Co-expression of luxF and luxG to enhance the brightness of lux operon.

According to previous studies, Lux CDEABE genes from Photorhabdus luminescens enabled organisms emit luminescence without any additional source of substrates, which aligns perfectly with our goal to design a sustainable source of light: LAMPS. So, we utilized this basic lux operon as the base of our design. Additionally, the intensity of light is an essential criterium of LAMPS, leading us to involve two other parts: LuxF(Part:BBa_K4594002) and LuxG(Part:BBa_K4594003). Previous study has shown that LuxF would activate luciferases (consists of LuxA and LuxB) by binding to the inhibitor of luciferases. And LuxG also contributes to the luminescence with extra substrates FMNH2. In a conclusion, in this section we combined basic lux operon, luxG and luxF to form luxCDABEGF(Part:BBa_K4594004) and proved the composite part enabled E.coli to emit stronger luminescence than the ordinary lux operon, which is the basis of our whole design LAMPS.

Figure 1 The mechanism of light enhancement by luxF and luxG

Initially, the lux operon was available in a pGEN plasmid from ADD-GENE (code: 44918), with an em7 promoter. To increase proteins production for better observation, we selected host cells BL21(DE3), and the vector was changed into pET28a for an efficient T7 promoter. So, we cloned lux operon from the pGEN plasmid by PCR.

And we got the CDS of LuxF(Part:BBa_K4594002) and LuxG(Part:BBa_K4594003) in NCBI, which we synthesized in Gene Script within plasmid pET28a.

As we had obtained luxCDABE, luxF, luxG and pET28a vector, we then assembled them together by Gibson Assembly. According to Figure 2A, we obtained two fragments(luxG and luxCDABE) with homology arms added during PCR process, and the linear vector of pET28a-luxF. Then we got pET28a-luxCDABEGF(Part:BBa_K4594004) as shown in Fig2B.

As the recombinant plasmid had been accessed, we transferred it to E.coli DH5`\alpha` to magnify it. And we extracted the plasmids from it and testified it by Sanger sequencing. The sequencing result reflected that it was correct (Figure 2C).

Figure 2A Scheme representing the major steps to obtain pET28a-luxCDABEGF. Step1: Perform PCR to linearize the vector, amplify gene frament and add homology arms. Step2: Conduct Gibson Assembly to assemble luxCDABE_fragment, luxG_fragment and pET28a-luxF linear vector.

Figure 2B the final plasmid graph of pET28a-luxCDABEGF.

Figure 2C The sequencing result of pET28a-luxCDABEGF

Then the plasmid was transferred into BL21(DE3), a kind of competent cells adapted for protein expression. BL21(DE3) with pET28a_luxCDABEGF was first cultured on LB plates(Kanamycin) for 12h under 37℃. Then three monoclonals were respectively cultured in 5ml LB(Kanamycin) overnight. The overnight cultures were added to new 5ml LB(Kanamycin) with a ratio of 1:100 for about 3 hours under 37℃. And the growing curve was detected before induction.

When OD600 of cultures reached 0.6-0.8, the induction was initiated by 0.2 mM IPTG, followed by a 16-hour culture under 23℃. After that, we observed luminescence with a 3-second exposure by a camera. Here's the results (Fig 3A). Additionally, we magnified the culture in a 200ml LB for a better phenomenon. Both culture and detection manners were the same as above (Fig 2B).

Figure 3  The luminous E.coli BL21(DE3) with pET28a-luxCDABEGF, which was cultured for 16 hours under 23℃ after being induced by 0.2mM IPTG. The photo was captured by an ordinary phone camera with an exposure time for 3 seconds.

We detected the luminescence not only by naked eyes, but also by a luminous spectrum assay. The result was shown in Fig 4. The first column was luminous BL21(DE3) with an OD600 adjusted to 2.2 by additional LB; the second column was the negative control (BL21(DE3) transformed with original pET-28a with an OD600 of 2.2); and the third column was luminous BL21(DE3) with an unchanged OD600 of 4.2. It was obvious that the luminescence was successfully detected.

Figure 4 Luminescent test by a multimode plate reader. The first column was luminous BL21 with an OD600 adjusted to 2.2 by additional LB; the second column was the negative control (BL21 transformed with original pET-28a with an OD600 of 2.2); and the third column was luminous BL21 with an unchanged OD600 of 4.2. The test clearly showed that BL21(DE3) with pET28a-luxCDABEGF emitted strong luminescence.

Also, we conducted an experiment to determine the best IPTG concentration for luminescence. So, we set a gradient of 0, 0.05, 0.1, 0.2, 0.4, 0.8, 1.6, 3.2(mM). Both culture and detection manners were the same as above. Here comes the result (Figure 5). Results showed that the best IPTG concentration for induction was between 0.05mM and 0.2mM.

Figure 5 Effect of the inducer concentration on light emission. 3 replications were set for each IPTG concentration, and the main value was used. OD600 was adjusted to 2 for each sample. The first 8 columns were luminous BL21(DE3) with different IPTG concentration and the negative control was BL21(DE3) transformed with an original pET28a plasmid. All BL21(DE3) underwent a 16-hour culture under 23℃ after being induced by 0.2 mM IPTG.

In order to confirm whether luxG and luxF can enhance the brightness, we constructed plasmids with and without luxF,G and compared the luminescence of them. Here are the plasmid graphs of them. And after culturing and induction, manners of which were the same as those above, we got photos and luminescence data for comparison (Fig5A, 5B) Results clearly showed luxF and luxG strongly enhanced the brightness of designed BL21(DE3) (Figure 6A). And luminescence assay excitingly showed the addition of luxF and luxG increased the luminescence by 58% (Figure 6B). So, our design is definitely feasible.

Figure 6A The result of luminescence assay for BL21(DE3) with pET28a-luxCDABE and BL21(DE3) with pET28a-luxCDABEGF (n=6). The result showed that the addition of luxF and luxG increased the luminescence by 58%. All BL21(DE3) underwent a 16-hour culture under 23℃ after being induced by 0.2 mM IPTG. And the assay was conducted by a plate reader.

Figure 6B Comparing luminescence of BL21(DE3) with pET28a-luxCDABE or pET28a-luxCDABEGF. Images were taken by an ordinary phone camera. The left three tubes were BL21(DE3) with pET28a-luxCDABE and the right three tubes were BL21(DE3) with pET28a-luxCDABEGF. It was clear that luminescence can be enhanced by additional luxF and luxG. All BL21(DE3) underwent a 16-hour culture under 23℃ after being induced by 0.2 mM IPTG.

In this part, we successfully constructed pET28a-luxCDABEGF, and the host cell was enlightened as we have designed. Also, we determined the best concentration of IPTG to induce expression. And excitingly, we testified the capacity of luxF and luxG to enhance brightness, with a convincing data of 58%! It's undeniable that to enhance the brightness of LAMPS by luxF and luxG is feasible.

In order to increase the brightness and change the colour of the Lux operon system using the principle of biofluorescence resonance energy transfer (BRET). We ligated the yellow fluorescent protein cp157Venus (Part: BBa_K4594005) at the C-terminal of LuxB with a linker (Glu-Leu) to become a new fusion fluorescent protein LuxB:cp157Venus (Part: BBa_K4594006), and in this way constructed pET-28a_lacO-LuxA-LuxB (Part: BBa_K4594013) and pET-28a_lacO-LuxA-LuxB:cp157Venus (Part: BBa_K4594014) two plasmids, they were plasmids amplified by DH5`\alpha` strain and transfected into BL21 (DE3) strain, cultured at 37°C and induced to express at 24°C. E. coli transfected with pET-28a_lacO-LuxA-LuxB successfully luminesced, but E. coli with the fusion protein LuxB:cp157Venus did not luminesce. We speculate that the target protein expression was unsuccessful due to the occurrence of a gene mutation.

pET-28a_lacO-LuxA-LuxB

Since the BRET principle has been proven in various literature, we firmly believe that our experimental principle is correct. Therefore, we will now turn our attention to our ultimate goal. Immediately after that, in order to find brighter fluorescent proteins after passing BRET, we first queried and screened the FPbase database for the two brightest proteins with fluorescence data and with excitation light between 490-520 nm and named them as A1 (Part: BBa_K4594007) and A2 (Part: BBa_K4594008). We also mapped the fluorescence data of the fluorescent proteins with incomplete information from amino acid sequences to fluorescence data by using the trained Long Short-Term Memory (LSTM) system and then filtered out the brightest two proteins named B1 (Part: BBa_K4594026) and B2 (Part: BBa_K4594027) from the mapped fluorescent proteins.

We synthesised the four genes A1, A2, B1, and B2 by GenScript and fused them at the C-terminal of LuxB in LuxCDABEGF using the same linker (Glu-Leu) to form four different fusion fluorescent proteins, and then the constructed plasmids were transferred into BL21(DE3) for induced expression.

After expression, it was found that the four LuxCDABEGFs fused with the new proteins A1, A2, B1, and B2 emitted light of significantly different colours, thus proving the feasibility of the BRET system. However, since the BRET efficiency is related to the sixth power of the distance, the brightness of the fusion proteins is weaker compared to the original LuxCDABEGF at non-optimal distances (all four proteins use the same linker as before).

Afterwards, we conducted emission spectroscopy. The spectral data revealed that all four fused fluorescent proteins exhibited either a bimodal trend or a flat-head peak trend. The latter was due to the excitation light and emission light being close in frequency, and the brightness levels were fairly consistent across the board. The results of the spectral data proved the effectiveness of the LSTM system.

LuxB:A1 fluorescent protein after BRET the highest peak of emission spectrum should be around 525nm. The experimental results show a double peak. The peak at 490nm is the highest peak of emission light of the first protein. The experimental results are in accordance with the theoretical values.

The spectrum of LuxCDABEGF fused with A1

LuxB:A2 fluorescent protein should have the highest peak of the emission spectrum after BRET around 512nm. The experimental results show a double peak. The peak at 490nm is the highest peak of the emission light of the first protein. The experimental results are in accordance with the theoretical values.

The spectrum of LuxCDABEGF fused with A2

LuxB:B1 fluorescent protein after fluorescence resonance energy transfer the highest peak of the emission spectrum should be around 507nm, the peak at 490nm is the highest peak of the emission light of the first protein, but due to the first highest peak of the emission light is close to the second one, so the experimental double peaks are not obvious, and it becomes a flat head peak at 490-510nm. The experimental results are in line with the theoretical values. As far as the vertical axis RLU value is concerned, it is not much different from the A series fluorescent proteins and is in the same order of magnitude, which is in line with the prediction of the LSTM system and proves that the model prediction is successful.

The spectrum of LuxCDABEGF fused with B1

LuxB:B2 fluorescent protein after fluorescence resonance energy transfer the highest peak of the emission spectrum should be around 506nm, the experimental results show a double peak, the peak at 490nm is the highest peak of the emission light of the first protein. The experimental results are in line with the theoretical values. As for the RLU value of the vertical axis, it is not much different from the A series fluorescent proteins and is in the same order of magnitude, which is consistent with the prediction of the LSTM system and proves that the model prediction is successful.

The spectrum of LuxCDABEGF fused with B2

Finally, in order to obtain fluorescent proteins with brighter excitation light between 490-520 nm, we decided to break away from nature and create fluorescent proteins by ourselves using computer technology. We used Generative Adversarial Network to generate the amino acid sequences of the candidate fluorescent proteins, and then screened the generated sequences by LSTM system, and finally got the two brightest fluorescent proteins and named them as C1 (Part: BBa_K4594028) and C2 (Part: BBa_K4594029) . Since the effectiveness of the LSTM system has been proved in the previous experiments, the fluorescent proteins created as C1 and C2 have a certain degree of credibility.

After consulting the literature, we found that PCC7942 has some degree of ampicillin resistance. We inoculated algal culture in 7 groups of 5 ml liquid media with ampicillin concentrations ranging from 0-120 μg/ml, monitoring the growth. The goal was to identify a concentration of the antibiotic that prevents contaminants and allows good cyanobacterial growth. The specific data is as follows:

We found that, except for the group with a concentration of zero antibiotics, all other groups of cyanobacteria exhibited significant decline in growth. This indicates that directly adding antibiotics to the culture medium cannot prevent cyanobacterial contamination.

We inoculated algal culture in 8 groups of 1 ml liquid media with ampicillin concentrations ranging from 0-100 μg/ml and monitored their growth. We then streaked the culture from groups 6 and 7 onto antibiotic-free BG-11 agar plates. After 6-7 days, green colonies were observed, and single colonies were picked and cultured in 5 ml of antibiotic-free BG-11 liquid media while monitoring their growth. We observed poor growth, with OD750 reaching its highest value of 0.26 after 16 days. We also validated this by streaking on antibiotic-free LB agar and found that contaminants were still present.

After identification, we found that our cyanobacteria were contaminated by a nitrate-reducing bacterium, Nitratireductor sp.

The original light intensity used was 10,000 lux. After consulting with Professor Zhao Quanyu (see Human Practice for details), we learned that excessive light intensity, combined with antibiotic-containing culture medium, could induce stress in cyanobacteria. This may have been one of the reasons for the long-term transformation failures. Ultimately, we adopted a light intensity of 4,000-6,000 lux, and we were able to achieve normal growth of cyanobacteria on solid culture media.

Based on previous experiments and discussions with Professor Zhao Quanyu from Nan Jing Tech University, we speculated that the initial cell density of inoculation could affect the growth rate of cyanobacteria. This may have been a reason why cyanobacterial growth was poor when isolating colonies in liquid culture media. Therefore, we inoculated HL7942 cyanobacterial strains from solid culture media into liquid media with volumes of 0.5, 1, 2, and 5, and measured growth curves.

We assumed that OD750 was linearly related to cell density and multiplied OD750 by the volume to obtain the relative quantity of cells. The graph is as follows:

It seems that inoculated in 1mL or 2mL is optimal.

This experiment provided preliminary validation, but it was not repeated, and the grouping was coarse. Further investigation is needed.

We started with the plasmid pUC57-NS3-2-lacUV-cscB-lacI-KanR-NS3-1[BBa_K4115045] left by the shanghaitech_China 2022 team as the basis and used Gibson Assembly to construct our transformation plasmid, as shown in the diagram.

original plasmid pUC57-NS3-2-lacUV-cscB-lacI-KanR-NS3-1

Our plasmid construction underwent two rounds of Gibson Assembly. We replaced CscB with sfGFP(with LVA tag)[BBa_K4115001] containing a degradation tag, and at the same time, we replaced the lacUV promoter with the final output promoter PKaiBC[BBa_K4594011], which is an intrinsic rhythm promoter in cyanobacteria. Finally, we constructed the plasmid pUC57_NS3-2-PKaiBC-sfGFP-lacI-KanR-NS3-1.[BBa_K4594019]

constructed plasmid pUC57_NS3-2-PKaiBC-sfGFP-lacI-KanR-NS3-1.

After constructing the plasmid, we attempted to insert them into the cyanobacterial genome. Although the filamentous cyanobacteria can naturally take up foreign DNA, they cannot stably maintain plasmids.

After mixing the culture with the plasmid and overnight incubation, we plated the mixture on solid agar medium containing kanamycin to select for recombinant cyanobacterial colonies. After multiple experiments, we finally obtained cyanobacterial colonies on kanamycin-containing solid agar medium. Unfortunately, these cyanobacteria grew too densely, making it impossible to pick single clones.

First, we designed primers at the NS3 ends and performed PCR experiments, revealing that the transformed HL7942 did not have the desired fragment inserted, but PCC7942 showed a different pattern from the negative control group. Therefore, we scraped a bit of cyanobacteria transformed with the pUC57_NS3-2-PKaiBC-sfGFP-lacI-KanR-NS3-1[BBa_K4594019] plasmid as a template and designed primers on the left side of PkaiBC and the right side of sfGFP for PCR. After electrophoresis, we observed a clear and single band of the correct length (approximately 1236 bp).

Similarly, using cyanobacteria transformed with the pUC57_NS3-2-PKaiBC-luxCDABEFG-KanR-NS3-1[BBa_K4594020]. plasmid as a template, we designed primers at both ends of the lux gene cluster, and we also obtained the correct band (approximately 7212 bp). This indicates that there were successfully transformed PCC7942 colonies on several plates, but possibly due to antibiotic inactivation or insufficient concentration, these solid agar plates did not provide adequate selection, leading to the co-growth of wild-type cyanobacteria with the successfully transformed ones.

Since neutral integration sites, PKaiBC, and other elements are present in the cyanobacterial genome, we aimed to obtain these gene fragments from cyanobacteria using PCR technology. We attempted various methods, including repeated freeze-thawing, sonication, and genome extraction, to prepare PCR templates. Although the sonication method achieved some success in obtaining the NS2 element (as seen in the figure below), it may be because sonication randomly disrupts the cyanobacterial genome, and this method of preparing templates was not successfully reproduced.

The first lane to the eighth lane represents ultrasonication for 0s, 20s, 30s, 60s, 90s, 120s, 180s, and 240s, respectively. Only the result of ultrasonication for 30s was positive.

Ultimately, we referenced the PCR method from Toulouse_INSA-UPS in 2021 and attempted to prepare PCR templates by repeated freeze-thawing at -80°C/60°C. We successfully obtained the NS3 fragment from HL7942, and this method produced clear and bright bands in PCR for three different samples.

We obtained a pure and improved PCC7942 strain from Luan Guodong of the Institute of Hydrobiology for use in transformation experiments. Use the natural transformation method (see protocol for details) to perform the transformation operation, and spread the bacterial solution on a BG11 solid plate containing 10 μg/mL kanamycin. Colonies grew after about 2 weeks. As shown below:

Due to lack of time, we were unable to perform characterization experiments on PKaiBC, so colony PCR was performed to verify whether the target gene was transferred.

We first used primers at both ends of NSIII and primers at both ends of the PkaiBC-sfGFP fragment to perform colony PCR, which were set as group 1 and group 2 respectively. Group 1 is expected to obtain a fragment band inserted into NSIII by homologous recombination, with a length of approximately 7000 bp, while Group 2 is expected to obtain a PKaiBC-sfGFP fragment band with a length of approximately 1200 bp.

The results of gel electrophoresis after PCR are as follows:

Lanes 1, 2, and 3 from left to right are the selected successfully transformed colonies, lane 4 is the blank control, and the far right is the DNA Marker.

It can be seen that compared with the control group, the two groups have obvious bands of about 1200 bp, which proves that the transformation plasmid we constructed has been successfully transferred.

However, almost no correct bands were obtained in group 1. Article evidence shows that Synechococcus PCC7942 is polyploid [1]. We hypothesized that there may still be a blank NSIII fragment (approximately 1800 bp) in the successfully transformed strain, and the blank The fragment binds to the primer more readily than the inserted foreign fragment and undergoes chain growth catalyzed by the polymerase.

However, we obtained single-colony transformants and demonstrated the success of the transformation.

[1] Watanabe S. Cyanobacterial multi-copy chromosomes and their replication. Biosci Biotechnol Biochem. 2020 Jul;84(7):1309-1321. doi: 10.1080/09168451.2020.1736983. Epub 2020 Mar 11. PMID: 32157949.