and MTs transcribed in acidic soil. For the latter, we included nosR, which encodes NosR suggested to 134 be involved in electron delivery to NosZ in organisms with nosZ cladeI; nosL, which encodes 135 a chaperone delivering Cu to the NosZ apo-protein (both nosZ clades); and the ORF nosDFY 136 (both nosZ clades) encoding NosD, suggested to be involved in NosZ maturation, and the 137 ABC-transporter NosFY [44][45][46].  Nitrate (as KNO3) was dissolved in autoclaved MilliQ water which was added to reach 80% 151 of the soil's water holding capacity (WHC) and 6.2-7.1 mM NO3in soil moisture. Thus, at 152 the onset of the incubation, the total amount of NO3per vial was 37 or 26 μmol NO3in 153 Soil3.8 or Soil6.8, respectively (see also [28]. The vials were immediately made anoxic by 154 6 cycles of gas evacuation and He filling [38], and incubated at 15 °C. Gases (CO2, O2, NO, 155 N2O and N2) were measured in headspace every three hours using an autosampler linked to a 156 GC and NO analyser [48]. At each gas sampling time point, one replicate vial of each soil Nucleic acid extraction. DNA and RNA were extracted from frozen samples using the 166 method of [43] who also tested several commercial kits for these soils without success. 167 Briefly, 3 x 0.2 g of soil was taken at time 0 (at the start of anoxic incubation) for DNA 168 extraction, and at selected time points (0.5-27 h) during anoxic incubation for RNA 169 extraction. Lysis was performed with glass beads in CTAB extraction buffer and phenol-170 chloroform-isoamyl alcohol (25:24:1), using a FastPrep-24 instrument. After ethanol 171 precipitation, the nucleic acids were resuspended in DEPC-treated nuclease-free water 172 purified with the OneStep PCR Inhibitor Removal Kit (Zymo Research, Irvine, USA), then 173 split into a fraction for DNA and one for RNA. The DNA fraction was further purified using 174 the Genomic DNA Clean & Concentrator kit (Zymo Research), then kept at -20 °C until use. 175 The RNA fraction was digested using TURBO DNA-free DNase kit (Ambion, Life 176 Technologies) according to the manufacturer's instructions, then purified using the RNA 177 Clean & Concentrator-5 kit (Zymo Research). Quantitative PCR (qPCR) using primers 178 targeting the 16S rRNA gene (described below) was used to assess the presence of residual 179 genomic DNA (gDNA) in the purified RNA fractions (defined by signal detected in the qPCR 180 at ≤ 35 cycles), and only RNA fractions free of gDNA was used for further analysis. The 181 purified and DNA-free RNA fractions were reverse transcribed using the Maxima Reverse 182 Transcriptase with random hexamer primers (Thermo Scientific), according to the 183 manufacturer's instructions. Primers targeting the 16S rRNA or nosZ genes (described below) 184 were used in qPCR to assess the quality (defined by uninhibited amplifiability) of purified 185 DNA and reverse-transcribed cDNA. 186 Sequencing the metagenome (MG), metatranscriptome (MT), and 16S rRNA genes. 187 Triplicate DNA and duplicate RNA samples were sent for metagenomic and 188 metatranscriptomic sequencing at The Roy J. Carver Biotechnology Center (CBC)/W. M. 189 Keck Center for Comparative and Functional Genomics at the University of Illinois at 190 Urbana-Champaign, using HiSeq 2500 technology. All nucleic acids were shipped in a liquid 191 nitrogen vapour dry shipper (Cryoport) and arrived within 5 days (the Cryoport Express 192 dewar is able to maintain the temperature at -150 °C during shipment for 10 days Reads derived from specific genes and meeting the assigned quality cutoffs were extracted 229 from read sets using filterbyname from the BBTools suite of programs. The extracted reads 230 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in products, the fluorescence signal was measured during the final step of each cycle, at 82 °C. 249 The detection limit of each qPCR run was 5 copies per microliter of reaction [43], which was 250 approximately 4 × 10 2 copies g -1 soil (ww).

253
Kinetics of denitrification intermediates depict a pH-dependent response to anoxia. 254 The denitrification kinetics of the two soils during 45 h of anoxic incubation are shown in Fig.   255 1, in which the sampling occasions for MT analyses are also indicated. A more complete 256 description of the incubation experiment is given by [28], including detailed analyses of 257 production/reduction rates of the denitrification intermediates/end products over 70 h. The 258 analysis included a careful mineral N budget analysis demonstrating 100% recovery of NO3 --259 N as N2 for the soil with pH=6.8, which suggests negligible reduction of nitrate to ammonium 260 in this soil. For Soil 3.8, the recovery as N-gas (N2 + N2O +NO) was lower (77%), but abiotic 261 nitrosylation of organic material accounted for 17% (see Table 2 in [28], thus in total 94% 262 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint was accounted for, leaving only 6% of the NO3 --loss that could possibly be ascribed to 263 DNRA. 264 The two soils showed striking differences in their accumulation of NO2and N2O, while 265 they had very similar NO kinetics (Fig. 1A). The NO2concentration in Soil3.8 was 20-50 µM 266 in the soil moisture during the entire anoxic incubation, except for the 36-40 h period when it 267 reached ~100 µM. This corroborates the general notion that NO2concentrations are low in 268 acidic soils. In Soil6.8, on the other hand, NO2steadily increased from the beginning, Soil bacterial community composition differed by pH but was stable during incubation. 288 In the 16S rRNA gene amplicon analysis >99.29 % of all sequenced reads were annotated as 289 bacterial, about 0.004 % were unclassified, and the rest belonged to Archaea, which 290 represented ≤0.60 % of the reads in Soil6.8 and ≤1.03 % in Soil3.8. Principal component 291 analysis of the bacterial 16S rRNA gene reads (Fig. S1A) clearly separated the reads from the 292 two soils along PC1, which explained 94% of the total variation and also showed that the 293 variation between replicate samples was low. The most abundant classified phyla in both soils 294 were Proteobacteria, Acidobacteria, Actinobacteria, Planctomycetes, Verrucomicrobia and 295 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint Bacteroidetes. A breakdown of Proteobacteria showed that Alphaproteobacteria were most 296 abundant in both soils, followed by Beta-, Gamma-and Deltaproteobacteria (Fig. S1B) Since the two soils accumulated different amounts of denitrification intermediates, we 325 calculated ratios of MG reads representing the four steps of denitrification (Table 1), to 326 examine if the genetic potential for production/consumption of the different intermediates was 327 related to the net production seen in Fig. 1. The NAR/NIR ratios were similar in the two soils 328 (4.68±0.27 for Soil3.8 and 4.32±0.12 for Soil6.8), which, by itself, is not consistent with the 329 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint higher net production of NO2in Soil6.8. The NOR genes were more abundant than NIR in 330 Soil3.8, with a NIR/NOR ratio of 0.49 ±0.03 compared to 0.92 ± 0.03 in Soil6.8. NIR and 331 NOR were more abundant than NOS in both soils with NIR/NOS ratios of 2.83 and 2.07 and 332 NOR/NOS ratios of 5.81 and 2.23 in Soil3.8 and Soil6.8, respectively. The two nosZ clades (I 333 and II) differed in gene abundance in the two soils with higher values for clade I in Soil3.8 334 and vice versa in Soil6.8 (Fig. 3A). The nosZ clade I/clade II ratio (based on RPM values) 335 was 28.1 in Soil3.8 and 0.25 in Soil6.8 (Table 1). 336 We also analysed accessory genes of the nos operon, with a focus on clade I, using a  (Fig. 4), which is in accordance with the comparable abundance of NOS (nosZ clade 346 I+II) in the two soils (12.5 RPM in Soil3.8 and 20.8 in Soil6.8, Table 1). The gene reads for 347 nosF, which is also part of the nos operon both in clade I and II [45], were 10-45 times higher 348 than for the other accessory genes which points to uncertainties in the databases for this gene. 349 In addition to the canonical denitrification genes, we examined reads from the two  It could be argued that the ratios of transcripts representing the reductases responsible for 380 the production and consumption of the various denitrification intermediates are more suitable 381 to use for comparison with phenotypic data than actual RPM values (transcript read 382 abundances and selected ratios are given in Table 1 CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint same time, this ratio for Soil6.8 was 0.6±0.2. These ratios reflect the higher number of NOR 397 reads in Soil3.8, combined with higher reads for NOS in Soil6.8. 398 Transcript reads of nos accessory genes were detected in both soils at all sampling 399 occasions (Fig. 4). The transcriptional activity increased for all these genes between 0.5 and 3 400 h, which is a strong indication that the lack of N2O reduction in Soil3.8 was not caused by any 401 transcriptional control mechanism. The nosR transcript abundances were comparable in the 402 two soils. 403 We also examined the MT with regard to the two DNRA genes nfrA and nirB, expecting 404 low read abundance since no DNRA activity was discerned from the gas analyses. 405 Surprisingly, these two genes were transcribed at levels comparable to the NIR and NOR 406 genes, with sometimes high read values for nirB (Fig. S2).  Table 1) were compared to qPCR-results in samples from the 411 same soil incubation (Fig. 5), although more time points were included in the qPCR analysis. 412 We targeted nirK, nirS and nosZ clade I using standard primer pairs (see Materials and corresponding transcript had a unique taxonomic profile that varied by soil pH (Fig. 6).

430
. CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint Proteobacteria, Actinobacteria and Bacteroidetes were the most abundant phyla of denitrifiers 431 in the MG and MT in both soils. Reads assigned to these phyla were detected for most 432 denitrification genes and transcripts with the exceptions of Bacteroidetes, which were not 433 found among narG and nirS reads, and Actinobacteria, which were not found among the nirS, 434 cnor and nosZ clades I and II reads. Several other phyla such as Firmicutes, Chlamydiae, 435 Nitrospira, Spirochaetes and Verrucomicrobia were represented by reads only from one or a 436 few genes/transcripts. 437 The MG reads of narG, the most abundant among the denitrification genes, were 438 dominated by Proteobacteria, mostly the classes Alpha-and Betaproteobacteria, and 439 Actinobacteria (Table S1). Second most abundant in both soils were napA reads, which were 440 mainly derived from Proteobacteria. Organisms belonging to these phyla also dominated the 441 transcriptional activity for narG and napA genes. narG reads from Nitrospira were high as 442 well in Soil6.8. MG reads derived from the NO2reduction gene nirK, which was far more 443 abundant than nirS in both soils (Table 1) (Table S1). Similar as for the nirK and nirS genes, the qnor genes dominated strongly 451 over cnor genes in the MG, and the qnor genes belonged to several phyla while the cnor reads 452 were mostly from Proteobacteria. The Delta-and Gammaproteobacteria had the highest 453 number of qnor MG reads in both soils, but it was the Betaproteobacteria that dominated 454 among the transcripts showing comparatively high and immediate transcription in Soil3.8. 455 Interestingly, qnor MG and MT reads from Acidobacteria and Planctomycetes were detected 456 in both soils at relatively high abundance. Apart from a few acidobacterial nosZ clade II 457 reads, reads from these two phyla were not found for the other denitrification genes in the MG 458 or MT. 459 The total number of nosZ MG reads was higher in Soil6.8 than in Soil3.8 (212 vs 152, 460 Table S1). As expected, nosZ clade I MG and MT reads were only from Proteobacteria in 461 both soils. nosZ clade II MG reads were more diverse, especially in Soil6.8, and were 462 comprised of not only Proteobacteria, mostly Deltaproteobacteria, but also several other phyla 463 including Bacterioidetes, which was the most highly represented group, as well as Firmicutes, 464 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in also seen in the present study, where the PCA of the 16S rRNA gene sequences clearly 480 separated the two soils (Fig. S1A). The sequence analysis also revealed a slight but significant The accumulation of NO2and N2O in the two soils shows contrasting patterns, with low 490 NO2and high N2O levels in Soil3.8 and vice versa for Soil6.8. The pattern of NO 491 accumulation was more similar in these two soils, as presented in [28]. In the present study 492 we investigated if the contrasting denitrification phenotypes of the two soils could be 493 predicted from the abundance and transcription of the denitrification genes. Net production of 494 denitrification intermediates will take place as soon as the reduction rate of one of the 495 denitrification steps surpasses that of the following step. The NO3reduction rate (VNAR) in 496 Soil6.8 grossly exceeded the NO2reduction rate (VNIR), as long as nitrate was present (Fig.   497 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint 1B and [28]). The MT analysis showed higher transcription of NAR (napA+narG) than of 498 NIR (nirK+nirS) (Fig. 3B) with NAR/NIR ratios between 1.6 and 3.4 (Table 1). The ratio was 499 similarly high for Soil3.8, however (2.0 and 2.3 for the two time points measured in this soil), 500 despite the near-absence of NO2accumulation in this soil. Thus, no direct link was found 501 between the transcript ratio NAR/NIR and NO2accumulation or the VNAR/VNIR ratio, as 502 affected by pH. In theory, abiotic NO2decomposition could be the primary reason for the 503 marginal transient accumulation of NO2in Soil3.8, but this was refuted by the careful 504 analyses of Lim et al. [28], who reached the conclusion that enzymatic NO2reduction was 505 the major sink for NO2in Soil 3.8. One explanation for the marginal NO2accumulation in 506 Soil3.8 could be that NO2reductase has a low pH optimum, as shown in a study by Abraham 507 et al. [72] where measured NirK activity in vitro was ~4 times higher at pH 4.2 than at pH 7. 508 Thus, Soil3.8 could have a much lower VNAR/VNIR ratio than Soil6.8, despite the nearly equal 509 NAR/NIR-ratio for the two soils. The results demonstrate that transcript numbers or ratios of 510 transcript numbers are poor predictors of metabolic activities in soils. Even more evident is 511 the discrepancy between gene numbers and activity. The NAR/NIR ratio in the metagenome 512 was almost identical in the two soils (Table 1) CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint According to the MG analysis the abundance of NIR genes was similar in the two soils, 532 with strong dominance of nirK over nirS, especially in Soil3.8 where nirS genes were almost 533 absent (Table 1; Fig 3A;). This contradicts the general conception that nirS is more abundant 534 in most environments, which is derived from primer-based studies [38,76,77]. However, 535 while nirS is mainly found in Proteobactera, nirK is spread among taxonomically diverse 536 groups, many of them being non-proteobacterial denitrifiers [16,17,78], which was also 537 found in the present study (Fig. 6). Transcripts of nirK dominated over nirS also in the MT 538 (Fig. 3B), and represented a diverse range of phyla including, in addition to Proteobacteria, 539 also Actinobacteria, Bacteroidetes and Firmicutes (Fig. 6). This is in accordance with the 540 results from a non-primer based study of another agricultural soil by [77]. conditions. It is likely that this applies in particular to the functional genes nirK and qnor, 549 which are probably more easily transferred horizontally than the other denitrification genes, 550 and it can be speculated that they comprise a pool of genes that circulate between organisms 551 depending on their needs. 552 The complete recovery of all added NO3as N2 in Soil6.8 (Fig. 1A) indicated minimal or 553 no conversion of NO2to NH4 + and thus that the NO2produced from NAR activity was not 554 used by DNRA organisms. This is surprising, taking into account the relatively high 555 abundance of reads from nrfA and nirB genes and transcripts (Fig. S2). Likewise, DNRA was 556 probably insignificant in Soil3.8, since 94 % of the NO3loss in this soil could be accounted 557 for by N-gas + nitrosylation (see [28]. In theory, full recovery of NO3reduction as N-gas 558 could be obtained in a system with equal rates of DNRA and anammox. However, members 559 of Brocardiales, which comprises the anammox-Planctomycetes [80], were scarce in the 16S 560 rRNA gene analysis (<0.008% of 16S reads in all SoilA samples), which lends little support 561 to this hypothesis. 562 The net production of NO was similar in the two soils despite the substantial abiotic 563 reduction of NO2that contributed to NO emissions from Soil3.8 [28], in addition to the 564 enzymatic NO2reduction taking place in both soils. The strong control of NO in Soil3.8, seen 565 . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint from the calculated VNOR activity (Fig. 1B), is in agreement with the 2.5 times increase in 566 qnor transcript abundance between 0.5 and 3 h (Fig. 2B) and NIR/NOR ratios <1 (Table 1). In 567 line with this, the gene abundance of qnor was also high in Soil3.8, almost 2 times its 568 abundance in Soil6.8 (Fig. 3A). This corroborates the metagenome results by Roco et al. [60] 569 from acidic soils showing dominance of qnor over all other denitrification genes, suggesting 570 that low pH selects for this gene. Interestingly, cnor apparently played a minor role in 571 controlling NO, especially in Soil3.8 where gene abundance was close to 0 and transcripts 572 were almost undetectable (Fig. 3). The taxonomic analysis of qnor genes and transcripts (Fig.   573 6) demonstrated that this gene, similarly as nirK, is widely distributed over different phyla, 574 with the largest representation in the Proteobacteria, Actinobacteria, Bacteroidetes, 575 Chlamydiae, Acidobacteria and Planctomycetes. qNor is the product of a single structural 576 gene, norB, existing alone or in a small operon [81], and it is conceivable that this gene/gene 577 cluster is more readily transferred horizontally than the bigger cnor operon. Furthermore, 578 recent evidence suggests that qNor is electrogenic [82], as opposed to cNor, and it can 579 therefore be speculated that it provides an extra, energetic advantage to its host organism. 580 Taken together, it is not unreasonable to conclude that, between the two functionally 581 redundant NO reductases related to denitrification, it is qNor that plays the major role in soils with neutral and slightly alkaline pH generally emitted more HONO than acidic soils. 593 Our results show more than 10 times higher HONO production from Soil3.8 compared to 594 nearly complete lack of N2O reduction in Soil3.8 during the first 35 h of incubation (Fig. 1A).
This severely delayed N2O reduction in acidic soil corroborates other studies of soils from the 601 same site and is in line with the growing evidence for a strong negative correlation between 602 soil pH and N2O/(N2O+N2) product ratios [12,33], suggested to be due to impaired maturation 603 of the NosZ enzyme under acidic conditions (pH<6.1) [37-39]. The present study detected 604 transcripts from both nosZ clades in Soil3.8, which suggests that the problem of producing 605 functional NosZ under low pH conditions applies to both clades. Moreover, the taxonomic 606 analysis showed that this problem is general to a diverse range of bacteria (Fig. 6), which adds CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint has also been suggested to be involved maturation of the CuZ site [45,86,87]. Taken 634 together, gene reads for all the accessory nos genes analyzed in this study were detected in 635 both soils, which was not unexpected. 636 Transcript reads encoding all accessory nos genes of clade I were detected in Soil3.8, 637 which could be taken to indicate that the organisms in this soil had the tools in place for NosZ 638 function, including nosZ transcriptional activation and electron transfer to NosZ (by NosR) 639 and CuZ site maturation (by NosL). To conclude, there were no obvious issues with the 640 genetic potential or the transcriptional activity of the nos operon which could explain the 641 delayed N2O reduction in Soil3.8. CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint The analyses of the MG and MT demonstrated that diverse microorganisms transcribed 667 nosZ genes from both clade I and II in the acidic soil, but this was not followed by any 668 detectable Nos function until >30 h after the start of the incubation. The results did not reveal 669 any obvious lack of genes or transcripts of nos accessory genes involved in NosZ maturation. 670 Although not exhaustive, this information provides a new building block towards an 671 understanding of the impaired N2O reduction under acidic conditions. 672 The discrepancies between qPCR and -omics based estimates of genes and transcripts 673 provides strong evidence that primers commonly used in denitrification studies only capture a 674 fraction of the community that carries these genes. The problem likely arises since genes such CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; Figure 1. Kinetics of NO2and N-gases, calculated enzyme rates, and kinetics of NO2and HNO2 during anoxic incubation of the soil with pHCaCl2 = 3.8 and 6.8, amended with NO3 -. Panel A shows the measured amounts of NO2 -, NO, N2O and N2. Sampling for metatranscriptome analyses are indicated by red arrows (0.5 and 3 h for Soil3.8; 0.5, 3, 9, 12 and 27 h for Soil6.8). Panel B shows the reduction rates for the different denitrification steps VNAR (NO3 -→ NO2 -), VNIR (NO2 -→NO), VNOR (NO→N2O) and VNOS (N2O→N2), all given as µmol N vial -1 h -1 . The values were based on the measured kinetics of NO2 -, NO, N2O and N2 and corrected for abiotic decomposition of NO2as published previously by Lim et al. (2018). Abiotic NO2decomposition was significant only in Soil3.8. The figure is adapted from graphs shown in Lim et al., 2018, based on the same data set.
. CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ;  . CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint Figure 5. Amplification-based quantification (qPCR) of genes and transcripts (copies g -1 soil, wet weight). (A) Quantification of gene copies. Primers targeting the 16S rRNA (27F/518R), nirK (517F/1055R), nirS (cd3aF/R3cd) and nosZ cladeI (nosZF/1622R) genes were able to detect gene copies in both soils (B) mRNA transcripts of nirK, nirS, and nosZ. Red= Soil3.8; Blue= Soil6.8.
. CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint Figure 6. Denitrification gene and transcript prevalence as summed reads across replicate samples (n=3 for MG; N=2 for MT). Annotated genes (MG; DNA sampled at start of incubation) and transcripts (MT) sampled after 0.5 and 3 h anoxic incubation (Soil3.8) and 0.5-27 h anoxic incubation (Soil6.8) (for sampling points also see Fig. 1A).
. CC-BY-NC-ND 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted November 26, 2020. ; https://doi.org/10.1101/2020.11.26.399899 doi: bioRxiv preprint Table 1. Abundances of gene and transcript reads in the metagenome (DNA) and metatranscriptome (RNA) of two soils with pH 3.8 and 6.8 during anoxic incubation with NO3 -. NAR=nitrate reductases, sum of narG and napA abundancies; NIR=nitrites reductase, sum of nirK and nirS abundancies; NOR = nitric oxide reductases, sum of qnor and cnor abundancies; NOS= nitrous oxide reductases; sum of clade I and clade II abundancies. DNA samples were taken just before the start of the incubation (t=0 h, Fig. 1A). RNA samples were taken after 0.5 and 3 h incubation (Soil 3.8) and after 0.5-27 h incubation (Soil 6.8) (Fig.  1A). All values are in reads per million (RPM). Ratios of reads from selected genes/transcripts are also given. DNA: Average values are given for abundance of individual genes, values in parenthesis are standard deviation; n=3. RNA: Duplicate metatranscriptome samples were analyzed from each sampling time except for the 9 h sample from Soil6.8 for which one samples was lost. Individual values for read abundances and ratios are shown except for the nosZ clades for which data from duplicate samples were pooled. Ratios including NOS are therefore calculated from average values. na= not applicable.  (2018). For this examination reads from duplicate samples in the MT were pooled before the analysis, thus replicates are not reported. Calculations of ratios were based on the same database as for the other denitrification genes, which was based on individual replicates. For detailed descriptions see Materials and Methods section.