Abstract
In their natural habitats, microbes rarely exist in isolation; instead,
they thrive in consortia, where various interactions occur. In this
study, a defined synthetic co-culture of the cyanobacterium S.
elongatus cscB, which supplies sucrose to the heterotrophic P. putida
cscRABY, is investigated to identify potential interactions. Initial
experiments reveal a remarkable growth-promoting effect of the
heterotrophic partner on the cyanobacterium, resulting in an up to 80%
increase in the growth rate and enhanced photosynthetic capacity. Vice
versa, the presence of the cyanobacterium has a neutral effect on P.
putida cscRABY, highlighting the resilience of pseudomonads against
stress and their potential as co-culture partners. Next, a suitable
reference process reinforcing the growth-promoting effect is
established in a parallel photobioreactor system, which sets the basis
for the analysis of the co-culture at the transcriptome, proteome, and
metabolome levels. In addition to several moderate changes, including
alterations in the metabolism and stress response in both microbes,
this comprehensive multi-OMICs approach strongly hints towards the
exchange of further molecules beyond the unidirectional feeding with
sucrose. Taken together, these findings provide valuable insights into
the complex dynamics between both co-culture partners, indicating
multi-level interactions, which can be employed for further
streamlining of the co-cultivation system.
Subject terms: Industrial microbiology, Industrial microbiology,
Microbial communities
__________________________________________________________________
This study investigates interspecies interaction in a synthetic
co-culture consisting of the sucrose-secreting cyanobacterium S.
elongatus cscB and its sucrose-consuming partner P. putida cscRABY by
employing a comprehensive multi-OMICs approach.
Introduction
Over the last two decades, a paradigm shift has started in
biotechnology, expanding beyond the historically-grown focus on
single-species or so-called axenic cultures. This change involves the
development of controllable co-cultures comprising two or more species
within the same reaction vessel. Considering that microbes barely live
separately in nature expands the bioproduction of value-added compounds
through novel pathways and strategies. Furthermore, co-cultures enable
us to understand synergistic effects and uncover otherwise hidden
behaviours, allowing for targeted utilisation of microbial
capabilities^[38]1,[39]2.
The dynamics of a community or a co-culture are determined by
interactions of individual cells, which normally act in their local
niches. Most microbes have evolved to thrive in the presence of
neighbouring species, which can present either potential threats or
offer benefits, and they have adapted accordingly^[40]3. Those
interactions can be divided into four general classes: mutualism,
neutralism, commensalism, and parasitism^[41]4. For instance, a
mutualistic interaction, where all community members profit, can appear
as cross-feeding or syntrophy. Here, one partner does not completely
metabolise a given substrate, which then, in turn, is accessible to
another partner. The latter might remove toxins or harmful gases,
creating the environment needed for the entire community^[42]4.
The design of a synthetic co-culture is not trivial, as multiple
aspects and parameters, such as medium composition, need to be
considered^[43]5,[44]6, and in the best-case scenario, resulting
co-cultures should align with a sustainable bioprocess. In most cases,
it is intended that the microbes live under the premise of the division
of labour, which allows different traits of microbes to complement each
other in a profitable way^[45]3,[46]7. One example of doing so is
pairing heterotrophs with phototrophs^[47]8. This composition of
microbes is prevalent in nature, which can be observed in lichens and
microbial mats^[48]9. Here, the photoautotrophic member, such as
cyanobacteria or algae, uses solar power to fix CO[2] into organic
carbon, a portion of which then in turn is accessible to the
heterotrophic members of the community. In the case of synthetic
co-cultures, the heterotrophic partner can be employed to convert the
carbon source provided into value-added products by engineered and
optimised metabolic pathways. The model organism Synechococcus PCC 7942
naturally accumulates sucrose as a compatible solute when exposed to
elevated NaCl concentrations. This trait was exploited to construct the
sucrose secreting strain S. elongatus cscB by genomic integration of
the cscB gene encoding a H^+/sucrose symporter^[49]10. The resulting
strain secretes sucrose into the surrounding medium with a rate of up
to 28 mg L^−1 per hour^[50]10,[51]11. Up to now, several robust
synthetic co-cultures have been constructed employing S. elongatus
cscB, and valuable compounds like α-amylase, polyhydroxyalkanoates
(PHAs), or isoprene have been successfully produced^[52]11–[53]13. The
metabolic capacity of the co-culture processes could be expanded
depending on the heterotrophic partner used, such as Escherichia coli,
Pseudomonas putida, Bacillus subtilis, or Saccharomyces cerevisiae.
In contrast to natural communities that are highly complex structures
with overlaying interactions, synthetic co-cultures are very well
suited to study interactions between and within the involved
species^[54]5. The inherent definition of a synthetic defined
co-culture is that the partners have not evolved together or, at least,
that the connection between them has not evolved driven by nature.
Therefore, we assume that non-engineered feedback and/or communication
between the partner organisms have their origin in a general answer
adopted from their individual natural habitats. In this study, we set
out to investigate those non-engineered interactions that might guide
us towards a better understanding of synthetic co-cultures in general
and how to design them. To this end, we used a synthetic co-culture
consisting of the cyanobacterium S. elongatus cscB and the
soil-bacterium P. putida cscRABY, which was recently employed for PHA
production in our lab^[55]12. In this co-culture, P. putida cscRABY has
been engineered to transport and metabolise sucrose by the integration
of the cscRABY operon into the chromosome^[56]14. In the work described
here, first, a suitable reference experiment was set up, which allowed
us to compare the co-culture with the respective axenic cultures of S.
elongatus cscB or P. putida cscRABY to identify and investigate the
putative interaction of the co-culture partners with each other.
Furthermore, we set out to analyse the co-culture not only on a
physiological level but also on the transcriptome, proteome, and
metabolome level by employing a multi-OMICs approach, which has been
demonstrated to be a powerful tool to decipher hidden traits of
synthetic communities^[57]13,[58]15.
Results and discussion
Influence of the co-culture partners on each other’s growth
It was frequently observed that cyanobacteria grow more efficiently in
co-cultivation with heterotrophic bacteria both in natural^[59]9,[60]16
and in synthetic co-cultures^[61]11. To this end, we investigated the
growth of both strains in the co-culture compared to the axenic
cultures in different scales and conditions (see Supplementary Note
S[62]1 and Supplementary Fig. [63]S1). To analyse the influence of P.
putida cscRABY on the initial growth of S. elongatus cscB in 12-well
plates at a 1.6 mL scale, we differentiated between a “SuSec-ON” and a
“SuSec-OFF” status of the synthetic connection, brought about by the
inducible exchange of sucrose (Fig. [64]1a, b). In the SuSec-ON
situation, the sucrose secretion by S. elongatus cscB is induced, which
is not the case in the SuSec-OFF situation. Here, an additional batch
of 1 g L^−1 sucrose was added to all cultures, including the axenically
grown S. elongatus cscB, to support heterotrophic growth and to
identify effects independent of the synthetic connection. To
investigate the influence of different inoculation ratios
(phototroph:heterotroph), the co-culture was inoculated with varying
amounts of P. putida cscRABY to reach S. elongatus to P. putida cell
ratios of 1:1, 1:10^−3, and 1:10^−5 and after 24 h cell counts of both
strains were determined.
Fig. 1. Physiological influence of the co-culture partners on each other’s
growth.
[65]Fig. 1
[66]Open in a new tab
a Influence of the heterotrophic partner on the cell count of S.
elongatus cscB in axenic culture and in different co-cultures with
decreasing P. putida cscRABY inoculation density after 24 h when
sucrose secretion is induced (SuSec-ON status) or b when the growth of
the heterotrophic partner is supported by an external sucrose batch
(SuSec-OFF status). Data is normalised to the start cell count (t[0]).
c Influence of the phototrophic partner on the cell count of P. putida
cscRABY after 24 h of growth in axenic culture or in three co-cultures
with decreasing S. elongatus cscB inoculation density in the SuSec-OFF
situation. Data is normalised to the start cell count (t[0]). d
Comparison of P. putida cscRABY cell count after 24 h grown in light
and dark, as axenic-culture and in two co-cultures with different
cyanobacterial inoculation cell counts in the SuSec-OFF situation.
Experimental conditions: 12-well plates with 1.6 ml BG11^+ supplemented
with 150 mM NaCl, 25 °C (or 30 °C), 120 rpm, 22 µE or darkness and no
additional aeration with the addition of 0.1 mM IPTG (for a) or
1 g L^−1 sucrose (for b–d). In all cases, data is derived from n = 3
biologically independent experiments (individual data points shown as
circles, x represent outliers), and error bars represent the standard
deviation (based on a sample) of the replicates, calculated using the
n − 1 method.
As shown in Fig. [67]1, in all co-cultures, cyanobacterial cell counts
were higher compared to the axenic culture, suggesting that the
presence of P. putida cscRABY promotes the initial growth of S.
elongatus cscB (left bar in Fig. [68]1a, b). This effect was
independent of whether the synthetic connection via the sucrose feed
(SuSec) was ON or OFF; however, it was more pronounced in the SuSec-OFF
situation. Higher cyanobacterial cell counts are reached in the
SuSec-OFF case due to a general reduction of growth when sucrose
secretion is induced. An influence of the inoculation ratio can only be
observed in the SuSec-OFF situation, where the positive impact of the
presence of P. putida cscRABY on the initial growth of the
cyanobacterium was less pronounced at an inoculation ratio of 1:1
compared to the situation with fewer P. putida cscRABY cells. A
possible explanation might be that the higher P. putida cscRABY cell
densities reached within these 24 h caused a more substantial shading
on S. elongatus cscB, reducing light availability for photosynthesis in
the cyanobacterium.
In a co-culture study by Hays et al., S. elongatus cscB was found to
have a significant negative impact on various heterotrophs,
particularly on the gram-positive bacterium Bacillus subtilis^[69]11.
In contrast to this, we could not identify an apparent effect of high
densities of S. elongatus cscB on the growth of P. putida cscRABY
within 24 h, and, if any, there might be a tendency towards weaker
growth of P. putida cscRABY when inoculated with less of cyanobacterial
cells (Fig. [70]1c). In previous studies, reactive oxidative species
(ROS:
[MATH: O2
− :MATH]
,
[MATH: OH⋅
:MATH]
,
[MATH: H2<
mrow>O2 :MATH]
) produced by S. elongatus cscB were shown to be the most invasive
substances for heterotrophic growth^[71]11,[72]13,[73]17. Thus, as ROS
are side products of photosynthesis, we performed the co-cultivations
with or without light in 12-well plates at 1.6 mL scale, and compared
them to axenic cultures of P. putida cscRABY grown under equal
conditions. Two different cyanobacterial cell densities were used for
inoculation in the co-culture, as shown in Fig. [74]1d. Still, no
statistically significant difference in the heterotrophic growth could
be detected (unpaired T-test, α = 0.05). Therefore, the formation of
ROS through photosynthesis had no detectable influence on the growth of
P. putida cscRABY under the conditions tested.
Influence of illumination, induction, and inoculation time of P. putida
cscRABY on S. elongatus cscB
To analyse the co-culture in more detail, we switched the cultivation
to the HD-9.100 CellDeg platform system, which permits parallel
co-cultivations under comparable conditions and ensures high
reproducibility. We started by analysing the effect of the illumination
profile, the induction of sucrose secretion, and the time point of
induction on the growth of S. elongatus cscB (Fig. [75]2). With a
constant illumination of 150 µmol m^−2 s^−1 and without IPTG induction,
S. elongatus cscB grew with a rate of 0.064 ± 0.001 h^−1. The presence
of 0.1 mM IPTG in the culture, however, nearly halved the growth rate
to 0.036 ± 0.004 h^−1 (Fig. [76]2a and Supplementary Note S[77]2). This
effect is not due to negative feedback of the sucrose accumulated in
the medium, as confirmed by growing the cells in the presence of
sucrose (Supplementary Note S[78]3 and Supplementary Fig. [79]S2), but
rather suggests a rechannelling of the fixed carbon into sucrose
secretion instead of biomass formation. This is in accordance with
previous studies, where it was reported that induction of the CscB
symporter reduces the biomass accumulation of S. elongatus cscB while
increasing the total carbon fixation^[80]10,[81]18.
Fig. 2. Influence of induction of sucrose secretion and of the illumination
profile.
[82]Fig. 2
[83]Open in a new tab
a Axenic growth of S. elongatus cscB with different time points of
induction of the cscB expression (non-induced, initially induced, or
induction after two days) with constant light at 150 µmol m^−2 s^−2 and
b with an exponential light setting. With an exponential light profile,
initial induction of sucrose secretion led to photobleaching and the
cultivation was stopped after ~90 h. The grey area indicates the data
points used to calculate the growth rates (Supplementary Note [84]S2).
The arrows mark the time points of induction. c Growth of axenic S.
elongatus cscB with and without induction of cscB expression, and of
the co-cultures inoculated with P. putida cscRABY after 16 h (1* = day
1), 46 h (2* = day 2), or after 65 h (3* = day 3) under constant light
conditions and d with an exponential light setting. The grey area
indicates the data points used for calculating the growth rates shown
in Table [85]1. e Ratio of phototroph:heterotroph over time with an
exponential light profile and inoculation of P. putida cscRABY to the
co-culture at day 1 or day 2. Individual data points are represented by
white circles. Experimental conditions a, b, e: BG11^+ supplemented
with 150 NaCl and 0.1 mM IPTG, when cscB expression was induced.
Constant light: 150 µmol m^−2 s^−2 for 160 h; Exponential light: 24 h
at 120 µmol m^−2 s^−2 followed by an exponential rising with a doubling
time t[d] = 52 h. Experimental conditions c, d: 25.5–34 °C, 2% CO[2],
BG11^+ + 150 mM NaCl, volume 95 mL; Const.120 = constant light profile
with 120 µmol m^−2 s^−2 and Const.50 = constant light profile with
50 µmol m^−2 s^−2, exponential light: 120 µmol m^−2 s^−2 constant for
24 h followed by exponential rising with t[d] = 52 h. Data is derived
from n = 3 biologically independent experiments, and error bars
represent the standard deviation (based on a sample) of the replicates,
calculated using the n − 1 method.
Next, we analysed the influence of an exponential light profile on the
cyanobacterial growth behaviour (Fig. [86]2b). As observed with the
constant illumination, induction of the sucrose secretion resulted in a
decreased growth rate (Supplementary Table [87]S1). However, here, the
effect of induction was way more severe than in conditions of constant
light, as cultures did not just have reduced growth rates but also went
into a state of photobleaching manifested as visible pigment loss after
80 h at ~300 µmol m^−2 s^−1. By shifting the induction to day 2,
instead of adding the inducer from the beginning of the cultivation,
this effect could be avoided, and growth rates reverted to what was
observed in the absence of IPTG (Supplementary Table [88]S1).
Furthermore, a higher amount of sucrose could be measured in the
supernatant of the culture, in which cscB expression was induced on day
2. These cultures accumulated three times more sucrose after 63.5 h
than those growing from the beginning in the presence of the inducer.
This trend persisted, and after 87.5 h, the level of sucrose
accumulated had risen to 0.75 ± 0.38 g L^−1, while the cultures grown
in the presence of IPTG from the beginning exhibited only a slight
increase (see Supplementary Note S[89]4 and Supplementary Fig. [90]S3).
Taken together, the interplay between the time point of induction and
the illumination profile emerged as a significant factor influencing
the growth behaviour of the cyanobacterium. We hypothesise that a
prolonged phase after inoculation without induction of cscB expression
facilitated a better adaptation of S. elongatus cscB to the
environmental conditions^[91]19. This photoacclimatisation, in turn,
could lead to notable differences in photosynthetic activity, resulting
in enhanced growth, increased sucrose accumulation, and protection
against photobleaching.
After analysing the cyanobacterial growth in axenic cultures, we set
out to study the co-cultivation with P. putida cscRABY. We observed
that under the detrimental conditions of an exponential light profile
and initial induction of sucrose secretion, the presence of P. putida
cscRABY in the co-culture rescued cscB-expressing S. elongatus cscB
from photobleaching and also led to a higher cyanobacterial growth rate
(Fig. [92]2c). Additionally, the time point of inoculation of P. putida
cscRABY had an influence on the growth behaviour and growth rate of S.
elongatus cscB. The addition of P. putida cscRABY to the culture within
the first 50 h (inoc. day 1, inoc. day 2) had a positive effect on the
growth of the cyanobacterium. For instance, when P. putida cscRABY was
inoculated on day 1, S. elongatus cscB exhibited a 32% higher growth
rate in comparison to the cultures not expressing cscB and a 61% higher
growth rate compared to the cyanobacterial cells expressing cscB
(Table [93]1). Inoculation on day three, however, could no longer
restore the growth behaviour to the one observed with S. elongatus cscB
grown without IPTG.
Table 1.
Growth rates of S. elongatus cscB and P. putida cscRABY in axenic
cultures and co-cultures with constant illumination (Const.120:
120 µmol m^−2 s^−2, Const.50: 50 µmol m^−2 s^−2) or an exponential
light profile (Expo.) and different inoculation time points (Inoc.) of
the heterotrophic partner; mean and standard deviation are derived from
three different cultures cultivated in parallel
Condition S. elongatus growth rate, h^-1
Light Inoc. Non-induced Axenic, induced Co-culture
Const.120 Day 1 n.p.^a 0.027 ± 0.006 0.065 ± 0.005
Const.50 Day 1 0.020 ± 0.003 0.010 ± 0.001 0.056 ± 0.003
Expo. Day 1 0.057 ± 0.002 0.033 ± 0.001 0.084 ± 0.001
Expo. Day 2 0.055 ± 0.002 0.044 ± 0.006 0.066 ± 0.02
Expo. Day 3 0.051 ± 0.002 0.031 ± 0.006 0.046 ± 0.002
[94]Open in a new tab
^aNot performed.
These observations prompted the question of why the presence of P.
putida cscRABY has a growth-promoting effect on S. elongatus cscB. One
possible explanation could be the physical protection from high light
intensities. For S. elongatus PCC 7942, the parental strain of the
derivative used in this study, light intensities higher than 400 µmol
m^−2 s^−1 are already considered as high light and intensities of
200 µmol m^−2 s^−1 can induce oxidative stress, which is noticeably
lower than for other cyanobacteria^[95]20,[96]21. To test this
hypothesis, we chose two non-harmful constant light intensities,
50 µmol m^−2 s^−1 (low) and 120 µmol m^−2 s^−1 (medium), and compared
the growth of S. elongatus cscB in axenic culture to that in the
co-culture (Fig. [97]2d). We observed a similar growth-promoting effect
of the presence of the co-culture partner, as described above with the
exponential light profile, even though the presence of the
heterotrophic partner under these low light conditions should rather
hamper the growth of the cyanobacterium due to shading from light. For
instance, with a constant illumination of 50 µmol m^−2 s^−1, the
presence of P. putida cscRABY prevented S. elongatus cscB from an early
entry into the stationary phase and increased its growth rate by 64%
compared to the cells not expressing cscB and even by 82% compared to
cells expressing cscB (Table [98]1). The same trend was observed with a
medium light intensity. From these experiments, we concluded that the
growth-promoting effect of the presence of P. putida cscRABY on S.
elongatus cscB cannot be reduced to the mere protection from high light
intensities but rather is the result of a more complex interplay of
different factors. As S. elongatus cscB grown under non-inducing
conditions was not drastically affected by the different light
profiles, we assumed that sucrose secretion contributes to the stress
perceived by S. elongatus cscB that finally leads to photobleaching and
reduced growth.
In order to see whether the population ratio provides a hint at the
origin of the positive effect on the cyanobacterium, we determined the
cell ratio of phototrophic to heterotrophic cells in the co-cultures
(Fig. [99]2e). Although the initial inoculation ratio was the same with
34% of P. putida cscRABY to 66% of S. elongatus cscB, varying
phototroph:heterotroph cell ratios adjusted themselves over time,
consequently leading to different shading conditions in each
experiment. Nevertheless, the overall behaviour of the co-cultures was
comparable irrespective of the time point of addition of the
heterotrophic partner (compare Fig. [100]2c), hinting again towards
more complex processes being involved in the growth promoting effect
observed by the addition of P. putida cscRABY. In order to gain a
deeper understanding of these processes and the possible interplay
between P. putida cscRABY and S. elongatus cscB, which extends beyond
the mere exchange of sucrose, we chose the exponential light profile
for a comprehensive OMICs-driven investigation.
Co-culture reference experiments for multi-OMICs analysis
To analyse the interplay between P. putida cscRABY and S. elongatus
cscB and to find a hint at what might be the origin of the remarkable
increase in the growth rate of S. elongatus cscB grown in co-culture,
we aimed to compare the co-culture with the axenic cultures of S.
elongatus cscB and P. putida cscRABY, respectively. Therefore, we set
up a reference procedure in the 9-fold parallel photobioreactor system,
which allowed us to have comparable conditions in all three different
setups, each in biological triplicates, for the analysis of the
transcriptome, proteome, and metabolome (Fig. [101]3a).
Fig. 3. Reference experiment for the multi-OMICs analysis.
[102]Fig. 3
[103]Open in a new tab
a Schematic visualisation of the reference experiment consisting of
three different settings, each performed in triplicates. The settings
are: Axenic S. elongatus cscB culture (green), the co-culture (green
and yellow), and P. putida cscRABY axenic culture (yellow). Red stages
represent the feed with sucrose for axenically grown P. putida cscRABY.
b Two reference experiments (Experiment I - light blue and Experiment
II - dark blue) with the co-culture of S. elongatus cscB and P. putida
cscRABY (shaded grey) and the respective axenic cultures. Depicted is
the normed cell count; growth rates of P. putida cscRABY in the
co-culture and the axenic culture are represented as small bar charts
(calculated between 18–65 h). The dashed boxes indicate the sampling
points for the multi-OMICs. Experimental conditions: 25–32 °C, 2%
CO[2], 95 mL BG11^+ + 150 mM NaCl, volume 95 mL; exponential light
profile: 120 µmol m^−2 s^−2 constant for 24 h followed by exponential
rising with t[d] = 52 h. Data is derived from n = 3 biologically
independent experiments, and error bars represent the standard
deviation (based on a sample) of the replicates, calculated using the
n − 1 method.
To achieve comparable growth rates of P. putida cscRABY in axenic
cultures and in the co-culture, we adjusted an external sucrose feed
for axenic P. putida cscRABY cultivations that mimicked the
cyanobacterial sucrose secretion. In order to calculate the sucrose
feeding rate, the biomass formation was connected with the sucrose
uptake by the heterotrophic partner (calculations see Supplementary
Note S[104]5). However, achieving consistency was challenging as the
growth rate of P. putida cscRABY in the co-culture showed a high
variance between different experiments, assumingly due to variations in
the temperature. Therefore, for the OMICs, two experiments that showed
comparable temperature profiles (see Supplementary Note S[105]6 and
Supplementary Fig. [106]S4) have been chosen, from now on referred to
as EI and EII (Fig. [107]3b). Samples for transcriptomics and
proteomics were derived from EII and samples for metabolomics from EI
at 27.0 °C or 27.4 °C, respectively. Thus, for each of the OMICs
experiments, external conditions, such as light or temperature, were
identical, as all samples were taken from the same experiment conducted
in the 9-fold parallel photo-bioreactor. Growth of the phototrophic
partner was highly reproducible in the co-culture compared to the
axenic culture (Fig. [108]3b, Supplementary Note S[109]7 and
Supplementary Fig. [110]S5) in both experiments. As observed before,
the growth was enhanced in the co-cultures compared to the axenic
cultures. An early entry of the cyanobacterial cells into the
stationary phase and visible photobleaching after 85 h was prevented.
Furthermore, cells of S. elongatus cscB grown in co-culture had a
smaller size, which fits well with the more than doubled specific
growth rate (Table [111]S2 and Supplementary Note S[112]8).
The growth of P. putida cscRABY depends on the cyanobacterial sucrose
secretion in the co-culture and on the sucrose feed in the axenic
culture. Although the growth rates of P. putida cscRABY were different
between both experiments, they did not differ within the same
experiment (Supplementary Table [113]S2). This was important for the
-OMICs, as these were the conditions to be compared. No differences in
the cell size of P. putida cscRABY in co-cultures or axenic cultures
could be observed, which aligns with the expected outcome due to the
same growth rate (Supplementary Fig. [114]S6). We assumed sucrose to be
the growth limiting factor in both cultures, and in fact, in the axenic
cultures of P. putida cscRABY and the co-culture, no sucrose could be
detected at the time point when the samples for the OMICs were taken
(Supplementary Fig. [115]S7).
As P. putida cscRABY is limited by the carbon source in all cultures,
we set out to analyse the medium components citrate and phosphate and
other commonly known overflow metabolites of P. putida. We detected a
transient accumulation of acetate and ethanol, but both substances were
completely taken up again at the end of the process, assumingly by P.
putida cscRABY itself, as S. elongatus cscB does not harbour typical
genes for acetate uptake. The BG11^+ medium contains citric acid, which
was completely consumed within 13.5 hours in all cultures. However,
after 40 h, citric acid concentration increased again, particularly in
cultures with P. putida cscRABY. At first glance, this is
counterintuitive, as under carbon limitation, cells should coordinate
their energy and maintenance needs. However, a recent study showed that
E. coli secrets overflow metabolites to address carbon-nutrient
imbalances^[116]22. Thus, apart from resulting from cell lysis, this
might also be the case here (Supplementary Fig. [117]S7).
A gross estimation of the photosynthetic activity of S. elongatus cscB
is possible, as the amount of carbon fixed can be approximated by
taking into account biomass production and sucrose accumulation
(Supplementary Note S[118]10). It has already been described by others
that inducing sucrose secretion resulted in increased overall CO[2]
fixation in S. elongatus cscB^[119]10,[120]18,[121]23. This is
explained by the idea that heterologously implemented sucrose
production serves as a sink for the carbon captured in the Calvin
cycle, thereby alleviating the so-called photosynthetic sink
limitation. Sink limitation describes the situation when photosynthetic
activity is reduced due to insufficient withdrawal of products from the
Calvin cycle^[122]17. In the co-culture, the constant pull on the
sucrose production by the heterotroph seems to lead to a more efficient
utilisation of the captured carbon and thereby contributes to
overcoming sink limitation. We observed that in the co-culture, the
photosynthetic activity of S. elongatus cscB was even higher than under
inducing conditions, as the growth rate surpassed that of the cells not
expressing cscB, and as additionally heterotrophic growth was supported
by sucrose secretion (Supplementary Table [123]S3 and Supplementary
Fig. [124]S8).
Overview of multi-OMICs in the co-culture process
As described above, the presence of P. putida cscRABY in the co-culture
had a positive effect on the growth of S. elongatus cscB. In the next
step, we aimed to get insights into the inter-species interaction in
the co-culture. This is not only interesting from a fundamental
research perspective but will also contribute to enhancing our
understanding of co-culture stability and, eventually, even help to
improve the production of value-added products in co-cultures in
general. Therefore, transcriptomics, proteomics, and metabolomics were
performed (see Supplementary Note S[125]11). Samples for multi-OMICs
were taken at ~60 h and processed as described in the methods section.
The time point was chosen to be in the second half of the growth phase
before cells entered the stationary phase to ensure a sufficiently high
number of cells for the analysis (compare Fig. [126]3a). The co-culture
was considered as the case of interest and compared to the axenic
cultures, which were considered as the controls. This provided us with
a snapshot to describe the cellular status at the time of sampling. In
all the datasets, we could pinpoint distinct clusters that were
specific for either the co-culture or the axenic cultures,
respectively. Exemplarily, this is demonstrated in Fig. [127]4a, b for
the metabolome data, which was used for a principal component analysis.
Thresholds for differently expressed genes (DEGs) and differently
abundant proteins (DAPs) were set at |log[2](FC)| >1.0 and a p-value
(adjusted, false discovery rate (FDR) corrected) <0.05. For different
abundant metabolites, the threshold was the adjusted p-value of 0.05
and a mean difference of 0.3.
Fig. 4. Principal component analysis (PCA) of metabolites.
[128]Fig. 4
[129]Open in a new tab
a RP-MS and b HILIC-MS measurements. The positive mode is depicted in
circles, and the negative mode is depicted in squares. Green = S.
elongatus cscB, yellow = P. putida cscRABY, blue = co-culture,
purple = controls. c, d Volcano plots of the transcriptomics. e, f
Volcano plots of the proteomics. In c and e, the co-culture is compared
to the axenically grown P. putida cscRABY; in d and f, the co-culture
is compared to the axenically grown S. elongatus cscB. Transcripts and
proteins in red circles meet the threshold for both log[2]-FC and
p-value. Transcripts and proteins depicted in blue fulfil the threshold
only for the p-value, while transcripts or proteins in green meet the
threshold only for log[2]-FC. Grey represents transcripts and proteins
that do not meet both thresholds. The thresholds are indicated in all
plots by dashed lines.
Comparing the transcriptome of the co-culture to the axenic culture, in
P. putida cscRABY, a total of 488 differently expressed genes (DEGs)
were identified, of which 303 were up-regulated, and 145 were
down-regulated (Fig. [130]4c). In S. elongatus cscB a total of 790 DEGs
were identified. Of these 324 genes were found to be up-regulated,
while 466 were down-regulated in comparison to axenically grown cells
(Fig. [131]4d). Gene set enrichment analysis based on the KEGG
Orthology database indicated that in P. putida cscRABY, pathways
involved in arginine biosynthesis, carbon metabolism, glyoxylate and
dicarboxylate metabolism, as well as two-component systems, were
impacted by the presence of the co-culture partner (Supplementary
Fig. [132]S9). For S. elongatus cscB, pathways involved in
arginine/proline metabolism, photosynthesis, pyruvate, glycolipid, and
biosynthesis of secondary metabolites were identified in the analysis
to be the most affected ones.
The analysis of the proteome revealed 69 proteins in P. putida cscRABY
to be more abundant, and 27 proteins showed less abundance when
compared to the axenic culture (Fig. [133]4e). In S. elongatus cscB, a
total of 92 proteins were identified to be more abundant in co-culture,
and 91 proteins were less abundant (Fig. [134]4f). The majority of
proteins identified in P. putida cscRABY belong to the category amino
acid metabolism and transport, and the majority identified in S.
elongatus cscB belong to the group of photosynthesis or the category
stress. This trend was also observed in the transcriptome (see above).
In general, the match between transcriptomics and proteomics regarding
the identification of processes and direction of regulation is the
range of what is described as the regular magnitude^[135]24
(Supplementary Fig. [136]S10).
Analysing the metabolome, in total, 876 features could be identified in
the cells with the HILIC-MS measurements (− and + MS-mode), and 1013
features were identified with RP-MS (− & + MS-mode). When comparing the
features obtained in the co-culture grown cells to those identified P.
putida cscRABY grown in axenic culture, 336 features for HILIC (− & +
MS-mode) and 427 for RP (− & + MS-mode) fulfilled the conditions set
(see Supplementary Fig. [137]S11 for Volcano plots). A comparison of
the features identified in the co-culture grown cells to those obtained
in S. elongatus cscB grown in axenic culture revealed 143 features that
fulfilled the threshold set for HILIC-MS measurement (− & + MS-mode)
and 254, which fulfilled it for the RP-MS (− & + MS-mode). Most
features were more abundant in the co-culture than in the respective
axenic culture. By reference measurements, some metabolites could be
identified (Supplementary Fig. [138]S12). Most of them participate in
sugar metabolism or belong to the group of phospholipids, amino acids,
or fatty acids.
Cellular processes affected by the co-culture partner: Core metabolism and
photosynthesis
The analysis of the multi-OMICs data yielded a large number of genes
and proteins that were differentially expressed or abundant when
comparing co-cultivated cells to those cultivated in axenic cultures.
Additionally, on the metabolome level, some differences were
identified. To get a first overview of the cellular processes that were
mainly affected by the presence of the respective co-culture partner,
we sorted the DEGs, DAPs, and metabolites into different groups
according to their putative function (Fig. [139]5). In P. putida
cscRABY the presence of the phototrophic partner led to changes in
various cellular processes, namely in the core metabolism, transport of
amino acids, nitrogen, small acids, and sugars, but also in the general
stress response, detoxification, and degradation. In S. elongatus cscB,
we found that the presence of the heterotrophic partner likewise had an
effect on the core metabolism but also on photosynthesis, which was
somehow expected as S. elongatus cscB exhibited a higher growth rate in
the co-culture. Furthermore, other processes connected to stress,
detoxification, or transport of sulphur and iron were affected. In the
following, these processes will be discussed in more detail.
Fig. 5. Overview of the cellular processes affected by the co-cultivation.
Fig. 5
[140]Open in a new tab
In S. elongatus cscB (green), the major impact of the presence of the
heterotrophic partner was found in the core metabolism and
photosynthesis. Furthermore, stress or detoxification-related processes
and the transport of sulphur and iron were also affected. In P. putida
cscRABY (yellow), processes in the core metabolism, transport of amino
acids, nitrogen, small acids, and sugars, as well as in stress,
detoxification and degradation were affected by the presence of the
phototrophic partner.
In P. putida cscRABY several genes, proteins, and metabolites with a
potential function in the core metabolism were identified to be
affected by the presence of S. elongatus cscB in the co-cultivation.
More specifically, on the proteome as well as on the transcriptome
level, processes that are connected to the amino acid (AA) synthesis or
degradation, the TCA cycle, or to the fatty acid (FA) metabolism were
affected (Fig. [141]6a, b). Additionally, some genes encoding proteins
involved in the Entner-Doudoroff-Embden-Meyerhof-Parnas (EDEMP) cycle
were differentially regulated (Fig. [142]6b). On the metabolome level,
the metabolites identified mainly belonged to the group of amino acids
(Fig. [143]6c). As a rhizobacterium, P. putida is specialised for the
uptake and metabolisation of amino acids^[144]25. In line with this, it
was not surprising that the expression of genes, as well as the
abundance of proteins connected to amino acid metabolism, was affected
by the presence of the co-culture partner. On the proteome level, the
asparagine synthetase AsnB, responsible for the conversion of aspartic
acid into asparagine, showed the biggest differences with a log[2]-FC
3.9 (Fig. [145]6a). The corresponding transcript asnB, however, was
down-regulated (Fig. [146]6b, PP_2453 log[2]-FC -3.1). This
demonstrates a prevailing issue of multi-OMICs analysis, which is that
the correlations between proteomics and transcriptomics are only
moderate^[147]24. However, these differences may also indicate a
post-transcriptional control. On the metabolome level, the amino acids
L-phenylalanine, L-glutamic acid, L-aspartic acid and L-glutamine were
identified to be more abundant in the co-culture compared to P. putida
cscRABY grown axenically (Fig. [148]6c and Supplementary
Fig. [149]S12). However, the metabolites were identified in all cells
grown in co-culture and, therefore, cannot be assigned specifically to
one of the co-culture partners. Another sector of the core metabolism
that was affected is the TCA cycle. On the proteome level, lower
protein abundances of Idh and SucA/B/D went together with a higher
abundance of the proteins AceA and GlcB, hinting towards a shut-down of
the TCA cycle and a redirection of the metabolic flux through the
glyoxylate cycle (Fig. [150]6a). This is known to happen in P. putida
when degradation of aromatics or xenobiotics is necessary ^[151]26. In
the transcriptome, the contrary is observed with a slight up-regulation
of Idh (PP_4012 log[2]-FC 1.9) combined with the down-regulation of the
transcript encoding AceA (PP_4116 log[2]-FC -2.64).
Fig. 6. Changes in the central carbon metabolism of P. putida cscRABY grown
in co-culture.
[152]Fig. 6
[153]Open in a new tab
a Proteome level: Differentially abundant proteins (DAPs) are
highlighted in orange (more abundant) or grey (less abundant).
Superscript numbers indicate the log[2]-fold change (log[2]-FC). b
Transcriptome level: DEGs encoding proteins putatively involved in AA
(amino acid) synthesis and degradation, or the TCA (tricarboxylic acid)
cycle, the EDEMP (Entner-Doudoroff-Embden-Meyerhof-Parnas)-cycle, or
the FA (fatty acid) synthesis and degradation. The blue shaded area
indicates the p-value of 0.05. c Metabolome level: metabolites
identified by reference measurements comparing the co-culture to P.
putida cscRABY and grouped into pathways by a pathway enrichment
analysis. Data was obtained by using the metabolic pathway analysis of
MetaboAnalyst 5.0. Abbreviations: 1 = Phe/Tyr Metabolism, amino acids
are abbreviated with the three-letter code, and M stands for
metabolism.
Taken together, the core metabolism of P. putida cscRABY is influenced
by the presence of the co-culture partner, particularly affecting
processes belonging to the amino acid metabolism and the TCA cycle. In
general, the central metabolism can reflect different metabolic states
of the cell, as it was described for cells growing on mixtures of
carbon sources^[154]26. Differences in the environmental conditions,
brought about by cultivation in the co-culture, might, therefore, lead
to changes in the core carbon metabolism or in its periphery, such as
the amino acid or fatty acid metabolism. Alterations in the metabolism
were further corroborated by the identification of secretion and
re-uptake of some metabolites, such as transient accumulation of
ethanol, acetate, and citrate in the supernatant during different
cultivation phases (Supplementary Note S[155]9).
When looking at the processes affected in S. elongatus cscB by the
presence of P. putida cscRABY, it has to be kept in mind that the
growth rates of axenically grown cells and cells grown in co-culture
differ by a factor of about two. Variations in transcripts or proteins
can be the consequence of the higher growth rate or arise from the
interaction with the co-culture partner, which encompasses specific
interactions as well as non-specific effects, such as shading or the
response to metabolic signals, which might arise from secreted
metabolites and/or consumed resources. These effects might also be
entangled, as a positive interaction could lead to a higher growth
rate. The higher growth rate of S. elongatus cscB in the co-culture is
reflected by an up-regulation of many growth-associated genes, such as
ribosomes, tRNAs, and polymerases, as it was observed at the
transcriptional level (Supplementary Note S[156]12 and Supplementary
Fig. [157]S13). This was not the case for P. putida cscRABY, showing
that a similar growth rate in axenic culture and co-culture leads to a
similar expression pattern of these genes. The higher abundance of
amino acids in the metabolome of the co-culture, which was already
mentioned above, could also be the consequence of the increased
metabolic activity of the cyanobacterium. However, attributing core
metabolites to a specific co-culture partner is not possible. Analysing
the core metabolism of S. elongatus cscB in more detail, only a few
proteins showed different abundance. They can be grouped into proteins
being involved in the porphyrin metabolism, the Calvin cycle, or
photosynthesis (Fig. [158]7a). Interestingly, the heterologously
expressed sucrose transport protein CscB was less abundant in S.
elongatus cscB when grown in co-culture. This is on the first glace
counterintuitive, as its transcription is regulated by the IPTG
inducible P[lacUV5] promoter and should, therefore, be constant. No
information on its transcript level is available, as the cscB gene was
not included in the transcriptomic analysis. We explain the lower
abundance of the CscB protein by the higher growth rate of S. elongatus
cscB in the co-culture. Assuming the total amount of the CscB protein
produced remains constant, but cells divide more rapidly, the CscB
protein is distributed across a larger number of cells, which results
in a lower abundance. In the metabolome of the co-culture grown cells,
we detected a higher amount of disaccharide, which could be sucrose,
which might hint towards a higher accumulation of sucrose in the
cytoplasm of the cyanobacterial cells due to decreased export activity.
Other metabolites identified in the metabolome of co-culture grown
cells include compounds assigned to the biosynthesis of amino acids or
purine and pyrimidine metabolism (Fig. [159]7c).
Fig. 7. Changes in the central carbon metabolism and photosynthesis of S.
elongatus cscB.
[160]Fig. 7
[161]Open in a new tab
a Proteome level: Schematic cell of S. elongatus cscB with changes in
the amino acid (AA) metabolism, porphyrin metabolism, (CC) Calvin
cycle, and in photosynthesis (TCA = tricarboxylic acid cycle). DAPs are
shown in orange (more abundant) or grey (less abundant). Superscript
numbers indicate the log[2]-fold change (log[2]-FC). b Changes in the
photosynthesis apparatus on the proteome level: PSII = photosystem II,
PQ = plastoquinone, Cyt-bf[6] = cytochrome bf[6], PC = Plastocyanin,
PSI photosystem I, FNR = ferrodoxin-NADP^+ reductase, and
Fd = ferrodoxin/flavodoxin. DAPs are shown in orange (more abundant) or
grey (less abundant). Superscript numbers indicate the log[2]-fold
change (log[2]-FC). c Metabolome level: Data were obtained by using the
metabolic pathway analysis of MetaboAnalyst 5.0. Abbreviations:
1 = Glutathione metabolism, 2 = Glyoxylate and dicarboxylate
metabolism. d Changes in the photosynthesis apparatus on the
transcriptome level: DEGs are subdivided into the groups pigments &
PSII, ETC = electron transfer chain, and PSI. The blue shaded area
indicates a p-value of 0.05.
Looking at genes and proteins involved in photosynthesis, the effect of
the different growth rates, when grown axenically or in co-culture,
becomes even more obvious. In the proteome, the most pronounced change
was the increased abundance of phycobiliproteins. Additionally, the
pigment-proteins phycocyanin and allophycocyanin, which are present in
the light-harvesting complex, were also more abundant (Fig. [162]7b).
In general, the light harvesting complexes are connected to
photosynthetic activity and growth. However, here, only a few proteins
of the Photosystem I (PSI), Photosystem II (PSII), and the connecting
electron chain consisting of the NAD(P)H-dehydrogenase-like complex
(NDH) and Plastocyanin (PQ) were detected. In the transcriptome, the
opposite effect was observed: Genes encoding proteins forming the PSI
and PSII and the phycobiliproteins were down-regulated, whereby genes
coding for the connecting NDH/PQ complex were up-regulated
(Fig. [163]7d). A likely explanation for the difference observed is
provided again by the different growth rates observed in the two
culture conditions. The axenically grown cells of S. elongatus cscB
displayed reduced growth and manifested phenotypically evident stress
effects. At the time point of sampling, cells grew linearly, suggesting
a non-constant growth rate and, consequently, a dynamic state of
cellular processes, whereas the cells grown in co-culture exhibited
exponential growth with a constant growth rate, indicating an
intracellular steady state of transcripts, proteins, and metabolites.
As a result, transcripts and proteins may be differently affected when
comparing the co-culture to the axenic cultures of S. elongatus cscB.
Photosynthesis is highly regulated, for instance, by the PSI:PII
ratio^[164]27. In order to cope with excess energy, photosynthetic
organisms regulate their electron transport chain (ETC) to prevent the
production of ROS. Another mechanism for encountering photooxidative
stress in high-light conditions involves the protein pair IsiA and
IsiB. Both proteins were more abundant in the co-culture, IsiA with a
log[2]-FC of 5.7 and IsiB with a log[2]-FC of 4.5, which was amongst
the highest increases detected at the protein level (Fig. [165]7b).
IsiA is annotated as an iron stress induced chlorophyll-binding protein
and IsiB as a flavodoxin. Consistent with iron induced stress, two
other proteins linked to iron limitation, IdiA and IrpA were notably
more abundant in the co-culture grown cyanobacteria (see Table [166]2).
However, we could not detect a stronger iron limitation for S.
elongatus cscB in the co-culture, as iron supplementation or limitation
had no discernible effect on the cultures (Supplementary Note S[167]13
and Supplementary Fig. [168]S14), nor was the higher protein abundance
of IsiA and IsiB reflected in the transcriptome. Additionally, at the
proteome level, some proteins associated with porphyrin biosynthesis
were identified, with the majority showing higher abundance
(Fig. [169]7a), potentially linked to observed variations in
photosynthesis.
Cellular processes affected by the co-culture partner: Transport
Microbes often rely on their capacity to efficiently utilise a wide
range of resources, which can be a critical factor in their
competitiveness and overall performance in relation to other
microorganisms. Consequently, they have developed numerous strategies
for acquiring compounds from their surrounding environment. Most
microbial interactions require a form of uptake of substrates, signals
in diverse forms, or toxins. In line with this, we have identified
various transporters to be differentially regulated in the co-culture
versus axenic cultures (Fig. [170]8). In general, we observed that in
P. putida cscRABY, transporters are more likely up-regulated, whereas
the opposite is the case in S. elongatus cscB. This might originate in
the individual lifestyles of each of the co-culture partners, as a
strict autotrophic and anabolic mode needs less transport of organic
carbon compounds compared to a heterotrophic lifestyle.
Fig. 8. Differential expression of transporters in the co-culture.
[171]Fig. 8
[172]Open in a new tab
a DEGs in P. putida cscRABY grouped according to the transported
substrate: AA (amino acids), nitrogen, small acids, and sugars. The
blue area indicates a p-value of 0.05. b DEGs in S. elongatus cscB
classified into two major groups: iron and sulphur transport. The blue
area indicates a p-value of 0.05. c Proteome level: DAPs in P. putida
cscRABY (above) and S. elongatus cscB (below) connected to transport.
The colour bar indicates the changes in the protein abundance.
Superscript numbers indicate the log[2]-fold change (log[2]-FC).
In P. putida cscRABY at the transcriptome level, the DEGs related to
transport are diverse and include genes encoding putative transporters
for amino acids, nitrogen, small acids, and sugars (Fig. [173]8a).
While in the group of amino acid transporters, the transcription of the
corresponding genes showed regulation in both directions, in the group
of nitrogen transport transcription was down-regulated and in the
groups of small acids and sugar transport, transcription was mainly
induced. The most highly up-regulated transcript was PP_4604, found in
the group of amino acid transport. It encodes a putative transporter
belonging to the EamA family, which in E. coli is related to transport
of cysteine-derivatives^[174]28. Directly downstream of this gene, a
gene encoding an AraC-type regulator (PP_4605, log[2]-FC 5.3), was
found to be highly up-regulated as well. A putative connection between
these two genes is predicted by the string-database. Down-regulation
was observed for some genes encoding ABC-transporters for
glutamate/aspartate uptake (gltJ log[2]-FC -2.6, gltP log[2]-FC −2.4,
and gltK log[2]-FC −2.7), whereas genes encoding proteins connected to
the transport of other amino acids derivatives, such as putrescine or
spermidine (potA ATB-binding and PP_0412 substrate binding with a
log[2]-FC of 1.5 and 1.0) were up-regulated. At the proteome level, the
latter one, PP_0412, was also identified to be more abundant.
Additionally, the proteins SpuD and YhdW, annotated as polyamide
transporter, were found to be more abundant in P. putida cscRABY
(Fig. [175]8c). The down-regulation of the transcription of genes
encoding nitrogen and urea transporter, for instance amtB encoding an
ammonium transporter, or urtABC, coding for a urea transporter, fit
well to the downregulation of the ureABCD cluster encoding a urease for
urea degradation. In line with this, at the protein level, the global
regulators NtrB and NtrC, which are responsible for nitrogen
regulation, are less abundant.
At first glance, it is not intuitive, that many genes involved in
transport are differentially regulated in the co-culture, as,
neglecting the small amount of citrate in BG11^+ medium, the sole
carbon source is sucrose secreted by the phototrophic partner. However,
at a global ecological scale, cyanobacteria including Synechococcus
spp., are well-known to drive marine bacterial communities because they
are the main suppliers of organic matter due to cell death, cell lysis
and leakiness to photosynthate or exudates^[176]16,[177]29. In
artificial seawater medium (nutrient rich) Synechococcus cultures
accumulated up to 200 µg mL^−1 carbohydrates and 400 µg mL^−1
proteins^[178]16. Furthermore, it is described that cyanobacteria can
secret amino acids and other components. For example, in the
supernatant of S. elongatus CCMP 1631 tryptophan and phenylalanine were
found^[179]5,[180]30. Moreover, P. putida is able to colonise plant
roots and was shown to exhibit advanced chemotaxis towards polyamides,
which are a component of complex root extrudates^[181]31,[182]32. Thus,
in a more general view, it is plausible that transporters for carbon,
carbon-nitrogen compounds, or nitrogen in P. putida cscRABY are
affected by the presence of the phototrophic co-culture partner.
In S. elongatus cscB the DEGs encoding proteins related to transport
mainly belong to the group of transporters for iron or sulphur
(Fig. [183]8b). Almost all of them were down-regulated, with the
exception of Synpcc7942_0197, which was the most highly up-regulated
gene, encoding a putative folate/biopterin family MFS transporter
(log[2]-FC 2.8). Pterins are ubiquitously occurring molecules, which
are needed by cyanobacteria for pigment generation, phototaxis, and UV
protection^[184]33. In the group of genes related to iron transport,
the futABC operon encoding siderophores responsible for iron uptake
(futA2 log[2]-FC −1.6, futB, log[2]-FC −3.1, and futC log[2]-FC −1.74)
was down-regulated (Fig. [185]8b). In the group of genes encoding
sulphur transporters, the strongest down-regulation was observed for
Synpcc7942_1681, annotated to encode a sulphate/sulfonate transporter.
Synpcc7942_1682, and Synpcc7942_1722, also encoding putative
sulphate/sulfonate transporters, were likewise down-regulated. Other
genes encoding putative sulphite exporters were slightly up-regulated
(Synpcc7942_0935, Synpcc7942_0238). Sulphur is an essential element for
microbes and participates in iron-sulphur clusters, a common co-factor
of proteins, in many important physiological processes including
photosynthesis, DNA/RNA modification, and purine metabolism^[186]34.
Sulphite is cell toxic and arises from the intracellular breakdown of
metabolic products, including sulphur-containing amino acids, which
boosts ROS generation^[187]35. Regulation of the transcription of genes
potentially involved in sulphur or sulphite transport hints towards
differences in the complex processes of sulphur homeostasis in S.
elongatus cscB, when grown in co-culture with P. putida cscRABY. In
conclusion, iron and sulphur transport seems to be down-regulated in S.
elongatus cscB when growing together with the heterotrophic partner.
On the protein level, only three proteins associated with transport
were identified to be differentially abundant. Two of them, annotated
as substrate-binding protein of an iron transport system
(Synpcc7942_1409) and as a hypothetical porin (major outer membrane
protein, Synpcc7942_1607) showed a higher abundance whereas another
porin (Synpccc7942_1635) was less abundant in the co-culture grown S.
elongatus cscB cells (Fig. [188]8c).
Cellular processes affected by the co-culture partner: Detoxification,
degradation and stress
Next, we analysed the group of regulated genes and proteins, that can
be functionally related to detoxification, degradation, and stress.
Hays et al. had observed a negative effect of S. elongatus cscB on the
growth of respective heterotrophic partner^[189]11, however as
described above, we have not seen this effect on P. putida cscRABY in
small-scale experiments. In the co-culture setup, both organisms
experience multiple situations that could cause different types of
stress. One situation they have to cope with is the increased salinity
conferring high ionic strength and external osmotic pressure.
Additionally, light can induce oxidative stresses and the adaptation to
changes in the illumination can be a further stress factor. However,
these external factors are comparable in both conditions, the axenic
cultures and the co-cultivation, thus differences that are identified
in transcript or protein abundance related to stress signals are
regarded to be specific for the presence of the respective co-culture
partner.
Our data indicate that several processes assumingly connected to stress
in P. putida cscRABY are affected by the presence of S. elongatus cscB
(Fig. [190]9). More specifically, processes involved in the degradation
of compounds, in the stress response induced by light, in the efflux of
(toxic) substances, or in the general stress response were impacted. A
general trend towards up-regulation of transcription was observed. For
instance, the transcription of the genes belonging to the ben- and
cat-operons encoding enzymes responsible for the degradation of
benzoate were up-regulated (Fig. [191]9a, b). The gene encoding the
AraC-type regulator BenR, however, was slightly down-regulated
(log[2]-FC −1.2).
Fig. 9. Effects on detoxification, degradation, and stress response.
[192]Fig. 9
[193]Open in a new tab
a Schematic visualisation of the degradation of benzoate and
assimilation in the TCA (tricarboxylic acid) cycle by proteins encoded
by the benABCD and catABC-operon in P. putida cscRABY cells grown in
co-culture. Blue shading marks differential expression of the
corresponding genes, and grey arrows indicate regulation by BenR and
CatR. b DEGs annotated to be involved in aromatic degradation and
light-induced stress in P. putida cscRABY grown in co-culture. c DEGs
annotated to be involved in efflux (detoxification) or general stress
in P. putida cscRABY grown in co-culture. d Heat map for metabolites
measured with HILIC in positive (+) and negative (-) ionisation modes.
Mean difference is presented in a colour code from red highest (2.1) to
yellow lowest (0.33). Shown is the comparison of the metabolites
identified in the co-culture to the respective axenic cultures of S.
elongatus cscB or P. putida cscRABY. e DEGs annotated to be involved in
ROS detoxification or glyoxylate degradation of S. elongatus cscB. f
DEGs annotated to be connected to efflux and stress in S. elongatus
cscB.
The transcription of a gene cluster (PP_0738 to PP_0742) that might be
connected to light-induced stress was found to be up-regulated
(Fig. [194]9b): One of its genes (PP_0739) encodes a putative
deoxyribodipyrimidine photolyase and another PP_0740 encodes a putative
MerR family transcriptional regulator of light-inducible genes, known
as PplR1^[195]36. As the illumination was identical for the axenic
culture of P. putida cscRABY and the co-culture, the light intensity
per cell assumingly was lower in the co-culture due to the higher cell
densities. Therefore, changes in the expression of these genes might be
traced back to a different stress situation caused by the presence of
the co-culture partner.
Most of the genes that were differentially expressed and are associated
with the efflux of substances are up-regulated (Fig. [196]9c). Many are
annotated to encode putative resistance-nodulation-division (RND)
efflux pumps, such as Mex-RND and TolC-RND, which are responsible for
the removal of toxic compounds. The highest up-regulation of
transcription was detected for the genes PP_2817 and PP_2731 encoding
putative multidrug efflux pumps with a log[2]-FC of 2.1 or 3.2,
respectively. Furthermore, genes encoding proteins that can be
connected to perceiving and combatting stress were differentially
regulated, most of them showed upregulation in the co-culture. As
mentioned above, ROS derived from the cyanobacterium’s photosynthesis
is likely to be one of the major stress factors for heterotrophic
partners. In line, one gene encoding a catalase (PP_2887 log[2]-FC 1.2)
was found to be slightly up-regulated and, on the protein level, the
catalase KatG was more abundant. However, only one of the two major
cellular ROS degrading regulators, SoxR (PP_2060 log[2]-FC 1.1), was
found to be marginally up-regulated. Another mechanism in the
antioxidant defence is the glutathione metabolism^[197]11,[198]37.
However, no genes encoding proteins associated with glutathione
metabolism could be identified to be differentially regulated, though,
on the metabolome level, metabolites belonging to the glutathione
metabolism were detected in the co-culture cells (Fig. [199]9e). The
transcription of genes belonging to the cop and czc-operons was mainly
up-regulated (Supplementary Note S[200]14 and Supplementary
Table [201]S4). Their gene products are involved in copper homeostasis
and the cytoplasmic detoxification of copper and silver ions, a vital
process controlled by the CopR/CopS two-component system (Supplementary
Fig. [202]S15). Taken together, these findings indicate that P. putida
cscRABY experienced a general stress situation, which is also
corroborated by the up-regulation of genes encoding putative
transcriptional regulators connected to stress^[203]38 (e.g. PP_0740
log[2]-FC of 3.3). We assume that the heterotrophic partner still has
some capacity left to react to stresses, as certain stress answers, for
example the glutathione metabolism, do not yet seem to be affected by
the co-cultivation. This underlines that P. putida cscRABY is a
well-fitting co-culture partner for S. elongatus cscB due to its
natural tolerance towards all different kind of stresses.
By analysing the genes, proteins, and metabolites related to the stress
response in S. elongatus cscB we have identified processes involved in
redox reactions, efflux, general stress and ion homeostasis
(Supplementary Note S[204]15 and Supplementary Table [205]S5). As
already mentioned above, one of the key compounds to combat redox
stress is glutathione, and oxidised glutathione (GSSG), L-glutamate and
γ-glutamylglutamic acid and these were more abundant in the co-culture
cells compared to either axenic culture (Fig. [206]9d). the
transcription of genes encoding putative glutathione peroxidases
(Synpcc7942_0437 and H6G84_RS08920) or a thioredoxin peroxidase tpxA
(log[2]-FC 1.3) were up-regulated in cyanobacterial cells grown in the
co-culture (Fig. [207]9e). Another way to handle oxidative stress is by
the Glutathione-independent degradation of H[2]O[2], performed
enzymatically by catalases, peroxidases, and superoxide dismutase.
Interestingly, the transcript of the catalase KatG was considerably
down-regulated with a log[2]-FC -5.36 in the co-culture growing
cyanobacterium. However, the superoxide dismutase SodB was up-regulated
with a log[2]-FC 1.5 in S. elongatus cscB grown in co-culture
(Fig. [208]9f).
As already observed for the heterotrophic partner, alterations were
also found in the sector of efflux processes and general stress
(Fig. [209]9f). DEGs encoding different types of efflux transporters,
such as HlyD-family efflux transporters (Synpcc7942_1224) or RND efflux
transporters (Synpcc7942_1869, Synpcc7942_1870) were mostly
up-regulated in the co-culture, as was the transcription of genes that
encode proteins putatively involved in stress response. In general, in
cyanobacteria, high-light-inducible proteins (Hlip) are expressed in
response to various exogenous stresses, including already moderate
light intensity^[210]39,[211]40. In this study, two transcripts
encoding these proteins were also found to be up-regulated
(Synpcc7942_1997 and Synpcc7942_1120) (Fig. [212]9f). Three FtsH
proteases, responsible for protein homeostasis of the thylakoid
membrane in photooxidative stress situations, were up-regulated on the
transcriptional level (Synpcc7942_1820, Synpcc7942_0998, and
Synpcc7942_0942), but on the protein level the proteins FtsH, FtsH.1,
and FtsH.3 (Synpcc7942_0297, Synpcc7942_0942, and Synpcc7942_0998) were
less abundant in the co-culture (Fig. [213]9f for transcripts and
Supporting Information S[214]15 for proteins). This is another example
of the discrepancies between transcriptomics and proteomics
results^[215]24.
Taken together, both co-culture partners showed differential regulation
in processes connected to various stresses, detoxification and
degradation when grown together. Even though S. elongatus cscB suffered
from severe photobleaching and reduced growth in the axenic culture,
the transcription of many genes encoding proteins involved in stress
response was mostly up-regulated in the co-culture and not vice versa.
In summary, the multi-OMICs analysis of the co-culture provided us with
a snapshot of the cellular status at the time of sampling and revealed
multi-layered signals of small changes. Thus, in addition to the
synthetic connection by sucrose, more links have to be integrated into
the mechanistic model of the co-culture. We propose to incorporate the
competition for common resources, such as medium components, including
citrate and various salts, as we have observed transient uptake or
accumulation of citrate, acetate, and ethanol. Furthermore, in both
organisms, the ion homeostasis was unbalanced, which might indicate
limitations or reduced accessibility of ions through advanced
scavenging strategies of the respective co-culture partner. This
highlights the requirement for careful medium optimisation in
co-cultures in general. The up-regulation of transport processes,
particularly for amino acids and degradation of aromatic compounds in
P. putida cscRABY, suggests the exchange of molecules belonging to
these groups. Further studies will be directed to investigate the
metabolites in the supernatant to determine whether amino acids or
other compounds accumulate. As this phenomenon has been previously
reported in growing cultures of S. elongatus CCMP 1631^[216]30 and is
commonly observed in marine cyanobacteria consortia^[217]16. Setting up
a precise mechanistic model of co-cultures will contribute to better
controllability and stability in multi-species processes and enable
upscaling and exploitation for biotechnological applications. However,
it is challenging to translate the results of a general grow-associated
classification of microbial interactions, e.g.,
positive/neutral/negative, and the comprehensive results obtained by a
multi-OMICs analysis into quantitative, predictive models.
Nevertheless, combining phototrophic and heterotrophic organisms holds
great potential for co-culture applications, as it combines different
metabolic regimes and thus can link CO[2] fixation to diverse metabolic
traits. The ability to efficiently utilise and recycle carbon offers
innovative solutions to address environmental and industrial
challenges, making these partnerships a promising avenue for future
biotechnological advancements. Our findings contribute to a deeper
understanding of co-culture dynamics and may, at the end of the day,
contribute to harnessing the benefits of synergistic interactions
between different microorganisms in biotechnological endeavours.
Materials and methods
Strains and culture preparation
The sucrose metabolising strain, Pseudomonas putida EM178 att::Tn7
cscRABY^[218]14 harbouring the cscRABY-operon was used as the
co-culture partner for Synechococcus elongatus PCC 7942 cscB^[219]10.
S. elongatus cscB pre-cultures were first grown in BG11^+
medium^[220]41 under continuous illumination of 22 µmol photons s^−1
m^−2, 30 °C, and 120 rpm in an orbital shaker (Multitron Pro from
Infors HT, Switzerland) without additional aeration. After reaching the
stationary phase, the cultures were transferred to BG11^+ medium
supplemented with 150 mM NaCl, inoculated with a 1:20 ratio, and grown
under the same conditions. These salt-adapted phototrophic cultures
were used for all experiments.
Pre-cultures of P. putida cscRABY were grown in 3 mL LB-medium at 30 °C
and 220 rpm^[221]14, and subsequently transferred to a second
pre-culture consisting of 3 mL BG11^+ medium with 3 g L^−1 sucrose. In
the reference experiments, the stationary P. putida cscRABY cultures
were transferred into BG11^+ medium supplemented with 150 mM NaCl and
1–3 g L^−1 sucrose in 100 mL shake flasks and grown under the same
conditions as the pre-cultures. The cultures were centrifuged at
4000 × g for 5 min. and then resuspended in fresh BG11^+ supplemented
with 150 mM NaCl before being added to the process vessels.
Physiological investigation of co-culture in 12-well plates (1.6 mL scale)
To investigate potential interactions, experiments were performed in
12-well plates (Brand GmbH, Germany) at a 1.6 mL scale. S. elongatus
cscB was inoculated to an OD[750] of 0.05 in BG11^+ medium supplemented
with 150 mM NaCl, and cells were acclimated for two days to salt and
other conditions (25–30 °C, 120 rpm, 20 photons µmol m^−2 s^−2,
incubator Multitron Pro from Infors HT from Switzerland). No additional
aeration was provided, and the plates were sealed with laboratory film
to prevent water evaporation. Water loss was considered by verifying
the volume left in the wells at the end of the experiment. Gene
expression of the transporter CscB in cyanobacterial cultures was
induced with 0.1 mM Isopropyl ß-D-1-thiogalactopyranoside (IPTG), and
an extra amount of sucrose of 1 g L^−1 was added to support
heterotrophic growth at the beginning of the experiment. Co-cultures
were started by inoculating different cell counts of P. putida cscRABY
to achieve different phototroph:heterotroph ratios. For experiments in
darkness, the plates were covered in tinfoil.
Physiological investigation of the co-culture in the membrane reactor
The CellDEG HDC 9.100 Universal Platform (CellDEG GmbH, Germany)
consisting of 9 cultivators (HD100 Cultivator) mounted to the platform,
an orbital-shaker, and a control unit was used. A partial CO[2]
pressure of 2% and different light profiles for the high-power LED
light sources (RX-400 LED light Source from Valoya) were implemented
(e.g. constant light of 50 or 120 µmol photons s^−1 m^−2 and
exponential light t[d] = 52 h). Cells were grown in BG11^+ supplemented
with 150 mM NaCl and 0.1 mM IPTG for induction of sucrose permease CscB
of the phototrophic partner. At the beginning of the process, the pH
was set to 7.5, and no further control occurred. The overall volume of
each reactor was 95 mL, and water loss through condensation was
considered by monitoring the weight of the membrane reactors during the
processes. The process was started by inoculating the cyanobacterium
from a salt-adapted culture to an OD[750] of 0.1–0.2 in the cultivation
vessel. P. putida cscRABY was added to the co-cultures to an OD[600] of
0.05–0.01. Cell count, optical density, and sucrose concentration were
analysed by daily sampling of 1–2 mL of the culture broth.
Reference experiment with different settings for comparative OMICs analysis
The reference experiment consisted of three different settings in
biological triplicates. The setting were the axenic cultures of S.
elongatus cscB, the axenic culture of P. putida cscRABY and the
co-cultures, thus 9 samples in total. The experiment started with an
acclimatisation phase for the phototrophic partner (Start OD[750] of
0.1–0.2) under constant light (120 µmol photons s^−1 m^−2). After
14–18 h, the co-culture was started by inoculating P. putida cscRABY to
an OD[600] of 0.05. A sucrose feed supplied P. putida cscRABY axenic
cultures with external carbon. Therefore, a cap was designed to enable
feeding and in situ sampling from the membrane reactor. The sucrose
secreted by S. elongatus cscB grown in co-culture was estimated and
used to define the sucrose feeding rate for the axenic cultures of P.
putida cscRABY (see Supplementary Note S[222]5 for the calculation). A
batch sucrose of 0.1 g L^−1 was provided at the beginning, mimicking
the initial sucrose production of the phototrophic partner. The axenic
cultures of S. elongatus cscB were handled as described above. After
~60 h samples for multi-OMICs analysis were taken, centrifuged at
4000 rpm for 5 min (10 mL proteomics) or 13,000 rpm 1 min (1 mL
metabolomics and transcriptomics) at 4 °C in a centrifuge 5418R from
Eppendorf and subsequently stored at – 80 °C.
Sample preparation and analytical methods
Samples of the processes were directly used to determine the optical
density at 750 nm (600 nm) and cell counts. For further analysis, cells
were separated from the medium by centrifugation at 13,000 rpm for
30 s. in a centrifuge 5418R from Eppendorf. Cell counting was carried
out as previously described^[223]12. High-performance liquid
chromatography (HPLC) was used to quantify sugars, medium components
and common overflow metabolites. The Agilent 1100 series, Waldbronn,
Germany with a Shodex SH 1011 column was used for sugar analysis and a
Shimadzu LC2030C Plus with a Bio Rad aminex HPX-87H column for other
metabolites. The flow rate for the sugar analysis was 0.45 mL min^−1
with 0.5 mM sulfuric acid, the column was heated to 30 °C, and the
refractive index (RI) detector to 50 °C. For analysing medium
components and overflow metabolites, a flow rate of 0.6 mL min^−1 was
used with the same aqueous solvent and an RI temperature of 40 °C.
Concentrations were calculated by integration of the peak area of each
peak and correlation to the corresponding standards.
Multi OMICs methods
Transcriptomics
Samples were sent on dry ice to Eurofins genomic in Konstanz, Germany,
for RNA isolation, sequencing, and initial bioinformatic analysis.
Results were verified and visualised using the Galaxy platform and
R-studio.
Metabolomics
Samples were extracted from the cell pellets and separated using two
types of columns. A UPLC BEH Amide 2.1 × 100 mm, 1.7 µm analytic column
(Waters, Eschborn Germany) with a 400 µL min^−1 flow rate for
hydrophilic interaction liquid chromatography (HILIC) and a Kinetex
XB18 2.1 ×100 mm, 1.7 µm (Phenomenex, Aschaffenburg Germany) for
reverse phase chromatography (RP) with a 300 µL min^−1 flow rate. A
volume of 5 µL per sample was injected. The autosampler was cooled to
10 °C, and the column oven heated to 40 °C. MS settings in the positive
mode were as follows: Gas 1 55 psi, Gas 2 65 psi, Curtain gas 35 psi,
temperature 500 °C, Ion Spray Voltage 5500 V, declustering potential
80 V. The mass range of the TOF MS and MS/MS scans were 50–2000 m/z and
the collision energy was ramped from 15–55 V. MS settings in the
negative mode were as follows: Gas 1 55 psi, Gas 2 65 psi, Cur 35 psi,
temperature 500 °C, Ion Spray Voltage –4500 V, declustering potential
–80 V. The mass range of the TOF MS and MS/MS scans were 50–2000 m/z
and the collision energy was ramped from –15–55 V. The data was
collected in the data-dependent-acquisition mode. A more detailed
description of the procedure and data analysis can be found in
Supplementary Note S[224]16.
Proteomics
Proteins were extracted from cell pellets and trypsin-digested peptide
desalting was processed using Bond Elut OMIX C18 tips (Agilent
Technologies) following the manufacturer’s instructions. Liquid
chromatography-tandem mass spectrometry (LC-MS/MS) proteome analysis
was performed using reverse-phase LC on a Dionex Ultimate 3000 RSLC
nano 2 system coupled online to a Q Exactive HF mass spectrometer
(Thermo Scientific). A more detailed description of the procedure and
data analysis can be found in Supplementary Note S[225]16.
Statistics and reproducibility
All data shown in this work is derived from three biological
replicates, i.e. three different cultures that were inoculated from
three different precultures, each derived from an individual clone. The
mean and the standard deviation were calculated from these three
replicates. The 9-fold parallel membrane reactor system allowed to grow
all cultures at the same time, reducing variability due to other
environmental factors as temperature. The reference experiment was run
twice (EI and EII), and data are shown for both experiments
(Fig. [226]3 and Supplementary Note S[227]7). OMICs data is also
derived from three biological replicates each, and details on the data
analysis are specified in Supplementary Note S[228]16.
Reporting summary
Further information on research design is available in the [229]Nature
Portfolio Reporting Summary linked to this article.
Supplementary information
[230]Supplementary Information^ (4.4MB, pdf)
[231]42003_2024_6098_MOESM2_ESM.docx^ (15KB, docx)
Description of additional Supplementary Materials
[232]Supplementary Data 1^ (1.9MB, xlsx)
[233]Supplementary Data 2^ (320.1KB, xlsx)
[234]Reporting Summary^ (1.8MB, pdf)
Acknowledgements