Team:Nottingham/Design

Project

Design



So you know what we want to do, but how are we going to do it?

Project
Design

Aims

Our project was split into the following objectives, spanning all of our modelling, wet lab and hardware-building activities:

  • Selection of suitable reporters
  • Establishment of C. sporogenes as an appropriate strain for acetone production
  • Engineering of C. sporogenes reporter and control strains
  • Reporter expression assays
  • Development of a technology for the detection of acetone

Clostridium sporogenes as a Surrogate Strain

We are using a strain of Clostridium sporogenes (NCIMB 10696) as a surrogate to Clostridium botulinum.


Handling of toxinogenic C. botulinum strains requires specialised training that could not have been completed within the given iGEM project timeframe. Therefore, we used a safe strain of C. sporogenes (NCIMB 10696) as a surrogate to Clostridium botulinum. C. sporogenes is widely employed in the food industry as a non-toxic surrogate to group I proteolytic strains of C. botulinum, thanks to a 99% sequence similarity between the two strains.[1] Although C. sporogenes does not produce the botulinum neurotoxin and cannot be used in food preservation studies concerned with preventing toxin production, it is frequently used in other types of testing, e.g., for the validation of high-pressure thermal treatments (HPT) in low acid shelf stable products.[2]


In order to turn C. sporogenes into a more effective model organism, capable of mimicking toxin production, we engineered it to express BotR, the sigma-factor responsible for activating toxin transcription in C. botulinum, from a chromosomally-located gene copy.

Selection of Suitable Reporters

When choosing our reporters, we considered the following criteria:

  • A substance that is detectable, e.g. fluorescent or volatile.
  • A substance that is relatively harmless to humans.
  • A substance that is not readily found in food or food packaging.
  • A substance that does not kill the host strain.

Acetone

GusA

FAST

Acetone Reporter System


Acetone could be an ideal reporter as it fulfils all of the above criteria. Being a volatile gas, it can be detected using an electronic nose/gas sensor in the headspace of food packaging.


The acetone-production operon (APO) consisted of the following genes:

  • thiolase (thl)
  • CoA-transferase subunit A/B (ctfA and ctfB)
  • acetoacetate decarboxylase (adc)

However, given that acetone production has not been reported in C. sporogenes (nor C. botulinum) previously, two alternative protein-based reporters were also considered for validation of our circuit design, in case acetone production was not successful.


Reporter proteins are a well-known limitation for anaerobic molecular microbiology, given that classical fluorescent reporter proteins such as green fluorescent protein (GFP) are inactive in anaerobic bacteria, requiring molecular oxygen for maturation.[4]


When selecting alternative reporters, we looked for reporters that would not require cell lysis, to parallel as closely as possible the acetone reporter system.

Establishment of C. sporogenes as a Suitable Strain for Acetone Production

To establish whether C. sporogenes could support the production of acetone, the wet and dry lab teams worked closely together.


Modelling Informing Genetic Engineering Decisions


To formulate a plan for engineering C. sporogenes to produce acetone, the modelling team created a structural model with the following aims in mind:

  • Determine whether C. sporogenes has any metabolic pathways capable of producing. If not, the model will assist in the engineering of C. sporogenes to produce acetone.
  • Identify the best pathway to add (e.g. the most thermodynamically feasible and efficient in anaerobic conditions)
  • Determine whether any acetone degradation pathways are present in the genome, since their presence could result in acetone being immediately consumed, and therefore interfere with the reporter being detected.
  • Assess biologically relevant flux distributions of the network with acetone production included. E.g. Is acetone likely to be produced?

When constructing our new model of C. sporogenes, we concluded that no genes for acetone production were present in the genome. Interestingly, however, all other genes in the widely studied Acetone-Butanol-Ethanol (ABE) pathway were present. As part of the ABE pathway, a gene encoding for a thiolase was found in the genome. However, this is reportedly hypothesised as being a bottleneck in the pathway, however we could not confirm this using our dynamical model (see modelling report for more details). Nevertheless, we added the thiolase from C. acetobuylicum to try and improve acetone production. The model was capable of synthesising all four products observed in our experiments using the wild type, namely, acetate, ethanol, butyrate and butanol.


We assessed four different pathways involving acetone and identified the CtfA/B-Adc pathway, which completes the ABE pathway in C. sporogenes, as the most promising candidate. Interestingly, the genome of the model Type A strain of C. botulinum ATCC 3502 was found to contain copies of the ctfA/B genes, while lacking adc. This raised the question of whether C. botulinum’s ctfA/B could support acetone production in combination with adc and thl from C. acetobutylicum. Therefore, we created 2 alternative acetone-production operon (APO) versions: one comprising thl, ctfA/B and adc from C. acetobutylicum and another with the ctfA/B genes replaced for those from C. botulinum.


During the construction process we also concluded that no acetone degradation reactions were present in either genomes.


The CtfA/B-Adc pathway was added to the structural model. Flux balance analysis was then used to confirm that a steady state solution existed for acetone production. The model predicts that acetate is the favourite product. Importantly, however, when acid excretion is blocked, to mimic the effect of a decreasing pH, acetone becomes the optimal product, which supports the idea to use acetone as a reporter.

For more details, visit the Modelling page.


In parallel to our modelling, the wet lab team undertook the following investigations:


Growth Characterisation


Before engineering the strain, we tried to understand the growth profile of wildtype C. sporogenes over the course of 48 hrs in its standard laboratory medium (TYG). Generated growth data were subsequently used in modelling a better fit to the growth equations. This allowed us to calculate a more accurate maximum growth rate to confirm whether not the structural model was correct.


Does C. sporogenes grow in a suitable pH range for acetone production?


Given that pH is known to regulate the adc gene of the acetone-production operon,[9] also investigated the pH profile of the strain, parallel to its growth.


Does C. sporogenes grow in real food media?


In addition to our initial growth analysis, we carried out additional growth curves to assess the growth of C. sporogenes in real food media. For example, based on the advice of Prof. Mike Peck, we grew C. sporogenes in carrot juice to model a previous outbreak of botulism in 2006. This outbreak was caused by the inappropriate storage of the juice product. Our growth curves showed C. sporogenes did successfully grow in carrot juice, reassuring us that our organism can have real-life applications.


Does C. sporogenes make acetate and/or acetone?


By sampling culture supernatant at different timepoints during growth of the wildtype strain and performing GC/HPLC analyses, it was experimentally confirmed that C. sporogenes produces acetate but no acetone, thus confirming the predictions of the structural model.


Does acetone kill C. sporogenes?


Another important consideration in choosing a suitable reporter, is to ensure that the reporter is harmless to the bacteria producing it. To determine the highest concentration of acetone that C. sporogenes can withstand, we conducted an acetone kill curve based on the recommendation of Prof. Liz Sockett. C. sporogenes was subjected to varying levels of acetone over 48hrs, after which OD and CFU were measured.


Investigating Different Media for Optimal Acetone Production


In our initial growth analysis of wildtype C. sporogenes we observed a gradual increase in pH over time, as expected due to the production of acetate (a weak acid). However, the observed pH was still higher than the optimum pH of around 5.5, required for the Adc enzyme to function properly for the production of acetone.[9]


To investigate how we could achieve this optimum, we grew C. sporogenes in 4 different types of media: TYG whose pH was corrected to 5.5 using HCl, TYG with 30 mM glucose, TYG with 60 mM glucose and TYG with 100 mM glucose. We expected C. sporogenes to convert greater amounts of glucose into more acetate, thereby resulting in a bigger pH drop.

It was found that TYG with 30 mM glucose was the best medium to use in order to obtain the desired pH.

Engineering of C. sporogenes Reporter and Control Strains

Splitting the botR- and Reporter-Expression Modules


The botR gene encodes an alternative sigma factor responsible for transcriptionally activating the genes of the botulinum neurotoxin cluster in C. botulinum,[10] as shown in Figure 1. It follows that placing reporter expression under the control of a BotR cognate promoter (e.g., Pntnh) was a logical way of mimicking toxin expression in C. sporogenes.


Figure 1.

Neurotoxin gene cluster in C. botulinum. PbotR activates transcription of the botR sigma factor. PNTNH is the promoter driven by BotR which activates neurotoxin production in C. botulinum. PHA33 is the promoter which transcribes the genes for the haemaglutinin operon.


Therefore, to mimic toxin expression in C. sporogenes two components were required:

  • BotR expression
  • Reporter expression under the control of a BotR-activated promoter

Due to the large size of the acetone production operon, it would not have been feasible to accommodate the botR- and reporter-expression modules on the same plasmid. Therefore, we decided to use a split configuration whereby the botR-expression module would be integrated in the genome of C. sporogenes, whereas the reporter-expression module would be provided on a suitable plasmid.


This configuration also benefited from increased flexibility, as the same reporter plasmids could be assayed in the three different botR-expressing strains (discussed below), without additional cloning. This allowed us to construct several controls for each of the reporters and botR-expressing strains and provided us with more data from which to draw conclusions for our proof-of-concept studies.


Given that in C. botulinum the entire toxin gene cluster is chromosomally located, ultimately, both modules could be integrated in the genome, to generate more accurate reporter transcription and translation data. However, genomic integration of all constructs of interest would not be feasible in the timeframe of this iGEM project, rendering the split configuration approach a suitable alternative.


The botR-Expression Module


Introduction of the botR gene with its native promoter (PbotR) in the genome of C. sporogenes would allow us to simulate, as closely as possible, botulinum toxin production in this organism. We obtained the PbotR and botR sequences from the genome of C. botulinum ATCC 3502, the C. botulinum type A toxin model strain. The PbotR-botR module was assembled with strong bidirectional terminators upstream (Tfad) and downstream (Tfdx), to isolate it from chromosomal read-through.


The pyrKDE locus was chosen as the genomic location for integration of the botR-expression module, being a well-validated locus for insertion of heterologous cargo in clostridial strains, including C. sporogenes.[11][12][13] We chose to use the recently published RiboCas method for CRISPR/Cas9-mediated genomic integration of the botR-expression module.[14] We designed our editing cassettes so that genomic integration of the module would simultaneously remove the pyrE gene and its upstream native RBS (20 bp upstream).


Given that PbotR might be under the influence of unknown environmental stimuli, to validate our design, we also chose to integrate botR with its native RBS downstream of the pyrD gene, to be expressed by the native promoter of the pyrKDE operon (PpyrKDE). Under the control of this promoter, botR expression (and therefore reporter expression) was anticipated to occur constitutively, serving as a BotR-dependent positive control. 


Finally, botR was also placed under the control of a lactose-inducible system from C. perfringens(PLAC).[15] This allowed us to test the minimum and maximum obtainable reporter expression range as well as gain an insight into the BotR-dose dependence of reporter expression, when placed downstream of a BotR-activated promoter. This information could then inform the range of detection for the sensor in our electronic nose.


In total, three different botR-expression modules were assembled as shown in Figure 2.


Figure 2. The three different botR-expression modules integrated in the genome of C. sporogenes in the PyrKDE locus. PbotR: the native reporter construct with the native botR promoter (PbotR) upstream of botR; PpyrKDE a botR dependent constitutive control with the native RBS of botR upstream of botR, whose expression is driven by the PpyrKDEpromoter of the PpyrKDEoperon; PLAC lactose-inducible promoter system from C. Perfringens driving the expression of botR. Tfdx and Tfad were incorporated where applicable to isolate the modules from chromosomal read-through. Tbgar is a native terminator associated with the PLAC system.

The BotR-Activated Reporter-Expression Module


Standard Clostridium-E. coli shuttle plasmids pMTL82151 (high copy Gram negative replicon) and pMTL82121 (low copy Gram negative replicon) were chosen for the expression of our reporter modules.[16] Suspected toxicity associated with GusA expression in E. coli, warranted the use of the low copy pMTL82121 plasmid; all other constructs were assembled in the high copy vector backbone.


Figure 3.

The modular shuttle plasmid used for reporter expression in C. sporogenes. Relevant features for plasmid transfer, replication, maintenance and assembly are annotated.


The chosen plasmid backbones are comprised of the following basic modules, shown in Figure 3.

  1. Gram-positive replicon: pBP1
  2. CatP: chloramphenicol/thiamphenicol antibiotic resistance marker
  3. Gram negative replicon: ColE1 (pMTL82151) or p15a (pMTL82121)
  4. Multiple Cloning Site

Reporter genes, namely the C. acetobutylicum and C. botulinum APO, FAST and gusA sequences, were placed downstream of Pntnh,, the BotR-activated promoter natively associated with the expression non-toxic non-haemagglutinin component of the neurotoxin complex. A BotR-independent, positive control was also constructed by placing reporter expression downstream of the strong, constitutive Pfdx promoter from C. sporogenes. A promotorless version of each reporter was also constructed, as a negative control, allowing us to discount basal levels of transcription that would occur through transcriptional leakage.


Figure 4. The different reporter–expression modules assembled. PNTNH : BotR-activated promoter, used to mimic toxin expression in the presence of BotR (reporter strain); Pfdx: strong, constitutive, clostridial promoter, used as a BotR-independent positive control; No promoter: promoter-less version, used as a negative control. All constructs on pMLT82121 high copy vector apart from gusA which is on pMLT82151.

The constructed reporter-expression plasmids were conjugated in the previously generated botR-expressing strains, according to the following table, to yield the desired reporter and various control strains:

Host strain/

Reporter module

ΔpyrE::PpyrKDE-botR

ΔpyrE::PbotR-botR

ΔpyrE::PLAC-botR

Wildtype (WT)

Pntnh-reporter

Constitutive, BotR-dependent, positive control

Reporter strain mimicking toxin expression

Inducible BotR-dependent control

BotR-dependent negative control (in the absence of botR, no Pntnh transcription is expected)

Pfdx-reporter

x

x

x

Constitutive, BotR-independent, positive control

No promoter-reporter

x

x

x

BotR-independent negative control

Reporter Expression Assays


Constructed strains were grown in the TYG/30 mM glucose medium, previously selected for optimal acetone production, alongside each other. Samples were taken at representative timepoints of growth, namely at 0 h (lag phase), 4 h (early exponential phase), 8 h (late exponential/early stationary), 24 h and 48 h (late stationary/death phase). Reporter expression and OD measurements were performed for all time points.


Acetone production was measured in the culture supernatant of constructed strains, using HPLC or GC, depending on instrument availability.


FAST and GusA expression were measured after addition of the TFAmber fluorogen and 4-MUG substrate, respectively, to pelleted cells.


Switching reporter expression ON/OFF


To validate our reporter strains (ΔpyrE::PbotR-botR + pMTL82151/pMTL82121-Pntnh-reporter), we wanted to test whether environmental stimuli known to regulate toxin production in C. botulinum could influence reporter expression in a similar way in the engineered strains of C. sporogenes. For instance, literature indicated arginine as a repressor of toxin production while having no effect in the growth in C. botulinum.[17] Therefore, we assayed our FAST-expressing strains in media with and without arginine to investigate changes in reporter expression.


In principle, any of the three reporters could have been chosen for this experiment, since all three had generated similar expression profiles at that point; we selected the FAST reporter, as it offered the most reliable, time-saving and user-friendly way of measuring reporter expression.

Development of a Technology for the Detection of Acetone


We needed some method of detecting acetone produced by our mutant strain of Clostridium sporogenes. We decided to build an ‘electronic nose’ that could record concentrations of acetone in the air and output a reading.


After receiving advice from Rapid Electronics, we decided to use the gas sensor TGS822 in our electronic nose. One downside of this sensor is that, as well as detecting ketones, it detects alcohols. C. sporogenes produces ethanol which will potentially affect the sensor’s acetone readings. Fortunately, the modelling team found that based on the temperatures the bacteria would be operating at, it is unlikely that ethanol is volatile enough to be in the headspace.


For more details on the design of our electronic nose, visit the Hardware page.

References

1. Brown, J.L., Tran-Dinh, N., and Chapman, B., Clostridium sporogenes PA 3679 and its uses in the derivation of thermal processing schedules for low-acid shelf-stable foods and as a research model for proteolytic Clostridium botulinum. J Food Prot, 2012. 75(4): p. 779-92.
2. García-Parra, J., González-Cebrino, F., Cava, R., and Ramírez, R., Effect of a different high pressure thermal processing compared to a traditional thermal treatment on a red flesh and peel plum purée. Innovative Food Science & Emerging Technologies, 2014. 26: p. 26-33.
3. Streett, H.E., Kalis, K.M., and Papoutsakis, E.T., A Strongly Fluorescing Anaerobic Reporter and Protein-Tagging System for Clostridium Organisms Based on the Fluorescence-Activating and Absorption-Shifting Tag Protein (FAST). Appl Environ Microbiol, 2019. 85(14).
4. Mordaka, P.M. and Heap, J.T., Stringency of Synthetic Promoter Sequences in Clostridium Revealed and Circumvented by Tuning Promoter Library Mutation Rates. ACS Synth Biol, 2018. 7(2): p. 672-681.
5. Lee, J., Jang, Y.S., Choi, S.J., Im, J.A., Song, H., Cho, J.H., Seung do, Y., Papoutsakis, E.T., Bennett, G.N., and Lee, S.Y., Metabolic engineering of Clostridium acetobutylicum ATCC 824 for isopropanol-butanol-ethanol fermentation. Appl Environ Microbiol, 2012. 78(5): p. 1416-23.
6. Gheshlaghi, R., Scharer, J.M., Moo-Young, M., and Chou, C.P., Metabolic pathways of clostridia for producing butanol. Biotechnology Advances, 2009. 27(6): p. 764-781.
7. Blast.ncbi.nlm.nih.gov., BLAST: Basic Local Alignment Search Tool. 2019 [accessed; Available from: https://blast.ncbi.nlm.nih.gov/Blast.cgi.
8. Biopython.org., Biopython. 2019 [accessed; Available from: https://biopython.org/.
9. Ho, M.-C., Ménétret, J.-F., Tsuruta, H., and Allen, K.N., The origin of the electrostatic perturbation in acetoacetate decarboxylase. Nature, 2009. 459: p. 393.
10. Raffestin, S., Dupuy, B., Marvaud, J.C., and Popoff, M.R., BotR/A and TetR are alternative RNA polymerase sigma factors controlling the expression of the neurotoxin and associated protein genes in Clostridium botulinum type A and Clostridium tetani. Mol Microbiol, 2005. 55(1): p. 235-49.
11. Ng, Y.K., Ehsaan, M., Philip, S., Collery, M.M., Janoir, C., Collignon, A., Cartman, S.T., and Minton, N.P., Expanding the repertoire of gene tools for precise manipulation of the Clostridium difficile genome: allelic exchange using pyrE alleles. PloS one, 2013. 8(2): p. e56051.
12. Heap, J.T., Ehsaan, M., Cooksley, C.M., Ng, Y.-K., Cartman, S.T., Winzer, K., and Minton, N.P., Integration of DNA into bacterial chromosomes from plasmids without a counter-selection marker. Nucleic Acids Res, 2012. 40(8): p. e59-e59.
13. Heap, J.T., Theys, J., Ehsaan, M., Kubiak, A.M., Dubois, L., Paesmans, K., Mellaert, L.V., Knox, R., Kuehne, S.A., Lambin, P., and Minton, N.P., Spores of Clostridium engineered for clinical efficacy and safety cause regression and cure of tumors in vivo. Oncotarget, 2014. 5(7): p. 1761-9.
14. Cañadas, I.C., Groothuis, D., Zygouropoulou, M., Rodrigues, R., and Minton, N.P., RiboCas: A Universal CRISPR-Based Editing Tool for Clostridium. ACS Synth Biol, 2019. 8(6): p. 1379-1390.
15. Hartman, A.H., Liu, H., and Melville, S.B., Construction and characterization of a lactose-inducible promoter system for controlled gene expression in Clostridium perfringens. Appl Environ Microbiol, 2011. 77(2): p. 471-8.
16. Heap, J.T., Pennington, O.J., Cartman, S.T., and Minton, N.P., A modular system for Clostridium shuttle plasmids. J Microbiol Methods, 2009. 78(1): p. 79-85.
17. Fredrick, C.M., Lin, G., and Johnson, E.A., Regulation of Botulinum Neurotoxin Synthesis and Toxin Complex Formation by Arginine and Glucose in Clostridium botulinum ATCC 3502. Appl Environ Microbiol, 2017. 83(13).