Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
1
Uncultivated DPANN archaea are ubiquitous inhabitants of global oxygen deficient zones with 1
diverse metabolic potential 2
3
Irene H. Zhanga#*, Benedict Borera, Rui Zhaoa, Steven Wilbertb, Dianne K. Newmanb, and 4
Andrew R. Babbina*
5
aDepartment of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of 6
Technology, Cambridge, MA, USA 7
bDivisions of Biology and Biological Engineering and Geological and Planetary Sciences, 8
California Institute of Technology, Pasadena, California, USA 9
10
Running Head: DPANN archaea in global marine oxygen deficient zones 11
12
#Address correspondence to Irene H. Zhang, izhang@mit.edu and Andrew R. Babbin, 13
babbin@mit.edu 14
*Present address: Department of Biological Sciences–Marine and Environmental Biology, 15
University of Southern California, Los Angeles, CA, USA 16
17
18
abstract word count: 222 19
importance statement word count: 146 20
text word count: 4996 21
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
2
Abstract 22
Archaea belonging to the DPANN superphylum have been found within an expanding number of 23 environments and perform a variety of biogeochemical roles, including contributing to carbon, 24 sulfur, and nitrogen cycling. Generally characterized by ultrasmall cell sizes and reduced 25 genomes, DPANN archaea may form mutualistic, commensal, or parasitic interactions with 26 various archaeal and bacterial hosts, influencing the ecology and functioning of microbial 27 communities. While DPANN archaea reportedly comprise 15–26% of the archaeal community 28 within marine oxygen deficient zone (ODZ) water columns, little is known about their metabolic 29 capabilities in these ecosystems. We report 33 novel metagenome-assembled genomes belonging 30 to DPANN phyla Nanoarchaeota, Pacearchaeota, Woesarchaeota, Undinarchaeota, Iainarchaeota, 31 and SpSt-1190 from pelagic ODZs in the Eastern Tropical North Pacific and Arabian Sea. We 32 find these archaea to be permanent, stable residents of all 3 major ODZs only within anoxic 33 depths, comprising up to 1% of the total microbial community and up to 25–50% of archaea. 34 ODZ DPANN appear capable of diverse metabolic functions, including fermentation, organic 35 carbon scavenging, and the cycling of sulfur, hydrogen, and methane. Within a majority of ODZ 36 DPANN, we identify a gene homologous to nitrous oxide reductase. Modeling analyses indicate 37 the feasibility of a nitrous oxide reduction metabolism for host-attached symbionts, and the small 38 genome sizes and reduced metabolic capabilities of most DPANN MAGs suggest host-39 associated lifestyles within ODZs. 40 41 Importance 42 43 Archaea from the DPANN superphylum have diverse metabolic capabilities and participate in 44 multiple biogeochemical cycles. While metagenomics and enrichments have revealed that many 45 DPANN are characterized by ultrasmall genomes, few biosynthetic genes, and episymbiotic 46 lifestyles, much remains unknown about their biology. We report 33 new DPANN metagenome-47 assembled genomes originating from the 3 global marine oxygen deficient zones (ODZs), the 48 first from these regions. We survey DPANN abundance and distribution within the ODZ water 49 column, investigate their biosynthetic capabilities, and report potential roles in the cycling of 50 organic carbon, methane, and nitrogen. We test the hypothesis that nitrous oxide reductases 51 found within several ODZ DPANN genomes may enable ultrasmall episymbionts to serve as 52 nitrous oxide consumers when attached to a host nitrous oxide producer. Our results indicate 53 DPANN archaea as ubiquitous residents within the anoxic core of ODZs with the potential to 54 produce or consume key compounds. 55 56
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
3
Introduction 57
In recent years, metagenomics has enabled the discovery of several prokaryotic 58
superphyla lacking pure culture representatives (1–3). One of these novel groups is the DPANN 59
archaea, named after the first members of the expanding superphylum (Diapherotrites, 60
Parvarchaeota, Aenigmarchaeota, Nanoarchaeota and Nanohaloarchaeota) which has come to 61
include at least ten putative phyla (4, 5). The DPANN archaea are characterized by ultrasmall 62
cell sizes (~0.1–1.5 µm), reduced genomes (~1.5 Mb), and limited metabolic capacities (6). 63
These features, along with several enrichments and visualizations of DPANN archaeal-host 64
associations (7–9), suggest a symbiotic or commensal lifestyle of DPANN archaea with diverse 65
microbial hosts. If DPANN indeed exist in partnership with others, this would explain why they 66
have been challenging to cultivate in isolation. 67
Since their discovery, DPANN archaea have been found in a variety of diverse 68
environments, including hydrothermal vents (10), freshwater and hypersaline lakes (11, 12), 69
groundwater (13, 14), terrestrial hot springs (15), marine sediments and water columns (10, 16, 70
17), and the Black Sea (18). Archaea writ large play crucial roles in global biogeochemical 71
cycles, such as in ammonia oxidation (19), methane cycling (20), and organic carbon scavenging 72
(21), and DPANN archaea have been found to possess genes for sulfur cycling and organic 73
substrate degradation (13, 16). Additionally, DPANN archaea in anoxic environments may form 74
consortia with methanogens and contribute to anaerobic carbon cycling (22). However, despite 75
their widespread abundance, distribution, and diversity (accounting for about half of all archaeal 76
diversity (6)), the ecological and biogeochemical roles of DPANN archaea are not fully 77
understood. Culture-independent techniques have only begun to unravel the importance of these 78
previously-overlooked microorganisms within their expanding list of habitats. 79
Amplicon surveys have detected the presence of DPANN archaea within both sediments 80
beneath oxygen deficient zones (ODZs) (23) and the ODZ water column itself (24). The three 81
major oceanic ODZs are located in the eastern tropical North Pacific (ETNP), the eastern tropical 82
South Pacific (ETSP), and the Arabian Sea. Oxygen profiles in these regions display rapid 83
decreases from surface saturation to below the detection limit of trace oxygen sensors (<10 nmol 84
L-1) between 50–100 m depth, a region termed the oxycline (25, 26). Oxygen concentrations then 85
remain below detection and with no vertical gradient for approximately 200–800 m (27), 86
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
4
although the ODZ thickness varies greatly across each basin (28–30). Due to these unique 87
features, ODZ water columns contain multiple biogeochemical gradients that support diverse 88
microbial assemblages performing nitrogen, carbon, and sulfur cycling (31). In particular, these 89
regions disproportionately contribute to marine nitrogen cycling, accounting for about 30% of 90
marine fixed nitrogen loss despite containing only 0.1–0.2% of oceanic volume (32, 33). 91
ODZs are characterized by prevalent denitrification, i.e. the microbially-mediated 92
stepwise reduction of nitrate to dinitrogen gas. This anaerobic respiratory metabolism occurs via 93
reductases encoded by a suite of widely distributed genes (34). The last step of denitrification, 94
the reduction of N2O to N2, is catalyzed by nitrous oxide reductase encoded by nos. Two clades 95
of the nos catalytic subunit nosZ have been found, a typical clade I nosZ associated with 96
complete denitrifiers defined by an N-terminal twin-arginine translocation (TAT) motif, and an 97
atypical clade II nosZ associated with partial denitrifiers defined by an N-terminal Sec-type motif 98
(35). Both variants contain conserved copper-binding sites CuA and CuZ, although CuZ sites of 99
clade II nosZ homologs exhibit greater variability and less conservation (36). Recent studies 100
reveal clade II nosZ predominates within ODZs, occurs within diverse marine taxa including 101
archaea, and may be associated with low oxygen and enhanced N2O affinity (36). Because N2O 102
depletes ozone and is a potent greenhouse gas, organisms with atypical nosZ variants, including 103
archaea, merit interest as potential N2O sinks. 104
Increasing attention has been focused on ODZ archaeal communities (37–39), such as 105
members of Thermoproteota (including former Marine Group I Thaumarchaeota) and 106
Thermoplasmatota (including former Marine Group II archaea) (40, 41). However, little is 107
known about ODZ DPANN archaea, despite reports that they may comprise up to 15–26% of 108
total archaeal reads in these regions (24). Challenges in cultivation of these environmental 109
microbes limit our understanding of the metabolic capabilities of clades such as DPANN that 110
lack cultured representatives. Accordingly, the contribution of DPANN archaea in ODZ 111
microbial assemblages and biogeochemical cycling, as well as the abundance, distribution, 112
metabolism, ecology, and phylogeny of these archaea remain open questions. Using genome-113
resolved metagenomics, we recover 33 genomes belonging to DPANN phyla Nanoarchaeota, 114
Pacearchaeota, Woesarchaeota, Undinarchaeota, and Iainarchaeota from the ETNP and Arabian 115
Sea ODZs. We characterize the metabolic capabilities of these archaea, place them within the 116
existing phylogeny of known DPANN, and determine their relative abundances and distributions 117
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
5
within and across global ODZs. Our results demonstrate that DPANN are a ubiquitous portion of 118
the microbial community within ODZs and comprise several lineages with diverse metabolic 119
potential. 120
Materials and Methods 121
Sample collection, sequencing, metagenome assembly, and binning 122
Sampling and sequencing methods for public ETNP metagenomes are described in Fuchsman et 123
al. 2017 (39), Glass et al. 2015 (42), and Tsementzi et al. 2016 (43). Sampling and sequencing 124
methods for public ETSP metagenomes are described in Stewart et al. 2012 (44) and Ganesh et 125
al. 2014 (45). Raw reads per metagenome were retrieved from the Sequence Read Archive 126
(SRA) using the following NCBI BioProject IDs: PRJNA350692 (Fuchsman ETNP 127
metagenomes), PRJNA254808 (Glass ETNP metagenomes), PRJNA323946 (Tsementzi ETNP 128
metagenomes), PRJNA68419 (Stewart ETSP metagenomes), and PRJNA217777 (Ganesh ETSP 129
metagenomes. Sampling locations for each metagenome were visualized using Python 3.7.12 and 130
the cartopy package. These were plotted against global oxygen concentrations from 300 m below 131
sea surface from Ocean Data Atlas 2018 (Figure 1A). 132
Trimming of raw reads, metagenome assembly, and binning methods are described 133
elsewhere (46). Metagenome-assembled genomes (MAGs) were defined as bins with completion 134
>50% and contamination <10% according to CheckM (47), although these statistics based on 135
single-copy genes may underestimate the true completeness of DPANN archaea MAGs due to 136
their limited genome sizes. Taxonomy was assigned to all MAGs using GTDB-tk v1.7.0 with the 137
classify_wf workflow (48). 33 MAGs belonging to DPANN phyla were annotated with 138
PROKKA v1.14.6 (49) against the HAMAP (50) and Pfam databases (51) using the --kingdom 139
Archaea flag. The full set of ODZ MAGs, as well as all DPANN MAGs generated in this study, 140
were deposited under NCBI BioProject ID PRJNA955304. 141
Published DPANN MAGs and genomes were manually downloaded from the Joint 142
Genome Institute (JGI), while published DPANN from the National Center for Biotechnology 143
Information (NCBI_ were downloaded using the EntrezDirect utility. These DPANN MAGs and 144
genomes were assessed for completeness and contamination with CheckM v1.0.12 (47), and 145
detailed taxonomy was determined with GTDB-tk v1.7.0 (48). MAGs and genomes below 50% 146
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
6
completion and above 10% contamination, along with those that did not taxonomically classify 147
within DPANN phyla, were pruned and remaining genomes were dereplicated with dRep v3.2.2 148
(52) with the -sa 0.99 flag to remove redundant genomes. 149
TARA Oceans MAGs were retrieved from Delmont et al. 2018 (53). To determine if 150
DPANN MAGs were present within the TARA Oceans collection, we reclassified the 957 non-151
redundant TARA Oceans MAGs with GTDB-tk v1.7.0 (48). These taxonomies were then 152
searched for the presence of any DPANN phyla. 153
For dereplicated ODZ DPANN MAGs, coverage mapping was performed with CoverM 154
using the flags minimap2-sr --min-read-aligned-percent 50 --min-read-percent-identity 0.95 --155
min-covered-fraction 0 (https://github.com/wwood/CoverM). Relative abundances of 156
dereplicated MAGs resulting from CoverM mapping were visualized using R v4.1.3 and the 157
packages phyloseq, ggplot2, and dplyr. 158
Gene searching, metabolic analysis, and tree building 159
Unique published DPANN and all ODZ DPANN MAGs were queried for 76 archaea-specific 160
single copy genes, which were aligned using GtoTree v1.6.31 with the -H Archaea -G 0.25 flags 161
(54). We created a phylogenetic tree based on the output archaeal single copy gene alignment 162
with IQ-Tree v1.6.12 (55) using the WAG+R6 model and 1000 ultrafast bootstraps (56). 163
To determine metabolic capabilities of ODZ DPANN, we used Anvi’o 7.1 (57). Briefly, 164
for each DPANN MAG we generated a contigs database with anvi-gen-contigs-database. For 165
metabolic predictions, we ran anvi-run-kegg-kofams to search against the KOfam HMM 166
database of Kyoto Encyclopedia of Genes and Genomes (KEGG) orthologs (58) and 167
automatically assign hits above the KEGG bitscore thresholds for each KOfam profile. 168
Additionally, we ran anvi-run-ncbi-cogs to search against the NCBI Clusters of Orthologous 169
Groups (COGs) database (59) and identified archaeal single-copy core genes using anvi-run-170
hmms -I Archaea 76. To predict the presence or absence of metabolic pathways, we ran anvi-171
estimate-metabolism on each MAG. We annotated a metabolic pathway as present if over 70% 172
of the genes in the pathway are present in a MAG, and partially present if 33–70% of the genes 173
in a pathway are present in a MAG. Additionally, we searched for annotations of genes of 174
interest within PROKKA annotations for each MAG, particularly for genes involved in 175
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
7
fermentation, aerobic or anaerobic respiration, and energy metabolism. Sequences belonging to 176
genes of interest were retrieved from each MAG and further inspected. 177
Protein sequences belonging to positive hits for denitrification genes from ODZ DPANN 178
MAGs were obtained for nosZ. We extracted and aligned with MAFFT v7.450 using the --auto 179
and --leavegappyregion parameters. Alignments were visualized in JalView 2.11.2.6 (60) and 180
inspected for alignment quality and the conservation of key enzymatic regions for nosZ. 181
Prediction of membrane-bound regions, protein localization, and protein structure were 182
determined via DeepTMHMM (61). To create a protein tree for nosZ, bacterial and archaeal 183
nosZ-encoded protein sequences were obtained from NCBI using the query esearch -db protein -184
query "NosZ" | efetch -format fasta, and sequences under 200 amino acids and over 800 amino 185
acids were removed. In addition, cytochrome c oxidase subunit II proteins from bacteria and 186
archaea were downloaded from NCBI using the queries esearch -db protein -query "cytochrome 187
c oxidase subunit ii [PROT] AND bacteria [ORGN]" | efetch -format fasta and esearch -db 188
protein -query "cytochrome c oxidase subunit ii [PROT] AND archaea [ORGN]" | efetch -format 189
fasta. Sequences under 200 amino acids and over 800 amino acids were removed. To remove 190
redundant or very similar sequences, Usearch v11 was used to cluster NCBI nosZ and Cox2 191
sequences at 90% amino acid identity with the flags -cluster_fast -id 0.9 -centroids (62). From 192
clustered Cox2 sequences, 15 Cox2 sequences from taxonomically diverse organisms were 193
chosen at random. These selected Cox2 were concatenated with clustered nosZ sequences and 194
DPANN nosZ sequences and aligned with MAFFT v7.450 (63) using the --auto and --195
leavegappyregion flags. The resulting Cox2 and nosZ alignment was trimmed with trimAl 1.4.1 196
with the -automated1 flag (64). We used the trimmed alignment to create a maximum likelihood 197
protein tree using IQ-Tree v1.6.12 with 1000 ultrafast bootstraps. 198
Methods for modeling producer and consumer dynamics 199
We used COMSOL (v5.6) to simulate the concentration field and associated uptake rate around a 200
two-cell system consisting of a producer and consumer in three-dimensional space. In this 201
simulation, the producer cell is represented as a sphere with a constant normalized concentration 202
of 1 on its surface. The consumer cell on the other hand is represented by a sphere with a 203
normalized concentration of 0 on its surface. We represent all aqueous concentrations as relative 204
concentrations between producer and consumer cells to keep the system universal. We then 205
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
8
strategically vary the relative radius of the producer (R) and consumer (r) cells, and the distance 206
between the surface of the two cells (d) to disentangle the influence of these different factors on 207
the relative substrate uptake of the consumer cell. A list of parameters and their values can be 208
found in Table S1 which we cross-combine to create a total of 100 simulations. The whole 209
simulation domain is a square domain of 20 µm side length with the consumer and producer cells 210
equidistant from the center in the horizontal plane. All simulations are solved for steady-state 211
concentration fields and the total uptake rate of the consumer cell is calculated directly in 212
COMSOL. 213
Genetic engineering methods and determining N2O concentrations 214
We inserted DNA sequences derived from DPANN putative nitrous oxide reductase genes into a 215
Pseudomonas aeruginosa strain PA14 model system on a plasmid integrated into the genomic 216
attTn7 site (65). Production and consumption of N2O by these cultures was quantified using a 217
microelectrode (Unisense, Denmark). Details are in Supplementary Methods. 218
Results 219
DPANN within the ODZ archaeal community 220
From a set of 962 MAGs >50% completion and <10% contamination binned from the ETNP and 221
Arabian Sea ODZs (46), 33 MAGs were taxonomically assigned to the DPANN superphylum, 222
with 23 Woesarchaeota, 2 Pacearchaeota, 2 Nanoarchaeota, 1 Iainarchaeota, 3 Undinarchaeota, 223
and 2 MAGs assigned to SpSt-1190, also known as Candidatus Altiarchaeota. The novel SpSt-224
1190 phylum was previously characterized in hydrothermal vents (10) but not in marine water 225
columns. While our Woesearchaeota and Pacearchaeota MAGs were classified by GTDB-tk as 226
members of Nanoarchaeota, phylogenetic analyses confirm their placement within these phyla 227
(Figure 2). DPANN MAGs mapped to ODZ metagenomes within all three ODZs, including 228
ETSP and ETNP metagenomes spanning multiple cruises, sampling sites, and years (Figure 1A). 229
However, no DPANN MAGs were recovered from oxygenated surface metagenomes from the 230
ETNP. Searching the TARA Oceans dataset comprising 957 non-redundant MAGs from co-231
assemblies from the global surface oceans (<10 m depth) and deep chlorophyll maxima (10–100 232
m depth) reveal no MAGs belonging to ODZ DPANN groups, and only 2 DPANN MAGs, both 233
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
9
of which originated from the Red Sea and were assigned to Halobacteriota. The remainder of the 234
87 archaeal MAGs from TARA Oceans were assigned to either Thermoplasmatota or 235
Thermoproteota. From 962 ODZ MAGs, 169 archaeal MAGs include 133 Thermoplasmatota or 236
Thermoproteota, 2 Hydrothermarchaeota, and 1 Methanobacteriota to complement the 33 237
DPANN archaea. The average completion of retrieved ODZ DPANN MAGs was 75% with an 238
average contamination of 2.6%. Dereplication at 99% average nucleotide identity (ANI) resulted 239
in 16 unique MAGs (1 Iainarchaeota, 2 Nanoarchaeota, 1 SpSt-1190, 3 Undinarchaeota, 1 240
Pacearchaeota, and 8 Woesarchaeota). 241
The archaeal population within ODZs, as represented by mapping to non-redundant 242
archaeal ODZ MAGs, peaks within the oxycline above the ODZ core, with archaeal MAGs 243
comprising 14% of the microbial community at 80 m depth in the ETNP (Figure S1A). At these 244
depths, Thermoplasmatota and Thermoproteota dominate, with DPANN MAGs present at 0.25% 245
relative abundance or less. DPANN MAGs also mapped to few reads from surface metagenomes 246
from either the ETNP or ETSP, indicating low or absent populations of ODZ DPANN groups in 247
surface waters (Figure 1B). No surface waters from the Arabian Sea were sampled. Within the 248
anoxic ODZ core, DPANN archaea comprise about 25–50% of the archaeal community (Figure 249
S1B). Highest relative abundances occur around 200 m depth in the ETNP (about 1% of the total 250
community and 50% of the archaeal community) and the Arabian Sea (about 0.8% of the total 251
community and 27% of the archaeal community) (Figures 1B, S1B). In the ETSP, relative 252
abundances are lower (about 0.3% of the total community and 25% of the archaeal community), 253
but peak approximately at the same depths (100–200 m). While abundances and distributions are 254
similar across the various ODZs, the Arabian Sea harbors a comparatively larger proportion of 255
Nanoarchaeota, although Woesarchaeota are still the most abundant fraction in general. The 256
ETSP and ETNP are primarily dominated by Woesarchaeota with smaller contributions by 257
Pacearchaeota, SpSt-1190, and Undinarchaeota (Figure 1B). 258
A phylogenetic tree of ODZ DPANN MAGs along with MAGs and genomes retrieved 259
from NCBI and JGI reveals that ODZ Woesarchaeota MAGs fall within one primary clade, 260
although several Woesarchaeota MAGs branch within other groups (Figure 2). Sister taxa falling 261
next to ODZ Woesarchaeota derive from Mariana Trench surficial sediments, coral reefs, and 262
groundwater metagenomes from NCBI. However, ODZ Pacearchaeota and Nanoarchaeota 263
MAGs fall within distinct clades, indicating that these MAGs, despite belonging to the same 264
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
10
phyla, are not closely related. The two SpSt-1190 MAGs from ODZs branch outside of the SpSt-265
1190 clade, and potentially form a distinct lineage from other SpSt-1190. 266
Carbon metabolism within ODZ DPANN archaea 267
A metabolic analysis of ODZ DPANN MAGs shows diverse metabolic capabilities across 268
MAGs, but limited metabolic and biosynthetic pathways within each MAG (Figure 3). Metabolic 269
capabilities described are based upon annotations against the KEGG and COGs databases and 270
require functional verification. While these annotations are predictions only, they offer estimates 271
of metabolic potential for these uncultured organisms. Regarding anabolic synthesis, MAGs 272
belonging to Nanoarchaeota have the most limited capabilities, with absent glycolysis of 3-273
carbon compounds, no TCA cycle genes, no pentose phosphate pathways, no pathways detected 274
for the biosynthesis of most amino acids, and limited biosynthetic pathways for purine 275
nucleotides. The absence of these pathways, even when considering these as partial genomes, 276
suggests extremely limited abilities to synthesize purines, amino acids, lipids, vitamins, and other 277
necessary cellular components. Other DPANN MAGs possess more metabolic capabilities, 278
although most lack evidence of complete glycolysis, TCA cycle, and pathways for synthesis of 279
multiple essential amino acids. Our draft Woesarchaeota genomes possess partial or complete 280
capabilities for the last stages of glycolysis, the non-oxidative or reductive portions of the 281
pentose phosphate pathway, and pyruvate oxidation. Additionally, most are capable of partial or 282
complete purine and pyrimidine biosynthesis. Other central carbon archaeal pathways vary, with 283
most MAGs lacking the shikimate pathway for biosynthesis of aromatic amino acids (66), the 284
biosynthesis pathway for the ubiquitous cofactor coenzyme A, and the DeLey-Douderoff 285
pathway for galactose utilization, which is analogous to the Entner-Doudoroff pathway (67). 286
Several MAGs lack the ability to synthesize the intermediate phosphoribosyl diphosphate 287
(PRPP) used in building nucleotides, some amino acids, and essential cofactors (68), as well as 288
biosynthesis pathways for isoprenoids (Figure 3). 289
In accordance with other published DPANN archaea (4, 7, 8), the genome sizes of most 290
ODZ DPANN are small, averaging 1.05 Mb. The exception are MAGs belonging to SpSt-1190, 291
which have genome sizes of 4 Mb. DPANN MAGs encode for a number of transporters, 292
including ones for zinc, iron, magnesium, and other metals, biotin transporters, SemiSWEET 293
transporters for cellular uptake and translocation of sugars, and other ABC-type transporters. 294
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
11
Peptidases, particularly signal peptidases and membrane-bound peptidases, are also widespread. 295
7 DPANN MAGs from 4 phyla contain genes for Type II or IV secretion systems associated with 296
protein transport and DNA exchange across membranes. Additionally, 3 DPANN MAGs 297
encoded a murein-like lytic transglycosylase (1, 69). Normally absent in archaea, these large 298
proteins bind and degrade peptidoglycan strands such as in bacterial cell walls (70). 299
Several DPANN MAGs possess the 3-oxoacyl-ACP reductase FabG, enoyl-ACP 300
reductase FabI, and 3-hydroxyacyl-ACP dehydratase FabZ. These acyl carrier protein (ACP) 301
fatty acid biosynthesis genes are typically found within bacteria and eukaryotes, which possess a 302
bacterial pathway for lipid biosynthesis, the methylerythritol phosphate (MEP) pathway, while 303
typical archaea use the non-homologous mevalonate (MVA) pathway. This “lipid divide” is a 304
central distinguishing feature between archaea and bacteria (3). We find a distinction between 305
Woesarchaeota and Nanoarchaeota MAGs, which possess the MEP pathway, while SpSt-1190, 306
Undinarchaeota, Pacearchaeota, and Iainarchaeota contain genes for the MVA pathway. Further 307
BLAST searching of DPANN ACP pathway proteins against the NCBI non-redundant protein 308
database reveal high sequence similarity to other protein sequences from DPANN archaea, 309
although the next-highest scoring hits primarily belong to bacteria. Additionally, 5 310
Woesarchaeota DPANN MAGs carry genes for cyclopropane fatty acid phospholipid synthesis, 311
which are used in stabilizing bacterial phospholipid membranes (71) but have not been 312
previously reported in archaea. 313
Energy metabolism and nutrient cycling within ODZ DPANN archaea 314
Lactate, malate, and pyruvate dehydrogenases were present in 7 out of 8 unique Woesarchaeota 315
and 2 out of 3 unique Undinarchaeota MAGs, indicating fermentative capabilities (Figure 3). 316
These genes were absent in most Pacearchaeota, Nanoarchaeota, SpSt-1190, and Iainarchaeota 317
MAGs. Two DPANN MAGs also carried the A and B subunits of assimilatory anaerobic sulfite 318
reductase, but none carried dissimilatory sulfur cycling genes. Additionally, several DPANN 319
MAGs contain desulfoferredoxin, manganese superoxide dismutases, and thioredoxin, which are 320
involved in antioxidant systems (14), despite living in anoxic water columns. Formaldehyde 321
assimilation genes were found within a number of Woesarchaeota MAGs, suggesting the ability 322
to use one-carbon compounds for growth. MAGs belonging only to SpSt-1190 also encoded 323
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
12
nearly-complete pathways for methanogenesis, along with a number of other methane 324
metabolisms including the ribulose monophosphate pathway and methanofuran biosynthesis. 325
Several Woesarchaeota MAGs encoded hydrogenases, with 1 MAG encoding an FeFe-326
type hydrogenase potentially used in fermentative metabolism, while 2 encode an NiFe-type 327
hydrogenase that may catalyze hydrogen oxidation for energy, which has recently been shown to 328
be widespread among archaea (71) and marine bacteria (72). Additionally, several ODZ DPANN 329
contain genes for urea cycling. These metabolic capabilities indicate diverse roles in carbon, 330
sulfur, hydrogen, and nitrogen cycling for DPANN archaea within ODZs. 331
Potential nitrous oxide reduction capability within ODZ DPANN 332
Within our 33 DPANN MAGs, 21 encoded a gene annotated as nitrous oxide reductase (nosZ), 333
which catalyzes the reduction of N2O to N2. An HMM search against the HMM profile from 334
validated nosZ sequences returns expectation values between 8.4 × 10-17 and 2.2 × 10-5 and bit 335
scores from 57.9–20.1, compared to canonical nosZ e-values of less than 1.2 × 10-33 and bit 336
scores > 113. In comparison, cytochrome c oxidase subunit II proteins returned expectation 337
values of 0.003 – 7.5 × 10-5 and bit scores of 13–18.4. While bit score cutoffs vary, a bit score 338
>50 is considered almost always significant (73). 339
However, other denitrification genes were absent within these MAGs. A gene tree built 340
with canonical nosZ from bacteria and archaea, the DPANN nosZ-like protein, and the closely 341
related homologue cytochrome c oxidase subunit II protein (Cox2) indicates DPANN nosZ-like 342
genes comprise a monophyletic clade branching in between Cox2 and clade II Sec-type nosZ 343
(Figure 4). Further investigation of multiple-sequence protein alignments of nosZ-encoded 344
nitrous oxide reductase and the DPANN nosZ-like protein reveal the presence of a conserved 345
copper-binding site, the CuA site, which has been reported within nosZ and Cox2 proteins. This 346
site is exemplified by the C1X3C2X3H binding motif (74–76). The CuZ catalytic site typically 347
found within nosZ is not found within the DPANN nosZ-like proteins, and DPANN nosZ-like 348
proteins are shorter (56–617 amino acids) than canonical nosZ proteins (200–796 amino acids, 349
although sequences vary in completeness). The CuZ site, which lacks a specific conserved motif, 350
is characterized by 7 conserved histidine residues (35, 76, 77). Within 2 DPANN MAGs, we find 351
a short cupredoxin-like domain protein directly upstream of the nosZ-like protein containing 5 352
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
13
conserved histidine residues, which clusters with clade II nosZ sequences containing the CuZ site 353
within the protein phylogeny (Figure 4). 354
The mature NosZ protein resides in the periplasmic space in known denitrifiers (35). 355
Predictions of protein location for NCBI nosZ sequences and the DPANN nosZ-like gene 356
indicate both contain a signal sequence followed by the majority of the protein located outside 357
the inner membrane, perhaps indicating function more similar to nosZ (Figure S2). In contrast, 358
the Cox2 protein contains two transmembrane regions. A heterologous complementation test 359
performed by separately introducing 3 DPANN nosZ-like sequences into a Pseudomonas 360
aeruginosa
∆
nosZ mutant did not yield significant differences in N2O consumption between P. 361
aeruginosa
∆
nosZ with DPANN nosZ-like gene insertion vs. the
∆
nosZ parent strain (Figure S3). 362
One strain carrying a putative DPANN nosZ-like gene variant displayed reduced N2O 363
concentrations compared to the
∆
nosZ parent, indicating potential N2O consumption, but this 364
difference was not statistically significant (p = 0.26). Our positive control, the wild type P. 365
aeruginosa PA14 containing a functional nosZ gene, displayed reduced N2O concentrations (p < 366
0.001), validating the ability of our experimental system to detect N2O accumulation. 367
N
2O in the ODZ typically exists at bulk nanomolar concentrations (78), raising the 368
question of whether specialization on N2O consumption is metabolically feasible. However, local 369
N2O concentrations may vary in the presence of N2O producers such as partial denitrifiers 370
carrying upstream denitrification capabilities. We simulate the conditions under which local N2O 371
concentrations differ from bulk conditions by varying a set of parameters representing the 372
distance between the N2O consumer and N2O producer and the size ratio of the two cells (Figure 373
5). Generally, two conditions favor elevated N2O uptake rates for the consumer normalized to 374
cell volume: when the consumer cell is small relative to the producer cell (Figure 5B) and when 375
the distance between the producer and consumer cells is small (Figure 5C). Rates are normalized 376
to cell volume to reflect the important consideration that resource requirements are proportional 377
to cell size (79). Consumer cell size has a decisive influence on N2O uptake rate as a smaller cell 378
size both increases the uptake rate normalized to cell volume and a large producer-to-consumer 379
size ratio surrounds the consumer cell within the diffusive boundary layer of the producer cell. 380
For example, an increase in the ratio of consumer-to-producer radii from ~ 0.1 to ~1 results in an 381
average 100-fold decrease in N2O uptake for attached consumer cells. Similarly, increasing the 382
producer cell size three-fold increases the attached consumer N2O uptake rate between ~14% for 383
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
14
small consumer cells to 67% for larger consumer cells. Incredibly, increasing distances between 384
cells from 0 µm to 0.1 µm reduces the maximal N2O uptake rate by ~65% on average, and a 385
consumer cell merely 2 µm from a producer receives 93% less than those attached (Figure 5C). 386
Therefore, when bulk N2O is low as in the ODZ (80), consumer cells experience high N2O supply 387
only when they are in physical contact with an N2O producer (d = 0 µm) and when they are 388
much smaller in size relative to the producer, as would be the case for episymbiotic DPANN 389
archaea. 390
Discussion 391
DPANN archaea were found to be a stable resident population within all 3 permanent pelagic 392
ODZs. Abundances of DPANN archaea, including Nanoarchaeota, SpSt-1190, Iainarchaeota, 393
Woesarchaeota, and Undinarchaeota, increase as oxygen decreases, while few or no DPANN 394
archaea were found in the surface oceans (Figure 1B). While a few population differences are 395
found between ODZs, Woesarchaeota are the dominant phylum within all three ODZs, with 396
Nanoarchaeota in the Arabian Sea and Pacearchaeota, Undinarchaeota, and SpSt-1190 in the 397
ETNP and ETSP forming the second most abundant groups (Figure 1B). 398
ODZ DPANN archaea are phylogenetically and metabolically diverse and group together 399
with other DPANN from non-ODZ environments, although several Woesarchaeota cluster within 400
the same clade (Figure 2). Similar to DPANN across various environments (4, 10, 12, 13), most 401
ODZ DPANN have small genome sizes and limited capacity for biosynthesis of essential amino 402
acids and nucleotides, limited energetic capabilities, and partial or absent pentose phosphate 403
pathways despite overall high MAG completion estimates (Figure 3). Additionally, completion 404
metrics may underestimate the completeness of DPANN MAGs due to their limited genomes 405
and high number of absent genes considered essential in other organisms. Numerous studies have 406
reported microscopy images of environmental DPANN attached to host cells (7, 13, 81). While 407
most ODZ DPANN genomes suggest a host-associated rather than free-living lifestyle, SpSt-408
1190 genomes average 4 Mb in size, possess a number of biosynthesis pathways, and carry 409
pathways for methanogenesis. These unique archaea may represent free-living DPANN 410
organisms (82–84) involved in methane cycling. The identification of ODZ DPANN hosts, 411
whether a single host, various hosts, or a community, may hold keys to their distribution and 412
survival within ODZ environments. 413
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
15
Studies have suggested a role of DPANN archaea for carbon cycling, such as by 414
scavenging organic carbon in the form of nucleotides, lipids, and amino acids (21, 85), 415
participating in the exchange of carbon compounds with hosts, and even directly parasitizing 416
upon hosts (12). In addition, some may perform fermentation and consume or produce acetate 417
(6). We find conserved pathways for amino acid salvage and fermentation across ODZ DPANN 418
genomes (Figure 3). While various sugar, protein, and DNA transporters indicate potential 419
resource exchange with host cells, the existence of a peptidoglycan-degrading enzyme and 420
secreted peptidases within several MAGs may point to a potentially parasitic relationship 421
between host and DPANN cell. Future experimental tests will be needed to clarify these 422
metagenomic predictions. 423
While other nitrogen cycling genes are absent, a majority of ODZ DPANN carry a gene 424
similar to the nitrous oxide reductase gene nosZ that catalyzes the reduction of nitrous oxide 425
(N2O) to N2. Further investigation of this gene, annotated as nitrous oxide reductase, indicates 426
the presence of a conserved CuA copper-binding site typical of nosZ and cytochrome c oxidase 427
subunit II (75, 76) (Figure 4). The cellular location of the protein product of the DPANN nosZ-428
like gene is postulated as outside of the membrane, possibly in the periplasmic space (Figure S2). 429
DPANN archaea are thought to possess two membranes (86), and canonical nosZ is a 430
periplasmic protein unlike the membrane-bound cytochrome c oxidase subunit II (74). 431
Cytochrome c oxidase performs the last step of aerobic respiration, but no other elements of 432
aerobic respiration were found within these archaea (Figure 3). The CuZ catalytic center, 433
typically found upstream of the CuA center in nosZ, is absent within DPANN nosZ-like genes. 434
The CuZ center lacks a consensus motif, but is characterized by 7 histidine residues that bind 435
copper ions (87). While the majority of DPANN MAGs possess several acyl carrier protein 436
genes for fatty acid biosynthesis surrounding the nosZ-like gene, 2 DPANN MAGs encode a 437
protein containing 5 histidine residues directly upstream of the nosZ-like gene. This protein, 438
annotated to the same family as nosZ, groups phylogenetically with clade II nosZ sequences 439
(Figure 4) and may perform a function related to that of the CuZ site. This hypothetical histidine-440
rich region was absent within other DPANN MAGs, and was not included within the 441
complementation test. The activity of these or other proteins within these genomes may be 442
required for N2O reduction. While the function of putative nosZ-like genes within DPANN 443
archaea remain hypothetical, evidence suggests the involvement of these genes in N2O reduction 444
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
16
or another redox process with metabolic or physiological importance due to their conservation 445
within these small, streamlined genomes. 446
Complementation of P. aeruginosa
∆
nosZ with DPANN nosZ-like genes did not result in 447
significant N2O consumption. While heterologous complementation may offer convincing 448
evidence for the function of unknown genes, negative results are difficult to interpret. Large 449
evolutionary distances between DPANN archaea and the gram-negative bacterium P. 450
aeruginosa, likely resulting in different intracellular conditions, may inhibit the proper 451
transcription, translation, or maturation of the DPANN NosZ-like protein. The protein may also 452
be adapted to specific environmental conditions necessary for its activity, which differ from 453
those used during standard cultivation of P. aeruginosa. Deletion and complementation of the 454
nosZ-like gene within native DPANN archaea would be an ideal functional test, but currently no 455
cultured representatives or genetic toolkits are available for these organisms, limiting our 456
knowledge of many of their metabolic features to predictions from gene annotations. 457
N
2O exists in nanomolar concentrations in ODZs compared to the higher concentrations 458
of nitrate and nitrite (78), posing challenges for N2O-reducing specialists lacking upstream 459
denitrification genes. However, an N2O-consuming lifestyle may be feasible if local N2O 460
concentrations are elevated in proximity to an N2O source, such as a partial denitrifier lacking 461
nosZ. Previous studies have indicated widespread occurrence of partial denitrifiers lacking nosZ 462
within ODZ regions (39, 46). We tested this scenario by modeling the local flux of N2O from a 463
producer (the source) to an N2O consumer (Figure 5). N2O uptake rate of the consumer is 464
elevated 100-fold when the two cells are in physical contact vs. when they are a short distance of 465
2 µm away (Figure 5C). This increase in N2O uptake rate drops off steeply, however, as the 466
consumer-to-producer cell size ratio increases (Figure 5). Under low bulk N2O concentrations, 467
partial denitrifiers may provide elevated local N2O only to much smaller surface-attached 468
episymbiotic N2O consumers. DPANN archaea within ODZs, similar to those found within other 469
environments (7, 8, 13), potentially exist as host-associated episymbionts and likely possess 470
small cell sizes. The average cell volume of DPANN archaea has been reported as 0.004 µm3
471
(13) while average marine bacterial cell volume has been reported at up to 0.096
μ
m3 (88), 472
resulting in a consumer-to-producer cell size ratio of < 0.05. Thus, DPANN archaea may be 473
uniquely adapted to consume N2O and other resources that are scarce under bulk conditions but 474
locally elevated in proximity to host cells. 475
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
17
DPANN archaea possess a high number of unknown or unannotated genes, representing 476
“microbial dark matter.” Within our ODZ DPANN, we found over 20,000 hypothetical proteins 477
across all MAGs. Further studies, possibly using genetic manipulations, isolation or enrichment 478
cultures, imaging, and computational proteomics approaches are required to characterize the 479
functions of putative or hypothetical proteins. The expanding knowledge of these organisms may 480
make these questions more tractable in the near future. At a large scale, the scavenging of 481
carbon, potential nitrogen, sulfur, and hydrogen cycling capabilities, and ecological effects on 482
host populations via symbiosis or parasitism by DPANN archaea in the ODZs warrants future 483
investigation. 484
Acknowledgements 485
We thank Dr. Xin Sun (Carnegie Institution for Science) and Dr. Bess B. Ward, Dr. Amal 486
Jayakumar, and Dr. Samantha G. Fortin (Princeton University) for sample collection, DNA 487
extractions, and providing the resulting metagenomics data we used to assemble MAGs for this 488
study. Funding for this project came from Simons Foundation award 622065 and National 489
Science Foundation award OCE-2142998 to ARB. IZ was supported in part by an MIT School of 490
Science MathWorks Science Fellowship. Grants to SW (from the National Science Foundation 491
Graduate Fellowship Program) and to DKN (from the NIH, R01 HL152190-03) also contributed. 492
We are additionally grateful for the generosity of Dr. Bruce Heflinger in supporting the bablab, 493
including this work. 494
Author contributions 495
IHZ and ARB conceptualized this study. IHZ assembled metagenomes and MAGs, conducted 496
bioinformatics analyses, and drafted the paper. BB and ARB conceived and carried out analyses 497
regarding the N2O uptake model. RZ provided bioinformatics and overall guidance. DKN and 498
SW conceived the heterologous complementation test for the nosZ homologs and provided all 499
strains used in this study, and SW performed genetic engineering within the Pseudomonas model 500
system. 501
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
18
References 502 503 1. Rinke C, Schwientek P, Sczyrba A, Ivanova NN, Anderson IJ, Cheng J-F, Darling A, 504 Malfatti S, Swan BK, Gies EA, Dodsworth JA, Hedlund BP, Tsiamis G, Sievert SM, Liu 505 W-T, Eisen JA, Hallam SJ, Kyrpides NC, Stepanauskas R, Rubin EM, Hugenholtz P, 506 Woyke T. 2013. Insights into the phylogeny and coding potential of microbial dark matter. 507 7459. Nature 499:431–437. 508
2. Hug LA, Baker BJ, Anantharaman K, Brown CT, Probst AJ, Castelle CJ, Butterfield CN, 509 Hernsdorf AW, Amano Y, Ise K, Suzuki Y, Dudek N, Relman DA, Finstad KM, 510 Amundson R, Thomas BC, Banfield JF. 2016. A new view of the tree of life. 5. Nat 511 Microbiol 1:1–6. 512
3. Castelle CJ, Banfield JF. 2018. Major New Microbial Groups Expand Diversity and Alter 513 our Understanding of the Tree of Life. Cell 172:1181–1197. 514
4. Dombrowski N, Lee J-H, Williams TA, Offre P, Spang A. 2019. Genomic diversity, 515 lifestyles and evolutionary origins of DPANN archaea. FEMS Microbiol Lett 366:fnz008. 516
5. Dombrowski N, Williams TA, Sun J, Woodcroft BJ, Lee J-H, Minh BQ, Rinke C, Spang A. 517 2020. Undinarchaeota illuminate DPANN phylogeny and the impact of gene transfer on 518 archaeal evolution. 1. Nat Commun 11:3939. 519
6. Castelle CJ, Brown CT, Anantharaman K, Probst AJ, Huang RH, Banfield JF. 2018. 520 Biosynthetic capacity, metabolic variety and unusual biology in the CPR and DPANN 521 radiations. 10. Nat Rev Microbiol 16:629–645. 522
7. Huber H, Hohn MJ, Rachel R, Fuchs T, Wimmer VC, Stetter KO. 2002. A new phylum of 523 Archaea represented by a nanosized hyperthermophilic symbiont. 6884. Nature 417:63–67. 524
8. Baker BJ, Comolli LR, Dick GJ, Hauser LJ, Hyatt D, Dill BD, Land ML, VerBerkmoes 525 NC, Hettich RL, Banfield JF. 2010. Enigmatic, ultrasmall, uncultivated Archaea. Proc Natl 526 Acad Sci U S A 107:8806–8811. 527
9. Comolli LR, Banfield JF. 2014. Inter-species interconnections in acid mine drainage 528 microbial communities. Front Microbiol 5:367. 529
10. Cai R, Zhang J, Liu R, Sun C. 2021. Metagenomic Insights into the Metabolic and 530 Ecological Functions of Abundant Deep-Sea Hydrothermal Vent DPANN Archaea. Appl 531 Environ Microbiol 87:e03009-20. 532
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
19
11. Cabello-Yeves PJ, Zemskaya TI, Zakharenko AS, Sakirko MV, Ivanov VG, Ghai R, 533 Rodriguez-Valera F. 2020. Microbiome of the deep Lake Baikal, a unique oxic 534 bathypelagic habitat. Limnol Oceanogr 65:1471–1488. 535
12. Vigneron A, Cruaud P, Lovejoy C, Vincent WF. 2022. Genomic evidence of functional 536 diversity in DPANN archaea, from oxic species to anoxic vampiristic consortia. ISME 537 Commun 2022 21 2:1–10. 538
13. He C, Keren R, Whittaker ML, Farag IF, Doudna JA, Cate JHD, Banfield JF. 2021. 539 Genome-resolved metagenomics reveals site-specific diversity of episymbiotic CPR 540 bacteria and DPANN archaea in groundwater ecosystems. Nat Microbiol 2021 63 6:354–541 365. 542
14. Gios E, Mosley OE, Weaver L, Close M, Daughney C, Handley KM. 2023. Ultra-small 543 bacteria and archaea exhibit genetic flexibility towards groundwater oxygen content, and 544 adaptations for attached or planktonic lifestyles. 1. ISME Commun 3:1–14. 545
15. Chen L-X, Méndez-García C, Dombrowski N, Servín-Garcidueñas LE, Eloe-Fadrosh EA, 546 Fang B-Z, Luo Z-H, Tan S, Zhi X-Y, Hua Z-S, Martinez-Romero E, Woyke T, Huang L-N, 547 Sánchez J, Peláez AI, Ferrer M, Baker BJ, Shu W-S. 2018. Metabolic versatility of small 548 archaea Micrarchaeota and Parvarchaeota. ISME J 12:756–775. 549
16. Lannes R, Cavaud L, Lopez P, Bapteste E. 2021. Marine Ultrasmall Prokaryotes Likely 550 Affect the Cycling of Carbon, Methane, Nitrogen, and Sulfur. Genome Biol Evol 551 13:evaa261. 552
17. Liu X, Wang Y, Gu J-D. 2021. Ecological distribution and potential roles of 553 Woesearchaeota in anaerobic biogeochemical cycling unveiled by genomic analysis. 554 Comput Struct Biotechnol J 19:794–800. 555
18. Suominen S, Dombrowski N, Sinninghe Damsté JS, Villanueva L. 2021. A diverse 556 uncultivated microbial community is responsible for organic matter degradation in the 557 Black Sea sulphidic zone. Environ Microbiol 23:2709–2728. 558
19. Könneke M, Bernhard AE, de la Torre JR, Walker CB, Waterbury JB, Stahl DA. 2005. 559 Isolation of an autotrophic ammonia-oxidizing marine archaeon. 7058. Nature 437:543–560 546. 561
20. Boetius A, Ravenschlag K, Schubert CJ, Rickert D, Widdel F, Gieseke A, Amann R, 562 Jørgensen BB, Witte U, Pfannkuche O. 2000. A marine microbial consortium apparently 563 mediating anaerobic oxidation of methane. 6804. Nature 407:623–626. 564
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
20
21. Li M, Baker BJ, Anantharaman K, Jain S, Breier JA, Dick GJ. 2015. Genomic and 565 transcriptomic evidence for scavenging of diverse organic compounds by widespread deep-566 sea archaea. Nat Commun 2015 61 6:1–6. 567
22. Liu X, Li M, Castelle CJ, Probst AJ, Zhou Z, Pan J, Liu Y, Banfield JF, Gu J-D. 2018. 568 Insights into the ecology, evolution, and metabolism of the widespread Woesearchaeotal 569 lineages. Microbiome 6:102. 570
23. Besseling MA, Hopmans EC, Boschman RC, Sinninghe Damsté JS, Villanueva L. 2018. 571 Benthic archaea as potential sources of tetraether membrane lipids in sediments across an 572 oxygen minimum zone. Biogeosciences 15:4047–4064. 573
24. Sollai M, Villanueva L, Hopmans EC, Keil RG, Damsté JSS. 2019. Archaeal sources of 574 intact membrane lipid biomarkers in the oxygen deficient zone of the eastern tropical south 575 Pacific. Front Microbiol 10:765. 576
25. Tiano L, Garcia-Robledo E, Dalsgaard T, Devol AH, Ward BB, Ulloa O, Canfield DE, 577 Peter Revsbech N. 2014. Oxygen distribution and aerobic respiration in the north and south 578 eastern tropical Pacific oxygen minimum zones. Deep Sea Res Part Oceanogr Res Pap 579 94:173–183. 580
26. Revsbech NP, Larsen LH, Gundersen J, Dalsgaard T, Ulloa O, Thamdrup B. 2009. 581 Determination of ultra-low oxygen concentrations in oxygen minimum zones by the STOX 582 sensor. Limnol Oceanogr Methods 7:371–381. 583
27. Kwiecinski JV, Babbin AR. 2021. A high-resolution atlas of the Eastern Tropical Pacific 584 oxygen deficient zones. Glob Biogeochem Cycles 35:e2021GB007001. 585
28. Ward BB, Devol AH, Rich JJ, Chang BX, Bulow SE, Naik H, Pratihary A, Jayakumar A. 586 2009. Denitrification as the dominant nitrogen loss process in the Arabian Sea. Nature 587 461:78–81. 588
29. Babbin AR, Keil RG, Devol AH, Ward BB. 2014. Organic matter stoichiometry, flux, and 589 oxygen control nitrogen loss in the ocean. Science 344:406–408. 590
30. Babbin AR, Buchwald C, Morel FMM, Wankel SD, Ward BB. 2020. Nitrite oxidation 591 exceeds reduction and fixed nitrogen loss in anoxic Pacific waters. Mar Chem 224:103814. 592
31. Ulloa O, Canfield DE, DeLong EF, Letelier RM, Stewart FJ. 2012. Microbial oceanography 593 of anoxic oxygen minimum zones. Proc Natl Acad Sci U S A 109:15996–16003. 594
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
21
32. Codispoti LA, Brandes JA, Christensen JP, Devol AH, Naqvi SWA, Paerl HW, Yoshinari 595 T. 2001. The oceanic fixed nitrogen and nitrous oxide budgets: moving targets as we enter 596 the anthropocene? S2. Sci Mar 65:85–105. 597
33. DeVries T, Deutsch C, Rafter PA, Primeau F. 2013. Marine denitrification rates determined 598 from a global 3-D inverse model. Biogeosciences 10:2481–2496. 599
34. Zumft WG. 1997. Cell Biology and Molecular Basis of Denitrification. Microbiol Mol Biol 600 Rev 61:533–616. 601
35. Sanford RA, Wagner DD, Wu Q, Chee-Sanford JC, Thomas SH, Cruz-García C, Rodríguez 602 G, Massol-Deyá A, Krishnani KK, Ritalahti KM, Nissen S, Konstantinidis KT, Löffler FE. 603 2012. Unexpected nondenitrifier nitrous oxide reductase gene diversity and abundance in 604 soils. Proc Natl Acad Sci U S A 109:19709–19714. 605
36. Bertagnolli AD, Konstantinidis KT, Stewart FJ. 2020. Non
‐
denitrifier nitrous oxide 606 reductases dominate marine biomes. Environ Microbiol Rep 12:681–692. 607
37. Lam P, Kuypers MMM. 2011. Microbial nitrogen cycling processes in oxygen minimum 608 zones. Annu Rev Mar Sci 3:317–345. 609
38. Lüke C, Speth DR, Kox MAR, Villanueva L, Jetten MSM. 2016. Metagenomic analysis of 610 nitrogen and methane cycling in the Arabian Sea oxygen minimum zone. PeerJ 4:e1924. 611
39. Fuchsman CA, Devol AH, Saunders JK, McKay C, Rocap G. 2017. Niche partitioning of 612 the N cycling microbial community of an offshore oxygen deficient zone. Front Microbiol 613 8. 614
40. Rattanasriampaipong R, Zhang YG, Pearson A, Hedlund BP, Zhang S. 2022. Archaeal 615 lipids trace ecology and evolution of marine ammonia-oxidizing archaea. Proc Natl Acad 616 Sci 119:e2123193119. 617
41. DeLong EF. 2021. Exploring Marine Planktonic Archaea: Then and Now. Front Microbiol 618 11:616086. 619
42. Glass JB, Kretz CB, Ganesh S, Ranjan P, Seston SL, Buck KN, Landing WM, Morton PL, 620 Moffett JW, Giovannoni SJ, Vergin KL, Stewart FJ. 2015. Meta-omic signatures of 621 microbial metal and nitrogen cycling in marine oxygen minimum zones. Front Microbiol 622 6:998. 623
43. Tsementzi D, Wu J, Deutsch S, Nath S, Rodriguez-R LM, Burns AS, Ranjan P, Sarode N, 624 Malmstrom RR, Padilla CC, Stone BK, Bristow LA, Larsen M, Glass JB, Thamdrup B, 625
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
22
Woyke T, Konstantinidis KT, Stewart FJ. 2016. SAR11 bacteria linked to ocean anoxia and 626 nitrogen loss. Nature 536:179–183. 627
44. Stewart FJ, Ulloa O, DeLong EF. 2012. Microbial metatranscriptomics in a permanent 628 marine oxygen minimum zone. Environ Microbiol 14:23–40. 629
45. Ganesh S, Parris DJ, DeLong EF, Stewart FJ. 2014. Metagenomic analysis of size-630 fractionated picoplankton in a marine oxygen minimum zone. ISME J 8:187–211. 631
46. Zhang IH, Sun X, Jayakumar A, Fortin SG, Ward BB, Babbin AR. 2023. Partitioning of the 632 denitrification pathway and other nitrite metabolisms within global oxygen deficient zones. 633 1. ISME Commun 3:1–14. 634
47. Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW. 2015. CheckM: 635 assessing the quality of microbial genomes recovered from isolates, single cells, and 636 metagenomes. Genome Res 25:1043–1055. 637
48. Chaumeil P-A, Mussig AJ, Hugenholtz P, Parks DH. 2020. GTDB-Tk: a toolkit to classify 638 genomes with the Genome Taxonomy Database. Bioinformatics 36:1925–1927. 639
49. Seemann T. 2014. Prokka: rapid prokaryotic genome annotation. Bioinformatics 30:2068–640 2069. 641
50. Pedruzzi I, Rivoire C, Auchincloss AH, Coudert E, Keller G, de Castro E, Baratin D, Cuche 642 BA, Bougueleret L, Poux S, Redaschi N, Xenarios I, Bridge A. 2015. HAMAP in 2015: 643 updates to the protein family classification and annotation system. Nucleic Acids Res 644 43:D1064–D1070. 645
51. Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, Tosatto 646 SCE, Paladin L, Raj S, Richardson LJ, Finn RD, Bateman A. 2021. Pfam: The protein 647 families database in 2021. Nucleic Acids Res 49:D412–D419. 648
52. Olm MR, Brown CT, Brooks B, Banfield JF. 2017. dRep: a tool for fast and accurate 649 genomic comparisons that enables improved genome recovery from metagenomes through 650 de-replication. ISME J 11:2864–2868. 651
53. Delmont TO, Quince C, Shaiber A, Esen ÖC, Lee ST, Rappé MS, McLellan SL, Lücker S, 652 Eren AM. 2018. Nitrogen-fixing populations of Planctomycetes and Proteobacteria are 653 abundant in surface ocean metagenomes. Nat Microbiol 3:804–813. 654
54. Lee MD. 2019. GToTree: a user-friendly workflow for phylogenomics. Bioinformatics 655 35:4162–4164. 656
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
23
55. Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ. 2015. IQ-TREE: a fast and effective 657 stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol 658 32:268–274. 659
56. Hoang DT, Chernomor O, von Haeseler A, Minh BQ, Vinh LS. 2018. UFBoot2: improving 660 the ultrafast bootstrap approximation. Mol Biol Evol 35:518–522. 661
57. Eren AM, Kiefl E, Shaiber A, Veseli I, Miller SE, Schechter MS, Fink I, Pan JN, Yousef 662 M, Fogarty EC, Trigodet F, Watson AR, Esen ÖC, Moore RM, Clayssen Q, Lee MD, 663 Kivenson V, Graham ED, Merrill BD, Karkman A, Blankenberg D, Eppley JM, Sjödin A, 664 Scott JJ, Vázquez-Campos X, McKay LJ, McDaniel EA, Stevens SLR, Anderson RE, 665 Fuessel J, Fernandez-Guerra A, Maignien L, Delmont TO, Willis AD. 2021. Community-666 led, integrated, reproducible multi-omics with anvi’o. 1. Nat Microbiol 6:3–6. 667
58. Kanehisa M, Goto S. 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic 668 Acids Res 28:27–30. 669
59. Tatusov RL, Fedorova ND, Jackson JD, Jacobs AR, Kiryutin B, Koonin EV, Krylov DM, 670 Mazumder R, Mekhedov SL, Nikolskaya AN, Rao BS, Smirnov S, Sverdlov AV, 671 Vasudevan S, Wolf YI, Yin JJ, Natale DA. 2003. The COG database: an updated version 672 includes eukaryotes. BMC Bioinformatics 4:41. 673
60. Waterhouse A, Procter J, Martin D, Clamp M, Barton G. 2009. Jalview version 2: a 674 multiple sequence alignment and analysis workbench. Bioinforma Oxf Engl 25:1189–91. 675
61. Hallgren J, Tsirigos KD, Pedersen MD, Armenteros JJA, Marcatili P, Nielsen H, Krogh A, 676 Winther O. 2022. DeepTMHMM predicts alpha and beta transmembrane proteins using 677 deep neural networks. bioRxiv https://doi.org/10.1101/2022.04.08.487609. 678
62. Edgar RC. 2010. Search and clustering orders of magnitude faster than BLAST. 679 Bioinformatics 26:2460–2461. 680
63. Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment software version 7: 681 improvements in performance and usability. Mol Biol Evol 30:772–780. 682
64. Capella-Gutiérrez S, Silla-Martínez JM, Gabaldón T. 2009. trimAl: a tool for automated 683 alignment trimming in large-scale phylogenetic analyses. Bioinformatics 25:1972–1973. 684
65. Wilbert SA, Newman DK. 2022. The contrasting roles of nitric oxide drive microbial 685 community organization as a function of oxygen presence. Curr Biol 32:5221-5234.e4. 686
66. Zucko J, Dunlap WC, Shick JM, Cullum J, Cercelet F, Amin B, Hammen L, Lau T, 687 Williams J, Hranueli D, Long PF. 2010. Global genome analysis of the shikimic acid 688
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
24
pathway reveals greater gene loss in host-associated than in free-living bacteria. BMC 689 Genomics 11:628. 690
67. Tästensen J-B, Johnsen U, Reinhardt A, Ortjohann M, Schönheit P. 2020. D-galactose 691 catabolism in archaea: operation of the DeLey–Doudoroff pathway in Haloferax volcanii. 692 FEMS Microbiol Lett 367:fnaa029. 693
68. Hove-Jensen B, Andersen KR, Kilstrup M, Martinussen J, Switzer RL, Willemoës M. 2016. 694 Phosphoribosyl Diphosphate (PRPP): Biosynthesis, Enzymology, Utilization, and 695 Metabolic Significance. Microbiol Mol Biol Rev 81:10.1128/mmbr.00040-16. 696
69. Castelle CJ, Wrighton KC, Thomas BC, Hug LA, Brown CT, Wilkins MJ, Frischkorn KR, 697 Tringe SG, Singh A, Markillie LM, Taylor RC, Williams KH, Banfield JF. 2015. Genomic 698 Expansion of Domain Archaea Highlights Roles for Organisms from New Phyla in 699 Anaerobic Carbon Cycling. Curr Biol 25:690–701. 700
70. Castelle CJ, Méheust R, Jaffe AL, Seitz K, Gong X, Baker BJ, Banfield JF. 2021. Protein 701 Family Content Uncovers Lineage Relationships and Bacterial Pathway Maintenance 702 Mechanisms in DPANN Archaea. Front Microbiol 12. 703
71. Peters JW, Schut GJ, Boyd ES, Mulder DW, Shepard EM, Broderick JB, King PW, Adams 704 MWW. 2015. [FeFe]- and [NiFe]-hydrogenase diversity, mechanism, and maturation. 705 Biochim Biophys Acta BBA - Mol Cell Res 1853:1350–1369. 706
72. Lappan R, Shelley G, Islam ZF, Leung PM, Lockwood S, Nauer PA, Jirapanjawat T, Ni G, 707 Chen Y-J, Kessler AJ, Williams TJ, Cavicchioli R, Baltar F, Cook PLM, Morales SE, 708 Greening C. 2023. Molecular hydrogen in seawater supports growth of diverse marine 709 bacteria. Nat Microbiol 1–15. 710
73. Pearson WR. 2013. An Introduction to Sequence Similarity (“Homology”) Searching. Curr 711 Protoc Bioinforma Ed Board Andreas Baxevanis Al 0 3:10.1002/0471250953.bi0301s42. 712
74. Zhang L, Wüst A, Prasser B, Müller C, Einsle O. 2019. Functional assembly of nitrous 713 oxide reductase provides insights into copper site maturation. Proc Natl Acad Sci U S A 714 116:12822–12827. 715
75. Kroneck PMH. 2018. Walking the seven lines: binuclear copper A in cytochrome c oxidase 716 and nitrous oxide reductase. JBIC J Biol Inorg Chem 23:27–39. 717
76. Farrar JA, Zumft WG, Thomson AJ. 1998. CuA and CuZ are variants of the electron 718 transfer center in nitrous oxide reductase. Proc Natl Acad Sci 95:9891–9896. 719
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
25
77. Zhang L, Bill E, Kroneck PMH, Einsle O. A [3Cu:2S] cluster provides insight into the 720 assembly and function of the CuZ site of nitrous oxide reductase. Chem Sci 12:3239–3244. 721
78. Bourbonnais A, Letscher RT, Bange HW, Échevin V, Larkum J, Mohn J, Yoshida N, 722 Altabet MA. 2017. N2O production and consumption from stable isotopic and 723 concentration data in the Peruvian coastal upwelling system. Glob Biogeochem Cycles 724 31:678–698. 725
79. Kempes CP, van Bodegom PM, Wolpert D, Libby E, Amend J, Hoehler T. 2017. Drivers of 726 Bacterial Maintenance and Minimal Energy Requirements. Front Microbiol 8. 727
80. Babbin AR, Bianchi D, Jayakumar A, Ward BB. 2015. Rapid nitrous oxide cycling in the 728 suboxic ocean. Science 348:1127–1129. 729
81. Sakai HD, Nur N, Kato S, Yuki M, Shimizu M, Itoh T, Ohkuma M, Suwanto A, Kurosawa 730 N. 2022. Insight into the symbiotic lifestyle of DPANN archaea revealed by cultivation and 731 genome analyses. Proc Natl Acad Sci 119:e2115449119. 732
82. Bird JT, Baker BJ, Probst AJ, Podar M, Lloyd KG. 2016. Culture Independent Genomic 733 Comparisons Reveal Environmental Adaptations for Altiarchaeales. Front Microbiol 7. 734
83. Li L, Liu Z, Zhou Z, Zhang M, Meng D, Liu X, Huang Y, Li X, Jiang Z, Zhong S, 735 Drewniak L, Yang Z, Li Q, Liu Y, Nan X, Jiang B, Jiang C, Yin H. 2021. Comparative 736 Genomics Provides Insights into the Genetic Diversity and Evolution of the DPANN 737 Superphylum. mSystems 6:e00602-21. 738
84. Probst AJ, Weinmaier T, Raymann K, Perras A, Emerson JB, Rattei T, Wanner G, Klingl 739 A, Berg IA, Yoshinaga M, Viehweger B, Hinrichs K-U, Thomas BC, Meck S, Auerbach 740 AK, Heise M, Schintlmeister A, Schmid M, Wagner M, Gribaldo S, Banfield JF, Moissl-741 Eichinger C. 2014. Biology of a widespread uncultivated archaeon that contributes to 742 carbon fixation in the subsurface. 1. Nat Commun 5:5497. 743
85. Lipsewers YA, Hopmans EC, Sinninghe Damsté JS, Villanueva L. 2018. Potential 744 recycling of thaumarchaeotal lipids by DPANN Archaea in seasonally hypoxic surface 745 marine sediments. Org Geochem 119:101–109. 746
86. Gfrerer S, Winkler D, Novion Ducassou J, Couté Y, Rachel R, Gescher J. 2022. A 747 Micrarchaeon Isolate Is Covered by a Proteinaceous S-Layer. Appl Environ Microbiol 748 88:e01553-21. 749
87. Brown K, Tegoni M, Prudêncio M, Pereira AS, Besson S, Moura JJ, Moura I, Cambillau C. 750 2000. A novel type of catalytic copper cluster in nitrous oxide reductase. 3. Nat Struct Biol 751 7:191–195. 752
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
26
88. Palumbo AV, Ferguson RL, Rublee PA. 1984. Size of Suspended Bacterial Cells and 753 Association of Heterotrophic Activity with Size Fractions of Particles in Estuarine and 754 Coastal Waters. Appl Environ Microbiol 48:157–164. 755
756 757
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
27
Figure Captions 758 759 Figure 1: A) Locations of metagenomes from ETNP, ETSP, and Arabian Sea used for 760 metagenome assembly, MAG binning, and relative abundance mapping B) Relative abundances 761 of DPANN MAGs across metagenome samples, color-coded by phylum-level taxonomy. 762 763 Figure 2: Species tree of DPANN MAGs and genomes from JGI, NCBI, ODZs, and TARA 764 Oceans collections, colored by phylum-level taxonomy. Only DPANN phyla containing ODZ or 765 TARA MAGs are shown. Black outlined circles indicate ODZ MAGs, blue outlined circles 766 indicate TARA Oceans MAGs, and circles without outlines indicate NCBI or JGI genomes. Stars 767 next to tips indicate the presence of a putative nosZ-like gene. Numbers by nodes correspond to 768 bootstrap supports. 769 770 Figure 3: Metabolic analysis of unique DPANN MAGs. Circles show the presence/absence of 771 key metabolic pathways, grouped by color according to general metabolism categories. Darker 772 circles indicate >70% of genes within the pathway are present, while lighter circles indicate 773 partial pathways (33–70% present). White circles indicate <33% of genes are present, and the 774 pathway is considered absent. Completion/contamination and size of MAGs are shown on the 775 right. 776 777 Figure 4: Protein tree of DPANN nosZ-like proteins (green) within the larger tree of canonical 778 nosZ proteins (typical TAT type in teal, atypical Sec type in orange, type unknown in pink). Tree 779 is rooted on cytochrome c oxidase subunit II (Cox2) proteins, shown in yellow. Diamonds at 780 nodes correspond to ultrafast bootstrap supports, while numbers are SH-alrt values. Sequence 781 motifs for the conserved CuA copper-binding site for each protein are displayed at right. 782 783 Figure 5: A) Schematic showing the spatial N2O concentration for two inter-cell distances of d = 784 0 µm (attached) and d = 2 µm (free-living). The relative surface N2O concentration for the 785 producer is set to 1, while the relative surface N2O concentration of the consumer is set to 0. The 786 radius of the producer, radius of the consumer, and distance between the cells are varied 787 according to values in Table S1. B) Volume-normalized uptake rate of N2O for the consumer at 0 788 µm separation (attached) and 2 µm separation (free-living) for all values of the consumer and 789 producer cell sizes. Numbers indicate the actual volume-normalized uptake rates (multiplied by 790 10-12). C) Uptake rates as a function of the inter-cell distance normalized to the attached scenario 791 of the same consumer-producer cell size combination. A value of, e.g., 0.2 indicates that this 792 combination of producer and consumer cell size shows a reduction of 80% in the consumer N2O 793 uptake rate at this distance compared to if they were attached. n = 20 simulations plotted for each 794 bar, with box representing ± 1 s.d. and the whiskers showing ± 2 s.d. 795 796
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
28
Supplementary Material 797
798
Uncultivated DPANN archaea are ubiquitous inhabitants of global oxygen deficient zones with 799
diverse metabolic potential 800
801
Irene H. Zhanga#*, Benedict Borera, Rui Zhaoa, Steven Wilbertb, Dianne K. Newmanb, and 802
Andrew R. Babbina*
803
aDepartment of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of 804
Technology, Cambridge, MA, USA 805
bDivisions of Biology and Biological Engineering and Geological and Planetary Sciences, 806
California Institute of Technology, Pasadena, California, USA 807
808
#Address correspondence to Irene H. Zhang, izhang@mit.edu and Andrew R. Babbin, 809
babbin@mit.edu 810
811
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
29
Supplementary Methods 812 813 Sequence selection and genetic engineering 814 815 DPANN nosZ-like genes from ODZ metagenome-assembled genomes were extracted and 816 aligned with MAFFT-linsi v7.450 using the --leavegappyregion parameters (1). Three 817 representative sequences were obtained by clustering the DPANN nos-like genes at 90% 818 nucleotide identity with vsearch (2). Putative nosZ-like gene sequences were optimized for 819 the Pseudomonas aeruginosa genome using the IDT codon optimization tool. These fragments 820 were synthesized as gBlocks from Twist Bioscience. Using gBlocks as template, the fragments 821 were amplified with homologous overhangs using PCR primers (Table 4.S1). Products were 822 cloned into the pJM220 plasmid linearized by KpnI and HindIII digestion using Gibson 823 assembly (NEB). Putative nosZ-like gene containing plasmids were introduced at the attTn7 site 824 in the PA14(
∆
nosZ) strain genome (3, 4). Strains with integrated plasmids will drive gene 825 expression using a rhamnose-inducible promotor when grown in the presence of l-rhamnose. 826 827 Growth conditions and N2O reduction tests 828 Three Pseudomonas aeruginosa PA14
∆
nosZ strains containing the putative nosZ-like genes 829 were grown on LB plates at 37oC overnight along with the parent wild-type PA14 (positive 830 control) and PA14
∆
nosZ (negative control). Colonies were inoculated into 10 mL Luria-Bertani 831 (LB) broth within serum vials. Serum vials containing each strain, along with a cell-free control, 832 were capped and sealed post-inoculation and grown within a 37oC incubator with shaking at 100 833 rpm. A total of 4 replicates were performed for each strain. After cultures used up the available 834 oxygen to reach stationary growth phase, sterile 50 µM nitrate and 0.02% l-rhamnose w/v were 835 injected into each serum bottle to drive denitrification and the expression of the nosZ-like genes. 836 After 24 hours of growth at 37oC, N2O concentrations were measured using a Unisense N2O 837 microelectrode connected to the Unisense Field Multimeter according to manufacturer protocols. 838 N2O concentrations were normalized to the average N2O readout of the PA14
∆
nosZ control. 839 840 Supplementary References 841 842 1. Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment software version 7: 843 improvements in performance and usability. Mol Biol Evol 30:772–780. 844
2. Rognes T, Flouri T, Nichols B, Quince C, Mahé F. 2016. VSEARCH: a versatile open 845 source tool for metagenomics. PeerJ 4:e2584. 846
3. Choi K-H, Schweizer HP. 2006. mini-Tn7 insertion in bacteria with single attTn7 sites: 847 example Pseudomonas aeruginosa. Nat Protoc 1:153–161. 848
4. Jeske M, Altenbuchner J. 2010. The Escherichia coli rhamnose promoter rhaP(BAD) is in 849
Pseudomonas putida KT2440 independent of Crp-cAMP activation. Appl Microbiol 850
Biotechnol 85:1923–1933. 851
852
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
30
Supplementary Tables 853
854 putnos1_DDOPOKPK_
00765_Opt putnos2_APMGHGFK_0
0118_Opt putnos3_DAILHJJC_
00524_Opt
Forward
primer CAGGAATTCCTCGAG
AAGCTTATGAAGAAT
AAAGTCCTCATCATC
CAGGAATTCCTCGAG
AAGCTTATGAAGAAG
TACCTGCTCATC
CAGGAATTCCTCGAG
AAGCTTATGCTCGTC
GGGATCGTGTC
Reverse
primer GGGAACTGGTGGTCT
CGTAAGGTACCTCGC
GAATCAGAACG
CGTTCTGATTCGCGA
GGTACCTTATTCGACC
ACGAGGGTCC
CGTTCTGATTCGCGA
GGTACCTTACTTCACG
ATCAGCTTCCC
855 Table S1: Primers used to amplify DPANN nosZ-like gene fragments with overhangs for each of 856 the 3 representative gene sequences. 857 858
Parameter Va lu es
Producer radius R [µm] [0.5, 0.75, 1.0, 1.25, 1.5]
Consumer radius r [µm] [0.05, 0.1, 0.25, 0.5]
Cell distance d [µm] [0, 0.1, 0.5, 1, 2]
859 Table S2: List of parameters used in model including specific values used for all simulations. 860 861
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
31
Supplementary Figures 862 863
864 Figure S1: A) Percentage of the total community assigned to the Archaea domain, with DPANN 865 archaeal phyla grouped together (teal). Other archaea are colored by phylum. B) Proportion of 866 total archaeal reads belonging to the DPANN superphylum vs. other archaeal taxa, with DPANN 867 archaeal phyla grouped together (orange). Other archaea are colored by phylum. All bars are 868 scaled to 100%. 869
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
32
870 Figure S2: Prediction of protein location for A) canonical nosZ-encoded protein B) DPANN 871 nosZ-like protein C) cytochrome c oxidase subunit II. Protein topologies and locations were 872 predicted using DeepTHHMM. Probability scores on the y-axis indicate predicted probability of 873 the protein region as belonging to a signal peptide sequence (orange), cytoplasmic (inside 874 membrane, pink), membrane-associated (red), or outside the membrane such as in the 875 periplasmic space (blue). Amino acid positions are indicated on the x-axis. 876
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
33
877 Figure S3: N2O concentrations after 24 hours of anaerobic growth in LB supplemented with 50 878 µM nitrate and 0.02% rhamnose to drive expression from the DPANN nosZ-like gene within the 879 P. aeruginosa PA14
∆
nosZ background. N2O concentrations are normalized to the average N2O 880 concentration for the parent PA14
∆
nosZ control. The putnos1, putnos2, and putnos3 strains 881 correspond to the 3 representative nosZ-like genes after clustering DPANN nosZ-like sequences 882 at 90% nucleotide identity. Wild-type PA14 and a cell free control are also shown. 883
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
Figures
Figure 1: A) Locations of metagenomes from ETNP, ETSP, and Arabian Sea used for
metagenome assembly, MAG binning, and relative abundance mapping B) Relative abundances
of DPANN MAGs across metagenome samples, color-coded by phylum-level taxonomy.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
Figure 2: Species tree of DPANN MAGs and genomes from JGI, NCBI, ODZs, and TARA
Oceans collections, colored by phylum-level taxonomy. Only DPANN phyla containing ODZ or
TARA MAGs are shown. Black outlined circles indicate ODZ MAGs, blue outlined circles
indicate TARA Oceans MAGs, and circles without outlines indicate NCBI or JGI genomes. Stars
next to tips indicate the presence of a putative nosZ-like gene. Numbers by nodes correspond to
bootstrap supports.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
Figure 3:
Metabolic analysis of unique DPANN MAGs. Circles show the presence/absence of
key metabolic pathways, grouped by color according to general metabolism categories. Darker
circles indica
te >70% of genes within the pathway are present, while lighter circles indicate
partial pathways (33–
70% present). White circles indicate <33% of genes are present, and the
pathway is considered absent. Completion/contamination and size of MAGs are shown o
n the
right.
of
er
ate
he
he
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
Figure 4: Protein tree of DPANN nosZ-like proteins (green) within the larger tree of canonical
nosZ proteins (typical TAT type in teal, atypical Sec type in orange, type unknown in pink). Tree
is rooted on cytochrome c oxidase subunit II (Cox2) proteins, shown in yellow. Diamonds at
nodes correspond to ultrafast bootstrap supports, while numbers are SH-alrt values. Sequence
motifs for the conserved CuA copper-binding site for each protein are displayed at right.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
Figure 5: A) Schematic showing the spatial N2O concentration for two inter-cell distances of d =
0 µm (attached) and d = 2 µm (free-living). The relative surface N2O concentration for the
producer is set to 1, while the relative surface N2O concentration of the consumer is set to 0. The
radius of the producer, radius of the consumer, and distance between the cells are varied
according to values in Table S1. B) Volume-normalized uptake rate of N2O for the consumer at 0
µm separation (attached) and 2 µm separation (free-living) for all values of the consumer and
producer cell sizes. Numbers indicate the actual volume-normalized uptake rates (multiplied by
10-12). C) Uptake rates as a function of the inter-cell distance normalized to the attached scenario
of the same consumer-producer cell size combination. A value of, e.g., 0.2 indicates that this
combination of producer and consumer cell size shows a reduction of 80% in the consumer N2O
uptake rate at this distance compared to if they were attached. n = 20 simulations plotted for each
bar, with box representing ± 1 s.d. and the whiskers showing ± 2 s.d.
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted October 30, 2023. ; https://doi.org/10.1101/2023.10.30.564641doi: bioRxiv preprint
