a year in the life of a thrombolite: comparative … · 2019. 8. 17. · a year in the life of a...

A year in the life of a thrombolite: comparativemetatranscriptomics reveals dynamic metabolicchanges over diel and seasonal cycles

Artemis S. Louyakis,1 Hadrien Gourl�e,1,2

Giorgio Casaburi,1 Rachelle M. E. Bonjawo,1

Alexandrea A. Duscher1 and Jamie S. Foster1*1Department of Microbiology and Cell Science,

University of Florida, Space Life Sciences Lab,

Merritt Island, FL, USA.2Department of Animal Breeding and Genetics,

Global Bioinformatics Centre, Swedish University of

Agricultural Sciences, Uppsala, Sweden.

Summary

Microbialites are one of the oldest known ecosys-

tems on Earth and the coordinated metabolisms and

activities of these mineral-depositing communities

have had a profound impact on the habitability of the

planet. Despite efforts to understand the diversity

and metabolic potential of these systems, there has

not been a systematic molecular analysis of the tran-

scriptional changes that occur within a living

microbialite over time. In this study, we generated

metatranscriptomic libraries from actively growing

thrombolites, a type of microbialite, throughout diel

and seasonal cycles and observed dynamic shifts in

the population and metabolic transcriptional activity.

The most transcribed genes in all seasons were asso-

ciated with photosynthesis, but only transcripts

associated with photosystem II exhibited diel cycling.

Photosystem I transcripts were constitutively

expressed at all time points including midnight and

sunrise. Transcripts associated with nitrogen fixa-

tion, methanogenesis and dissimilatory sulfate

reduction exhibited diel cycling, and variability

between seasons. Networking analysis of the meta-

transcriptomes showed correlated expression

patterns helping to elucidate how metabolic interac-

tions are coordinated within the thrombolite

community. These findings have identified distinctive

temporal patterns within the thrombolites and will

serve an important foundation to understand the

mechanisms by which these communities form and

respond to changes in their environment.

Introduction

Microbially induced mineralization is a critical process that

has helped shape Earth’s geological landscape (Dupraz

et al., 2009). This process occurs via the metabolic activi-

ties of lithifying microbial communities that facilitate the

precipitation and/or the entrapment of inorganic materials,

which together form depositional structures known as

microbialites (Burne and Moore, 1987; Reid et al., 2000;

Dupraz and Visscher, 2005; Dupraz et al., 2009). The abil-

ity of these microbialite-forming communities to alter their

biological and geological environments has enabled these

ecosystems to adapt to a wide range of environmental con-

ditions over the course of Earth’s history (Walter, 1994;

Grotzinger and Knoll, 1999). Microbialites are located

throughout the globe in a wide range of habitats (e.g.,

lacustrine, marine and hypersaline) and their ecological

complexity is just now beginning to be revealed through

advancements in high-throughput sequencing of the taxo-

nomic and functional gene diversity (Breitbart et al., 2009;

Khodadad and Foster, 2012; Petrash et al., 2012;

Bernhard et al., 2011; Mobberley et al., 2013; Russell

et al., 2014; Valdespino-Castillo et al., 2014; Sagha€ı et al.,

2015; Casaburi et al., 2016; Cerqueda-Garcia and Falcon,

2016; Chagas et al., 2016; Ruvindy et al., 2016; Warden

et al., 2016; White et al., 2016; Alcantara-Hern�andez et al.,

2017; Louyakis et al., 2017).

There are two major categories of microbialites, includ-

ing stromatolites, with well-laminated internal structures

(Walter, 1994; Reid et al., 2000) and thrombolites, which

have irregular clotted fabrics (Aitken, 1967; Kennard and

James, 1986). Although there has been extensive

research on the metabolisms and organisms associated

with formation of stromatolites (Visscher et al., 2000;

Visscher et al., 1998; Stolz et al., 2001; Burns et al., 2004;

Papineau et al., 2005; Havemann and Foster, 2008; Allen

et al., 2009; Baumgartner et al., 2006; Foster et al., 2009;

Foster and Green, 2011; Goh et al., 2011; Berelson et al.,

Received 9 September, 2017; revised 12 December, 2017;accepted 12 December, 2017. *For correspondence. E-mail [email protected]; Tel. 321-525-1047.

VC 2017 Society for Applied Microbiology and John Wiley & Sons Ltd

Environmental Microbiology (2018) 20(2), 842–861 doi:10.1111/1462-2920.14029

2013; Foster and Green, 2011; Khodadad and Foster,

2012; Pepe-Ranney et al., 2012; Mata et al., 2012; Casa-

buri et al., 2016; Ruvindy et al., 2016; Suosaari et al.,

2016), far fewer studies have examined thrombolite-

forming communities (Desnues et al., 2008; Planavsky

et al., 2009; Myshrall et al., 2010; Mobberley et al., 2012;

Edgcomb et al., 2013; Mobberley et al., 2013; Gleeson

et al., 2015; Warden et al., 2016; Louyakis et al., 2017).

One ecosystem, where both microbialite types co-occur

and form calcium carbonate build-ups, is the island of

Highborne Cay, The Bahamas (Myshrall et al., 2010). In

these carbonate-based microbialites, it is the community

composition coupled with the type and rate of the metabo-

lism that alters the microenvironment, thereby creating

conditions that either drive the alkalinity towards precipita-

tion or dissolution of minerals (for review Dupraz et al.,

2009). For example, some metabolisms (e.g., oxic and

anoxygenic photosynthesis, and some forms of sulfate

reduction) increase the local pH thereby promoting condi-

tions for mineral deposition, whereas other metabolisms

(e.g., aerobic respiration, dissimilatory nitrate reduction

and sulfide-oxidation) can result in a net loss of carbonate

(Dupraz and Visscher, 2005; Visscher and Stolz, 2005). In

addition to these intrinsic factors driving local alkalinity, the

precipitation of carbonate is also dependent on the pres-

ence of exopolymeric substances (EPS). EPS material

creates an organic matrix within the microbialite-forming

mat community and serves as integral nucleation sites

(Braissant et al., 2007; Dupraz et al., 2009; Planavsky

et al., 2009). The negatively charged acidic groups within

the EPS matrix bind free Ca21, and through modification

or degradation by heterotrophic microbes, the Ca21 is

released. The local increase in free Ca21 and alkalinity

results in biologically induced mineralization within the mat

communities. However, the specific metabolic interactions

and underlying molecular pathways of these metabolisms

have not been fully elucidated in microbialites.

Despite similarities in the overall processes of biologi-

cally induced mineralization, Bahamian stromatolites

undergo an iterative growth process through microbial mat

cycling and environmental controls (Visscher et al., 2000;

Bowlin et al., 2012), whereas thrombolites form as a result

of the activity of bundles of calcified cyanobacteria fila-

ments, primarily the taxa Dichothrix spp. of the family

Rivulariaceae (Planavsky et al., 2009; Louyakis et al.,

2017). On the surface of the Bahamian thrombolite plat-

forms (Fig. 1A), there are numerous nodular mat

formations called ‘button’ mats (Fig. 1B) (Myshrall et al.,

2010; Mobberley et al., 2013) that are comprised of verti-

cally orientated Dichothrix spp. filaments (Fig. 1C). The

filaments serve as ‘hot-spots’ for carbonate deposition

(Fig. 1D) and previous isotopic analyses have revealed the

precipitate is derived from photosynthetic activity (Planav-

sky et al., 2009; Louyakis et al., 2017).

Although the taxonomic complexity and metabolic poten-

tial of thrombolite-forming communities is now emerging,

the transcriptional activities are not well known. Only a sin-

gle metatranscriptome analysis of a living microbialite has

been completed (Mobberley et al., 2015), and, although it

targeted the Bahamian thrombolites, it was limited to a sin-

gle time-point (Mobberley et al., 2015). The previous RNA-

Seq study revealed that, at noon, there was pronounced

transcriptional activity that formed a distinctive vertical

0

5

10

15

20

25

30

35

0 2 4 6 8

10 12 14 16

October November December January February March April May June July August September

16

14

12

10

8

6

4

2

0

Day

light

(h)

Temp Max

Temp Min

35

30

25

20

15

10

5

0

Temperature (˚C

)

October

Novem

ber

Decem

ber

January

February

March

April

May

June

JulyA

ugust

Septembe r

E

DC

BA

Fig. 1. Thrombolite of Highborne Cay, The Bahamas.

A. Thrombolite platforms located in the intertidal zone. Red box

reflected area magnified in (B). Bar, 1 m.

B. Thrombolite ‘button’-forming mats found on the surface of the

platforms. Green box represents area magnified in (C). Bar, 2 cm.

C. Cross section of ‘button’ mat depicting the extensive field of

vertically orientated Dichothrix filaments. Blue box reflects area

magnified in (D). Bar, 3 mm.

D. Light micrograph of Dichothrix bundles showing extensive

carbonate precipitation along the filaments. Bar, 1 mm.

E. Maximum and minimum air temperatures for the duration of

experiment from October 2013 thru September 2014. Grey area

reflects the hours of daylight during the same time period. Asterisks

indicate months in which the thrombolite-forming mats were

sampled. [Colour figure can be viewed at wileyonlinelibrary.com]

A year in the life of a thrombolite 843

VC 2017 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology, 20, 842–861

http://wileyonlinelibrary.com

gradient of metabolisms within the thrombolite-forming

mats (Mobberley et al., 2015), however, how these activi-

ties change over the diel cycle and throughout the year is

not yet known. In this study, we provide the first transcrip-

tional analysis of a lithifying microbial mat system

throughout the diel cycle replicated over the course of a

year. A comparative metatranscriptomic approach was

used to elucidate the activity of the microbial consortium

associated with the thrombolite of Highborne Cay. Overall,

this study provides important snapshots of the thrombolites

throughout the day and year to more fully characterize the

microbiome of these lithifying systems, as well as under-

stand how these communities interact and coordinate their

activities to drive the formation of these structures in an

ever-changing environment.

Results and discussion

To examine how the metabolic activities of thrombolite-

forming mats of Highborne Cay change over time, meta-

transcriptomic libraries were generated throughout the diel

cycle and replicated over the course of one year. The

thrombolite-forming button-mats from Site 4 (Andres and

Reid, 2006) were sampled every 6 h, for two consecutive

days, in the months of October (2013), March (2014) and

August (2014). The diel time points were chosen to reflect

the peak of photosynthetic (12:00) and respiration (00:00)

activities of the button-mat communities (Myshrall et al.,

2010) as well as the light/dark transitions at approximately

sunrise (06:00) and sunset (18:00). Although day length

did fluctuate between the different months (Fig. 1E), the

objective was to ensure a consistent sampling regime

between the treatments. The months were chosen as they

reflect the major seasonal transitions in The Bahamas, fall

(October), winter/early spring (March) and summer

(August) and previous studies have ascertained key envi-

ronmental changes (e.g., temperature, burial events,

photosynthetic active radiation) between these Bahamian

seasons (Bowlin et al., 2012). At the time of collection, the

water temperature was 288C in October 2013, 248C in

March 2014 and 308C in August 2014. The full dynamic

ranges of both water and air temperatures throughout

October 2013 to September 2014 are presented in

Fig. 1E.

In total, 24 libraries were constructed for four time-points

over three seasons with two biological replicates for each.

More than 1 billion reads were recovered and after quality

filtering, 687 million rRNA and 305 million mRNA reads

were retained (Table 1). Sequence data from all seasons

and time-points was assembled using Trinity v2.2.0 (Garber

et al., 2011). A total of 890 921 contigs were recovered with

reads mapped to them. Contigs were annotated to 13 983

KEGG Orthologies (KOs; Supporting Information Table

S1). Genes were considered differentially expressed (DE)

Table 1. Metatranscriptome data summary for four timepoints over three seasons sampled from Highborne Cay thrombolite-forming mats.

Sample Month Timepoint # of reads # of post-QC readsa % rRNA % remaining % alignedb

oct06_1 October 6:00 39609094 33689382 80.4 19.6 59.4

oct12_1 October 12:00 77945514 66083045 71.6 28.4 60.4

oct18_1 October 18:00 71433480 59414453 70.9 29.1 55.4

oct00_1 October 0:00 39379596 31584081 79.7 20.3 55.2

oct06_2 October 6:00 35044080 29260586 71.0 29.0 46.0

oct12_2 October 12:00 75711876 65456261 56.2 43.8 46.8

oct18_2 October 18:00 46851856 41098015 71.7 28.3 46.3

oct00_2 October 0:00 75785140 65230103 76.2 23.8 46.2

mar06_1 March 6:00 42286936 37108069 84.0 16.0 61.9

mar12_1 March 12:00 43962956 36332970 53.5 46.5 42.9

mar18_1 March 18:00 18040772 15057539 55.1 44.9 44.6

mar00_1 March 0:00 30358562 26183784 56.7 43.3 47.2

mar06_2 March 6:00 50796354 42258844 54.8 45.2 44.7

mar12_2 March 12:00 34809204 29685603 73.6 26.4 45.6

mar18_2 March 18:00 47205018 40876339 44.1 55.9 47.3

mar00_2 March 0:00 78846408 68104175 55.5 44.5 46.4

aug06_1 August 6:00 38970248 27423407 73.9 26.1 38.3

aug12_1 August 12:00 31367092 25207290 86.9 13.1 56.1

aug18_1 August 18:00 63575558 46390757 69.7 30.3 41.3

aug00_1 August 0:00 46398488 31001238 71.3 28.7 32.1

aug06_2 August 6:00 51792988 42670313 74.7 25.3 53.8

aug12_2 August 12:00 55154298 46063430 75.8 24.2 59.5

aug18_2 August 18:00 55090470 45178031 81.3 18.7 68.6

aug00_2 August 0:00 49895344 41203480 82.5 17.5 64.0

a. Quality trimming completed using Trimmomatic within the Trinity pipeline.b. Percents are non-rRNA portion of raw reads aligned to Trinity assembly using Bowtie2.

844 A. S. Louyakis et al.


with p-values�0.01 through a NoiSEQBio DE analysis.

Mapping of reads to assembled contigs resulted in low

overlap between diel time points and seasons. The paucity

of sequenced genomes and metagenomes from these sys-

tems does represent a limitation with regard to annotation

and assembly. Additionally, we also acknowledge that only

a single year is represented in this study and that there

may be year-to-year variation within the Highborne Cay

ecosystem. Despite this potential caveat, the differences

between the time points do suggest that the transcriptional

activities within mat community are dynamic throughout the

day and year.

Two replicate metatranscriptomes were generated for

each time point and reproducibility was assessed using

Pearson and Spearman’s rank correlation coefficients.

Pearson correlations for biological replicates ranged from

0.061 to 0.793, with a mean of 0.487 and all p-value

<10211, and Spearman’s rank correlation ranged from

0.296 to 0.506, with a mean of 0.446 and all p-value

<10211 (Supporting Information Table S2). The standard

Pearson correlation for technical replicates of a transcrip-

tome is� 0.9, while no standard has yet been devised for

biological replicates (Conesa et al., 2016). Although efforts

were made to collect samples from across the thrombolite

platform in Site 4 (Fig. 1A), the lower correlation values

likely reflect some spatial heterogeneity within the

thrombolite-forming community. Despite the potential het-

erogeneity, distinctive, reproducible trends were observed

throughout the thrombolite-forming community over time.

Taxonomic dynamics within thrombolite-forming buttonmat community

During library construction, the samples underwent rRNA

depletion to help improve mRNA recovery. To test whether

rRNA depletion would have a biasing effect on the recov-

ered taxonomic data, a parallel study was completed using

a previously published dataset from samples collected and

handled identically with the exception that one was rRNA

depleted and the other was not (NCBI Bioproject ID:

PRJNA261361) (Mobberley et al., 2015). A Kruskal-Wallis

test (p>0.319) showed no significant difference between

rRNA depleted and total RNA samples suggesting the

depletion step did not significantly alter the relative taxo-

nomic abundances. Additionally, Pearson and Spearman

tests showed strong linear correlation (0.965 and 0.892

correlation coefficients, respectively, p-values< 10216),

also suggesting that depletion has no distinguishable effect

on community structure.

The metatranscriptome data was mined for 16S rRNA

transcripts to compare relative abundances of bacteria and

archaea within the community over diel and seasonal

cycles. The use of rRNA transcript levels as a proxy for

microbial activity does have limitations due to differences

in growth rates, life strategies and environmental condi-

tions (Blazewicz et al., 2013). However, rRNA-based

analyses can yield important insight into the community

dynamics and provide qualitative information about those

taxa that have the potential for protein synthesis and how

they fluctuate over time. Together, analyses identified a

total of 23 bacterial phyla, three archaeal phyla and an

additional 26 candidate phyla that exhibited enriched rRNA

transcript expression within the mats over the observed

diel and seasonal cycles. A cladogram and heat map

depicting the differential expression of the 16S rRNA

genes for each of the time points revealed extensive

changes in transcript abundance throughout the day and

between seasons with the node size reflecting the total

16S rRNA transcript number for a given taxa (Fig. 2). Of

the many phyla represented within the thrombolite

communities the most diversely represented within the

metatranscriptomes were the Proteobacteria, Bacteroi-

detes and Cyanobacteria regardless of season (Fig. 2). A

detailed list of those taxa with the highest levels of 16S

rRNA transcripts through the diel cycle and seasons are

presented in Table 2, Supporting Information Fig. S2 and

Table S3. Although Proteobacteria (specifically Gammapro-

teobacteria, Alphaproteobacteria and Deltaproteobacteria)

populations represented the most diverse assemblage of

taxa, the individual taxa that were the most enriched

within the thrombolite-forming mat transcriptomes

were the Cyanobacteria (Table 2, Supporting Information

Table S3).

Regardless of season and time of day, the cyanobacte-

rium with the highest level of transcripts within the

community shared sequence similarity with the Xenococ-

caceae, a family of coccoid cyanobacteria within the Order

Pleurocapsales. This result was visually confirmed with

light- and epifluorescence microscopy, where coccoid cya-

nobacteria were observed in high numbers in the upper cm

of the thrombolite-forming mats (data not shown). This

taxon, along with Pseudoanabaenaceae, which were also

highly represented within the metatranscriptome (Table 2),

are abundant in other lithifying mat systems, such as

Shark Bay, Western Australia (Goh et al., 2009; Suosaari

et al., 2016), Lake Alchichica, Mexico (Ka�zmierczak et al.,

2011) and in Lake Van, Turkey (Kazmierczak et al., 2009).

Coccoid cyanobacteria have been shown to have impor-

tant roles in lithifying systems, as carbonate precipitation

initiates near clusters of coccoid cyanobacteria in some

hypersaline mats (Dupraz et al., 2004). Additionally,

numerous microfossils, dating back to the Archean, have

been interpreted as calcified, colonial coccoid cyanobacte-

ria (Kazmierczak and Altermann, 2002; Altermann et al.,

2006) suggesting their importance in biologically induced

mineralization within the thrombolites.

Unsurprisingly, another highly represented cyanobacte-

rial taxa within the recovered metatranscriptome was the



family Rivulariaceae, which harbors the filamentous

Dichothrix spp., known to serve as hot spots for carbon-

ate precipitation within the thrombolites (Planavsky

et al., 2009; Louyakis et al., 2017). Much like the coc-

coid cyanobacteria, the filamentous strains exhibited

very high numbers of transcript levels regardless of sea-

son. Although the cyanobacteria transcript levels peaked

at noon, high levels of 16S rRNA transcripts were

observed at night (Table 2). The accumulation or main-

tenance of rRNA content during the dark periods has

been observed in numerous cultured cyanobacterial

studies (Lepp and Schmidt, 1998; Zinser et al., 2009;

Beck et al., 2014) and may provide the cyanobacteria

with a competitive advantage to respond quickly during

the onset of the light period in the thrombolite

communities.

BA

C

Archaea

Archaea

Bacteria

Bacteria

γ

α

α

γ

δ δ

γ

α

δ

α

γ

γγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγγ

ααααααααααααααααααα

δδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδδ

Archaea

Bacteriaα

γ

δ

Fig. 2. Composition of thrombolite active taxa as visualized with GraPhlAn. Cladograms represent metatranscriptome bacterial and archaealdiversity for each season,

A. October 2013, (B) March 2014 and (C) August 2014. Cladogram was constructed using total abundances for the seasons with size of

nodes increasing with abundance and colored by phylum. The innermost node (white) represents domain, then phylum, class, order, family,

genus and species on the outermost node. Outer rings represent relative abundance heatmaps for each time-point, innermost is 06:00 (red),

then 12:00 (green), 18:00 (orange) and 00:00 (blue) on the outside. [Colour figure can be viewed at wileyonlinelibrary.com]




Tab

le2.

Rela

tive

abundance

of

recove

red

16S

rRN

Atr

anscri

pts

(means

show

n)

within

thro

mbolit

es

ove

rdie

land

seasonalcycle

s.

Tota

lTra

nscripts

per

Mill

ion

Octo

ber

06:0

0O

cto

ber

12:0

0O

cto

ber

18:0

0O

cto

ber

00:0

0P

hylu

mC

lass

Ord

er

Fam

ily

471786

100352

129120

149798

92516

Cyanobacte

ria

Oscill

ato

riop

hycid

eae

Chro

ococcale

sX

enococcaceae

422906

124770

69827

87993

140317

Cyanobacte

ria

Nosto

cophycid

eae

Stigonem

ata

les

Riv

ula

riaceae

398288

141842

52996

58360

145090

Cyanobacte

ria

Nosto

cophycid

eae

Nosto

cale

sN

osto

caceae

270337

94635

35655

41652

98394

Cyanobacte

ria

Synechococcophycid

eae

246991

84135

35088

41728

86040

Cyanobacte

ria

Nosto

cophycid

eae

Nosto

cale

s

242077

73127

38795

50385

79769

Cyanobacte

ria

Synechococcophycid

eae

Pseudanabaenale

sP

seudanabaenacea

e

188047

35170

59275

60406

33197

Pro

teobacte

ria

Alp

hapro

teobacte

ria

Rhodobacte

rale

sR

hodobacte

raceae

184577

44170

33857

72400

34150

Cyanobacte

ria

Oscill

ato

riop

hycid

eae

Chro

ococcale

s

160582

43984

30519

36718

49362

Cyanobacte

ria

76399

8070

18973

39575

9781

Pro

teobacte

ria

Deltapro

teobacte

ria

Myxococcale

s

67910

16515

27914

8729

14752

Pro

teobacte

ria

Alp

hapro

teobacte

ria

Rhiz

obia

les

Hyphom

icro

bia

ceae

60652

11838

19604

17986

11225

Pro

teobacte

ria

Alp

hapro

teobacte

ria

44223

3066

9659

28937

2561

Cyanobacte

ria

Oscill

ato

riop

hycid

eae

Oscill

ato

riale

sP

horm

idia

ceae

43677

6986

12938

16185

7569

Chlo

roflexi

Anaero

lineae

SB

R1031

A4b

41721

6142

15348

11002

9229

Bacte

roid

ete

s[S

apro

spirae]

[Sapro

spirale

s]

Sapro

spiraceae

38867

11817

11684

7627

7739

Cyanobacte

ria

Synechococcophycid

eae

Pseudanabaenale

sP

seudanabaenacea

e

34210

8815

8697

9152

7546

Pla

ncto

mycete

sP

hycis

phaera

eP

hycis

phaera

les

28062

6092

8043

7688

6239

Pla

ncto

mycete

sP

lancto

mycetia

Pirellu

lale

sP

irellu

laceae

26449

4659

8707

8012

5071

Pro

teobacte

ria

Deltapro

teobacte

ria

Myxococcale

sO

M27

Tota

lTra

nscripts

per

Mill

ion

Marc

h06:0

0M

arc

h12:0

0M

arc

h18:0

0M

arc

h00:0

0P

hylu

mC

lass

Ord

er

Fam

ily

699229

161877

193626

171318

172407

Cyanobacte

ria

Oscill

ato

rio

phycid

eae

Chro

ococcale

sX

enococcaceae

205100

46561

48116

58289

52134

Pro

teobacte

ria

Alp

hapro

teobacte

ria

Rhodobacte

rale

sR

hodobacte

raceae

190576

38152

56897

48108

47419

Cyanobacte

ria

Oscill

ato

rio

phycid

eae

Chro

ococcale

s

187291

66563

55305

42924

22498

Cyanobacte

ria

Nosto

cophycid

eae

Stigonem

ata

les

Riv

ula

riaceae

153678

58452

44890

33011

17324

Cyanobacte

ria

Nosto

cophycid

eae

Nosto

cale

sN

osto

caceae

137779

44032

42180

28172

23396

Cyanobacte

ria

Synechococcophycid

eae

Pseudanabaena

les

Pseudanabaenacea

e

110081

42697

31578

23601

12204

Cyanobacte

ria

Synechococcophycid

eae

108551

21519

34193

25054

27784

Pro

teobacte

ria

Deltapro

teobacte

ria

Myxococcale

s

103295

40250

28379

22655

12011

Cyanobacte

ria

Nosto

cophycid

eae

Nosto

cale

s

96388

19946

23715

30490

22237

Pro

teobacte

ria

Alp

hapro

teobacte

ria

77839

12260

17150

28055

20373

Pla

ncto

mycete

sP

lancto

mycetia

Pirellu

lale

sP

irellu

laceae

74885

25602

20515

19265

9502

Cyanobacte

ria

65301

16617

17808

15958

14918

Cyanobacte

ria

Synechococcophycid

eae

Pseudanabaena

les

Pseudanabaenacea

e

63135

17221

12648

15682

17582

Pro

teobacte

ria

Alp

hapro

teobacte

ria

Rhiz

obia

les

Hyphom

icro

bia

ceae

53011

12636

11410

11411

17554

Cyanobacte

ria

Oscill

ato

rio

phycid

eae

Chro

ococcale

sC

yanobacte

riaceae

48757

12677

11162

9677

15241

Chlo

roflexi

Anaero

lineae

SB

R1031

A4b

48450

10228

10674

11782

15765

Bacte

roid

ete

s[S

apro

spirae]

[Sapro

spirale

s]

Sapro

spiraceae

47608

12837

12355

10712

11703

Pro

teobacte

ria

Deltapro

teobacte

ria

Myxococcale

sO

M27

47133

9663

13716

13186

10569

Pla

ncto

mycete

sP

hycis

phaera

eP

hycis

phaera

les



Other highly represented taxa within the thrombolite

metatranscriptomes included the alphaproteobacterial

families Rhodobacteraceae and Rhizobiales, deltaproteo-

bacterial order Mycoccocales and the Planctomycetes

order Phycisphaerales, which are common in a wide range

marine ecosystems and are thought to play important roles

in biochemical cycling (Baker et al., 2015; Simon et al.,

2017). These taxa exhibited relatively consistent expres-

sion of rRNA throughout the diel cycle and over the course

of the year (Table 2, Supporting Information Table S3).

Those enriched taxa that did exhibit seasonal changes

in 16S rRNA expression were the Bacteroidetes, Chloro-

flexi and Firmicutes. In the Bacteroidetes, the families

Saprospiraceae, Cytophagaceae and the cluster BME43

exhibited increased relative abundance in March and Octo-

ber. Heterotrophic Saprospiraceae are known to hydrolyse

complex carbon compounds and may be contributing to

the degradation of exopolymeric substances (Fernandez-

Gomez et al., 2013; McIlroy and Nielson, 2014). The deg-

radation of EPS is an essential process to release calcium

ions during the precipitation of calcium carbonate (Dupraz

et al., 2009) and have been reported in a wide range of lith-

ifying hypersaline mat systems (Mobberley et al., 2012;

Schneider et al., 2013). In the warmer month of August,

however, there was an overall decrease in the representa-

tion of Bacteroidetes, with the exception of the family

Flammeovirgaceae, which has been shown to exhibit ther-

motolerance (Tang et al., 2016) and form associations with

cyanobacteria in marine ecosystems, such as corals (Den

Uyl et al., 2016).

There were also seasonal changes in the orders of the

Chloroflexi and Firmicutes (Table 2). Among the Chloro-

flexi, the anoxic non-phototrophic Anaerolineae (Class

SBR1031) were enriched in March and October, but lower

in August, whereas the photoheterotrophic Chloroflexales

(Candidate family Chlorothrixaceae) were enriched in

August. These differences in the relative abundance may

simply reflect differences in light levels throughout the

year. In the Firmicutes, the families Bacillaceae and Plano-

coccaceae were also enriched at sunset and midnight in

August (Table 2) and have been isolated from a wide range

of extreme environments including salterns, permafrost,

hyperalkaline lakes and microbial mats (Reddy et al.,

2002; Sorensen et al., 2005). Members of both families

have been shown to have a high tolerance to elevated UV

and salt conditions (Ordonez et al., 2009; Paul et al., 2015;

See-Too et al., 2017) and the increased relative abun-

dance in August may reflect resistance to elevated UV

radiation and desiccation stress in the thrombolites during

the summer.

Although Bacteria dominated the metatranscriptome at

all time points and seasons, Archaea transcripts were

recovered with similarity to three phyla, the methanogenic

Euryarchaeota, ammonia-oxidizing Crenarchaeota and theTab

le2.

cont.

Tota

lTra

nscri

pts

per

Mill

ion

August

06:0

0A

ugust

12:0

0A

ugust

18:0

0A

ugust

00:0

0P

hylu

mC

lass

Ord

er

Fam

ily

603314

215328

253340

71615

63030

Cya

nobacte

ria

Oscill

ato

riop

hycid

eae

Chro

ococcale

sX

enococcaceae

436844

144003

201465

53328

38048

Cya

nobacte

ria

Oscill

ato

riop

hycid

eae

Chro

ococcale

s

328293

14328

2687

213623

97656

Fir

mic

ute

sB

acill

iB

acill

ale

sB

acill

aceae

326312

11124

2960

223642

88587

Pro

teobacte

ria

Gam

mapro

teobacte

ria

Ente

robacte

riale

sE

nte

robacte

riaceae

236338

44024

13856

40771

137687

Cya

nobacte

ria

Nosto

cophycid

eae

Stigonem

ata

les

Riv

ula

riaceae

193247

56177

69714

20780

46576

Cya

nobacte

ria

Synechococcophycid

eae

Pseudanabaenale

sP

seudanabaen

aceae

137630

62496

48752

8005

18376

Cya

nobacte

ria

Synechococcophycid

eae

Pseudanabaenale

sP

seudanabaen

aceae

130254

17075

7593

38773

66813

Cya

nobacte

ria

Nosto

cophycid

eae

Nosto

cale

sN

osto

caceae

81215

23799

18666

7107

31642

Cya

nobacte

ria

77846

23864

18918

19758

15307

Pla

ncto

mycete

sP

hycis

phaera

eP

hycis

phaera

les

72014

9553

5003

18148

39310

Cya

nobacte

ria

Nosto

cophycid

eae

Nosto

cale

s

68374

26915

27687

3530

10241

Pro

teobacte

ria

Alp

hapro

teobacte

ria

68177

25647

19291

9740

13499

Pro

teobacte

ria

Deltapro

teobacte

ria

Myxococcale

s

57704

20834

19602

1564

15704

Pro

teobacte

ria

Alp

hapro

teobacte

ria

Rhodobacte

rale

sR

hodobacte

race

ae

49418

7442

3577

12960

25437

Cya

nobacte

ria

Synechococcophycid

eae

44860

7166

3694

25487

8512

Chlo

roflexi

Chlo

roflexi

Chlo

roflexale

s[C

hlo

roth

rixaceae]

44710

15155

21369

4785

3401

Cya

nobacte

ria

Oscill

ato

riop

hycid

eae

Chro

ococcale

sC

yanobacte

riaceae

43108

1404

355

27268

14081

Fir

mic

ute

sB

acill

iB

acill

ale

sP

lanococcaceae

30239

8098

4768

13410

3963

Bacte

roid

ete

sC

yto

phagia

Cyto

phagale

sF

lam

meovirgaceae

30142

9551

7905

3376

9311

Pla

ncto

mycete

sP

lancto

mycetia

Pirellu

lale

sP

irellu

laceae



newly proposed archaeal phyla Parvarchaeota (Rinke

et al., 2013) (Supporting Information Table S3). Archaea

had very low transcriptional activity in all seasons com-

pared to Bacteria, but were highest in August, comprising

0.138% of recovered rRNA transcripts. Of the three phyla,

the Parvarchaeota were the most abundant subgroup and

the orders YLA114 and WCHD3–30 have been observed

in other microbial mat systems including the hypersaline

smooth mats of Shark Bay, Western Australia (Wong et al.,

2017) and geothermal spring microbialite-forming mats in

Romania (Coman et al., 2015), but their role within these

systems is currently unknown.

Metabolic activity of thrombolite-forming community

Functional genes grouped by KEGG Orthology (KO) classi-

fications were compared using principle component

analysis (PCA, Euclidian distance) based on normalized

and batch corrected using the package RNA-Seq by Expec-

tation Maximization (RSEM) estimated counts (Fig. 3; Li

and Dewey, 2011). When all functional genes were consid-

ered, PCA plots showed no significant clustering based on

time of day (Fig. 3A) or season (Fig. 3B), suggesting that

the majority of functional genes within the thrombolites are

constitutively expressed. This assessment was confirmed

when the number of unique gene counts grouped by KOs

were compared, as depicted in the Venn diagrams (Fig. 3C

and D), revealing that the majority of unique genes were

expressed at all time points (KO 5 4898) and during all

three seasons (KO 5 6013). There were, however, unique

gene counts associated with each diel time point with the

highest number recovered from the midnight samples

(KO 5 773) followed by noon (KO 5 556). With respect to

the seasons, March (KO 5 1063) had the highest number of

unique gene counts followed by August (KO 5 703), sug-

gesting that some pathways within the thrombolite-forming

communities do differentially respond to changes in their

environment over diel and seasonal cycles.

To begin to elucidate those pathways that might be dif-

ferentially expressed over time, several key metabolisms

were examined that have been typically associated with

the promotion or dissolution of carbonate (Visscher and

Stolz, 2005; Dupraz et al., 2009). These metabolisms

include photosynthesis, nitrogen fixation, denitrification,

methanogenesis and dissimilatory sulfate reduction

(Fig. 4). The data were normalized for within sample

counts (transcripts per million, TPM), across samples

(trimmed mean of M-values, TMM) and batch corrected,

allowing for comparisons between times and seasons.

These normalized data were used for all downstream anal-

yses. The specific genes used for these metabolic

comparisons are listed in Supporting Information Table S4.

�

�

�

�

��

�

�

�

�

�

�

��

�

�

�

�

�

� �

�

�

�

−100 −50 0 50 100

−10

0−

500

5010

0

�

�

�

��

�

�

�

�

amo

�

�

�

�

��

�

�

�

�

�

�

��

�

�

�

�

�

� �

�

�

�

−100 −50 0 50 100

−100

−50

050

100

�

�

�

�

�

�

�

t06t12t18t24

06:0012:0018:0000:00

AugustMarchOctober

BA10

050

0-5

0-1

00

100

500

-50

-100

100500-50-100100500-50-100%91CP%91CP

PC2

6%

PC2

6%

06:00 00:00

18:0012:00C D

August

MarchOctober

703

13236013

1063856476

463

441

413 832

474

773

218

279

450

4898207

428 332

213

556

387

Fig. 3. Relationships between seasonsand diel time-points.

Principle component analysis of

thrombolite metatranscriptomes based on

distribution of KEGG Orthology colored by

(A) time or (B) season. Venn diagrams

displaying unique genes counts grouped

by KEGG Orthology (C) time comparisons

between seasons and (D) comparisons

within each season. [Colour figure can be

viewed at wileyonlinelibrary.com]




Expression of photosynthetic pathways during diel andseasonal cycles

For photosynthesis, the transcripts were differentiated based

on Photosystem I (PSI) and II (PSII) and showed pro-

nounced differences (Fig. 4). Transcripts associated with

PSII genes exhibited a pronounced diel cycle with expres-

sion increasing rapidly after sunrise and peaking at noon,

mirroring many previous studies in both cyanobacterial cul-

tures and non-lithifying microbial mat communities (Fig. 4;

Steunou et al., 2008; Ludwig and Bryant, 2011). For exam-

ple, in cyanobacteria, the expression of psbA gene, which

encodes for the D1 reaction centre protein, has long been

known to respond to the quality and quantity of light (Kulkarni

and Golden, 1994). Genes for PSII (e.g., psbA, psbC, psbD,

psbF and psbE) and photosynthesis antenna proteins (e.g.,

apcD) were some of the most highly expressed transcripts

at noon regardless of the season (Fig. 5). Although PSII

associated gene expression levels drop precipitously at sun-

set and midnight, PSII-associated transcripts were still

present in high numbers (Fig. 4). At night, psbA transcripts

have been shown to be more stable with a half-life of approx-

imately 7 h and play a role in regulating D1 production at the

translational level (Mohamed and Jansson, 1991). Addition-

ally, some diazotrophic coccoid cyanobacteria isoforms of

psbA are transcribed at night to form nonfunctional D1 pro-

teins thereby preventing oxygen evolution during the

nitrogen fixation period (Wegener et al., 2015).

In contrast, transcripts associated with PSI showed no

diel cycling and maintained a relatively high, but constant,

level of expression even at night. Several studies using cul-

tured cyanobacteria have shown that under high light

intensity almost all of the PSI genes are downregulated

(Hihara et al., 2001; Ludwig and Bryant, 2011) to lower the

susceptibility of the cells to damage (Hihara et al., 1998),

giving the appearance of no diel rhythmn. Cyanobacteria

often regulate their PSII/PSI ratios to rapidly adjust to ever

changing cellular redox states and maintain a balance

between CO2 fixation and generation of reductants to

avoid excess production of reactive oxygen and nitrogen

species, which can cause photooxidative damage to the

cells (Ludwig and Bryant, 2011).

Cyanobacteria have also evolved efficient mechanisms

to acquire the inorganic carbon needed for photosynthesis.

Genes associated with carbon concentrating mechanisms

(CCM) were identified in the thrombolite-forming communi-

ties and some exhibited differential diel expression. The

complete family of genes encoding CCM proteins

(ccmKLMNOP) were recovered but exhibited no differential

expression between the treatments, suggesting no pro-

nounced diel or seasonal cycling. Genes associated with

NADPH dehydrogenase (e.g., ndhD, ndhM, ndhN) were

enriched at noon in all seasons (Fig. 5) and, in cultures,

have been shown to be associated with CO2 uptake and

are upregulated under high light and low CO2 conditions

!"

#!!"

$!!"

%!!"

&!!"

'!!!"

'#!!"

")(#'")*%")*#'")(%

%()"%()"!"

#!!!"

%!!!"

'!!!!"

'$!!!"

'&!!!"

'#*)" %*)" '#()"

!"

$!!!!"

%!!!!"

&!!!!"

'!!!!"

#!!!!"

ma21mp6mp21ma6

50,000

40,000

30,000

20,000

10,000

006:00 12:00 18:00 00:00

norm

aliz

ed r

ead

coun

ts photosystem II

06:00 12:00 18:00 00:00

norm

aliz

ed r

ead

coun

ts

photosystem I18,000

14,000

10,000

6,000

2,000

time of day

#!!!"

$!!!"

%!!!"

&!!!"

'!!!!"

'#!!!"

06:00 12:00 18:00 00:00

nitrogen fixation12,000

10,000

8,000

6,000

4,000

2,000

AugustOctober March

!"

#!!"

$!!"

%!!"

&!!"

'!!"

(!!"

)!!"

*!!"

(+," #$-," (-," #$+,"

time of day06:00 12:00 18:00 00:00

800

600

400

200

0

denitrification

!"

#!"

$!"

%!"

&!"

'!!"

'#!"

'$!"

'%!"

'&!"

#!!"

time of day06:00 12:00 18:00 00:00

methanogenesis

1000

800

600

400

2000

time of day

$!!!"

%!!!"

&!!!"

'!!!"

06:00 12:00 18:00 00:00

dissimilatory sulfate reduction4,000

3,000

2,000

1,000

0

time of day

time of day

1200

Fig. 4. Diel and seasonal cycling of major pathway gene expression presenting combined normalized abundances for all the genes in eachmetabolism.

October (green, dotted line); March (red, dotted circles); August (blue, solid line). [Colour figure can be viewed at wileyonlinelibrary.com]




(Hihara et al., 2001). Although not typically present in

marine cyanobacteria, transcripts associated with the high

affinity bicarbonate transport system cmpABCD peaked at

dawn (Badger and Price, 2003), albeit at low expression

levels. A low affinity SulP-type bicarbonate transporter,

bicA, exhibited higher expression levels peaking at 06:00

with the highest expression in March (Fig. 5). This trans-

porter has been identified across a range of marine

cyanobacteria and BicA has been shown to act indepen-

dently in the uptake of bicarbonate, simplifying the process

of carbon concentration within the cyanobacteria carboxy-

some (Price et al., 2004). The increased expression of

bicA at dawn has been observed in several Microcystis

strains as well as non-lithifying mats (Jensen et al., 2011;

Sandrini et al., 2016), and may be light-dependent (Price

et al., 2010). Further examination of these contigs may

reveal novel homologs of the bicA gene within the

thrombolite-forming communities.

In addition to oxygenic photosynthesis, anoxygenic pho-

tosynthesis can locally increase the pH thereby creating a

microenvironment that is favorable to promoting carbonate

precipitation. Evidence for diel cycling of anoxygenic photo-

synthesis was also observed as genes encoding the

photosynthetic reaction centre (puf operon) and bacterio-

chlorophyll biosynthesis (bchC, bchG), which are common

to a diverse group of anoxygenic phototrophs from the Pro-

teobacteria and Chloroflexi, were primarily expressed at

midnight (Fig. 5). Specifically, pufL and pufM, which

encode for the light and medium polypeptide genes of the

reaction centre respectively, as well as the pufC gene,

which encodes the cytochrome c subunit of the complex,

were significantly expressed at midnight regardless of sea-

son (Fig. 5). A similar trend of the enrichment of transcripts

associated with anoxygenic phototrophy at night has been

observed in both cultured Roseobacters (Yurkov and

Beatty, 1998) and in coastal marsh systems (Gifford et al.,

2014). Evidence with cultured strains indicates that the

transcription of the puf operon is light sensitive, but not

oxygen sensitive (Nishimura et al., 1996) suggesting the

differential expression of the puf operon genes at night is a

strategy to avoid photo-induced damage within the

thrombolites.

Nitrogen cycling within thrombolite-forming mats

With regard to nitrogen cycling within the thrombolite-

forming mats, metatranscriptome analysis revealed a pro-

nounced diel cycling of nitrogenase transcripts suggesting

that nitrogen fixation is the primary N source for the throm-

bolite communities. In August and October, the overall

expression of the Mo-nitrogenase structural genes (nifD,

nifH, nifK) peaked at 18:00, whereas in March, these

genes peaked at midnight (Fig. 4). Additionally, several

transcripts encoding nitrogenase FeMo cofactors (nifB)

and stabilizing proteins (nifW) were highly differentially

expressed at midnight regardless of season (Fig. 5). These

differences in nitrogenase reductase expression corre-

spond to non-lithifying cyanobacterial mats from

Mushroom Spring in Yellowstone and may reflect the

increase accumulation of fixed carbon during August and

October where the days are longer, thereby providing

more available energy for the onset of nitrogenase tran-

scription (Steunou et al., 2008). In the thrombolites, the

majority of the nitrogenase transcripts were produced by

coccoid (Order Chroococcales) and filamentous (Order

Nostocales) cyanobacteria regardless of season, although

there were differences in the diel peak of transcripts. The

nitrogenase transcripts derived from the coccoid cyano-

bacteria were enriched at both 18:00 and 00:00, whereas

nif transcripts derived from the filamentous cyanobacteria

peaked primarily at midnight. The H2 generated from the

nitrogen fixation may be recovered by the cyanobacteria

through the presence of a bidirectional hydrogenase (hox)

that is highly expressed at night (Fig. 5).

The metatranscriptomic results indicated that the pri-

mary sink for N in the thrombolite community was

assimilatory nitrate reduction, as transcript levels for deni-

trification (< 700 TPM) and ammonium oxidation (< 300

TPM) were very low at all time points and seasons. It is

possible that the sampling regime did not fully capture diel

transcriptional activity, as transcripts associated with deni-

trification pathways (e.g., nirS, nosZ, norC), although

relatively stable by mRNA standards, typically have a half-

life extending to only 30 min (Hartig and Zumft, 1999).

There was, however, an observable trend of an increase in

transcripts associated with denitrification at 00:00 and

06:00, but it was not statistically significant (Fig. 4; see

Table S4 for complete list). Overall, these results suggest

that denitrification and ammonium oxidation do not repre-

sent primary sinks for N within the thrombolite

communities.

Expression of anaerobic respiration metabolisms

Although photosynthesis and nitrogen fixation processes

dominated the metatranscriptomes, there was evidence of

a diverse array of respiration processes, including metha-

nogenesis and dissimilatory sulfate reduction. For

example, transcripts related to methanogenesis were

recovered from several Archaea and exhibited diel and

seasonal trends (Fig. 2, Supporting Information Table S2).

In August, transcript levels were highest at 00:00, whereas

in October and March the highest number of methanogen-

esis transcripts were recovered at 06:00 (Fig. 4;

Supporting Information Table S4). The majority of

recovered methanogenic Euryarchaeota within the

thrombolite-forming mats appeared to be methylotrophic

(Orders Methanobacterales, Methanosarcinales). For



Fig. 5. Volcano plots visualizing gene expression changes between 00:00 (up on left) and 12:00 (up on right) or 18:00 (up on left) and 06:00(up on right) for each season, October, March and August.

Points represent unique KEGG Orthologies; log2(FC)< 5 and FDR adjusted p-value< 0.01, green; log2(FC)> 5 and FDR adjusted p-

value< 0.01, orange; log2(FC)> 5 and FDR adjusted p-value< 0.001, red; no significance, black; FC, fold change. [Colour figure can be

viewed at wileyonlinelibrary.com]




example, transcripts of methyltransferase genes associ-

ated with methanogenesis from methanol (e.g., mtaC) and

methylamines (e.g., mttB, mttC) were recovered, primarily

from August samples and peaked in expression at 06:00

(data not shown). Methylotrophic methanogenesis is more

varied and may enable these taxa to overcome competition

for H2 by other metabolisms (e.g., sulfate reduction). Addi-

tionally, numerous transcripts associated with aceticlastic

methanogenesis were also recovered (e.g., acsCD, ppa,

ackA) but these transcripts showed no significant differen-

tial expression patterns between time points or seasons

(data not shown). These results suggest that the methano-

gens can use a range of compounds as substrates,

whereas electrons can come from H2 or through methyl

disproportionation, enabling greater metabolic versatility

within the thrombolite-forming community.

Another key metabolism observed within the mat that

plays an important role in carbonate precipitation is sulfate

Fig. 6. Gene expression network using Bray-Curtis dissimilarity (cutoff: 0.45).

Each circle is a KEGG orthologue (KO) and lines connect similar expression patterns (i.e., shorter lines indicate a closer connection). Only

KOs associated with Energy Metabolism pathways are displayed. [Colour figure can be viewed at wileyonlinelibrary.com]




reduction (Dupraz and Visscher, 2005). In the adjacent

stromatolites at Highborne Cay, there is extensive biogeo-

chemical evidence to suggest sulfate reduction increases

localized alkalinity, thereby promoting carbonate precipita-

tion (Visscher et al., 2000; Baumgartner et al., 2009).

Microelectrode profiling in the thrombolites have shown

that the distribution of sulfide is detectable but its abun-

dance changes at different depths (Louyakis et al., 2017).

Numerous taxa typically associated with sulfate reduction

have been recovered from the thrombolites (Myshrall

et al., 2010; Mobberley et al., 2012; Louyakis et al., 2017),

but their transcriptional activity throughout the diel cycle is

not yet known.

One of the key steps in sulfate reduction is the direct

reduction of sulfite to sulfide by dissimilatory sulfite reduc-

tase (dsrAB) (Rees, 1973). However, previous

metagenomic analyses recovered very few dsrAB genes

from the thrombolite-forming populations (Mobberley et al.,

2013; Mobberley et al., 2015). Therefore, the transcription

of genes typically associated with this pathway was exam-

ined here (Fig. 4). Transcripts of dsrAB were recovered

from Desulfovibrio, the archaea Archaeoglobus and the

phototrophic sulfur bacterium Allochromatium but exhibited

relatively low read counts (< 700 TPM) with transcript lev-

els peaking at midnight in August. Transcripts for dsrAB

associated genes, including APS reductase (aprAB) and

sulfate adenylyltransferase (sat), were also recovered at a

higher level from a more diverse consortium of taxa, which

also included Desulfovibrionales and Chroococcales, but

these genes can have alternative functions in addition to

sulfate reduction. Interestingly, numerous transcripts were

also recovered for anaerobic sulfite reductases (asrC) orig-

inally described in Salmonella and Clostridium (Hallenbeck

et al., 1989; Czyzewski and Wang, 2012). Additionally,

thiosulfate (phsA) and tetrathionate reductases (ttrB) were

highly expressed within the thrombolites at night (Figs 5

and 6). Together, these results suggest that much like the

stromatolites, dissimilatory sulfate reduction is an impor-

tant component of the sulfur cycle in the thrombolite-

forming community, but is likely a multi-step process with

intermediate sulfur species (e.g., trithionate pathway;

Kobayashi et al., 1969) rather than the direct reduction of

sulfate to sulfide by dissimilatory sulfite reductase (Rees,

1973). These pathways may be highly dependent on the

nature of electron donors and the environmental

conditions.

The oxidation of inorganic sulfur compounds within the

thrombolite-forming mats is less clear, as transcripts asso-

ciated with pathways, such as the SOX pathway, were in

low abundance at all time points. However, transcripts

associated with sulfide:quinone oxidoreductase (sqr) gene,

which can catalyse the oxidation of sulfide to elemental

sulfur in a wide range of phototrophic and chemotrophic

bacteria (Ghosh and Dam, 2009), were observed in higher

numbers at periods of low oxygen concentration (00:00

and 06:00), but exhibited no seasonal differences (data not

shown). These results may suggest a higher prevalence of

anaerobic oxidation of sulfide at night, which has been

observed in the adjacent stromatolites and may promote

carbonate precipitation (Visscher et al., 2000). More

detailed comparative metagenomics and metatranscrip-

tomics are required to identify the full range of sulfur

cycling within the thrombolite-forming communities.

Network analysis of metabolic pathways

To begin to assess the overarching architecture of the tran-

scriptional network within the thrombolites, all transcripts

(regardless of differential expression) encoding genes

associated with metabolism were clustered by KEGG and

mapped with a weighted approach using Phyloseq (Fig. 6)

(McMurdie and Holmes, 2013). The topography of the net-

work indicated two primary clusters with high levels of

connectivity within the transcriptome. In cluster 1, path-

ways associated with sulfur and nitrogen cycling were

highly intertwined. Specifically, transcripts associated with

ammonia oxidation (e.g., amoAB), dissimilatory nitrate

reduction [e.g., nitrous oxide reductase (e.g., nosZ), nitrate

reductase (e.g., narGH) and nitrite reductases (e.g., nifK)],

as well as several nitrite oxidation (e.g., norB) reactions

clustered extensively with transcripts associated with dis-

similatory sulfur reduction genes. These dissimilatory

sulfur reduction transcripts were associated with thiosulfate

reductase (e.g., phsA), tetrathionate reductase (ttrA),

dimethyl sulfoxide (DMSO) reductase (dsmAB), dimethyl

sulfide dehydrogenase (e.g., ddhAC), dimethylsuloniopro-

prionate (DMSP) lyases (e.g., dddL) and an anaerobic

sulfite reductase (asrA). Although this network does not

reflect the differential expression of these genes within the

thrombolites, these results reveal respiration versatility and

suggest that these sulfur and nitrogen pathways may be

tightly coordinated and regulated. Several previous studies

have shown that some sulfide oxidizing bacteria (e.g., Thi-

obacillus spp.) under anaerobic conditions can use nitrate

as the terminal electron acceptor rather than oxygen

(Visscher and Taylor, 1993). For example, the oxidation of

sulfur and thiosulfate can be linked to nitrate and nitrite

reduction in both cultivars (Aminuddin and Nicholas, 1973)

as well as in oceanic oxygen minimum zones (Azhar et al.,

2014), suggesting that in the thrombolite-forming mats the

sulfur and nitrogen cycle are closely coupled.

Interestingly, transcripts associated with dissimilatory

sulfite reductases (dsrA and dsrB), which directly reduce

sulfite to sulfide, were not associated with cluster 1 and

were located within an uncompressed section of the net-

work, suggesting that the expression of these genes are

distinctly regulated. The regulation of dsrAB is highly vari-

able between organisms and in many organisms is not



dependent on sulfide levels (Grimm et al., 2010). In some,

sulfide-oxidizing taxa expression of dsrAB is increased

under sulfur-oxidizing conditions and under high levels of

reduced sulfur compounds (Chan et al., 2008; Grimm

et al., 2010). However, in several purple and green sulfur

bacteria, as well as some sulfide oxidizers, dsrAB have

been shown to be constitutively expressed (Beller et al.,

2006).

The transcriptional network also revealed a second clus-

ter of transcripts primarily associated with photosynthesis

(PSII and PSI), nitrogen fixation (e.g., nifDK), ammonia

assimilation (e.g., gltBD) and assimilatory sulfur pathways,

such as cysteine biosynthesis (e.g., cysECHIJKM), sug-

gesting these metabolisms exhibit extensive connectivity

within the thrombolite-forming mats. The transcripts asso-

ciated with this cluster are diverse representing a wide

range of taxa associated with these core metabolisms

within the thrombolites. Together, the networking maps

reveal that, with regard to energy metabolism, the

thrombolite-forming mats exhibit tightly coordinated

expression patterns amongst the diverse assemblage of

taxa. Although no taxa-specific clusters are observed

within these metabolic clusters, deeper examination of indi-

vidual gene nodes may help elucidate the potential cross-

talk that occurs between individual taxa and identify how

the community interactions are coordinated. These coordi-

nated activities may serve as important regulators of the

biogeochemical processes that drive the precipitation of

carbonate within the thrombolite-forming communities.

Conclusions

Microbialites represent an important ecosystem to under-

stand the metabolic processes that lead to microbially

induced mineralization. Therefore, examining the ecosys-

tem dynamics of living thrombolites is essential to gain a

more in-depth assessment of these ancient ecosystems.

Our comparative analysis of thrombolite metatranscrip-

tomes provided a new perspective on how microbes

coordinate their metabolic activities across diel and sea-

sonal cycles. First, this study demonstrated that the two

most transcriptionally active taxa within the thrombolites

were the filamentous Rivulariaceae, which have been pre-

viously shown to serve as hot spots for precipitation within

the thrombolites (Planavsky et al., 2009; Louyakis et al.,

2017), and the Xenococcaceae. These results suggest

that these coccoid cyanobacteria may be working syner-

gistically with the Dichothrix spp. to help facilitate the onset

of carbonate precipitation within the thrombolite-forming

mats and this putative relationship will need further

exploration.

Second, our results provide evidence for dynamic tem-

poral patterns of gene expression demonstrating

distinctive diel and seasonal metabolism fluxes that occur

in responses to changing environmental cues. Lastly, the

network mapping of the metabolism transcripts revealed a

high level of connectivity between nitrogen and sulfur

cycling pathways as well as the photosynthesis and assim-

ilatory sulfur pathways, indicating these are among the

most tightly regulated metabolic processes within the

thrombolites. Together, our findings underscore the impor-

tance of elucidating the relative contribution of the different

taxa and metabolisms within thrombolite-forming ecosys-

tems as it will lead to a better understanding of the

underlying interactions and controls required for microbially

induced mineralization.

Experimental procedures

Sample collection

Samples of thrombolite-forming mats were collected in Octo-ber 2013, March 2014 and August 2014 from the island ofHighborne Cay, Exumas, The Bahamas (768490 W, 248430N).

For each season samples were collected for two consecutivesunny days at four time-points: sunrise (06:00), midday(12:00), sunset (18:00) and midnight (00:00). Three replicatebutton formations (Fig. 1B) were collected for each time-pointusing a sterile scalpel and placed immediately in RNAlater

(Life Technologies, Grand Island, NY) before being trans-ported to Space Life Sciences Lab, Merritt Island, FL wherethey were stored at 2808C until processing. Metadata forannual temperature and day lengths were attained fromarchives from Weather Underground (https://goo.gl/qkXGc7).

RNA isolation, purification and cDNA synthesis

Upon thawing, samples were gently pressed and centrifuged

multiple times for removal of RNAlater. A modified PowerBio-film RNA Isolation Kit (Qiagen, Carlsbad, CA, USA) was usedto extract total RNA from �100 mg of sample per reaction.Modifications included adding BFR1, ß-mercaptoethanol

(2.8%) and BFR2 solution directly to the bead tube prior toadding the mat. Additionally, to ensure adequate lysing of cya-nobacterial filaments without compromising the overall qualityof the RNA, the bead tubes underwent homogenization in aBioSpec Mini-Beadbeater-8 in three, 1 min increments with 1

min rests on ice in-between. The bead tubes were then heatedto 658C for 5 min in a water bath and vortexed for 10 min.BFR3 was increased to 200 ll and added to the supernatant.The manufacturer’s protocol was followed for the remainingsteps. An additional DNAse treatment, the Turbo DNA-free Kit

(Ambion, Applied Biosystem Business, CA), was used toremove residual DNA and the samples were purified usingRNA Clean and Concentrator-5 (Zymo Research, Orange,CA, USA).

Nucleic acid concentrations were measured with a Qubit

2.0 Fluorometer (Invitrogen, Life Technologies, Carlsbad, CA,USA). Quality of the RNA was visualized on an Agilent 2100Bioanalyser with an RNA 6000 Nano Chip (Agilent Technolo-gies, Santa Clara, CA, USA). To ensure the RNA extracts didnot contain any enzymatic inhibitors, samples were amplified

using real-time PCR. Universal primers, targeting the 16SrRNA gene (Khodadad et al., 2011), with an iTaq Universal



https://goo.gl/qkXGc7

SYBR Green One-Step kit (BioRad, Hercules, CA, USA) were

used on a CFX96 Touch platform (BioRad, Hercules, CA,

USA) as previously described (Casaburi et al., 2017). Overall,

approximately 3–5 extractions for each of the collected throm-

bolite button formations (n 5 3) were pooled to acquire the

minimum 1 lg of RNA needed for each library preparation.

Generation and sequencing of RNA libraries

Recovered RNA first underwent rRNA depletion with Ribo-

Zero Gold rRNA Removal Kit (Epidemiology, Illumina, San

Diego, CA, USA) followed by purification with RNA Clean and

Concentrator-5. Synthesis of cDNA libraries was accom-

plished using ScriptSeq v2 RNA-Seq Library Preparation Kit

(Illumina, San Diego USA) following manufacturers protocol.

Libraries were purified with Axygen AxyPrep Mag kit (Axygen

Biosciences, Union City, CA, USA) and quantified before

sequencing. For all time-points (n 5 4) and seasons (n 5 3),

duplicate libraries were generated for a total of 24 libraries

that were sequenced at the University of Florida’s Interdisci-

plinary Centre for Biotechnology Research using NextSeq

500/550 High Output v2 Kit (paired end, 150 cycles, insert

size 550 bp) on a NextSeq 500 sequencing system (Illumina,

San Diego, CA, USA). Sequences of thrombolite cDNA were

deposited in the NCBI Sequence Read Archive (SRR6006825

though SRR6006848). BioSample accession numbers are

consecutive from SAMN07554631 through SAMN07554654

under BioProject PRJNA305634.

Sequence quality control, assembly, annotation

and mapping

A comprehensive flowchart of the bioinformatic analysis is

described in Supporting Information Fig. S1. Briefly, sequen-

ces were trimmed using Trimmomatic v0.35 with a threshold

of Q5 (Bolger et al., 2014) and rRNA was removed using Sort-

MeRNA, as previously described (Kopylova et al., 2012).

FastQC was used to analyse read quality (Andrews, 2014).

Taxonomic analysis was completed on the quality-filtered 16S

rRNA gene sequences isolated from each sample using

Quantitative Insights Into Microbial Ecology (QIIME v1.9.1)

(Caporaso et al., 2010b). Operational taxonomic units (OTUs)

were assigned to the reads at 97% identity against the Green-

genes database v13.8 (DeSantis et al., 2006) using the

UCLUST method (open reference OTU picking) within QIIME.

Representative reads were aligned with PyNAST (v1.2.2)

(Caporaso et al., 2010a) to the Greengenes Core reference

alignment and a phylogenetic tree was built with FastTree

(v2.1.3) (Price et al., 2013). The resulting biom data was visu-

alized using GraPhlAn v0.9.7 (Asnicar et al., 2015) after

removal of unassigned records. The networking analysis

included only transcripts with> 10 hits to a specific gene and

was completed using the R package Phyloseq v1.19.1

(McMurdie and Holmes, 2013).

The remaining sequences were assembled using Trinity

v2.2.0 with the following parameters: fastq assembly (left read

file contained forward and unpaired reads), minimum contig

length of 75 bp and normalized reads (Garber et al., 2011).

Detonate v1.11 was used to score assembly tools under

altered parameters to facilitate selection for downstream

analysis (Li et al., 2014). Alignment was completed using

Bowtie2 v2.2.9 (Langmead and Salzberg, 2012) and RSEM

v1.2.7 (Li and Dewey, 2011) estimation was used for counts of

sample replicates, as well as pooled samples. RSEM esti-

mates were rounded to nearest integer, length corrected

(transcripts per million method), trimmed mean of M-values

(TMM) adjusted for normalized expression values (EdgeR

v3.16.5) (Robinson et al., 2010; McCarthy et al., 2012), and

batch corrected with ARSyNseq. NOISeq v2.18.0 (Tarazona

et al., 2011; Tarazona et al., 2015) in Bioconductor v3.4 was

used for PCA plotting (Euclidian distance) as well as differen-

tial expression analysis with the following noiseqbio

parameters: norm 5 ‘tmm’, k 5 0.5, r 5 50, filter 5 1,

nclust 5 15, a0per 5 0.9 and random.seed 5 12 200. Venn

diagrams were generated using the R package, VennDiagram

v1.6.17 (Chen and Boutros, 2011). The metatranscriptome

assembly was annotated using Trinotate v3.0.1 using com-

plete pipeline (http://trinotate.github.io). KEGG terms were

associated with their respective pathways using the R pack-

age KEGGREST (Tenenbaum, 2017).

Acknowledgements

The authors would like to thank Kimberly Cranmore for her

technical assistance as well as Drs. Brelan Moritz and Keelnat-

ham Shanmugam for their advice on RNA extraction. ASL was

supported by the NSF Graduate Research Fellowship Pro-

gram. HG was supported by the Developing a European Ameri-

can NGS Network program. Collection permits were provided

through the Ministry of Agriculture, Marine Resources and

Local Government of The Bahamas. The research efforts were

supported by the NASA Exobiology and Evolutionary Biology

program element (NNX12AD64G) awarded to JSF.

Author contributions

ASL and JSF designed the experiments and conducted

the fieldwork; ASL, RMEB and JSF performed the experi-

ments; ASL, HG, GC, AAD and JSF analysed the data;

ASL and JSF wrote the manuscript.

Conflict of Interest

The authors declare no conflict of interest.

References

Aitken, J.D. (1967) Classification and environmental signifi-

cance of cryptalgal limestones and dolomites, with illustra-

tions from the Cambrian and Ordovician of southwestern

Alberta. J Sediment Petrol 37: 1163–1178.Alcantara-Hern�andez, R.J., Valdespino-Castillo, P.M.,

Centeno, C.M., Alcocer, J., Merino-Ibarra, M., and Falc�on,

L.I. (2017) Genetic diversity associated with N-cycle path-

ways in microbialites from Lake Alchichica, Mexico. Aquat

Microb Ecol 78: 121–133.Allen, M.A., Goh, F., Burns, B.P., and Neilan, B.A. (2009)

Bacterial, archaeal and eukaryotic diversity of smooth and

pustular microbial mat communities in the hypersaline

lagoon of Shark Bay. Geobiology 7: 82–96.



http://trinotate.github.io

Altermann, W., Kazmierczak, J., Oren, A., and Wright, D.T.

(2006) Cyanobacterial calcification and its rock-building

potential during 3.5 billion years of Earth history. Geobiol-

ogy 4: 147–166.

Aminuddin, M., and Nicholas, D.J. (1973) Sulphide oxidation

linked to the reduction of nitrate and nitrite in Thiobacillus

denitrificans. Biochim Biophys Acta 325: 81–93.

Andres, M.S., and Reid, R.P. (2006) Growth morphologies of

modern marine stromatolites: a case study from Highborne

Cay, Bahamas. Sediment Geol 185: 319–328.Andrews, S. (2014) FastQC a quality-control tool for high-

throughput sequence data [WWW document]. URL http://

www. In Bioinformaticsbabraham. ac. uk/projects/fastqc.Asnicar, F., Weingart, G., Tickle, T.L., Huttenhower, C., and

Segata, N. (2015) Compact graphical representation of phylo-

genetic data and metadata with GraPhlAn. PeerJ 3: e1029.Azhar, M.A., Canfield, D.E., Fennel, K., Thamdrup, B., and

Bjerrum, C.J. (2014) A model-based insight into the cou-

pling of nitrogen and sulfur cycles in a coastal upwelling

system. J Geophys Res Biogeosci 119: 264–285.Badger, M.R., and Price, G.D. (2003) CO2 concentrating

mechanisms in cyanobacteria: molecular components, their

diversity and evolution. J Exp Bot 54: 609–622.Baker, B.J., Lazar, C.S., Teske, A.P., and Dick, G.J. (2015)

Genomic resolution of linkages in carbon, nitrogen, and sul-

fur cycling among widespread estuary sediment bacteria.

Microbiome 3: 14.

Baumgartner, L.K., Reid, R.P., Dupraz, C., Decho, A.W.,

Buckley, D.H., Spear, J.R., et al. (2006) Sulfate reducing

bacteria in microbial mats: changing paradigms, new dis-

coveries. Sediment Geol 185: 131–145.Baumgartner, L.K., Spear, J.R., Buckley, D.H., Pace, N.R.,

Reid, R.P., Dupraz, C., and Visscher, P.T. (2009) Microbial

diversity in modern marine stromatolites, Highborne Cay,

Bahamas. Environ Microbiol 11: 2710–2719.Beck, C., Hertel, S., Rediger, A., Lehmann, R., Wiegard, A.,

Kolsch, A., et al. (2014) Daily expression pattern of protein-

encoding genes and small noncoding RNAs in Synechocystis

sp. strain PCC 6803. Appl Environ Microbiol 80: 5195–5206.,.Beller, H.R., Chain, P.S., Letain, T.E., Chakicherla, A.,

Larimer, F.W., Richardson, P.M., et al. (2006) The genome

sequence of the obligately chemolithoautotrophic, faculta-

tively anaerobic bacterium Thiobacillus denitrificans.

J Bacteriol 188: 1473–1488.Berelson, W.M., Corsetti, F.A., Pepe-Ranney, C., Hammond,

D.E., Beaumont, W., and Spear, J.R. (2011) Hot spring sili-

ceous stromatolites from Yellowstone National Park:

assessing growth rate and laminae formation. Geobiology

9: 411–424.Bernhard, J.M., Edgcomb, V.P., Visscher, P.T., McIntyre-

Wressnig, A., Summons, R.E., Bouxsein, M.L., et al. (2013)

Insights into foraminiferal influences on microfabrics of

microbialites at Highborne Cay, Bahamas. Proc Natl Acad

Sci USA 110: 9830–9834.Blazewicz, S.J., Barnard, R.L., Daly, R.A., and Firestone,

M.K. (2013) Evaluating rRNA as an indicator of microbial

activity in environmental communities: limitations and uses.

ISME J 7: 2061–2068.Bolger, A.M., Lohse, M., and Usadel, B. (2014) Trimmomatic:

a flexible trimmer for Illumina sequence data. Bioinformatics

30: 2114–2120.

Bowlin, E.M., Klaus, J., Foster, J.S., Andres, M., Custals, L.,

and Reid, R.P. (2012) Environmental controls on microbial

community cycling in modern marine stromatolites. Sedi-

ment Geol 263–264: 45–55.Braissant, O., Decho, A.W., Dupraz, C., Glunk, C., Przekop,

K.M., and Visscher, P.T. (2007) Exopolymeric substances of

sulfate-reducing bacteria: interactions with calcium at alka-

line pH and implication for formation of carbonate minerals.

Geobiology 5: 401–411.Breitbart, M., Hoare, A., Nitti, A., Siefert, J., Haynes, M.,

Dinsdale, E., et al. (2009) Metagenomic and stable isotopic

analyses of modern freshwater microbialites in Cuatro

Ci�enegas, Mexico. Environ Microbiol 11: 16–34.Burne, R.V., and Moore, L.S. (1987) Microbialites: organosedi-

mentary deposits of benthic microbial communities. Palaios

2: 241–254.Burns, B.P., Goh, F., Allen, M., and Neilan, B.A. (2004) Micro-

bial diversity of extant stromatolites in the hypersaline

marine environment of Shark Bay, Australia. Environ Micro-

biol 6: 1096–1101.

Caporaso, J.G., Bittinger, K., Bushman, F.D., DeSantis, T.Z.,

Andersen, G.L., and Knight, R. (2010a) PyNAST: a flexible

tool for aligning sequences to a template alignment. Bioin-

formatics 26: 266–267.Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K.,

Bushman, F.D., Costello, E.K., et al. (2010b) QIIME allows

analysis of high-throughput community sequencing data.

Nat Methods 7: 335–336.Casaburi, G., Duscher, A.A., Reid, R.P., and Foster, J.S.

(2016) Characterization of the stromatolite microbiome from

Little Darby Island, The Bahamas using predictive and

whole shotgun metagenomic analysis. Environ Microbiol

18: 1452–1469.Casaburi, G., Goncharenko-Foster, I., Duscher, A.A., and

Foster, J.S. (2017) Transcriptomic changes in an animal-

bacterial symbiosis under modeled microgravity conditions.

Sci Rep 7: 46318.

Cerqueda-Garcia, D., and Falcon, L.I. (2016) Metabolic poten-

tial of microbial mats and microbialites: autotrophic capabili-

ties described by an in silico stoichiometric approach from

shared genomic resources. J Bioinform Comput Biol 14:

1650020.Chagas, A.A.P., Webb, G.E., Burne, R.A., and Southam, G.

(2016) Modern lacustrine mcirobialites; towards a synthesis

of aquous and carbonate geochemistry and minerology.

Earth Sci Rev 162: 338–363.

Chan, L.K., Weber, T.S., Morgan-Kiss, R.M., and Hanson, T.E.

(2008) A genomic region required for phototrophic thiosulfate

oxidation in the green sulfur bacterium Chlorobium tepidum

(syn. Chlorobaculum tepidum). Microbiology 154: 818–829.Chen, H., and Boutros, P.C. (2011) VennDiagram: a package

for the generation of highly-customizable Venn and Euler

diagrams in R. BMC Bioinformatics 12: 35.Coman, C., Chiriac, C.M., Robeson, M.S., Ionescu, C.,

Dragos, N., Barbu-Tudoran, L., et al. (2015) Structure, min-

eralogy, and microbial diversity of geothermal spring micro-

bialites associated with a deep oil drilling in Romania. Front

Microbiol 6: 253.Conesa, A., Madrigal, P., Tarazona, S., Gomez-Cabrero, D.,

Cervera, A., McPherson, A., et al. (2016) A survey of best

practices for RNA-seq data analysis. Genome Biol 17: 13.



http://www

http://www

Czyzewski, B.K., and Wang, D.N. (2012) Identification and

characterization of a bacterial hydrosulphide ion channel.

Nature 483: 494–497.Den Uyl, P.A., Richardson, L.L., Jain, S., Dick, G.J., and

Kellogg, C.A. (2016) Unraveling the physiological roles of

the cyanobacterium Geitlerinema sp. BBD and other Black

Band Disease community members through genomic anal-

ysis of a mixed culture. PLoS One 11: e0157953.DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie,

E.L., Keller, K., et al. (2006) Greengenes, a chimera-

checked 16S rRNA gene database and workbench compat-

ible with ARB. Appl Environ Microbiol 72: 5069–5072.Desnues, C.G., Rodriguez-Brito, B., Rayhawk, S., Kelley, S.,

Tran, T., Haynes, M., et al. (2008) Biodiversity and biogeog-

raphy of phages in modern stromatolites and thrombolites.

Nature 452: 340–345.Dupraz, C., and Visscher, P.T. (2005) Microbial lithification in

marine stromatolites and hypersaline mats. Trends Micro-

biol 13: 429–438.

Dupraz, C., Visscher, P.T., Baumgartner, L.K., and Reid, R.P.

(2004) Microbe-mineral interactions: early carbonate precip-

itation in a hypersaline lake (Eleuthera Island, Bahamas).

Sedimentology 51: 1: 21.Dupraz, C., Reid, R.P., Braissant, O., Decho, A.W., Norman,

R.S., and Visscher, P.T. (2009) Processes of carbonate

precipitation in modern microbial mats. Earth Sci Rev 96:

141–162.Edgcomb, V.P., Bernhard, J.M., Beaudoin, D., Pruss, S.,

Welander, P.V., Schubotz, F., et al. (2013) Molecular indica-

tors of microbial diversity in oolitic sands of Highborne Cay,

Bahamas. Geobiology 11: 234–251.Fernandez-Gomez, B., Richter, M., Schuler, M., Pinhassi, J.,

Acinas, S.G., Gonzalez, J.M., and Pedros-Alio, C. (2013)

Ecology of marine Bacteroidetes: a comparative genomics

approach. ISME J 7: 1026–1037.

Foster, J.S., and Green, S.J. (2011) Microbial diversity in mod-

ern stromatolites In Cellular Origin, Life in Extreme Habitats

and Astrobiology: Interactions with Sediments. Seckbach,

J., and Tewari, V. (eds). Springer, pp. 385–405.Foster, J.S., Green, S.J., Ahrendt, S.R., Hetherington, K.L.,

Golubic, S., Reid, R.P., and Bebout, L. (2009) Molecular

and morphological characterization of cyanobacterial diver-

sity in the marine stromatolites of Highborne Cay, Bahamas.

ISME J 3: 573–587.Garber, M., Grabherr, M.G., Guttman, M., and Trapnell, C.

(2011) Computational methods for transcriptome annotation

and quantification using RNA-seq. Nat Methods 8:

469–477.

Ghosh, W., and Dam, B. (2009) Biochemistry and molecular

biology of lithotrophic sulfur oxidation by taxonomically and

ecologically diverse bacteria and archaea. FEMS Microbiol

Rev 33: 999–1043.Gifford, S.M., Sharma, S., and Moran, M.A. (2014) Linking

activity and function to ecosystem dynamics in a coastal

bacterioplankton community. Front Microbiol 5: 185.Gleeson, D.B., Wacey, D., Waite, I., O’donnell, A.G., and

Kilburn, M.R. (2015) Biodiversity of living, non-marine,

thrombolites of Lake Clifton, Western Australia. Geomicro-

biol J 33: 850–859.Goh, F., Allen, M.A., Leuko, S., Kawaguchi, T., Decho, A.W.,

Burns, B.P., and Neilan, B.A. (2009) Determining the

specific microbial populations and their spatial distribution

within the stromatolite ecosystem of Shark Bay. ISME J 3:

383–396.Goh, F., Jeon, Y.J., Barrow, K., Neilan, B.A., and Burns, B.P.

(2011) Osmoadaptive strategies of the archaeon Halococ-

cus hamelinensis isolated from a hypersaline stromatolite

environment. Astrobiology 11: 529–536.

Grimm, F., Dobler, N., and Dahl, C. (2010) Regulation of dsr

genes encoding proteins responsible for the oxidation of

stored sulfur in Allochromatium vinosum. Microbiology 156:

764–773.Grotzinger, J.P., and Knoll, A.H. (1999) Stromatolites in Pre-

cambrian carbonates: evolutionary mileposts or environ-

mental dipsticks?. Annu Rev Earth Planet Sci 27: 313–358.Hallenbeck, P.C., Clark, M.A., and Barrett, E.L. (1989) Char-

acterization of anaerobic sulfite reduction by Salmonella

typhimurium and purification of the anaerobically induced

sulfite reductase. J Bacteriol 171: 3008–3015.Hartig, E., and Zumft, W.G. (1999) Kinetics of nirS expression

(cytochrome cd1 nitrite reductase) in Pseudomonas

stutzeri during the transition from aerobic respiration to

denitrification: evidence for a denitrification-specific nitrate-

and nitrite-responsive regulatory system. J Bacteriol 181:

161–166.Havemann, S.A., and Foster, J.S. (2008) Comparative charac-

terization of the microbial diversities of an artificial microbia-

lite model and a natural stromatolite. Appl Environ Microbiol

74: 7410–7421.Hihara, Y., Sonoike, K., and Ikeuchi, M. (1998) A novel gene,

pmgA, specifically regulates photosystem stoichiometry in

the cyanobacterium Synechocystis species PCC 6803 in

response to high light. Plant Physiol 117: 1205–1216.Hihara, Y., Kamei, A., Kanehisa, M., Kaplan, A., and Ikeuchi,

M. (2001) DNA microarray analysis of cyanobacterial gene

expression during acclimation to high light. Plant Cell 13:

793–806.Jensen, S.I., Steunou, A.S., Bhaya, D., Kuhl, M., and

Grossman, A.R. (2011) In situ dynamics of O2, pH and cya-

nobacterial transcripts associated with CCM, photosynthe-

sis and detoxification of ROS. ISME J 5: 317–328.Kazmierczak, J., and Altermann, W. (2002) Neoarchean bio-

mineralization by benthic cyanobacteria. Science 298:

2351.Kazmierczak, J., Altermann, W., Kremer, B., Kempe, S., and

Eriksson, P.G. (2009) Mass occurrence of benthic coccoid

cyanobacteria and their role in the production of

Neoarchean carbonates of South Africa. Precambrian Res

173: 79–92.

Ka�zmierczak, J., Kempe, S., Kremer, B., L�opez-Garcia, P.,

Moreira, D., and Tavera, R. (2011) Hydrochemistry and

microbialites of the alkaline crater lake Alchichica, Mexico.

Facies 57: 543–570.Kennard, J.M., and James, N.P. (1986) Thrombolites and stro-

matolites: two distinct types of microbial structures. Palaios

1: 492–503.Khodadad, C.L., and Foster, J.S. (2012) Metagenomic and

metabolic profiling of nonlithifying and lithifying stromatolitic

mats of Highborne Cay, The Bahamas. PLos One 7:

e38229.Khodadad, C.L., Zimmerman, A.R., Green, S.J., Uthandi, S.,

and Foster, J.S. (2011) Taxa-specific changes in soild



microbial composition induced by pyrogenic carbon amend-

ments. Soil Biol Biochem 43: 385–392.Kobayashi, K., Tachibana, S., and Ishimoto, M. (1969)

Intermediary formation of trithionate in sulfite

reduction in a sulfate-reducing bacteirum. J Biochem

(Tokyo) 65: 155.Kopylova, E., Noe, L., and Touzet, H. (2012) SortMeRNA: fast

and accurate filtering of ribosomal RNAs in metatranscrip-

tomic data. Bioinformatics 28: 3211–3217.Kulkarni, R.D., and Golden, S.S. (1994) Adaptation to high

light intensity in Synechococcus sp. strain PCC 7942: regu-

lation of three psbA genes and two forms of the D1 protein.

J Bacteriol 176: 959–965.Langmead, B., and Salzberg, S.L. (2012) Fast gapped-read

alignment with Bowtie 2. Nat Methods 9: 357–359.

Lepp, P.W., and Schmidt, T.M. (1998) Nucleic acid content of

Synechococcus spp. during growth in continuous light and

light/dark cycles. Arch Microbiol 170: 201–207.

Li, B., and Dewey, C.N. (2011) RSEM: accurate transcript

quantification from RNA-Seq data with or without a refer-

ence genome. BMC Bioinformatics 12: 323.Li, B., Fillmore, N., Bai, Y., Collins, M., Thomson, J.A.,

Stewart, R., and Dewey, C.N. (2014) Evaluation of de novo

transcriptome assemblies from RNA-Seq data. Genome

Biol 15: 553.Louyakis, A.S., Mobberley, J.M., Vitek, B.E., Visscher, P.T.,

Hagan, P.D., Reid, R.P., et al. (2017) A study of the micro-

bial spatial heterogeneity of Bahamian thrombolites using

molecular, biochemical, and stable isotope analyses. Astro-

biology 17: 413–430.Ludwig, M., and Bryant, D.A. (2011) Transcription profiling of

the model cyanobacterium Synechococcus sp. strain PCC

7002 by Next-Gen (SOLiD) sequencing of cDNA. Front

Microbiol 2: 41.Mata, S.A., Harwood, C.L., Corsetti, F.A., Stork, N.J., Eilers,

K., Berelson, W., and Spear, J.R. (2012) Influence of gas

production and filament orientation on stromatolite microfa-

bric. Palaios 27: 206–219.McCarthy, D.J., Chen, Y., and Smyth, G.K. (2012) Differential

expression analysis of multifactor RNA-Seq experiments

with respect to biological variation. Nucleic Acids Res 40:

4288–4297.McIlroy, S.J., and Nielson, P.H. (2014) The family saprospira-

ceae. In The Prokaryotes. Berlin, Heidelberg: Springer, pp.

863–889.McMurdie, P.J., and Holmes, S. (2013) phyloseq: an R pack-

age for reproducible interactive analysis and graphics of

microbiome census data. PLos One 8: e61217.Mobberley, J.M., Ortega, M.C., and Foster, J.S. (2012) Com-

parative microbial diversity analyses of modern marine

thrombolitic mats by barcoded pyrosequencing. Environ

Microbiol 14: 82–100.

Mobberley, J.M., Khodadad, C.L., and Foster, J.S. (2013) Met-

abolic potential of lithifying cyanobacteria-dominated throm-

bolitic mats. Photosynth Res 118: 125–140.

Mobberley, J.M., Khodadad, C.L., Visscher, P.T., Reid, R.P.,

Hagan, P., and Foster, J.S. (2015) Inner workings of throm-

bolites: spatial gradients of metabolic activity as revealed by

metatranscriptome profiling. Sci Rep 5: 12601.Mohamed, A., and Jansson, C. (1991) Photosynthetic elec-

tron transport controls degradation but not production of

psbA transcripts in the cyanobacterium Synechocystis

6803. Plant Mol Biol 16: 891–897.Myshrall, K., Mobberley, J.M., Green, S.J., Visscher, P.T.,

Havemann, S.A., Reid, R.P., and Foster, J.S. (2010) Bio-

geochemical cycling and microbial diversity in the modern

marine thrombolites of Highborne Cay, Bahamas. Geobiol-

ogy 8: 337–354.Nishimura, K., Shimada, H., Ohta, H., Masuda, T., Shioi, Y.,

and Takamiya, K. (1996) Expression of the puf operon in an

aerobic photosynthetic bacterium, Roseobacter denitrifi-

cans. Plant Cell Physiol 37: 153–159.Ordonez, O.F., Flores, M.R., Dib, J.R., Paz, A., and Farias,

M.E. (2009) Extremophile culture collection from Andean

lakes: extreme pristine environments that host a wide diver-

sity of microorganisms with tolerance to UV radiation.

Microb Ecol 58: 461–473.Papineau, D., Walker, J.J., Mojzsis, S.J., and Pace, N.R.

(2005) Composition and structure of microbial communities

from stromatolites of Hamelin Pool in Shark Bay, Western

Australia. Appl Environ Microbiol 71: 4822–4832.

Paul, D., Kumbhare, S.V., Mhatre, S.S., Chowdhury, S.P.,

Shetty, S.A., Marathe, N.P., et al. (2015) Exploration of

microbial diversity and community structure of Lonar Lake:

the only hypersaline meteorite crater lake within basalt rock.

Front Microbiol 6: 1553.Pepe-Ranney, C., Berelson, W.M., Corsetti, F.A., Treants, M.,

and Spear, J.R. (2012) Cyanobacterial construction of hot

spring siliceous stromatolites in Yellowstone National Park.

Environ Microbiol 14: 1182–1197.

Petrash, D.A., Gingras, M.K., Lalonde, S.V., Orange, F.,

Pecoits, E., and Konhauser, K.O. (2012) Dynamic controls

on accretion and lithification of modern gypsum-dominated

thrombolites, Los Roques, Venezuela. Sediment Geol

245–246: 29–47.Planavsky, N., Reid, R.P., Andres, M., Visscher, P.T., Myshrall,

K.L., and Lyons, T.W. (2009) Formation and diagenesis of

modern marine calcified cyanobacteria. Geobiology 7: 566–

576.Price, G.D., Woodger, F.J., Badger, M.R., Howitt, S.M., and

Tucker, L. (2004) Identification of a SulP-type bicarbonate

transporter in marine cyanobacteria. Proc Natl Acad Sci

USA 101: 18228–18233.Price, M.N., Dehal, P.S., Arkin, A.P., and Poon, A.F.Y. (2010)

FastTree 2–approximately maximum-likelihood trees for

large alignments. PLos One 5: e9490.Price, G.D., Pengelly, J.J., Forster, B., Du, J., Whitney, S.M.,

von Caemmerer, S., et al. (2013) The cyanobacterial CCM

as a source of genes for improving photosynthetic CO2 fixa-

tion in crop species. J Exp Bot 64: 753–768.Reddy, G.S., Prakash, J.S., Vairamani, M., Prabhakar, S.,

Matsumoto, G.I., and Shivaji, S. (2002) Planococcus ant-

arcticus and Planococcus psychrophilus spp. nov. isolated

from cyanobacterial mat samples collected from ponds in

Antarctica. Extremophiles 6: 253–261.

Rees, C.E. (1973) A steady-state model for sulphur isotope

fractionation in bacterial reduction processes. Geochim Et

Cosmochim Acta 37: 1141–1162.Reid, R.P., Visscher, P.T., Decho, A.W., Stolz, J.F., Bebout,

B.M., Dupraz, C., et al. (2000) The role of microbes in

accretion, lamination and early lithification of modern

marine stromatolites. Nature 406: 989–992.



Rinke, C., Schwientek, P., Sczyrba, A., Ivanova, N.N.,

Anderson, I.J., Cheng, J.F., et al. (2013) Insights into the

phylogeny and coding potential of microbial dark matter.

Nature 499: 431–437.

Robinson, M.D., McCarthy, D.J., and Smyth, G.K. (2010)

edgeR: a Bioconductor package for differential expression

analysis of digital gene expression data. Bioinformatics 26:

139–140.Russell, J.A., Brady, A.L., Cardman, Z., Slater, G.F., Lim,

D.S., and Biddle, J.F. (2014) Prokaryote populations of

extant microbialites along a depth gradient in Pavilion Lake,

British Columbia, Canada. Geobiology 12: 250–264.Ruvindy, R., White, R.A., III, Neilan, B.A., and Burns, B.P.

(2016) Unravelling core microbial metabolisms in the hyper-

saline microbial mats of Shark Bay using high-throughput

metagenomics. ISME J 10: 183–196.Sagha€ı, A., Zivanovic, Y., Zeyen, N., Moreira, D., Benzerara,

K., Deschamps, P., et al. (2015) Metagenome-based diver-

sity analyses suggest a significant contribution of non-

cyanobacterial lineages to carbonate precipitation in mod-

ern microbialites. Front Microbiol 6: 797.Sandrini, G., Tann, R.P., Schuurmans, J.M., van Beusekom,

S.A., Matthijs, H.C., and Huisman, J. (2016) Diel variation in

gene expression of the CO2-concentrating mechanism dur-

ing a harmful cyanobacterial bloom. Front Microbiol 7: 551.Schneider, D., Arp, G., Reimer, A., Reitner, J., Daniel, R., and

Balcazar, J.L. (2013) Phylogenetic analysis of a

microbialite-forming microbial mat from a hypersaline lake

of the Kiritimati atoll, Central Pacific. PLoS One 8: e66662.See-Too, W.-S., Ee, R., Madhaiyan, M., Kwon, S.-W., Tan,

J.Y., Lim, Y.L., et al. (2017) Planococcus versutus sp. nov.,

isolated from soil. Int J Syst Evol Microbiol 67: 94–950.Simon, M., Scheuner, C., Meier-Kolthoff, J.P., Brinkhoff, T.,

Wagner-Dobler, I., Ulbrich, M., et al. (2017) Phylogenomics

of Rhodobacteraceae reveals evolutionary adaptation to

marine and non-marine habitats. ISME J 11: 1483–1499.Sorensen, K.B., Canfield, D.E., Teske, A.P., and Oren, A.

(2005) Community composition of a hypersaline endoeva-

poritic microbial mat. Appl Environ Microbiol 71: 7352–

7365.Steunou, A.S., Jensen, S.I., Brecht, E., Becraft, E.D.,

Bateson, M.M., Kilian, O., et al. (2008) Regulation of nif

gene expression and the energetics of N2 fixation over the

diel cycle in a hot spring microbial mat. ISME J 2: 364–378.Stolz, J.F., Feinstein, T.N., Salsi, J., Visscher, P.T., and Reid,

R.P. (2001) TEM analysis of microbial mediated sedimenta-

tion and lithification in modern marine stromatolites. Amer

Mineral 86: 826–833.

Suosaari, E.P., Reid, R.P., Playford, P.E., Foster, J.S., Stolz,

J.F., Casaburi, G., et al. (2016) New multi-scale perspec-

tives on the stromatolites of Shark Bay, Western Australia.

Sci Rep 6: 20557.Tang, M., Chen, C., Li, J., Xiang, W., Wu, H., Wu, J., et al.

(2016) Fabivirga thermotolerans gen. nov., sp. nov., a novel

marine bacterium isolated from culture broth of a marine

cyanobacterium. Int J Syst Evol Microbiol 66: 1095–1099.Tarazona, S., Garcia-Alcalde, F., Dopazo, J., Ferrer, A., and

Conesa, A. (2011) Differential expression in RNA-seq: a

matter of depth. Genome Res 21: 2213–2223.Tarazona, S., Furio-Tari, P., Turra, D., Pietro, A.D., Nueda,

M.J., Ferrer, A., and Conesa, A. (2015) Data quality aware

analysis of differential expression in RNA-seq with NOISeq

R/Bioc package. Nucleic Acids Res 43: e140.Tenenbaum, D. (2017) KEGGREST: Client-side REST access

to KEGG. In R package version 1161.Valdespino-Castillo, P.M., Alcantara-Hernandez, R.J., Alcocer,

J., Merino-Ibarra, M., Macek, M., and Falcon, L.I. (2014)

Alkaline phosphatases in microbialites and bacterioplankton

from Alchichica soda lake, Mexico. FEMS Microbiol Ecol

90: 504–519.Visscher, P.T., and Stolz, J.F. (2005) Microbial mats as bio-

reactors: populations, processes and products. Palaeo-

geogr Palaeoclimatol Palaeoecol 219: 87–100.Visscher, P.T., and Taylor, B.F. (1993) A new mechanism for

the aerobic catabolism of dimethyl sulfide. Appl Environ

Microbiol 59: 3784–3789.Visscher, P.T., Reid, R.P., Bebout, B.M., Hoeft, S.E.,

Macintyre, I.G., and Thompson, J.A. (1998) Formation of

lithified micritic laminae in modern marine stromatolites

(Bahamas): the role of sulfur cycling. Amer Mineral 83:

1482–1493.Visscher, P.T., Reid, R.P., and Bebout, B.M. (2000) Microscale

observations of sulfate reduction: correlation of microbial

activity with lithified micritic laminae in modern marine stro-

matolites. Geology 28: 919–922.Walter, M.R. (1994) Stromatolites: the main geological source

of information on the evolution of the early benthos. In Early

Life on Earth Nobel Symposium. Bengston, S. (ed). New

York, NY: Columbia University Press, pp. 270–283.

Warden, J.G., Casaburi, G., Omelon, C.R., Bennett, P.C.,

Breecker, D.O., and Foster, J.S. (2016) Characterization of

microbial mat microbiomes in the modern thrombolite eco-

system of Lake Clifton, Western Australia using shotgun

metagenomics. Front Microbiol 7: 1064.

Wegener, K.M., Nagarajan, A., and Pakrasi, H.B. (2015) An

atypical psbA gene encodes a sentinel D1 protein to form a

physiologically relevant inactive photosystem II complex in

cyanobacteria. J Biol Chem 290: 3764–3774.White, R.A., III, Chan, A.M., Gavelis, G.S., Leander, B.S.,

Brady, A.L., Slater, G.F., et al. (2016) Metagenomic analysis

suggests modern freshwater microbialites harbor a distinct

core microbial community. Front Microbiol 6: 1531.Wong, H.L., Visscher, P.T., White Iii, R.A., Smith, D.L.,

Patterson, M.M., and Burns, B.P. (2017) Dynamics of

archaea at fine spatial scales in Shark Bay mat micro-

biomes. Sci Rep 7: 46160.Yurkov, V.V., and Beatty, J.T. (1998) Aerobic anoxygenic pho-

totrophic bacteria. Microbiol Mol Biol Rev 62: 695–724.Zinser, E.R., Lindell, D., Johnson, Z.I., Futschik, M.E.,

Steglich, C., Coleman, M.L., et al. (2009) Choreography of

the transcriptome, photophysiology, and cell cycle of a mini-

mal photoautotroph, prochlorococcus. PLoS One 4: e5135.

Supporting information

Additional Supporting Information may be found in the

online version of this article at the publisher’s web-site:

Fig. S1. Methodology flowchart outlining the experimental

parameters and bioinformatics tools used in the analysis of

the metatranscriptome data.

Fig. S2. Taxonomic histograms for each sample to family

level or higher if unable to classify to family. Legend



identifies thirty most abundant taxa over all libraries and

lists in order of bar plot from top to bottom. Note, two colors

of Chroococcales reflect both ‘unclassified’ and ‘other’

designations.Table S1. Summary of Trinity assembly statisticsTable S2. Correlation coefficients for sample pairs; repli-

cates outlined in bold, Pearson correlation on bottom (blue

scale) and Spearman correlation on top (green scale).

Table S3. Relative abundance of active taxa within throm-

bolites over diel and seasonal cycles.Table S4. Genes used to generate metabolism profiles inFigure 4.Table S5. Expressed genes in thrombolites with normalizedcounts over diel and seasonal cycles and number of ortho-

logs detected in Trinity assembly.



a year in the life of a thrombolite: comparative … · 2019. 8. 17. · a year in the life of a...

Documents