Cellulopolis aims to optimize the production of bacterial cellulose (BC) by bacteria of the genus Komagataeibacter, through two main approaches: strain domestication and its gene editing, and the adoption of alternative culture media. The laboratory arm of the project is subdivided into cultivation in alternative media, establishing conditions for genetic manipulation, developing a toolkit for Komagataeibacter rhaeticus AF1 engineering, producing BC in defined shapes, and testing BC as a scaffold for tissue culture.
Our design was focused on planning, building (partial) and testing (partial) novel parts and strains. To achieve engineering success utilizing a novel chassis, our main challenge was to construct a GoldenGate compatible toolkit, composed of promoters with different strengths, RBS, and terminator, in addition to level 1 and 2 backbones with different origins of replication, making its manipulation available by other research groups. From these parts, we propose a solution to increase the efficiency in the production of cellulose, from the modulation of the operon of cellulose production of the species, through the adoption of a promoter inducible by light. We also evaluated different compositions of the culture medium used, proposing the adoption of more sustainable and economically viable alternatives for the growth of the strain.
From the inception of the team Unicamp_Brazil, the goal was to contribute towards a more sustainable planet. Hence, when we decided to optimize BC production from waste available in large quantities in our region, we contacted BioPolMat (UNIARA), BioSmart, and HB consortium, which cultivates BC on agroindustry residues and were supplied with K. rhaeticus AF1. To the best of our knowledge, the genome of K. rhaeticus had never been edited, thus presenting a potential engineering challenge. However, the benefits of starting from a strain adapted to growth in our substrate of choice outweighed the drawbacks. As a backup strategy, we worked in parallel with one more Komagataeibacter strain, namely K. medellinensis ID13488.
Predictions from the flux balance analysis we performed using the Komagataeibacter xylinus genome-scale model (Rezazadeh et al, 2020), indicate that bacterial cellulose (BC) production competes with biomass accumulation, hence redirecting Komagataeibacter´s metabolism towards BC production should have detrimental effects on biomass accumulation. This negative growth pressure would lead to the selection of mutants with defects in BC-producing enzymes, as previously shown by Hur et al., 2020. Thus, we designed a bimodal strategy whereby we will engineer the bacterial cellulose biosynthetic gene cluster (bcs) to shut down BC production during biomass accumulation and induce BC production once desired.
The entire premise of our project is to minimize BC production costs, hence the use of a low-cost inducer is imperative. This requires the use of inducible promotors which can be switched on once the desired cell density is achieved. Literature searches revealed only a couple of inducible promoters functional in K. rhaeticus, such as LuxR (activated by acyl homoserine lactone - AHL) (Goosens et al., 2021). In spite of the good characterization, LuxR is not an option for scalable induction due to the cost of AHL. Fortunately we came across a paper describing a single component light responsive system for control of gene expression in E. coli: the LexRO (Li et al.,2020) . Under dark conditions LexRO dimers interact with an operator region upstream of the RBS, inhibiting transcription. When exposed to blue light, LexRO proteins dissociate and detach from the DNA, allowing transcription.
Based on Florea et al (Florea et al., 2016), we synthesized promotors with high (TUp_a), medium (TUp_b) and low (TUp_c) strengths. These were coupled to optimized RBS (Hur et al, 2020) (TU_RBS), and strong terminator (TU_term) (Florea et al 2016).
Komagataeibacter rhaeticus AF1 genome is highly GC rich, therefore we ordered gene synthesis to represent this codon preference. Unfortunately our custom parts arrived only a couple of weeks prior to the wiki freeze so the composite parts required for validation of the LexRO in K. rhaeticus were not ready on time. Our plan is to assemble 3 plasmids encoding multiple transcriptional units each: LexRO under the control of strong, medium and weak promotors and lex operator controling the expression of a purple reporter (for details and references, see parts section). These will be transformed into K. rhaeticus and will be cultivated in dark (foil covered flask or plate) or light (custom constructed light box - for details see hardware session) conditions. Colour profile of K. rhaeticus cells should indicate the levels if LexRO expression efficient in inhibiting transcription under dark condition (white cells) and allowing transcription under blue light (purple cells).
Considering that cellulose is a polysaccharide formed by units of glucose, we decided to quantify glucose consumption from the bacterial growth media throughout their growth, as well as the formation of short-chain carbohydrates. With this approach, we can observe if the glucose monomer from the media is being consumed to produce glucose polymers.
Carbohydrate quantification from K. rhaeticus cultivation media was performed through High-Performance Liquid Chromatography analysis. Oligosaccharides from the samples were separated and quantified by ion exchange chromatography (HPAEC).The sugars were eluted in a gradient of sodium acetate in a mobile phase of 100 mM NaOH. After separation, íon detection was performed by Pulsed Amperometry (PAD).
Samples were prepared by collecting aliquots of supernatant of liquid culture media every day for 5 days of bacteria cultivation and diluted in 500x in ultra-pure water. For quantification, samples were calibrated with a standard cellooligosaccharide solution (glucose, gluconic acid, cellobiose, cellotriose, cellotetraose, cellopentaose, and cellohexaose) and correspondent peak areas were integrated, resulting in concentration values.
K. xylinus is arguably the most studied Komagatabacter species for which there are tested engineering tools and genome-scale metabolic models (Ryngajłło et al 2020). K. rhaeticus iGEM strain was the first organism sequenced by an iGEM team (Imperial College London) and for which an engineering toolbox was developed (Goosens et al., 2021).
The genomes of K. rhaeticus AF1 and K. medellinensis ID13488 have been sequenced (Hernández-Arriaga et al. 2019; Santos et al. 2019), however, to the best of our knowledge neither strain has been previously manipulated. Hence, to permit engineering and optimization of BC production, we began by evaluating the susceptibility of each strain to the 4 classic antibiotics: ampicillin, kanamycin, spectinomycin and chloramphenicol.
According to the literature, standard replication origins do not work efficiently in Komagataeibacter, therefore we selected backbones with the broad spectrum origins RK2 (BBa_J428346; BBa_J428366), pBBR1(BBa_J428347; BBa_J428367) or RSF1010 (BBa_J428349; BBa_J428369) for further work (Fricke et al 2021).
To evaluate the ameneability of K. rhaeticus AF1 and K. medellinensis ID13488 to genetic manipulations, we transformed both strains by electroporation with BBa_J428346, BBa_J428347 and BBa_J428349, as suggested by Florea, 2016. Initially we struggled with poor transformation efficiency and contamination. After several rounds of frustrated transformation attempts, using a new optimized protocol we acchieved transformation success.
Unfortunately the parts suplied with the iGEM distribution kit had few selection markers, with the only level 2 backbones with broad spectrum origins harbouring spectinomycin resistance cassette, which, to the best of our knowledge, has not been shown to be effective in Komagataeibacter. Therefore we deviced SapI Golden Gate based strategy to replace the KanR marker from level 1 plasmid and replace the SpecR marker from level 2 plasmids with AmpR and CmR. Plasmid backbones BBa_J428346, BBa_J428347, BBa_J428349 , BBa_J428366, BBa_J428367, and BBa_J428369 were amplified with primers JUMP-mF and JUMP-mR and antibiotic resistance cassettes were amplified with AmpR_F, AmpR_R, CmR_F and CmR_R using plasmids BBa_J428385 or BBa_J428357 as templates (Figure y).
To allow the assembly of multiple transcriptional units we are modifying the cloning site of level 2 vectors with replication origins functional in Komagataeibacter, by replacing the region between BsaI sites of plasmids derived from BBa_J428366 and BBa_J428367 with an adapter suitable for simultaneous cloning 4 transcriptional units by GoldenGate with SapI. This is done by amplifying the red fluorescent cassettefrom BBa_J04452, and cloning by GoldenGate with BsaI digestion into BBa_K4435304, BBa_K4435305, BBa_K4435311 and BBa_K4435312, generating plasmids BBa_K4435306, BBa_K4435307, BBa_K4435313 and BBa_K4435314.
The replacement of the green fluorescent markers from plasmids BBa_K4435304, BBa_K4435305, BBa_K4435311 and BBa_K4435312 with a red fluorescent marker flanked by the TU adapters allows the visual selection of sucessfull constructs.
In order to investigate the efficiency of LexRO as a transcriptional repressor in K. rhaeticus AF1, we planned a series of plasmids where lexOp control transcription of reporters such as BBF10K_003335, BBF10K_003349, BBF10K_003360, BBF10K_003344 or BBF10K_003362.
Genes encoding BC-producing enzymes and BC transporters are organized into different bcs operons (Orlovska et al, 2021). The type I cellulose synthase operon (bcsI) comprises four genes, bcsA, bcsB, bcsC and bcsD, flanked by the cellulose synthesis modulators bcsZ, bcsH and bglX (Römling and Galperin 2015). We decided to control BC production by introducing a LexRO binding region upstream of cmcAx (upstream construct) and directly upstream of bcsA (mid construct), the cellulose synthase encoding gene.
Previous work (Goosens et al., 2021) demonstrated that 1 kb of homology arms upstream and downstream of the desired insertion site is sufficient for efficient targeting of the construct to the desired K. rhaeticus chromosomal location. It’s possible to check how we design the homology arms in the parts page (link).
Using designed primers, we PCR amplified 1 kb of homology 5´of the upstream insertion site (TU1a) of both K. rhaeticus AF1 and K. medellinensis (Km) and cloned the by Golden Gate into SapI sites of BBa_J428381. We also PCR amplified and cloned 1 kb of homology 5´of the mid insertion site (TU1b) of both K. rhaeticus AF1 and K. medellinensis (Km) into BBa_J428381.
To allow precise targeting of our constructs we also PCR amplified and will clone 1 kb of homology 3´of the upstream insertion site (TU4a_CDS), and 1 kb of homology 3´of the mid insertion site (TU4b_CDS) from both species.
Metabolic engineering for the optimum synthesis of the desired product frequently requires multiple modifications which could be limited to a few selectable markers. To overcome this problem we constructed an AmpR marker cassette flanked by FRT sites (TU2), which are recognized by the yeast Flp recombinase, promoting the excision of the DNA region located between the FRTs. As K. rhaeticus genome is highly GC-rich, the competition´s sponsor IDT synthesized and FLP gene (FLPe) (Akbudak and Srivastava, 2011) with codon usage optimized for the host.
Altogether, we would like to integrate a cassette composted of 1 kb homology arm (TU1), recyclable antibiotic resistance marker (TU2), LexRO transcriptional unit (TU3), followed by a LexRO responsive promoter controlling the expression of cmcax or bcsA (TU4 including 1 kb downstream homology arms). To that end, it is necessary to PCR amplify the upstreams arms TU1 and genomic DNA fromK. rhaeticus AF1 or K. medellinensis, and clone into BBa_J428381 by GoldenGate with SapI Type IIs restriction enzyme and T4 DNA ligase.
Our TU2 type part was constructed by PCR amplification of the AmpR cassette from iGEM part BBa_J428385, flanked by FRT recombination sites, SapI restriction sites, and TU2 type overhangs, utilizing primers. This PCR product was cloned into BBa_J428382 by GoldenGate with SapI Type IIs restriction enzyme and T4 DNA ligase.
TU3 type parts encoding different strength promoters, strong RBS, synthetic codon-optimized LexRO encoding gene, and strong terminator are being constructed by cloning individual parts into pSB1C3SA, pSB1C3SB, pSB1C3C, and pSB1C3SD, respectively, followed by GoldenGate TU3 assembly with SapI and T4 DNA ligase into BBa_J428383.
TU4 type parts encoding the synthetic LexRO binding site with RBS, followed by either cmcax or bcsA (1 kb 3´homology) are being constructed by Golden gate assembly of BBa_K4435016 with the PCR products of amplifications from genomic DNA from either K. rhaeticus AF1 or K. medellinensis.
Recombination of FRTs require the expression of FLP recombinases, therefore we design a strategy to express either yeast FLP (BBF10K_000210) or newly designed enhanced FLP, called FLPe, which was codon optimized for Komagataeibacter (BBa_K4435002). We are cloning the recombinases under the control of strong promotors (BBa_K4435105; BBa_K4435106) for transformation into Komagataeibacter once correct genomic integration of the bcs locus cassettes is verified.
Synthesis of cellulose by the bcsI complex follows a series of steps that include glucose (GLC) import, conversion into D-glucose-6-phosphate (G6P) by glucokinase (GK), which is then converted into D-glucose-1-phosphate (G1P) by UTP-glucose-1-phosphate uridylyltransferase (galU), followed by UDP-glucose (UDPG) production by hosphoglucomutase (pgm), which requires UTP produced by nucleoside diphosphate pyrophosphorylase (ndp). UDPG is a substrate for cellulose synthesis by bcsA, which requires cyclic diguanylic acid (C-di-GMP) produced by diguanylate cyclase (dgc).
It has been postulated that expression levels of GK, galU, pgm, ndp and/or dgc could redirect glucose towards cellulose production, increasing final yields. To identify the limiting enzyme in the pathway and device a strategy for maximum BC production, we designed primers and amplified genes encoding GK, galU, pgm and ndp from genomic DNA of K. rhaeticus AF1 and K. medellinensis. These are being cloned into BBa_J428381, BBa_J428382, BBa_J428383, and BBa_J428384, under the control of the strong constitutive promotor BBa_K4435012, RBS BBa_K4435015 and terminator BBa_K4435017 (level 1 composite parts BBa_K4435120 to BBa_K4435127). Parts BBa_K4435120 to BBa_K4435123 will be combined into BBa_K4435313 to produce BBa_K4435217 and parts BBa_K4435124 to BBa_K4435127 will be combined into BBa_K4435313 to produce BBa_K4435218. BBa_K4435217 and BBa_K4435218 will be transformed into Komagataeibacter and BC production quantified. If there is an increase in cellulose production, we will test individual parts to determine the limiting enzyme in the pathway. All these used primers and amplified genes are disponible at parts page (link).
| Primers F | Sequence | Primer R | Sequence |
|---|---|---|---|
| galU_CDS_F | GGTCTCTAATGGTGATTAAGCCCCTTAAAAAAGCC | galU_CDS_R | GGTCTCTAAGCTCATTTATACTTTTTCAGGAATTCCC |
| pgm_CDS_F | GGTCTCTAATGCCCAGCATAAGCCCGTTTGC | pgm_CDS_R | GGTCTCTAAGCTCAGCCCACCGTCTCAGCC |
| ndp_CDS_F | GGTCTTAATGGCAGTCGAACGTACCCTC | ndp_CDS_R | GGTCTCTAAGCTCAGGGCAGGATTTCGGTGC |
| GK_CDS_F | GGTCTTAATGAAAGAGATCGTAGCAGTTGATATCG | GK_CDS_R | GGTCTCTAAGCTTAGCGATAAGCCTCGGCAAAGG |
C-di-GMP is an essential activator of bcsA, hence we ordered with IDT the gene encoding the enzyme dgc, responsible for c-di-GMP synthesis. This is being cloned into the pSB1C3C backbone (new basic part BBa_K4435003). BBa_K4435003 will be cloned into BBa_J428384, with BBa_K4435012, BBa_K4435015 and BBa_K4435022 (new composite part BBa_K4435128). This will be combined with level 1 parts type BBa_J428381, BBa_J428382 and BBa_J428383 in BBa_K4435306. Level 2 composite parts will be transformed into Komagataeibacter for BC production.