Tobias Pfennig – RWTH Aachen, Germany
Homepage: https://www.cpbl.rwth-aachen.de/cms/CPBL/Die-Juniorprofessur/Unser/~wljpm/Tobias-Pfennig/
Capturing Cyanobacterial Photosynthesis – A spectrum-dependent mathematical model
Well-designed mathematical models complement experimental scientific work. The mathematical representation of reaction networks allows for a detailed and systematic investigation of the system. The network’s simplification and breakdown also give a new perspective to working on the whole organism. Cyanobacteria are of high economic value and becoming a promising tool in biotechnological production. However, not all aspects of photosynthesis in cyanobacteria could be experimentally elucidated, yet. Furthermore, despite their evolutionary bond to plants, the structure and components of photosynthetic electron transport differ with a high impact on the overall dynamics, prohibiting the usage of established plant-based models. Therefore, targeted mathematical models might be highly beneficial.
We have developed an ordinary differential equation-based photosynthesis model in Synechocystis sp. PCC 6803. It dynamically tracks the major photosynthetic processes, from light capture to electron transport and carbon fixation. Notably, we included a full-spectrum light description to mimic lab growth conditions. We used simple kinetic rate laws where no crucial regulatory mechanisms were apparent, and the model was parameterized using dedicated measurements and literature values. With our model, we can reproduce key spectrometric experiments and photosynthetic dynamics. These include the simulated redox state of electron carriers, flux through alternative electron pathways, and dynamic fluorescence signals. We also investigated alternative mechanisms for state transitions, for which no consensus exists yet. Notably, the model showed that the irradiance light color determines the distribution of metabolic control and, therefore, biotechnological targets.
With this model, we integrate systems-level knowledge of photosynthesis in cyanobacteria and provide a theoretical framework for further complex investigation.
This talk is part of the CyanoWorld online seminar series, hosted by Nicolas Schmelling and Ilka Axmann, and was held on Zoom on October 12th, 2023.
Follow Us on Twitter: https://twitter.com/CyanoWorld1
Thank you very much Nick forly letting me present this nice Community I’m Nick I’m a PhD student in the junior professorship for computational life science at AA arur and as Nick very r that I am interested in photosynthesis of C bacteria and modeling it and therefore kind of continuing the legacy
Of my profet and that is exactly what I want to talk to you today I will talk about our attempt at shedding light on cab bacterial photosynthesis and more precisely about our recently published metical model of photosynthesis in cissus PCC 683 to start off I would like
To tell you the story of how we came to create and this project is rooted at the roots of rice originally we were interested in the growth of rice in patties like these flood and under these conditions the whole lower plant body is mer and there have been multiple studies
Finding that there is Caceria living onside of these roots colonizing them and there are also instances where SC Fields here are deliberately inoculate with cerial strats that is because the colonized plants show significantly increased growth particularly in the roots and be great natural fertilizers as probably a lot of us are even
Researching we were interested in this symbiosis as it involves two phototropic organisms as compared to the usual phototroph heterotroph Dynamic so the first first we ask ourselves is if the sand bacteria are indeed actively photosynthesize because being underwater drastically changes the light of course that they receive as compared to the
Plant not only will there rece less light due to the water’s absorption and reflection these phenomena of the water are also wave length pendence you can see that the absorption is much stronger in the red actual range spittering the other way around round so the solar
Spectrum that you see at the top will not be what the S bacterium sees now our question became how the San would behave under such different light quality conditions and being computational life scientist we wanted to use a mechanical model to look what the Caceria would do under these light conditions but sightly
No appropriate model was available for what exactly we were interested in and so we decided to create our own model of s v photosynthesis the electron trans chain model this is an overview of processes in the model all of the orange errors are reactions involving electrons
Blue are protons and these are the main reactions that we implemented according to literature each compound here for example plustic keone and each reaction they are all described in mathematical terms and that means for our reaction we have certain kinetic rate equations that tell us how fast is a certain direction
Going at a particular time time Point depending on the compound concentration and if a certain very important direction is missing here then tell us because we would really like to know otherwise I would just quickly go over the main characteristics that we identify as the main ones of course the
Two photos systems which s the light and drive photosynthesis and we estimate the excitation rates for these photos systems from these simulated light absorption and for that we also include a description of the films the light har inten but also of light adaption mechanisms like the orange cite protein
And the state transitions this will be important later we also include thoc restoration because as you probably know for bacteria it’s quite special that the respiration is intercepting with the photosynthetic electron transferred it shares a membrane and so electrons inserted from glycolysis directly integrate into photosynthetic electron
Flow and that of course affects all of the other electron cles to complete the picture we also include terminal oxidases which reduce oxygen back to water not all ter oxid are just active in respiration some also in normal photosynthetic electron flux we also include the Caron conserving mechanism which increases the CO2 concentration
For higher efficiency CO2 fixation and in our model that is the concentration increased by constant Factor but it does consider the cytoplasm pH to calculate how much C2 is actually usable for fixation we lastly carbon fixation and H respiration which offers a very large Downstream pop waste downam of
Photosynthesis and since they’re so large we lump them into single reactions which try to estimate the effect of both pathways on photosynthesis and estimate the rate without adding trans reactions to them and these two far reactions have substate dependency and are regulated by the Redux overall the model is dynamic
Over time so we can simulate how these concentrations evolve because it’s ordinary differential equation based we use the principle of Parson we try to keep everything simple by using simple kinetics mass action or M man we lar the large Pathways which I talked about and overall the model State reasonably size
With six differential equations and 23 reactions included and overall we needed 80 parameters and 62 of them are actually from literature directly take we also estimate the chorophyll like fluoresence to simulate florescence experiments and you will see just a moment an example of that as I told you
In the beginning which is quite important to me we approximate the IR Radiance absorption for that we use a wlink dependent function that takes into account the different pigments of s bacteria and that means we can use any light quality quantity and pigment content in our simulations which if you
Think about the beginning it’s exactly what we wanted to answer that is important because the fontos and S bacteria have significantly different light preferences photos system 2 absorbs mind the Reds together with the fil zones while photo system one has more chor fill absorbs in the blue and
Our Red Spectrum as the two photo assistance Mark entrance and exit points if the Spectrum changes that also drastically Alters the way that the electrons take through here in general we differentiate four different Pathways the elron can take that is the linear pathway from water splitting to NPH the
Water water cycle from water and then back to reducing oxygen to water electron set are included by glycolysis through the respiratory pathway and then cyclic electrons which are excited at PS1 get inserted back into the chain and so around and these popway create different amount ATP and dph which in
Turn affect the downstream metabolism especially the Caron fixation on the cin this is why this is the first thing we looked at the C2 fixation and also the electron flates with it at the top right you can see different light Spector you would commonly find in bi technological
Applications of course solar light but also different artificial Lights flent lamps different Neds and here on the left you can see one of our simulation results where we had these different light sources at different light intensities here on the X and then simulated the C2 consumption as an nitter of photosynthetic activity and
You can see directly that there is major differences between these different lights that the same light intensity the white light that we simulated each of these cases is not simply white light it’s quite different and if you want to look more into why that is we can also
Look into the cellular process into the reactions that take place at PA that the electrons take and here for example you can see that in the very beginning electrons mostly go the linear path to NPH also bit with a cyclic but in a certain point at higher light intensity
The linear electron flow stagnates and cyclic is replaced by the water water cycle which is known from literature that the water water cyle kind of overflow valve electrons highs we can also predict these Pathways for simulated muts and here you see a highlight adaptive Flav protein which is
A major play in this water water cycle and and the water water cycle is strongly reduced and overall you see that at highlights the whole electron flow gets imped which again makes sense for a highlight adaptive protein such electron processes are hard to measure experimentally to cross check our
Simulations however there is a method of exploiting the fluoresence of florify and that is the photosynthesis measurement via pulse amplitude modulation just a short explanation of what we measure here actually if you excite chlorophyll bya light absorption if can react with oxygen to create reacted oxygen species which the cell
Usually wants to avoid so it instead controls it through three quenching mechanisms that is energy can be released as fluoresence that molecule does itself through photosynthesis of course which means that electrons cast into the transfer chain and also a heat dissipation which can actively op passively engaged for example in
Highlight adaption at this floresent here you can see we can measure as this red shifted signal aparted from the absorbance and by measuring the amount of fluoresence and for example blocking the photosynthesis we can estimate how these relative amounts of fluoresence photosynthesis and heat dissipation change how these different crunching
Mechanisms engage or disengage and how do we do it well we use a light induction protocol you can see here the fluoresence of the cell and that in different lights first in dark then in light again and in dark and overall we measure ground fluoresence and also we
Give strong light spikes which tell us something about how St the Chlor can even flourish and what important us are these Dynamics these Dynamics informs about the photosynthetic State about the steady stage to synesis usually and how different adaption processes engage and relax therefore we selected this kind of
Measurement as good comparative data for a model and you can see such a comparison of experimental data in red and the simulation in black and important is that these Dynamics are captured we simulated it first in dark and then in different light colors and even different light intensities and all
Of that engages or disengages different light caption responses of the cell and as I said important to us are these dynamics that the ground fence increases or decreases the way the experimental data does it and also that the peak fluorescent does these are the important and we know that this simulation is
Perfect there are certain imperfections especially the underestimation of front fluoresence in state 2 or an underestimation of the peak fluoresence during and EQ but overall what’s important to us is that we captured the description of the protos synthetic processes and these adaption processes and all of that can be traced to the
Fluoresence which we estimate so we think that the description can be of course improved but it is quite good already and since we model light absorption in a pigment dependent manner we can also simulate the behavior of differently adapted cells here you can see different simulations of cells which
Were grown at monochromatic light sources which you see here at the right and the pigment content of these cells was measured and then we simulated their reaction to always the same light protocol that you saw just now you can see again that the fluoresence responses of these differ quite widely not just in
The height of the fluoresence response that also P past certain processes for example you engage or even don’t engage and these simulations illustrate how the light quality quantity and also the cell absorption into plenty and how important is to consider all of these factors when discussing photosynthesis next we wanted to
Investigate aspects that could be useful for biotechnological occation and that is finding reactions controlling the metabolic system and also simulating Target here you can see what we’re going to use next we’re going to simulate a monochromes so we overlay here the absorption Spectre of the most important photosynthetic pigments for example
Chlorophyll here in solid black betac cortine alop pipil proteins and the colored spikes here Mark monochromatic light sources that we simulate for the cing plops in these four graphs that you see here you can see the metabolic control that different pathway have on the steady state carbon fixation that
Means we simulated the model under first different monochromatic light sources and also different intensities and then slightly vared the flow through single rections within the core Pathways that you see a light El for respiration rubiscos col fixation and the terminal oxidases and the stronger the color in
Each of these pots the more effect the small perturbances ha on the carbon F and you can see for example that the lighter action glow from photosystem 2 to photosystem one at the heights control under the blue violet and the far red light colors that means that it
Is mostly them controlling when we’re in the absorption spectrum of florif while rabisco on the other hand the main carbon fixation enzyme only gain control under higher light intensities and not just that but also wavelength specific and we see that here in this or monitor chromatic light that the rabisco gains
Control earliest and if you look at the others rabisco and terminal oxidases they had relatively little control throughout the whole process but small pedants we see also match for example between Risco and the ter oxidases meaning that the mechanisms behind all of that are still connected all of these
Process are just isolated overall if you look at it we see that the control shifts from The Source reaction so the photos systems to the SN reactions that is rabisco for increasing light and often it would be wise to Target halfways with a high control for engineering however we see that the
Control is not always give it shifts with both the intensity and also depends on the light color that we give the whole thing and therefore light is an important factor when we engineer new strengths and especially if you want to base that on models because we have to
Keep that in mind which of a Pathways might actually be in controll room in the next step we also simulate other potential targets for bike production and how the light would affect those so here we added to the model Su reactions that drain NPH ATP and Pi carbon in ratios that correspond to
Biotechnological Target compounds on the left NPH and on the right BP so purely and then in the middle we see isoprene and glycogen where the isoprene production and glycogen more ATP and we see again that the light wavelength that we give the cells governs of the optimal
Light intensity for our production is so where it is most yellow and the production patterns that we see here so the patterns where production is highest isop and glycogen also seem like the mixture of an adph ATP actually also a bit of the rabisco control that we saw
Before and that lets us think that the mph to ATP ratio that these compounds have actually governs the optimal light conditions in both light color and light intensity and if we just look at ISO we see that in general the company production was highest in the red orange
Light lower in blue and mostly absent in between and a recent study found white similar results here an icpr synthesizing cicis strain had the high growth rate and intermediate IOP production in red light and the strong growth deficit in blue so not a high yield pH interestingly though these
Experiments also showed a high yield in green light and an even higher in Violet which even surpassed the black we assume that since we didn’t take into account pigment adaption in these analyses and the background that these adaptions allowed the cells to surpass our predicted production rates but that
Is also still under investigation now leaving the biotechnological application we also see potential in the model to investigate hypothetical mechanisms in photosynthetic machinery for that as an example we took the state transitions that I mentioned earlier State transitions are adaptive mechanism that is light Spectra which differentially
Excite our two photosystems to and one and it is usually measured as a change in the photosystem vers and especially the ratio between the foressence of both of them and it’s usually regarded as a mechanism to balance the photosystems excitation levels if one of them is more
Excited than the others the problem is that the correct mechanism hasn’t been yet pound by consensus and there are four main mechanisms proposed for this the photos system to quenching model a photo 2 is OV excited this overe excitation is the stest heat the spin over model where this overe excitation
Is drained to photosystem one so passed on the fism mobile model where fism move between the two photos systems and the PVS Detachment model where it detaches from the photos and interestingly according to literature they detach in the state where at least photos 2 receives less excitation and to investigate these hypothesis we
Implemented these four mechanisms in our model and used simple kinetic rate laws to describe that but since we didn’t know the correct parameter values for this we instead simulated 1,000 different models with a range of parameter values and then evaluated the distribution of simulation results first we tested if the mechanisms could
Provide a light adaptive effect that means that it could alleviate the over reduction of the plasto pool when po two so entrance point is overx and in the thought you can see are four different model thees and on the Y AIS the pq4 oxidation by activation of this
Mechanism so that means the higher the value The more stress relieves this mechanism gave the model and you can see the distribution of our 1,000 models here through the box plots and more in the relative terms of through the violent but so where the violence are wider there is more of our simulation
Results in that area and you can see that the mechanisms that all of them bought the PBS Detachment model were potentially advantages to the cell so they reduce the over reduction of the plaster keenon we furthermore tested if these models could replicate the shift INF floresent which I told you in the
Beginning are typically associated with State transitions here the spill over model showed a much lower response than expected not as much fluoresence difference as we would see in usual experiments and so assuming the true mechanisms behind State transitions would provide both a Redux balancing effect and also the typical florescence
Pattern that we would find we conclude that the photo crunching model and the PBS mobile models are more likely to be the main mechanism that is present in s bacteria rather than the spill over model and the PVS Detachment model with this I would already like to conclude
And come to summary I showed you s photosynthesis model which we curated to represent the out transfer chain in detail the reactions are characteristic ones that we find for C bacteria and we included a high adaptability of both reactions to simulate mutants and also light handling to simulate different
Light conditions and also pigment conditions we can simulate the photosynthetic electron fluxes and also called amplitude modulations pyramids which are well captured on the model and when we went more into the model we also saw how the light spectrum governance where the control in our system lies and
Especially between the source at the sink reactions and for ey technological Target reaction that these optimal light conditions can dier which very likely is connected to the ATP to each ratio we also saw that we can test mechanistic hypothesis with the model by implementing the hypothesis first and
Then examining the simulations of them by which we could support two of original for proposes for State Transitions and this model is already publicly available and submitted as a preprint so if you were interested you could test it out right now however in the future would also like to make this
Model even more attractive to use for research groups espe especially in the biotechnological field that means you want to simulate a broader range of biotechnological Target reactions like hydrogen production or PHP or compound oxygenation in whole cell biocatalysis we also want to investigate mechanisms and Pathways that were right now in the
Current model version simplified or even not included like the CO2 acquisition mechanisms and the effect that CO2 has on for example pH or the Highlight damage reactions protomotive Force you name it and we also think that with this model we could simulate how a single cell could behave within a bioreactor
Where it travels throughout different layers of cells which of course also shifts the light that the cell receives lastly we are open for suggestions if you have a particular process or data that you think would fit the scope or model then please talk to us and maybe
We could find something to work together I thank you all for attention I also want to thank my pror an chinska and especially from all group Andreas nil and Elena kman from cheia Dr Tomas zro and Yan CH from dor Dr and Dr G B from
Hungary who all worked with us on making this preprint on publishing it who gave us the data and of course we’re also heavily funded especially by the deut of for commanda of the European Union and Ministries in czechia and Hungary and now if there’s any questions I will be
Very open for them and thank you for attention