234
For the fi rst time, we have provided an accurate compari-
son of fi ve most used EM and TEM protocols for virus stud-
ies based on freshwater communities collected in Lakes
Pavin and Aydat. Our results revealed that the quantitative
determination of viruses was different according to the type
of fl uorochrome used (Bettarel et al. 2000 ). Nowadays,
SYBR Green I is the most used fl uorochrome, when coupled
with a powerful antifading, it allows to improve the observa-
tion of viruses in EM (Suttle and Furhman 2010 ). However,
it is important to specify that the latter targets preferentially
dsDNA viruses and has very few affi nity with ssDNA and
RNA viruses for which the use of SYBR Gold would be
more suitable. The fi rst protocol for high throughput count-
ing of free-fl oating viruses using fl ow cytometry and SYBR
Green I stain was developed for pelagic oligotrophic marine
samples (Marie et al. 1997 , 1999 ), and was later on opti-
mized for both marine (Brussaard 2004 ) and freshwater sam-
ples (Duhamel and Jacquet 2006 ).
14.3.3 Concentration of Viruses
for Microscopic and Molecular Genetic
Analyses
The genetic and biological diversity of aquatic viruses
remains largely unexplored, mainly because of the method-
ological diffi culties related to the observation of environ-
mental nanoparticles and the weakness of environmental
viral gene banks. The study of free-fl oating viruses in aquatic
systems indeed requires a concentration method for estimat-
ing their number and their diversity based on both morpho-
logic and genomic features. The fi rst generation of methods
used came from the study of human, animal and plant patho-
genic viruses. These methods are based on physicochemical
approaches such as (i) adsorption of phages onto micropo-
rous fi lters charged positively or negatively, (ii) the use of
precipitation agents like calcium phosphate, aluminium,
magnesium or acid precipitation, and (iii) the use of organic
fl occulation (cf. Colombet et al. 2007 ). The main disadvan-
tage of these techniques for natural samples is their selectiv-
ity because viruses may have different adsorptive properties,
while electrostatic interactions may affect the viability of
concentrated viruses.
As described above, Børsheim et al. ( 1990 ) have proposed
the fi rst concentration method based on the ultracentrifuga-
tion of viruses contained in few milliliters of samples directly
onto grids, followed by staining before observation under
TEM. Disadvantages of this method include the risk of viral
disruption due to the high centrifugation speed, the effects of
suspended particles and ‘virucidal’ substances, and the low
initial volume of experimental samples that may render TEM
observations tedious according to the high magnifi cations
used. The method is thus costly and time-consuming for
large volume of samples, and therefore cannot be recom-Fig. 14.4 Frequencies of visibly infected bacteria recorded in lakes,
along a trophic gradient.◇ correspond to short-term studies (<5 days),
◆ correspond to long-term studies (>3 months) (Sources: [a] Hofer and
Sommaruga ( 2001 ); [b] Tapper and Hicks ( 1998 ); [c] Bettarel et al.
( 2003a , b ); [d, k] Bettarel et al. ( 2004 ); [e] Mathias et al. ( 1995 ); [f]
Hennes and Simon ( 1995 ); [g] Wilhelm and Smith ( 2000 ); [h] Simek
et al. ( 2001 ); [i] Weinbauer and Höfl e ( 1998 ); [j] Fischer and Velimirov
( 2002 ) )T. Sime-Ngando et al.