Perchlorate‐specific proteomic stress responses of Debaryomyces hansenii could enable microbial survival in Martian brines [original]

RESEARCH ARTICLE

Perchlorate-specific proteomic stress responses of

Debaryomyces hansenii could enable microbial survival in

Martian brines

Jacob Heinz

| Joerg Doellinger

| Deborah Maus

| Andy Schneider

Peter Lasch

| Hans-Peter Grossart

4,5

| Dirk Schulze-Makuch

1,4,6,7

Center for Astronomy and Astrophysics, RG

Astrobiology, Technische Universität Berlin,

Berlin, Germany

Robert Koch-Institute, Centre for Biological

Threats and Special Pathogens, Proteomics

and Spectroscopy (ZBS6), Berlin, Germany

Robert Koch-Institute, Metabolism of

Microbial Pathogens (NG2), Berlin, Germany

Department of Plankton and Microbial

Ecology, Leibniz-Institute of Freshwater

Ecology and Inland Fisheries (IGB), Stechlin,

Germany

Institute for Biochemistry and Biology,

Potsdam University, Potsdam, Germany

GFZ German Research Center for

Geosciences, Section Geomicrobiology,

Potsdam, Germany

School of the Environment, Washington

State University, Pullman, Washington, USA

Correspondence

Jacob Heinz, Center for Astronomy and

Astrophysics, RG Astrobiology, Sekr. ER 3-2,

Technische Universität Berlin, Hardenbergstr.

36A, 10623 Berlin, Germany.

Email: [email protected]

Funding information

Deutsche Forschungsgemeinschaft,

Grant/Award Number: 455070607

Abstract

If life exists on Mars, it would face several challenges including the pres-

ence of perchlorates, which destabilize biomacromolecules by inducing

chaotropic stress. However, little is known about perchlorate toxicity for

microorganisms on the cellular level. Here, we present the first proteomic

investigation on the perchlorate-specific stress responses of the halotolerant

yeast Debaryomyces hansenii and compare these to generally known salt

stress adaptations. We found that the responses to NaCl and NaClO

induced stresses share many common metabolic features, for example, sig-

nalling pathways, elevated energy metabolism, or osmolyte biosynthesis.

Nevertheless, several new perchlorate-specific stress responses could be

identified, such as protein glycosylation and cell wall remodulations, pre-

sumably in order to stabilize protein structures and the cell envelope. These

stress responses would also be relevant for putative life on Mars, which—

given the environmental conditions—likely developed chaotropic defence

strategies such as stabilized confirmations of biomacromolecules or the for-

mation of cell clusters.

INTRODUCTION

Life as we know it requires energy and access to

CHNOPS (carbon, hydrogen, nitrogen, oxygen, phos-

phorus, sulfur), trace elements, and liquid water. On

Mars, energy would be provided to putative life chemi-

cally or via sunlight, carbon is accessible through the

thin but CO

-rich atmosphere, and other essential ele-

ments are abundant in the regolith (Clark et al., 2021).

Availability of liquid water, however, is strongly

restricted due to the low atmospheric pressure of

approximately 6 mbar and mostly subzero tempera-

tures on Mars (Martínez & Renno, 2013). One of the

few possibilities to generate liquid water in the Martian

near surface is the formation of temporarily stable

brines via deliquescence, a process in which a hygro-

scopic salt absorbs water from the atmosphere and dis-

solves within that water (Hallsworth, 2020). It has been

Received: 10 June 2022 Accepted: 27 July 2022

DOI: 10.1111/1462-2920.16152

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,

provided the original work is properly cited.

Environ Microbiol. 2022;24:5051–5065. wileyonlinelibrary.com/journal/emi 5051

shown that deliquescent water is sufficient to drive the

metabolism of halotolerant methanogenic archaea

(Maus et al., 2020). Intriguingly, several hygroscopic

salts have been detected on Mars (Davila et al., 2010).

Among those are very deliquescent and freezing point

depressing perchlorates (ClO



), which are widely dis-

tributed on the Martian surface (Clark &

Kounaves, 2016) but appear in natural environments

on Earth only occasionally in hyperarid deserts (Catling

et al., 2010; Kounaves et al., 2010).

Brines formed via deliquescence provide diverse

challenges for microbial life. High salt concentrations

lead to osmotic stress and reduced water activity,

which is a measure for the amount of unbound water

molecules in a solution available for biological pro-

cesses (Hallsworth et al., 2021). Furthermore, salts can

induce ion-specific stresses like interferences with the

cell’s metabolism or changes in cell permeability

through variations in ionic hydration shells (Waajen

et al., 2020). Some anions like perchlorate additionally

evoke chaotropic stress, that is, they destabilize bioma-

cromolecules like proteins (Ball & Hallsworth, 2015),

presumably through nonlocalized attractive dispersion

forces (Hyde et al., 2017). In Pseudomonas putida,it

has been shown that chaotropic solute-induced water

stress mainly leads to upregulation of proteins involved

in stabilization of biological macromolecules and mem-

brane structure (Hallsworth et al., 2003). Furthermore,

it has been demonstrated that chaotropic effects can

be neutralized to some extent by the presence or bio-

production of kosmotropes or compatible solutes

(Bhaganna et al., 2010; Cray et al., 2015). However,

detailed research on proteomic responses to chaotropic

stress induced by perchlorate is still lacking.

Here, we present a proteomic study investigating

the perchlorate-specific stress response of Debaryo-

myces hansenii to evaluate the physiological adapta-

tions required for microorganisms to thrive in the

Martian near surface. The halotolerant yeast

D. hansenii has been chosen as a model organism as it

has been described earlier to tolerate the highest per-

chlorate concentrations reported to date (Heinz

et al., 2020,2021). This yeast provides a large meta-

bolic toolset to counteract salt stress, such as the high-

osmolarity glycerol (HOG) pathway, which enables

stress signalling and concomitant biosynthesis of glyc-

erol (Prista et al., 2016), which acts as compatible sol-

ute and antagonizes osmotic as well as chaotropic

stress (Bhaganna et al., 2016). Its close relation to the

intensively studied bakery yeast Saccharomyces cere-

visiae greatly facilitates the annotation of proteins and

thus prediction of their functions.

It has been found previously that the eukaryotic

species thus far investigated showed a higher perchlo-

rate tolerance than prokaryotes (Heinz et al., 2020).

Even though an evolutionary development of eukary-

otes on Mars might be considered unlikely due to the

relatively short habitable window of Mars, it can be

assumed that Martian microorganism—if they exist—

would have adapted over longer time scales to the

increasing aridity, salinity, and perchlorate concentra-

tions (Davila & Schulze-Makuch, 2016) and developed

defence strategies at least as efficiently and complex

as observed in D. hansenii, which is not even exposed

to high perchlorate concentrations in its natural

environment.

For the investigation of the proteome of D. hansenii,

we choose a recently developed proteomics protocol

called Sample Preparation by Easy Extraction and

Digestion (SPEED) which enables sample-type inde-

pendent deep proteome profiling with high quantitative

accuracy and precision (Doellinger, Blumenscheit,

et al., 2020; Doellinger, Schneider, et al., 2020).

This is the first study investigating perchlorate-

specific stress responses with an untargeted proteomic

approach to provide novel and fundamental under-

standing of the required cellular adaptation mecha-

nisms for life in perchlorate-rich, chaotropic habitats on

Earth, Mars, and beyond.

EXPERIMENTAL PROCEDURES

Microbial cultures

The halotolerant yeast D. hansenii (DSM 3428) was

obtained from the Leibniz Institute DSMZ—German

Collection of Microorganisms and Cell Cultures. A stock

culture was grown aerobically without shaking at 25C

(optimum growth temperature) in liquid DMSZ growth

medium #90 (3% (w/v) malt extract, 0.3% (w/v) soya

peptone) and was frequently re-inoculated. Addition-

ally, four different salt-containing liquid growth media

(DSMZ #90) were prepared having a molal (mol/kg) salt

concentration of either 1.5 mol/kg NaClO

, 2.4 mol/kg

NaCl, 2.4 mol/kg NaClO

, or 3.9 mol/kg NaCl. The latter

two concentrations (2.4 mol/kg NaClO

and 3.9 mol/kg

NaCl) represent the almost highest concentrations of

the respective salt enabling growth of D. hansenii. The

maximum growth-enabling concentrations reported to

date are 2.5 mol/kg NaClO

and 4.0 mol/kg NaCl

(Heinz et al., 2021). We choose slightly lower concen-

trations to guarantee reproducible growth of the cul-

tures and to generate sufficient biomass for protein

extraction. The other two salt concentrations (1.5 mol/

kg NaClO

and 2.4 mol/kg NaCl) represent moderate

salt concentrations of the respective salt (i.e. approx.

62 mol% of the respective maximum salt concentra-

tions enabling growth). The availability of two treat-

ments with the same molal salt concentrations (2.4 mol/

kg NaCl and 2.4 mol/kg NaClO

) allowed for an addi-

tional comparison of cellular stress responses to the

two different salt species at the same osmolality. The

growth media were prepared by mixing the media

5052 HEINZ ET AL.

components, the respective salt and water, followed by

pH adjustment (pH 5.6) and sterile filtration. All treat-

ments (no salt, 1.5 mol/kg NaClO

, 2.4 mol/kg NaCl,

2.4 mol/kg NaClO

, and 3.9 mol/kg NaCl) were inocu-

lated as biological triplicates, that is, for each treatment

three different samples were inoculated. The salt-free

treatment and the samples containing 1.5 mol/kg

NaClO

and 2.4 mol/kg NaCl were inoculated with the

salt-free stock culture. Hence, the two saline treatments

with moderate salt concentrations (1.5 mol/kg NaClO

and 2.4 mol/kg NaCl) experienced a salt shock after

inoculation. Since the respective salt shock would be

too intense in 2.4 mol/kg NaClO

and 3.9 mol/kg NaCl

treatments to enable growth, these samples were inoc-

ulated with long-term adapted cultures already grown

at the respective salt concentration (Figure 1A).

Sample preparation for proteomics

Protein extraction was conducted using the recently

developed filter-aided Sample Preparation by Easy

Extraction and Digestion (fa-SPEED) protocol

(Doellinger, Schneider, et al., 2020). Cells were centri-

fuged for 3 min at 5.000 gafter reaching late expo-

nential growth phase, which is approximately 1 day for

salt-free treatments, 3 days for 1.5 mol/kg NaClO

and

2.4 mol/kg NaCl, 6 days for 2.4 mol/kg NaClO

, and

7 days for 3.9 mol/kg NaCl (Figure 1B). Cell pelleting in

3.9 mol/kg NaCl samples was incomplete (turbid super-

natant) but sufficient for further protein extraction. The

reason for incomplete pelleting is presumably an

electrostatic repulsion of cells because dilution of addi-

tional test samples with water did not result in larger

pellets but gently stirring with a grounded metal rod

before centrifugation did. The cell pellets were washed

three times with phosphate buffer saline (PBS) followed

by cell lysing with 50 μl trifluoroacetic acid (TFA) for

3 min at 70C. Afterwards, samples were neutralized

with 500 μl 2 M tris(hydroxymethyl)aminomethane

(TRIS) solution. After adding 55 μl reduction/alkylation

buffer (100 mM tris(2-carboxyethyl)phosphine/400 mM

2-Chloracetamid), the samples were incubated at 95C

for 5 min.

Protein concentrations were determined by turbidity

measurements at 360 nm using GENESYS™10S UV–

Vis spectrophotometer (Thermo Fisher Scientific). The

50 μg of proteins was diluted to 40 μl using a 10:1 (v/v)

mixture of 2 M TrisBase and TFA, mixed with 160 μl

acetone and incubated for 2 min at RT. For samples

containing less than 50 μg proteins per 40 μl sample,

the volumes of sample and acetone were increased at

constant sample/acetone ratio until 50 μg protein/

sample were reached. Afterwards, proteins were cap-

tured on Ultrafree

-MC (0.5 ml) centrifugal devices,

0.2 μm, PTFE (Merck) at 5000 gfor 2 min. The sam-

ples were washed successively with 200 μl 80% (v/v)

acetone, 200 μl 100% acetone and 200 μl n-pentane at

5000 gfor 2 min each.

Subsequently, 40 μl of digestion buffer (50 mM

ammonium bicarbonate) containing trypsin (1:25

[enzyme-to-protein ratio] Trypsin Gold, Mass Spectrom-

etry Grade [Promega]) was added to the filter contain-

ing the proteins followed by incubation at 37C for 20 h.

The sample solution containing the digested proteins

was centrifuged at 5.000 gfor 2 min and the filter

was washed subsequently with 40 μl digestion buffer

FIGURE 1 Workflow of the inoculation procedure and

corresponding growth curves. (A) A salt-free stock culture of

D. hansenii was frequently re-inoculated into fresh growth medium.

An aliquot of this culture was used to inoculate growth media with

moderate salt concentrations (2.4 mol/kg NaCl and 1.5 mol/kg

NaClO

). To obtain cell growth at even higher salt concentrations, a

stepwise concentration increase was needed for each inoculation

step. The maximum salt concentrations used in this study were

3.9 mol/kg NaCl and 2.4 mol/kg NaClO

. (B) Growth curves of all

samples used in this study (n=3). Cells were harvested for protein

extraction in the late exponential growth phase of the respective

treatments (label with ‘X’).

PERCHLORATE-SPECIFIC MICROBIAL STRESS RESPONSES 5053

containing 0.1% (v/v) TFA. The 10% (v/v) TFA solution

was added until the pH of the samples reached approx-

imately 2. Peptides were desalted using the Pierce™

Peptide Desalting Spin Columns (Thermo Scientific)

according to the manufacture’s protocol no. 2162704.

The desalted samples were dried in a vacuum concen-

trator. The dried peptides were dissolved in 0.1% (v/v)

formic acid and quantified by measuring the absor-

bance at 280 nm using an Implen NP80 spectropho-

tometer (Implen, Munich, Germany).

Liquid chromatography and mass

spectrometry

Peptides were analysed on an EASY-nanoLC 1200

(Thermo Fisher Scientific, Bremen, Germany) coupled

online to a Q Exactive™HF mass spectrometer

(Thermo Fisher Scientific). One microgram of peptides

was separated on a PepSep column (15 cm length,

75 μm i.d., 1.9 μm C18 beads, PepSep, Denmark)

using a stepped 30 min gradient of 80% (v/v) acetoni-

trile (Solvent B) in 0.1% (v/v) formic acid (Solvent A) at

300 nl/min flow rate: 5%–11% (v/v) B in 2:49 min,

11%–29% (v/v) B in 18:04 min, 29%–33% (v/v) B in

3:03 min, 33%–39% (v/v) B in 2:04 min, 39%–95%

(v/v) B in 0:10 min, 95% (v/v) B for 2:50 min, 95%–0%

(v/v) B in 0:10 min and 0% (v/v) B for 0:50 min. Column

temperature was kept at 50C using a butterfly heater

(Phoenix S&T, Chester, PA, USA). The Q Exactive™

HF was operated in a data-independent (DIA) manner

in the m/z range of 345–1650. Full scan spectra were

recorded with a resolution of 120,000 using an auto-

matic gain control (AGC) target value of 3 10

with a

maximum injection time of 100 ms. The full scans were

followed by 62 DIA scans of dynamic window widths

using an overlap of 0.5 Th (Doellinger, Blumenscheit,

et al., 2020). DIA spectra were recorded at a resolution

of 30,000 using an AGC target value of 3 10

with a

maximum injection time of 55 ms and a first fixed mass

of 200 Th. Normalized collision energy (NCE) was set

to 27% and default charge state was set to 3. Peptides

were ionized using electrospray with a stainless-steel

emitter, I.D. 30 μm (PepSep, Denmark) at a spray volt-

age of 2.1 kV and a heated capillary temperature

of 275C.

Data analysis and statistical information

Protein sequences of Debaryomyces hansenii

(UP000000599, downloaded 16/10/20), were obtained

from UniProt (UniProt Consortium, 2019). A spectral

library was predicted for all possible peptides with strict

trypsin specificity (KR not P) in the m/z range of 350–

1150 with charge states of 2–4 and allowing up to one

missed cleavage site using Prosit (Gessulat

et al., 2019). Input files for library prediction were gen-

erated using EncyclopeDIA (Version 0.9.5) (Searle

et al., 2018). The data were analysed using the pre-

dicted library with fixed mass tolerances of 10 ppm for

and 20 ppm for MS

spectra using the ‘robust LC

(high accuracy)’quantification strategy. The false dis-

covery rate was set to 0.01 for precursor identifications

and proteins were grouped according to their respec-

tive genes. The resulting pg_matrix.tsv file was used

for further analysis in Perseus (Version 1.6.5.0)

(Tyanova et al., 2016).

The same programme was used to z-normalize pro-

tein abundances followed by ANOVA (FDR =0.01)

and post hoc testing (FDR =0.05). Subsequently, the

abundances of biological triplicates were median aver-

aged, and the relative log2-fold changes of the salt-

containing (saline) treatments compared to the salt-free

control were calculated. The results were filtered for

significant pairs of the salt-free samples and at least

one of the saline treatments and were then plotted into

a hierarchical clustered heatmap. Additionally, volcano

plots have been generated with the same software after

t-test of the z-normalized protein abundances. Protein

groups of interest were annotated and analysed with

the STRING database (https://string-db.org/)

(Szklarczyk et al., 2021) regarding enriched metabolic

pathways and the formation of functional protein

clusters.

RESULTS

In order to distinguish the perchlorate-specific stress

response of D. hansenii to the stress caused by NaCl,

proteomes of cell cultures containing either NaClO

NaCl or no additional salts in growth medium DSMZ

#90 were analysed. Two different salt concentration

regimes were investigated (Figure 1A). At moderate

salt concentrations (1.5 mol/kg NaClO

and 2.4 mol/kg

NaCl), growth was obtained by inoculation with a salt-

free culture to provoke a salt shock response. How-

ever, the highest salt concentrations used in this study

(2.4 mol/kg NaClO

and 3.9 mol/kg NaCl) only enabled

growth when cells were long-term adapted to stepwise

increasing salt concentrations (Figure 1A). All samples

were prepared as biological triplicates and cells were

harvested in the late exponential growth phase in order

to obtain sufficient biomass for protein extraction

(Figure 1B). It should be noted that while 2.4 mol/kg

NaClO

is already close to the growth-limiting NaClO

concentration (similar growth rate to 3.9 mol/kg NaCl),

2.4 mol/kg NaCl represents a readily feasible NaCl con-

centration with a growth rate similar to 1.5 mol/kg

NaClO

In total, 2713 proteins were detected representing a

bulk coding sequence coverage of approximately 43%.

Through analysis of variance (ANOVA, FDR ≤0.01) of

5054 HEINZ ET AL.

the z-normalized protein abundances, the expression

of 1099 proteins was found to be significantly different

between the five different treatment types (one salt-free

control and four salt-exposed treatments). The salt con-

centration (moderate vs. high) had a stronger impact on

the intensity of protein expression than the type of

anion as can be seen from the comparison of protein

abundances of all replicates, which show similar protein

expressions for the same salt concentration regimes

(Figure 2A). This is confirmed by the principal compo-

nent analysis (PCA), which revealed a clear clustering

of the replicates of each treatment in dependence on

salt concentration and type of anions (Figure 2B). While

the physiological response to different salt concentra-

tions clustered along principal component 1 and

explains 55% of the observed differences, the salt spe-

cies had a lower impact on the variability (15%), as

treatments exposed to chloride or perchlorate spread

along the principal component 2.

Post hoc testing (FDR ≤0.05) revealed 1068 pro-

teins to be significantly regulated in at least one of the

salt-exposed samples compared to the salt-free treat-

ment. The log2-fold changes of these proteins in the

saline treatments compared to the salt-free control

were plotted in a heatmap with upregulated proteins

coloured red and downregulated protein shown in

FIGURE 2 Results of the proteomic analyses. (A) Abundances of upregulated (upper plot) and downregulated (lower plot) proteins

expressed in all investigated samples (three replicates for each treatment as indicated in the space between the two plots). (B) Principal

component analysis (PCA) demonstrating clear clustering of all biological triplicates in dependence of salt concentration and type of anion.

(C) Heat map including all proteins passing ANOVA (FDR ≤0.01) and post hoc test (FDR ≤0.05) generated by the Perseus software after

hierarchical clustering. Upregulated proteins (compared to the salt-free treatment) are coloured red and downregulated proteins are shown in

green. Two exemplarily perchlorate-specific clusters are highlighted in pink for upregulated and in cyan for downregulated proteins. (D) Volcano

plot visualizing perchlorate-specific regulated proteins with a high (FDR ≤0.012) and a lower significance (0.012 ≤FDR ≤0.05). Significantly

regulated metabolic pathways were analysed with the STRING database.

PERCHLORATE-SPECIFIC MICROBIAL STRESS RESPONSES 5055

Loading more pages...