Results
Promoter of B488_05770 was deduced to possess specificity to LdtR
LdtR is the only MarR family member transcriptional factor in Liberibacter asiaticus[1], the panthogen that caused citrus greening. LdtR is specific to CLas and its close family members. In an attempt to obtain a biosensor that specifically and efficiently respond to LdtR, we proposed to find promoters specifically activated by LdtR. Previous research have revealed regulatory relationships between LdtP promoter, ZnuA2 promoter and LdtR [1,2,3]. P_LdtP (B488_10900) and P_ZnuA2 (B488_07960) were subsequently selected as candidates for the sensor. Previous research sequenced whole transcriptome of Liberibectar crescens (Lcr) with functional LdtR or inhibited LdtR [1]. By analyzing the published dataset and comparing differentially expressed genes (DEGs) with the threashold of foldchange < 0.9 or > 1.1, adjust p value < 0.05, we identified 121 down-regulated genes and 131 up-regulated genes (Figure 1). Under the condition of LdtR inhibition, down-regulation of expression might indicate potential positive regulatory roles of LdtR on the gene's promoter. Interestingly, the gene with unknown function B488_05770 was dramatically down regulated with a fold change of 0.123 when LdtR was inhibited (Figure 1). This suggested strong posibility that P_B488_05770 might positively and specifically respond to LdtR. Therefore, B_488_05770 promoter was selected as one of the sensor candidates for further investigation.
Figure 1. DEGs under conditions with and without intact LdtR. DEGs were identified with the threashold of foldchange < 0.9 or > 1.1, adjust p value < 0.05. Volcanal plot was generated by ggplot2. The arrow annotated gene B488_05770.
P_B488_05770 and P_LdtR mediated sensor specifically detected LdtR
E. coli is a classic substitution of unculturable CLas bacteria for molecualar research, such as the study of viral factor OptA [4]. To construct a substitutive model organism our LdtR-centered research, we introduced the LdtR transcriptional factor into E. coli BL21 (DE3). The protein coding sequences of LdtR (HA tagged) was retrieved from Genbank (Gene ID: 8223535), which were codon-optimized for E. coli expression using Genscript's Gensmart codon optimization services (Genscript, China), which were cloned into kanamycin resistant pRSF plasmid backbone. The recombinant plasmid pRSF-LdtR were subsequently transformed into E. coli BL21 (DE3) and selected by kanamycin. To check the expression status of LdtR, western blot targetting HA tag on LdtR and Commassie Blue staining assays were performed, the results of which suggested constitutive expression of the LdtR transcriptional factor in E. coli BL21 (DE3), condirming the success construction of LdtR expressing strain for further studies (Figure 2).
CLasMV1 is a natural phage that lives on CLas, which obtained a symbiosis relationship for unknown reasons [5,6]. The capsid protein of CLasMV1 constructs the outer layer of CLasMV1, which is essential for CLas recognition and invasion. In order to conquer the problem of lacking phage genome material, we only selected the essential part of CLasMV1, which is capsid protein, to create a phagemid expression system for testing. CLasMV1 capsid protein sequence (FLAG tagged) was acquired from Genbank (GenBank: CP045566.1), cloned into Ampicillin resistant pBluescript backbone. Promoters of LdtP, ZnuA2, and B488_05770 were cloned to control CLasMV1 capsid protein (BBa_K4587198, BBa_K4587197, BBa_K4587196) (Figure 2). Recombinant phagemids was either transformed into previously constructed LdtR-expressing E. coli strains or wild type E. coli BL21 strains containing no plasmid. Capsid protein expression was detected and quantified with Western Blot (targeting FLAG tag, Figure 2). P_ZnuA2 showed greatest expression efficiency under the activation of LdtR, however, the considerable expression quantity without LdtR suggested non-specificity of this candidate sensor. The specificity of P_LdtP was obviously higher with almost no expression without LdtR, but its efficiency was relatively low. Interestingly, P_B488_05770 showed specificity to LdtR, and a moderate efficiency in expression exhibition with the activation of LdtR (Figure 2). The severe expression leakage of P_ZnuA2 controlled expression cassette suggested unspecificity of P_ZnuA2 as a sensor. In contrast, P_LdtP and P_B488_05770, especially P_B488_05770 seemed to be specifically activated by LdtR, which were selected as sensors of CLas to be further optimized.
Figure 2. LdtR expressing strain and sensor selection. a. schematic view of LdtR expression cassette that was cloned in pRSF; b. schematic view of sensor-capsid expression cassette that was cloned in pBluescript; c. Coomassie blue staining results. Bacteria were induced with or without IPTG for 12 hours under 25°C. LdtR-HA has the size of ~26kDa. LdtR-OE: LdtR expressing E. coli strain; WT: Wild type E. coli BL21 (DE3); d. Western blot result with rabbit anti-HA antibody as the primary antibody; Bacteria were induced with or without IPTG for 12 hours under 25°C. LdtR-HA has the size of ~26kDa. LdtR-OE: LdtR expressing E. coli strain; WT: Wild type E. coli BL21 (DE3); e. Western blot results with rabbit anti-FLAG antibody as primary antibody. Bacteria were induced with IPTG for 12 hours under 25°C. Capsid-FLAG has the size of ~55kDa. LdtR: the LdtR expressing E. coli strain background; P_LdtP: Bacteria with phagemids featuring LdtP promoter; P_ZnuA2: Bacteria with phagemids featuring ZnuA2 promoter; P_B488_05770: Bacteria with phagemids featuring B488_05770 promoter.
RinA amplification system significantly magnified signal conferred by P_B488_05770 mediated sensor
To magnify the expression of our downstream genes including AMP and dsRNA, we simplified and adopted a RinA amlifier system from iGEM 2019 BEAS China to better cater to our needs [7]. Sequences of RinA amplifier and RinA-speciific promoter were acquired from Genbank (NCBI Reference Sequence: YP_001285353.1) and cloned into ampicillin resistant pBluescript phagemid backbone containing AcGFP reporter under control of P_LdtP or P_B488_05770 (BBa_K4587195, BBa_K4587194) (Figure 3). The recombinant phagemids were transformed into previously constructed LdtR-expressing E. coli BL21 strains. Expression level of signal GFP was detected by fluorescence intensity at 489nm. RinA has proven to amplify the signal around 20-fold in P_B488_05770 sensor system and 2-fold in P_LdtP sensor system (Figure 3). Given varying performances of RinA amplifier in P_LdtP mediated and P_B488_05770 mediated sensors, P_B488_05770 mediated sensor amplified by RinA system was obviously the optimal system to specifically sense the presence of LdtR and efficiently activate downstream genes.
Figure 3. RinA amplifier amplifies signal conferred by sensor. a. Schematic view of sensor-RinA amplifier-GFP cassette that was cloned into pBluescript; b. Fluorescence intensity at 488nm of sensor-RinA amplifier-GFP system after 24h in LdtR expressing E.coli strain.
Sequence and structural optimization of MaSAMP by ProteinOpti
The amino acid sequence of MaSAMP was found in the appendix of Huang et al., 2021 [7]. Codon optimization was done using VectorBuilder to reach rational GC% and high Codon Adaption Index (CAI) for the E. coli expression system (Figure 4). The amino acid sequence of MaSAMP was found in the appendix of Huang et al., 2021[7]. Amino acid optimization was conducted via substituting amino acids in the hydrophilic face with single of multiple lysine. From the result of single and double substitution, the lyticity hardly changes, which indicates that our substitution will not enhance lysing human cells via software calculating. Total hydrophobicity slightly increases after substitution while net charge increases immensely. The total positive charges promote MaSAMP to bind with the negative cell membrane. As a result, we arranged the mutants based on the net charge from low to high (top to bottom) and selected 6 candidates with highest net charge for testing. For efficiency, we chemically synthesized the second alpha helix of the six candidates considering the second alpha helix is the main function domain. The 3D hydrophobic Moment vector will be further used in figuring out the interaction between MaSAMP mutants and bacteria membrane.
Figure 4. Obtain MaSAMP and Optimization. a. The pasted sequence is the amino acid sequence of MaSAMP. The improved DNA has 47.55% GC content and 0.98 Codon Adaption Index (CAI); b. Red characters point out the position substituted. Blue characters are the main function domain (second alpha helix). The following biophysical parameters are calculated via our assembled toolkit – ProteinOpti.
AMPs could be biosynthesized for massive tests
Since chemical peptide synthesis is expensive, time consuming, and not practical for massive testing assays, we proposed to biologically synthesize AMPs for our usages. Detailed description of the sythesis genetic module can be found in [Enginnering Success]. In brief, AMP gene sequences were marked by GFP as reporter, as well as SUMO tag for toxicity concealment. After massively synthesis by high efficient T7 RNApol-promoter E. coli system, AMP could be purified after protease cleavage of SUMO. Recombinant plasmids were constructed by cloning designed sequences BBa_K4587209 (FAMP), BBa_K4587210 (AMP), and BBa_K4587215 (BAMP) separately into pET-30a(+) with Kanamycin-resistant gene at the Ndel & Xhol restriction enzyme sites. The recombinant plasmids were subsequently transformed into BL21 (DE3) selected by kanamycin (Figure 5). The size of plasmid FAMP, AMP, and BAMP are 6513 bp, 6477 bp, and 6498 bp respectively, while the size of empty plasmid pET-30a(+) is 5422 bp. Clear band size differences between recombinant plasmids and the pET-30a(+) empty plasmid were found in the gel electrophoresis results after plasmid extraction (Figure 5). Two or more bands in one lane were prone to be various conformations of plasmids such as supercoiled, nicked circular, and linear conformation. These results indicated successful construction of bacteria for AMP bio-manufacturing.
After induction with 1mM IPTG at 37°C for 12 hours, green fluorescence signal from GFP were visualized by fluorescent microscope. Strong green fluorescence signal was exhibited in both the FAMP and AMP group, indicating the successful expression of GFP in the corresponding recombinant plasmids (Figure 5). The AMP group showed a strong green fluorescence signal even without IPTG induction. The green fluorescence signal was relatively weak compared to the FAMP and AMP group. To further confirm the expression of our fusion protein with the exact size, Tris-SDS-PAGE and multiple western blots were performed for each group (Figure 5). According to our design, the size of the fusion protein for FAMP, AMP, and BAMP group were 47.4 kDa, 47.4 kDa, and 48.2 kDa respectively. Large amount of expression was observed in both supernatant and pellet of AMP group after ultrasonication compared to the control groups (Figure 5). Much thicker bands were observed in FAMP and AMP groups, indicating higher levels of fusion protein expression, which were confirmed by Tris-SDS-PAGE results (Figure 5). Therefore, we proceeded our experiment with the FAMP and AMP.
Figure 5. Consturction of AMP biosynthesis system. a. Transformation success of recombinant plasmids into E. coli BL21 (DE3). 50uM Kanamycin was used for selection; b. Gel electrophoresis of recombinant plasmids extracted from transformed E. coli that were amplified overnight; c. Fluorescence microscopic imaging of engineered E. coli BL21.FAMP; d. Western blot of both supernatant and pellet of E. coli BL21 with recombinant plasmid (AMP) after ultrasonication using 6X His tag antibodies, while E. coli having pET-30a(+) plasmid without our designed sequence (Empty) and without plasmid were tested as controls. Expression under both IPTG induction and without induction was examined; e. Western blot of supernatant of E. coli BL21 with recombinant plasmids FAMP, AMP and BAMP. Expression under both IPTG induction and without induction was examined; f. Tris-SDS-PAGE and Coomassie brilliant blue staining results of recombinant protein expression in E. coli BL21 with recombinant plasmids AMP and FAMP. pET-30a(+) was tested as a control; g. Western blot result of recombinant protein expression in E. coli BL21 with recombinant plasmids AMP and FAMP using Anti-6X His tag antibodies. The red rectangle indicates the recombinant protein size and expression.
Purified AMP (showed high lethality to Lcr)
Large-scale extraction and purification of our fusion protein were performed after successful small amount extraction. Recombinant E. coli strains were diluted to OD600=0.1 in 100 mL LB respectively in 400 mL conical flask and cultured at 37°C to reach the OD600=0.4~0.6. IPTG induction was followed for another 6~8 hrs. After induction with 1mM IPTG at 37°C for 12 hours, whole protein were extracted with High-Pressure Homogenizer. For protein purification, we used the His-tag Protein Purification Kit produced by Beyotime to purify the fusion protein in the FAMP group. The bands in lane 5 of Figure 6 were purified fusion protein. However, after SUMO protease cleavage, no significant band size change and no band near 7kD was observed, which means the MaSAMP was not successfully released from our fusion protein (Figure 6). After troubleshooting, we found that most of our fusion protein was in the pellet of E. coli lysate. For the AMP group, we did not find any fusion protein in the supernatant of E. coli lysate, while the pellet had protein of the correct size (Figure 6). For the FAMP group, we can also see abundant fusion protein in the pellet of lysate compared to the supernatant, and no thick band for purification product (Figure 6). Therefore, we adjusted our protocol to conduct inclusion body purification to purify our fusion protein.
Figure 6. Bio-manipulation of AMP. a. 10% Tris-SDS-PAGE gel. Lane 1: Supernatant after ultrasonication and centrifugation; Lane 2: Supernatant after chelation; Lane 3: Supernatant after first wash; Lane 4: Supernatant after second wash; Lane 5: Supernatant after elution; Lane 6: SUMO protease digestion at 4°C; Lane 7: SUMO protease digestion at 30°C; b. 18% Tris-SDS-PAGE gel. Lane 1: Supernatant after ultrasonication and centrifugation; Lane 2: Supernatant after chelation; Lane 3: Supernatant after first wash; Lane 4: Supernatant after second wash; Lane 5: Supernatant after elution; Lane 6: SUMO protease digestion at 4°C, +salt; Lane 7: SUMO protease digestion at 4°C, -salt; c. 10% Tris-SDS-PAGE gel of recombinant E. coli lysate in AMP group. Lane 1: Mixture of supernatant and pellet after using High-Pressure Homogenizer; Lane 2: supernatant of lysate; Lane 3: pellet of lysate; d. 10% Tris-SDS-PAGE gel of recombinant E. coli lysate and purified product after washing and eluting using gradient imidazole solution. Lane 1: Pellet after lysing cells and centrifuging; Lane 2: Beads after final elution; Lane 3: First elution (0.3M imidazole solution); Lane 4: 0.03M imidazole solution Washing; Lane 5: Second elution using 0.3M imidazole solution; Lane 6: Supernatant after extraction; Lane 7: 0.06M imidazole solution; Lane 8: Lysis solution washing.
Optimized AMP 6 showed high lethality to Lcr
SAMP_5: BBa_K4587222, SAMP_6: BBa_K4587224) of the MaSAMP second alpha-helix generated by the ProteinOpti software (developed by DKU iGEM team) were synthesized by GeneScript and dissolved in DMSO to make 30 µM SAMP solutions. Equal volumes of DMSO and 30 µM SAMP solutions (including unmutated SAMP and 6 mutants) were added to the Lcr culture at OD600=0.4, and 3 replicates for each group were made. After 5 hours, the OD600 of each culture was measured using Microplate Reader (Figure 7). The Abs OD600 of SAMP_4 and SAMP_6 was lower than that of SAMP, exhibiting better sterilizing effect on Lcr compared to unmutated SAMP. Other mutants have a variation of sterilizing effects, but they are not as effective as unmutated
Figure 7. The measurement of the OD600 absorption for SAMP mutants. Mean (±1 SD) OD600 of Lcr culture in different SAMP solutions after 5 hours of culturing (n=3). SAMP showed a significant sterilizing effect compared to the control group DMSO (p=0.003215<0.05), while our mutants SAMP_4 and SAMP_6 had more significant sterilizing effect (p=0.0019, p=0.002947) than SAMP.
dsRNA can be biosynthesized for massive tests
We designed our primers for our targeted genes- BBa_K4587003 (DCR), BBa_K4587006 (CP450), BBa_K4587009 (MP20), BBa_K4587012 (CP64), BBa_K4587015 (DM2). The lengths of these genes are respectively 399bp, 1836bp, 757bp, and 480bp. The mRNA sequences of these primers (both forward and reverse) are BBa_K4587004 (DCR forward), BBa_K4587005 (DCR reverse), BBa_K4587007 (CP450 forward), BBa_K4587008 (CP450 reverse), BBa_K4587010 (MP20 forward), BBa_K4587011 (MP20 reverse), BBa_K4587013 (CP64 forward), BBa_K4587014 (CP64 reverse), BBa_K4587016 (DM2 forward), BBa_K4587017 (DM2 reverse). Figure 8a shown below demonstrates the successful amplification of CP450 sense and antisense sequence
We constructed our plasmids by inserting our target genes (BBa_K4587006, BBa_K4587009, BBa_K4587012, BBa_K4587015) respectively into L4440 plasmid with T7 promoter and Ampicillin resistance gene at the KpnI & HindIII (for sense sequence) and SacI & XbaI (for antisense sequence) restriction enzyme sites. The gel results as shown in Figure proves that both restriction enzymes work well since we can see that the band of the uncut vector shows a smaller band than the +RE ones, meaning that the RE ones are linear instead of circular, though concentration problems exist. The size of the undigested L4440 plasmid is 2790bp. A hairpin structure is formed since there’s a linker between the sense and antisense sequence.
We successfully transformed different L4440 plasmids (DM2, MP20, CP64, CP450 respectively) into the HT115 E. coli strain following the Transformation protocol and cultured them. Small colonies can be seen in the following figures (Figure 8b), which are the successfully cultured HT115 E. coli that can produce hairpin RNA with IPTG induction. However, for Figure 8b, it demonstrates the successfully constructed L4440 plasmid with the CP450 antisense in DH5α bacteria.
We successfully produced hpRNA for MP20, CP64, and DM2 genes by transforming the HT115 E. coli strain and utilizing the CTAB miniprep protocol (Figure 8c).
Figure 8. Success biosynthesis of dsRNA. E. coli HT115 strains with dsRNA synthesis functions were constructed and efficiently produce dsRNA. a. Gel-electrophoresis of testing the function of restriction enzymes; b. Success transformation of L4440 + CP450, DM2, MP20 with T7, DM2 with T7, CP64, CP64 with T7, and MP20 (left to right, top to down); c. Gel electrophoresis result of hpRNA induction from plasmids extracted from HT115(DE3) bacteria, which were induced with IPTG and treated with RNase A.
References
1: Pagliai FA, Coyle JF, Kapoor S, Gonzalez CF, Lorca GL. LdtR is a master regulator of gene expression in Liberibacter asiaticus. Microb Biotechnol. 2017 Jul;10(4):896-909. doi: 10.1111/1751-7915.12728. Epub 2017 May 15. PMID: 28503858; PMCID: PMC5481520.
2: Pagliai FA, Pan L, Silva D, Gonzalez CF, Lorca GL. Zinc is an inhibitor of the LdtR transcriptional activator. PLoS One. 2018 Apr 10;13(4):e0195746. doi: 10.1371/journal.pone.0195746. PMID: 29634775; PMCID: PMC5892913.
3: Pagliai FA, Gardner CL, Bojilova L, Sarnegrim A, Tamayo C, Potts AH, Teplitski M, Folimonova SY, Gonzalez CF, Lorca GL. The transcriptional activator LdtR from 'Candidatus Liberibacter asiaticus' mediates osmotic stress tolerance. PLoS Pathog. 2014 Apr 24;10(4):e1004101. doi: 10.1371/journal.ppat.1004101. PMID: 24763829; PMCID: PMC3999280.
4: Hao G, Boyle M, Zhou L, Duan Y. The intracellular citrus huanglongbing bacterium, 'Candidatus Liberibacter asiaticus' encodes two novel autotransporters. PLoS One. 2013 Jul 11;8(7):e68921. doi: 10.1371/journal.pone.0068921. PMID: 23874813; PMCID: PMC3708911.
5: Zhang L, Li Z, Bao M, Li T, Fang F, Zheng Y, Liu Y, Xu M, Chen J, Deng X, Zheng Z. A Novel Microviridae Phage (CLasMV1) From "Candidatus Liberibacter asiaticus". Front Microbiol. 2021 Oct 13;12:754245. doi: 10.3389/fmicb.2021.754245. PMID: 34721359; PMCID: PMC8548822.
6: Wang C, Fang F, Li Y, Zhang L, Wu J, Li T, Zheng Y, Xu Q, Fan S, Chen J, Deng X, Zheng Z. Biological Features and In Planta Transcriptomic Analyses of a Microviridae Phage (CLasMV1) in "Candidatus Liberibacter asiaticus". Int J Mol Sci. 2022 Sep 2;23(17):10024. doi: 10.3390/ijms231710024. PMID: 36077424; PMCID: PMC9456138.
7: Huang CY, Araujo K, Sánchez JN, Kund G, Trumble J, Roper C, Godfrey KE, Jin H. A stable antimicrobial peptide with dual functions of treating and preventing citrus Huanglongbing. Proc Natl Acad Sci U S A. 2021 Feb 9;118(6):e2019628118. doi: 10.1073/pnas.2019628118. PMID: 33526689; PMCID: PMC8017978.