*2.3. Evaluation of the Internalization of B12-ASOs*

To examine the internalization of both conjugates in *E. coli* K12 and assess if association of the ASOs to the B<sup>12</sup> could have hampered B12-promoted uptake, bacteria were observed under an epifluorescence microscope, after incubation with each of the Cy3-labeled conjugates or controls (B12, ASOgapmer, and ASOsteric).

As expected, almost no fluorescent bacteria were detected when ASOgapmer and ASOsteric were used alone (Figure 4–ASOgapmer and ASOsteric, Cy3 line). In contrast, it is clear that the conjugation of B<sup>12</sup> to either ASO significantly increased the amount of fluorescently labeled *E. coli* K12, with all cells becoming fluorescent (Figure 4, B12-ASOgapmer and B12-ASOsteric). The same was observed for the B<sup>12</sup> control (Figure 4, B12). Figure 4 shows images obtained at 30 µM, but a similar pattern was obtained for the lower concentration tested (15 µM, Figure S2). the conjugation of B12 to either ASO significantly increased the amount of fluorescently labeled *E. coli* K12, with all cells becoming fluorescent (Figure 4, B12-ASOgapmer and B12- ASOsteric). The same was observed for the B12 control (Figure 4, B12). Figure 4 shows images obtained at 30 μM, but a similar pattern was obtained for the lower concentration tested (15 μM, Figure S2).

To examine the internalization of both conjugates in *E. coli* K12 and assess if association of the ASOs to the B12 could have hampered B12-promoted uptake, bacteria were observed under an epifluorescence microscope, after incubation with each of the Cy3-la-

As expected, almost no fluorescent bacteria were detected when ASOgapmer and ASOsteric were used alone (Figure 4–ASOgapmer and ASOsteric, Cy3 line). In contrast, it is clear that

*Antibiotics* **2021**, *10*, x FOR PEER REVIEW 5 of 12

beled conjugates or controls (B12, ASOgapmer, and ASOsteric).

**Figure 4.** Interaction of Cy3-labeled ASOs, B12, and B12 conjugates (concentration of 30 μM) with *E. coli* K12 after 4 h. Bacteria are counterstained with 4′,6′-diamidino-2-phenylindole (DAPI). Images are representative of three independent experiments (using duplicates in each). Scale bar represents 5 μm. **Figure 4.** Interaction of Cy3-labeled ASOs, B12, and B<sup>12</sup> conjugates (concentration of 30 µM) with *E. coli* K12 after 4 h. Bacteria are counterstained with 40 ,60 -diamidino-2-phenylindole (DAPI). Images are representative of three independent experiments (using duplicates in each). Scale bar represents 5 µm.

These results point toward the B12-promoted association of the conjugates with the bacterial cells. However, the experimental distinction between membrane-associated and internalized molecules in bacteria remains a difficult task, given the small size of bacteria, which challenges the resolution limit of most standard equipment, including fluorescence microscopes [36]. These results point toward the B12-promoted association of the conjugates with the bacterial cells. However, the experimental distinction between membrane-associated and internalized molecules in bacteria remains a difficult task, given the small size of bacteria, which challenges the resolution limit of most standard equipment, including fluorescence microscopes [36].

Therefore, in an attempt to understand if the conjugates were internalized or membrane adhered, as well as to quantify their relative distribution, the bacterial cells were fractionated. A series of washing steps with a gradient of Triton X-100 concentrations was used to differentiate the membrane fraction from the cytosol [37]. These fractions were quantified using a fluorometer. Figure 5 clearly shows that only a small fraction of B12 and B12 conjugates completely penetrate the bacterial envelope into the cytosol (only 12%, 9%, and 6%, respectively, for the unconjugated B12, B12-ASOgapmer, and B12-ASOsteric), while more than 80% remain adhered to the membrane in all cases. The presence on the periplasm is not relevant (only ~3% of the conjugates were retained in this matrix, which was not statistically different from the cytosol. *p* > 0.05, Figure S3 [38]), which indicates that the BtuB at the outer membrane (OM) is likely the limiting factor for conjugate internalization into Therefore, in an attempt to understand if the conjugates were internalized or membrane adhered, as well as to quantify their relative distribution, the bacterial cells were fractionated. A series of washing steps with a gradient of Triton X-100 concentrations was used to differentiate the membrane fraction from the cytosol [37]. These fractions were quantified using a fluorometer. Figure 5 clearly shows that only a small fraction of B<sup>12</sup> and B<sup>12</sup> conjugates completely penetrate the bacterial envelope into the cytosol (only 12%, 9%, and 6%, respectively, for the unconjugated B12, B12-ASOgapmer, and B12-ASOsteric), while more than 80% remain adhered to the membrane in all cases. The presence on the periplasm is not relevant (only ~3% of the conjugates were retained in this matrix, which was not statistically different from the cytosol. *p* > 0.05, Figure S3 [38]), which indicates that the BtuB at the outer membrane (OM) is likely the limiting factor for conjugate internalization into the cytosol. On the contrary, 40 ,60 -diamidino-2-phenylindole (DAPI), a small and cell-permeant DNA intercalating dye, was majorly localized at the cytosol (Figure 5), as expected. Nonetheless, a small fraction was also present in the membrane (Figure 5), which can occur especially in non-fixed cells [39].

the cytosol. On the contrary, 4′,6′-diamidino-2-phenylindole (DAPI), a small and cell-permeant DNA intercalating dye, was majorly localized at the cytosol (Figure 5), as expected. Nonetheless, a small fraction was also present in the membrane (Figure 5), which can oc-

cur especially in non-fixed cells [39].

**Figure 5.** Cellular localization of B12 conjugates, B12, and DAPI control in *E. coli* K12. B12 conjugates and B12 are mainly found on the OM, while the DAPI control is mostly associated with the cytosol. No significant differences were observed between the different internalized conjugates and between the conjugates and the B12 control (*p* > 0.05). Significant differences were observed between the membrane and cytosol-associated compounds (*p* ≤ 0.0001, \*\*\*\*). The fluorescence of each fraction present in the DAPI control is significantly different from the tested counterparts (*p* ≤ 0.0001). Results are presented as mean values and respective standard deviation from three independent assays (using duplicates in each). **Figure 5.** Cellular localization of B<sup>12</sup> conjugates, B12, and DAPI control in *E. coli* K12. B<sup>12</sup> conjugates and B<sup>12</sup> are mainly found on the OM, while the DAPI control is mostly associated with the cytosol. No significant differences were observed between the different internalized conjugates and between the conjugates and the B<sup>12</sup> control (*p* > 0.05). Significant differences were observed between the membrane and cytosol-associated compounds (*p* ≤ 0.0001, \*\*\*\*). The fluorescence of each fraction present in the DAPI control is significantly different from the tested counterparts (*p* ≤ 0.0001). Results are presented as mean values and respective standard deviation from three independent assays (using duplicates in each).

From the obtained fractionation results, it can be concluded that the microscopy fluorescence observed in Figure 4 is predominantly derived from conjugates associated with the OM of *E. coli* K12, rather than conjugates internalized in the cytosol, where the ASO would hybridize the *acpP* mRNA target. The inability of B12 to serve, in this study, as an efficient trojan-horse for the internalization of ASOs explains the lack of antimicrobial activity of the conjugates observed in Figure 3. It is possible that the uptake of B12 is strongly limited by the activity of BtuB, which is present in the OM. From the obtained fractionation results, it can be concluded that the microscopy fluorescence observed in Figure 4 is predominantly derived from conjugates associated with the OM of *E. coli* K12, rather than conjugates internalized in the cytosol, where the ASO would hybridize the *acpP* mRNA target. The inability of B<sup>12</sup> to serve, in this study, as an efficient trojan-horse for the internalization of ASOs explains the lack of antimicrobial activity of the conjugates observed in Figure 3. It is possible that the uptake of B<sup>12</sup> is strongly limited by the activity of BtuB, which is present in the OM.

The uptake of B12 is regulated by the expression/repression of the BtuB, the locus encoding for the B12 receptor [40]. B12 acts as a cofactor for methionine synthesis, necessary for growth [41]. In *E. coli*, it has been estimated that the methionine synthesis requires very low levels of B12 (20 molecules per cell) [42], while there are hundreds of thousands of mRNA copies of the *acpP* gene [43]. In our work, we have used a much higher concentration of B12 than the amount that *E. coli* needs for methionine synthesis. Hence, the difference between the amount of internalized B12 conjugates and the high amount of copies of the essential gene we aimed to inhibit may explain the lack of effectiveness of the conju-The uptake of B<sup>12</sup> is regulated by the expression/repression of the BtuB, the locus encoding for the B<sup>12</sup> receptor [40]. B<sup>12</sup> acts as a cofactor for methionine synthesis, necessary for growth [41]. In *E. coli*, it has been estimated that the methionine synthesis requires very low levels of B<sup>12</sup> (20 molecules per cell) [42], while there are hundreds of thousands of mRNA copies of the *acpP* gene [43]. In our work, we have used a much higher concentration of B<sup>12</sup> than the amount that *E. coli* needs for methionine synthesis. Hence, the difference between the amount of internalized B<sup>12</sup> conjugates and the high amount of copies of the essential gene we aimed to inhibit may explain the lack of effectiveness of the conjugates.

gates. In addition, it is also important to reflect on the future of this strategy, considering that *in vivo*, the number of internalized conjugates will probably be even lower since the host cells, as well as other bacteria from the microbiome, will compete for B12. Moreover, most *in vivo* infections are associated not with single-cell but with clustered cells organized in biofilms [44,45]. Therefore, the bioavailability of B12 conjugates may also be limited by interactions with the extracellular matrix. Nonetheless, genes encoding for virulent In addition, it is also important to reflect on the future of this strategy, considering that in vivo, the number of internalized conjugates will probably be even lower since the host cells, as well as other bacteria from the microbiome, will compete for B12. Moreover, most in vivo infections are associated not with single-cell but with clustered cells organized in biofilms [44,45]. Therefore, the bioavailability of B<sup>12</sup> conjugates may also be limited by interactions with the extracellular matrix. Nonetheless, genes encoding for virulent characteristics, such as the biofilm formation, are usually present in lower amounts of copies. Thus, it would be relevant to study the effect of B12-ASO conjugates targeting these genes in the future. In addition, the in vivo competition for B<sup>12</sup> will favor bacteria with an improved affinity toward B12, as it has been found for some bacteria in the gut possessing an additional lipoprotein (BtuG) [46]. The use of B<sup>12</sup> conjugates to target infections caused by such bacteria possessing BtuG could be considered in future biofilm studies.

#### **3. Materials and Methods**

## *3.1. Materials*

All basic reagents used were purchased from commercial sources (Sigma-Aldrich, Søhus, Denmark) and used as received. Specific reagents and chemicals included LNA phosphoramidite monomers (Innovassynth Technologies, Maharashtra, India), DNA phosphoramidite monomers (Sigma-Aldrich, St Louis, MO, USA), 20OMe phosphoramidite monomers, 30 -PT-amino-modifier C6, BCN *N*-hydroxysuccinimide ester (Glen Research, Sterling, VA, USA), DBCO-sulfo-Cy3 (Jena Bioscience, Jena, Germany) and Vitamin B<sup>12</sup> (Carbosynth, Compton, U.K).

#### *3.2. Synthesis and Design of the ASOs*

ASOs were designed to target the gene *acpP*, an essential gene coding for a protein involved in fatty acid biosynthesis. The particular *acpP* target region for the ASOs was selected based on previous studies [22,27,47]. Two different ASOs were synthesized: (i) an LNA/20OMe chimera, designed for steric blocking (ASOsteric), and (ii) an LNA/DNA chimera (ASOgapmer), designed to recruit RNase (sequences are represented in Figure 1a). Since LNA and 20OMe substitutions increase the duplex stability, the ASOsteric was designed to be shorter than the ASOgapmer. ASOs were synthesized under anhydrous conditions using a PerSpective Biosystems Expedite 8909 nucleic acid synthesizer, as described elsewhere [48]. The synthesis was performed on a 1 µmol scale, using a 30 -PT-aminomodifier C6 support, with the following conditions: trichloroacetic acid in CH2Cl<sup>2</sup> (3:97) as detritylation reagent, 0.25 M 4,5-dicyanoimidazole (DCI) in CH3CN as an activator, acetic anhydride in THF (9:91, *v*/*v*) as cap A solution, *N*-methylimidazole in THF (1:9, *v*/*v*) as cap B solution, and a thiolation solution containing 0.0225 M xanthan hydrate in pyridine/CH3CN (20:90, *v*/*v*). The coupling time was 4.6 min for both monomers. To obtain labeled ASOs, Cy3 phosphoramidite was added to the 50 end in anhydrous CH3CN (0.1 M) and activated by tetrazole with a 15 min coupling time. The stepwise coupling yields were determined by the UV absorbance (at 500 nm) of dimethoxytrityl cations (DMT<sup>+</sup> ) that were released after each coupling. The resulting ASOs were purified by reverse-phase HPLC (RP-HPLC), using a Waters System 600 HPLC equipment, equipped with a Waters XBridge BEH C18-column (5 µm, 100 nm × 19 mm). Their composition and purity (>85%) were confirmed by MALDI-TOF MS and ion-exchange HPLC analysis, respectively. Finally, the purified ASOs were labeled by reaction with BCN *N*-hydroxysuccinimide ester I in carbonate buffer (ASO:BCN = 1:2.5 equivalent) for 2 h. BCN labeled ASOs were desalted using NAP-10 Sephadex columns and purified by RP-HPLC. Their composition and purity (>95%) were confirmed by MALDI-TOF MS and analytical reverse-RP-HPLC, respectively. Concentrations of purified oligonucleotides were determined by UV absorption measurements at 600 nm.
