RESEARCH ARTICLE

An in situ method for cultivating microorganisms using a double

encapsulation technique
Eitan Ben-Dov1,2, Esti Kramarsky-Winter3,4 & Ariel Kushmaro1,5
1
Department of Biotechnology Engineering, Ben-Gurion University of the Negev, Be’er-Sheva, Israel; 2Achva Academic College, M.P. Shikmim, Israel;
3
Department of Zoology, Tel-Aviv University, Tel-Aviv, Israel; 4Pacific Biomedical Research Center, University of Hawaii at Manoa, Honolulu, HI, USA; and
5
National Institute for Biotechnology, Ben-Gurion University of the Negev, Be’er-Sheva, Israel

Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023

Correspondence: Ariel Kushmaro,
Department of Biotechnology Engineering,
National Institute for Biotechnology, Ben-
Gurion University of the Negev, PO Box 653,
Beer-Sheva 84105, Israel. Tel.: 1972 8 647
9024; fax: 1972 8 647 2983; e-mail:
arielkus@bgu.ac.il
Received 25 January 2009; revised 8 March
2009; accepted 18 March 2009.
First published online 21 April 2009.
DOI:10.1111/j.1574-6941.2009.00682.x
Editor: Riks Laanbroek
Keywords
unculturable microorganisms; microbial
diversity; cultivating technique.
Abstract
The lack of cultured microorganisms represents a bottleneck for advancement in
microbiology. The development of novel culturing techniques is, therefore, a
crucial step in our understanding of microbial diversity in general, and the role of
such diversity in the environment, in particular. This study presents an innovative
method for cultivating microorganisms by encapsulating them within agar
spheres, which are then encased in a polysulfonic polymeric membrane and
incubated in a simulated or natural environment. This method stimulates growth
of the entrapped microorganisms by allowing them access to essential nutrients
and cues from the environment. It allows for the discovery of microorganisms
from dilutions that are 10–100-fold greater than possible with conventional
plating techniques. Analysis of microorganisms grown in such spheres incubated
in and on a number of different substrates yielded numerous novel ribotypes. For
example, spheres incubated on the mucus surface of a Fungiid coral yielded
numerous ribotypes, with only 50% sharing similarity (85–96%) to previously
identified microorganisms. This suggests that many of the species represent novel
ribotypes. Hence, the technique reported here advances our ability to retrieve and
successfully culture microorganisms and provides an innovative tool to access
unknown microbial diversity.

portant biological functions of the microorganisms. For

Introduction
The field of microbiology has been limited by the restricted
ability to culture microorganisms in the laboratory. Indeed,
it is estimated that 4 99% of microorganisms have not yet
been cultured (Staley & Konopka, 1985; Wagner et al., 1993;
Amann et al., 1995). The failure of microbial culturing is due
to a number of reasons. Medium toxicity, auxotrophy for
missing nutrients, strict dependence on mutualistic interac-
tions with other bacteria and suppression of slow-growing
bacteria by overgrowing bacteria can all arrest microorgan-
ism growth (Manome et al., 2001; Giovannoni et al., 2007).
Currently, only 26 of the c. 52 identifiable major lineages, or
phyla, within the domain Bacteria have cultivated represen-
tatives. The other remaining candidate divisions are repre-
sented only by environmental DNA sequences (Rappé &
Giovannoni, 2003).
There are, however, several limitations to using molecular
analysis alone. rRNA gene phylogenies cannot reveal im-
example, such approaches do not address unique innova-
tions in cell function that may have driven the evolution of a
certain microbial group. Moreover, metagenomic techni-
ques, either alone or together with expression vectors, are
unable to provide enough information regarding important
biochemical activities of such organisms in the environment
(Rappé & Giovannoni, 2003; Cottrell et al., 2005). The
development of novel culturing techniques (Manome et al.,
2001; Connon & Giovannoni, 2002; Kaeberlein et al., 2002;
Zengler et al., 2002; Akselband et al., 2006; Bollmann et al.,
2007; Giovannoni et al., 2007; Ingham et al., 2007) is,
therefore, crucial to expanding our understanding of micro-
bial diversity and of the roles played by microorganisms in
the environment.
In this report, we describe a method for cultivating
microorganisms by encapsulating them in polysulfone-
coated agar spheres, and subsequently incubating the
spheres in simulated or natural environments. Using this

FEMS Microbiol Ecol 68 (2009) 363–371 

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
364
approach, we succeeded in isolating and cultivating a
number of previously unknown microorganisms.
Materials and methods
Sample collection
Mucus from Fungia granulosa corals (maintained in aquaria
with artificial seawater) was sampled by collecting c. 1-mL
mucus in a sterile disposable 50-mL polypropylene centri-
fuge tube. In addition, samples of mucus from healthy
individual F. granulosa corals were collected from the Red
Sea (Gulf of Eilat) at a depth of 10–15 m, in front of the
Inter-University Institute for Marine Science, Eilat, Israel
(29151 0 N, 34194 0 E). Bacteriological peel loops were used to
collect the coral surface microlayer. The loops were inserted
into sterile 50-mL polypropylene centrifuge tubes, sealed
underwater, brought to the surface and immediately placed
on ice. Soil samples were collected from the Halutzah area of
the Negev desert, Israel, using sterile plastic bags. The
samples were transported on ice to the laboratory at
Ben-Gurion University. Wastewater samples were obtained
from a laboratory-scale wastewater bioreactor (filled with
waste water from the Ramat-Hovav industrial park in the
Negev desert, Israel).
Culturing technique
To dilute the samples so as to allow insertion of approxi-
mately one bacterium per sphere, environmental samples
(from the laboratory scale wastewater bioreactor, soil or
mucus from healthy F. granulosa) were estimated for total
bacterial numbers by 4 0 ,6 0 -diamidino-2-phenylindole (DAPI)
staining (final concentration, 2 mg mL1) and counting by
epifluorescence microscopy or directly under a phase micro-
scope. For the coral mucus samples, the number of cultur-
able CFU were determined by plating 100-mL samples of
undiluted mucus or mucus diluted in filtered artificial
seawater on duplicate Marine Agar 2216 (HiMedia Labora-
tories, Mumbai, India) plates (50% of recommended final
concentration) and incubated in a dark/light (12 h) regime
at 22 1C. Soil samples (1 g) were vigorously mixed with
0.5 mL sterile tap water to obtain an initial suspension. The
soil and wastewater samples were then diluted by sterile tap
water. CFU counts of undiluted and diluted samples of soil
suspension and wastewater were obtained by plating 100-mL
samples on duplicate Luria–Bertani (LB) (HiMedia Labora-
tories) and nutrient broth agar (Laboratorios Conda,
Madrid, Spain) plates, incubated at 30 1C.
Polysulfone preparation and its permeability
Polysulfone in the form of granules [Sigma-Aldrich,
St. Louis, MO; Cat No. 42830-2; MW 35 000 (LS)] was

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
E. Ben-Dov et al.
solublized in N,N-dimethylformamide (DMF) (Merck KGaA,
Darmstadt, Germany) solvent to obtain 10%, 15%, 20% and
25% (w/v) solutions. Permeability of the polysulfone coat-
ing layer for soluble nutrients was evaluated by mixing
7% (w/v) Blue Dextran-2000 (Pharmacia, Uppsala, Sweden)
within agar spheres (1%). Each sphere was incubated in
1 mL of sterile tap water at room temperature. Polysulfone
membrane-coated and -uncoated spheres without blue
dextran served as controls. At 1-h intervals, release of color
into the surroundings was measured using a microplate
reader (Model ELx808; Bio-Tek Instruments, Winooski, VT)
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
at 630 nm. Release of Blue Dextran was determined over a
24-h window.
Sphere production
Evaluation of the encapsulation method was carried out
with pure cultures of Escherichia coli. The bacteria were
diluted up to 104 cells mL1 and mixed at a ratio of 1 : 1 with
warm (48 1C) 2% bacteriological agar to form agar spheres
that were coated by polysulfone (10% and 20% in DMF).
Coated spheres were then immersed in ethanol (99%) and
flamed for few seconds (to prevent contamination of the
incubating medium) and incubated in LB broth for 24 h in a
water shaker (200 r.p.m.) at 30 1C.
Samples from mucus, wastewater or soil were 10-fold
diluted up to 109 of initial concentrations and mixed with
warm (48 1C) 2% marine or bacteriological autoclaved agar,
respectively, at a 1 : 1 ratio in order to entrap bacteria within
the spheres. For control samples, warm agar was mixed with
sterile water in the same ratio. Approximately, 1–3-mm
diameter spheres (controlled using filter tips with different
pipette diameters) were formed by dripping agar droplets
into cold mineral oil. Excess mineral oil was removed from
the resulting agar spheres upon washing twice with sterile
water and slight drying on Whatman filter paper. The agar
spheres were then inserted into the polysulfone solution at
various concentrations [10–25% (w/v)] for a short time
(20–30 s) and then transferred into sterile water to complete
formation of the polymeric membrane coating and to wash
away surplus DMF. Polysulfone-coated agar spheres pre-
pared in the same dilutions, either free or in mesh bags were
incubated for at least 3 weeks in the appropriate environ-
ment. Spheres containing mucus-derived samples were
incubated in aquaria with artificial seawater and in the
presence of the coral F. granulosa. Spheres containing
wastewater-derived samples were incubated in wastewater
bioreactors for a period of 3 weeks. Spheres containing soil-
derived samples were placed in disposable cups filled with
wet soil from the collection site. This approach reduced agar
dehydration and enabled superior transport of nutrients and
excretes between the sphere and the environment through
the membrane. After incubation, the agar spheres were
FEMS Microbiol Ecol 68 (2009) 363–371
In situ method for cultivating microorganisms
ethanol flamed to remove external contaminants. Such
flaming does not harm the membrane barrier. The coated
spheres were then cut, the embedded agar spheres removed,
flattened and scanned microscopically for the presence of
microbial colonies at magnifications of  400 and  1000.
In order to enrich and isolate microorganisms, the coated
original agar spheres were transected by sterile scalpel and
the agar spheres were removed, diluted and remixed with
warm agar to form new spheres, as above. The coated
spheres were then repeatedly reincubated in the appropriate
environment. Following every two to three transfers, the
samples were streaked onto appropriate agar plates (e.g.
marine agar for marine environment).
Electron microscopy
For scanning electron microscopy (SEM), the polysulfone
spheres were washed, dehydrated, dried overnight, sputter
coated with gold and examined using a JEOL JSM 5600
scanning electron microscope.
DNA extraction
After appropriate incubation, the agar spheres (coated by
polysulfone) were ethanol flamed to remove external con-
taminants and then transected using a sterile scalpel to
liberate the agar spheres. DNA was then extracted from the
agar-encased microcolonies using a chemical lysis protocol
with a cetyltrimethylammonium bromide solution, followed
by phenol/chloroform extraction (Ausubel et al., 1995) or
using a NucleoSpin food purification kit (Macherey-Nagel,
Düren, Germany). Genomic DNA was eluted using
20–40 mL of elution buffer or double-distilled water. Con-
centrations were determined with an ND-1000 UV–
Vis spectrophotometer (NanoDrop Technologies, Wilmington,
DE) and stored at  20 1C.
PCR amplification
Total DNA was amplified with a Mastercycler gradient
thermocycler (Eppendorf, Westbury, NY) by PCR using
specific 16S rRNA gene primers for bacteria [forward
primer, 8F (GGATCCAGACTTTGAT(C/T)(A/C)TGGCTC
AG) was shortened from 5 0 end, and reverse primer, 1512R
(GTGAAGCTTACGG(C/T)TAGCTTGTTACGACTT), both
taken from Felske et al. (1997)]. For Archaea, Arch-21F [5 0 -
TTCCGGTTGATCCYGCCG-3 0 ; taken from DeLong (1992),
and modified by shortening two bases from the 3 0 end to
avoid self-primer dimering] and Arch-915R [5 0 -GTGCTCC
CCCGCCAATTC-3 0 ; taken from Amann et al. (1995), and
modified by shortening two bases from the 3 0 end to avoid
internal annealing between two primers in some archaeal
sequences], were used. For eukaryotes, primers NSF4/18 (5 0 -
CTGGTTGATYCTGCCAGT-3 0 ) and NSR1787/18 (5 0 -CYG
FEMS Microbiol Ecol 68 (2009) 363–371
365
CAGGTTCACCTACRG-3 0 ) from the European Ribosomal
RNA Database (http://www.psb.ugent.be/rRNA/index.html)
were used. To eliminate dimer formation or nonspecific
annealing, the universal primers were tested using the
AMPLIFY 1.0 program. Primers used in the PCR amplifications
were obtained from Sigma-Genosys. Reaction mixtures
included a 12.5-mL ReddyMix (PCR Master mix containing
1.5 mM MgCl2 and 0.2 mM of each deoxynucleoside tripho-
sphate) (ABgene, Surrey, UK), 1 pmol each of the forward
and reverse primers, 1–2 mL of the sample preparation, and
water to make up the total volume to 25 mL. An initial
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
denaturation-hot start of 4 min at 95 1C was followed by 30
cycles of the following pattern: 94 1C for 30 s, 53–56 1C for
40 s and 72 1C for 105 s. A final soak at 72 1C for 20 min
concluded the reaction.
Clone library construction and sequencing
The PCR products were purified by electrophoresis on 0.8%
agarose (Sigma), stained with ethidium bromide, and visua-
lized with a UV transilluminator. An c. 1.5-kbp heterologous
16S rRNA gene products were excised from the gel, and the
DNA products were purified from the gel slice using a Wizard
PCR prep kit (Promega, Madison, WI). The gel-purified PCR
products were cloned into the pGEM-T Easy cloning vector,
as specified by Promega and transformed into calcium
chloride-competent E. coli DH5a cells, according to the
manufacturer’s instructions and standard techniques.
Plasmid DNA was isolated from individual clones using a
Wizard Plus SV minipreps DNA purification system (Pro-
mega). Aliquots from a subset of the samples of purified
plasmid DNA were digested with the restriction enzyme
EcoRI (MBI Fermentas) for 4 4 h at 37 1C. The digested
products were separated by electrophoresis on 1% agarose
(agarose low electroendosmosis; Hispanagar, Spain). After
staining with ethidium bromide, the bands were visualized
using a UV transilluminator to select clones containing the
appropriately sized insert.
Restriction fragment length polymorphism (RFLP) of
DNA inserts from these clones was achieved through double
digestion with EcoRI and RsaI, StyI, SphI or EcoRV restric-
tion enzymes (MBI Fermentas). The digest products were
then separated by electrophoresis on a 2–2.5% agarose,
stained with ethidium bromide, and the RFLP patterns were
used to identify unique clones to be submitted for sequence
analysis. In addition to the 8F and NSF4/18 primers,
universal internal primers, i.e. 519F (5 0 -CAGCMGCCGCG
GTAATWC-3 0 ) (Connon & Giovannoni, 2002) and 968F
(5 0 -AACGCGAAGAACCTTAC-3 0 ) (Felske et al., 1996) for
prokaryotic 16S rRNA gene sequences, and NSF573/19
(5 0 -CGCGGTAATTCCAGCTCCA-3 0 ) and NSF963/18
(5 0 -TTRATCAAGAACGAAAGT-3 0 ) for eukaryotic 18S rRNA
gene sequences (http://www.psb.ugent.be/rRNA/index.html),

c2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
366
were used for complete sequencing of some bacterial clones.
Sequencing was performed using ABI PRISM dye termina-
tor cycle sequencing ready reaction kit with AmpliTaq DNA
polymerase FS and an ABI model 373A DNA sequencer
(Perkin-Elmer).
Sequence analyses
FastGroupII was used to dereplicate the 16S rRNA gene
sequence libraries for subsequent analyses (Yu et al., 2006)
by comparing all the sequences in a data set to each other,
grouping similar sequences together and outputting a
representative sequence from each group. For this study,
optimal strategies for dereplicating sequences were used,
including trimming ambiguous bases from the 5 0 end of the
sequences and downstream of the conserved Bact517 site
(5 0 -CCAGCAGCCGCGGTAAT-3 0 ), as well as grouping se-
quences Z97% identical to each other (Yu et al., 2006).
The representative 16S rRNA gene sequences of each
group were first compared with those in the GenBank
database, using BLAST (http://www.ncbi.nlm.nih.gov/BLAST/
BLAST.cgi) or CLASSIFIER version 1.0 (to assign 16S rRNA
gene sequences to a taxonomical hierarchy), available at the
Ribosomal Database Project II website (Cole et al., 2005), to
find closely related sequences for subsequent analyses. To
control the occurrence of possibly chimeric sequences, all
sequenced clones were analyzed by the CHIMERA CHECK
program of the RDP database (version 2.7).
Sequences were aligned using CLUSTALW with the MEGA
version 3.1 package (Kumar et al., 2004) and positions not
(a) (b)
(c) (d) (e)

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
E. Ben-Dov et al.
sequenced in all isolates or those with alignment uncertain-
ties were removed. Phylogenetic trees were constructed by
the Neighbor-Joining method (Saito & Nei, 1987) with the
MEGA package (Kumar et al., 2004). Bootstrap resampling
analysis (Felsenstein, 1985) for 100 replicates was performed
to estimate the confidence levels of tree topologies.
Nucleotide sequence accession number
The DNA sequences described here have been deposited in
the NCBI GenBank database under accession numbers
DQ097275–DQ097307, DQ073056, DQ073057, DQ073059
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
and EF371005–EF371010 for coral microorganisms and
DQ099468–DQ099490 for soil and wastewater bacteria.
Results
Feasibility of the encapsulation method
The method presented here involves cultivating microor-
ganisms, singly or in mixed culture by encapsulating them
within agar spheres coated by polysulfonated polymeric
membranes (Fig. 1b). The coated spheres are then incubated
in various test environments (Fig. 1a). This method allows
essential nutrients (including those excreted by mutualistic
bacteria) to be accessed by the entrapped microorganism
while retaining that microorganism and allowing it to grow
in pure or mixed culture (Fig. 1c–f).
The permeability of the polysulfone coating layer for
soluble nutrients was evaluated using Blue Dextran (MW,
2  106) embedded in agar spheres that were incubated in
Fig. 1. Culturing of microorganisms within the
spheres incubated on the surface of the
(f)
coral Fungia granulosa: (a) F. granulosa with
coated agar spheres (scale bar = 2 cm).
(b) Polysulfone-coated agar sphere (  60; scale
bar = 1 mm). In the inset (  12 000; scale
bar = 5 mm), the surface of the sphere as seen by
SEM is shown. (c–f) Phase contrast images
(  1000; scale bar = 10 mm) showing various
microcolonies grown within agar spheres after
3–6 weeks of incubation.
FEMS Microbiol Ecol 68 (2009) 363–371
In situ method for cultivating microorganisms 367

sterile tap water. Release of Blue Dextran, as determined
spectrophotometrically, was 0.445 mg mL1 h1 with 10%
polysulfone-coated spheres and proportionally decreased
with an increase in the concentration of the polysulfone
coating solution, up to 25% (0.069 mg mL1 h1) (Fig. 2a).
SEM showed that pore size, type, compaction and the number
of layers in the coating depended on polymer concentration
[Fig. 1b (inset) and Fig. 2b].
Thus, having confirmed the permeability of the sphere
coats, the feasibility of this method was first evaluated by
encapsulation of E. coli cells. The results showed an increase
in bacterial growth in the spheres from 104 cells mL1 up to
concentration results in increased permeability. Coated
spheres treated by ethanol flaming (to prevent contamina-
tion) before and after incubation in LB broth display the
same E. coli cell concentrations as untreated coated spheres.
However, ethanol flaming might be harmful to some
microbial communities; within the spheres, other methods
such as sterilization via UV exposure could be used. This
experiment also demonstrated that bacteria do not escape
from the agar spheres through the polymeric membrane
prepared from 10% polysulfone/DMF solution, and vice
versa (i.e. E. coli do not pass through the polysulfone coat

Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023

from LB broth into unoccupied agar spheres). In addition,
107 and 109 cells mL1 as the concentration of the polysulfo- to determine if DMF had any deleterious effects, E. coli was
nated polymer in the solvent (DMF) was reduced from 20% grown on LB medium containing different concentrations
to 10%, respectively. This implies that a decrease in polymer of DMF. DMF at concentrations up to 1% did not affect the

(a)
0.6
Blue dextran diffusion rate (mg mL –1h–1)

0.5

0.4

0.3

0.2

0.1

0
10 15 20 25
Polysulfone concentration (%)

(b) A B

C D
Fig. 2. (a) Diffusion rate of Blue Dextran
embedded in agar spheres coated with different
concentrations of polysulfone. Bar heights
represent means based on average samples of
eight spheres, while error bars represent SDs.
(b) SEM of cross-sections of coated spheres
(  40; scale bar = 1 mm) formed by different
concentrations (A, 10%; B, 15%; C, 20%; D,
25%) of polysulfone/DMF solution.

FEMS Microbiol Ecol 68 (2009) 363–371 

c2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
368
growth rate of E. coli, although 5% DMF reduced growth to
1/3 and 10% DMF completely stopped growth (data not
shown). Because agar spheres coated by polymeric polysul-
fone containing traces of DMF washed away immediately
after sphere formation (see Materials and methods), the
solvent likely has only a negligible effect on encapsulated
microorganisms.

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
E. Ben-Dov et al.
Soil and wastewater samples
16S rRNA gene clone libraries constructed from different
spheres inoculated with soil-derived bacteria and incubated
for 3 weeks in soil revealed a diverse microbial community
(Fig. 3, in green). For example, sequence HH.Sph.064
(DQ099481) showed 99% identity with Pseudomonas
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
Fig. 3. Phylogenetic tree based on 16S rRNA
gene sequences that were retrieved from
agar spheres obtained from Fungia granulosa
mucus (in blue), soil (in green) and industrial
wastewater (in red). The trees were
constructed by the Neighbor-Joining method
(Saito & Nei, 1987) with the MEGA package
(Kumar et al., 2004), using partial 16S rRNA
gene sequences. The bar represents five
substitutions per 100 nucleotide positions.
Bootstrap probabilities (Felsenstein, 1985)
are indicated at branch nodes. The NCBI/
GenBank accession numbers for reference
strains are shown in parentheses.
FEMS Microbiol Ecol 68 (2009) 363–371
In situ method for cultivating microorganisms
gessardii (AF074384) (Verhille et al., 1999) which was
isolated from natural mineral waters. Another sequence,
BG.Sph052 (DQ099469), showed 88% and 92% identity
with Agrobacterium sp. (AJ295674) and an uncultured
alphaproteobacterium (AY695728), respectively, indicating
this sequence as corresponding to a possible representative
of a new ribotype.
Molecular analysis of spheres inoculated with wastewater-
derived bacteria, containing a mixture of a few microcolo-
nies, also revealed several novel and interesting sequences
(Fig. 3, in red). For example, one microcolony, RH.Sph067
(DQ099484), displayed 98% identity with Alcaligenes sp.
(AY346141), a phenol-degrading strain that was previously
isolated from an activated sludge system (Zhang et al.,
2004). Comparison of enumeration of microbial cell num-
bers from soil and wastewater environmental samples by
CFU with most-probable-number method (by 10-fold di-
luted samples in agar spheres) revealed that the latter
technique (108–109 cells mL1) was about 10–200-fold high-
er than the former (5  106–107 cells mL1) (Table 1).
Marine (coral)-derived samples
The most-studied environment in this project was the sur-
face mucus layer of the Red Sea coral F. granulosa. Conven-
tional plating techniques of coral mucus samples on marine
agar plates revealed the appearance CFU from dilution of
samples (up to 105), whereas colonies of microorganisms
from agar spheres were detected from dilutions 10 to 100-
fold greater (Table 1). Partial analysis of several spheres from
these dilutions yielded 31 bacterial (Fig. 3, in blue), two
archaeal and eight eukaryotic (Fig. 4) distinct ribotypes.
Because about 50% of the sequences shared identity with
previously identified 16S and 18S rRNA gene sequences of
cultured microorganisms at the 85–96% level, it is probable
that they represent novel ribotypes. One example,
Table 1. Measurement of microorganisms from different environmental
samples by CFUs, microscopic observation and by bacterial culturing in
spheres
Microbial cell number (cells mL1)
Sample CFU Microscope Spheres
Wastewater 5  106 2  109 109
Soil 107 109 108
Coral mucus 3  105 2  107 107
CFU for coral mucus samples obtained on marine agar plates incubated
in a dark/light (12 h) regime at 22 1C, and for soil samples and waste-
water samples on LB and nutrient broth agar plates at 30 1C. Bacteria
from environmental samples were counted by epifluorescence micro-
scopy (DAPI staining) or directly by phase microscopy. Enumeration of
bacterial cell numbers from subsequent 10-fold diluted samples within
agar spheres, measured after 3 weeks of incubation by scanning
microscopically for the presence of microbial microcolonies.
FEMS Microbiol Ecol 68 (2009) 363–371
369
RS.Sph.005 (DQ097279), demonstrated 93% similarity to a
marine gammaproteobacterium (AY386344) (Fig. 3, in
blue), previously isolated by dilution to extinction in
nutrient-deprived medium (Cho & Giovannoni, 2004).
Other sequences, such as RS.Sph.008 (DQ097282),
RS.Sph.009 (DQ097283) and RS.Sph.018 (DQ097292),
showed lower similarities (89–91%) to known cultured
bacteria (Roseobacter sp., AY536579; Cytophaga sp.,
AJ431254). Sequence RS.Sph.002 (DQ097276) demon-
strates 88% similarity to Opitutus sp. (X99392), a genus
belonging to the division Verrucomicrobia, isolated by dilu-
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
tion culture technology (Janssen et al., 1997). These rela-
tively low similarities to known cultured stains demonstrate
the importance of this encapsulation technique in retrieving
novel bacteria from the environment.
This technique also allows for the isolation and culture of
unicellular Eukaryotic organisms (e.g. zooxanthellae, stra-
menopiles and fungi). Using this encapsulation method, we
were able to isolate, identify and cultivate a number of novel
single-celled eukaryotes from F. granulosa mucus. Molecular
analysis of the organisms grown in the spheres revealed a
diverse community of eukaryotic microorganisms with
representatives of various stromanopiles genera (Fig. 4).
For example, we identified microorganisms related to the
Labyrinthula sp. (Euc-1) (Figs 1f and 4). The Euc-1
(DQ073056) sequence demonstrates 87% similarity to the
Labyrinthula sp. AN-1565 (AB022105) and Labyrinthula sp.
L59 (AB095092), known to produce n-6 docosapentaenoic
acid (n-6 DPA), a long chain polyunsaturated fatty acid
(Kumon et al., 2003).
In general, streaking of subsamples from open primary
spheres containing developed microcolonies onto agar
plates did not result in colony development, suggesting that
those microorganisms were apparently uncultivable using
standard microbiological techniques.
Discussion
To establish proof of concept of our method of in situ
incubation, we assessed the cultivability of environmental
samples of microorganisms from three different environ-
ments (i.e. coral surface, soil and wastewater). The experi-
ments described here demonstrate that this method allows
bacterial growth in a number of different simulated or
natural environments (Figs 1, 3 and 4). The small size of the
coated spheres (c. 1–3 mm in diameter) enabled the transfer
of nutrients and excretes (inflow and outflow) between the
sphere and the environment (see Fig. 2 for outflow). In
addition, the membrane coating remained intact over a
period of several months of incubation in the respective
environments, allowing for the culturing of slow-growing
microorganisms. Moreover, we found that we could culture
microorganisms at concentrations that were at least 10-fold

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
370
100 Euc-3
100 Nitzschia thermalis (AY485458)
Euc-4
100 Amphiprora paludosa (AY485468)
78 Lab-1
100 Pinguiococcus pyrenoidosus (AF438324)
Euc-2
100 Rhizochromulina marina (U14388)
100 Euc-7
Euc-6
100 Codonosiga gracilis (AY149897)
Lab-3
Thraustochytrium multirudimentale AB022111)
Aplanochytrium stocchinoi (AJ519935)
Euc-1
84 Labyrinthula sp. AN-1565 (AB022105)
86 Labyrinthula sp. L59 (AB095092)
Euc-5
100 Perkinsiella -like sp. (AY163355)
0.05
less than the concentration needed to obtain colonies on
agar plates (Table 1).
Over the last decade, culture-independent surveys have
provided an enormous amount of information on the
diversity, function and distribution of the uncultivated
majority of the microbial world. Information about the
nutritional requirements of a given microorganism, in situ
enzyme expression and its genetic potential all can be used
to predict optimal conditions for cultivation (Giovannoni
et al., 2007; Ingham et al., 2007). Several studies have used
novel techniques for selective isolation of nongrowing
bacteria by incubation in low-nutrient broth (Manome
et al., 2001; Connon & Giovannoni, 2002), incubation in
low-nutrient flux conditions (Manome et al., 2001; Kaeber-
lein et al., 2002; Zengler et al., 2002) or exponential
miniaturization of microbial cultures, such as the ‘micro-
Petri chip’ that contains a million microchambers and can
support the growth of various, previously uncultivable
microorganisms (Ingham et al., 2007). Application of one
such technique, agar gel microdroplets (GMD), was shown
to be particularly suitable for isolating certain slow-growing
bacteria by elimination of overgrowth by other bacteria
(Manome et al., 2001; Zengler et al., 2002; Akselband et al.,
2006). Application of GMD and flow cytometry for selective
enrichment of nongrowing Leuconostoc mesenteroides cells
was demonstrated by Manome et al. (2001). Zengler et al.
(2002) incubated GMDs (occupied by single encapsulated
cells from seawater or soil samples) in growth columns,
where low-nutrient media were pumped through filter
membranes. These techniques have, however, proven to be
limited because there are no barriers between the GMD and
the environment, thus preventing the feasibility of using this
method for in situ incubation. An additional approach used
for in situ incubation uses diluted marine microorganisms
(from intertidal marine sediment) mixed with warm agar
made with seawater. These mixtures are placed in a diffusion

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
E. Ben-Dov et al.
Stramenopiles
Fig. 4. Phylogenetic tree based on 18S rRNA
gene sequences that were retrieved from agar
Fungi/Metazoa
spheres (in bold) obtained from Fungia granulosa
mucus. The trees were constructed by the
Neighbor-Joining method (Saito & Nei, 1987)
Stramenopiles with the MEGA package (Kumar et al., 2004),
using partial 18S rRNA gene sequences. The bar
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
represents five substitutions per 100 nucleotide
positions. Bootstrap probabilities (Felsenstein,
Euglenozoa
1985) are indicated at branch nodes. The NCBI/
GenBank accession numbers for reference strains
are shown in parentheses.
chamber made of stainless steel washers between two poly-
carbonate membranes glued to the washers (Kaeberlein
et al., 2002; Bollmann, et al., 2007). These chambers were
then incubated on the surface of marine sediment kept in
aquaria. Although this technique yielded a maximum of
40% recovery (Kaeberlein et al., 2002), it is limited by the
relatively large size of the washers, which prevent compar-
ison of microniches and has been proven to work in aquatic
environments. Because of the small size of the spheres, the
inflexible nature and the selective permeability of the poly-
sulfone membrane, our technique is feasible both in in situ
(in aquatic as well as terrestrial environments) and labora-
tory incubations.
This study demonstrates the applicability of our techni-
que both in terms of environmental microbial diversity
assessment and increased culturability of environmentally
derived microorganisms. Our double encapsulation techni-
que hence offers an additional and novel tool, providing
access to previously unknown microbial diversity.
Acknowledgements
This work was supported by ISF grant nos 511/02 and 1169/
07 and grant no. WT-501 from the BMBF-MOST Coopera-
tion in Water Technologies. We thank N. Siboni for his help
in sample collection and technical support and the Inter-
University Institute in Eilat for use of their facilities.
References
Akselband Y, Cabral C, Castor TP, Chikarmane HM & McGrath P
(2006) Enrichment of slow-growing marine microorganisms
from mixed cultures using gel microdrop (GMD) growth assay
and fluorescence-activated cell sorting. J Exp Mar Biol Ecol
329: 196–205.
FEMS Microbiol Ecol 68 (2009) 363–371
In situ method for cultivating microorganisms
Amann RI, Ludwig W & Schleifer KH (1995) Phylogenetic
identification and in situ detection of individual microbial cells
without cultivation. Microbiol Rev 59: 143–169.
Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG,
Smith JA & Struhl K (1995) Short Protocols in Molecular
Biology, 3rd edn. John Wiley and Sons Inc., New York.
Bollmann A, Lewis K & Epstein SS (2007) Incubation of
environmental samples in a diffusion chamber increases the
diversity of recovered isolates. Appl Environ Microb 73:
6386–6390.
Cho JC & Giovannoni SJ (2004) Cultivation and growth
characteristics of a diverse group of oligotrophic marine
Gammaproteobacteria. Appl Environ Microb 70: 432–440.
Cole JR, Chai B, Farris RJ, Wang Q, Kulam SA, McGarrell DM,
Garrity GM & Tiedje JM (2005) The Ribosomal Database
Project (RDP-II): sequences and tools for high-throughput
rRNA analysis. Nucleic Acids Res 33: D294–D296.
Connon SA & Giovannoni SJ (2002) High-throughput methods
for culturing microorganisms in very-low-nutrient media yield
diverse new marine isolates. Appl Environ Microb 68: 3878–3885.
Cottrell MT, Waidner LA, Yu L & Kirchman DL (2005) Bacterial
diversity of metagenomic and PCR libraries from the Delaware
River. Environ Microbiol 7: 1883–1895.
DeLong EF (1992) Archaea in coastal marine environments.
P Natl Acad Sci USA 89: 5685–5689.
Felsenstein J (1985) Confidence limits of phylogenies: an
approach using the bootstrap. Evolution 39: 783–791.
Felske A, Engelen B, Nübel U & Backhaus H (1996) Direct
ribosome isolation from soil to extract bacterial rRNA for
community analysis. Appl Environ Microb 62: 4162–4167.
Felske A, Rheims H, Wolterink A, Stackebrandt E & Akkermans
AD (1997) Ribosome analysis reveals prominent activity of an
uncultured member of the class Actinobacteria in grassland
soils. Microbiology 143: 2983–2989.
Giovannoni SJ, Foster RA, Rappé MS & Epstein S (2007) New
cultivation strategies bring more microbial plankton species
into the laboratory. Oceanography 20: 62–69.
Ingham CJ, Sprenkels A, Bomer J, Molenaar D, van den Berg A,
van Hylckama Vlieg JET & de Vos WM (2007) The micro-Petri
dish, a million-well growth chip for the culture and high-
throughput screening of microorganisms. P Natl Acad Sci USA
104: 18217–18222.
Janssen PH, Schuhmann A, Morschel E & Rainey FA (1997)
Novel anaerobic ultramicrobacteria belonging to the
FEMS Microbiol Ecol 68 (2009) 363–371
371
Verrucomicrobiales lineage of bacterial descent isolated by
dilution culture from anoxic rice paddy soil. Appl Environ
Microb 63: 1382–1388.
Kaeberlein T, Lewis K & Epstein SS (2002) Isolating
‘‘uncultivable’’ microorganisms in pure culture in a simulated
natural environment. Science 296: 1127–1129.
Kumar S, Tomura K & Nei M (2004) MEGA3: integrated software
for molecular evolutionary genetics analysis and sequence
alignment. Brief Bioinform 5: 150–163.
Kumon Y, Yokoyama R, Yokochi T, Honda D & Nakahara T
(2003) A new labyrinthulid isolate, which solely produces
n-6 docosapentaenoic acid. Appl Microbiol Biot 63:
Downloaded from https://academic.oup.com/femsec/article/68/3/363/431828 by guest on 25 February 2023
22–28.
Manome A, Zhang H, Tani Y, Katsuragi T, Kurane R & Tsuchida T
(2001) Application of gel microdroplet and flow cytometry
techniques to selective enrichment of non-growing bacterial
cells. FEMS Microbiol Lett 197: 29–33.
Rappé MS & Giovannoni SJ (2003) The uncultured microbial
majority. Annu Rev Microbiol 57: 369–394.
Saito N & Nei M (1987) The neighbor-joining method: a new
method for constructing phylogenetic trees. Mol Biol Evol 4:
406–425.
Staley JT & Konopka A (1985) Measurements of in situ activities
of nonphotosynthetic microorganisms in aquatic and
terrestrial habitats. Annu Rev Microbiol 39: 321–346.
Verhille S, Baida N, Dabboussi F, Hamze M, Izard D & Leclerc H
(1999) Pseudomonas gessardii sp. nov. and Pseudomonas
migulae sp. nov., two new species isolated from natural
mineral waters. Int J Syst Bacteriol 49: 1559–1572.
Wagner M, Amann R, Lemmer H & Schleifer KH (1993) Probing
activated sludge with oligonucleotides specific for
proteobacteria: inadequacy of culture-dependent methods for
describing microbial community structure. Appl Environ
Microb 59: 1520–1525.
Yu Y, Breitbart M, McNairnie P & Rohwer F (2006) FastGroupII:
a web-based bioinformatics platform for analyses of large 16S
rRNA gene libraries. BMC Bioinformatics 7: 57.
Zengler K, Toledo G, Rappé M, Elkins J, Mathur EJ, Short JM &
Keller M (2002) Cultivating the uncultured. P Natl Acad Sci
USA 99: 15681–15686.
Zhang X, Gao P, Chao Q, Wang L, Senior E & Zhao L (2004)
Microdiversity of phenol hydroxylase genes among phenol-
degrading isolates of Alcaligenes sp. from an activated sludge
system. FEMS Microbiol Lett 237: 369–375.

c 2009 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved