Introduction
An environment’s ability to support microbial life is determined by physio-chemical characteristics such as temperature, pressure, pH, ionic composition, water activity, radiation and nutrient availability (Rothschild and Mancinelli, Reference Rothschild and Mancinelli2001; Pikuta et al., Reference Pikuta, Hoover and Tang2007; Dartnell, Reference Dartnell2011; Harrison et al., Reference Harrison, Hallsworth and Cockell2015; Merino et al., Reference Merino, Aronson, Bojanova, Feyhl-Buska, Wong, Zhang and Giovannelli2019). The biochemical adaptations to individual environmental extremes have been extensively studied; however, multi-extreme adaptations remain poorly characterised (Konings et al., Reference Konings, Albers, Koning and Driessen2002; Jaenicke and Sterner, Reference Jaenicke and Sterner2006; Oren, Reference Oren2006; Baker-Austin and Dopson, Reference Baker-Austin and Dopson2007; Krulwich et al., Reference Krulwich, Sachs and Padan2011; Siliakus et al., Reference Siliakus, van der Osst and Kengen2017). Understanding how multiple environmental parameters interact to ultimately influence habitability is of immediate interest in disparate fields such as microbial ecology, astrobiology and biotechnology (Nichols et al., Reference Nichols, Olley, Garda, Brenner and McMeekein2000; Rothschild and Mancinelli, Reference Rothschild and Mancinelli2001; Tanaka et al., Reference Tanaka, Burgess and Wright2001; Pikuta et al., Reference Pikuta, Hoover and Tang2007; Bowers et al., Reference Bowers, Mesbah and Wiegel2009; Rampelotto, Reference Rampelotto2010; Harrison et al., Reference Harrison, Agel and Cockell2017).
Investigations into the limits of life usually focus on an organism’s capacity to cope with individual extremes in isolation and the limits imposed by individual environmental parameters have been defined in detailed studies (DasSarma and Arora, Reference DasSarma and Arora2002; Jaenicke and Sterner, Reference Jaenicke and Sterner2006; Krulwich, Reference Krulwich2006; Oren, Reference Oren2006; Baker-Austin and Dopson, Reference Baker-Austin and Dopson2007; Krulwich et al., Reference Krulwich, Sachs and Padan2011; Clarke, Reference Clarke2014; Merino et al., Reference Merino, Aronson, Bojanova, Feyhl-Buska, Wong, Zhang and Giovannelli2019). Yet research rarely focuses on multiple environmental parameters that more closely resemble a natural extreme environment. In some cases, it may be that one extreme mitigates the deleterious effect of another. For example, the barophilic archaeon Thermococcus barophilus was shown to have an increased growth rate at 85°C at higher pressure (40 MPa) compared with atmospheric pressure, as well as requiring 15–17.5 MPa for growth between 95 and 100°C (Marteinsson et al., Reference Marteinsson, Birrien, Reysenbach, Vernet, Marie, Gambacorta, Messner, Sleytr and Prieur1999), and the bacterium Shewanella gelidimarina displays an increased temperature range when cultured under 2.5 and 4.0% [wt/vol] concentrations of NaCl compared to 1.0% NaCl (Nichols et al., Reference Nichols, Olley, Garda, Brenner and McMeekein2000). Furthermore, culturing Listeria monocytogenes with 1.5 M NaCl increased its maximum growth temperature by 6°C, which was directly correlated with increased proteome stability, thus demonstrating a biophysical basis for the effects of multiple extremes (Anderson et al., Reference Anderson, Hedges, Jones and Cole1991).
One explanation for these interacting effects is that physical and chemical effects on cellular biochemistry are often mediated through the same biomolecules. For example, the composition of the cytoplasmic membrane of bacteria is a common biochemical basis for adaptation to different extremes. Membrane composition can be altered in response to osmotic stress (Padan et al., Reference Roberts2001; Konings et al., Reference Konings, Albers, Koning and Driessen2002; Roberts, Reference Roberts2005; Aston and Peyton, Reference Aston and Peyton2007; Bremer and Krämer, Reference Bremer and Krämer2019), but other environmental parameters that influence membrane integrity and composition are temperature, pH and hydrostatic pressure, causing changes in permeability and fluidity (Nichols et al., Reference Nichols, Olley, Garda, Brenner and McMeekein2000; Konings et al., Reference Konings, Albers, Koning and Driessen2002; Siliakus et al., Reference Siliakus, van der Osst and Kengen2017). Similarly, it has been shown that mannosyl-glycerate is accumulated in some archaea in response to temperature, salinity and pressure extremes, with the greatest accumulation under multiple extremes, suggesting that the production of this solute is used to cope with different extremes (Cario et al., Reference Cario, Jebbar, Thiel, Kervarec and Oger2016). Thus, we would predict that in general, the imposition of one environmental extreme would influence the limits to which an organism could cope with another.
One implication of these laboratory findings is that in natural environments, we might expect that in some cases, the challenges of adapting to multiple extremes may limit microbial diversity. For example, the extremely dynamic, ultra-acidic, high temperature lake in the active Poás crater (Laguna Caliente), Costa Rica, is thought to have low diversity on account of combined extremes of pH and temperature (Hynek et al., Reference Hynek, Rogers, Antunovich, Avard and Alvarado2018). In the Dallol Rift, Ethiopia, geochemically varied hot springs have been investigated for their microbial diversity (Belilla et al., Reference Belilla, Moreira, Jardillier, Reboul, Benzerara, López-García, Bertolino, López-Archilla and López-García2019) where extremes of high ions, high temperatures and acidity in different combinations have been proposed to act together to limit life in these environments.
Despite these observations, systematic studies to investigate how combined extremes such as salinity, temperature and pH act in combination to change the limits of life compared to the individual extremes in the same organism are surprisingly rare in the literature. Previous studies have addressed the combined effects of salinity, temperature and pH (Cole et al., Reference Cole, Jones and Holyoak1990; Pang et al., Reference Cole, Jones and Holyoak2020). However, these studies investigated the interaction of extremes on survival or on the growth rates at sub-lethal values. Relatively simple experiments to characterise the effect of individual extremes on organisms and combinations of those extremes would be useful in both mapping the limits to microbial growth and in motivating more detailed biochemical studies of how physical and chemical stressors combine to influence microbial limits to growth.
In this study, we undertook an investigation of the effects of a combination of extremes in temperature, salinity (NaCl) and pH on the growth of the psychrotolerant, moderately halophilic bacterium Halomonas hydrothermalis. We chose these parameters because they have been the focus of substantial investigation as individual limits of life (Tolner et al., Reference Tolner, Poolman and Konings1997; DasSarma and Arora, Reference DasSarma and Arora2002; Jaenicke and Sterner, Reference Jaenicke and Sterner2006; Oren, Reference Oren2006; Aston and Peyton, Reference Aston and Peyton2007; Baker-Austin and Dopson, Reference Baker-Austin and Dopson2007; Krulwich et al., Reference Krulwich, Sachs and Padan2011; Méndez-García et al., 2015; Jin and Kirk, Reference Jin and Kirk2018). The aim of this work was to test the hypothesis that combinations of extreme environmental parameters can restrict the growth limits of this organism more than the individual extremes on their own and to quantify those interactions.
Methods and Materials
Strain
H. hydrothermalis DSM-15725, originally isolated from a hydrothermal vent, was obtained from the German collection of Microorganisms (DSMZ, Braunschweig, Germany).
Growth assays
Growth cultures were obtained by growing the bacterium in minimal marine media (MMM) (Östling et al., Reference Östling, Goodman and Kjelleberg1991) using 0.5% glucose composition, with 1.63% [wt/vol] NaCl (pH 8). Aliquots were prepared with 25% [vol/vol] glycerol, stored at −80°C and subsequently used to inoculate MMM agar (above recipe with the addition of 1.5% Agar Bacteriological No. 1) which were grown at 30°C for 48 hours and stored at 4°C until use. Starter cultures were prepared by transferring cells from agar plates to 5 mL MMM broth (1.63% NaCl [wt/vol] pH 8) in a loosely capped 15 mL tube (Sarstedt, Nümbrecht, Germany). The culture was grown for 24 hours in a shaking incubator (30°C, 120 rpm) and diluted with fresh MMM to a final cell density equivalent to an optical density at 600 nm (OD600) of 0.2. Optical density measurements for growth assay starter cultures were obtained with a volume of 1 mL using the DR 5000 UV-Vis Spectrophotometer (Hach Company, Düsseldorf, Germany).
Growth assays were started by adding 10 µL starter culture to fresh MMM broth at a range of salinities (n = 21) to a total volume of 200 µL in each well of a 96-well micro-plates with the accompanying lid placed over the whole plate (Greiner Bio-One, Frickenhausen, Germany). Twenty-one variations of MMM were used with 6.49%, 6.58%, 6.66%, 6.75%, 6.84%, 6.93%, 7.01%, 7.10%, 7.19%, 7.28%, 7.37%, 7.45%, 7.54%, 7.63%, 7.72%, 7.80%, 7.89%, 7.98%, 8.07%, 8.16% and 8.24% NaCl [wt/vol]. To alleviate pipetting error, aliquots for each salinity were prepared in 50 mL Falcon tubes by mixing MMM with 0% and 10% NaCl [wt/vol] to the required final NaCl concentration and subsequent 96-well plate preparations were made. Three variations of culture media based on MMM under salinity ranges were prepared at pH values of 8, 7 and 6 using 1 M hydrochloric acid to lower pH. The pH was measured using a Jenway 3510 pH metre (Jenway, Fisher, UK).
All cultures were incubated at 30, 40, 41, 42, 43, 44 and 45°C. These temperatures span the reported optimal (30°C) and supra-optimal (40°C) values for H. hydrothermalis under otherwise optimal conditions, with the additional exploration of further supra-optimal (41-45°C) temperatures. To ensure that the pH of the culture medium did not vary over the temperature range tested, 50 mL aliquots of each media pH variation were stored at each temperature used in this experiment for a period of 48 hours and the pH was measured.
Cell density was measured every 10-minutes over a 24-hour period by OD600 measurement using a Synergy 2 microplate reader (BioTek Instruments Inc., Vermont, USA) shaken continuously at 1080 rpm. H. hydrothermalis achieves stationary phase growth and reaches death phase within 24-hr. In this study, we chose to define ‘no growth’ to be where the OD600 value did not exceed 0.01 after the 24-hr incubation period. We cannot rule out extremely slow growth at lower optical densities, but this allowed us to make comparisons between conditions. Maximal OD600 values reported in this study correspond to the maximal OD600 measurement achieved over the 24-hour period. Each culture condition was observed in triplicate within a micro-plate and triplicate micro-plates monitored per temperature and pH combination, providing n = 9 measurements for each culture variation. Medium only controls were included (n = 6) for each plate and used to perform a background subtraction from experimental wells. Because of very low maximal OD600 values attained under multiple extremes, it is necessary to establish a cut off point for assumed propagation. We chose to measure maximal optical density as a proxy for total cell biomass achieved in a given unit of time. OD600 readings can in principle be altered by cell size and shape, but we consider it is to be a sufficiently robust measure of growth for a single organism under the well-defined conditions of our experiment and we verified this in exponential growth and under optimal growth conditions by calibrating OD600 against colony-forming units (cfus). In addition, we were interested in the limits of growth. The reduction of optical density changes to below a threshold representing measurable growth would not be influenced by small changes in optical density measurements that could be caused by, for example, cell morphology. Furthermore, this method allowed us to take a sufficient number of measurements in a reasonable time.
Calibration curve for OD versus growth
A calibration curve based on colony counting was obtained for H. hydrothermalis to demonstrate that OD600 values were proportional to cell population in the exponential phase.
H. hydrothermalis was cultured in MMM. Colony Forming Units (CFUs) per mL were obtained every hour over a 24-hour growth period. For each time point, 1 mL of sample was removed from the culture and serial dilutions (10-1 to 10-9) created by transferring media into sterile Falcon tubes, adding phosphate-buffered saline (PBS) and vortexing after each dilution transfer. OD600 values were also obtained. One hundred microlitres of each dilution was then pipetted onto a sterile MMM agar plate and spread across the surface using a sterile single-use spreader. Plates were incubated for 48 hours in a Memmert GmbH incubator under optimal temperatures. Plates were inspected and the dilution showing between 30 and 300 clear colonies was selected for counting. CFUs were calculated, correcting for dilution.
A standard curve for CFU versus OD600 was then plotted for the exponential phase. Linear regression analysis was undertaken using Microsoft Excel to obtain a value of R2 to assess the relationship between CFU and OD600 measurements.
Data analysis
For comparisons of maximal OD600 values under optimal temperature conditions (30°C) a two-way analysis of variance (ANOVA) was performed with salinity and pH as the factors. Maximum OD600 values under supra-optimal temperature conditions (40-45°C) were compared using three-way ANOVA with temperature, salinity and pH as the factors. To meet model assumptions a Levene’s test was performed to assess to the equality of variance. To alleviate heteroscedasticity of variance, measurements obtained for cells incubated under all culture conditions were Box-Cox transformed (λ = 0.17 and 0.07 for 30°C and 40-45°C, respectively).
For confirmation of where the differences occur between each combination of temperature, salinity and pH, the three-way ANOVA and Box-Cox transformed data were performed in conjunction with a Tukey’s honestly significant difference (HSD) test. All statistical analysis performed using RStudio v1.1.453 (R Development Core Team, 2015). Levene’s test, Box-Cox transformation and Tukey HSD were completed using the ‘car’, ‘MASS’ and ‘agricolae’ packages, respectively.
Results
Effects of salinity and pH on limits of growth at optimal temperature
Under the optimal temperature condition of 30°C, H. hydrothermalis displayed growth up to and including the highest concentration of NaCl tested in this study (8.24%) at all pH values tested in this study (6-8) (Figure 1). There is notably less growth at pH 6 when compared with pH 7 and 8, however some growth occurred at pH 6 by the criterion employed here since maximal OD600 values for H. hydrothermalis obtained under pH 6 conditions were above the 0.01 optical density threshold for this study at all salinities tested under optimal temperature conditions (Figure 1). At optimal temperatures, maximal growth was greater at pH 7.0 compared to pH 8.0 at any given salinity. The effect of NaCl and pH (6-8) on maximal OD600 values of H. hydrothermalis under optimal temperature conditions was demonstrated by a two-way ANOVA and Tukey’s HSD test comparing maximal OD600 values. Significant effects of NaCl concentration and pH on maximal OD600 values were observed both individually and in combination (Table 1). Tukey’s post hoc was used to compare the means of each culture condition. Maximal OD600 values at pH 6 were significantly different to all values at pH 7 and 8.
Figure 1. Maximal OD 600 values of H. hydrothermalis cultures (n = 9) under optimal temperature conditions (30°C). Maximal OD 600 values for H. hydrothermalis cultures obtained under a variation of salinities (6.44 – 8.18%, n = 21) and pH (6 – 8, n = 3). Data shown as means and standard error of the mean.
Table 1. Two-way ANOVA results for maximal OD600 Values at 30°C. These data correspond with the interactions between the factor’s salinity and pH. Cultures were incubated under a range of salinities (n = 21) and pH values of 8, 7 and 6. Data were Box-Cox transformed to alleviate heteroscedasticity of variance
Combined effects of pH, salinity and supra-optimal temperature on limits to growth
Culturing at supra-optimal temperatures revealed a temperature-salinity-pH interplay not seen under optimal temperatures (Figures 2 and 3). At pH 8, H. hydrothermalis displayed the highest growth temperature of 43°C, while tolerating 6.58% NaCl. Decreasing the temperature to 42 and 41/40°C increased the salinity limit to 7.01 % and 8.24 % respectively. We note here that the cultures grew at temperatures exceeding the previously published temperature tolerance limit for this organism (40°C) (Kaye et al., Reference Kaye, Márquez, Ventosa and Baross2004). At 44 and 45°C, no growth was observed under any pH or NaCl concentration assayed.
Figure 2. Maximal OD 600 values of H. hydrothermalis cultures (n = 9) at supra-optimal temperatures. Values were obtained for cultures grown under a temperature range of 40 – 45°C (n = 5), variations in salinity (6.44 – 8.18%, n = 21) and different pH values (6 – 8, n = 3). Views of the maximal OD 600 values are displayed from two arbitrarily selected angles.
Figure 3. Heatmaps displaying maximal OD 600 values of H. hydrothermalis cultures (n = 9) under a range of temperatures (40 – 45°C, n = 5) and salinities (6.44 – 8.18%, n = 21) at pH 7 (a) and pH87 (b). Data omitted for pH 6 as no OD 600 value was above 0.01.
At pH 7, the temperature-salinity limits to growth observed were: 7.37% at 40°C, 6.75% at 41°C and 6.58% at 42°C, with no growth between 43 and 45°C (Figures 2 and 3). At pH 6 no growth was observed at any NaCl concentration at supra-optimal temperatures (40 – 45°C) (Figure 2).
The reduction of maximal OD600 values of H. hydrothermalis under the combinations of the three different extremes, supra-optimal temperatures, NaCl concentration and pH (6-8) was demonstrated by three-way ANOVA and Tukey’s HSD test comparing maximal OD600 values. Significant effects of temperature, NaCl concentration and pH were observed both individually and in combination (Table 2). Tukey’s post hoc was used to compare the maximum growth means of each culture condition. Under culture conditions at pH 8, at 40°C the most pronounced significant differences in maximal OD600 values were between 6.49 – 6.75% and 7.37 – 8.24% NaCl (excluding 7.45%). Differences observed at 41°C where salinity ranges between 6.49 – 6.75% NaCl were significantly different to those obtained at 7.98% NaCl, and at 42°C where maximal OD600 values obtained between 6.49 – 6.58% NaCl were significantly different to those between 7.10 – 8.24% NaCl. There were shown to be no significant differences between maximal OD600 values within temperatures 43 – 45°C.
Table 2. Three-Way ANOVA results for maximal OD600 Values at 40-45°C. These data correspond with the interactions between the factor’s temperature, salinity and pH. Cultures were incubated under a range of temperatures (n = 5), salinities (n = 21) and pH values of 8, 7 and 6. Data were Box-Cox transformed to alleviate heteroscedasticity of variance
At pH 7, at 40°C significant differences in the maximum growth observed were observed at 6.49% and between 7.45 – 7.89% NaCl (excluding 7.89%). Owing to low maximal OD600 values obtained between pH 7 at 41°C – 45°C, post-hoc analysis shows there to be no significant differences. We observed differences in the NaCl concentrations at which maximum OD600 values were observed between pH conditions.
Growth calibration curve
Analysis indicated a strong relationship between CFU and OD600 values, with a R2 value of 0.9768 (Supplemental Figure 1).
Discussion
Understanding the limits to microbial growth under multiple extremes remains a surprisingly unexplored area of microbiology. While substantial work has elucidated the biophysical limits to life under individual extremes (Tolner et al., Reference Tolner, Poolman and Konings1997; DasSarma and Arora, Reference DasSarma and Arora2002; Konings et al., Reference Konings, Albers, Koning and Driessen2002; Jaenicke and Sterner, Reference Jaenicke and Sterner2006; Oren, Reference Oren2006; Aston and Peyton, Reference Aston and Peyton2007; Krulwich et al., Reference Krulwich, Sachs and Padan2011; Mesbah and Wiegel, Reference Mesbah and Wiegel2012; Méndez-García et al., Reference Méndez-García, Peláez, Mesa, Sánchez, Golyshina and Ferrer2015; Merino et al., Reference Merino, Aronson, Bojanova, Feyhl-Buska, Wong, Zhang and Giovannelli2019), our knowledge of life’s limits diminishes significantly when we consider combined extremes. A previous report by Harrison et al. (Reference Harrison, Gheeraert, Tsigelnitskiy and Cockell2013) examined the laboratory growth data of 67 prokaryotic strains to produce three-dimensional maps of the limits of microbial growth under the extremes of NaCl, temperature and pH, but the lack of data on the combined effects of these stresses meant that these maps represent the limits of life only under single imposed stresses.
While it seems intuitively trivial that growth under multiple extremes would produce different results than individual extremes imposed alone, there are very few laboratory experiments that systematically examine these differences for a given organism and under extremes of relevance to the natural environment.
There are various reasons for the lack of information on combined extremes. Technologically, it is a challenge to develop suits of equipment which can probe the facets of life under extreme conditions across multiple scales, from proteins to cells to communities. More fundamentally, the primary challenge is that environmental extremes are not experienced by life in isolation, and as such the synergistic effects of multiple environmental extremes are hard to disentangle. For example, the currently defined upper temperature limit for life at 122 ºC is established at above ambient pressure (20 MPa) (Takai et al., Reference Takai, Nakamura, Toki, Tsunogai, Miyazaki, Miyazaki, Hirayama, Nakagawa, Nunoura and Hrikoshi2008). At atmospheric pressure (101 kPa), the upper temperature limit is necessarily limited by the upper temperature limit of liquid water stability at 100 ºC. Our concept of the theoretical upper temperature limit for life (Daniel and Cowan, Reference Daniel and Cowan2000) is implicitly a multi-extreme measurement which includes a pressure extreme.
In this study, we sought to investigate how the growth conditions of the hydrothermal bacterium H. hydrothermalis under the combined extremes of temperature, pH and salinity compared to the limits of growth under each extreme individually. Our motivation was to study, quantitatively, the extent to which imposing multiple extremes influenced the growth limit compared to each extreme in isolation.
Although we could have selected an extreme polyextremophile, we sought to use an organism whose growth limits were within the operating conditions of standard research equipment, and that had relatively broad pH, temperature and salinity tolerances, allowing us to easily explore growth under varied multiple extremes. The organism has also been the subject of previous multiple extreme studies (Harrison et al., Reference Harrison, Hallsworth and Cockell2015). H. hydrothermalis exhibits cellular division between 2°C and 40°C (optimal growth at 30°C), NaCl concentrations between 0.5% and 22% (wt/vol)(optimal range of 4% to 7% (wt/vol)) and pH between 5 and 12 (optimal range of 7 to 8) (Kaye and Baross, Reference Kaye and Baross2004). Owing to its ability to propagate under a broad range of temperature, salinity and pH ranges, H. hydrothermalis is a model organism for investigating the limits of life under a combination of extreme parameters.
We found that under optimal temperature conditions, changing the pH and increasing the salinity reduced the maximal growth of H. hydrothermalis¸ as one might expect. However, the effects of temperature, pH and salinity became more nuanced as the temperatures became supra-optimal. We demonstrated that culturing at pH 8 increased the growth limit for temperature and salinity, despite reducing the maximum growth under optimal conditions compared to pH 7. H. hydrothermalis growth at 43°C was permissible at pH 8, and a 0.87% wt/vol increase in salinity tolerance at 40°C was observed when cells were cultured at pH 8 versus pH 7.
Qualitatively similar observations were made by Nichols et al. (Reference Nichols, Olley, Garda, Brenner and McMeekein2000) who showed that the temperature growth range was higher at the more extreme (lower) water activity for Shewanella gelidimarina although peak growth rate was lower. They attribute these changes to changes in the membrane. A hypothesis to explain our observations could be that membrane adaptation at more extreme pHs (for example, greater packing of lipids to achieve better proton regulation) might extend the salinity tolerance range by countering osmotic stress, but could lower the efficiency of chemiosmosis, thus reducing final growth yield.
The conditions under which cells did not grow demonstrates the intricacy of examining multiple environmental parameters simultaneously. Our results show how small changes in one parameter can influence the limits of another. This shows the complexity of the ‘sliding-scales’ of environmental parameters, where increasing one variable (temperature) can reduce the tolerance for another (salinity), or increasing a variable (pH) can lead to increased tolerance for another (temperature).
When cultured at pH 6, the combined effect of pH, salinity and supra-optimal temperature further acted to restrict growth of H. hydrothermalis and the maximal OD600 values obtained demonstrate that under these conditions, pH completely dominated growth. At pH 6, the limitation to growth is such that additional limitations to growth imposed by NaCl have no additional discernible effect on cell physiology, in which case at these combined extremes one individual extreme can dominate the limits to growth.
At supra-optimal temperatures, we observed the organism’s increasing sensitivity to NaCl. One potential explanation for these data could be changes in the organism’s membranes. For example, supra-optimal temperature adaptations in mesophilic bacteria can involve structural modification of the membrane such as adopting a more fluid and permeable structure by increasing the chain length of lipid acyl chains and the degree of saturation (Konings et al., Reference Konings, Albers, Koning and Driessen2002; Siliakus et al., Reference Siliakus, van der Osst and Kengen2017). An increase in permeability of the H. hydrothermalis membrane at supra-optimal temperatures could explain the increase in the sensitivity to NaCl at increasing temperatures and the interaction between these two extremes, similarly with pH.
We found that pH exerted a stronger effect than NaCl on growth at optimal temperatures. The pH is known to be able to dominate microbial community composition and abundance (Merino et al., Reference Merino, Aronson, Bojanova, Feyhl-Buska, Wong, Zhang and Giovannelli2019) and adaptations to both alkaline and acidic pH rely on maintaining intracellular neutrality through proton uptake or efflux over the cell membrane (Baker-Austin and Dopson, Reference Baker-Austin and Dopson2007; Krulwich et al., Reference Krulwich, Sachs and Padan2011; Siliakus et al., Reference Siliakus, van der Osst and Kengen2017). In this organism, impairment of intracellular pH homeostasis by changing the extracellular pH may dominate cellular responses compared to osmotic regulation.
Understanding the multifaceted nature of life’s responses to multiple environmental extremes is further frustrated by trying to pinpoint the biophysical basis. Considering that cells are made of a diverse range of biomolecules of various stabilities, sizes and charges, finding the exact site where environmental extremes exert their effects is challenging, even more so when it is not consistent. For example, it has been shown by various groups that NaCl can both increase or decrease proteome stability (Anderson et al., Reference Anderson, Hedges, Jones and Cole1991; Lee and Kaletunç, Reference Lee and Kaletunç2005), and more broadly ionic effects vary amongst proteins and even vary depending on the buffers in which they are assayed (Salis and Monduzzi, Reference Salis and Monduzzi2016).
The statistical analysis conducted here offers a useful way disentangle the effect of environmental parameters on cellular growth, both individually and in all combinations. The three-way ANOVA confirmed that the pH, temperature and salinity all exerted statistically significant effects on growth, and in each of their combinations.
Our data show that we can only interpret the presence or absence of organisms in the natural environment by understanding the combined effects of extreme conditions. Figure 4 shows a simplified depiction of the data obtained in this study at pH 8.0. It illustrates that there exist combinations of salinity and temperature which prevented growth, yet each individual extreme of salinity and temperature was tolerable to the organism when combined with lower temperature or salinity, respectively. Separate investigations of the limits of individual extremes of salinity or temperature might lead us to predict that point ‘X’ was permissive for growth, contrary to the real biophysical effects of the multi-extreme conditions. Furthermore, such experiments would be necessary to explain the lack of particular organisms in environments in which individual extremes might a priori be predicted to be non-limiting to microbial growth.
Figure 4. A conceptual illustration of the limits of H. hydrothermalis growth at pH 8.0 described in this work (simplification of Figure 3a). Point ‘X’ marks a combination of extremes that resulted in no growth, yet the two individual parameters of temperature and salinity at this point do not on their own prevent the growth of the organism when combined with either lower salinity or temperature, respectively.
These data have implications for the biosphere, suggesting that the limits to life may be more restricted than those defined by growth under single extreme would suggest. Few natural environments on the planet exhibit only one extreme (Hynek et al., Reference Hynek, Rogers, Antunovich, Avard and Alvarado2018; Belilla et al., Reference Belilla, Moreira, Jardillier, Reboul, Benzerara, López-García, Bertolino, López-Archilla and López-García2019) and thus attempts to define the biosphere’s limits based on the artificiality of single extremes may mislead our view of the limits to life in its totality. By contrast, in some instances, if one extreme mitigates another extreme (Nichols et al., Reference Nichols, Olley, Garda, Brenner and McMeekein2000) we may underestimate the capacities of the biosphere.
Finally, our results have applications to assessing the habitability of other planetary bodies. For example, when considering the theoretical habitability of planetary bodies that have been shown to have past or present liquid water, such as Mars, the subsurface oceans of the Saturnian moon Enceladus and the Jovian moon Europa, one must consider the net effect of all environmental parameters imposed within the respective habitat. For example, brines that contain high levels of salts such as sulfates and perchlorates, allowing them to remain liquid in the near surface of Mars, have been proposed as locations to test for the presence of habitable conditions on Mars (Stevens et al., Reference Stevens, Childers, Fox-Powell, Nicholson, Jhoti and Cockell2019). Other extremes such as temperature (Cavicchioli, Reference Cavicchioli2002), pH (Fairén, Reference Fairén2010), desiccation (Stevens et al., Reference Stevens, Childers, Fox-Powell, Nicholson, Jhoti and Cockell2019), water activity (Martin-Torres et al., Reference Martin-Torres, Zorzano, Valentín-Serrano, Harri, Genzer, Kemppinen, Rivera-Valentin, Jun, Wray, Madsen, Goetz, McEwen, Hardgrove, Renno, Chevrier, Mischna, Navarro-González, Martínez-Frías, Conrad, McConnochie, Cockell, Berger, Vasavada, Sumner and Vaniman2015), atmospheric pressure and UV radiation (Cockell et al., Reference Cockell, Catling, Davis, Kepner, Lee, Snook and McKay2000) and their changes over time (Ehlman et al., Reference Kaye, Márquez, Ventosa and Baross2016) will play a role. As on Earth, the assessment of any one of these parameters as lying within the growth and reproductive capacities of known organisms does not mean that the environment, when all its stresses are imposed on an organism, is habitable.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1473550425100153
Acknowledgements
This work was supported through the Science and Technologies Facilities Council (STFC) grant number ST/V000586/1 and an EPSRC studentship for A.W.D. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising from this submission.
Open access
