Experimentally-validated correlati
on analysis reveals new anaerobic
1
methane oxidation partnerships with
consortium-level heterogeneity in
2
diazotrophy
3
Kyle S. Metcalfe*, Ranjani Murali*, Sean W. Mullin, Stephanie A. Connon, and Victoria J.
4
Orphan
5
*Authors contributed equally to this work.
6
7
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA
8
USA
9
10
1
Correspondence: K.S. Metcalfe (kmetcalf@caltech.edu) or V.J. Orphan
11
(vorphan@caltech.edu), Division of Geological and Planetary Sciences, California Institute of
12
Technology, 1200 E. California Blvd., Mail Code 100-23, Pasadena, CA 91125, USA.
13
14
Abstract
15
Archaeal anaerobic methanotrophs (‘ANME’) and sulfate-reducing Deltaproteobacteria (‘SRB’)
16
form symbiotic multicellular consortia capable of anaerobic methane oxidation (AOM), and in so
17
doing modulate methane flux from marine sediments. The specificity with which ANME
18
associate with particular SRB partners
in situ
, however, is poorly understood. To characterize
19
partnership specificity in ANME-SRB consortia, we applied the correlation inference technique
20
SparCC to 310 16S rRNA Illumina iTag amplicon libraries prepared from Costa Rica sediment
21
samples, uncovering a strong positive correlation between ANME-2b and members of a clade of
22
Deltaproteobacteria we termed SEEP-SRB1g. We confirmed this association by examining 16S
23
rRNA diversity in individual ANME-SRB consortia sorted using flow cytometry and by imaging
24
ANME-SRB consortia with fluorescence
in situ
hybridization (FISH) microscopy using newly-
25
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
2
designed probes targeting the SEEP-SRB1g clade. Analysis of genome bins belonging to
26
SEEP-SRB1g revealed the presence of a complete
nifHDK
operon required for diazotrophy,
27
unusual in published genomes of ANME-associated SRB. Active expression of
nifH
in SEEP-
28
SRB1g and diazotrophic activity within
ANME-2b/SEEP-SRB1g consortia was then
29
demonstrated by microscopy using hybridization chain-reaction (HCR-) FISH targeting
nifH
30
transcripts and by FISH-nanoSIMS experim
ents. NanoSIMS analysis of ANME-2b/SEEP-
31
SRB1g consortia incubated with a headspace containing CH
4
and
15
N
2
revealed differences in
32
cellular
15
N-enrichment between the two partners that varied between individual consortia, with
33
SEEP-SRB1g cells enriched in
15
N relative to ANME-2b in one consortium and the opposite
34
pattern observed in others,
indicating both ANME-2b and SEEP-
SRB1g are capable of nitrogen
35
fixation, but with consortium-specific variation in whether the archaea or bacterial partner is the
36
dominant diazotroph.
37
38
Introduction
39
The partnership between anaerobic, methanotrophic Archaea (ANME) and their associated
40
sulfate-reducing bacteria (SRB) is one of the most biogeochemically-important symbioses in the
41
deep-sea methane cycle [1, 2]. As a critical component of methane seep ecosystems,
42
multicellular consortia of ANME and associated SRB consume a significant fraction of the
43
methane produced in marine sediments, using sulfate as a terminal electron acceptor to perform
44
the anaerobic oxidation of methane (AOM) [1–4]. ANME-SRB consortia are thought to perform
45
AOM through the direct extracellular transfer of electrons between ANME and SRB [5–7]. Along
46
with symbiotic extracellular electron transfer, ANME-SRB consortia also exhibit other traits of
47
mutualism such as the sharing of nutrients. For example, members of the ANME-2 clade have
48
been reported to fix and share N with partner bacteria [8–11], but the extent to which
49
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
3
diazotrophic capability might vary
across the diverse clades of ANME and associated SRB is
50
the focus of ongoing research.
51
52
Comparative studies of ANME [12] and associated SRB [13, 14] genomes from multiple ANME-
53
SRB consortia have revealed significant diversity across clades, particularly for SRB genomes
54
falling within subclades
of the SEEP-SRB1 [14], common SRB
partners to ANME [15]. However,
55
the implications of symbiont diversity for metabolic adaptation in ANME-SRB consortia are
56
obscured by the absence of clearly-established ANME-SRB pairings in the environment. A
57
framework defining these pairings would address this gap in knowledge. Establishing this
58
framework for partnership specificity in ANME-SRB consortia—being the preference that certain
59
ANME exhibit for specific SRB partners—would shed light on the extent to which ANME or SRB
60
physiology may differ in consortia constituted of different ANME-SRB pairs.
61
62
As an aspect of ANME or SRB physiology that may differ in different ANME-SRB pairings,
63
nitrogen anabolism has been observed to be involved in the symbiotic relationship between
64
partners [8, 9] and has been shown to influence niche differentiation of different ANME-SRB
65
consortia via ni
trate assimilation ability [16]. Previous
evidence documenting
active diazotrophy
66
in ANME-SRB consortia by nitrogenase expression [8] and
15
N
2
fixation by nanoSIMS indicated
67
that ANME-2 are the primary diazotrophs in ANME-SRB consortia and supply fixed nitrogen to
68
SRB partners [8–10]. The diazotrophic potential of syntrophic SRB, however, and their role in
69
nitrogen fixation within consortia is poorly understood. Evidence from SRB genomes [14] and
70
the expression of unidentified nitrogenase sequences in methane seep sediments [8] suggested
71
that seep associated SRB may fix nitrogen, opening up the possibility of variation in diazotrophic
72
activity among taxonomically-distinct ANME-SRB consortia.
73
Previous research characterizing the diversity of partnerships in ANME-SRB consortia have
74
employed fluorescence microscopy, magnetic separation by magneto-FISH, and single-cell
75
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
4
sorting techniques (e.g. BONCAT-FACS) that are robust against false positives, but are often
76
limited in statistical power. Fluorescence
in situ
hybridization (FISH) has helped to establish the
77
diversity of ANME-bacterial associations, with ANME constituting four diverse polyphyletic
78
clades within the Methanomicrobia:
ANME-1a/b [4, 17–20], ANME-2a,b,c [3, 20–22] , ANME-2d
79
[23, 24], and ANME-3 [20, 25, 26]. ANME-associated SRB have also observed by FISH to be
80
diverse, representing several clades of Deltaproteobacteria including the
81
Desulfococcus/Desulfosarcina
(DSS) clade [3–6, 15, 19–22, 27–33], two separate subclades
82
within the Desulfobulbaceae [16, 25, 26
], a deeply-branching group termed the SEEP-SRB2
83
[34], and a thermophilic clade of Desulfobacteraceae known as HotSeep-1 [34, 35]. These FISH
84
studies documented associations for different ANME-SRB consortia, including partnerships
85
between members of ANME-1 and SEEP-SRB2 [13] or HotSeep-1 [7, 13, 35], ANME-2a and
86
SEEP-SRB1a [15], ANME-2c and SEEP-SRB1a [5],
SEEP-SRB2 [13, 34], or Desulfobulbaceae
87
[29], and ANME-3 and SEEP-SRB1a [15] or Desu
lfobulbaceae [25, 26].
Conspicuously, SRB
88
found in consortia with ANME-2b have only been identified broadly as members of the
89
Deltaproteobacteria targeted by the probe S-C-dProt-0495-a-A-18 (often referred to as
Δ
495) [5,
90
31, 36], leaving little known about the specific identity of this SRB partner. Visualizing ANME-
91
SRB partnerships by FISH has been a valuable aspect of AOM research, but FISH requires the
92
design of probes with suff
icient specificity to
identify partner organisms and thus will only detect
93
partnerships consisting of taxa for which phylogenetic information is known [22]. Magneto-FISH
94
[29, 37, 38] or BONCAT-enabled fluorescence-activated cell sorting (BONCAT-FACS) of single
95
ANME-SRB consortia [39] complement FISH experiments by physical capture (via magnetic
96
beads or flow cytometry, respectively) and sequencing of ANME and associated SRB from
97
sediment samples. These studies corroborated some of the patterns observed from FISH
98
experiments, showing associations between ANME-2 and diverse members of the DSS [39].
99
Magneto-FISH and BONCAT-FACS observations of ANME-SRB pairings are also highly robust
100
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
5
against false positives but can lack the statistical power conferred by more high-throughput
101
approaches that is necessary to establish a general framework for partnership specificity.
102
103
Recently, a number of correlation analysis techniques have been introduced in molecular
104
microbial ecology studies, providing information about patterns of co-occurrence between 16S
105
rRNA OTUs or ASVs recovered from environm
ental iTag [40] diversity surveys [41–43].
106
Correlation analysis performed on 16S rRNA amplicon surveys provides a complementary
107
method to Magneto-FISH and/or BONCAT-FACS that can be used to develop hypotheses about
108
potential microbial interactions. While predictions of co-occurrence between phylotypes from
109
these correlation analysis techniques have been reported in a number of diverse environments,
110
they are rarely validated through independent approaches, with a few notable exceptions [44].
111
112
Here, we present a framework for ANME-SRB partnership specificity, using correlation analysis
113
of 16S iTag amplicon sequences from a large-scale survey of seafloor methane seep sediments
114
near Costa Rica to predict potential ANME-SRB partnerships. A partnership between ANME-2b
115
and members of an SRB group previously not known to associate with ANME (SEEP-SRB1g)
116
was hypothesized by correlation analysis and independently assessed using FISH and amplicon
117
data from BONCAT-FACS-sorted ANME-SRB consortia. With this new framework, we were
118
able to identify a nov
el partnership between ANME-2b and SEEP-SRB1g and map predicted
119
physiological traits of SEEP-SRB1g genomes onto partnership specificity with ANME-2b. Our
120
approach led us to formulate new hypothes
es regarding how SEEP-SRB1g physiology may
121
complement ANME-2b physiology, focusing on nitrogen fixation in SEEP-SRB1g. We
122
demonstrate in this study that the symbiotic relationship between ANME and associated SRB
123
can vary depending on the nature of the partner taxa and affirm the importance of characterizing
124
individual symbiont pairings in understanding AOM symbiosis.
125
126
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
6
Materials and Methods
127
Here, we present an abridged description of the methods used in this study. A full description
128
can be found in the Supplemental Materials and Methods.
129
130
Sample origin and processing
131
Pushcore samples of seafloor sediment were collected by DSV
Alvin
during the May 20-June 11
132
2017 ROC HITS Expedition (AT37-13) aboard R/V
Atlantis
to methane seep sites southwest of
133
Costa Rica [45–47]. After retrieval from the seafloor, sediment pushcores were extruded aboard
134
R/V
Atlantis
and sectioned at 1-3 cm intervals for geochemistry and microbiological sampling
135
using published protocols [21, 48]. Samples for DNA extraction were immediately frozen in
136
liquid N
2
and stored at -80 ̊C. Samples for microscopy were fixed in 2% paraformaldehyde for
137
24 h at 4 ̊C. A full list of samples used in this study can be found in Supplemental Table 1 and
138
additional location and geochemical data can be found at https://www.bco-
139
dmo.org/dataset/715706.
140
141
DNA extraction and iTag sequencing
142
DNA was extracted from 310 samples of Costa Rican methane seep sediments and seep
143
carbonates (Supp. Table 1) using the Qiagen PowerSoil DNA Isolation Kit 12888 following
144
manufacturer directions modified for sediment and carbonate samples [21, 49]. The V4-V5
145
region of the 16S rRNA gene was amplified using archaeal/bacterial primers, 515F (5'-
146
GTGYCAGCMGCCGCGGTAA-3') and 926R (5'-CCGYCAATTYMTTTRAGTTT-3') with Illumina
147
adapters [50]. PCR reaction mix was set up in duplicate for each sample with New England
148
Biolabs Q5 Hot Start High-Fidelity 2x Master Mix in a 15 μL reaction volume with annealing
149
conditions of 54°C for 30 cycles. Duplicate PCR samples were then pooled and 2.5 μL of each
150
product was barcoded with Illumina NexteraXT index 2 Primers that include unique 8-bp
151
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
7
barcodes. Amplification with barcoded primers used annealing conditions of 66°C and 10
152
cycles. Barcoded samples were combined into a single tube and purified with Qiagen PCR
153
Purification Kit 28104 before submission to Laragen (Culver City, CA, USA) for 2 x 250 bp
154
paired end analysis on Illumina’s MiSeq platform. Sequence data was submitted to the NCBI
155
Sequence Read Archive as Bioproject PRJNA623020. Sequence data was processed in QIIME
156
version 1.8.0 [51] following Mason, et al. 2015 [52]. Sequences were clustered into
de novo
157
operational taxonomic units (OTUs) with 99% similarity [53], and taxonomy was assigned using
158
the SILVA 119 database [54]. The produced table of OTUs detected in the 310 methane seep
159
sediment and seep carbonate amplicon libraries was analyzed using the correlation algorithm
160
SparCC [41]. To examine phylogenetic placement of SRB 16S rRNA gene amplicon sequences
161
predicted by network analysis to associate with particular ANME subgroup amplicon sequences,
162
a phylogeny was constructed using RAxML-HPC [55] on XSEDE [56] using the CIPRES
163
Science Gateway [57] from full-length 16S rRNA sequences of Deltaproteobacteria aligned by
164
MUSCLE [58]. Genomes downloaded from the IMG/M database were searched using tblastn.
165
Chlorophyllide reductase BchX (WP011566468) was used as a reference sequence for a tblastn
166
nifH
search using BLAST+. Genome trees were constructed using the Anvi’o platform [59] using
167
HMM profiles from a subset [60] of ribosomal protein sequences and visualized in iTOL [61].
168
169
FISH probe design and microscopy
170
A new FISH probe was designed in ARB [62]. This new probe, hereafter referred to as Seep1g-
171
1443 (Supp. Table 2), was designed to complement and target 16S rRNA sequences in a
172
monophyletic “
Desulfococcus
sp.” clade. Based on phylogenetic analysis (see below), this clade
173
was renamed SEEP-SRB1g. Seep1g-1443 was order
ed from Integrated DNA Technologies
174
(Coralville, IA, USA). FISH r
eaction conditions
were optimized for Seep1g-1443, with optimal
175
formamide stringency found to be 35% (Supp. Fig. 1). FISH and hybridization chain reaction
176
(HCR-) FISH was performed on fixed ANME-SRB consortia using previously published density
177
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
8
separation and FISH protocols [22]. FISH was performed overnight (18 hr) using modifications
178
(G. Chadwick, pers. comm.) to previously-published protocols [29, 39, 63, 64]. Structured-
179
illumination microscopy (SIM) was performed on FISH and HCR-FISH (see below) experiments
180
to image ANME-SRB consortia using the Elyra PS.1 SIM platform (Zeiss, Germany) and an
181
alpha Plan-APOCHROMAT 100X/1.46 Oil DIC M27 objective. Zen Black software (Zeiss) was
182
used to construct final images from the structured-illumination data.
183
184
mRNA imaging using HCR-FISH
185
Hybridization chain reaction FISH (HCR-FISH) is a powerful technique to amplify signal from
186
FISH probes [65, 66]. The protocol used here was modified from Yamaguchi and coworkers
187
[67].
nifH
initiators, purchased from Molecular Technologies (Pasadena, CA, USA; probe
188
identifier “nifH 3793/D933”) or designed in-house (Supp. Table 2) and ordered from Integrated
189
DNA Technologies, were hybridized to fixed ANME-SRB consortia. Hairpins B1H1 and B1H2
190
with attached Alexa647 fluorophores (Molecular Technologies) were added separately to two 45
191
μL volumes of amplification buffer in PCR tubes and snap cooled by placement in a C1000
192
Touch Thermal Cycler (BioRad, Hercules, CA, USA) for 3 min at 95 ̊C. After 30 min at room
193
temperature, hairpins were mixed and placed in PCR tubes along with hybridized ANME-SRB
194
consortia. Amplification was performed for 15 min at 35 ̊C. Similar results were observed when
195
the HCR-FISH v3.0 protocol established by Choi et al. [68] was used.
196
197
Stable Isotope Probing and nanoSIMS
198
Methane seep sediments containing abundant
ANME-2b and SEEP-SRB1g consortia (Supp.
199
Fig. 4) were used in stable isotope probing (SIP) experiments to test for diazotrophic activity by
200
SEEP-SRB1g. N sources were removed from the s
ediment slurry by wash
ing with ar
tificial
201
seawater without an N source (see Supplemental Materials and Methods). Two anoxic
202
incubations were pressurized with 2.8 bar CH
4
with 1.2 mL
15
N
2
at 1 bar, approximately
203
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
9
equivalent to 2% headspace in 20 mL CH
4
at 2.8 bar (Supp. Table 3). Positive control
204
incubations (
n
= 2) were amended with
15
NH
4
Cl and were further pressurized with 2.8 bar CH
4
205
and 1.2 mL natural-abundance N
2
at 1 bar. Incubations were periodically checked for AOM
206
activity via sulfide production using the Cline assay [69] and were chemically fixed for FISH-
207
nanoSIMS analysis [70] after 9 months. Fixed ANME-SRB consortia were separated from the
208
sediment matrix and concentrated following published protocols [5]. Samples were then
209
embedded in Technovit H8100 (Kulzer GmbH, Germany) resin according to published protocols
210
[5, 31] and thin sections (2 μm thickness) were prepared using an Ultracut E microtome
211
(Reichert AG, Austria) which were mounted on Teflon/poly-L-lysine slides (Tekdon Inc., USA).
212
FISH reactions were performed using Seep1g-1443 and ANME-2b-729 probes as described
213
above, with the omission of 10% SDS to prevent detachment of section from slide (G.
214
Chadwick, pers. comm.), and slides were imaged and mapped for subsequent nanoSIMS
215
analysis using a Zeiss Elyra PS.1 platform. After removal of DAPI-Citifuor by washing following
216
published protocols [70], slides were cut to fit into nanoSIMS sample holders and sputter-coated
217
with 40 nm Au using a Cressington sputter coater. NanoSIMS was performed using a Cameca
218
NanoSIMS 50L housed in Caltech’s Microanalysis Center, and data was analyzed using
219
look@nanoSIMS [71].
220
221
Results
222
16S rRNA correlation analysis predicts a s
pecific association between ANME-2b and SEEP-
223
SRB1g
224
Correlation analysis applied to 16S rRNA gene amplicon libraries has been frequently used to
225
identify interactions between microorganisms based on the co-occurrence of their 16S rRNA
226
sequences in different environments or conditions [72–75]. Here, we applied correlation analysis
227
to 16S rRNA amplicon libraries prepared from Costa Rican methane seep sites (Supp. Table 1)
228
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
10
to explore partnership specificity between ANME and associated SRB. QIIME processing of
229
amplicon sequences prepared from 310 Costa Rican methane seep sediment and seep
230
carbonate samples yielded 3,052 OTUs after filtering in R. A table of read abundances for these
231
OTUs across the 310 samples was analyzed by SparCC to calculate correlation coefficients and
232
significance for all possible 4,658,878 OTU pairs using 100 bootstraps (Fig. 1). Of these pairs,
233
9.7% (452,377) had pseudo-
p
-values < 0.01, indicating the coefficients for each of these
234
correlations exceeded that calculated for that same OTU pair in any of the 100 bootstrapped
235
datasets [41]. The taxonomic assignment of the constituent OTUs of correlations with pseudo-
p
236
< 0.01 were then inspected, where 18% (81,459) of correlations with pseudo-
p
< 0.01 describe
237
those involving ANME. Of these, 32% occur between ANME and OTUs assigned to three main
238
taxa:
Desulfococcus
sp., SEEP-SRB1a, and SEEP-SRB2 (Fig. 1)
. A complete list of significant
239
correlations, their coefficient values, OTU identifiers, and accompanying taxonomy assignments
240
can be found in Supplemental Table 4.
241
242
16S rRNA phylogenetic analysis revealed the SILVA-assigned “
Desulfococcus
sp.” OTUs
243
comprise a sister clade to the SEEP-SRB
1a that is distinct from cultured
Desulfococcus
sp.
244
(e.g.
D. oleovorans
and
D. multivorans
, see below). We therefore reassigned the
Desulfococcus
245
OTUs to a new clade we termed SEEP-SRB1g
following the naming scheme outlined for seep-
246
associated SRB in Schreiber, et al. (e.g. SEEP-SRB1a through -SRB1f) [15]. Furthermore,
247
statistically-significant correlations between OTUs of ANME and SRB taxa suggested that
248
ANME-SRB partnerships in the Costa Rica seep samples could be classified into the following
249
types: ANME-1 with SEEP-SRB1a or SEEP-SRB2, ANME-2a with SEEP-SRB1a, ANME-2b
250
with SEEP-SRB1g, ANME-2c with SEEP-SR
B1a or SEEP-SRB2, and ANME-3 with SEEP-
251
SRB1a (Fig. 1). While physical association between different ANME lineages and
252
Deltaproteobacterial clades SEEP-SRB1a and
SEEP-SRB2 had been well-documented [5, 13,
253
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
11
15, 31, 34], members of the SEEP-SRB1g
had not previously been identified as a potential
254
syntrophic partner with methanotrophic ANME.
255
256
These associations were further examined by detailed network analysis in which the table of
257
correlations with pseudo
p
-values < 0.01 was further filtered to contain only those correlations
258
with coefficients (a measure of correlation strength) in the 99
th
percentile of all significant
259
correlations. A network diagram in which nodes represent OTUs and edges between nodes
260
represent correlations was constructed with force-directed methods [76], where edge length
261
varied in inverse proportion to correlation strength. A subregion of this network focused on
262
ANME-associated OTUs is presented in Figure 2a. Cohesive blocks, subsets of the graph with
263
greater connectivity to other nodes in the block than to nodes outside [77], were calculated and
264
revealed 3 primary blocks of ANME and SRB OTUs. Visualization of these 3 blocks by a chord
265
diagram [78] further highlighted
the taxonomic identity of ANME-S
RB OTU pairs in these blocks:
266
ANME-1 or ANME-2c (one OTU with mean read count < 10) and SEEP-SRB2, ANME-2a or
267
ANME-2c and SEEP-SRB1a, and ANME-2b or ANME-2a and SEEP-SRB1g (Fig. 2b). The
268
predicted associations
between ANME-2c and SEEP-SRB2 and between ANME-2a and SEEP-
269
SRB1g were relatively more rare than the other associations; only one rare ANME-2c OTU
270
(mean read count < 10) and four uncommon ANME-2a OTUs (mean read count < 100) were
271
predicted between SEEP-SRB2 and SEEP-SRB1g, respectively. Inferred partnership specificity
272
in two of the blo
cks has been previously corroborated by FISH
studies, namely t
hat exhibited by
273
ANME-1 with SEEP-SRB2 [13, 34], ANME-2
c with SEEP-SRB1a [5], and ANME-2a with SEEP-
274
SRB1a [15]. The partnership between SEEP-SRB1g and ANME-2b, however, had no
275
precedent, as the only previous FISH descriptions of ANME-2b had placed it with a partner
276
Deltaproteobacterium with taxonomy not known beyond the phylum level [5, 31].
277
278
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
12
Common patterns of association observed in network analysis and in single ANME-SRB
279
consortia
280
To test if ANME-SRB partnership specificity observed in our iTag correlation analysis from seep
281
samples (Figs. 1, 2) was consistent with data collected from individually-sorted ANME-SRB
282
consortia after BONCAT-FACS [39], we constructed a phylogeny with full-length and amplicon
283
16S rRNA sequences from ANME-associated SRB including SEEP-SRB1g (Fig. 3; Supp. Fig
284
5). 16S rRNA iTag amplicon sequences from the network analysis (Fig. 2) and from BONCAT-
285
FACS sorted consortia (Fig. 3; [39]) were then annotated by ANME subtype and identity of
286
associated phylotypes. In the BONCAT-FACS dataset, 8 out of 11 (72%) of the consortia with
287
ANME-2b OTUs had corresponding deltaproteobacterial OTUs that belonged to the SEEP-
288
SRB1g clade (Fig. 3). Similarly, of the Deltaproteobacteria OTU sequences from the BONCAT-
289
FACS sorted consortia affiliated with SEEP-SR
B1g 89% (8/9) had ANME
-2b as the archaeal
290
partner (Fig. 3). Notably, we found that these SEEP-SRB1g sequences were also highly-similar
291
to published full-length 16S rRNA clone library sequences (e.g. NCBI accession AF354159)
292
from seep sediments where ANME-2b phylotypes were also recovered [21]. A
≗
2
-test for
293
independence was performed on 16S rRNA OTUs recovered from (39) to test the null
294
hypothesis that the presence of a given SRB taxon in a FACS sort is independent of the type of
295
ANME present in the sort. This test demonstrated that the SRB taxon found in a given sort was
296
dependent on the ANME also present in the sort,
≗
2
= 30.6 (
d.f.
= 6,
n
= 30),
p
< 0.001. The
297
pattern of association between ANME and SRB OTUs in individual BONCAT-FACS-sorted
298
ANME-SRB consortia thus corroborated the inference from network analysis that ANME-2b and
299
SEEP-SRB1g OTUs exhibit significant partnership
specificity. On t
he basis of amplicon
300
sequence associations found from the BONCAT-FACS sorting dataset as well as those
301
displayed by correlation analysis of amplicons from Costa Rica methane seeps, we designed a
302
set of independent experiments to test the hypothesis that ANME-2b form syntrophic
303
partnerships with the previously-undescribed SEEP-SRB1g deltaproteobacteria.
304
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
13
305
FISH experiments show SEEP-SRB1g in association with ANME-2b
306
Specific oligonucleot
ide probes were designed and test
ed for the SEEP-SRB1g clade (Supp.
307
Fig. 1) and FISH experiments were used to
validate the predicted ANME-2b—SEEP-SRB1g
308
partnership. Simultaneous application of FISH
probes targeting SEEP-SRB1a, the dominant
309
deltaproteobacterial partner of ANME (S
eep1a-1441 [15]), the newly designed SEEP-SRB1g
310
probe (Seep1g-1443, this work), and a probe targeting ANME-2b (ANME-2b-729 [39])
311
demonstrated that ANME-2b predominantly form consortia with SEEP-
SRB1g, appearing as
312
large multicellular consortia in seep sediment samples from different localities at Costa Rica
313
methane seep sites (see Supplemental Materials and Methods for site details) that also contain
314
ANME-2a (Fig. 4b) and ANME-2c (Fig. 4f). ANME-2b was not observed in association with
315
SEEP-SRB1a (Figs. 4a, 4e), and SEEP-SRB1g was not observed in associ
ation with ANME-2a
316
(Fig. 4d) or ANME-2c (Fig. 4h) when FISH probes ANME-2a-828 or ANME-2c-760 [20] were
317
substituted for ANME-2b-729 (
n
≈
100 consortia). Instead, SEEP-
SRB1a was found in consortia
318
with ANME-2a (Fig. 4c) and ANME-2c (Fig. 4g), consistent with previous reports [15].
319
320
Genomic potential for N
2
fixation in sulfate-reducing SEEP-SRB1g deltaproteobacteria
321
Given the importance of diazotrophy in the functioning of ANME-SRB syntrophy, we screened
322
metagenome-assembled genome bins (MAGs) of SEEP-SRB1g for the presence of the
323
nitrogenase operon. A genome tree constructed from previously published MAGs from Hydrate
324
Ridge and Santa Monica Basin [14, 39] revealed that two closely related MAGs
325
(Desulfobacterales sp. C00003104, and C00003106) originally classified as belonging to the
326
Seep-SRB1c clade [14] possessed the nitrogenase operon (Fig. 5). These MAGs did not
327
possess 16S rRNA genes, precluding 16S rRNA-based taxonomic identification. A more
328
detailed look at these reconstructed genomes revealed that the nitrogenase along with a suite of
329
other genes were unique to th
is subclade and missing in
other SEEP-SRB1c MAGs [14],
330
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
14
suggesting they may represent a distinct lineage. In effort to connect these nitrogenase
331
containing SRB MAG’s with representative 16S rRNA sequences, we examined mini-
332
metagenome data from individual BONCAT-FACS sorted ANME-SRB consortia which each
333
contained 16S rRNA gene sequences for the ANME and bacterial partner [39]. A genome tree
334
containing deltaproteobacterial MAGs from Skennerton, et al. [14] and reconstructed
335
deltaproteobacterial genomes from the BONCAT-FACS sorts [39] revealed one SRB genome
336
from a FACS-sorted consortium (Desulfobacterales sp. CONS3730E01UFb1, IMG Genome ID
337
3300009064) was closely related to the two putative Seep-SRB1c MAGs containing the
338
nitrogenase operon (Fig. 5). The 16S rRNA amplicon sequence (NCBI accession KT945234)
339
associated with the Desulfobacterales sp. CONS3730E01UFb1 genome was used to construct
340
a 16S rRNA phylogeny and confirm
ed to cluster within the SEEP-
SRB1g clade, providing a link
341
between the 16S rRNA and associated nitrogenase sequences in this lineage (Fig. 3). Given
342
that Desulfobacterales sp. CONS3730E01UFb1, C00003104, and C00003106 genomes
343
appeared highly similar on the genome tree (Fig. 5), we reassigned the previously published
344
Desulfobacterales sp. C00003104 and C00003106 MAGs to the SEEP-SRB1g. Notably, the
345
other 16S rRNA amplicon sequence sampled from the sorted consortium CONS3730E01UF
346
(NCBI accession KT945229) was assigned to ANME-2b [39].
347
348
As noted above, these SEEP-SRB1g MAGs were remarkable for the presence of the
nifHDK
349
operon involved nitrogen fixation, which had previously not been an area of focus in previous
350
analyses of ANME-associated SRB genomes (Fig. 5). A re-analysis of published
nifH
cDNA
351
sequences from methane seep sediments revealed sequences that were nearly identical to the
352
SEEP-SRB1g
nifH
(NCBI accession KR020451-KR020457, [8]) suggesting active transcription
353
of SEEP-SRB1g
nifH
under
in situ
conditions (Fig. 6). An analysis of published methane seep
354
metaproteomic data [14] also indicated acti
ve translation of ni
trogenase by SEEP-SRB1g,
355
corroborating evidence from cDNA libraries [8]. Additionally, other
nifH
cDNA sequences in this
356
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
15
study were found to be identical to ni
trogenase sequences occurring in 18 SEEP-SRB1a
357
unpublished metagenome bins (Supp. Fig. 6) demonstrating that at least some of the syntrophic
358
SEEP-SRB1a SRB partners also possess and actively express
nifH
.
359
360
Single-cell nifH expression visualized by HCR-FISH and
15
N
2
FISH-nanoSIMS experiments
361
confirm involvement of SEEP-SRB1g in N
2
-fixation
362
The dominant role of ANME-2 in nitrogen fixation reported by previous studies [8–10] motivated
363
our examination of whether the sulfate-reduc
ing SEEP-SRB1g partners of ANME-2b were also
364
involved in diazotrophy, either in concert with the ANME-2b partner, or perhaps as the sole
365
diazotroph in this AOM partnership. Using the
nifH
sequences from SEEP-SRB1g, we designed
366
a specific mRNA-targeted probe set to use in whole-cell hybridization chain reaction FISH
367
(HCR-FISH) assays (Supp. Table 2). HCR-FISH allows for signal amplification and improved
368
signal-to-noise ratio compared to FISH, and has been used in single cell mRNA expression
369
studies in select microbial studies [79–81]. Prior to this study, however, HCR-FISH had not been
370
applied to visualize gene expression in ANME-SRB consortia from methane seep sediments. In
371
the context of experiments with sediment-dwelling ANME-SRB consortia, HCR-FISH provided
372
adequate amplification of the signal to detect expressed mRNA above the inherent background
373
autofluorescence in sediments. Using our HCR-FISH probes targeting SEEP-SRB1g
nifH
374
mRNA together with the standard 16S rRNA targeted oligonucleotide FISH probes Seep1g-
375
1443 (targeting SEEP-SRB1g) and ANME-2b-729 (targeting ANME-2b), we successfully
376
imaged
nifH
mRNA transcripts by SEEP-SRB1g cells
in ANME-2b—SEEP-SRB1g consortia in
377
a sediment AOM microcosm experim
ent (Fig. 7). Positive HCR-FISH
nifH
hybridization in this
378
sample was observed to be exclusively associ
ated with the SEEP-SRB1g bacterial partner in
379
ANME-2b consortia (
n
= 5), and not observed in ANME-2b stained cells nor in ANME-2a or -2c
380
consortia, supporting the specificity of this assay. Negative control experiments for the HCR-
381
FISH reaction were also performed in which SEEP-SRB1g
nifH
initiator probes were added to
382
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
16
the assay, but the fluorescent amplifier hairpins were absent. In this case, there is no
383
fluorescent signal in either the bacteria or archaeal partners in ANME-2b aggregates confirming
384
that there is no native autofluorescence in Seep-SRB1g that could be responsible for the signal
385
observed in the HCR-FISH experiments (Supp. Fig. 7f-j). We performed another control without
386
the initiator probes that bind the mRNA but with the addition of the fluorescent amplifier hairpins.
387
As in the other negative control, we observed limited non-specific binding of the hairpins that
388
were easy to differentiate from the positively-hybridized SEEP-SRB1g (Supp. Fig. 7a-e).
389
Occasionally, highly localized and small spots of hairpins were observed (Supp. Fig 7d) but
390
these dots were mostly localized outside of aggregates and did not align with either bacteria or
391
archaea in consortia (e.g. Figure 7d). Confirmation of
nifH
expression using HCR-FISH
392
corroborated evidence from
cDNA libraries (Fig. 6) that SEEP-SRB1g actively express
nifH
,
393
providing support for diazotrophy in the sulfate-reduci
ng partner in ANME-2b—SEEP-SRB1g
394
consortia.
395
396
To confirm active diazotrophy by ANME-2b-associated SEEP-SRB1g, we prepared stable
397
isotope probing incubations of methane seep sediments recovered from a Costa Rica methane
398
seep. These nitrogen-poor sediment incubations were amended with unlabeled methane and
399
15
N
2
and maintained in the laboratory at 10ºC under conditions supporting active sulfate-coupled
400
AOM (see Supplemental Materials and Methods). Sediments with abundant ANME-SRB
401
consortia were sampled after 9 months of incubation and separated consortia were analyzed by
402
nanoSIMS to measure single cell
15
N enrichment associated with diazotrophy within ANME-
403
2b—SEEP-SRB1g consortia. Representativ
e ANME-2b—SEEP-SRB1g consortia (
n
= 4) were
404
analyzed by FISH-nanoSIMS and shown to be significantly enriched in
15
N relative to natural
405
abundance values (0.36%; Fig. 8). Among the consortia analyzed, the
15
N fractional abundance
406
in ANME-2b cells were often higher than that measured in SEEP-SRB1g, with ANME-2b cells
407
on the exterior of an exceptionally large consortium (Fig. 8b-c) featuring
15
N fractional
408
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
17
abundance of 1.73% ± 0.14 (number of ROIs,
n
= 72), significantly enriched relative to that
409
measured in SEEP-SRB1g cells in the exterior, 0.77% ± 0.09 (
n
= 58). This indicated that
410
ANME-2b were often the primary diazotroph in consortia, consistent with previous reports from
411
ANME-2–DSS consortia [8–11]. Notably,
however, in one ANME-2b—SEEP-SRB1g
412
consortium, the SEEP-SRB1g cells were more enriched in
15
N relative to the associated ANME-
413
2b cells, with ANME-2b cells containing 1.34% ± 0.13
15
N (
n
= 82) and SEEP-SRB1g containing
414
3.02% ± 0.20
15
N (
n
= 22, Fig. 8i), suggesting that under certain circumstances the sulfate-
415
reducing partner may serve as the primary diazotroph. This pattern suggests diazotrophic
416
flexibility in ANME-2b—SEEP-
SRB1g consortia in whic
h one partner–ANME-2b or SEEP-
417
SRB1g–can serve as the primary diazotroph in the consortium. Additionally, a gradient in
15
N
418
enrichment in a the large ANME-2b consortium was observed in which clusters of ANME-2b
419
cells associated with the interior of the consortia were significantly more enriched in
15
N relative
420
to ANME-2b clusters near the aggregate exterior, with
15
N fractional abundances for ANME-2b
421
cells in the exterior of 1.73% ± 0.14 (
n
= 72), significantly higher than those measured for
422
ANME-2b cells in the interior, 2.64% ± 0.14 (
n
= 116). Notably, no equivalent gradient was
423
observed in the SEEP-SRB1g partner, with SEEP-SRB1g cells in the exterior displaying
15
N
424
fractional abundances of 0.77% ± 0.09 (
n
= 58) compared with those measured on the interior,
425
0.78% ± 0.09 (
n
= 62).
426
427
Discussion
428
The symbiotic relationship between ANME and associated SRB, originally described by Hinrichs
429
[17], Boetius [4], and Orphan [21], has been the focus of extensive study using FISH [5, 7, 13,
430
15, 25, 26, 29, 34, 35], magneto-FISH [29, 37, 38], and BONCAT-FACS [39], techniques that
431
have provided insight into the diversity of partnerships between ANME and SRB. While these
432
fluorescence-based approaches offer direct confirmation of physical association between taxa
433
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
18
and are thus useful for characterizing partnership specificity, they are often constrained by
434
sample size and are comparatively lower-throughput than sequencing-based approaches. Next-
435
generation Illumina iTag sequencing of 16S rRNA amplicon sequences offers advantages in
436
terms of throughput and is rapidly becoming a standard approach in molecular microbial
437
ecology studies. Correlation analysis performed on these large iTag datasets can be an
438
effective hypothesis-generating tool for identifying microbial interactions and symbioses in the
439
environment [75], but most studies employing this approach stop short of validating predictions.
440
As correlation analysis of iTag datasets is known to be sensitive to false positives due to the
441
compositional nature of 16S rRNA amplicon libraries [41, 42, 82], specific correlations predicted
442
between taxa should be corroborated when possible by independent approaches.
443
444
In this study, we used correlation analysis of 16S rRNA iTag data from 310 methane seep
445
sediment and carbonate samples on the Costa Rican Margin to identify well-supported (pseudo-
446
p
-values < 0.01) positive correlations between specific OTUs commonly observed in seep
447
ecosystems. Our analysis identified strong correlations between syntrophic partners previously
448
described in the literature, such as that
between members of the SEEP-SRB1a and ANME-
449
2a/ANME-2c clades and between ANME-1 and SEEP-SRB2
[5, 7, 13, 15, 25, 26, 29, 34, 35],
450
and uncovered previously unrecognized relationships between members of the ANME-2b clade
451
and OTUs affiliated with an uncultured Desulfobacterales lineage, SEEP-SRB1g (Figs. 1-3). We
452
then validated the specificity of the ANME-2b and SEEP-SRB1g association by FISH (Fig. 4).
453
454
The specificity of the association between ANME-2b and SEEP-SRB1g appeared to extend
455
beyond Costa Rica methane seeps and is likely a widespread phenomenon, as this association
456
was also recovered from BONCAT-FACS datasets originating from methane seep sites off of
457
Oregon, USA (Hydrate Ridge) and from the Santa Monica Basin, California, USA. Our
458
observations of ANME-2b—SEEP-SRB1g partnership specificity in numerous samples is
459
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
19
consistent with published observations of other ANME-SRB partnerships, where consortia
460
composed of specific ANME and SRB clades have been observed in seep ecosystems
461
worldwide [15]. Notably, the syntrophic
relationship between ANME-2b and SEEP-SRB1g
462
appears to be highly specific (Fig. 2), as FISH observations from sediment samples from
463
multiple Costa Rica methane seep sites (Supp. Table 1) did not show ANME-2b in consortia
464
with other bacteria bes
ides the SEEP-SRB1g (Fig. 4). This
is in contrast with SEEP-SRB1a
465
which, in these same experiments, was found to form associations with both ANME-2a and
466
ANME-2c, indicative of this SRB syntroph having a broader capacity for establishing
467
associations with methanotrophic ANME. Members of the diverse ANME-2c lineage also
468
appeared to display partnership promiscuity in our network analysis, with positive correlations
469
observed between ANME-2c OTUs and both
SEEP-SRB1a and SEEP-SRB2 OTUs (Fig. 2).
470
This predicted partnership flexibility in the network analysis was again corroborated by FISH
471
observations of ANME-2c—SEEP-SR
B1a consortia (Fig. 4) and pr
ior reports of ANME-2c in
472
association with SEEP-SRB2 from Guaymas Basin sediments [13]. These collective data
473
suggest that partnership specificity varies among different clades of ANME and SRB, which may
474
be the result of physiological differences and/or molecular compatibility, signal exchange, and
475
recognition among distinct ANME and SRB that shape the degree of specificity between
476
particular ANME and SRB partners, as has been observed in other symbiotic associations [83–
477
85]. The degree of promiscuity or specificity for a given syntrophic partner may be influenced by
478
the co-evolutionary history of each partnership, with some ANME or SRB physiologies requiring
479
obligate association with specific partners. A more detailed examination of the genomes of
480
ANME-2b and SEEP-SRB1g alongside targeted ecophysi
ological studies may
provide clues to
481
the underlying mechanism(s) driving specificity within this ANME-SRB consortia. Comparative
482
investigations with ANME-2a and -2c subgroups may similarly uncover strategies enabling
483
broader partner association, perhaps with preference for a SRB partner shaped by
484
environmental variables rather than through pre-existing co-evolutionary relationships.
485
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
20
486
An initial genomic screening of SEEP-SRB1g offered
some insight into t
he distinct metabolic
487
capabilities of the SRB partner which may contribute to the association with ANME-2b. The
488
observation of a complete nitrogenase operon in 3 different SEEP-SRB1g genome bins
489
suggested the potential for nitrogen fixation, a phenotype not previously described for ANME-
490
associated SRB (Fig. 5). While previous work
on nitrogen utilization by
ANME-SRB consortia
491
has focused on diazotrophy performed by ANME-2 [8–10], environmental surveys of seep
492
sediments have noted active expression of nitrogenase typically associated with
493
Deltaproteobacteria [8, 86]. In these studies, the specific microorganisms associated with the
494
expressed nitrogenase in methane seep sediments were not identified. Prior to our findings
495
presented here, diazotrophy by ANME-associated SRB had not been demonstrated. A
496
phylogenetic comparison of the
nifH
sequences associated with
SEEP-SRB1g with sequences
497
of the expressed deltaproteobacte
rial-affiliated (i
.e. Group III)
nifH
transcripts reported in Dekas,
498
et al. [8] revealed a high degree of sequenc
e similarity, with SEEP-SRB1g related
nifH
among
499
the most highly expressed (Figs. 5-6).
Explicit tests for nitrogenase expression using HCR-FISH
500
and active diazotrophy using stable isotope probing and FISH-nanoSIMS confirmed the
501
involvement of SEEP-SRB1g in nitrogen fixati
on. Of the 4 ANME-2b—SEEP-SRB1g consortia
502
analyzed by FISH-nanoSIMS, one had significantly more
15
N enrichment in the SEEP-SRB1g
503
partner relative to the ANME-2b, while the other 3 displayed higher cellular
15
N enrichment in
504
the ANME-2b partner (Fig. 8). This pattern supported our inference of diazotrophic flexibility
505
within ANME-2b—SEEP-SRB1g consortia in which
either the ANME or the SRB partner can
506
serve as the primary diazotroph in the consortium. Additionally, our detection of nitrogenase
507
operons in the reconstructed genomes of t
he dominant syntrophic SRB partner, SEEP-SRB1a
508
(Supp. Fig. 6), suggests the potential for nitrogen fixation may extend to other bacterial partners
509
as well and merits further investigation. Re-examination of nitrogen fixation in these
510
partnerships with new FISH probes and nanoSIMS at
single-cell resolution will further illuminate
511
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
21
the full diversity of diazotrophic activity among ANME-SRB consortia and the associated
512
environmental/ physiological controls.
513
514
The factors responsible for determining which partner becomes the primary diazotroph in
515
ANME-2b—SEEP-SRB1g consortia
requires targeted study, but
our preliminary data suggest
516
this may be influenced in part by the relative position of ANME-2b or SEEP-SRB1g cells,
517
particularly within large (>50 μm) ANME-2b—
SEEP-SRB1g consortia. Previous studies of
518
nitrogen fixation in ANME-SRB consortia found no correlation between consortia size and
519
diazotrophic activity in consortia with diameters < 10 μm [10], but larger consortia such as those
520
presented here have not been examined at single-cell resolution. Additionally, consortia with the
521
morphology observed here, in which ANME-2b cells form multiple sarcinal clusters surrounded
522
by SEEP-SRB1g (Figs. 4b, 8), have not been t
he specific focus of nanoSIMS analysis but
523
appear to be the common morphotype among ANME-2b—SEEP-SRB1g consortia [31]. The
524
frequency with which this mor
photype is observed in ANME
-2b—SEEP-SRB1g consortia may
525
be related to the underlying physiology of this specific partnership. NanoSIMS analysis of a
526
particularly large ANME-2b—SEEP-SRB1g consorti
um (~200 μm) with this characteristic
527
morphology (Fig. 8a-f) revealed a gradient in diazotrophic activity in which ANME-2b cells
528
located in the interior of the consortium incorporated far more
15
N from
15
N
2
than ANME-2b cells
529
near the exterior. This pattern may be related to variations in nitrogen supply from the external
530
environment, as similar patterns of nutrient depletion with increasing depth into microbial
531
aggregates have been predicted in modeling studies of nitrate uptake in
Trichodesmium
sp. [87]
532
and directly observed by SIMS in stable isotope probing studies of carbon fixation in biofilm-
533
forming filamentous cyanobacteria [88]. In these examples, modeling and experimental results
534
document declining nitrate or bicarbonate ion availability inwards toward the center of the
535
aggregates resulting from nitrate or bicarbonate consumption. An analogous process may occur
536
in large ANME-2b—SEEP-SRB1g consortia, where cells situated closer to the exterior of the
537
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
22
consortium assimilate environmental NH
4
+
, increasing nitrogen limitation for cells within the
538
consortium core. Intere
stingly, the singl
e consortium in which the SEEP-SRB1g partner was the
539
inferred primary diazotroph featured SEEP-SRB1g cells in the core of this consortium with
540
ANME-2b cells toward the exterior (Fig. 8). The current nanoSIMS dataset is small and
541
determining the biotic and environmental factors that influence which partner serves as the
542
primary diazotroph in ANME-2b—SEEP-SRB1g consortia necessitates further study, but a
543
reasonable hypothesis is
that the proximity of cells in a given ANME-2b—SEEP-SRB1g
544
consortium relative to the consortium exterior (and NH
4
+
availability in the surrounding pore
545
fluid) influences the spatial patterns of diazotrophic activity by both ANME and SRB in large
546
consortia. The concentration of ammonium in seep porefluids can be highly variable over
547
relatively small spatial scales (e.g. between 50 - 300 μM [80]), and rates of diazotrophy
548
estimated from laboratory incubations of methane seep sediment samples indicate different
549
threshold concentrations of NH
4
+
(aq)
above which diazotrophy ceases, as low as 25 μM [89] to
550
100-1000 μM [90–92]. In the large consortia observed here, this threshold [NH
4
+
(aq)
] may be
551
crossed within the consortium as NH
4
+
is assimilated by cells at the consortium exterior,
552
inducing nitrogen limitation and diazotrophy by ANME or SRB near the consortium core. Given
553
the importance of diazotrophy in ANME-SRB consortia for nitrogen cycling at methane seep
554
communities [10, 89], future work should test this hypothesis with SIP incubations with
15
N
2
555
under variable [NH
4
+
(aq)
].
556
557
The observed variation in
diazotrophic activity in ANME-2b or SEEP-SRB1g cells may also be
558
the result of phenotypic heterogeneity [93] wi
thin the multicellu
lar ANME-2b—SEEP-SRB1g
559
consortia, in which expr
ession of the nitrogenase operon
that ANME-2b and SEEP-SRB1g
560
partners both possess is an emergent behavior resulting from the spatial organization of ANME-
561
2b and SEEP-SRB1g cells within the consortium.
On the basis of nanoSIMS observations of
562
heterogeneous diazotrophy in clonal
Klebsiella oxytoca
cultures, phenotypic heterogeneity was
563
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
23
inferred to confer selective advantage on microbial communities by enabling rapid response to
564
environmental fluctuations [94]. Similar heterogeneity in
nif
expression by ANME-2b or SEEP-
565
SRB1g cells may provide partners with resilie
nce against changes in
environmental nitrogen
566
supply. Corroborating these observations in diverse ANME-SRB consortia and direct coupling of
567
single-cell mRNA expression with nanoSIMS-acquired
15
N enrichment would further inform the
568
degree to which relative arrangement of the partners and spatial structure within a consortium
569
plays a significant role in determining the mode of nutrient or electron transfer between partners.
570
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint
Correlation analysis reveals new AOM partnerships
24
Conclusions
571
Here, we present an effective approach to detect novel pairings of microbial symbionts by
572
coupling correlation analysis of 16S rRNA amplicon libraries with FISH and BONCAT-FACS
573
experiments, going beyond amplicon sequencing-based hypothesis generation to experimental
574
validation of hypothesized partnerships using microscopy and single-cell sorting techniques.
575
Correlation analysis performed on a 16S amplicon survey of methane seep sediments near
576
Costa Rica uncovered a novel and highly-specific ANME-SRB partnership between ANME-2b
577
and SEEP-SRB1g. This partnership specificity was then validated by FISH, and further
578
corroborated by 16S rRNA amplicon sequences
from BONCAT-FACS-sorted single ANME-SRB
579
consortia from methane seep sediments near Costa Rica, Hydrate Ridge, and Santa Monica
580
Basin in California. Preliminary genomic
screening of representative genomes from SEEP-
581
SRB1g uncovered potential for nitrogen fixation in these genomes. Examination of published
582
nifH
cDNA clone libraries [8] and transcriptomic data [14] prepared from methane seep
583
sediments demonstrated that S
EEP-SRB1g actively expresses
nifH
in vivo
. Co-localization of
584
signal for
nifH
mRNA and SEEP-SRB1g 16S rRNA by HCR-FISH further corroborated active
585
transcription of
nifH
by SEEP-SRB1g. FISH-nanoSIMS analysis of ANME-2b—SEEP-SRB1g
586
consortia grown with
15
N
2
headspace documented
15
N incorporation in SEEP-SRB1g cells,
587
suggesting that SEEP-SRB1g may fix nitrogen as well. Future work should focus on examining
588
unique aspects of each ANME-SRB partnership to improve our understanding of the diversity of
589
anaerobic methane oxidation symbioses endowed by evolution.
590
.
CC-BY-NC-ND 4.0 International license
author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is the
.
https://doi.org/10.1101/2020.04.12.038331
doi:
bioRxiv preprint