Vitamin B12 as a carrier of peptide nucleic acid (PNA) into bacterial cells.

Short modified oligonucleotides targeted at bacterial DNA or RNA could serve as antibacterial agents provided that they are efficiently taken up by bacterial cells. However, the uptake of such oligonucleotides is hindered by the bacterial cell wall. To overcome this problem, oligomers have been attached to cell-penetrating peptides, but the efficiency of delivery remains poor. Thus, we have investigated the ability of vitamin B12 to transport peptide nucleic acid (PNA) oligomers into cells of Escherichia coli and Salmonella Typhimurium. Vitamin B12 was covalently linked to a PNA oligomer targeted at the mRNA of a reporter gene expressing Red Fluorescent Protein. Cu-catalyzed 1,3-dipolar cycloaddition was employed for the synthesis of PNA-vitamin B12 conjugates; namely the vitamin B12 azide was reacted with PNA possessing the terminal alkyne group. Different types of linkers and spacers between vitamin B12 and PNA were tested, including a disulfide bond. We found that vitamin B12 transports antisense PNA into E. coli cells more efficiently than the most widely used cell-penetrating peptide (KFF)3K. We also determined that the structure of the linker impacts the antisense effect. The results of this study provide the foundation for developing vitamin B12 as a carrier of PNA oligonucleotides into bacterial cells.

Introduction

The rapid development and spread of antimicrobial resistance motivates the search for new antibiotics. In principle, the use of modified sequence-specific oligonucleotides as steric blockers of bacterial RNA or DNA seems a promising strategy. Traditional antisense strategies use short oligonucleotides that hybridize with complementary mRNA sequences through Watson-Crick base pairing and block translation[1, 2]. The advantage of this approach is that oligomer sequences can be rapidly redesigned if bacterial resistance arises due to mutation of the target. Since natural oligonucleotides are rapidly degraded in the intracellular environment, chemically-modified oligonucleotides such as Peptide Nucleic Acids (PNA)[3] have been used (Fig. 1a).

Figure 1

(a) Chemical structure of a PNA oligomer and its complementary pairing scheme with natural RNA. (b) Vitamin B12 in the form of cyanocobalamin used in this study. R5 refers to the fifth carbon of the ribose moiety, while R5′-OH stands for the primary hydroxyl group according to the nomenclature of cobalamin[4].

PNA oligomers containing a pseudo-peptide instead of a sugar-phosphate backbone show improved nuclease resistance, lower toxicity and increased affinity of hybridization with natural nucleic acids[5]. PNAs have been successfully tested as antimicrobials in a variety of bacterial species[2, 6, 7]. To inhibit bacterial growth, the mRNAs of several essential genes have been targeted, including acpP, the gene for the acyl carrier protein, gyrA encoding DNA gyrase subunit A, and murA and fabI, genes involved in cell wall and fatty acid biosynthesis, respectively[3, 8, 9].In a separate approach, functional fragments of both 23S and 16S rRNA have been verified as targets for antisense PNAs[10-13].

The most serious drawback, which currently precludes the use of this approach, is poor delivery of the oligonucleotides into bacterial cells. The bacterial cell wall prevents the efficient uptake of short oligonucleotides from the environment[14] and non-invasive delivery of PNAs is extremely difficult. To overcome this problem, chemical conjugation of PNA to a variety of cell-penetrating peptides has been tested[7, 15, 16]. The most commonly used conjugate of PNA with the peptide (KFF)3K was delivered with high efficiency in vitro. However, the activity of (KFF)3K-PNA conjugates was dramatically decreased in the presence of serum in eukaryotic cells[17]. Furthermore, as a cationic peptide (KFF)3K possesses hemolytic activity at the very low concentration of 40 µg/ml (for comparison, the antibiotic polymyxin B is not hemolytic at 1500 µg/ml)[18]. In addition, (KFF)3K initiates histamine release in some mammalian cells, which leads to the development of an inflammatory response and causes pruritis[6]. For these reasons (KFF)3K is not a good delivery agent for neutral oligonucleotides and its future medical application is doubtful. A variety of other membrane-penetrating peptides have been tested in vitro, including TAT[15] and (RXR)4XB (X – 6-aminohexanoic acid, B – β-alanine)[2], oncocin[19], and others[20-22] but with variable results. All of these peptides are cationic and amphipathic[23], and their efficiency as PNA transporters is still far from perfect[24]. Moreover, we recently observed that conjugation with (KFF)3K decreased the ability of PNA to efficiently hybridize with an RNA hairpin even though the peptide is positively charged and thus attracted by the negatively charged RNA[13]. Also, conjugation of (KFF)3K with PNA (targeted at a functional site in 23S rRNA) decreased the level of inhibition of protein production in an E. coli cell-free transcription/translation system as compared to PNA alone. Therefore, there is still a pressing need for an effective carrier system to deliver PNAs to bacterial cells.

Vitamin B12 (cobalamin) is a natural organometallic molecule (Fig. 1b)[25]. It is an essential nutrient cofactor in mammalian metabolism[26]. Vitamin B12 cannot be synthesized within the human body so must be included in the diet[27], which makes this molecule an attractive and easy-to-administer candidate as a drug carrier. In recent studies, vitamin B12 has been used as a delivery vehicle in mammalian cells[25] and applied to increase the bioavailability of different therapeutics including proteins[28] and anti-cancer drugs[29]. Most aerobic bacteria require vitamin B12 for growth, but only a few species are able to produce it[30]. A variety of microorganisms are capable of its uptake[30], especially members of the Enterobacteriaceae family, Bacillus subtilis and Group A streptococci[31, 32]. Escherichia coli (E. coli) and Salmonella enterica serovar Typhimurium (S. Typhimurium) cells actively transport vitamin B12 using a cascade of membrane proteins[33]. The import of vitamin B12 into these cells involves a rapid energy-independent phase in which the molecule associates with the receptor BtuB in the outer membrane. This initial stage is followed by a slower energy-dependent process involving other membrane proteins. Furthermore, Salmonella species possess a second independent system for the passage of vitamin B12 across the outer membrane[34]. The presence of fairly well characterized vitamin B12 uptake mechanisms in bacteria makes it an attractive candidate for a carrier. To exploit the natural uptake properties of vitamin B12, the molecule has to be modified and conjugated to PNA oligomers. It was shown that conjugation at certain positions dramatically alters the binding properties of vitamin B12 in mammalian cells[35]. To successfully use the vitamin B12 pathway in the transport of PNA, the conjugate must still be recognized by the B12 uptake mechanism, and the PNA has to interact with its target and cause the desired effect.

We have investigated the ability of the non-peptidic carrier vitamin B12 to transport PNA oligomers into E. coli and S. typhimurium cells. Vitamin B12 was conjugated to a PNA oligomer targeted at the mRNA of the reporter gene mrfp1 [36], expressing Red Fluorescent Protein (RFP), in E. coli and S. Typhimurium cells. Decreases in RFP fluorescence were monitored to detect any PNA activity within cells treated with these compounds. Since the uptake of conjugates may depend on the linker, we tested different linker types and spacer lengths, including a cleavable disulfide bond linker.

Results and Discussion

System to monitor inhibition of mrfp1 mRNA translation

To provide a convenient system for comparative studies of the effect of antisense PNA in E. coli and S. Typhimurium, we constructed pBBR(rfp), an RFP reporter vector optimized for expression in Enterobacteriaceae (Fig. 2a). Gene mrfp1 was chosen as the reporter because any antisense effect could be readily assessed by examining red fluorescence of the cells (Figure S1). The anti-mrfp1 PNA was designed to target the region of the mRNA overlapping the translation start codon, which was shown to be sensitive to antisense inhibition[37], plus part of the ribosome binding site (RBS) B0034 (http://parts.igem.org/Part:BBa_B0034) (Fig. 2b). This RBS is recognized as strong and highly efficient (http://parts.igem.org/Part:BBa_K1017202). To ensure that the PNA sequence was specific for the mrfp1 mRNA, we examined off-target gene complementarity using online sequence analysis tools - GenoList[38] and RiboScanner[10]. We also verified the physicochemical properties of the PNA oligomer, which could affect its solubility (PNA Tool http://pnabio.com/).

Figure 2

(a) Schematic representation and construction strategy of mrfp1 expression plasmid pBBR(rfp). The red rectangle surrounds the mRNA target site for the PNA sequence. KmR refers to the kanamycin resistance marker; gray arrows represent the lacZ gene encoding β-galactosidase, disrupted by the insert (RBS and mrfp1). The plasmid contains two other structural elements: rep and mob which were not included in the figure. (b) Anti-mrfp1 PNA sequence. c) Control scrambled PNA. The carrier is either (KFF)3K or vitamin B12.

Synthesis of vitamin B12- PNA conjugates

We designed and synthesized a series of conjugates to be delivered into bacterial cells, by attaching the PNA oligomer to vitamin B12, incorporating both cleavable and non-cleavable linkers between the molecules. 1,3-dipolar cycloaddition was used to prepare four such conjugates (Fig. 3a-b). In one conjugate, alkyne-PNA was directly coupled to vitamin B12 possessing an azide moiety at 5′ position – B12-N3 (Fig. 3a)[39], while in the other three, a spacer (alkyl- or PEG-type) was incorporated into the vitamin B12 structure via the carbamate bond[40, 41], and this was then coupled to the alkyne-PNA (Fig. 3b). To introduce a cleavable linker, Cys-PNA was reacted with the B12-SS-Py derivative to produce a conjugate containing a disulfide bridge (Fig. 3c)[42]. As controls, we also synthesized PNA conjugates (anti-mrfp1 and scrambled sequences) with the (KFF)3K peptide. The azide derivative of this peptide ((KFF)3K-N3) was attached to the alkyne-PEG5-PNA via 1,3-dipolar cycloaddition (Fig. 3d)[43]. See Methods for detailed synthesis procedures.

Figure 3

Synthesis of vitamin B12 and (KFF)3K conjugates with PNA. The spacer is marked in blue.

In order to achieve the desired inhibitory effect in cells, the conjugates need to be stable. Thus we examined the stability of the vitamin B12-PNA conjugates in vitro in the bacterial Davis Minimal Broth[44] medium and fetal bovine serum. The resulting HPLC chromatograms did not show any appreciable differences before and after incubation so all conjugates were considered stable in the presence of biological media.

Verification of the system

To evaluate the potential of antisense PNA to inhibit translation of the mrfp1 mRNA transcript, we cultured bacteria in the presence of anti-mrfp1 PNA conjugated to the (KFF)3K peptide ((KFF)3K-PNA, Fig. 3d). This cell-penetrating peptide is frequently employed to transport PNA into bacterial cells[8], so it was used as a control. As shown in Fig. 4, we observed a significant decrease in red fluorescence in bacteria grown in the presence of (KFF)3K-PNA (relative to untreated cells). At all tested concentrations of this conjugate, the level of RFP was reduced by about 70% in E. coli. However, in S. Typhimurium the production of RFP decreased in a dose-dependent manner following (KFF)3K-PNA treatment. In addition, the antisense effect of (KFF)3K-PNA on RFP production in S. Typhimurium reached almost 100% at concentrations of ≥8 µM. This variation between two members of the Enterobacteriaceae probably results from phenotypic and genotypic differences[45, 46], such as structural diversity in the core regions of the lipopolysaccharides[47] which may affect PNA transport. To confirm that the decrease in RFP expression was caused by the anti-mrfp1 PNA, we tested a scrambled PNA sequence (Fig. 2c). RFP production was unchanged following treatment with this altered PNA (Fig. 4). Furthermore, there was no inhibition of mrfp1 expression when either free (KFF)3K or the PNA sequence without the carrier peptide were applied (Fig. 4). Therefore, the inhibitory effect that we observed was dependent on the PNA sequence delivered in a conjugate.

Figure 4

Inhibition of RFP synthesis after treating cells with (KFF)3K-PNA, (KFF)3K-PNA (scrambled), PNA anti-mrfp1 or free (KFF)3K. For the sequence of the PNA targeted at the mRNA transcript encoding RFP, see Fig. 2. The mean values from three independent experiments are plotted, with error bars indicating the standard error. The fluorescence intensity given in relative fluorescence units [RFU] correlates well with the cellular RFP level (see Methods). Data for all tested concentrations are shown in Table S1. The differences between RFU obtained for (KFF)3K-PNA and each of the other compounds are highly significant with P ≤ 0.001.

Interestingly, E. coli cells treated with PNA only exhibit higher fluorescence intensity than cells treated with (KFF)3K-PNA(scrambled) or (KFF)3K (Fig. 4). We hypothesize that this effect may be due to electrostatically-driven interactions of (KFF)3K with RFP that could influence the detected level of fluorescence. The (KFF)3K peptide is positively charged (with a total net charge of +4e) and the RFP protein is negatively charged (total net charge of −5e and pI of 6.1[48]). On the other hand PNA is neutral so is not attracted by the RFP protein.

To rule out unspecific toxicity of the compounds we monitored their effect on bacterial growth. The measurements of OD600 after overnight incubation of bacterial cells with the above compounds (at concentrations required to deliver PNA to cells) did not indicate bacterial growth inhibition and thus antibacterial activity. The minimal inhibitory concentration for free (KFF)3K is >20 μM in both E. coli and S. Typhimurium[10, 13]. Surprisingly, after treatment with PNA only, we observed a slightly higher level of OD600 than for untreated cells but this observation corroborates with the observed RFU (Fig. 4).

Inhibition of RFP synthesis by vitamin B12 conjugates with anti-mrfp1 PNA

We next investigated the ability of five different vitamin B12 derivatives conjugated to the anti-mrfp1 PNA to transport this PNA into Gram-negative bacterial cells and inhibit RFP synthesis. The tested compounds shown in Fig. 4 vary in the way in which vitamin B12 was connected to the PNA. In four of the compounds, vitamin B12 and the PNA were conjugated via the triazole ring, which is stable in the biological environment and, most importantly, in the broth used for culturing bacteria (Fig. 5a–d)[49]. These compounds differed in the length of the spacer. In conjugate B12-SS-PNA (Fig. 5e), vitamin B12 was connected to PNA via a disulfide bond that can be reduced by glutathione (GSH), an antioxidant molecule widely distributed in bacteria. The level of antisense inhibition of RFP translation by these conjugates was determined by measuring changes in the red fluorescence of E. coli and S. Typhimurium cells following overnight treatment (see Methods).

Figure 5

Structures and naming of vitamin B12-PNA conjugates. The carrier is at the PNA N-terminus and PNA C-terminus contains Lys.

Antisense effects were detected with all of the vitamin B12-PNA conjugates tested on E. coli and S. Typhimurium (Fig. 6). Moreover, we observed that the structure and length of the linker and spacer affected the delivery of PNA. The conjugates that most effectively inhibited RFP synthesis had either no spacer (B12-PNA) or the longest spacer (B12-(CH2)12-PNA), especially in E. coli cells. The two other compounds with the triazole ring as a linker, i.e. B12-(CH2)6-PNA and B12-PEG2-PNA, caused a lesser antisense effect. Vitamin B12 with the degradable linker B12-SS-PNA inhibited mrfp1 gene expression with the lowest efficiency. This was anticipated because disulfide bonds are generally not stable in the cytosolic compartments of bacteria or eukaryotic cells, due to their chemically reducing nature[50].

Figure 6

Relative fluorescence measured in cells after treatment with conjugates of vitamin B12-PNA targeted at the mRNA encoding RFP. Data for all tested concentrations are shown in Table S2. The differences between RFU of B12-(CH2)12-PNA(scrambled) and the rest of tested conjugates are highly significant with P ≤ 0.001. In addition, for E. coli: the difference between B12-(CH2)12-PNA and B12-(CH2)6-PNA is significant with P ≤ 0.01 and the difference between B12-(CH2)12-PNA and B12-PEG2-PNA is marginally significant (P ≤ 0.05). The differences between other tested conjugates are not significant (P > 0.05).

Similar results were obtained for both E. coli and S. Typhimurium, although in the latter bacterium, the inhibition of RFP production by all compounds was less effective (~75% at most). In addition, the nature of the linkers and spacers present caused no significant difference in the inhibitory activity of the vitamin B12-PNA conjugates in S. Typhimurium. This different effect in these two bacteria might be due to differences in the structure of the cell wall as well as vitamin B12 requirements for growth and the membrane transport systems involved in uptake[34, 51].

As a control for these experiments, we used the scrambled PNA (Fig. 2c) attached to vitamin B12 via –(CH2)12 (Fig. 5d), i.e. the linker that inhibited RFP production most effectively when conjugated to the complementary PNA. Free vitamin B12 was used as a second control. Neither B12–(CH2)12-PNA(scrambled) (Fig. 6) nor free vitamin B12 (data not shown) had any inhibitory effect on mrfp1 expression.

In addition, the vitamin B12-PNA conjugates did not affect bacterial growth. There was no decrease in OD600 upon overnight treatment with the conjugates (up to the tested 16 μM concentrations). Free vitamin B12 has no antibacterial activity at concentrations up to 100 μg/ml[52]. However, it has some growth-promoting properties[53], which resulted in a slight increase of OD600 while increasing the concentration of vitamin B12.

In E. coli for all vitamin B12-PNA conjugates, we observed a concentration-dependent decrease in red fluorescence (Fig. 6). In contrast, with the (KFF)3K-PNA conjugate, we saw an initial drop in fluorescence at 0.125 µM, followed by a steady level of inhibition at higher concentrations (Fig. 4 and Table S1). A similar inhibitory effect was observed when all conjugates were applied at 2 µM. In E. coli, at concentrations between 4 and 16 µM, the decrease in fluorescence produced by the most effective vitamin B12-(CH2)12-PNA constructs was greater than for (KFF)3K-PNA (Figure S2).

The conjugate of PNA with the (KFF)3K peptide produced a different effect in S. Typhimurium (Fig. 4). In addition, the (KFF)3K-PNA and vitamin B12-PNA conjugates showed comparable dose-dependent inhibition of RFP fluorescence in this bacterium (Figs 4 and 6). A similar efficiency of RFP inhibition was produced by (KFF)3K-PNA and B12-(CH2)12-PNA at 1 µM (Figure S2). Overall, the delivery efficiency of PNA to S. Typhimurium cells was better for PNA conjugated with (KFF)3K than vitamin B12, when applied at concentrations of between 2 and 16 µM.

Overall, in E. coli for the conjugates with the longest B12-(CH2)12-PNA and shortest B12-PNA linker we observed that PNA transport efficiency by vitamin B12 is slightly better than by the (KFF)3K peptide but for S. Typhimurium the effect was the opposite. One reason could be that the uptake occurs only up to certain concentrations of the conjugate that are too low for the PNA to achieve its antisense effect in cells (and these maximal concentration of vitamin B12 uptake is lower in S. Typhimurium than E. coli). Also, we have recently observed that conjugation of PNA with the (KFF)3K peptide hinders PNA hybridization efficiency with a complementary RNA strand[13]. To determine if the attachment of vitamin B12 affects hybridization of a PNA oligomer with complementary RNA, we performed polyacrylamide gel electrophoresis (PAGE) experiments. We used the best working conjugates and an RNA oligomer with the 5′AGGAGAAAUACUAGAUGGCU3′ sequence corresponding to the targeted mRNA transcript (with the fragment complementary to PNA underlined). With the secondary structure prediction programs (MFold[54], RNAfold[55], and Sfold[56]) we verified that this RNA sequence most probably does not acquire any secondary structure and does not self-interact. The mRNA fragment was incubated in water solution containing either PNA or the B12-PNA and B12-(CH2)12-PNA conjugates (in a 1:1 ratio) and assayed by PAGE in non-denaturing conditions. We found that after adding PNA to RNA (Figure S3, lane 2) the band from free RNA disappeared confirming that free PNA binds to RNA. However, after adding PNA conjugated to vitamin B12 to RNA (Figure S3, lanes 3 and 4), the band from the unbound RNA was still present. This means that although the band from the complex is visible, the attachment of vitamin B12 to PNA somehow interferes with the formation of the complex. Thus similar as we previously observed for the (KFF)3K carrier[13], vitamin B12 may also decrease the ability of PNA to bind complementary RNA.

In summary, we have demonstrated that vitamin B12 can deliver PNA oligomers into E. coli and S. Typhimurium cells. However, attaching of vitamin B12 to PNA may influence PNA hybridization with a complementary fragment of the targeted mRNA transcript encoding RFP. Even though the attachment of vitamin B12 does not favor hybridization of PNA with mRNA the antisense effect of PNA is still visible. The vitamin B12-PNA conjugates were stable in bacterial media and serum and did not inhibit bacterial growth. In the future we plan to investigate whether vitamin B12 can also be used to transport PNA into the cells of Gram-positive bacteria.

Methods

Reagents and conditions

Commercial reagents and solvents were used as received from the supplier. Fmoc-XAL PEG PS resin for PNA synthesis was obtained from Merck and Fmoc/Bhoc-protected PNA monomers from Panagene. Nα-Fmoc protected L-amino acids were obtained from Novabiochem (Fmoc-Lys(Boc)-OH) and Sigma-Aldrich (Fmoc-Phe-OH, Fmoc-Cys(Trt)-OH, Fmoc-β-azido-Ala-OH). Rink-amide resin (TentaGel S RAM resin) for peptide synthesis was obtained from Sigma-Aldrich. All reactions were monitored using Reverse Phase-HPLC (RP-HPLC) techniques. Preparative chromatography was performed using C18 reversed-phase silica gel 90 Å (Sigma-Aldrich) with redistilled water and HPLC grade MeCN as eluents. The following conditions were used for HPLC: column – Eurospher II 100–5 C18, 250 mm × 4.6 mm with a precolumn, or Kromasil C18, 5 µm, 250 mm × 4.0 mm; pressure – 10 MPa; flow rate – 1 mL/min; room temperature; detection – UV/vis at wavelengths (λ) of 361 and 267 nm. HPLC methods, molecular masses and yields are given in Table 1. Details of the 1H and 13C NMR spectra are presented in Section S1.

Table 1

Retention times (tR) and molecular masses of the synthesized conjugates.

Nr	Conjugate	HPLC tR [min]a	HPLC method	Molecular mass [g/mol]		Yieldsc [%]
				Calculated	Foundb
1.	B12-PNA	15.2	0–50%/30 min	5339.4	5340.5	81.1
2.	B12-(CH2)6-PNA	17.1	10–45%/30 min	5482.6	5482.7	79.6
4.	B12-PEG2-PNA	15.5	0–50%/30 min	5514.6	5516.0	67.6
5.	B12-(CH2)12-PNA	20.0	0–50%/30 min	5566.8	5566.5	73.5
6.	B12-SS-PNA	14.7	0–50%/30 min	5353.3	5352.0	53.1
7.	B12-(CH2)12-PNA (scrambled)	20.6	0–50%/30 min	5690.2	5567.2	56.8
8.	(KFF3)K-PNA	20.6	0–50%/30 min	5691.2	5690.0	79.6
9.	(KFF)3K-PNA (scrambled)	20.9	0–50%/30 min	5690.1	5690.2	57.0

aThe product was analyzed by analytical RP-HPLC; bdata obtained from Q-TOF Premier cisolated yield on 1 µmol scale.

Preparation of vitamin B12 derivatives at the 5′ position

B12-PEG2-N3 and B12-(CH2)12-N3 were synthesized in the following way. Vitamin B12 (100 mg, 75 μmol) was dissolved in 2.5 mL of dry DMSO at 40 °C in an argon atmosphere. Solid CDT (50 mg, 300 μmol) was added and the solution stirred under argon. When full consumption of the substrate (monitored by RP-HPLC) had occurred (usually after 1.5 h) heating was removed and 100 µL of aminoazide NH2-PEG2-N3 or B12-(CH2)6-N3 was added in one portion, followed by 20 µL of NEt3. The resulting solution was stirred overnight and then it was poured into 50 mL of AcOEt and centrifuged. The pelleted precipitate was then washed with Et2O (2 × 15 mL). After drying in air, the precipitate was dissolved in water and purified by RP column chromatography with a mixture of MeCN and H2O as the eluent. The experimental details and the complete characterization of B12-(CH2)12-N3 and B12-PEG2-N3 derivatives are presented in Section S1, and Figures S4 and S5, respectively. Vitamin B12 derivatives B12-(CH2)12-N3, B12-PEG2-N3 and B12-(CH2)6-N3 were synthesized according to previously described procedures[39, 41, 42]. The obtained spectral data matched that in the literature. The NH2-(CH2)12-N3 linker was synthesized according to the procedure described in[57] and the NH2-PEG2-N3 and NH2-(CH2)6-N3 linkers were synthesized according to[41]. Again, the obtained spectral data matched that in the literature.

Synthesis of PNA oligomers

Cys-PNA was synthesized according to the procedure of Wierzba et al. (2016)[42]. PNA oligomers (alkyne-PNA, alkyne-PEG5-PNA, Cys-PNA, alkyne-PEG5-PNA scrambled) were synthesized manually using Fmoc chemistry at 10 µmol scale with a 2.5-fold molar excess of the Fmoc/Bhoc-protected monomers and 3-fold molar excess of the Fmoc-protected Lys, pentynoic acid, alkyne-PEG5-acid and polyethylene glycol-polystyrene resin (Fmoc-XAL PEG PS resin, amine groups loading of 190 μmol/g; this resin has a linker which yields a C-terminal amide upon TFA cleavage of PNA). In all syntheses Lys was the first monomer attached to resin. Monomers were activated by treatment with a 2-(1H-7-azabenzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HATU), N-methylmorpholine (NMM) and 2,6-lutidine (0.7:1:1.5) mixture using DMF/NMP (1:1, v/v) solution, and coupled for 40 min as active derivatives. A double coupling was performed. Fmoc deprotection was accomplished using 20% piperidine in DMF (2 × 2 min). After synthesis of the PNA backbone and removal of the N-terminal Fmoc, the pentynoic acid or alkyne-PEG5-acid were attached to the N-terminus. Acids were assembled as active derivatives in 3-fold molar excess by the use of HATU with the addition of HOAt and collidine (1:1:2), using the DMF/NMP (1:1, v/v) solution-coupling method for 2 h. Fmoc deprotection of amino acids was accomplished using 20% piperidine in DMF for 2 cycles (5 and 15 min). Removal of the protecting group and cleavage of PNA from the resin was performed by treatment with a TFA/triisopropylsilane/m-cresol (95:2.5:2.5; v/v/v) mixture for 60 min. The obtained crude oligomers were lyophilized and subsequently purified by RP-HPLC.

Synthesis of (KFF)3K-N3

Azido-peptide was synthesized by manual solid-phase peptide synthesis (SPPS) using the standard Fmoc/t-Bu chemistry on a 100 µmol scale with a 3-fold molar excess of the Fmoc-protected amino acids and Rink-amide resin (TentaGel S RAM resin, amine groups loading of 240 μmol/g; this resin has a linker which yields a C-terminal amide upon TFA cleavage of the peptide). Fmoc-protected amino acids were assembled as active derivatives in a 3-fold molar excess by the use of HATU with the addition of 1-hydroxy-7-azabenzotriazole (HOAt) and collidine (1:1:2), using the DMF/NMP (1:1, v/v) solution-coupling method for 2 h. Fmoc deprotection was accomplished using 20% piperidine in DMF for 2 cycles (5 and 15 min). Removal of the protecting group (Boc from Lys) and cleavage of peptide from the resin was performed by treatment with a TFA/triisopropylsilane/m-cresol (95:2.5:2.5; v/v/v) mixture for 60 min. The obtained crude peptide was lyophilized and subsequently purified by RP-HPLC.

Synthesis of PNA conjugates with carriers: vitamin B12 and (KFF)3K

The B12-SS-PNA conjugate was prepared by the method reported in ref. 42. Other conjugates were synthesized using copper-catalyzed azide-alkyne cycloaddition according to the procedures described in refs 39, 43. CuI (1.0 mg, 5 μmol) and TBTA (5.0 mg, 10 μmol) were dissolved in DMF/H2O (0.5 mL, 1:1 v/v) and stirred for 20 min. The respective azide-B12 or azide-peptide (3 µmol) and the respective alkyne-PNA (1 µmol) were then added and the reaction mixtures were stirred overnight. The mixtures were centrifuged to remove the catalyst and the solutions, containing the crude products, were then purified by RP-HPLC. Table 1 gives the experimental details and the physicochemical properties of the final compounds. According to HPLC analyses all reactions proceeded with conversion > 99%. HPLC and MS analysis of the B12-PNA and (KFF)3K-PNA products gave m/z values in accordance with their calculated molecular masses. Yields of the final isolated products were in the range 53–81% on 1 µmol synthesis scale. The purity, determined by RP-HPLC (267 nm), was ≥ 98% for all the conjugates. Mass spectra and RP-HPLC chromatograms are shown in Section S2, Figures S6–S13. To monitor the stability, the vitamin B12-PNA conjugates were added at 50 µM concentrations to either Davis Minimal Broth[44] medium or fetal bovine serum. After overnight incubation at 37 °C with shaking the RP-HPLC analyses were performed under the same conditions as described above.

Bacterial strains and growth conditions

E. coli TG1[58] was used for plasmid construction. For triparental mating, E. coli DH5α carried the helper plasmid pRK2013[59] and E. coli S17-1[60] was the mrfp1 plasmid donor. E. coli K-12 MG1655[61], and Salmonella enterica subsp. enterica serovar Typhimurium LT2-R (rifampicin resistant mutant of wild-type S. Typhimurium LT2[62]) were used in PNA delivery experiments. To prepare inocula, all strains were grown overnight in lysogeny broth (LB) at 37 °C with shaking. To monitor inhibition of mrfp1 gene expression the analyzed strains were grown in Davis Minimal Broth at 37 °C with shaking. Cultures were supplemented with appropriate antibiotics to prevent plasmid loss.

Construction of the RFP vector

Plasmid pSB3K3-RFP contains the mrfp1 gene with rbs_B0034, an efficient ribosomal binding site (http://parts.igem.org/Part:BBa_K1017202). Plasmid pBBR1MCS-2 is a cloning vector optimized for Enterobacteriaceae [63] and functional in the majority of Gram-negative bacteria. Plasmid pBBR(rfp), which constitutively expresses RFP, was constructed by ligating an EcoRI-SpeI restriction fragment containing the mrfp1 gene and rbs_B0034 from pSB3K3-RFP to pBBR1MCS-2 digested by the same endonucleases (Figure S14). Appropriate kits were used for the isolation of plasmid DNA from bacterial cells and its purification after enzymatic reactions (EURx, Gdańsk, Poland). Restriction enzymes and ligase were purchased from Thermo Scientific™. Common DNA manipulation methods were performed as described by Sambrook and Russell[58].

Introduction of plasmid pBBR(rfp) into bacterial cells

E. coli transformation

Chemical competent E. coli TG1 cells were prepared using the E. coli Transformer kit (A&A Biotechnology) following the supplied protocol. A 100 µL aliquot of frozen competent cells (stored at −80 °C) was thawed on ice, then 20 µL of ligation reaction were added and the cells gently mixed. Following incubation on ice for 45 min, the transformation mixture was placed in a thermoblock at 37 °C for 10 min. The mixture was then cooled on ice for 2 min and 100 μl were spread on an LB agar plate containing 50 µg/ml kanamycin. Plates were incubated overnight at 37 °C to permit growth of transformants. The expression of RFP was verified by the detection of red fluorecence when colonies were examined on a UV transilluminator.

Triparental mating with S. Typhimurium

Triparental mating[64] was used to mobilize plasmid pBBR(rfp) into the rifampicin resistant mutant of S. Typhimurium LT2 (LT2-R). Overnight cultures of E. coli S17-1 carrying mobilizable vector pBBR(rfp) (donor), E. coli DH5α carrying plasmid pRK2013 (helper) and S. Typhimurium LT2-R (recipient) were centrifuged. The pelleted cells were washed once in LB to remove antibiotics and then resuspended in LB. The donor, helper and recipient strains were then mixed in a volume ratio of 1:1:2, respectively. 100 µL of this cell mixture were spread on a plate of LB agar and incubated overnight at 37 °C. The bacteria were then washed off the plate and serially diluted 1:10, 1:100, and 1:1000 in LB. 100 µl samples of the undiluted cell suspension and each of the dilutions were plated on LB agar containing rifampicin (50 µg/ml) (selectable marker for the recipient strain) and kanamycin (50 µg/ml) and incubated overnight at 37 °C. Transconjugant colonies were identified by the detection of red fluorescence as described above (Figure S15). Restriction digest analysis of isolated plasmid DNA verified the presence of pBBR(rfp) in 5–10 fluorescent clones (Figure S14). To confirm that transconjugants were S. Typhimurium, 16 S rDNA Restriction Fragment Length Polymorphism (RFLP) analysis was performed.

Amplification of 16S rDNA genes and RFLP analysis

For DNA preparation, small loopfuls of S. Typhimurium or E. coli cells were suspended in lysis buffer (0.25% SDS, 50 mM NaOH) and heated at 99 °C for 10 min. The lysate was then diluted with distilled water and this was used as the template DNA in PCRs. Universal bacterial 16 S rRNA primers were included in the reactions: forward primer 27 F (5′- AGAGTTTGATCMTGGCTCAG -3′)[65] corresponding to positions 8 to 27 in E. coli rRNA and reverse primer 1492R[66] corresponding to positions from 1491 to 1509 (5′- GGTTACCTTGTTACGACTT-3′). The PCR reactions (25 µL) had the following composition: 1 µL of template DNA, 1 µL of each 16 S rRNA primer (0.5 µM), 12.5 µL Thermo Scientific™ DreamTaq™ (LifeTechnologies) DNA Polymerase mixture, 9.5 µL ddH20. The following thermocycle was employed in the PCR: initial denaturation at 95 °C for 180 s, followed by 20 cycles of 95 °C for 50 s, 53 °C for 50 s, and 72 °C for 90 s; then 15 cycles of 95 °C for 30 s, 46 °C for 30 s and 72 °C for 90 s, finishing with a final extension step at 72 °C for 10 min.

16s rDNA Restriction Fragment Length Polymorphism

Amplified 16 S rDNA PCR products were purified and then digested separately with the restriction enzyme SalI, which was predicted to produce different restriction fragment patterns for the two species (Figure 16 caption). A 16 S rDNA fragment amplified from E. coli was cleaved with the same enzyme as a control. The digested DNA fragments were analyzed by agarose gel (0.8%) electrophoresis (60 min at 100 V) and the gel was stained with ethidium bromide for 20 min before viewing on a UV transilluminator (Figure S16).

Determination of the level of red fluorescence

The effect of the different PNA conjugates and controls on the red fluorescence of bacterial cells was determined using a standard microdilution method[67] in concentration range 0–16 µM. Cultures grown to exponential phase in Davis Minimal Broth were diluted to ~5 × 105 CFU/ml. These cell suspensions were added to wells of sterile 96-well plates containing different concentrations of the tested compounds. Following overnight incubation at 37 °C with shaking, the plates were vigorously shaken (double orbital) for 10 sec and then the cell density (OD600) and fluorescence (λ excitation – 584 nm; λ emission – 610 nm) were measured using a plate reader (Microplate Reader Biotek Synergy H1MFDG). The cell suspensions were also examined by light and fluorescence microscopy using a Nikon Eclipse Ni-U microscope (Figure S1).

Fluorescence Data analysis

To obtain relative fluorescence values (RFU), the background media noise was subtracted from each data point for both cell density (OD600) and fluorescence. Then the background-adjusted fluorescence data were divided by the background-adjusted OD600 and the data were normalized to the untreated cells. This value is proportional to the amount of RFP per cell, which correlates with the PNA activity. Free vitamin B12 was not taken into account because it shows negligible fluorescence emission. To evaluate statistical significance the two-way ANOVA was used. A probability value of P ≤ 0.05 was considered indicative of a statistical significance.

PAGE analysis

Polyacrylamide gel electrophoresis (PAGE) was performed using the Mini-PROTEAN Tetra Cell system (Bio-Rad, Poland). Sample mixtures (200 pmol each) were transferred onto the gel using a solution of 40% glycerol in water (loading buffer). For 5 µL samples, 1 µL of loading buffer was added, then samples were electrophoresed under 60 V for 3 h on a 15% non-denaturing polyacrylamide gel, using 1x TBE as a running buffer (89 mM Tris-base, 89 mM borate, 2 mM EDTA, pH 8.3). Gels were stained by Stains-All (Sigma-Aldrich) and imaged using a Gel Doc XR + System (Bio-Rad).

Supporting Information