Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2025 Jun 4;59(23):11653–11665. doi: 10.1021/acs.est.4c05624

Transcriptional Difference of Deep-Sea Microorganisms under Different Sampling Methods

Yinghui He †, Hongxi Zhang ‡, Federico Baltar §, Yong Wang †,∥,*
PMCID: PMC12178249  PMID: 40465356

Abstract

The metabolic potential and activity of deep-sea microbes have not been fully explored by meta-transcriptomics using the samples obtained by different sampling methods. Here, we report active deep-sea microbes obtained by the methods of multiple in situ nucleic acid collection (MISNAC), in situ microbial filtration and fixation (ISMIFF), in situ microbial filtration without fixation (ISMIFU), and the Niskin bottle at a 1038 m depth in the South China Sea. Higher biodiversity and different dominant active microbial taxa in the metatranscriptomes were detected in the MISNAC and ISMIFF samples compared with the other two approaches. The transcriptional profiles of 40 conserved genes were similar between the MISNAC and ISMIFF samples, while the expression of a quarter of these genes was not detected in the ISMIFU sample. Genes related to the CO oxidation and nitrification processes were highly transcribed in the MISNAC and ISMIFF transcriptomes, whereas those for chemotaxis and low-oxygen adaptation were highly transcribed in the Niskin samples. Overall, our result highlights the importance of in situ sampling and preservation for more precise quantification of the ecological function of active deep-sea microbiomes.

Keywords: MISNAC, Niskin, in situ, metatranscriptome, two-component systems

graphic file with name es4c05624_0009.jpg

graphic file with name es4c05624_0007.jpg

Introduction

Microbes are major players involved in element cycles in the ocean. The majority of marine microorganisms, particularly those found in the deep sea, are difficult to cultivate in the laboratory. It is commonly observed that some microorganisms indigenous to the deep sea demonstrate optimal growth under high-pressure regimes.

Rapid decompression may adversely affect the morphology and enzymic inactivation of a piezophilic bacterial cell through various means, which renders further difficulties in the recovery of the deep-sea microbes for cultivation. Nevertheless, high-throughput sequencing technologies and metagenomics have been employed to understand community structure, metabolic potential, and phylogenomics of the deep-sea microorganisms. , Furthermore, deep-sea metatranscriptomics is critical for deciphering the activities and adaptation schemes of deep-sea microbes as well; however, environmental changes during sampling may result in the alteration of the gene expression profile, which prohibits insights into the real ecological roles of the microbial inhabitants. Prokaryotes can quickly shift their gene expression to cope with complex environmental fluctuations. As such, the varying conditions during recovery of a Niskin bottle to the ocean surface can lead to degradation of the in situ expressed RNA molecules and stimulation of a new set of gene transcripts under physicochemical differences and rapid growth of a subset of microbes in the sampling bottle. −

Different in situ sampling methods have been developed for high-quality RNA sampling from the deep ocean. An automatic flow injection sampler (AFIS) with in situ fixation revealed 30-fold higher abundance of thaumarchaeal ammonia monooxygenase gene transcripts, compared with a traditional water sampler. , In contrast, the transcription of the cellular stress resistance marker gene recA exhibited higher abundance in nonfixed samples relative to fixed ones at sampling depth of 120 m. Transcriptional profiles of certain taxa obtained using an in situ incubation device (MS-SID) that is able to collect and preserve samples in situ showed dramatically different transcriptional patterns compared to a Niskin sampler. Specifically, the genes involved in the category of post-translational modification, protein turnover, and chaperone proteins were found to have higher expression levels in the Niskin samples for Gammaproteobacteria and protists, while these genes exhibited lower expression levels in the Niskin samples for fungi. The alterations in pressure have a remarkable impact on the community structure and activity of microbes in the deep sea. Therefore, capturing the gene expression profile of the active microbes with a great contribution to the deep-sea biogeochemical cycles requires new sampling tools for the collection and preservation of high-quality in situ samples.

In our previous study, we developed the in situ microbial filtration and fixation (ISMIFF) apparatus and applied it in a microbial study of the Challenger Deep. With the capacity of in situ filtration and fixation, ISMIFF might avoid introduction of contaminant microbes from the ocean’s upper layers. Moreover, the most abundant prokaryotic taxa in our ISMIFF samples of the hadal water (>6000 m depth) were Marinimicrobia (SAR406), Alphaproteobacteria, Thaumarchaeota, and Chloroflexi, which were distinct from those detected in the Niskin samples. To further recover the RNA molecules indicative of in situ gene expression, we invented an automatic multiple in situ nucleic acid collection (MISNAC) with nine working units. The device was applied for in situ microbial filtration, cell lysis, and nucleic acid collection at different time intervals. Given the similarity in the microbial community, it remains to be elucidated whether in situ cell lysis and RNA extraction using MISNAC make a notable difference from in situ fixed/unfixed ISMIFF samples at the metatranscriptomic level. Here we report a comparison of metatranscriptomes for mesopelagic water column samples collected in the South China Sea (SCS) using Niskin bottles, ISMIFF, and MISNAC. The community structures of active microbes displayed similar compositions between the MISNAC and ISMIFF metatranscriptomes, as well as their gene transcriptional profiles. Some gene categories that are sensitive to environmental perturbations exhibited elevated activity in the samples collected via the Niskin bottle, compared to the MISNAC and ISMIFF samples. Most of the conserved genes in the transcriptome for the ISMIFF sample without fixation have been found to lack transcriptional activity. These results indicate that in situ sampling devices like ISMIFF and MISNAC are critical for preserving the samples for the studies of deep-sea microbial communities and their transcriptomic states with minimal disturbance from sample collection and handling.

Results

Distinct Community Structures of Active Microbes Revealed by Metatranscriptomes

Two Niskin bottles and MISNAC were mounted on a deep-sea lander to collect samples at a 1038 m depth in SCS (Table S1). We obtained six samples using different subunits of MISNAC: six absorption columns with in situ coextracted DNA/RNA (MISNAC1-6) (Figure S1A,B,E). An in situ microbial filtration and fixation (ISMIFF) sample and an in situ microbial filtration sample without RNALater fixation (ISMIFU) were obtained by the ISMIFF devices (Figure S1C,D,F). Two 10 L water samples (Niskin1 and Niskin2) were collected in Niskin bottles. The membrane filters with a pore size of 0.22 μm were used for the filtration of all of the samples. Co-extraction of DNA/RNA was performed for the ISMIFF, ISMIFU, and Niskin samples with the same protocol. Metagenomes and metatranscriptomes were then obtained for the 10 individual samples (Table S2).

Using 16S rRNA gene Illumina reads (16S miTags) in the metatranscriptomics data, we revealed that the community structures of active prokaryotes in the Niskin, ISMIFU, ISMIFF, and MISNAC samples differed among certain families (Figure A and Table S3). Saccharospirillaceae, Alteromonadaceae, Spongiibacteraceae, and Marinobacteraceae demonstrated higher relative abundances in the Niskin metatranscriptomes, which were averaged to be 11.89%, 9.53%, 8.25%, and 3.36% of the 16S miTags, respectively. The taxonomic assignment of the metagenomic reads also indicated a higher proportion of these microorganisms in the Niskin samples compared to the MISNAC samples (Table S4). The same taxa in the MISNAC samples showed relative abundances of 0.11%, 1.52%, 0.11%, and 0.03%, respectively. In the ISMIFF sample, the relative abundances of the corresponding taxa were found to be analogous to those observed in the MISNAC samples, with proportions of 0.17%, 1.00%, 0.36%, and 0.00%, respectively. In contrast, the MISNAC samples were remarkably enriched with 16S miTags sorted to Methylomonadaceae, Marinimicrobia, and Nitrosopumilaceae, accounting for a markedly higher proportion of 6.98%, 4.13%, and 3.42%, respectively. Relative abundances for these taxa in the Niskin samples were averaged to be 3.46%, 1.25%, and 0.19%, respectively. Notably, the relative abundance of Alteromonadaceae also differed greatly between the ISMIFF and ISMIFU samples. It constitutes 1.00% in the fixed sample ISMIFF but 28.23% in the unfixed sample ISMIFU. Principal coordinate analysis (PCoA) of community structures of the active prokaryotes further supports the difference in the compositions of active prokaryotes (Figure B). The ISMIFF sample clusters closely with the MISNAC samples in the PCoA plot, whereas the ISMIFU sample is distinctly separated.

1.

1

Open in a new tab

Relative abundance and diversity indices of active prokaryotes in the samples obtained by different methods based on 16S miTags extracted from metatranscriptomes. (A) The structures of prokaryotic communities at the family level in the samples gathered by different sampling methods (Table S3). (B) Principal coordinates analysis (PCoA) of prokaryotic community structures. (C) Alpha diversity indices of prokaryotic communities. MT: metatranscriptome; MISNAC1 to MISNAC6: the samples collected by MISNAC; ISMIFF: in situ microbial filtration and fixed sample; ISMIFU: in situ microbial filtration sample without fixation; Niskin1 and Niskin2: the samples collected by a Niskin bottle.

Analysis of alpha diversity metrics provided insights into the species richness and evenness of active prokaryotes between the sampling methods (Figure C). Our result showed that Chao1 was higher in the MISNAC and ISMIFF samples, compared to the ISMIFU and Niskin, suggesting higher species recovery by MISNAC and ISMIFF. With respect to the Shannon index, the compositional diversity of populations was higher in the MISNAC and ISMIFF samples than that of the others as well. Collectively, these alpha diversity indices demonstrated that there were more active prokaryotes in the samples collected by MISNAC and ISMIFF, compared with the unfixed ISMIFU and Niskin samples.

We investigated the community composition of active eukaryotic microbes across different samples using 18S rRNA gene Illumina reads (18S miTags) from the metatranscriptomics data, as well. The overwhelmingly dominant family identified in all the samples was Oligotrichia (Figure A and Table S5). The eukaryotic microbial communities in the MISNAC samples are similar to those of the ISMIFF sample. The MISNAC samples exhibited a more than 3-fold increase in the abundance of Choreotrichia, compared to the Niskin samples. Notably, the ISMIFU and Niskin samples displayed lower taxonomic diversity compared to the MISNAC samples. Several taxa with average relative abundance higher than 3% in the MISNAC samples, such as Choreotrichia, were almost absent in the Niskin samples (Figure A and Table S5). Additionally, a PCoA analysis revealed clear separation between ISMIFU and ISMIFF, and their large differences from the Niskin samples in the community structures of the active eukaryotes (Figure B). Furthermore, the Chao1 and Shannon indices indicated higher richness and evenness of active eukaryotic microorganisms in the MISNAC and ISMIFF samples compared to the Niskin and ISMIFU counterparts, suggesting a notable decrease in eukaryotic activity during the ISMIFU and particularly Niskin sampling (Figure C).

2.

2

Open in a new tab

Relative abundance and diversity indices of active eukaryotes in the examples obtained by different methods based on 18S miTags extracted from metatranscriptomes. (A) The structures of eukaryotic communities at the family level across different sampling methods (Table S5). (B) Principal coordinates analysis (PCoA) plot using eukaryotic community structures. (C) Alpha diversity indices for eukaryotic communities.

A total of 284 metagenome-assembled genomes (MAGs) were binned from the individual metagenomic assemblies (Table S6). A combination of the MISNAC MAGs was dereplicated and searched against the GTDB database to obtain average nucleotide identity (ANI) with their closest relatives, as well as that of the Niskin MAGs. A total of 46 and 9 MAGs for the MISNAC and Niskin samples, respectively, were associated with ANI below 95%, which is a threshold for a new species. For the MISNAC and ISMIFF metagenomes, 37.20% and 26.37% of the reads of the MISNAC and ISMIFF metagenomes were mapped on their MAGs, respectively. For the Niskin and ISMIFU metagenomes, the recruitment rates of the reads were 53.38% and 46.78%, respectively.

Gene Transcriptional Variations among the Samples Collected by Different Methods

We further investigated gene expression patterns in the samples. The gene annotation of the metagenomics data against the KEGG database allowed us to identify 6212 KEGG genes. Upon obtaining TPM values using the transcriptomic data for the genes, we conducted a correlation analysis of the TPM values among the samples. The analysis showed that the Niskin samples Niskin1 and Niskin2 clustered together, while the MISNAC samples were adjacent to the ISMIFF sample (Figure A).

3.

3

Open in a new tab

Transcriptomic variations among samples obtained by different sampling methods. (A) Heatmap with correlations and clusters of transcriptional levels (log 10 (TPM + 0.001)) of all the genes (n = 6212) among transcriptomic data for the different samples. (B) The bar plot illustrates the TPM variations for 40 single-copy genes across transcriptomic data sets obtained by MISNAC, ISMIFF, ISMIFU, and Niskin. (C) Heatmap showing the relationships between expression levels of the 40 single-copy genes. (D) Heatmap showing the changes in transcriptional levels, as indicated by TPM values, of 40 single-copy genes across the different samples.

To learn about the metabolic activity of the unfixed ISMIFU and Niskin samples, a total of 40 single-copy conserved genes , were used to estimate the overall transcriptional activity of active prokaryotic members. Analysis of these TPM values revealed significantly higher transcriptional levels of the Niskin samples (t-test, p < 0.05), compared to the MISNAC collections, indicating that some microorganisms might undergo rapid proliferation during the Niskin bottle recovery (Figure B,C). A correlation analysis showed highly similar levels of ISMIFF and MISNAC gene transcripts (Figure C). Furthermore, approximately a quarter of the 40 conserved genes were not transcribed in ISMIFU: i.e., those coding for FARSA, RP-L11, RP-L3, RP-L5, SARS, RP-S15, RP-S19, RP-S17, VARS, and SRP54 (Figure D and Table S7). We further confirmed zero TPM value for a majority of the ISMIFU genes (5699 out of 6212 genes).

To evaluate the contribution of the active deep-sea microbes to element cycles, we estimated the transcriptional activity of the genes involved in the Calvin cycle, CO oxidation, methane oxidation, nitrification, and dissimilatory sulfate reduction (Figure A and Table S8). The enzymes phosphoribulokinase and 1,5-bisphosphate carboxylase/oxygenase encoded by rbcL/rbcS and PRK are essential for a functional CBB cycle. The expression of the essential gene, rbcL, greatly differed between the MISNAC and the Niskin samples (TPM: 172.37 versus 0). The coxS, coxM, and coxL genes that encode small, medium, and large subunits of aerobic carbon-monoxide dehydrogenase are the main indicators of the carbon monoxide oxidation pathway. coxS, coxM, and coxL showed significantly higher expression in the MISNAC samples than in the Niskin ones (t-test; p = 0.009 for coxS, p = 0.01 for coxM, and p = 0.04 for coxL). pmoABC genes are vital for methane oxidation in the marine water column, and their average transcriptional levels were significantly higher in the MISNAC samples than those in the Niskin ones (t-test; p values are 0.02, 0.006, and 0.002 for pmoA, pmoB, and pmoC genes, respectively). Ammonia monooxygenase (Amo) is the key enzyme for ammonia oxidation that is involved in autotrophic and nitrification processes in the deep ocean. The average TPM values for amoA, amoB, and amoC genes in the MISNAC samples were significantly higher than those in the Niskin samples (t-test; p values are 0.006, 0.002, and 0.03 for amoA, amoB, and amoC genes, respectively). Copper-dependent nitrite reductase, NirK, catalyzes the key step in denitrification by converting nitrite to nitric oxide. It is also the key enzyme involved in NO2 reduction carried out by ammonia-oxidizing archaea (AOA), following ammonia oxidation. The average TPM value of the nirK gene in the MISNAC samples was significantly higher than that in the Niskin samples (t-test; p = 0.009). Dissimilatory sulfate reduction involves the reduction of sulfate to sulfite by a combination of sulfate adenylyltransferase and adenylyl-sulfate reductase (AprAB). Both aprA and aprB showed significantly higher expression in the MISNAC samples than the Niskin ones (t-test; p = 0.003 for aprA and p = 0.01 for the aprB gene). For the pathways of the Calvin cycle, CO oxidation, methane oxidation, nitrification, and dissimilatory sulfate reduction, the TPM values of the genes in the ISMIFF sample were also considerably higher than those for the Niskin1 and Niskin2 samples. Additionally, the TPM values for amoABC responsible for ammonia oxidation (695.35, 1477.07, and 1595.09 for amoABC, respectively) and pmoABC involved in methane oxidation (7657.55, 1845.19, and 3881.73 for pmoABC, respectively) in the ISMIFF transcriptome were remarkably different from those in the ISMIFU counterpart (0, 0, 0, 0, 0, and 146.37 for amoABC and pmoABC, respectively). Furthermore, it should be noted that over two-thirds (12/17) of the TPM values for the selected genes were zero for ISMIFU. Overall, the genes related to the Calvin cycle, CO oxidation, nitrification, methane oxidation, and dissimilatory sulfate reduction generally showed higher expression levels in the MISNAC and ISMIFF samples, in contrast to the Niskin and ISMIFU samples.

4.

4

Open in a new tab

Transcriptomic variations of different pathways for samples collected by different methods. (A) Transcriptomic variations of functional genes shown as log 10 (TPM + 0.001) for carbon and nitrogen metabolisms in different metatranscriptomes. (B) Sum of TPM for all transcribed genes assigned to each cluster of orthologous group (COG) category for a comparison between the MISNAC, ISMIFF, ISMIFU, and Niskin metatranscriptomes. GI, glycolysis and gluconeogenesis pathway; TCA, tricarboxylic acid cycle; ACE, acetate metabolism; rTCA, reductive tricarboxylic acid cycle; Cal, Calvin cycle; DNR, dissimilatory nitrate reduction; MO, methane oxidation.

We also compared the transcriptional activity of different categories of clusters of orthologous groups (COGs) among the samples (Figure B and Table S9). COG categories of “Cell motility”, “Signal transduction mechanisms”, and “Translation, ribosomal structure and biogenesis” exhibited higher expression in the Niskin samples, much higher than those of all the other samples. A considerable number of COGs are involved in environmental sensing and response. Almost ubiquitously present in microorganisms, the two-component systems (TCSs) perceive and transmit environmental information that is employed to trigger appropriate cellular responses. To understand the microbial response to the environmental changes in the unfixed Niskin samples, we calculated TPM values of TCS genes in the KEGG annotation using the meta-transcriptomes (Figure A,B and Table S10). We revealed the top 10 most abundant TCS gene transcripts in the MISNAC and Niskin samples. For the MISNAC samples, most of the highly expressed genes were predicted to function in carbon and nitrogen metabolisms, such as gln A encoding the nitrogen storage regulator, tctC encoding the C3 transport regulator, fecR encoding the transmembrane sensor, dctM encoding the C4 dicarboxylate transporter, and gln B encoding the nitrogen regulator. The ISMIFF and MISNAC metatranscriptomes resembled each other in the transcriptional patterns of these genes. For the Niskin transcriptomes, a majority of the highly expressed genes were associated with environmental sensing and flagellar motility. These genes are represented by chemoreceptor gene mcp, flagella synthesizer gene filC, negative regulator of flagellin synthesis gene flgM, and type IV pilus assembly gene pilA (Figure C and Table S10). Furthermore, the overall transcription of TCS genes in the ISMIFU sample was at a comparatively low level compared to that of the marker genes of carbon and nitrogen pathways, as the TPM values for most of the TCS genes (123/146) were zero with an average of 8.42 (Table S10). The correlation analysis confirmed the clustering of the ISMIFF and MISNAC samples into one group with an average TPM of their TCS genes of up to 182.97 (Figure B).

5.

5

Open in a new tab

Intersample heterogeneity in the transcriptional level of two-component genes. (A) Box plot illustrating the expression for two-component genes across transcriptomic data sets generated through different sampling methodologies. (B) Correlation heatmap visualizing differences in expression patterns of the two-component genes, complementing the analysis in (A). (C) Heatmap showing the transcriptomic variations of 10 highly expressed two-component system genes shown as log 10 (TPM + 0.001) among the MISNAC (left) and Niskin (right) samples, respectively.

Active Species Revealed by the Coverage Level of MAGs by Metatranscriptomic Reads

Our result of the 16S miTags mentioned above indicates that Saccharospirillaceae, Alteromonadaceae, Spongiibacteraceae, and Marinobacteraceae might proliferate rapidly in the Niskin bottles. To confirm the most active species in the different samples at the genome level, we leveraged the MAGs to recruit transcriptomic reads of the corresponding sample for an estimate of the transcriptional level of the species represented by the MAGs. Alteromonas macleodii, Oleibacter sp002331105, Glaciecola sp014872805, and Oleibacter sp006212115 were more predominant in Niskin1 and Niskin2 (on average, recruitment of 4.06%, 2.63%, 1.35%, and 1.52% of the metatranscriptomic reads by the corresponding MAGs, respectively) than the MISNAC samples (on average, recruitment of 0.0050%, 0.0004%, 0.0002%, and 0.0004% of the metatranscriptomic reads) (Figure A). For the ISMIFF sample, the proportion of these taxa was 0.0060%, 0.0030%, 0.2000%, and 0.1700%, respectively. The AOA species, Nitrosopelagicus sp000484935 and Nitrosopelagicus sp013390745, were more abundant in the MISNAC samples than in the Niskin samples (on average, recruitment of 0.62% and 0.45% of the MISNAC versus 0% and 0% of the Niskin metatranscriptomic reads).

6.

6

Open in a new tab

Relative abundance of eight MAGs representing the most actively transcribed species and their TPM for COG categories in the Niskin and MISNAC metatranscriptomes. (A) Transcriptional activity of the most active species is represented by the percentage of transcriptomic reads mapped to the corresponding MAG and differs among the sampling methods; the MAGs were ranked by their total TPM values calculated among all the Niskin and MISNAC metatranscriptomes. (B) The total TPM values of genes in each COG category for MISNAC (MIS) and Niskin (NIS) transcriptomes are shown in the heatmap for the eight most active prokaryotic species. The letters denote COG categories as below: E: amino acid transport and metabolism; G: carbohydrate transport and metabolism; D: cell cycle control, cell division, chromosome partitioning; N: cell motility; M: cell wall/membrane/envelope biogenesis; H: coenzyme transport and metabolism; Z: defense mechanisms; C: energy production and conversion; W: extracellular structures; S: function unknown; R: general function prediction only; P: inorganic ion transport and metabolism; U: intracellular trafficking, secretion, and vesicular transport; I: lipid transport and metabolism; X: mobilome prophages, transposons; F: nucleotide transport and metabolism; O: posttranslational modification, protein turnover, chaperones; L: replication, recombination, and repair; Q: secondary metabolites biosynthesis, transport, and catabolism; T: signal transduction mechanisms; K: transcription; J: translation, ribosomal structure, and biogenesis.

The transcriptional level of the COG categories of the above 8 highly transcribed MAGs (Figure B) confirmed higher activity ofA. macleodii, Oleibacter sp002331105, Glaciecola sp014872805, andOleibacter sp006212115 in the Niskin samples than in the MISNAC samples. The COG categories that exhibited the highest average TPM rankings in the Niskin samples for these species were “Translation, ribosomal structure and biogenesis,” “Energy production and conversion,” and “Posttranslational modification, protein turnover, chaperones.” However, a contrasting pattern was observed for the MAGs of Nitrosopelagicus, CG-Epi1 sp002506485, and P. autotrophicus A, as the COG category of “Amino acid transport and metabolism” was among the highest average TPM rankings when searching against the MISNAC metatranscriptomes.

Discussion

In this study, we uncovered different active microorganisms and their gene expression profiles in the deep-sea samples obtained by different sampling methods. While limitations associated with deep-sea sampling precluded the collection of replicate samples for the ISMIFF and ISMIFU approaches, it is still possible to gain some insights into the important question about the effects of in situ filtration and preservation on downstream data, particularly for gene expression. In comparison to metatranscriptomes for the samples collected using Niskin bottles, those for the samples with in situ nucleic acid extraction or in situ fixation treatment were more likely to capture new microbial taxa and to offer a precise depiction of the metabolic processes actively occurring within the deep-sea environment. For the Niskin samples, rapid growth of certain temperature- and pressure-sensitive bacteria and/or exposure of living bacteria to temperature and pressure changes during sample recovery likely resulted in high expression of TCS genes and low-oxygen adaptive genes in those samples. Here, Alteromonadales were shown to be the most active bacteria in our Niskin samples, which is consistent with our previous study. Members of the Alteromonadales order are characterized by a large cell size and a robust growth capacity under minimal nutritional conditions. Notably, Alteromonadales bacteria are characterized by relatively high optimal growth temperatures, compared to other deep-sea microorganisms. Probably, the prevalence of this group of bacteria in the Niskin samples was stimulated by increasing the temperature during the lander recovery onboard. Besides, hydrostatic pressure is also an important factor for Alteromonadales growth. It is reported that major members of Alteromonas were piezoresistive prokaryotes in the deep-sea community. When piezoresistive prokaryotes experience depressurization, a significant increase in metabolic activity of over 100-fold has been observed. The majority of genes highly expressed in A. macleodii are associated with the ribosome synthesis pathway in this study, confirming its remarkable proliferation and enhanced activity under depressurized conditions in our Niskin samples. Moreover, Alteromonas bacteria have been found on the surface of the ocean. Another possibility is that during the ascent of the Niskin bottles to the surface, dissolved gases in the water sample might escape from the Niskin, possibly introducing water and taxa from the upper water column into the collected sample.

It is crucial to minimize environmental disturbance during the deep-sea sample collection and preservation. Cells in the Niskin water samples of this study have obviously suffered from environmental disturbance including depressurization and temperature increase. This is reflected by the high gene expression of two-component systems in the Niskin samples of the current study. For the MISNAC samples, most of the highly expressed genes in TCS were related to carbon and nitrogen metabolisms; however, for the Niskin samples, the majority of the highly expressed genes were associated with environmental sensing and flagellar motility, which is similar to the previous study, as the relative abundance of genes involved in chaperones and cell motility in Gammaproteobacteria was higher in the MIS-SID data set than that in the Niskin data set. Additionally, it is worth noting that ccoO responsible for respiratory electron transport under low oxygen is among the top ten highly transcribed TCS genes in our Niskin samples, which suggests low oxygen content in the bottles likely due to the rapid proliferation of certain heterotrophic microorganisms, such as Saccharospirillaceae, Spongiibacteraceae, Alteromonadaceae, and Marinobacteraceae. The enhanced microbial activities probably depleted large amounts of oxygen in the microenvironments, particularly during the microbial degradation of organic particles within the Niskin bottle. Except forAlteromonas, we also identified high gene expression and dominance of Oleibacter and Glaciecola in the Niskin samples. It is reported that the optimum growth ofOleibacter occurs at 25–30 °C. When the Niskin bottle was recovered to the research vessel, the temperature increased gradually from 4 to 28 °C. This might result in the active proliferation of Oleibacter in the Niskin bottle. In previous studies, Alteromonas, Oleibacter, and Glaciecola have been reported as the dominant genera in the Niskin samples from the deep-seawater column of the Tropical Western Pacific Ocean and the North Atlantic. Therefore, the changes in the external environment possibly resulted in an elevation of environmental sensitivity and motility of microorganisms within the Niskin sampling bottles.

Pressure changes lead to diverse impacts on a deep-sea sample, while some microbial activities are more sensitive to the perturbation than others. In addition to the TCS genes, some others related to energy metabolism and carbon/nitrogen/sulfur cycles were also influenced by the sampling methods, as evidenced by the quantification of gene transcripts in the MS-SID and AFIS samples. Our data indicate that the contribution of ammonium oxidation and methane oxidation to the deep-sea ecosystem might be underestimated. The rapidly proliferating microorganisms might outcompete the dominant indigenous bacteria in the Niskin samples, probably resulting in declining transcriptional activity of these pathways specifically active in the deep sea. Additionally, alterations in external environmental factors, such as temperature and pressure, might also prohibit the metabolic responses of the indigenous deep-sea bacteria in this study. Hence, using the proper in situ sampling methods is fundamental for the capture and discovery of those short-half-life mRNA molecules.

The microbial community structures of active deep-sea eukaryotes also exhibited variations across different sampling methodologies. We found that Oligotrichia was the dominant active taxon among all the samples as shown in our previous study. Oligotrichia had been found in the Antarctic, North Pacific, Black Sea, and the Caribbean Sea, as well as SCS. , However, the abundance and diversity of viable microbial species in the Niskin and ISMIFU samples were markedly lower than those in the MISNAC samples of this study. Besides the difference in sampling volume, the discrepancy in biodiversity and transcriptomes might arise from the pressure decrease during the Niskin and ISMIFU recovery, preventing the timely release of dissolved gases from eukaryotic microorganisms and probably resulting in consequent cell explosion. , Furthermore, deep-sea organisms contain high levels of osmolytes to counteract high-pressure damage to cells. During the pressure reduction process, the osmolytes might not be adequately degraded, leading to cell fragmentation as well. , Furthermore, due to the presence of active organisms within the Niskin bottle, species composition might shift over time. The confined environment and limited resources could probably influence the natural food web and cause certain taxa to become dominant quickly. Here, we report that Oligotrichia was the dominant taxon in all our samples. However, some eukaryotes, such as bacterial consumer Choreotrichia, were a major family in the MISNAC samples but were almost missing from the Niskin samples. In the Niskin bottles with restricted water exchange and limited food availability, potential shifts of predator–prey relationships probably resulted in low competition of Choreotrichia in the Niskin bottles under changing conditions. In sum, to obtain an accurate snapshot of in situ eukaryotic microbial activity, it is important to collect and preserve samples while minimizing the impact of environmental changes on the samples.

Gene expression of the organisms under in situ and onboard treatments exhibited disparity in this study. Transcriptional activities of taxa and genes were similar between the MISNAC and ISMIFF samples, regardless of the fact that the microbes in the MISNAC samples were lysed in situ, and the ISMIFF sample was fixed in situ and lysed in the laboratory. In contrast, the ISMIFU sample was not fixed in situ, which resulted in a lack of transcripts for a majority of the genes. The lower expression of the 40 single-copy genes in the ISMIFU sample, compared to the ISMIFF and Niskin samples, suggests that without immediate killing and RNA stabilization, microbial cells remaining in the MISNAC filtration chamber would ultimately lose most of their metabolic activities. Additionally, the filtration process of the Niskin water onboard resulted in exposure of the deep-sea microbes to a completely different environment for a considerable period of time, which is probably causal to the degradation of the existing RNA and expression of a new set of genes. In spite of evidence from a single sample of ISMIFF, the strong divergence in the transcriptional pattern between the Niskin and in situ samples with RNA extraction/fixation (MINSAC and ISMIFF) implies the importance of instant chemical fixation or cell lysis for a deep-sea microbial sample. Certainly, in situ membrane filtration of deep-seawater samplers also has limitations. It might lead to cell lysis or significant alterations to cellular activities, both of which could decrease recovered transcripts in filtered and unpreserved samples for some taxa. Alternatively, predation may alter the recovery of some taxa if predatory cells on filters remain active. Additionally, it is inevitable that cells may be exposed to pressure on the membrane during the process of membrane filtration. This exposure could potentially alter the gene expression profile as cells are subjected to pressures, especially for a long filtration time. Future studies should investigate optimal filtration volumes and times that minimize their impact on the gene expression of active microbial communities. While ISMIFF and MISNAC sampling methods capture deep-sea microbial in situ activity more accurately than traditional Niskin sampling, artifacts arising from phage-induced cell death in active microbial communities may compromise sample integrity. This process could selectively eliminate specific bacterial groups, thereby altering the community structure. Furthermore, environmental DNA retained on filtration membranes may also distort the observed microbial composition. Altogether, a protocol is currently required to assess how well a sampling method reflects deep-sea in situ conditions.

In this study, we demonstrate that deep-sea sampling methods dramatically affect the results of the gene activity and element cycling of deep-sea microbiomes. The metatranscriptomes of MISNAC and ISMIFF with the capacity of RNA preservation showed highly diverse active microbes. Their gene expression profiles differed much from the results for the ISMIFU and Niskin samples, which are dominated by the genes involving TCS, ammonia oxidation, and C1 (methane, CO, and CO2) metabolisms. Therefore, these results suggest that in situ preservation and nucleic acid extraction should be a priority for quantitative studies of critical microbial activities. This is essential for understanding the critical microbial activities that influence carbon cycling and other nutrient utilization, which may have great impacts on global climate change. Developing a multifunctional sampler capable of recovering diverse sample types will facilitate a more comprehensive reconstruction of deep-sea ecological networks by tracing the interconnectivity across microbial communities.

Materials and Methods

Sampling

The samples were collected on November 25th of 2019 with the 39th deployment of the “Phoenix” lander at a 1038 m depth. The sampling location was at latitude 17°04.28′N and longitude 110°00.588′E of SCS. The MISNAC apparatus has been described elsewhere (Figure S1A,E). For the MISNAC samples, the seawater was automatically pumped into the filtration chamber and flowed through the 0.22 μm polycarbonate membrane (Millipore, Bedford, MA, USA) at the deep-sea site first. Next, 200 mL cell lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 10% GuHCl, 10% DTT, 1% Triton X-100, 0.2% Proteinase K and 0.1% Lysozyme) was injected into the membrane chamber, which replaces the seawater and lyses the cells (Figure S1B). To ensure complete cell digestion, the cell lysis buffer was maintained in the chamber for 30 min, based on our laboratory experimental results. Then, 90% ethanol was injected to mix with the cell lysis mixture, followed by the attachment of the precipitated nucleic acids (DNA/RNA) on adsorption columns (Tiangen, Beijing, China) in the terminal nucleic acid collection system. Six subunits in the MISNAC device were used for MISNAC sample collection. The ISMIFF sample was obtained by in situ water filtration through a 0.22 μm polycarbonate membrane using an ISMIFF device. Subsequently, 200 mL of RNALater (Ambion, Carlsbad, CA, USA) was added into the membrane chamber to preserve the microbes (Figure S1C,F). For the ISMIFU sample, seawater was filtered through a 0.22 μm polycarbonate membrane in the chamber of an ISMIFF device without the addition of any fixative or lysis solution (Figure S1D). The membranes of the ISMIFF and ISMIFU were stored frozen after retrieval upon return of the lander. For MISNAC, the filtration time was 4 h for each sample, during which approximately 40–80 L of seawater was filtered. Approximately 10 L of seawater was filtered by the ISMIFF devices. Due to the varying deep-sea environment, large organic particles and swimming zooplanktons might accumulate on the filter membrane, leading to partial clogging despite our prefiltration efforts using a mesh with a pore size of 1 mm. This might have resulted in variations in the filtered volume of the deep-seawater, despite consistent settings in the filtration time. After recovery of MISNAC onboard, the nucleic acids in the columns were rinsed with 2 mL of ddH2O, collected by centrifugation into a tube, and stored in a −20 °C freezer onboard. Two 10 L Niskin bottles were mounted on the lander for water collection. The seawater in the bottles was filtered through a 0.22 μm polycarbonate membrane onboard. All of the filtration membranes were temporarily stored in a −20 °C freezer onboard. Upon return to the laboratory, all of the samples were processed immediately for metagenomics and metatranscriptomics work.

Metatranscriptome and Metagenome Library Preparations

The ISMIFF, ISMIFU, and Niskin filtration membranes were cut into small pieces and soaked in the cell lysis buffer for 30 min. The nucleic acids were then precipitated with 90% ethanol and adsorbed using the nucleic acid binding columns, followed by rinsing and centrifugation as performed on the MISNAC columns. For DNA library preparation, all the nucleic acid samples were incubated with 0.1 ng/μL RNaseA (TaKaRa, Dalian, China) at 37 °C for 10 min to remove RNA fragments. The purity and quantity of the DNA samples were checked using a Qubit 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA). 100 ng of DNA was fragmented to 350 bp using a Covaris M220 Focused ultrasonicator (Covaris, Massachusetts, USA). Metagenomic sequencing libraries were then constructed using a VAHTS Universal DNA library prep kit for Illumina V3 (Vazyme, Nanjing, China) according to the manufacturer’s protocol. Although each MISNAC DNA/RNA coextraction yielded much more RNA (>2 μg) than the other samples, the same amount of RNA (10 ng) from different samples was used for the metatranscriptomic sequencing library construction. DNA fragments were removed from the MISNAC coextractions using a Turbo DNA-free Kit (Ambion, Carlsbad, CA, USA) according to the manufacturer’s instructions. To ensure complete DNA removal from the nucleic acid samples, the universal primers 341F (5′-CCTAYGGGRBGCASCAG-3′) and 802R (5′-TACNVGGGTATCTAATCC-3′) for the 16S rRNA gene were used for PCR amplification. Double-stranded cDNA was synthesized using an Ovation RNA-Seq System V2 Kit (Qiagen, Hilden, Germany). The libraries were sequenced on an Illumina NovaSeq6000 platform (Illumina, San Diego, USA) with a paired-end approach for 2 × 150 bp reads.

Sequencing and Processing of Raw Illumina Sequencing Data

The metagenomic and metatranscriptomic raw reads were trimmed to remove adapters using fastp (v.0.21.0) with parameters (−f 1 −t 0 –F 1 −T 0 −w 24 −c −q 20 −u 20 −g −W 5 −3 −l 50). After that, low-quality reads (>20% of read lengths had quality scores <20), unpaired reads, and reads <50 bp were removed. Then FastUniq was used to remove duplicates introduced by PCR amplification with default parameters. For the metagenomes, Bowtie2 was used with default parameters to map reads to a control sequencing data set from our lab and contamination databases, which include sequences from mouse, human, and common laboratory contaminant bacterial genomes downloaded from the NCBI, to filter out contamination reads. For the metatranscriptomes, SortMeRNA software was used to remove rRNA gene reads (5S, 5.8S, 16S, 18S, 23S, 28S) with an e-value cutoff of 1 × 10–5.

Taxonomic Classification and Biodiversity Estimates Using 16S and 18S miTags

The 16S and 18S rRNA miTags were identified from the qualified reads of metatranscriptomes using rRNA_HMM. The mapped fragments to the V4 region of the 16S rRNA gene and V9 region of the 18S rRNA gene were further identified using hmmsearch. The length threshold for miTags was set to 148 bp. The reads meeting the size criterion were extracted and subsequently imported into Qiime2 (v.2022.2) for biodiversity analysis and taxonomic classification. The number of reads was calculated by Qiime2 as shown in Table S11. To avoid potential contaminant taxa, 16S and 18S miTags were recruited from the metagenomic data for human skin and laboratory kits to form a control. We conducted the same filtration step as described above to remove contaminant miTags from the control. Kranken2 (v2.1.0) and Bracken were used for taxonomic classification of metagenomic reads.

We used DADA2 to correct Illumina sequencing errors in the rRNA gene reads. The 16S and 18S miTags were clustered into operational taxonomic units (OTUs) at a 99% similarity threshold using Qiime2. Using the Qiime2 classify-sklearn plugin against the SILVA 138 database, taxonomic classification and composition of the microbial communities were obtained. The representative reads of the most abundant OTUs in the unassigned group were further classified taxonomically by searching in the NCBI database using Blastn (e-value cutoff = 10). The relative abundances of the taxa were used for PCoA. We also used the diversity alpha-rarefaction function in Qiime2 to calculate the alpha diversity indices including Chao1, Shannon, and Simpson.

Metagenome Assembly, Genome Binning, and Gene Annotation

Metagenome assembly was conducted separately for each sample using metaSPAdes (v3.15.2) with parameters “-min-contig-len 1300 −k 21,33,55,77,99,127”. Genome binning was performed by MetaWRAP (v1.2.1) software. The bin refinement module was run with the default parameters. CheckM was used to identify and select MAGs with completeness greater than 50% and contamination less than 10% (metawrap bin_refinement “–c 50 −x 10”). All of the MAGs were then classified taxonomically with GTDB-Tk v.2.3.2 (GTDB release220). The ANI values of the MAGs were also calculated with GTDB-Tk v.2.3.2. After dereplication using dRep using an ANI threshold of 97%, the MAGs of the MISNAC and Niskin data sets were compared with the GTDB genomes to calculate ANI with the closest relatives again. The number of the MAGs associated with <95% ANI was counted. Protein prediction of the genomes was performed using Prodigal (v.2.6.3) with the option “–p meta”. KofamScan18 (v.1.1.0) and GhostKOALA (v.2.2) were used to assign KEGG orthologs (KO) to the predicted protein sequences.

Gene Expression Inferred from Metatranscriptomic Data

Gene expression levels were estimated by calculating TPM values using CoverM (v.0.6.1 https://github.com/wwood/CoverM) (“–m tpm -min-read-aligned-length 50 --min-read-percent-identity 0.97 --min-covered-overlap-fraction 10”). To normalize different metatranscriptomic data sets, we divided the TPM values of ORF clusters from the metatranscriptomes by the total count of 40 universally conserved, single-copy marker genes recommended by mOTUs2. ,

Estimation of Gene Expression Variations

The KEGG annotation of the MAGs binned from the metagenomes resulted in the functional characterization of 6212 genes. Transcript abundances for these genes, measured as TPM values across metatranscriptomic profiles, were subjected to Pearson correlation analysis to determine both p-values and R 2 metrics. Cluster analysis was then performed using the R 2 metrics to generate a heatmap visualizing relationships between the transcriptomes obtained by different methods. Besides, student’s tests were performed to identify the significantly differentiated gene expression patterns between the MISNAC and Niskin samples.

The 40 single-copy housekeeping genes previously identified in mOTUs2 , were selected to probe the activity of conserved prokaryotic genes. TPM values for these reference genes were calculated across meta-transcriptomes and were visualized via box plots. Pearson correlation analysis was conducted to determine p-values and R 2 coefficients between gene expression profiles. Clustering based on the R 2 metrics was then performed, and the results were displayed on a heatmap to complement the above correlation analysis. Besides, log 10 (TPM + 0.001) was used to transform the TPM values in the heatmap. Additionally, TCS genes selected based on the KEGG pathway ko02020 were also plotted using the same approach.

Identification of the Most Actively Transcribed COGs in Eight MAGs

The transcriptional level of the most active prokaryotes represented by MAGs was computed across MISNAC1-6, ISMIFF, ISMIFU, Niskin1, and Niskin2. The percentage of metatranscriptomic reads mapped on the MAGs was estimated by CoverM using “–m relative_abundance -min-read-aligned-length 50 --min-read-percent-identity 0.95 --min-covered-overlap-fraction 10” as an indicator of the transcriptional level. The recruitment rates of MISNAC, ISMIFF, ISMIFU, and Niskin samples were also calculated using CoverM with the same parameters. The top 8 MAGs exhibiting the highest average transcriptional levels among all the transcriptomes were then selected for further analysis of actively transcribed genes. The COGs in the MAGs were annotated using the COG classifier (https://github.com/moshi4/COGclassifier/). The TPM values of genes assigned to the same COG category were summed across the MAGs for the different samples. The expression differences of the COG categories for individual MAGs were then compared between those of the MISNAC and Niskin samples.

Supplementary Material

es4c05624_si_001.pdf (298KB, pdf)
es4c05624_si_002.xlsx (132.8KB, xlsx)

Acknowledgments

This study was supported by the National Natural Science Foundation of China (42376149), the Shenzhen Key Laboratory of Advanced Technology for Marine Ecology (ZDSYS20230626091459009), and the National Key Research and Development Program of China (2016YFC0302504).

All data are available in the main text or the Supporting Information. The raw metatranscriptomic sequencing data, assembled contigs, and MAGs generated in this study have been deposited in the NCBI database under the accession number PRJNA1109759. We have archived the non-rRNA metatranscriptomes and MAG data sets in the Zenodo repository. The data sets are accessible through the following Digital Object Identifier (DOI): https://zenodo.org/records/13381952 and https://zenodo.org/records/14230125. Furthermore, the analysis codes are publicly available at https://github.com/heyinghui22/Codes-for-In-situ-sampling-uncovers-seasonal-variability-of-active-deep-sea-microbes-.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.4c05624.

  • Schematic diagram of four deep-sea sampling methods (PDF)

  • Tables of sampling methods and mesopelagic water volumes; sequencing data yields for metagenomes and metatranscriptomes; relative abundance of prokaryotic taxa based on 16S miTags in metatranscriptomes; relative abundance of prokaryotic and eukaryotic taxa in metagenomes estimated by Kraken2; relative abundance of eukaryotic taxa in metatranscriptomes; quality metrics and characteristics of MAGs; TPM values of mOTUs2 marker genes in metatranscriptomes; representative genes and TPM values for carbon, nitrogen, and sulfur metabolism; TPM profiles of COG categories across MAGs in metatranscriptomes; TPM values of TCS genes across metatranscriptomes; and extracted read counts for V4 (16S rRNA) and V9 (18S rRNA) regions in metatranscriptomes (XLSX)

Y.H. and Y.W. conceived the study. H.Z. performed the experiment. Y.H. analyzed the data. Y.H. wrote the manuscript, and Y.W. and F.B. critically revised the manuscript.

The authors declare no competing financial interest.

References

  1. York A.. Marine biogeochemical cycles in a changing world. Nat. Rev. Microbiol. 2018;16(5):259. doi: 10.1038/nrmicro.2018.40. [DOI] [PubMed] [Google Scholar]
  2. Wang F., Li M., Huang L., Zhang X.-H.. Cultivation of uncultured marine microorganisms. Mar. Life Sci. & Technol. 2021;3(2):117–120. doi: 10.1007/s42995-021-00093-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cario A., Oliver G. C., Rogers K. L.. Exploring the deep marine biosphere: Challenges, innovations, and opportunities. Front. Earth Sci. 2019;7:1–9. doi: 10.3389/feart.2019.00225. [DOI] [Google Scholar]
  4. Zhou Y.-L., Mara P., Cui G.-J., Edgcomb V. P., Wang Y.. Microbiomes in the Challenger Deep slope and bottom-axis sediments. Nat. Commun. 2022;13(1):1515. doi: 10.1038/s41467-022-29144-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Taylor C. D., Edgcomb V. P., Doherty K. W., Engstrom I., Shanahan T., Pachiadaki M. G., Molyneaux S. J., Honjo S.. Fixation filter, device for the rapid in situ preservation of particulate samples. Deep Sea Res., Part I. 2015;96(FEB):69–79. doi: 10.1016/j.dsr.2014.09.006. [DOI] [Google Scholar]
  6. Wang Y., Gao Z. M., Li J., He L. S., Cui G. J., Li W. L., Chen J., Xin Y. Z., Cai D. S., Zhang A. Q.. Hadal water sampling by in situ microbial filtration and fixation (ISMIFF) apparatus. Deep Sea Res., Part I. 2019;144:132–137. doi: 10.1016/j.dsr.2019.01.009. [DOI] [Google Scholar]
  7. Edgcomb V., Taylor C., Pachiadaki M., Honjo S., Engstrom I., Yakimov M.. Comparison of Niskin vs. in situ approaches for analysis of gene expression in deep Mediterranean Sea water samples. Deep Sea Res., Part II. 2016;129:213–222. doi: 10.1016/j.dsr2.2014.10.020. [DOI] [Google Scholar]
  8. Feike J., Jürgens K., Hollibaugh J. T., Krüger S., Jost G., Labrenz M.. Measuring unbiased metatranscriptomics in suboxic waters of the central Baltic Sea using a new in situ fixation system. ISME J. 2012;6(2):461. doi: 10.1038/ismej.2011.94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Charvet S., Riemann L., Alneberg J., Andersson A. F., von Borries J., Fischer U., Labrenz M.. AFISsys - An autonomous instrument for the preservation of brackish water samples for microbial metatranscriptome analysis. Water Res. 2019;149:351–361. doi: 10.1016/j.watres.2018.11.017. [DOI] [PubMed] [Google Scholar]
  10. Garel M., Bonin P., Martini S., Guasco S., Roumagnac M., Bhairy N., Armougom F., Tamburini C.. Pressure-retaining sampler and high-pressure systems to study deep-sea microbes under in situ conditions. Front. Microbiol. 2019;10:453. doi: 10.3389/fmicb.2019.00453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gao Z. M., Huang J. M., Cui G. J., Li W. L., Li J., Wei Z. F., Chen J., Xin Y. Z., Cai D. S., Zhang A. Q.. et al. In situ meta-omic insights into the community compositions and ecological roles of hadal microbes in the Mariana Trench. Environ. Microbiol. 2019;21(11):4092–4108. doi: 10.1111/1462-2920.14759. [DOI] [PubMed] [Google Scholar]
  12. Wei Z. F., Li W. L., Li J., Chen J., Xin Y. Z., He L. S., Wang Y.. Multiple in situ nucleic acid collections (MISNAC) from deep-sea waters. Front. Mar. Sci. 2020;7:81. doi: 10.3389/fmars.2020.00081. [DOI] [Google Scholar]
  13. Chaumeil P.-A., Mussig A. J., Hugenholtz P., Parks D. H.. GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics. 2022;38(23):5315–5316. doi: 10.1093/bioinformatics/btac672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kanehisa M., Goto S.. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000;28(1):27–30. doi: 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Milanese A., Mende D. R., Paoli L., Salazar G., Ruscheweyh H.-J., Cuenca M., Hingamp P., Alves R., Costea P. I., Coelho L. P., Schmidt T. S. B., Almeida A., Mitchell A. L., Finn R. D., Huerta-Cepas J., Bork P., Zeller G., Sunagawa S.. Microbial abundance, activity and population genomic profiling with mOTUs2. Nat. Commun. 2019;10(1):1014. doi: 10.1038/s41467-019-08844-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Sunagawa S., Mende D. R., Zeller G., Izquierdo-Carrasco F., Berger S. A., Kultima J. R., Coelho L. P., Arumugam M., Tap J., Nielsen H. B., Rasmussen S., Brunak S., Pedersen O., Guarner F., de Vos W. M., Wang J., Li J., Dore J., Ehrlich S. D., Stamatakis A., Bork P.. Metagenomic species profiling using universal phylogenetic marker genes. Nat. Methods. 2013;10(12):1196–1199. doi: 10.1038/nmeth.2693. [DOI] [PubMed] [Google Scholar]
  17. Hügler M., Sievert S. M.. Beyond the Calvin cycle: autotrophic carbon fixation in the ocean. Ann. Rev. Mar. Sci. 2011;3:261–289. doi: 10.1146/annurev-marine-120709-142712. [DOI] [PubMed] [Google Scholar]
  18. Ragsdale S. W.. Life with carbon monoxide. Crit. Rev. Biochem. Mol. Biol. 2004;39(3):165–195. doi: 10.1080/10409230490496577. [DOI] [PubMed] [Google Scholar]
  19. Lieberman R. L., Rosenzweig A. C.. Biological methane oxidation: regulation, biochemistry, and active site structure of particulate methane monooxygenase. Crit. Rev. Biochem. Mol. Biol. 2004;39(3):147–164. doi: 10.1080/10409230490475507. [DOI] [PubMed] [Google Scholar]
  20. Wright C. L., Lehtovirta-Morley L. E.. Nitrification and beyond: metabolic versatility of ammonia oxidising archaea. ISME J. 2023;17(9):1358–1368. doi: 10.1038/s41396-023-01467-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Helen D., Kim H., Tytgat B., Anne W.. Highly diverse nirK genes comprise two major clades that harbour ammonium-producing denitrifiers. BMC Genomics. 2016;17(1):155. doi: 10.1186/s12864-016-2465-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kobayashi S., Hira D., Yoshida K., Toyofuku M., Shida Y., Ogasawara W., Yamaguchi T., Araki N., Oshiki M.. Nitric oxide production from nitrite reduction and hydroxylamine oxidation by copper-containing dissimilatory nitrite reductase (NirK) from the aerobic ammonia-oxidizing archaeon, Nitrososphaera viennensis. Microbes Environ. 2018;33(4):428–434. doi: 10.1264/jsme2.ME18058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Anantharaman K., Hausmann B., Jungbluth S. P., Kantor R. S., Lavy A., Warren L. A., Rappé M. S., Pester M., Loy A., Thomas B. C., Banfield J. F.. Expanded diversity of microbial groups that shape the dissimilatory sulfur cycle. ISME J. 2018;12(7):1715–1728. doi: 10.1038/s41396-018-0078-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Tatusov R. L., Galperin M. Y., Natale D. A., Koonin E. V.. The COG database: a tool for genome-scale analysis of protein functions and evolution. Nucleic Acids Res. 2000;28(1):33–36. doi: 10.1093/nar/28.1.33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jacob-Dubuisson F., Mechaly A., Betton J.-M., Antoine R.. Structural insights into the signalling mechanisms of two-component systems. Nat. Rev. Microbiol. 2018;16(10):585–593. doi: 10.1038/s41579-018-0055-7. [DOI] [PubMed] [Google Scholar]
  26. Harth G., Masleša-Galić S., Tullius M. V., Horwitz M. A.. All four Mycobacterium tuberculosis glnA genes encode glutamine synthetase activities but only GlnA1 is abundantly expressed and essential for bacterial homeostasis. Mol. Microbiol. 2005;58(4):1157–1172. doi: 10.1111/j.1365-2958.2005.04899.x. [DOI] [PubMed] [Google Scholar]
  27. Sweet G., Somers J., Kay W.. Purification and properties of a citrate-binding transport component the C protein of Salmonella typhimurium. Can. J. Biochem. 1979;57(6):710–715. doi: 10.1139/o79-089. [DOI] [PubMed] [Google Scholar]
  28. Braun V., Mahren S., Sauter A.. Gene regulation by transmembrane signaling. BioMetals. 2006;19:103–113. doi: 10.1007/s10534-005-8253-y. [DOI] [PubMed] [Google Scholar]
  29. Janausch I., Zientz E., Tran Q., Kröger A., Unden G.. C4-dicarboxylate carriers and sensors in bacteria. Biochim. Biophys. Acta. 2002;1553(1–2):39–56. doi: 10.1016/S0005-2728(01)00233-X. [DOI] [PubMed] [Google Scholar]
  30. Drepper T., Groß S., Yakunin A. F., Hallenbeck P. C., Masepohl B., Klipp W. J. M.. Role of GlnB and GlnK in ammonium control of both nitrogenase systems in the phototrophic bacterium Rhodobacter capsulatus. Microbiology. 2003;149(8):2203–2212. doi: 10.1099/mic.0.26235-0. [DOI] [PubMed] [Google Scholar]
  31. Salah Ud-Din A. I. M., Roujeinikova A.. Methyl-accepting chemotaxis proteins: a core sensing element in prokaryotes and archaea. Cell. Mol. Life Sci. 2017;74(18):3293–3303. doi: 10.1007/s00018-017-2514-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Jozwick A. K. S., LaPatra S. E., Graf J., Welch T. J.. Flagellar regulation mediated by the Rcs pathway is required for virulence in the fish pathogen Yersinia ruckeri. Fish Shellfish Immunol. 2019;91:306–314. doi: 10.1016/j.fsi.2019.05.036. [DOI] [PubMed] [Google Scholar]
  33. Xu S., Peng Z., Cui B., Wang T., Song Y., Zhang L., Wei G., Wang Y., Shen X.. FliS modulates FlgM activity by acting as a non-canonical chaperone to control late flagellar gene expression, motility and biofilm formation in Yersinia pseudotuberculosis. Environ. Microbiol. 2014;16(4):1090–1104. doi: 10.1111/1462-2920.12222. [DOI] [PubMed] [Google Scholar]
  34. Ronish L. A., Lillehoj E., Fields J. K., Sundberg E. J., Piepenbrink K. H.. The structure of PilA from Acinetobacter baumannii AB5075 suggests a mechanism for functional specialization in Acinetobacter type IV pili. J. Biol. Chem. 2019;294(1):218–230. doi: 10.1074/jbc.RA118.005814. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Bowman, J. P. ; McMeekin, T. A. . Alteromonadaceae. In Bergey’s Manual of Systematics of Archaea Bacteria; Wiley, 2015; pp 1–2. [Google Scholar]
  36. Amano C., Zhao Z., Sintes E., Reinthaler T., Stefanschitz J., Kisadur M., Utsumi M., Herndl G. J.. Limited carbon cycling due to high-pressure effects on the deep-sea microbiome. Nat. Geosci. 2022;15(12):1041–1047. doi: 10.1038/s41561-022-01081-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. López-Pérez M., Gonzaga A., Martin-Cuadrado A.-B., Onyshchenko O., Ghavidel A., Ghai R., Rodriguez-Valera F.. Genomes of surface isolates of Alteromonas macleodii: the life of a widespread marine opportunistic copiotroph. Sci. Rep. 2012;2(1):696. doi: 10.1038/srep00696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ducluzeau A.-L., Ouchane S., Nitschke W.. The cbb3 oxidases are an ancient innovation of the domain Bacteria. Mol. Biol. Evol. 2008;25(6):1158–1166. doi: 10.1093/molbev/msn062. [DOI] [PubMed] [Google Scholar]
  39. Teramoto M., Ohuchi M., Hatmanti A., Darmayati Y., Widyastuti Y., Harayama S., Fukunaga Y.. Oleibacter marinus gen. nov., sp. nov., a bacterium that degrades petroleum aliphatic hydrocarbons in a tropical marine environment. Int. J. Syst. Evol. Microbiol. 2011;61(2):375–380. doi: 10.1099/ijs.0.018671-0. [DOI] [PubMed] [Google Scholar]
  40. Sun Q., Song J., Li X., Yuan H., Ma J., Wang Q.. Bacterial vertical and horizontal variability around a deep seamount in the Tropical Western Pacific Ocean. Mar. Pollut. Bull. 2020;158:111419. doi: 10.1016/j.marpolbul.2020.111419. [DOI] [PubMed] [Google Scholar]
  41. Vázquez Rosas Landa M., De Anda V., Rohwer R. R., Angelova A., Waldram G., Gutierrez T., Baker B. J.. Exploring novel alkane-degradation pathways in uncultured bacteria from the North Atlantic Ocean. mSystems. 2023;8(5):e00619-23. doi: 10.1128/msystems.00619-23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Xu P., Yang M., He L., Zhang H., Gao Z., Jiang Y., Wang Y.. Discovery of prevalent Ciliophora, Discoba, and Copepoda protists in deep sea water by in situ nucleotide extraction. J. Mar. Sci. Eng. 2024;12(1):61. doi: 10.3390/jmse12010061. [DOI] [Google Scholar]
  43. Agatha S.. Global diversity of aloricate Oligotrichea (Protista, Ciliophora, Spirotricha) in marine and brackish sea water. PLoS One. 2011;6(8):e22466. doi: 10.1371/journal.pone.0022466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Gooday A. J., Schoenle A., Dolan J. R., Arndt H.. Protist diversity and function in the dark ocean–Challenging the paradigms of deep-sea ecology with special emphasis on foraminiferans and naked protists. Eur. J. Protistol. 2020;75:125721. doi: 10.1016/j.ejop.2020.125721. [DOI] [PubMed] [Google Scholar]
  45. Yancey P. H., Blake W. R., Conley J.. Unusual organic osmolytes in deep-sea animals: adaptations to hydrostatic pressure and other perturbants. Comp. Biochem. Physiol., Part A: Mol. Integr. Physiol. 2002;133(3):667–676. doi: 10.1016/S1095-6433(02)00182-4. [DOI] [PubMed] [Google Scholar]
  46. Santoferrara L. F., Alder V. V., McManus G. B.. Phylogeny, classification and diversity of Choreotrichia and Oligotrichia (Ciliophora, Spirotrichea) Mol. Phylogenet. Evol. 2017;112:12–22. doi: 10.1016/j.ympev.2017.03.010. [DOI] [PubMed] [Google Scholar]
  47. Bairoliya S., Koh Zhi Xiang J., Cao B.. Extracellular DNA in environmental samples: Occurrence, extraction, quantification, and impact on microbial biodiversity assessment. Appl. Environ. Microbiol. 2022;88(3):e0184521. doi: 10.1128/aem.01845-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Chen S., Zhou Y., Chen Y., Gu J.. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–i890. doi: 10.1093/bioinformatics/bty560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Xu H., Luo X., Qian J., Pang X., Song J., Qian G., Chen J., Chen S.. FastUniq: a fast de novo duplicates removal tool for paired short reads. PLoS One. 2012;7(12):e52249. doi: 10.1371/journal.pone.0052249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Salter S. J., Cox M. J., Turek E. M., Calus S. T., Cookson W. O., Moffatt M. F., Turner P., Parkhill J., Loman N. J., Walker A. W.. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol. 2014;12:1–12. doi: 10.1186/s12915-014-0087-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Kopylova E., Noe L., Touzet H.. SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data. Bioinformatics. 2012;28(24):3211–3217. doi: 10.1093/bioinformatics/bts611. [DOI] [PubMed] [Google Scholar]
  52. Huang Y., Gilna P., Li W.. Identification of ribosomal RNA genes in metagenomic fragments. Bioinformatics. 2009;25(10):1338–1340. doi: 10.1093/bioinformatics/btp161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Eddy S. R.. Accelerated Profile HMM Searches. PLoS Comput. Biol. 2011;7(10):e1002195. doi: 10.1371/journal.pcbi.1002195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Bolyen E., Rideout J. R., Dillon M. R., Bokulich N. A., Abnet C. C., Al-Ghalith G. A., Alexander H., Alm E. J., Arumugam M., Asnicar F., Bai Y., Bisanz J. E., Bittinger K., Brejnrod A., Brislawn C. J., Brown C. T., Callahan B. J., Caraballo-Rodriguez A. M., Chase J., Cope E. K., Da Silva R., Diener C., Dorrestein P. C., Douglas G. M., Durall D. M., Duvallet C., Edwardson C. F., Ernst M., Estaki M., Fouquier J., Gauglitz J. M., Gibbons S. M., Gibson D. L., Gonzalez A., Gorlick K., Guo J., Hillmann B., Holmes S., Holste H., Huttenhower C., Huttley G. A., Janssen S., Jarmusch A. K., Jiang L., Kaehler B. D., Kang K. B., Keefe C. R., Keim P., Kelley S. T., Knights D., Koester I., Kosciolek T., Kreps J., Langille M. G. I., Lee J., Ley R., Liu Y. X., Loftfield E., Lozupone C., Maher M., Marotz C., Martin B. D., McDonald D., McIver L. J., Melnik A. V., Metcalf J. L., Morgan S. C., Morton J. T., Naimey A. T., Navas-Molina J. A., Nothias L. F., Orchanian S. B., Pearson T., Peoples S. L., Petras D., Preuss M. L., Pruesse E., Rasmussen L. B., Rivers A., Robeson M. S. . 2nd, Rosenthal P., Segata N., Shaffer M., Shiffer A., Sinha R., Song S. J., Spear J. R., Swafford A. D., Thompson L. R., Torres P. J., Trinh P., Tripathi A., Turnbaugh P. J., Ul-Hasan S., van der Hooft J. J. J., Vargas F., Vazquez-Baeza Y., Vogtmann E., von Hippel M., Walters W., Wan Y., Wang M., Warren J., Weber K. C., Williamson C. H. D., Willis A. D., Xu Z. Z., Zaneveld J. R., Zhang Y., Zhu Q., Knight R., Caporaso J. G.. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol. 2019;37(8):852–857. doi: 10.1038/s41587-019-0209-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Lu J., Rincon N., Wood D. E., Breitwieser F. P., Pockrandt C., Langmead B., Salzberg S. L., Steinegger M.. Metagenome analysis using the Kraken software suite. Nat. Protoc. 2022;17(12):2815–2839. doi: 10.1038/s41596-022-00738-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Callahan B. J., McMurdie P. J., Rosen M. J., Han A. W., Johnson A. J. A., Holmes S. P.. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods. 2016;13(7):581–583. doi: 10.1038/nmeth.3869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Quast C., Pruesse E., Yilmaz P., Gerken J., Schweer T., Yarza P., Peplies J., Glockner F. O.. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41(D1):D590–D596. doi: 10.1093/nar/gks1219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Madden T.. The BLAST sequence analysis tool. NCBI Handb. 2013;2(5):425–436. [Google Scholar]
  59. Nurk S., Meleshko D., Korobeynikov A., Pevzner P. A.. metaSPAdes: a new versatile metagenomic assembler. Genome Res. 2017;27(5):824–834. doi: 10.1101/gr.213959.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Uritskiy G. V., DiRuggiero J., Taylor J.. MetaWRAP-a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome. 2018;6(1):1–13. doi: 10.1186/s40168-018-0541-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Parks D. H., Imelfort M., Skennerton C. T., Hugenholtz P., Tyson G. W.. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 2015;25(7):1043–1055. doi: 10.1101/gr.186072.114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Olm M. R., Brown C. T., Brooks B., Banfield J. F.. dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. ISME J. 2017;11(12):2864–2868. doi: 10.1038/ismej.2017.126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Kanehisa M., Sato Y., Morishima K.. BlastKOALA and GhostKOALA: KEGG tools for functional characterization of genome and metagenome sequences. J. Mol. Biol. 2016;428(4):726–731. doi: 10.1016/j.jmb.2015.11.006. [DOI] [PubMed] [Google Scholar]
  64. Kanehisa M., Sato Y., Kawashima M., Furumichi M., Tanabe M.. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2016;44(D1):D457–D462. doi: 10.1093/nar/gkv1070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Milanese A., Mende D. R., Paoli L., Salazar G., Ruscheweyh H. J., Cuenca M., Hingamp P., Alves R., Costea P. I., Coelho L. P., Schmidt T. S. B., Almeida A., Mitchell A. L., Finn R. D., Huerta-Cepas J., Bork P., Zeller G., Sunagawa S.. Microbial abundance, activity and population genomic profiling with mOTUs2. Nat. Commun. 2019;10(1):1014. doi: 10.1038/s41467-019-08844-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

es4c05624_si_001.pdf (298KB, pdf)
es4c05624_si_002.xlsx (132.8KB, xlsx)

Data Availability Statement

All data are available in the main text or the Supporting Information. The raw metatranscriptomic sequencing data, assembled contigs, and MAGs generated in this study have been deposited in the NCBI database under the accession number PRJNA1109759. We have archived the non-rRNA metatranscriptomes and MAG data sets in the Zenodo repository. The data sets are accessible through the following Digital Object Identifier (DOI): https://zenodo.org/records/13381952 and https://zenodo.org/records/14230125. Furthermore, the analysis codes are publicly available at https://github.com/heyinghui22/Codes-for-In-situ-sampling-uncovers-seasonal-variability-of-active-deep-sea-microbes-.

Articles from Environmental Science & Technology are provided here courtesy of American Chemical Society

RESOURCES