ASTROBIOLOGY

Volume 17, Number 4, 2017

ª Mary Ann Liebert, Inc.
DOI: 10.1089/ast.2016.1580

Liquid Water Restricts Habitability in Extreme Deserts

Charles S. Cockell, Sarah Brown, Hanna Landenmark, Toby Samuels,

Rebecca Siddall, and Jennifer Wadsworth

Abstract

Liquid water is a requirement for biochemistry, yet under some circumstances it is deleterious to life. Here,
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.

we show that liquid water reduces the upper temperature survival limit for two extremophilic photosynthetic
microorganisms (Gloeocapsa and Chroococcidiopsis spp.) by greater than 40C under hydrated conditions
compared to desiccated conditions. Under hydrated conditions, thermal stress causes protein inactivation as
shown by the fluorescein diacetate assay. The presence of water was also found to enhance the deleterious
effects of freeze-thaw in Chroococcidiopsis sp. In the presence of water, short-wavelength UV radiation
more effectively kills Gloeocapsa sp. colonies, which we hypothesize is caused by factors including the
greater penetration of UV radiation into hydrated colonies compared to desiccated colonies. The data predict
that deserts where maximum thermal stress or irradiation occurs in conjunction with the presence of liquid
water may be less habitable to some organisms than more extreme arid deserts where organisms can
dehydrate prior to being exposed to these extremes, thus minimizing thermal and radiation damage. Life in
extreme deserts is poised between the deleterious effects of the presence and the lack of liquid water. Key
Words: Deserts—Extremophiles—Stress—High temperatures—UV radiation—Desiccation. Astrobiology
17, 309–318.

1. Introduction Ultraviolet radiation also has deleterious effects on cya-

nobacteria, particularly in extreme deserts where during the

E xtreme deserts are characterized by sporadic avail-

ability of liquid water and often extreme thermal fluc-
tuations. In deserts near the equator, intense solar irradiation
is an additional stress. Among the denizens of extreme des-
erts, the cyanobacteria have been some of the best charac-
terized with respect to extreme physical stresses such as high
temperatures and UV irradiation. These organisms are found
to inhabit the interior and underside of rocks and the interior
of desert crusts (Berner and Evenari, 1978; Friedmann, 1980;
Broady, 1981; Nienow et al., 1988; Belnap and Lange, 2001;
Douglas, 2004), and many of these organisms exhibit extreme
desiccation tolerance (Potts, 1999; Büdel, 2011; Lüttge, 2011;
Rajeev et al., 2013). Insofar as they are responsible for primary
production in desert ecosystems, their response to stress is an
important determinant of the resilience of desert ecosystems to
extreme conditions.
High temperature is known to cause protein aggregation
and denaturation in cyanobacteria (Inouse et al., 2001). Al-
though they have a range of responses when metabolically
active to cope with heat stress, such as the production of heat
shock proteins (Suzuki et al., 2006), the upper temperature
limit for most cyanobacteria in a hydrated state is around
50C (Casteilli et al., 2009).
day organisms accumulate UV radiation damage. Using the
desert-dwelling cyanobacterium, Chroococcidiopsis, Cock-
ell et al. (2008) showed that these cells were killed within a
day by unattenuated surface fluxes of UV radiation. These
data demonstrate the importance of habitats that attenuate
UV radiation, such as the interior and subsurface of rocks.
UV radiation, particularly short-wavelength UVB (280–
320 nm) radiation, directly damages proteins and nucleic
acids (Quesada and Vincent, 1997) and causes indirect
damage through the production of reactive oxygen species
(Schulz and Scherer, 1999). The photosynthetic apparatus of
cyanobacteria is bleached and destroyed by UV radiation
(Castenholz and Garcia-Pichel, 2000).
It is generally assumed that, in deserts where organisms
are exposed to physical extremes such as extremes of tem-
peratures and UV radiation, the presence of liquid water is
beneficial to life (e.g., Lange et al., 1990; Pointing et al.,
2006). Insofar as metabolism (such as repair processes) and
the growth and replication of organisms depend on the
presence of liquid water, water must be available, even if
only transiently, for life-forms to subsist.
However, it is known that, at the biochemical level, water
is not always beneficial to life. Hydrolysis reactions are

UK Centre for Astrobiology, School of Physics and Astronomy, University of Edinburgh, Edinburgh, UK.

309
 310
testament to the potentially deleterious role that this solvent
can play at the molecular level (Ball, 2007). In this paper, we
show how liquid water can also be deleterious to life at the
organismal and ecological level, and we show that water
defines a window of habitability in extreme deserts that is
caused by a trade-off between the presence of liquid water for
beneficial biochemical reactions and the role of liquid water
in exacerbating cellular temperature and UV radiation–induced
damage. We discuss predictions that arise from these results.
2. Methods
2.1. Organisms
Two extremophilic cyanobacteria were used for the in-
vestigations.
Chroococcidiopsis strain 029 (isolated from cryptoendo-
lithic growth in Nubian sandstone in the Negev Desert): This
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
Gram-negative non-spore-forming cyanobacterium, which is
adapted to life in hot deserts, has been studied previously in the
laboratory in a variety of contexts and belongs to a genus
reported in diverse hot and cold deserts around the world
(Friedmann, 1980; Grilli Caiola et al., 1996; Billi, 2009; Billi
et al., 2010; Cockell et al., 2011; Baqué et al., 2013). The cells
do not form large colonies, but cells sometimes exist as mul-
tiple cell aggregates containing 3–6 cells, including tetrads.
Gloeocapsa sp.: This Gram-negative non-spore-forming
cyanobacterium was isolated in our laboratory by exposing
rocks obtained from cliff faces in Devon, UK, to conditions
in low Earth orbit on the outside of the Biopan orbital sat-
ellite and the International Space Station (Olsson-Francis
et al., 2010; Cockell et al., 2011). The Gloeocapsa species
studied here has a characteristic colonial growth habit
whereby cells multiply into an amorphous mass held to-
gether by polysaccharide containing several hundred to
thousands of cells. Colonies achieve sizes of several milli-
meters after 3–4 months’ growth in liquid media or on the
surface of agar plates. In some experiments, these native
colonial forms were studied. In other experiments, these
colonies were gently broken up into individual cells and cell
clumps by homogenization in a 7 mL borosilicate homoge-
nizer (Fisher Scientific, UK), taking care not to rupture the
cells. These cells are referred to as ‘‘separated cells.’’ Se-
paration was confirmed under bright field microscopy as
described for microscopy below.
Both organisms were cultivated in BG (Blue-Green)-11
liquid medium (Rippka et al., 1979) to prepare cells for the
experiments. For the enumeration of cell viability, organ-
isms were grown on BG-11 agar plates containing 2% w/v
Bacteriological agar No 1 (Oxoid, UK). All cultures and
plates were kept under a light intensity of 50 mmol/m2/s at a
temperature of 21C for 2 months, unless otherwise stated.
The organisms were grown under continuous light. This is
observed to allow for good growth (unpublished data), and it
is observed to be satisfactory for the growth of other cya-
nobacteria (e.g., Moore et al., 2007).
For experiments in which desiccated cells were used,
cultures of cells were aliquoted into Eppendorf tubes and
placed into a desiccation jar containing silica gel. Cells were
desiccated for 4 days until water within the tubes had been
removed and cells formed a dried pellet on the bottom of the
tube. We confirmed the desiccated state by observing that
there was no residual water associated with the cells under
COCKELL ET AL.
bright field microscopy (see below) and that the cells had
lost turgor associated with the hydrated state.
2.2. Microscopy
In some experiments, cells were observed by microscopy.
All cells were observed with a Leica DM4000B fluores-
cence microscope (Leica Microsystems, Germany). Images
were acquired with a Leica DFC450C camera. Auto-
fluorescence of cyanobacterial pigments was observed under
a rhodamine filter set (510–560 nm, emission 590 nm).
In some experiments, metabolic activity was assessed by
examining the activity of esterases. Esterases are a group of
intracellular enzymes present in all types of cells that are
involved in cell membrane formation. Esterase activity was
assessed by the ability of the cells to intracellularly cleave
fluorescein diacetate (FDA) to the fluorescent product,
fluorescein (e.g., Battin, 1997; Regel et al., 2002). FDA
(Sigma Chemicals, UK) was dissolved in acetone at a
concentration of 5 mg/mL and then added to cell suspen-
sions to a final working concentration of 5 mg/mL. Cells
were incubated at room temperature in the dark for 15 min
before examination under the microscope. For FDA visu-
alization, cells were excited at 485 nm with a bandpass of
22 nm and observed at the emission wavelength of 530 nm
(bandpass 30 nm). Esterase activity is indicated by cells
glowing a bright green color. Inactive cells displayed a red
coloration. In cases where cell metabolic viability was
quantified, cells were scored as ‘‘fluorescing’’ when they
showed a bright green coloration. Esterase activity was
expressed as the percentage of ‘‘enzyme inactive’’ cells (red
cells) in the observed population.
LIVE/DEAD stain was used according to the manufactur-
er’s instructions (Invitrogen, UK). The kit was used with a 1:1
ratio of SYTO 9 and propidium iodide. Cells were observed
under fluorescence microscopy as for FDA analysis (above).
2.3. Experiments
2.3.1. Thermal stress. To investigate the effects of
thermal stress on cell viability, Chroococcidiopsis sp. and
Gloeocapsa sp. (native colonies) were aliquoted into Ep-
pendorf tubes. Cells were either maintained in BG-11
medium or desiccated. Tubes of cells were placed into a
Grant GBD2 heat block (Grant Instruments, UK) at defined
temperatures. Following thermal stress, viable cells were
enumerated on BG-11 plates, and the loss of viability was
determined by expressing the cell viability as a ratio of the
number of viable thermally stressed cells to the control
(nonthermally stressed, either hydrated or desiccated) cells
calculated as a percentage.
In Experiment 1, cells were exposed to temperatures
from 40C to 90C for 15 min to evaluate the effects of
temperature on hydrated and desiccated Chroococcidiopsis
sp. (at 5C increments) and Gloeocapsa sp. (at 40C, 60C,
and 80C) cells. In this set of experiments, the colonial
growth form of Gloeocapsa sp. was used. We examined
fewer temperature points, as our objective was to deter-
mine whether, like Chroococcidiopsis, hydration resulted
in a lower temperature threshold for thermal stress.
In Experiment 2, the effects of thermal stress in Gloeo-
capsa sp. were further investigated under hydrated condi-
tions to determine with greater resolution the temperature at
 LIQUID WATER AND HABITABILITY IN EXTREME DESERTS
which thermal inactivation of cells occurs. In this experi-
ment, the viability of separated, hydrated cultures was tested
after heat shock treatment at varying temperatures of 2C
increments between 40C and 50C, for varying time in-
crements of 2, 5, 10, 15, and 30 min.
To determine how thermal stress can affect enzymatic
activity, the esterase assay was employed. Quantification of
the number of metabolically inactive cells was carried out
after thermal stress at 40C, 60C, and 80C for both
Chroococcidiopsis and Gloeocapsa sp. after different time
periods of thermal stress (5, 10, 15, and 30 min).
To determine whether loss of cell membrane integrity
might account for loss of viability under thermal stress in
hydrated cells, LIVE/DEAD stain was employed. Cells
stained with LIVE/DEAD were evaluated after 15 min
thermal stress at 40C, 45C, 50C, 55C, 60C, and 70C.
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
2.3.2. Ultraviolet radiation. To investigate the effects of
desiccation and liquid water availability on UV radiation
damage, colonies and separated cells of Gloeocapsa
sp. were placed into 4 cm diameter glass dishes and exposed
to short-wavelength (254 nm) UV radiation by using a 3 W
UV lamp (UV Products, California, USA) at a distance of
4 cm. UV radiation of wavelength 254 nm is not found in the
natural environment, but in this experiment we use it as a
proxy for general damage caused by natural short wavelengths
of UV radiation. Colonies were either submerged in a *2 mm
deep layer of BG-11 for hydrated cells or dried overnight on
the bottom of the glass disc for desiccation experiments. Fol-
lowing UV exposure, the desiccated cells were rehydrated.
Colonies and separated cells were either plated directly onto
BG-11 agar plates as described above or examined for es-
terase activity under the microscope as described above.
2.3.3. Freeze-thaw stress. To determine whether des-
iccation prior to freezing provides an advantage to cells,
aliquots of Chroococcidiopsis sp. were prepared in Eppen-
dorf tubes. One set comprised 250 mL hydrated cells, and
one set of desiccated cells was prepared by desiccating
250 mL of the same culture for 4 days in a desiccator. These
aliquots were stored at -20C, and every 24 h they were
warmed to 21C under laboratory conditions prior to being
refrozen. After each freeze-thaw cycle, a triplicate of both
hydrated and desiccated cells was removed and plated.
After 10 freeze-thaw cycles, aliquots of hydrated and
desiccated cells were exposed to 15 min thermal stress at
40C, 45C, and 50C to determine whether freeze-thaw
cycles increased susceptibility to thermal stress.
2.3.4. Long-term (159 months) desiccation and thermal
stress. A 1-month-old culture of Chroococcidiopsis sp.,
grown under the conditions described in Section 2.1, was
plated into BG-11 agar to form a lawn of cells. The plates
were incubated as described above for enumeration. Under
aseptic conditions, the plate lids were removed and the cells
allowed to desiccate over a period of 2 days. The lid was
replaced on the plates containing the desiccated sheet of
agar and cells, and the sheets were stored in darkness under
laboratory conditions for 159 months (February 2003 to
May 2016) with a temperature of between 15C and 21C
and a relative humidity of between 20% and 60%.
311
After 159 months of desiccation, the status of the cells
was examined on BG-11 to determine whether any of the
cells remained viable. They were also examined for the
presence of green pigmentation, autofluorescence, and es-
terase activity (as above). To study the effects of thermal
stress after long-term desiccation, segments of the desic-
cated sheets were rehydrated, and cells were exposed to
15 min of thermal stress at 40C, 45C, and 50C. Samples
of desiccated sheet were also exposed to 70C, 90C, and
100C to determine whether cells could survive long-term
desiccation and transient extreme thermal stress.
2.4. Statistical methods
In all experiments, triplicates of each condition were
examined. Student’s t test was used to determine significant
differences between given treatments with a p value of <0.1,
which is considered to be statistically significant.
3. Results
3.1. Thermal stress experiments
The effects of thermal stress on Chroococcidiopsis
sp. survivability at different temperatures in a desiccated and
hydrated state is shown in Fig. 1. Desiccated cells of
Chroococcidiopsis sp. showed survival to 90C, but hydrated
cells were completely killed at 60C. In the desiccated state,
between 45C and 55C, the number of colony-forming units
showed an increase in the number of viable cells.
The effects of thermal stress on Gloeocapsa sp. surviv-
ability at different temperatures in a desiccated and hydrated
state are shown in Fig. 2. Gloeocapsa sp. cell survival under
desiccated conditions was observed up to 80C, but no hy-
drated cells survived at 60C.
In Fig. 3, the results of experiments to investigate the
thermal stress of Gloeocapsa sp. at 2C increments between
FIG. 1. The effects of thermal stress on Chroococci-
diopsis viability at different temperatures in a desiccated
and hydrated state. Error bars are the standard deviation of
triplicate measurements. Viability is expressed as follows:
Number of viable cells after the given treatment (N)/number
of viable cells in the control (No) · 100%.
 312
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
FIG. 2. The effects of thermal stress on Gloeocapsa sp. on
survivability at different temperatures in a desiccated and
hydrated state. Error bars are the standard deviation of
triplicate measurements. Viability is expressed as follows:
Number of viable cells after the given treatment (N)/number
of viable cells in the control (No) · 100%.
40C and 50C are shown. After 30 min thermal stress at
40C and 42C, there was no significant difference in via-
bility compared to an untreated control (Student’s t test;
p = 0.98). At 44C, there was no significant difference to an
untreated control after 10 min of thermal stress (t test,
FIG. 3. The effects of thermal stress on hydrated Gloeo-
capsa sp. at 2C increments between 40C and 50C. Error
bars are the standard deviation of triplicate measurements.
Viability is expressed as follows: Number of viable cells
after the given treatment (N)/number of viable cells in the
control (No) · 100%.
COCKELL ET AL.
p = 0.47), but viability dropped significantly for 15 min
( p < 0.1) and 30 min stress ( p < 0.1). At 46C, significant
loss of viability was observed after 10 min and for longer
times thereafter ( p < 0.1 for 10 min and all time points
thereafter). At both 48C and 50C, a significant drop in
viability compared to the untreated control was observed
after 5 min stress ( p < 0.1). Some cells remained viable after
30 min stress at 50C.
In Fig. 4, the effects of thermal stress on the esterase ac-
tivity of Chroococcidiopsis sp. at 40C, 60C, and 80C are
shown. At 80C, after all time points, there was a large re-
duction in measurable enzyme-active cells at 5 min and
greater periods of thermal stress, which was statistically
significant (t test, p < 0.1). The control was the percentage of
enzyme-inactive cells in unstressed cells. Figure 5 shows the
corresponding data for Gloeocapsa sp. Thermal stress at 60C
and 80C caused a large reduction in measurable enzyme-
active cells after 5 min and for all longer time periods of
thermal stress, which was statistically significant ( p < 0.1). In
both Gloeocapsa sp. at 40C and Chroococcidiopsis sp. at
40C and 60C, there was a statistically significant (t test,
p < 0.1) reduction in the number of enzyme-inactive cells at
30 min. One possible explanation for this latter observation is
that thermal treatment can enhance the uptake of FDA into a
small proportion of cells that show inactivity not because of
inactive enzymes but because of insufficient uptake of FDA.
This might occur by partially solubilizing or weakening the
extracellular polysaccharide layer.
LIVE/DEAD stain was employed to investigate the
effects of thermal stress on cell membrane integrity. The
percentage of cells showing viability according to the test
was not statistically different in cells exposed to 15 min
thermal stress at 40C (89.7 – 4.0% compared to the con-
trols, which had 90.5 – 5.3% viability), 45C (89.6 – 5.2%),
and 50C (91.7 – 3.0%) (Student’s t test, p > 0.1). However,
at 55C (81.3 – 9.2% viability) there was just significantly
less viability [p < 0.1 (0.09)], and at 60C (29.0 – 17.6%)
FIG. 4. The effects of thermal stress on the esterase activity
of hydrated Chroococcidiopsis sp. at 40C, 60C, and 80C.
Error bars are the standard deviation of triplicate measure-
ments. Number of viable cells after the given treatment (N)/
number of viable cells in the control (No) · 100%.
 LIQUID WATER AND HABITABILITY IN EXTREME DESERTS
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
FIG. 5. The effects of thermal stress on the esterase activity
of hydrated Gloeocapsa sp. at 40C, 60C, and 80C. Error
bars are the standard deviation of triplicate measurements.
Number of viable cells after the given treatment (N)/number
of viable cells in the control (No) · 100%.
there was also significantly less viability compared to con-
trols ( p < 0.05). No cells showed viability according to the
test at 70C.
3.2. UV radiation
In Fig. 6, the effects of UV radiation on the viability of
Gloeocapsa sp. are shown for separated cells (isolated single
cells and small cell clumps) of the organism and colonial
forms of the organisms under desiccated and hydrated con-
ditions. At 5 min and higher exposure times, there is a sta-
tistically significant difference between all treatments (t test,
FIG. 6. The effects of UV radiation
(254 nm) on the viability of Gloeocapsa
sp. for separated (isolated single cells here
shown as homogenized colonies) organisms
and colonial forms under desiccated and
hydrated conditions. Letters above columns
cluster columns with no statistically signifi-
cant difference ( p > 0.1) for each time point.
Error bars are the standard deviation of
triplicate measurements. Number of viable
cells after the given treatment (N)/number of
viable cells in the control (No) · 100%.
313
all values p < 0.1). Hydrated single cells showed higher UV
tolerance than desiccated single cells. Desiccated colonies of
cells had the greatest UV tolerance of all the growth forms
and conditions. After 2 h exposure, only colonial forms sur-
vived, with desiccated colonies showing a significantly higher
survival than hydrated colonies (t test, p < 0.1).
Figure 7 shows representative images of the organism in
colonial growth form taken under epifluorescence micros-
copy following FDA treatment. UV radiation has the effect
of killing the cells on the outer layer of the colonies, but
cells interior to the colony show enzymatic activity after 2 h.
By contrast, colonies in which cells had been separated to
expose interior cells were all killed after 2 h.
3.3. Freeze-thaw stress
Freeze-thaw stress was deleterious to both desiccated and
hydrated Chroococcidiopsis sp. cells after 10 cycles (Fig. 8);
however, hydrated cells had a much greater loss of viability.
The mean viability of hydrated cells after 10 freeze-thaw
cycles was 0.29 – 0.17% of hydrated controls not exposed to
freeze-thaw cycles. The mean viability of desiccated cells
after 10 freeze-thaw cycles was 54.5 – 29.7% of desiccated
controls. Both of these losses of viability were statistically
significantly different from that of the controls ( p < 0.1). In
the first three freeze-thaw cycles, the number of estimated
viable cells in desiccated samples was greater than that of
the controls, which we attribute to disruption of tetrads and
multicellular growth aggregates.
After 10 freeze-thaw cycles, the thermal stress response
of the cells after 15 min at 40C, 45C, and 50C was
qualitatively similar to that of cells unexposed to freeze-
thaw (Fig. 1). The viability of hydrated cells, as indicated
by Student’s t test, that were subsequently exposed to stress
at 40C (0.26 – 0.10%) and 45C (0.34 – 0.10%) after 10
freeze-thaw cycles was not statistically lower than the via-
bility of cells that underwent 10 freeze-thaw cycles but
 314 COCKELL ET AL.
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.

FIG. 7. The effects of UV radiation on desiccated Gloeocapsa sp. cells. (A) Epifluorescent image of control colonies (white
arrow as an example) stained with FDA and exhibiting fluorescence on account of fluorescein production by esterase enzymatic
activity in the cells. (B) Control dead colony FDA stained and exhibiting no green fluorescence. Red fluorescence is caused by
chlorophylls. (C) Bright field image of colonies exposed for 2 h to UV radiation (254 nm). (D) Epifluorescence image of same
colonies as in (C) showing unaffected interior of colonies and red UV radiation–exposed surfaces of colonies (white box) where
cells show no enzymatic activity. (E) Magnified image of a colony showing enzymatically inactive surface (red) exposed to UV
radiation and underside with enzymatically active (green) cells. (F) Separated cells of Gloeocapsa cells after exposure to UV
radiation for 2 h, showing complete loss of enzymatic activity. Scale bars in (A) to (D), 100 mm; scale bars in (E) and (F), 20 mm.
The images are representative for three independent experiments.

were not exposed to heat stress. However, cells exposed to
15 min of 50C stress had significantly lower viability
(0.032 – 0.05%, p < 0.1). Desiccated cells exposed to 15 min
at 40C (6.16 – 1.21%), 45C (3.92 – 2.74%), and 50C
(5.29 – 1.66%) after 10 cycles of freeze-thaw showed no
significant loss of viability compared to cells that underwent
10 freeze-thaw cycles but were not exposed to heat stress
(Student’s t test, p < 0.1).
3.4. Long-term desiccation and thermal stress
Cells retained viability after 13 years and 3 months’ des-
iccation (Fig. 9). After this time, cells were visualized as cell
masses retaining green pigmentation and autofluorescence
caused by phycobiliproteins (Fig. 9A–9C). Some cells also
demonstrated esterase activity (Fig. 9D). After exposure to
thermal stress at 40C, cells showed no statistically significant
 LIQUID WATER AND HABITABILITY IN EXTREME DESERTS
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
FIG. 8. Effects of successive freeze-thaw cycles on the
loss of viability of desiccated and hydrated Chroococci-
diopsis sp. compared to hydrated and desiccated controls
(cells prior to the first freeze-thaw cycle). Number of viable
cells after the given treatment (N)/number of viable cells in
the control (No) · 100%.
loss of viability compared to controls (Student’s t test), but
cells exposed to 45C and 50C showed a significant (Stu-
dent’s t test, p < 0.1) loss of viability compared to controls
(1.22 – 2.11% and 2.1 – 3.00% of controls, respectively),
showing that long-term desiccation enhanced sensitivity of
cells to thermal stress at 45C compared to a fresh hydrated
culture, which normally showed no statistically significant
loss of viability at 45C compared to controls.
Long-term desiccated cells were successfully grown after
exposure to 15 min at 70C and 90C but with viability less than
0.1% of controls. No cells survived 100C (data not shown).
4. Discussion
Extreme hot and cold deserts are known to be unin-
habitable to many microorganisms owing to extreme des-
iccation and, in equatorial deserts, accumulated solar (UV)
radiation damage in surface environments (Nienow et al.,
1988). Extreme temperature fluctuations impose additional
stresses on desert organisms (Friedmann, 1980). The re-
quirement for liquid water to allow growth and biochemical
repair pathways (Mottl et al., 2007) leads to the paradigm
that liquid water is beneficial to life and that deserts where
liquid water is more available are therefore generally more
habitable (Warren-Rhodes et al., 2006).
In this work, we show how liquid water can be deleterious
in extreme deserts by increasing the damaging effects of
certain physical extremes. Liquid water, therefore, defines a
window of habitability in extreme environments where its
beneficial uses in growth and reproduction are balanced
against the advantages of its absence when those same or-
ganisms suffer exposure to extreme conditions, the damaging
biochemical effects of which are exacerbated by liquid water.
The presence of liquid water increased the mortality of
two different extremophilic cyanobacteria under high tem-
315
perature stress. We found that the threshold at which cells
were completely killed was increased by *40C when cells
were desiccated. Under desiccated conditions, thermal en-
ergy will be less effectively transferred to biomolecules
through their kinetic energy and thereby less likely to cause
disruption of essential molecular activity. However, at 50C
and 70C even desiccated Chroococcidiopsis sp. showed
some loss of viability after several hours of thermal stress.
Under hydrated conditions, it was found that the thermal
range within which loss of viability occurs was very narrow.
For example, after 30 min at 42C, Gloeocapsa sp. cells
showed no appreciable loss of viability. However, viability
began to be lost after the same time period in cells exposed
to 44C, whereas those at 48C and 50C had lost more than
90% viability after 5 min. These data are consistent with the
thermal limit of plant tissues. In Ranunculus glacialis, ir-
reversible damage was done to the plant’s chloroplasts at
45–46C. This damage was attributed to the thermal damage
of membranes manifested as the swelling of thylakoid
membranes and general changes in organelle ultrastructure
(Larcher et al., 1997). The mediation of heat shock through
changes in membrane integrity has been observed in other
organisms including microorganisms (Cossins, 1994; Vigh
et al., 1998), and the response to heat shock can involve
modifications to membrane composition (Balogi et al.,
2005). However, our data, acquired by using LIVE/DEAD
stain as a proxy for membrane integrity, show that cells
apparently remain viable according to this test even after
15 min thermal stress at 60C, after which no colonies can
be cultured on BG-11 plates. This shows that cell membrane
integrity remains even after cell viability is lost, which sug-
gests that the thermal cause of loss of viability in Chroo-
coccidiopsis sp. is not gross disruption of membrane integrity
but some other component of the cell. The data we obtained
suggest that LIVE/DEAD stain, at least in some cyano-
bacteria, may not be a reliable indicator of cell viability.
Investigations in which FDA was used to examine enzy-
matic activity of thermally stressed hydrated cells are consis-
tent with the idea that denaturation of proteins can play a role
in loss of viability (Inouse et al., 2001). Under hydrated con-
ditions, within 5 min at 60C, enzymatic activity was com-
pletely lost in Gloeocapsa sp. At 80C, similar results were
observed in Chroococcidiopsis sp.; however, esterase activity
was still observed at 60C, which suggests that not only might
there be differences in susceptibility of enzymatic activities
in different species of cyanobacteria, but also the loss of via-
bility can occur within the cells before all enzymatic activity
is completely lost. This is consistent with observations by
Cockell et al. (2005), who studied Chroococcidiopsis under a
simulated martian UV flux and found that cells lost viability
while retaining esterase activity. The cause of cell death in
cyanobacteria during thermal stress may involve oxidative
stress, protein denaturation, and loss of membrane integrity
(Cossins, 1994; Vigh et al., 1998; Balogi et al., 2005); how-
ever, the narrow temperature range in which cell viability is
lost suggests that critical degradation of a few specific mole-
cules is responsible for determining the threshold of thermal
stress. The lack of this effect in desiccated cells could be ac-
counted for by the more rigid nature of molecules that are less
easily denatured and the lack of a milieu to generate and move
around reactive oxygen species or to mediate fluid deteriora-
tion in membrane integrity.
 316 COCKELL ET AL.
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.

FIG. 9. Chroococcidiopsis sp. exposed to 13 years and 3 months’ desiccation. (A) Cells retained green pigmentation
visualized under bright field microscopy either as cell masses or as separated cells (B). (C) Cells also demonstrated
autofluorescence from chlorophylls. (D) Some cells demonstrated little or no esterase activity (red arrow); others dem-
onstrated clear esterase activity (white arrow). Images (B)–(D) are the same image under different filter sets.

Intriguingly, we observed an apparent increase in viable
Chroococcidiopsis sp. cell numbers after thermal stress up
to 55C in desiccated cells (Fig. 1) and after the first three
freeze-thaw cycles (Fig. 8). This finding might be attributed
to the growth pattern of the cells. They frequently form
tetrads during growth, and one plausible explanation for the
results is that desiccation or freeze-thaw causes separation
of the cells, which would lead to higher cell numbers when
determined by plate counts compared to the controls in
which tetrads and other multicell aggregates will generally
appear as a single colony. As the organism has a doubling
time of approximately 2–10 days, depending on growth
conditions, the observations are not accounted for by cell
growth during the course of the experiment.
We also investigated the influence of liquid water in UV
radiation damage. We found that separated cells of Gloeo-
capsa sp., when exposed to UV radiation, were more
quickly killed when desiccated compared to hydrated cells,
yet in a colonial growth form, we observed the opposite
effect. Cells had greater viability after UV irradiation when
in the desiccated state compared to the hydrated state.
When exposed to UV radiation as single cells, hydrated
cells have the advantage that they can actively repair ac-
cumulating damage, compared to desiccated cells, which
cannot actively repair this damage that results in more
rapidly accumulated lethal levels.
In contrast, in desiccated Gloeocapsa sp. colonies, cells
on the upper part of the colony protect those at greater depth
as shown by the esterase enzymatic assay. This is explained
by attenuation, by both direct absorption and scattering, of
UV radiation through the cell material. Even after 2 h ex-
posure to 254 nm radiation, there were still viable cells in
the interior of the colonies. However, when colonies are
hydrated, the cells are less compact, and we hypothesize that
UV radiation more effectively penetrates the colony interior
and kills the cells, which leads to a lower viability. When
hydrated, colonies were also liable to move around in the
fluid by convection, leading to more effective exposure on
the outside of the colonies. In extreme natural environments,
cyanobacteria usually adopt biofilm growth habits where
cells form a coherent layer either within rocks, under them,
or in desert crusts (Belnap and Lange, 2001). Thus, the data
we obtained with colonies are likely to be representative of
cells in the natural environment. One other way in which
liquid water can be damaging to cells under UV radiation
stress is by producing photolysis products such as reactive
 LIQUID WATER AND HABITABILITY IN EXTREME DESERTS
oxygen species, products that can be catalyzed through the
photo-Fenton reaction (Barrett and Baxendale, 1960; Garcı́a-
Fernández et al., 2012). This is likely to be one mechanism
for damage induced under hydrated conditions.
The deleterious effects of water are also observed at low
temperatures. Hydrated Chroococcidiopsis sp. were more
susceptible to damage caused by freeze-thaw cycles than
desiccated cells. One possible cause of this damage is the
formation of ice crystals that could damage cell membranes
or intracellular components. Previous research has sug-
gested that, under natural situations, cells might vitrify,
which would mitigate or prevent damage caused by ice
crystal formation (Clarke et al., 2013). Nevertheless, the
results show how liquid water can be deleterious in exac-
erbating the destructive effects of freeze-thaw.
We found that Chroococcidiopsis sp. desiccated for 13
years and 3 months retained viability and autofluorescence,
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
and some cells retained enzymatic activity. They were also
able to tolerate short-term 90C thermal stress similarly to
fresh desiccated cells. Chroococcidiopsis are known to tol-
erate desiccation (Potts, 1999) and have been recovered
from 30-year-old hypoliths. The organism is also known to
possess a range of physiological and biochemical properties,
such as thick cell envelopes, that enhance desiccation re-
sistance (Grilli Caiola et al., 1996; Billi, 2009) and a sus-
ceptibility to desiccation that can lead to death in some cells,
but not in others (Billi, 2009). However, we found that when
the cells were rehydrated they were more susceptible to
thermal stress, showing that long-term desiccation com-
promises the ability of cells to deal with the deleterious
effects of liquid water.
These experiments show that life in extreme deserts is
precariously poised between two extremes. In one extreme, a
lack of liquid water threatens cell death by accumulated un-
repaired damage. In the other extreme, the presence of liquid
water exacerbates certain physically and chemically induced
damage such as extreme temperature and UV radiation stress.
These observations result in some predictions about the
habitability of extreme deserts. In deserts where liquid water
is present contemporaneously with high temperatures and
UV radiation, damage to some organisms will be greater;
thus deserts with these attributes will be less habitable to
them compared to deserts where cells have a chance to
desiccate prior to exposure to these extremes. For example,
a desert where water is available in the form of fog in the
early morning might allow for cell growth and replication.
The dissipation of the fog prior to midday high temperatures
and maximum UV radiation flux would allow cells to des-
iccate and better survive these extremes. A desert where
liquid water persists in microhabitats until the afternoon or
where precipitation occurs shortly before the maximum of
UV irradiation and high temperatures, which would leave
cells hydrated during these extremes, would expose them to
greater stress. In addition, at the other end of the tempera-
ture scale, these data illustrate that, for deserts that experi-
ence diurnal freeze-thaw cycles, those that are characterized
by more extreme conditions whereby cells are desiccated
prior to freezing may be more habitable than deserts where
greater liquid water availability causes cells to enter freeze-
thaw cycles in a hydrated state. Thus, these observations
predict the possibility of deserts that are less habitable to
certain organisms than deserts that are more extreme and
317
have lower liquid water availability. Some of the best places
to isolate novel thermal and radiation-tolerant extremophiles
or polyextremophiles with which to investigate new bio-
chemical pathways are not necessarily the most extreme dry
deserts. Instead, deserts where organisms must adapt to the
extremes caused by physical stress combined with the
presence of liquid water may provide more interesting in-
sights into microbial stress responses and adaptations.
Finally, these experiments provide an example of how
water is not always beneficial for life. Although the solvent
has long been known to mediate damaging chemical reactions
in life including hydrolysis reactions (Ball, 2007), these data
show how the damage caused by the presence of liquid water
under certain extremes might manifest at the ecological scale.
Under the most challenging high-temperature conditions to be
found in hot deserts (or, more speculatively, planets at the end
of the habitable lifetimes transitioning into a greenhouse state;
O’Malley-James et al., 2015), life is involved in a precarious
balance in using its available energy and capabilities of ad-
aptation to deal with the deleterious effects of both the lack
and presence of liquid water.
Acknowledgments
H.L. received summer student funding support from the
School of Physics and Astronomy. The Microbiology Society
provided the Harry Smith Vacation Studentship for T.S.
References
Ball, P. (2007) Water as an active constituent in cell biology.
Chem Rev 108:74–108.
Balogi, Z., Török, Z., Balogh, G., Jósvay, K., Shigapova, N.,
Vierling, E., Vigh, L., and Horváth, I. (2005) ‘‘Heat shock
lipid’’ in cyanobacteria during heat/light-acclimation. Arch
Biochem Biophys 436:346–354.
Baqué, M., Scalzi, G., Rabbow, E., Rettberg, P., and Billi, D.
(2013) Biofilm and planktonic lifestyles differently support
the resistance of the desert cyanobacterium Chroococci-
diopsis under space and martian simulations. Orig Life Evol
Biosph 3:377–389.
Barrett, J. and Baxendale, J.H. (1960) The photolysis of liquid
water. Transactions of the Faraday Society 56:37–43.
Battin, T.J. (1997) Assessment of fluorescein diacetate hydro-
lysis as a measure of total esterase activity in natural stream
sediment biofilms. Sci Total Environ 198:51–60.
Belnap, J. and Lange, O. (2001) Biological Soil Crusts:
Structure, Function, and Management, Springer, New York,
p 503.
Berner, T. and Evenari, M. (1978) The influence of temperature
and light penetration on the abundance of the hypolithic algae
in the Negev Desert of Israel. Oecologia 33:255–260.
Billi, D. (2009) Subcellular integrities in Chroococcidiopsis
sp. CCMEE 029 survivors after prolonged desiccation re-
vealed by molecular probes and genome stability assays.
Extremophiles 13:49–57.
Billi, D., Viaggiu, E., Cockell, C.S., Rabbow, E., Horneck, G., and
Onofri, S. (2010) Damage escape and repair in dried Chroo-
coccidiopsis spp. from hot and cold deserts exposed to simulated
space and martian conditions. Astrobiology 11:65–73.
Broady, P.A. (1981) The ecology of chasmolithic algae at
coastal locations of Antarctica. Phycologia 20:259–272.
Büdel, B. (2011) Cyanobacteria: habitats and species. In Plant
Desiccation Tolerance, Ecological Studies 215, edited by U.
Lüttge, E. Beck, and D. Bartels, Springer, Berlin, pp 11–21.
 318
Casteilli, O., de la Cerda, B., Navarro, J.A., Hervas, M., and de
la Rosa, M.A. (2009) Proteomic analyses of the response of
cyanobacteria to different stress conditions. FEBS Lett 583:
1753–1758.
Castenholz, R.W. and Garcia-Pichel, F. (2000) Cyanobacterial re-
sponses to UV radiation. In The Ecology of Cyanobacteria: Their
Diversity in Time and Space, edited by B.A. Whitton and M. Potts,
Kluwer Academic, Dordrecht, the Netherlands, pp 591–611.
Clarke, A., Morris, G.J., Fonseca, F., Murray, B.J., Acton, E.,
and Price, H.C. (2013) A low temperature limit for life on
Earth. PloS One 8, doi:10.1371/journal.pone.0066207.
Cockell, C.S., Schuerger, A.C., Billi, D., Friedmann, E.I., and
Panitz, C. (2005) Effects of a simulated martian UV flux on
the cyanobacterium Chroococcidiopsis sp 029. Astrobiology
5:127–140.
Cockell, C.S., McKay, C.P., Warren-Rhodes, K.A., and Hor-
neck, G. (2008) Ultraviolet radiation-induced limitation to
epilithic microbial growth in arid deserts—dosimetric ex-
Downloaded by Macquarie University (Caul) from www.liebertpub.com at 03/18/18. For personal use only.
periments in the hyperarid core of the Atacama Desert. J
Photochem Photobiol 90:79–87.
Cockell, C.S., Rettberg, P., Rabbow, E., and Olsson-Francis, K.
(2011) Exposure of phototrophs to 548 days in low Earth
orbit: microbial selection pressures in outer space and on
early Earth. ISME J 5:1671–1682.
Cossins, A.R. (1994) Temperature Adaptation of Biological
Membranes, Portland Press, London.
Douglas, S. (2004) Microbial biosignatures in evaporite de-
posits: evidence from Death Valley, California. Planet Space
Sci 52:223–227.
Friedmann, E.I. (1980) Endolithic microbial life in hot and cold
deserts. Orig Life Evol Biosph 10:223–235.
Garcı́a-Fernández, I., Polo-Lopez, M.I., Oller, I., and Fernández-
Ibáñez, P. (2012) Bacteria and fungi inactivation using Fe3+/
sunlight, H2O2/sunlight and near neutral photo-Fenton: a
comparative study. Appl Catal B 121–122:20–29.
Grilli Caiola, M., Billi, D., and Friedmann, E.I. (1996) Effect of
desiccation on envelopes of the cyanobacterium Chroo-
coccidiopsis (Chroococcales). Eur J Phycol 31:97–105.
Inouse, N., Taira, Y., Emi, T., Yamane, Y., Kashino, Y., Koike,
H., and Satoh, K. (2001) Acclimation to the growth temper-
ature and high temperature effects on photosystem II and
plasma membranes in a mesophilic cyanobacterium. Sy-
nechocystis sp. P6803. Plant Cell Physiol 42:1140–1148.
Lange, O.L., Meyer, A., Zellner, H., Ullmann, I., and Wessels,
D.C.J. (1990) Eight days in the life of a desert lichen: water
relations and photosynthesis of Teloschistes capensis in the
coastal fog zone of the Namib Desert. Madoqua 17:17–30.
Larcher, W., Wagner, J., and Lütz, C. (1997) The effect of heat
on photosynthesis, dark respiration and cellular ultrastructure
of the arctic-alpine psychrophyte Ranunculus glacialis.
Photosynthetica 34:219–232.
Lüttge, U. (2011) Cyanobacteria: multiple stresses, desiccation-
tolerant photosynthesis and di-nitrogen fixation. In Plant
Desiccation Tolerance, Ecological Studies 215, edited by
U. Lüttge, E. Beck, and D. Bartels, Springer, Berlin, pp
23–42.
Moore, L.R., Coe, A., Zinser, E.R., Saito, M.A., Sullivan, M.B.,
Lindell, D., Frois-Moniz, K., Waterbury, J., and Chisholm,
S.W. (2007) Culturing the marine cyanobacterium Pro-
chlorococcus. Limnol Oceanogr 5:353–362.
Mottl, M., Glazer, B., Kaiser, R., and Meech, K. (2007) Water
and astrobiology. Chemie Erde 67:253–282.
Nienow, J.A., McKay, C.P., and Friedmann, E.I. (1988) The
cryptoendolithic microbial environment in the Ross Desert,
COCKELL ET AL.
Antarctica: light in the photosynthetically active region. Mi-
crob Ecol 16:271–289.
Olsson-Francis, K., de la Torre, R., and Cockell, C.S. (2010)
Isolation of novel extreme-tolerant cyanobacteria from a
coastal rock-dwelling microbial community using exposure to
low Earth orbit. Appl Environ Microbiol 76:2115–2121.
O’Malley-James, J.T., Greaves, J.S., Raven, J.A., and Cockell,
C.S. (2015) In search of future Earths: assessing the possi-
bility of finding Earth analogues in the later stages of their
habitable lifetimes. Astrobiology 15:400–411.
Pointing, S.B., Warren-Rhodes, K.A., Lacap, D.C., Rhodes,
K.L., and McKay, C.P. (2006) Hypolithic community shifts
occur as a result of liquid water availability along environ-
mental gradients in China’s hot and cold hyperarid deserts.
Environ Microbiol 9:414–424.
Potts, M. (1999) Mechanisms of desiccation tolerance in cya-
nobacteria. Eur J Phycol 34:319–328.
Quesada, A. and Vincent, W.F. (1997) Strategies of adaptation
by Antarctic cyanobacteria to ultraviolet radiation. Eur J
Phycol 32:335–342.
Rajeev, L., da Rocha, U.N., Klitgord, N., Luning, E.G., Fortney,
J., Axen, S.D., Shih, P.M., Bouskill, N.J., Bowen, B.P.,
Kerfeld, C.A., Garcia-Pichel, F., Brodie, E.L., Northen, T.R.,
and Mukhopadhyay, A. (2013) Dynamic cyanobacterial re-
sponse to hydration and dehydration in a desert biological soil
crust. ISME J 7:2178–2191.
Regel, R.H., Ferris, J.M., Ganf, G.G., and Brookes, J.D. (2002)
Algal esterase activity as a biomeasure of environmental deg-
radation in a freshwater creek. Aquatic Toxicol 59:209–223.
Rippka, R., Deruelles, J., Waterbury, J.B., Herdman, M., and
Stanier, R.Y. (1979) Generic assignments, strain histories and
properties of pure cultures of cyanobacteria. J Gen Microbiol
111:1–61.
Schulz, E.M. and Scherer, S. (1999) UV protection in cyano-
bacteria. Eur J Phycol 34:329–338.
Suzuki, I., Simon, W.J., and Slabas, A.R. (2006) The heat shock
response of Synechocystis PCC6803 analysed by tran-
scriptomics and proteomics. J Exp Bot 57:1573–1578.
Vigh, L., Maresca, B., and Harwood, J.L. (1998) Does the
membrane’s physical state control the expression of heat
shock and other genes? Trends Biochem Sci 23:369–374.
Warren-Rhodes, K.A., Rhodes, K.L., Pointing, S.B., Ewing, S.A.,
Lacap, D.C., Gómez-Silva, B., Amundson, R., Friedmann, E.I.,
and McKay, C.P. (2006) Hypolithic cyanobacteria, dry limit of
photosynthesis, and microbial ecology in the hyperarid Ata-
cama Desert. Microbial Ecol 52:389–398.
Address correspondence to:
Charles S. Cockell
UK Centre for Astrobiology
School of Physics and Astronomy
University of Edinburgh
Edinburgh, EH10 4EP
UK
E-mail: c.s.cockell@ed.ac.uk
Submitted 20 August 2016
Accepted 24 October 2016
Abbreviation Used
FDA ¼ fluorescein diacetate