UC San Diego
UC San Diego Previously Published Works
Title
Biogeography of the marine actinomycete Salinispora.
Permalink
https://escholarship.org/uc/item/3rc5k97b
Journal
Environmental microbiology, 8(11)
ISSN
1462-2912
Authors
Jensen, Paul R
Mafnas, Chrisy
Publication Date
2006-11-01
DOI
10.1111/j.1462-2920.2006.01093.x
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
Environmental Microbiology (2006)
8
(11), 1881–1888 doi:10.1111/j.1462-2920.2006.01093.x
© 2006 The Authors
Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd
Blackwell Publishing LtdOxford, UKEMIEnvironmental Microbiology1462-2912© 2006 The Authors; Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd
2006
8
1118811888
Original Article
Salin-
ispora biogeographyP. R. Jensen and C. Mafnas
Received 21 November, 2005; accepted 4 March, 2006. *For corre-
spondence. E-mail pjensen@ucsd.edu; Tel. (
+
1) 858 534 7322; Fax
(
+
1) 858 558 3702.
Biogeography of the marine actinomycete
Salinispora
Paul R. Jensen* and Chrisy Mafnas
Center for Marine Biotechnology and Biomedicine,
Scripps Institution of Oceanography, University of
California, San Diego, CA, USA.
Summary
Marine actinomycetes belonging to the genus
Salin-
ispora
were cultured from marine sediments collected
at six geographically distinct locations. Detailed phy-
logenetic analyses of both 16S rRNA and
gyrB
gene
sequences reveal that this genus is comprised of
three distinct but closely related clades correspond-
ing to the species
Salinispora tropica
,
Salinispora
arenicola
and a third species for which the name
‘
Salinispora pacifica
’ is proposed.
Salinispora areni-
cola
was cultured from all locations sampled and pro-
vides clear evidence for the cosmopolitan distribution
of an individual bacterial species. The co-occurrence
of
S. arenicola
with
S. tropica
and
S. pacifica
sug-
gests that ecological differentiation as opposed to
geographical isolation is driving speciation within the
genus. All
Salinispora
strains cultured to date share
greater than 99% 16S rRNA gene sequence identity
and thus comprise what has been described as a
microdiverse ribotype cluster. The description of this
cluster as a new genus, containing multiple species,
provides clear evidence that fine-scale 16S rDNA
sequence analysis can be used to delineate among
closely related species and that more conservative
operational taxonomic unit values may significantly
underestimate global species diversity.
Introduction
Bacterial biogeography remains an unresolved issue in
microbiology (Fenchel, 2003). Because of their small size,
high abundance and ease of dispersal, the prevailing
hypothesis in the field is that free-living bacteria are
not subject to geographical isolation and, without this
constraint, should exhibit a cosmopolitan distribution
(reviewed by Staley and Gosink, 1999; Martiny
et al
.,
2006). While it is widely accepted that bacterial genera
are widely distributed in their respective habitats (Hedlund
and Staley, 2004), there is currently little empirical support
for the ‘everything is everywhere but the environment
selects’ paradigm (De Wit and Bouvier, 2006) when
applied at the species level. Without a better understand-
ing of the extent to which geographical isolation affects
the population structure of individual bacterial species, it
will not be possible to effectively estimate global species
richness or to understand the forces driving speciation
among bacteria.
Little emphasis has been given to the study of microbial
biogeography (Cho and Tiedje, 2000), and as a result it is
not clear if similar populations occupy analogous environ-
ments on a global scale. The most outspoken support for
microbial cosmopolitanism comes from studies of microeu-
karyotes (Findlay, 2002); however, this support is based
largely on the analysis of protozoan morphospecies. As
might be expected, evidence for cosmopolitanism among
environmental prokaryotes includes taxa with robust sur-
vival strategies, such as the spore-forming genus
Bacillus
,
for which it has been shown that migration rates are suf-
ficiently high to prevent geographical isolation (Roberts
and Cohan, 1995). Additional evidence comes from a study
of fluorescent
Pseudomonas
strains where cosmopolitan-
ism was evident by the analysis of 16S rDNA and, to a
lesser extent, 16S-23S intergenic spacer regions (Cho and
Tiedje, 2000). Evidence for endemism was documented
at the infraspecific level among the same
Pseudomonas
strains when higher resolution genomic fingerprinting
methods were applied. Additional evidence for endemism
is found among prokaryotes inhabiting extreme environ-
ments where the barriers to surviving dispersal are high.
This includes gas vacuolated sea ice bacteria (Staley and
Gosink, 1999), the thermophilic archeon
Sulfolobus
(Whi-
taker
et al
., 2003) and the thermophilic cyanobacterium
Synecococcus
(Papke
et al
., 2003).
Any discussion of species-level bacterial biogeography
is affected by uncertainty surrounding the species concept
for bacteria (Cohan, 2002; Gevers
et al
., 2005). Recently,
it has been proposed that molecular sequence data can
be used to define natural units of bacterial diversity that
possess the fundamental properties of species (Cohan,
2002). These units can be recognized as clusters of
sequences that share greater similarity to each other than
to related sequences and are believed to delineate eco-
logically distinct populations or ecotypes (Cohan, 2002).
Ecotypes may arise through various processes (Gevers
et al
., 2005) including geographical isolation or natural
selection and can be difficult to resolve using highly con-
1882
P. R. Jensen and C. Mafnas
© 2006 The Authors
Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd,
Environmental Microbiology
,
8
, 1881–1888
served loci such as the 16S rRNA gene (Fox
et al
., 1992;
Palys
et al
., 1997; Staley and Gosink, 1999). This has led
to an increased reliance on protein coding genes and,
more recently, multilocus sequence analysis for the reso-
lution of intrageneric relationships (Gevers
et al
., 2005).
In several cases, it has been demonstrated that named
species are comprised of multiple ecotypes (Palys
et al
.,
2000), leading to the suggestion that the bacterial species
generally recognized today are in fact composites of mul-
tiple ecotypes each possessing the dynamic properties of
individual species (Cohan, 2002).
We recently reported the discovery of the actinomycete
genus
Salinispora
, which is widely distributed in tropical
and subtropical marine sediments (Mincer
et al
., 2002;
Maldonado
et al
., 2005). To date, two species have been
formally described (
S. arenicola
and
S. tropica
), and a
third (‘
S. pacifica
’) is proposed based on DNA–DNA
hybridization.
Salinispora
belongs to the Micromonospo-
raceae and is the first actinomycete genus known to
require seawater for growth. As these bacteria produce
resistant spores and have been cultured from worldwide
locations, they represent model organisms to test hypoth-
eses about bacterial biogeography and the processes that
drive speciation. In this paper, the phylogenetic relation-
ships of 152 strains were assessed using 16S rRNA and
gyrB
gene sequences. The results provide compelling evi-
dence that an individual bacterial species comprised
largely of 16S rDNA clones can have a cosmopolitan
distribution and that speciation within the genus
Salin-
ispora
is not due to geographical isolation.
Results
Salinispora
strains were cultivated from all six tropical/
subtropical locations sampled. These locations included
multiple collection sites within the Bahamas, where they
were originally discovered (Jensen
et al
., 1991), the US
Virgin Islands, the Red Sea, the Sea of Cortez, Palau and
Guam. In addition, strains were recently reported from the
sponge
Pseudoceratum clavata
collected from the Great
Barrier Reef (Kim
et al
., 2005) and the ascidian
Polysyn-
craton lithostrotum
collected from Fiji (He
et al
., 2001),
providing the first Southern hemisphere sites from which
Salinispora
strains have been recovered. Despite exten-
sive effort, we have yet to cultivate
Salinispora
strains from
temperate Pacific Ocean sediments collected off La Jolla,
California. They also do not appear to be among the
numerous
Micromonospora
strains recovered from North
Sea sediments (E. Helmke, pers. comm.). Detailed 16S
and
gyrB
phylogenetic analyses of 46
Salinispora
strains
clearly reveal that the genus, as we know it today, is
comprised of three distinct but closely related species
(Figs 1 and 2). Two of these,
S. arenicola
and
S. tropica
,
were recently described (Maldonado
et al
., 2005), while
‘
S. pacifica
’ is proposed based on
<
60% interspecies
DNA–DNA hybridization (performed by the DSMZ, Ger-
man Collection of Microorganisms and Cell Cultures,
Braunschweig).
Biogeographical distribution
The three
Salinispora
species vary in their biogeographi-
cal distributions (Figs 1 and 2).
Salinispora arenicola
has
Fig. 1.
Neighbour-joining phylogenetic tree created from 46 nearly
complete (1449 nucleotides) 16S rRNA gene sequences from
Salin-
ispora
strains cultured from worldwide locations. The three major
Salinispora
phylotypes, consisting of the two formally described
species
S. tropica
and
S. arenicola
and the proposed species
‘
S. pacifica
’, are clearly delineated. Type strains representing the five
Micromonospora
species most closely related to
Salinispora
, along
with
Micromonospora halophytica
, are included. Species names are
followed by strain number, strain source (BA, Bahamas; RS, Red Sea;
GU, Guam; PA, Palau; USVI, US Virgin Islands; SC, Sea of Cortez),
year of collection (89
=
1989, etc.) and accession number (for repre-
sentative sequences).
Propionibacterium propionicus
and
Blastococ-
cus aggregatus
were used as outgroups.
Actinoplanes auranticolor (ABO47491)
Catellatospora ferruginea (AF152108)
Micromonospora olivasterospora (X92613)
M. eburnea (AB107231)
M. halophytica (X92601)
M. aurantiaca (X92604)
M. rosaria (X92631)
M. endolithica (AJ560635)
S. tropica (CNB-440) BA 89 (AY040617)
S. tropica (CNB-476) BA 89
S. tropica (CNB-392) BA 89
S. tropica (CNB-536) BA 89 (AY040618)
S. tropica (CNH-898) BA 00 (AY040622)
S. tropica (CNR-699) BA 03
S. arenicola (CNS-205) PA 04
S. arenicola (CNS-051) PA 04
S. arenicola (CNB-458) BA 89
S. arenicola (CNB-527) BA 89
S. arenicola (CNR-107) GU 02 (AY464534)
S. arenicola (CNH-643) BA 99 (AY040619)
S. arenicola (CNH-646) BA 99 (AY040620)
S. arenicola (CNH-877) BA 00
S. arenicola (CNH-905) BA 00
S. arenicola (CNQ-817) GU 02
S. arenicola (CNR-921) PA 04
S. arenicola (CNQ-748) GU 02
S. arenicola (CNR-425) GU 02 (AY464533)
S. arenicola (CNQ-884) GU 02
S. arenicola (CNQ-976) GU 02
S. arenicola (CNR-005) GU 02
S. arenicola (CNR-416) GU 02
S. arenicola (CNR-075) GU 02
S. arenicola (CNH-719 RS 00
S. arenicola (CNH-718) RS 00
S. arenicola (CNH-725) RS 00 (AY040621)
S. arenicola (CNH-721) RS 00
S. arenicola (CNH-713) RS 00
S. arenicola (CNP-161) USVI 01
S. arenicola (CNP-173) USVI 01
S. arenicola (CNP-188) USVI 01
S. arenicola "A" (CNH-962 )SC 00
S. arenicola "A" (CNH-963) SC 00 (DQ224162)
S. arenicola "A" (CNH-941) SC 01 (DQ224163)
S. arenicola "A" (CNH-996) SC 01
S. arenicola "B" (CNP-152) SC 01 (DQ224164)
S. arenicola "B" (CNP-193) SC 01
S. arenicola "B" (CNP-105) SC 00
S. arenicola "B" (CNH-964) SC 00 (AY040623)
"S. pacifica" (CNH-732) RS 00 (DQ224165)
"S. pacifica" (CNR-114) GU 02 (DQ224161)
"S. pacifica" (CNQ-768) GU 02
"S. pacifica" (CNS-103) PA 04 (DQ224160)
"S. pacifica" (CNS-143) PA 04 (DQ092624)
"S. pacifica" A (CNS-055) PA 04 (DQ224159)
Blastococcus aggregatus (AJ430193)
Propionibacterium propionicus
0.001 substitutions/site
100
100
96
100
56
99
71
Salinispora
biogeography
1883
© 2006 The Authors
Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd,
Environmental Microbiology
,
8
, 1881–1888
a cosmopolitan distribution having been recovered from
all six of the locations sampled. It is also consistently the
most abundant species observed, representing 86% of
the 152 strains examined in this study.
Salinispora tropica
has the most restricted distribution having thus far only
been detected from the Bahamas, where it has been
consistently recovered over a 15 year period and repre-
sents seven of the 19 strains examined. Surprisingly, this
species was not recovered from the US Virgin Islands,
despite the examination of 20 strains from this site and its
proximity to the Bahamas. ‘
Salinispora pacifica
’ has been
recovered from Guam, Palau and the Red Sea, with only
one strain being recovered from the latter. This species is
also considerably less common than
S. arenicola
(three
of 59 strains from Guam, seven of 23 strains from Palau,
one of 18 strains from the Red Sea). Although more
widely distributed than
S. tropica
, ‘
S. pacifica
’ was absent
or remained below the detection limit in the Caribbean and
the Sea of Cortez and represents a second species that,
at present, appears to be geographically restricted relative
to
S. arenicola
. The
Salinispora
strains recently reported
from the Great Barrier Reef (Kim
et al
., 2005) fall within
both
S. arenicola
and ‘
S. pacifica
’, while the single strain
reported from Fiji (He
et al
., 2001) is identical (based on
16S rRNA gene sequence) to ‘
S. pacifica
’.
SSU rRNA gene diversity
There is a remarkable lack of intraclade diversity within
the three
Salinispora
species. Despite the inclusion of
strains isolated over a 15 year period from multiple col-
lections sites throughout the Bahamas, all
S. tropica
strains cultured to date share 100% sequence identity
throughout the 1479 base pairs examined (Table 1). This
absence of sequence variation could only be detected
once careful corrections were made for polymerase chain
reaction (PCR) and sequencing errors, including correc-
tions to previously reported data (Mincer
et al
., 2002;
Jensen
et al
., 2005). The 34
S. arenicola
strains exam-
Fig. 2.
Neighbour-joining phylogenetic tree created from 46 nearly
complete (1164 nucleotides)
gyrB
gene sequences from
Salinispora
strains cultured from worldwide locations. Labelling is similar to
Fig. 1.
Pseudoalteromonas haloplanktis
was used as an outgroup.
Pseudoalteromonas haloplanktis (AF007283)
Streptomyces mashuensis (AB072864)
Actinoplanes brasiliensis (AB014125)
Couchioplanes caeruleus (AB014138)
Actinoplanes cyaneus (AB014126)
Micromonospora aurantiaca (AB015621)
Micromonospora rosaria (IFO13697)
Micromonospora halophytica (AB014157)
Micromonospora carbonacea (AB014147)
Micromonospora olivasterospora (AB014159)
Micromonospora echinospora (AB014153)
Micromonospora purpurea (AB014160)
S. tropica (CNB-536) BA 89 (DQ228692)
S. tropica (CNH-898) BA 00 (DQ228683)
S. tropica (CNR-699) BA 03
S. tropica (CNB-476) BA 89
S. tropica (CNB-392) BA 89
S. tropica (CNB-440) BA 89 (DQ228684)
"S. pacifica" (CNS-143) PA 04 (DQ228693)
"S. pacifica" (CNH-732) RS 00 (DQ228687)
"S. pacifica" (CNS-103) PA 04 (DQ228685)
"S. pacifica" (CNQ-768) GU 02
"S. pacifica" (CNR-114) GU 02 (DQ228686)
"S. pacifica" A (CNS-055) PA 04 (DQ228691)
S. arenicola (CNH-877) BA 00
S. arenicola (CNP-161) USVI 01
S. arenicola (CNP-173) USVI 01
S. arenicola (CNH-905) BA 00
S. arenicola (CNH-646) BA 99 (DQ228680)
S. arenicola "A" (CNH-962 )SC 00
S. arenicola (CNQ-817) GU 02
S. arenicola (CNS-205) PA 04
S. arenicola "B" (CNH-964) SC 00 (DQ263620)
S. arenicola "B" (CNP-105) SC 00
S. arenicola "B" (CNP-193) SC 01
S. arenicola "B" (CNP-152) SC 01 (DQ228688)
S. arenicola (CNR-107) GU 02 (DQ228682)
S. arenicola (CNH-643) BA 99 (DQ228681)
S. arenicola (CNR-921) PA 04
S. arenicola (CNQ-748) GU 02
S. arenicola (CNQ-976) GU 02
S. arenicola (CNQ-884) GU 02
S. arenicola (CNR-425) GU 02 (DQ228679)
S. arenicola (CNH-721) RS 00
S. arenicola (CNH-713) RS 00
S. arenicola (CNH-718) RS 00
S. arenicola (CNH-725) RS 00 (DQ228678)
S. arenicola (CNH-719 RS 00
S. arenicola (CNR-005) GU 02
S. arenicola (CNR-075) GU 02
S. arenicola (CNR-416) GU 02
S. arenicola (CNS-051) PA 04
S. arenicola (CNP-188) USVI 01
S. arenicola "A" (CNH-963) SC 00 (DQ228690)
S. arenicola "A" (CNH-941) SC 01 (DQ228689)
S. arenicola "A" (CNH-996) SC 01
S. arenicola (CNB-458) BA 89
S. arenicola (CNB-527) BA 89
Dactylosporangium matsuzakiense (AB014104)
0.01 substitutions/site
100
100
100
100
100
100
66
100
Table 1.
Salinispora
intra- and interspecific genetic similarity (number of strains in parentheses after species identifier).
Species
% Similarity
gyrB
(DNA)
16S (rDNA)
gyrB
(DNA)
gyrB
(aa)
d
N
d
S
d
N
/
d
S
St (6) 100 (1479) 99.57 (1159) 99.23 (385) 3 2 1.50
Sa (34) 99.86 (1477) 96.13 (1119) 98.45 (382) 6 39 0.15
Sp (6) 99.86 (1477) 97.16 (1131) 99.23 (385) 5 30 0.17
St : Sp 99.59 (1473) 95.10 (1107) 98.71 (383) 5 52 0.10
Sa : Sp 99.26 (1468) 92.87 (1081) 96.65 (375) 13 70 0.19
St : Sa 99.53 (1472) 92.87 (1081) 96.39 (374) 14 69 0.20
Sa,
S. arenicola
; Sp, ‘
S. pacifica
’; St,
S. tropica
. 16S similarities generated from 1479 nucleotide positions,
gyrB
DNA similarities generated from
1164 nucleotide positions,
gyrB
amino acid (aa) similarities generated from 388 positions. Number of invariant positions in parentheses after per
cent similarities.
d
N
, non-synonymous nucleotide substitution;
d
S
, synonymous nucleotide substitution. Interspecific comparisons were made using
the type strains for each species.
1884
P. R. Jensen and C. Mafnas
© 2006 The Authors
Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd,
Environmental Microbiology
,
8
, 1881–1888
ined in detail possessed nearly identical sequences
(99.86% similarity) with the only variations arising from
strains cultured from the Sea of Cortez, all of which con-
tained one of two possible single nucleotide polymor-
phisms (Fig. 1) resulting in the subclades S. arenicola ‘A’
(12 strains observed) and S. arenicola ‘B’ (five strains
observed). None of the Sea of Cortez strains were a
perfect sequence match with the S. arenicola type strain
(CNH-643) providing extremely fine scale (one nucle-
otide) biogeographical resolution of these two Sea of
Cortez populations. Despite analysing partial sequence
data for an additional 96 S. arenicola strains, including
multiple representatives from all locations, no new
intraclade sequence diversity was detected. As with
S. arenicola, ‘S. pacifica’ intraclade similarity was 99.86%
(two variable nucleotide positions out of 1479 examined).
Both of these nucleotide variations occurred in strain
CNS-055 relative to the proposed type strain (CNS-143)
and delineate the ‘S. pacifica’ ‘A’ clade.
Interclade diversity among the three Salinispora spe-
cies was also low and places the entire genus into what
has been described as a microdiverse sequence cluster
(Acinas et al., 2004). Pairwise similarities (
BLAST bl2seq,
NCBI) reveal that S. tropica and S. arenicola share
99.53% 16S rRNA gene sequence identity (Table 1). This
is a difference of seven nucleotides of 1479 examined.
Salinispora tropica was found to differ from S. arenicola
subclades ‘A’ and ‘B’ by one additional nucleotide (99.46%
similarity). The greatest sequence differences occurred
between S. arenicola and ‘S. pacifica’ (11 nucleotides,
99.26%) and between S. arenicola subclade ‘A’ or ‘B’ and
‘S. pacifica’ (12 nucleotides, 99.19% similarity). The most
similar species were S. tropica and ‘S. pacifica’, which
differed by only six nucleotides (99.59% similarity).
Despite the high level of sequence identity, S. tropica and
S. arenicola have been classified as distinct species (Mal-
donado et al., 2005), while the classification of ‘S. pacifica’
as a third species is supported by DNA–DNA hybridization
experiments (Wayne et al., 1987) in which the proposed
type strain shared < 60% genomic similarity to S. tropica
and S. arenicola (data provided by the DSMZ). Salin-
ispora species share 96.50–96.60% similarity with
Micromonospora chalcea, the type strain for the genus
Micromonospora, and 97.28–97.56% similarity with
Micromonospora rosaria, the most closely related
Micromonospora species.
The majority of the Salinispora sequence diversity thus
far detected occurs in appropriately variable regions of the
SSU rRNA gene and, with the exception of CNS-055, in
multiple strains, providing strong evidence that these
changes are not due to PCR or sequencing errors. Nine
of 15 variable nucleotide positions occur in the V2 variable
region (Rijik et al., 1992) with all but one of these occur-
ring in non-conserved helixes. Of the remaining substitu-
tions, only a G⇔A hairpin loop transition (Escherichia coli
position 262) occurs in a conserved region (90–98%
among all bacteria).
gyrB phylogeny
The phylogenetic tree based on nearly complete gyrB
DNA sequences (1164 nucleotides) re-affirms the mono-
phyletic nature of the Salinispora clade and its separation
from other genera within the Micromonosporaceae
(Fig. 2). The three Salinispora phylotypes, corresponding
to S. tropica, S. arenicola and ‘S. pacifica’ are clearly
delineated providing additional phylogenetic support for
the separation of these taxa. No cryptic species were
detected from the analysis of this protein-coding gene,
and no variations were detected among any of the 46
strains in terms of species-specific 16S rDNA and gyrB
cladding patterns, although there is a difference in the
branching patterns in the two trees with the gyrB pattern
being better supported by bootstrap analysis.
As with the 16S rRNA gene sequence data, there was
a remarkably high level of gyrB sequence similarity within
the three Salinispora species (Table 1), with S. tropica
strains sharing 99.57% sequence identity, S. arenicola
strains sharing 96.13% sequence identity and ‘S. pacifica’
strains sharing 97.16% sequence identity. The interspe-
cies similarity was greatest between S. tropica and
‘S. pacifica’ (95.10%) and least between S. arenicola and
the other two phylotypes (92.87%). The closest Blastn
gyrB sequence match for all three of the Salinispora
phylotypes was M. rosaria (BAA89737) for which the
sequence identity was 89–90%. The Salinispora gyrB
sequence data were translated into 388 amino acids and,
as expected, both the intra- and interspecies amino acid
similarities are high (98.45–99.23% and 96.39–98.71%
respectively). Salinispora species share 90–92% amino
acid sequence identity with M. rosaria, the closest
BLASTp
(NCBI) match. The d
N
/d
S
ratio for S. tropica was approxi-
mately 10-fold greater than for the other two species;
however, this may be due to small sample size.
Effects of temperature on growth
In previous studies, we have observed that Salinispora
strains are capable of growth at 10°C but not at 4°C. To
test the effects of exposure to 4°C on Salinispora growth
and viability, seven strains were maintained at 4°C for 2,
4, 6, or 8 weeks then incubated at 25°C for 2 months. All
strains incubated at 4°C for 2 or 4 weeks showed no
reduction in growth relative to controls upon transfer to
25°C. After 6 weeks at 4°C however, all strains exhibited
reduced growth at 25°C with two strains (CNS-103 and
CNR-114) remaining reduced even after 2 months at this
temperature. After 8 weeks at 4°C, these same two strains
