Microb Ecol (2014) 68:740750
DOI 10.1007/s00248-014-0441-2
ENVIRONMENTAL MICROBIOLOGY
Linking Host Prokaryotic Physiology to Viral Lifestyle Dynamics
in a Temperate Freshwater Lake (Lake Pavin, France)
S. Palesse & J. Colombet & A. S. Pradeep Ram &
T. Sime-Ngando
Received: 6 January 2014 / Accepted: 20 May 2014 / Published online: 10 June 2014
# Springer Science+Business Media New York 2014
Abstract In aquatic ecosystems, fluctuations in environmental
conditions and prokaryotic host physiological states can strong-
ly affect the dynamics of viral life strategies. The influence of
prokaryote physiology and environmental factors on viral rep-
lication cycles (lytic and lysogeny) was investigated from April
to September 2011 at three different strata (epi, meta, and
hypolimnion) in the mixolimnion of deep volcanic temperate
freshwater Lake Pavin (France). Overall, the euphotic region
(epi and metalimnion) was more dynamic and showed signif-
icant variation in microbial standing stocks, prokaryotic phys-
iological state, and viral life strategies compared to the aphotic
hypolimnion which was stable within sampled months. The
prokaryotic host physiology as inferred from the nucleic acid
content of prokaryotic cells (high or low nucleic acid) was
strongly regulated by the chlorophyll concentration. The pre-
dominance of the high nucleic acid (HNA) prokaryotes (cells)
over low nucleic acid (LNA) prokaryotes (cells) in the spring
(HNA/LNA=1.2) and vice versa in the summer period (HNA/
LNA=0.4) suggest that the natural prokaryotic communities
underwent major shifts in their physiological states during
investigated time period. The increase in the percentage of
inducible lysogenic prokaryotes in the summer period was
associated with the switch in the dominance of LNA over
HNA cells, which coincided with the periods of strong resource
(nutrient) limitation. This supports the idea that lysogeny rep-
resents a maintenance strategy for viruses in unproductive or
harsh nutrient/host conditions. A negative correlation of per-
centage of lysogenic prokaryotes with HNA cell abundance
and chlorophyll suggest that lysogenic cycle is closely related
to prokaryotic cells which are stressed or starved due to un-
availability of resources for its growth and activity. Our results
S. Palesse : J. Colombet : A. S. Pradeep Ram (*) : T. Sime-Ngando
Laboratoire Microorganismes: Gnome et Environnement, UMR
CNRS 6023, Clermont Universit, Universit Blaise Pascal, BP
80026, 63171 Aubire Cedex, France
e-mail: Pradeep_Ram.ANGIA_SRIRAM@univ-bpclermont.fr
provide support to previous findings that changes in prokaryote
physiology are critical for the promotion and establishment of
lysogeny in aquatic ecosystems, which are prone to constant
environmental fluctuations.
Introduction
Viruses, primarily bacteriophages, represent the most abun-
dant and diverse biological entities in aquatic environments
[1, 2]. They play an important role in various ecological and
biogeochemical processes by regulating microbial diversity,
carbon and nutrients fluxes, and food web dynamics and
mediating lateral gene transfers [24]. Viruses are excellent
survivors in a wide range of environments, and one of the key
explanations for their omnipresence is through the existence
of several lifestyles, of which two major pathways, namely
lysis and lysogeny, are prevalent and mostly studied in aquatic
and phage-hosts systems [4, 5]. Lytic infections result in host
cell death via lysis following the replication of viral particles,
whereas lysogeny has been described as a means of survival
for viral populations, using the host as a refuge under chal-
lenging environmental conditions. During lysogenic infection,
the temperate phage can alternatively integrate into the hosts
genome (as a prophage) and can remain until an induction
event takes place, triggering the onset of viral production [6,
7]. Studies with cultured isolates and natural prokaryotic
communities indicated that lytic infection occurred preferen-
tially in more productive ecosystems characterized by high
host abundance and activity, whereas lysogeny is known to
thrive under unfavorable conditions such as low host abun-
dance and primary production [8, 9]. In aquatic systems, viral
life strategies are strongly influenced by environmental and
biological factors. More precisely, nutrient availability and
temperature have been described as major factors influencing
the two viral pathways [1014].
Linking Host Prokaryotic Physiology to Viral Lifestyle Dynamics
Very few studies have explored both viral lifestyles simul-
taneously in natural aquatic ecosystems and have resulted in
contrasting ecological scenarios. Spatiotemporal studies on
both lytic and lysogenic infections have either reported an
inverse relationship [6, 10, 1517], a positive relationship
[10, 18], or no correlation between both cycles [12, 19]. These
results strongly illustrate that the replacement patterns neither
occur in all natural systems or within a same ecosystem over
time, which emphasized that factors controlling those lifestyles
dynamics are not fully understood. More than the availability,
the quality in terms of physiological states and activity of
prokaryotic hosts may strongly influence the viral decision
toward lytic or lysogenic lifestyle. While most of the studies
have only explored the bulk prokaryotic production as proxy of
biological factors, very few studies have taken into consider-
ation the physiological states of prokaryotic hosts to study viral
life strategies in aquatic ecosystems[10, 14]. Maurice et al. [20]
have reported an interesting scenario where different signals of
prokaryote physiology could control the viral lifestyles in
Mediterranean coastal lagoons. Since lakes are more prone to
changing or fluctuating environmental conditions, prokaryotic
communities in such ecosystems tend to have diverse physiol-
ogy, activity, and community structure. Such constant fluctua-
tion in host physiology and activity could provide ideal setting
or platform to test the hypotheses linking prokaryotic host
physiology to viral lifestyle dynamics in lakes and compare
to those reported previously from marine systems [20].
Lake Pavin (altitude, 1,197 m above sea level), located in
the French Massif Central, is a meromictic and dimictic
oligomesotrophic crater mountain lake that experiences partial
overturns and that is characterized by a maximal depth of 92 m.
The mixolimnion which constitutes 060 m can be differenti-
ated into three different strata (namely, epi, meta, and hypolim-
nion), and these water masses are known to have contrasted
physicochemical conditions with different amplitude of vari-
ability over time (from highly variable to a more stable envi-
ronment) [21]. Therefore, this ecosystem could provide an ideal
environment to test the hypothesis linking host physiological
states and activity with phage life styles. Such studies on the
above are rare especially in a deep freshwater lake. In order to
test the hypothesis that changes in the physiology and activity
of prokaryotic hosts control the phage lifestyles in freshwater
lakes, a 6-month study was carried out in the three different
strata (epi, meta, and hypolimnion) of Lake Pavin. The objec-
tives of this study were (1) to investigate and compare the
prevalence and dynamics of the lytic and lysogenic cycles
among the three strata and (2) to determine the factors that
influence mechanisms of viral life strategies taking into account
the different host physiological states and metabolic activities.
Therefore, the present study is an attempt to simultaneously
investigate two viral lifestyles (lytic and lysogenic) on a spa-
tiotemporal scale and bring out the knowledge on hostvirus
interactions and bacterioplankton regulations.
741
Materials and Methods
Study Site and Sampling
Water samples from Lake Pavin were collected bimonthly
from April to September 2011 at the central reference location
in the lake (2 53 12 E, 45 29 41 N). The spring period
corresponded from the beginning of the study to early June,
and the summer period referred from late June to early Sep-
tember. The three different strata, namely the epilimnion (0
5 m), metalimnion (917 m), and hypolimnion (2045 m)
which were representative of the mixolimnion, were sampled
manually by a depth-integrated sampling method using an 8-l
Van Dorn bottle. Integrated water samples, which were repre-
sentative of each stratum, were collected in 20-l capacity
insulated carboys, which had been previously acid-washed
and rinsed three times with milliQ water and lake water. All
samples were collected in triplicates, i.e., from three indepen-
dent sampling operations. Utmost care was taken to transport
the water samples from the time of collection to laboratory.
During transportation, the water samples were kept in refrig-
erated boxes (filled with ice packs and coolants) and were well
protected from heat and light radiations. The time delay be-
tween sampling and sample processing in laboratory never
exceeded 2 h, and this gap was uniform on all sampling dates.
Water samples were processed immediately upon arrival to the
laboratory, and priority was given to unfixed samples. Water
samples that were fixed on the site with glutaraldehyde were
stored at 4 C and were later processed on the same day.
Physicochemical Analyses
Water temperature and dissolved oxygen profiles were deter-
mined in situ using an oxycal-SL 197 multiparameter probe
(WWT, Limonest, France). Phytoplankton biomass was esti-
mated in situ as micrograms equivalent to chlorophyll a (Chl)
per liter using a submersible fluorescence photometer
(Fluoroprobe; BBE-Moldaenke, Kiel, Germany). The device
was calibrated for the system studied so as to minimize any
likely errors associated with its measurements. Secchi depth
(Zs) measurements were used to estimate the euphotic depth
(Zeu) according to the relationship Zeu=2.42Zs, assuming
that 15 % of the incident light is transmitted to the Zs [22].
According to the above criteria, Zeu did not exceed 20 m in
the present study, and as a result, both the epilimnion and
metalimnion were considered to be located in the euphotic
zone and the hypolimnion in the aphotic zone.
Viral and Prokaryotic Abundance
Enumeration of viruses and prokaryotes in 0.02 m filtered
glutaraldehyde (0.5 % final concentration) preserved samples
were performed using a FACSCalibur flow cytometer (FCM)
742
(BD Sciences, San Jose, CA, USA) equipped with an air-
cooled laser, providing 15 mW at 488 nm with the standard
filter setup. The method used in our survey has been described
in detail by Marie et al.[23] and Brussaard [24, 25]. Briefly,
extracted samples were diluted with filtered TE buffer (10 mM
TrisHCl and 1 mM EDTA at pH 8) and stained with SYBR
Green I (10,000-fold dilution of commercial stock, Molecular
Probes, Oregon, USA). Mixture was incubated for 5 min,
heated for 10 min at 80 C in the dark, and cooled for 5 min
prior to analysis. Populations of viruses and prokaryotes dif-
fering in fluorescence intensity were distinguished on plots of
side scatter versus green fluorescence (530 nm wave length,
fluorescence channel 1 of the instrument). FCM list modes
were analyzed using CellQuest Pro software (BD Biosciences,
version 4.0). A blank was routinely examined to control for
contamination of equipments and reagents.
Cytometric Characteristics and Physiological Status
of Prokaryote Cells
Based on the differences in the individual cell fluorescent
(related to nucleic acid content) and in the side and forward
light scattering signal on the cytogram, at least two major
prokaryotic groups were distinguished: cells with low nucleic
acid content (LNA cells) and cells with high nucleic acid
content (HNA cells) [26]. Cell abundances were determined
for each of these groups.
Physiological status of prokaryotic cells was analyzed ac-
cording to their membrane integrity using the LIVE/DEAD
BacLightTM bacterial viability Kit (Invitrogen). This kit con-
tains a mixture of two types of stains: a red cell-impermeant
nucleic acid stain (propidium iodide) that only penetrate cells
with damaged membranes and a green cell-permeant nucleic
acid stain (SYTO 9) that acts as a counter stain for all cells.
The stain concentration and the incubation time have been
adjusted according to the communities studied: 200 l of stain
solution (2 stock solution) was added to 200 l of subsam-
ples of each container and incubated for 15 min in the dark at
room temperature prior to cytometric analysis. In this assay,
cells with compromised membranes are discriminated from
intact cells in a cytogram of red (FL3) versus green (FL1)
fluorescence. Membrane-compromised cells have higher red
than green fluorescence intensity, because of greater penetra-
tion of the red propidium iodide (PI), whereas membrane-
intact cells have higher green than red fluorescence [27].
Hereafter, the terms of live and dead will be used to and maximum burst size estimates at the start of the experiment.
designate membrane-intact and membrane-compromised
cells, respectively. Controls were routinely performed that
consisted of the same natural water sample heated at 80 C
for 30 min. The proportion of intact and damaged cells
was enumerated using flow cytometer (as mentioned in
previous section).
S. Palesse et al.
Lytic Phage Infection and Burst Size Estimates
Viral lytic infection was inferred from the frequency of infect-
ed cells (FIC) according to Sime-Ngando et al. [28]. Briefly,
prokaryotic cells in 8 ml of water samples (glutaraldehyde,
1 % final concentration) were collected on copper electron
microscope grids (400-mesh, carbon-coated formvar, Pelanne
Instruments, France) by ultracentrifugation (Beckman Coulter
SW40Ti swing-out-rotor at 70,000g for 20 min at 4 C).
Each grid was stained with uranyl acetate for 45 s and rinsed
thrice with 0.02-m filtered distilled water and dried on a filter
paper. Grids were examined using a JEOL1200 EX transmis-
sion electron microscope (TEM) operated at an acceleration
voltage of 80 kV at a magnification of 20,000 to 60,000 to
distinguish between prokaryotic cells with and without intra-
cellular viruses. At least 600 prokaryotes were inspected per
grid to determine the frequency of visibly infected cells
(FVIC). A cell was considered visibly infected when at least
five virus-like particles could be observed in its cytoplasm
[29]. FVIC counts were converted to the FIC using the fol-
lowing formula: FIC=9.524 FVIC3.256 [29] and thereafter
to viral-induced prokaryotic mortality (VIBM) using the fol-
lowing equation: VIBM=(FIC+0.6 FIC2)/11.2 FIC [30].
Burst size (BS, viruses per prokaryote) was estimated by
enumerating viral particles in infected cells, and the number
of infected cells examined was on average 20 cells per sample
(range, 1525).
Induction Assays for Lysogenic Prokaryotes
The frequency of lysogenic cells (FLCs) was obtained after
24-h incubation of water samples (20 ml) with and without
(controls) the inducing agent mitomycin C (1 g ml1 final
concentration, Sigma, MO, USA) in the dark at in situ temper-
ature [31]. Incubations were carried out in triplicates, and
subsamples were collected and fixed at t0 and t24 h with
glutaraldehyde (0.5 % final concentration) for enumeration of
total prokaryotic and viral abundances by flow cytometry using
SYBR Green I stain (as described previously). Induction of
lysogenic prokaryotes was determined from increases in viral
abundance and decreases in prokaryotic abundance in the pres-
ence of mitomycin C relative to controls. FLC within prokary-
otic communities was calculated as FLC=[(VAMC VAC)/
(BSPAt0)]100 where VAMC and VAC are the viral abun-
dance in mitomycin C-treated and control assays after incuba-
tions, respectively. PAt0 and BS are the prokaryotic abundance
Statistical Analysis
All statistical analyses of the results were performed using the
software package XL Stat 2013. Data (both log-transformed
and untransformed) were checked for normality using
Linking Host Prokaryotic Physiology to Viral Lifestyle Dynamics 743

ShapiroWilk Test. Due to the limitation of the assumptions of
homoscedasticity and homogeneity of variances among the
three strata within each variable, the nonparametric Kruskal
Wallis (KW) analysis of variance on rank tests followed by
Dunns post hoc multiple comparison was used to assess the
difference of each parameters between the epi, meta, and
hypolimnion. Comparisons between spring and summer sea-
sons were analyzed by nonparametric MannWhitney
Wilcoxon test for each stratum. Potential relationships be-
tween all microbial and environmental variables were tested
by Spearmans rank correlations analyses.
The magnitude of temporal fluctuations of each environ-
mental and biological variable among the different strata was
compared either with Bartletts test (normal distribution) or
Fligners test (nonnormal distribution) simultaneously with
FisherSnedecors test. In addition, principal component anal-
ysis (PCA) was done to determine whether each sampled layer
represents distinct environments according to the combination
of biotic and abiotic variables.
Results
Environmental Conditions
Water temperature and dissolved oxygen concentration were
significantly different between the three layers during the
sampled period (Table 1). Thermal stratification started in
early April and extended throughout the study period until
September 2011. Among the sampled layers, the epilimnion
showed stronger temperature fluctuations (p<0.001, Table 2)
with the values increasing from 9.1 C in April to 19.6 C in
late August (Fig. 1a). In the hypolimnion, water temperature
was low (mean=4.2 C) and varied less with sampled months
(Table 2). A clear stratification of oxygen concentration was
observed in the mixolimnion (Fig. 1b). Both the epilimnion
and metalimnion were generally well oxygenated
(>9.0 mg l1) and differed significantly (p<0.001) with the
hypolimnion where dissolved oxygen levels were the lowest
(mean=5.00.9 mg l1) (Table 1). The amplitude of temporal
changes in oxygen concentrations was the strongest in the
metalimnion (Table 2) where values fluctuated from 8.7 to
15.8 mg l1. Oxycline was located between 45 and 50 m
during the whole sampling period.
The euphotic region in Lake Pavin fluctuated from 8.7 to
20.0 m with the highest value observed in late August. There-
fore, according to Secchi measurements, both the epilimnion
and metalimnion fell in the euphotic region whereas the
hypolimnion was in the aphotic zone. Chl concentration pro-
file as a proxy of phytoplankton biomass showed that the
metalimnion was the most productive layer (mean=7.0
3.7 g l1) with significantly higher values (p<0.001) than
in the epilimnion (mean=2.41.2 g l1) and hypolimnion
(mean=1.60.8 g l1). Chl content decreased significantly
(p<0.01) in summer compared to spring in both epi and
metalimnion, while it was below the limit of detection in the
hypolimnion (Fig. 1c).

Table 1 Mean (range) environmental and microbial characteristics of epi, meta, and hypolimnion of Lake Pavin, AprilSeptember 2011

Parameters Mean (range)

EPI META HYPO

Water temperature (C)b 15.3 (9.219.6) 8.4 (4.814.8) 4.2 (4.14.3)

Dissolved oxygen (mg l1)b 9.6 (8.411.4) 12.7 (8.715.8) 5.0 (3.66.8)
Chlorophyll a (g l1) 2.3 (0.94.6)* 7.0 (1.412.9)a 1.5 (0.42.7)
Viral abundance (107 ml1) 2.8 (0.84.4)* 3.5 (1.56.6) 0.6 (0.21.1)a
Prokaryotic abundance (106 cells ml1) 1.8 (0.63.2)* 2.3 (0.74.0) 0.7 (0.41.0)a
LNA prokaryotic abundance (106 cells ml1) 1.1 (0.21.9)* 1.2 (0.42.6) 0.4 (0.20.5)a
HNA prokaryotic abundance (106 cells ml1) 0.7 (0.11.8)* 1.0 (0.32.3) 0.3 (0.10.6)a
Proportions of HNA cells (%) 39.7 (11.261.3) 44.0 (22.169.8) 42.4 (22.157.6)
Proportion of damaged prokaryotic cells (%) 32.8 (21.049.1) 37.3 (19.457.0) 38.0 (21.157.1)
Virus-to-prokaryote ratio 16.8 (7.527.1)* 16.3 (9.630.1) 8.9 (4.412.7)a
Frequency of infected cells (%) 16.3 (6.339.6)* 21.5 (5.836.6) 10.3 (2.434.8)a
Frequency of lysogenic cells (%) 1.9 (011.1)* 0.7 (05.3) 0.2 (01.4)a
Burst size (virus prokaryote1) 40 (19118) 42 (2373) 40 (16182)

*p<0.05
a
Indicates a significant difference between two of the three layers (Dunns test)
b
Significant differences between the three layers (Kruskal Wallis)
744 S. Palesse et al.

Table 2 Variances in environmental and microbiological parameters of epi, meta, and hypolimnion of Lake Pavin, AprilSeptember 2011

Parameters Observed variance p value

EPI META HYPO Bartletts test Fligners test

Water temperature 8.65* 2.68 0.01a p<0.0001

Dissolved oxygen 0.64* 4.69a 0.74 p<0.0001
Chlorophyll a 1.47* 13.80a 0.60 p<0.0001
Viral abundance (1013) 3.95* 13.18 0.58a p<0.0001
Prokaryotic abundance (1011) 5.07* 7.40 0.40a p=0.0064
HNA prokaryotic abundance (1011) 1.65* 3.44 0.18a p<0.0001
LNA prokaryotic abundance (1011) 2.44* 2.25 0.07a p<0.0001
Proportion of damaged prokaryotic cells 91.88 163.99 133.09 NS
Frequency of infected cells 59.55 46.56 45.13 NS
Frequency of lysogenic cellsb 7.60 1.61 0.07 p=0.011

NS not significant
*p<0.05
a
Indicate a significant difference of variance between two of the three layers (Fisher-Snedecor)
b
Significant difference between each of the three layers

Viral and Prokaryotic Abundance
We looked for evidence of time series variability in the abun-
dances of viruses (VA) and prokaryotes (PA) and the differences
between the sampled depths. During the studied period, both VA
and PA varied by an order of magnitude especially in the Zeus
(epilimnion and metalimnion). PA fluctuated from 0.6 to 4.0
106 cells ml1 (mean=2.00.8106 cells ml1) in the euphotic
region, and the values were significantly higher (p<0.001)
compared to the hypolimnion (mean=0.70.2106 cells ml1)
(Table 1). The fluctuations in PA were observed at all the
sampled depths with a decline in the middle of spring followed
by a sharp peak in June and thereafter declined gradually during
summer months which extended until September (Fig. 2a).

Fig. 1 Spatiotemporal variations

of water temperature, dissolved
oxygen, and chlorophyll
concentrations in the
mixolimnion of Lake Pavin, from
April to September 2011
Linking Host Prokaryotic Physiology to Viral Lifestyle Dynamics
2
3 early June and was followed by an increase in the LNA cell
Fig. 2 Spatiotemporal variability in the abundance of prokaryotes (a),
viruses (b), and virus to prokaryote ratio (c) in the three sampled depths of
Lake Pavin. Error bars represent SE (n=3)
VA fluctuated from 0.2 to 6.6107 ml1 in the euphotic
region, and the values were significantly higher (p<0.001) than
in the hypolimnion (Fig. 2b, Table 1). In both the epi and
hypolimnion, peaks in the PA was generally followed by the
increase in VA resulting in a strong correlation (r=0.78,
p<0.001) between the two variables (Table 3). The overall virus
to prokaryote ratio (VPR) ranged from 4.4 to 30.1 (mean=14.0
5.4) at the sampled depths. Although VPR did not differ
significantly between the epi and metalimnion, their mean
values were twofold higher than the hypolimnion (Table 1).
Prokaryotic Cell and Physiological Characteristics
Among the prokaryotes, the proportion of HNA cells fluctu-
ated from 11.2 to 69.7 % (mean=42.011.7 %) of the total PA
and did not differ significantly among the sampled depths
(Table 1). In terms of absolute abundance, HNA and LNA
cell abundance in the euphotic zone was significantly higher
(p<0.001) than that in the aphotic hypolimnion (Table 1) and
showed a stronger variability with the sampled months (Fig. 3
745
and Table 2). Indeed, the absolute abundance of HNA cells
differed from that of LNA with respect to months mostly in the
euphotic regions when HNA cells decreased after a peak in
abundance which attained maximum values in July
(Fig. 3a, b). The ratio between HNA and LNA cells was
significantly higher (p<0.001) in spring than in summer
(Fig. 3c).
Proportions of dead cells (PI+) fluctuated from 19.3 to
57.0 % (mean=36.311 %) and did not vary significantly
between the sampled depths (Fig. 4a). Values were signifi-
cantly higher (p<0.05) in spring especially in the meta and
hypolimnion (Fig. 4b) with the peak and value recorded in
April for the hypolimnion and in June for both the meta and
hypolimnion (Fig. 4a).
Lytic Infection and Lysogeny
Lytic viral infection, as determined from the percentage of
viral infected prokaryotic cells (i.e., FIC) by TEM analyses
varied significantly with sampled months in the euphotic and
in the aphotic zones (Table 2). Overall, FIC values ranged
from 5.8 to 39.6 % (mean=19.37.8 %) in the euphotic
region and was significantly higher (p<0.001) than in the
aphotic hypolimnion (mean=10.36.7 %) (Table 1). Multiple
peaks in FIC were observed in the metalimnion of which the
maximum was observed in late August coinciding with a
decrease in PA and corresponded to a VIBM level of 66.5 %
(Figs. 2a and 5a). In the hypolimnion, FIC remained very low
with minimum variation. BS of individual cells ranged from
16 to 182 viruses per prokaryote with an overall mean of 41.
No significant difference in BS estimates was observed be-
tween sampled depths.
On many sampling occasions, FLC was below the detec-
tion limit compared to lytically infected cells. Lysogenic cells
induced by mitomycin C were observed on 64 % of the
sampling dates in the epilimnion (7 out of 11 sampling dates),
36 % of the sampling dates in the metalimnion and in the
hypolimnion (4 out of 11 sampling dates). When detected, the
FLC was less prevalent and significantly lower (p<0.001)
when compared to FIC, fluctuating from 0.03 to 11.0 %
(Table 1). FLC was significantly higher (p<0.01) in the epi-
limnion (mean=1.92.7 %) and metalimnion (mean=0.7
1.3 %) than in the hypolimnion where values never exceeded
1.0 % (mean=0.20.3 %). Lysogeny was very low in spring
and never exceeded 2.0 % whereas in summer, values in-
creased sharply in the epi and metalimnion reaching the
highest maximal values in early September (Fig. 5b).
Correlation Analyses
The results of Spearmans rank correlation analyses between
environmental variables and microbial parameters located in
746 S. Palesse et al.

Table 3 Spearman rank correlation (r) between variable in the euphotic (EPI + META) (n=22) and aphotic zone (HYPO) (n=11) of Lake Pavin

O2 Chl VA PA HNA LNA %PI+ FLC

Chl 0.69***
0.63*
VA 0.67** NS
NS NS
PA 0.55** 0.67*** 0.78**
NS NS 0.75**
HNA 0.56** 0.80*** 0.66*** 0.84***
NS NS 0.79** 0.81**
LNA 0.52* NS 0.75*** 0.85*** 0.48*
NS NS NS 0.95*** 0.66
%PI+ NS 0.46* NS 0.54** 0.59** NS
NS NS NS NS NS NS
FLC 0.60** 0.61** NS 0.48* 0.60** NS NS
NS NS NS NS NS NS 0.63*
FIC NS NS NS NS NS NS NS NS
0.63* NS NS NS NS NS 0.61* 0.82**

Values for euphotic zone (EPI + META) are indicated in boldface and in italics for aphotic zone (HYPO)
O2 Oxygen, Chl Chlorophyll a, VA viral abundance, PA prokaryotic abundance, HNA high nucleic acid cells abundance, LNA low nucleic acid cells
abundance, PI+ proportion of damaged prokaryotes, FLC frequency of lysogenically infected cells, FIC frequency of infected prokaryotic cells, NS not
significant
*p<0.05; **p<0.01; ***p<0.001, levels of significance

the euphotic zone (i.e., epilimnion and metalimnion) and in
the aphotic zone (i.e., hypolimnion) are summarized in
Table 3. VA was significantly correlated with PA both in the
euphotic (r=0.78, p<0.001) and aphotic zones (r=0.75,
p<0.01). In the euphotic region, chlorophyll a was signifi-
cantly correlated with PA (r=0.67, p<0.001), and this vari-
ability was largely explained by the HNA cells (r=0.80,
p<0.001,) suggesting that HNA cells were the dominant
fraction in phytoplankton-rich waters and perhaps may repre-
sent the more active fraction of the community. Furthermore,
FLC was negatively correlated with chlorophyll concentration
(r=0.61, p<0.01), PA (r=0.48, p<0.05) and more particu-
larly with the HNA cell abundance (r=0.60, p<0.01). In
contrast, FLC was positively correlated with the proportions
of cells with damaged membrane (r=0.63, p<0.05) in the
aphotic depth (hypolimnion) only.
Finally, the plot of field observations from PCA showed
that the distributions of abiotic and biotic variables under
study were dependent on the vertical gradients (Fig. 6). The
two major axes of this plot explained 63 % of the total
variance. Along the two major axes of this PCA plot, signif-
icant differences in the studied variables were observed be-
tween euphotic zone (epi and metalimnion) and the aphotic
zone (hypolimnion) suggesting that both represented two
distinct environments. The euphotic region was more dynamic
and showed significant variability in the measured parameters
between spring and summer, whereas the aphotic hypolimni-
on was more stable with sampled months.
Discussion
The present investigation is one of the few that explores how
environmental factors, host physiology, and their variability
affect the type of pattern between the lytic and lysogenic
replication cycles on a time series scale in a natural temperate
freshwater system. On the above basis, spatiotemporal inves-
tigations which were carried out in the mixolimnion of Lake
Pavin suggest that the dynamics of viral replication cycles
were largely controlled by the shifts of physiological states
and activity of the main prokaryotic hosts as forced by the
contrasted vertical gradient related to resource availability.
Host Physiological Changes
As proposed by del Giorgio and Gasol [32], we focused on the
variation of two categories of prokaryotic metabolic properties
which are indicative of their physiology and activity, namely
the nucleic acid content and membrane integrity of prokary-
otic cells. The most striking feature in the temporal variability
of standing stocks in the euphotic region was the predomi-
nance of HNA over LNA cells in the spring and vice versa in
Linking Host Prokaryotic Physiology to Viral Lifestyle Dynamics
1
Fig. 3 Spatiotemporal variability in the abundance of high nucleic acid
(a), low nucleic acid prokaryotes (b), and their mean ratio (HNA/LNA)
between spring and summer season (c) in the sampled depths of Lake
Pavin. Levels of significant differences: *p<0.05, ***p<0.001
the summer period. This suggests that the natural prokaryotic
communities underwent major shifts in their physiological
states during investigated time period. The fact that HNA cells
dominated in chlorophyll-rich waters (as seen by positive
significant correlation between them) could probably suggest
that HNA cells may perhaps represent the metabolically active
fraction of prokaryotic community. Such similar observations
have been previously reported in a set of lakes from the same
geographical region [17], and other studies have specifically
suggested that HNA cells apparently depend more on
phytoplankton-derived dissolved organic carbon for its
growth and metabolism [33, 34].
In Lake Pavin, the late summer period toward autumn (i.e.,
clear water phase) is characterized by lowest annual concen-
trations of organic and inorganic nutrients. This is due to its
peculiarities such as small surface, low catchment to lake area
ratio, and imbalance between sublacustrine spring water in-
puts and the surface water outputs which leads to an apparent
deficit of 20 l s1 [21]. Moreover, geomorphological features
of the lake basin and the absence of riverine and agricultural
747
1
Fig. 4 Spatiotemporal variations of the proportion of prokaryotes with
compromised membranes (PI cells) (a) and their comparison between
spring and summer period (b) in the sampled depths of Lake Pavin.
Levels of significant differences: *p<0.05; ***p<0.001
activities in the vicinity of the lakes further have been shown
to drive Lake Pavin to organic and inorganic nutrient limita-
tion in the late seasonal cycle [11]. Such strong nutrient
limitation can serve as a source of physiological stress for
the thriving prokaryotic communities. The fact that LNA cells
dominated during the period of low nutrient conditions sug-
gests that the prokaryote community would largely be repre-
sented by less active, dormant, or starved cells. The decline of
HNA cells in early summer is essentially attributable to the
resource limitation which is required to sustain growth and
activity. Therefore, the nutritional quality of available re-
sources could be an important factor controlling the physio-
logical state and metabolic activity of prokaryotes especially
in the euphotic region, and drastic changes in their availability
can promote nutritive stress among prokaryotic communities.
The dominance of LNA cells in resource limited conditions
and HNA cells in nutrient relaxed conditions has been already
shown in oceanic environments [35]. In contrast to the eupho-
tic, the magnitude of variation in HNA and LNA cells was low
in the hypolimnion with latter being the dominant fraction
throughout the study period. As aphotic depths are character-
ized by very low concentrations of chlorophyll (Fig 1c), pro-
karyotic community thriving in this region maintain a rela-
tively low metabolic activity and look more adapted to star-
vation stresses.
The finding that the proportion of prokaryotes with com-
promised membrane (PI cells) was significantly higher in the
spring suggests that the prokaryotic community which
748
1
2
Fig. 5 Spatiotemporal variability in the frequency of viral-infected pro-
karyotic cells (a) and frequency of lysogenic cells (b) in the sampled
depths of Lake Pavin. Error bars represent standard deviation (n=3)
comprised of dominant HNA cells were preferentially targeted
and selectively lysed by the viruses in the euphotic region of
the water column. From a methodological perspective, there
has been a recent concern in evaluating the resistance of
prokaryotic cell membranes to treatment with oxidants, with
cell membrane of HNA cells being more sensitive and were
damaged more quickly than LNA cells suggesting that the two
prokaryotic subgroups have a different chemical constitution
[36]. Thus, we can believe that some HNA cells may react
quickly to any physicochemical changes in Lake Pavin due to
possible lower membrane permeability. Proportions of cells
with damaged membrane represented overall a minor fraction
of the whole prokaryotic community (<50 %) in the
Fig. 6 Plot of field observations
from a principal component
analysis (PCA) for the three
sampled depths between each
season (1=spring, 2=
summer)
S. Palesse et al.
mixolimnion and did not differ significantly among the sam-
ple depths. In aquatic ecosystems, there is a large fraction of
prokaryotic cells that retain their cellular integrity but that
nevertheless does not have measurable metabolic activity
[32]. So, as explained above, the changes in prokaryotic cells
due to starvation do not obligatory imply an increase of cells
with damaged membrane.
Dynamics of Viral Life Strategies in Relation to Host
Physiological State
Lytic infection was generally more prevalent at the sampled
depths, while lysogeny was apparently more occasional and
mostly detected during the summer period in the euphotic
layers. To determine the percentage of lysogenic prokaryotes,
the mutagen mitomycin C was used in the present study
because to date, it is the most frequently used artificial agent
to induce prophages, and it allows comparison between a large
majority of aquatic studies [6, 10, 11, 14, 37]. In the euphotic
zone, the lysogenic strategy was strongly regulated by nutrient
environment as seen by a significant negative correlation of
chlorophyll concentration with lysogenic cycles. The increase
in the detection of lysogenic prokaryotes in the epi and
metalimnion during the summer period coincided with the
switch in the dominance of LNA over HNA cells in the clear
water phase period. This could probably be associated to
lower metabolic activities of prokaryotes such as energy dep-
rivation and starvation due to the prevailing resource limita-
tion. Such scenario has often been reported among aquatic
ecosystems thereby lending support to the hypothesis that
viruses favored the establishment of lysogeny in the less
productive or nutrient-limited environments[14, 38].
The most striking feature is that physicochemical condi-
tions did not directly affect the viral activities, but in turn, they
impacted the physiological states and growth of prokaryotic
Linking Host Prokaryotic Physiology to Viral Lifestyle Dynamics
hosts. The physiological state of the hosts has been suggested
to be a potential factor that can regulate the viral life strategies
in aquatic systems [10, 20, 39]. However, the accurate nature
of this link remains unclear. In the euphotic region of Lake
Pavin, FLC was not correlated with the proportion of cells
with damaged membrane (PI cells) whereas the two variables
were positively correlated in the hypolimnion. Study conduct-
ed in Northern Canadian lakes of differing trophic status, also
reported a similar positive relationship between FLC and
damaged cells especially in the less productive lake suggest-
ing that lysogeny may constitute a strategy of refuge for viral
communities until favorable cellular condition recover for
viral production [10]. A negative correlation of FLC with
HNA cell abundance and chlorophyll suggests that lysogenic
cycle was established mainly in prokaryotic cells that were
starved. Consequently, transition from high to low nutrient
environment promotes a fundamental change in cell physiol-
ogy, with a switch from growth to maintenance [40]. Such
shifts in HNA and LNA with corresponding increase in per-
centage of lysogeny have been reported in coastal lagoon [20].
Starvation response of prokaryotes involves profound
changes in their macromolecular composition, with the syn-
thesis of specialized protective proteins such as alarmone
(p)ppGpp, namely guanosine tetraphosphate [41]. This pro-
tein activates metabolic changes upon starvation which in turn
causes the cell to divert resources away from growth and
division to promote survival under nutrient-limited conditions
[42, 43]. Therefore, we assume that such stringent response
that activates signaling proteins such as (p)ppGpp in prokary-
otes could be linked to the regulatory mechanisms that pro-
mote the establishment of lysogeny. Studies on phage host
systems have reported that lysogenization is significantly
more effective when lamda () phage infects nutrient starved
prokaryotes [44, 45] and that the modest levels of (p)ppGpp
within the infected host cells favor lysogeny whereas absence
or higher levels of ppGpp might favor viral lysis [46].
Our study supports the hypothesis as previously proposed
by Maurice et al. [10, 20] that viral lifestyles are highly
influenced by the physiological states of prokaryotic hosts,
and lysogeny is favored by the changes in the physiology of
hosts which in turn depend on the environmental conditions.
More particularly, the source and magnitude of stress tend to
regulate viral decision. For prokaryotic community that lives
in an environment with constant fluctuations of nutrient con-
ditions, the main physiological stress promoting lysogeny
would be closely related to nutritive stress, and the stringent
response would be a signal for viruses to establish a lysogenic
cycle. In a more stable ecosystem such as the hypolimnion, the
lower levels of lysogeny might result from continuous pro-
phage induction events due to highest levels of starvation.
Alternatively, it could be possible that the prokaryotic com-
munity could be adapted to limited resources conditions, and
not affected by this nutritive stress, favoring the viral lysis
749
decision. Our results suggest that there might exist a physio-
logical threshold level which could be related to the degree of
nutritive stress of the prokaryotic hosts, and an accumulation
of such stress could be an induction signal for the viruses that
can lyse its host to abandon the sinking ship before the host
dies [4]. Furthermore, it would be essential to focus studies
more on the lines of physiology and activity at the single cell
level in accordance to broad range of physiological parameters
(i.e., enzyme activity), if we want to understand more on the
mechanisms that regulate of viral life styles in natural aquatic
systems.
Acknowledgments SP was supported by PhD fellowship from the
French Conseil Rgional dAuvergne through the project CPER (Contrat
de Projet Etat-Rgion) Axe: Environnement.
References
1. Breitbart M, Rohwer F (2005) Here a virus, there a virus, everywhere
the same virus? Trends Microbiol 13:278284. doi:10.1016/j.tim.
2005.04.003
2. Suttle CA (2007) Marine viruses-major players in the global ecosys-
tem. Nat Rev Microbiol 5:801812. doi:10.1038/nrmicro1750
3. Fuhrman JA (1999) Marine viruses and their biogeochemical and
ecological effects. Nature 399:541548. doi:10.1038/21119
4. Weinbauer MG (2004) Ecology of prokaryotic viruses. FEMS
Microbiol Rev 28:127181. doi:10.1016/j.femsre.2003.08.001
5. Sime-Ngando T, Colombet J (2009) Viruses and prophages in aquatic
ecosystems. Can J Microbiol 55:95109. doi:10.1139/w08-099
6. Weinbauer MG, Brettar I, Hfle MG (2003) Lysogeny and virus-
induced mortality of bacterioplankton in surface, deep, and anoxic
marine waters. Limnol Oceanogr 48:14571465. doi:10.4319/lo.
2003.48.4.1457
7. Paul JH (2008) Prophages in marine bacteria: dangerous molecular
time bombs or the key to survival in the seas? ISME J 2:579589.
doi:10.1038/ismej.2008.35
8. Wommack KE, Colwell RR (2000) Virioplankton: viruses in aquatic
ecosystems. Microbiol Mol Biol Rev 64:69114. doi:10.1128/mmbr.
64.1.69-114.2000
9. Chibani-Chennoufi S, Bruttin A, Dillmann M-L, Brssow H (2004)
Phage-host interaction: an ecological perspective. J Bacteriol 186:
36773686. doi:10.1128/jb.186.12.3677-3686.2004
10. Maurice CF, Bouvier T, Comte J, Guillemette F, del Giorgio PA
(2010) Seasonal variations of phage life strategies and bacterial
physiological states in three northern temperate lakes. Environ
Microbiol 12:628641. doi:10.1111/j.1462-2920.2009.02103.x
11. Pradeep Ram AS, Sime-Ngando T (2010) Resources drive trade-off
between viral lifestyles in the plankton: evidence from freshwater
microbial microcosms. Environ Microbiol 12:467479. doi:10.1111/
j.1462-2920.2009.02088.x
12. Thomas R, Berdjeb L, Sime-Ngando T, Jacquet S (2012) Viral
abundance, production, decay rates and life strategies (lysogeny
versus lysis) in Lake Bourget (France). Environ Microbiol 13:616
630. doi:10.1111/j.1462-2920.2010.02364.x
13. Danovaro R, Corinaldesi C, DellAnno A, Fuhrman JA, Middelburg
JJ, Noble RT, Suttle CA (2012) Marine viruses and global climate
change. FEMS Microbiol Rev 35:9931034. doi:10.1111/j.1574-
6976.2010.00258.x
750
14. Payet JP, Suttle CA (2013) To kill or not to kill: the balance between
lytic and lysogenic viral infection is driven by trophic status. Limnol
Oceanogr 58:465474. doi:10.4319/lo.2013.58.2.0465
15. Colombet J, Sime-Ngando T, Cauchie HM, Fonty G, Hoffmann L,
Demeure G (2006) Depth-related gradients of viral activity in Lake
Pavin. Appl Environ Microbiol 72:44404445. doi:10.1128/AEM.
00021-06
16. Evans C, Brussaard CPD (2012) Regional variation in lytic and
lysogenic viral infection in the Southern Ocean and its contribution
to biogeochemical cycling. Appl Environ Microbiol 78:67416748.
doi:10.1128/aem.01388-12
17. Pradeep Ram AS, Palesse S, Colombet J, Thouvenot A, Sime-
Ngando T (2014) The relative importance of viral lysis and
nanoflagellate grazing for prokaryote mortality in temperate lakes.
Freshw Biol 59:300311. doi:10.1111/fwb.12265
18. Bettarel Y, Arfi R, Bouvier T, Bouvy M, Briand E, Colombet J,
Corbin D, Sime-Ngando T (2008) Virioplankton distribution and
activity in a tropical eutrophicated bay. Estuar Coast Shelf Sci 80:
425429. doi:10.1016/j.ecss.2008.08.018
19. Boras JA, Sala MM, Vzquez-Domnguez E, Weinbauer MG, Vaque
D (2009) Annual changes of bacterial mortality due to viruses and
protists in an oligotrophic coastal environment (NW Mediterranean).
Environ Microbiol 11:11811193. doi:10.1111/j.1462-2920.2008.
01849.x
20. Maurice CF, Bouvier C, de Wit R, Bouvier T (2013) Linking the lytic
and lysogenic bacteriophage cycles to environmental conditions, host
physiology and their variability in coastal lagoons. Environ
Microbiol 15:24632475. doi:10.1111/1462-2920.12120
21. Aeschbach-Hertig W, Hofer M, Schmid M, Kipfer R, Imboden DM
(2002) The physical structure and dynamics of a deep, meromictic
crater lake (Lac Pavin, France). Hydrobiologia 487:111136. doi:10.
1023/A:1022942226198
22. Wetzel RG, Likens GE (1995) Limnological analysis, 2nd edn.
Springer, New York
23. Marie D, Brussaard CPD, Thyrhaug R, Bratbak G, Vaulot D (1999)
Enumeration of marine viruses in culture and natural samples by flow
cytometry. Appl Environ Microbiol 65:4552
24. Brussaard CPD, Marie D, Bratbak G (2000) Flow cytometric detec-
tion of viruses. J Virol Methods 85:175182. doi:10.1016/s0166-
0934(99)00167-6
25. Brussaard CPD, Payet J, Winter C, Weinbauer MG (2010)
Quantification of aquatic viruses by flow cytometry. In: Wilhelm
SW, Weinbauer MG, Suttle CA (eds) Manual of aquatic viral ecol-
ogy, vol 11, ASLO, Texas, Chapter., pp 102109
26. Bouvier T, del Giorgio PA, Gasol JM (2007) A comparative study of
the cytometric characteristics of High and Low nucleic-acid
bacterioplankton cells from different aquatic ecosystems. Environ
Microbiol 9:20502066. doi:10.1111/j.1462-2920.2007.01321.x
27. Gasol JM, Zweifel UL, Peters F, Fuhrman JA, Hagstrm (1999)
Significance of size and nucleic acid content heterogeneity as mea-
sured by flow cytometry in natural planktonic bacteria. Appl Environ
Microbiol 65:44754483
28. Sime-Ngando T, Mignot J, Amblard C, Bourdier G, Devilettes C,
Quilblier-Lloberas C (1996) Caractrisation des particules virales
planctoniques dans un lac du Massif Central Franais: Aspects
mthodologiques et premiers rsultats. Int J Limnol 32:259263.
doi:10.1051/limn/1996025
29. Weinbauer MG, Winter C, Hfle MG (2002) Reconsidering trans-
mission electron microscopy based estimates of viral infection of
S. Palesse et al.
bacterio-plankton using conversion factors derived from natural com-
munities. Aquat Microb Ecol 27:103110. doi:10.3354/ame027103
30. Binder B (1999) Reconsidering the relationship between virally
induced bacterial mortality and frequency of infected cells. Aquat
Microb Ecol 18:207215. doi:10.3354/ame018207
31. Paul JH, Jiang SC (2001) Lysogeny and transduction. In: Paul JH
(ed) Methods in microbiology, vol 30. Elsevier Academic Press Inc,
San Diego, pp 105125
32. del Giorgio PA, Gasol JM (2008) Physiological structure and single-
cell activity in marine bacterioplankton. Microbial Ecology of the
Oceans. Wiley, New Jersy, pp 243298
33. Moran XAG, Bode A, Suarez LA, Nogueira E (2007) Assessing the
relevance of nucleic acid content as an indicator of marine bacterial
activity. Aquat Microb Ecol 46:141152. doi:10.3354/ame046141
34. Cuevas LA, Egge J, Thingstad TF, Tpper B (2011) Organic carbon
and mineral nutrient limitation of oxygen consumption, bacterial
growth and efficiency in the Norwegian Sea. Polar Biol 34:871
882. doi:10.1007/s00300-010-0944-3
35. Gasol JM, Simons AM, Kalff J (1995) Patterns in the top-down
versus bottom-up regulation of heterotrophic nanoflagellates in tem-
perate lakes. J Plankton Res 17:18791903. doi:10.1093/plankt/17.
10.1879
36. Ramseier MK, von Gunten U, Freihofer P, Hammes F (2011)
Kinetics of membrane damage to high (HNA) and low (LNA)
nucleic acid bacterial clusters in drinking water by ozone, chlorine,
chlorine dioxide, monochloramine, ferrate(VI), and permanganate.
Water Res 45:14901500. doi:10.1016/j.watres.2010.11.016
37. Tapper MA, Hicks RE (1998) Temperate viruses and lysogeny in
Lake Superior bacterioplankton. Limnol Ocean 43:95103. doi:10.
4319/lo.1998.43.1.0095
38. Clasen JL, Brigden SM, Payet JP, Suttle CA (2008) Evidence that
viral abundance across oceans and lakes is driven by different bio-
logical factors. Freshw Biol 53:10901100. doi:10.1111/j.1365-
2427.2008.01992.x
39. Williamson SJ, Houchin LA, McDaniel L, Paul JH (2002) Seasonal
variation in lysogeny as depicted by prophage induction in Tampa
Bay, Florida. Appl Environ Microbiol 68:43074314. doi:10.1128/
AEM.68.9.4307-4314.2002
40. Sigee DC (2005) Freshwater environments: the influence of physico-
chemical conditions on microbial communities. In: Sigee DC (ed)
Freshwater microbiology, vol 2, Wiley, Chichester, England,
Chapter., pp 47103
41. Wu J, Xie J (2009) Magic spot: (p) ppGpp. J Cell Physiol 220:297
302. doi:10.1002/jcp.21797
42. Kriel A, Bittner AN, Kim SH et al (2012) Direct regulation of GTP
homeostasis by (p)ppGpp: a critical component of viability and stress
resistance. Mol Cell 48:231241. doi:10.1016/j.molcel.2012.08.009
43. Boutte CC, Crosson S (2013) Bacterial lifestyle shapes stringent
response activation. Trends Microbiol 21:174180. doi:10.1016/j.
tim.2013.01.002
44. Wegrzyn G, Licznerska K, Wegrzyn A (2012) Phage lambdanew
insights into regulatory circuits. Adv Virus Res 82:155178. doi:10.
1016/b978-0-12-394621-8.00016-9
45. Wgrzyn G, Wgrzyn A (2005) Genetic switches during bacterio-
phage development. In: Kivie M (ed) Progress in nucleic acid
research and molecular biology. Academic Press, New York, pp 148
46. Potrykus K, Wegrzyn G, Hernandez VJ (2002) Multiple mechanisms
of transcription inhibition by ppGpp at the lambda rho(R) promoter. J
Biol Chem 277:4378543791. doi:10.1074/jbc.M208768200