Application of Safirinium N-Hydroxysuccinimide Esters to Derivatization of Peptides for High-Resolution Mass Spectrometry, Tandem Mass Spectrometry, and Fluorescent Labeling of Bacterial Cells

Mass spectrometry methods are commonly used in the identification of peptides and biomarkers. Due to a relatively low abundance of proteins in biological samples, there is a need for the development of novel derivatization methods that would improve MS detection limits. Hence, novel fluorescent N–hydroxysuccinimide esters of dihydro-[1,2,4]triazolo[4,3-a]pyridin-2-ium carboxylates (Safirinium P dyes) have been synthesized. The obtained compounds, which incorporate quaternary ammonium salt moieties, easily react with aliphatic amine groups of peptides, both in solution and on the solid support; thus, they can be applied for derivatization as ionization enhancers. Safirinium tagging experiments with ubiquitin hydrolysate revealed that the sequence coverage level was high (ca. 80%), and intensities of signals were enhanced up to 8-fold, which proves the applicability of the proposed tags in the bottom–up approach. The obtained results confirmed that the novel compounds enable the detection of trace amounts of peptides, and fixed positive charge within the tags results in high ionization efficiency. Moreover, Safirinium NHS esters have been utilized as imaging agents for fluorescent labeling and the microscopic visualization of living cells such as E. coli Top10 bacterial strain.


Introduction
The aim of proteomics is qualitative and quantitative analysis of proteins in complex samples. This field of study requires high-resolution separation methods and the selective identification which would facilitate peptides identification due to the presence of permanent positive charge, i.e., quaternary nitrogen atom within triazolinium moiety [48]. Herein, we wish to combine both the permanent positive charge within the core structure and distinctive optical properties of the Safirinium dyes to design a compact, readily available ionization and fluorescent tags. The present paper is devoted to the development and application of novel probes suitable for mass spectrometry analyses as well as live-cell imaging.

Chemistry
A panel of Safirinium derivatives bearing diverse alkyl groups within the triazolium moiety (Scheme 1) have been synthesized according to our previously reported procedures [42,43]. Hence, the corresponding isoxazolones 1 and 2 were subjected to tandem Mannich-electrophilic amination reactions with formaldehyde and various secondary amines, such as diethylamine (a), dioctylamine (b), and pyrrolidine (c), to give Safirinium P (pyridine) and Q (quinoline) products 3 and 4 , respectively. Next, the synthesized products obtained in zwitterionic forms were converted into hydrochlorides 3 and 4 with use of methanolic HCl solution, and subsequently reacted with NHS and N,N'-diisopropylcarbodiimide (DIC) as a coupling agent to yield a series of active NHS Safirinium esters 5 and 6. Scheme 1. Synthesis of Safirinium dyes as well as their reactive N-hydroxysuccinimide esters.
The model tripeptide, Ac-AKF-NH 2 , was tagged with Safirinium P esters 5a-c in anhydrous dimethylformamide (DMF) at room temperature to form the corresponding 7a-c derivatives (Scheme 2). The acetylated at the N-terminus peptide, which is also blocked at the C-terminus by conversion of the carboxyl group into an amide, reacts with NHS-activated agents, since it bears a free amine group in the lysine side chain. Since the tags with hydrophobic alkyl substituents increase the effectiveness of peptide ionization [24,28], in addition to the previously reported diethyl derivative 5a [42], we have synthesized compound 5b that incorporates n-octyl substituents on the N2 nitrogen atom of the triazolinium ring. Additionally, novel pyrrolidine derivative 5c was obtained for the reason that bicyclic spiro systems are resistant to Hoffman elimination, which significantly hinders mass spectra analysis for alkyl QAS under MS/MS conditions [49]. Scheme 2. The reaction of active Safirinium P N-hydroxysuccinimide (NHS) esters 5a-c with peptide Ac-AKF-NH 2 .
Tagging of a peptide immobilized on solid support was also completed. Hence, an exemplary tetrapeptide Ac-AAAK-Wang was labeled with reagent 5a and subsequently cleaved from the resin (Scheme 3). Scheme 3. Tagging of Ac-AAAK-Wang with 5a and its cleavage to yield compound 8.

Mass Spectrometry Analysis
Electrospray ionization (ESI)-MS and ESI-MS/MS analyses of compound 5a, AKF peptide conjugates 7a-c, as well as AAAK peptide conjugate 8 have been performed.

Analysis of Reactive NHS Ester 5a
ESI-MS spectrum of tag 5a was measured on ESI-FT-ICR-MS instrument and is shown in Figure 2. We identified its [M] + signal (m/z calc. 347.171, found 347.172); however, the spectrum also revealed an unusual peak distribution from 317 to 320 m/z. For their further identification, we have simulated the isotopic patterns ( Figure 3) of C 15 [50,51], leading to the alkene and H 2 elimination (panel b). The supposed structures are given in Figure 3. The simulated isotopic patterns are in accordance with the observed pattern (panel a).   In ESI-micrOTOF-Q MS/MS experiments with collision-induced dissociation (CID) fragmentation (parent ion 347.2) and collision energies from 5-30 eV ( Figure 4) unusual peak distribution from 317 to 320 m/z is observed, which varies depending on applied collision energies (CE). For the collision energies of 5 eV and 10 eV, the highest daughter signal appears at 319.130, which is characteristic for the product of Hoffman elimination, i.e., elimination of CH 2 =CH 2 from 5a ( Figure 3d). The second abundant fragmentation signal 318.122 corresponds to the radical cation ( Figure 3c). With rising collision energy, these two signals disappear, and in 25 eV, a product of charge remote fragmentation (317.124 as in Figure 3b) is observed. The spectra reveal also signal 289.080, which corresponds to the product of the structure given in Figure 4 (m/z calc. 289.093). The proposed mechanisms of fragmentation reactions, based on the observed patterns, are shown in Figure 5: scheme (a) illustrates the Hoffman elimination of ethylene from 5a, while equation (b) presents the simultaneous elimination of second ethylene molecule and H 2 during charge remote fragmentation. The homolytic cleavage that leads to the radical cation and ethylene radical formation is shown in scheme (c). Finally, equation (d) corresponds to the charge-remote fragmentation of compound 5a, which gives rise to the formation of the most abundant product (m/z 317.124) for high-energy CID. 2.2.2. Analysis of Conjugate 7a, i.e., Peptide Ac-AKF-NH 2 Tagged with Diethyl Derivative 5a Next, we have performed MS and MS/MS analyses of model peptide 7a and its HDX subjected analogue to evaluate deuterated Safirinium system in applications that involve isotopic tagging. Hence, conjugate 7a was incubated in 1% triethylamine (TEA)/D 2 O for 24 h at room temperature. Subsequently, after lyophilization, the sample was dissolved in water to examine the back-exchange process. In Figure 6, the following MS spectra are shown: (a) prior to HDX, (b) 24 h after the HDX procedure, (c) after re-dissolving in water, (d) 3 days after deuterium-hydrogen-exchange (DHX), and (e) 5 days after DHX. The reaction scheme is shown accordingly. Subsequently to HDX, the signal shifts 6 Da, which indicates that in the peptide conjugate 7a, six hydrogen atoms have been exchanged to deuterons. After deuterium-hydrogen exchange, isotopic signals shift slowly to the previous values, which indicates that the deuterons are labile and not resistant to back-exchange in neutral pH. The DHX process takes much more time than HDX, and after 5 days of incubation in water solution, less than half of the deuterons are fully exchanged to hydrogens. Although it was not possible to obtain a stable isotopologue using HDX, the synthesis of this type of isotopic tagging reagent may be feasible. Considering the fact that the formation of 5-membered triazolinium ring by means of the tandem Mannich-electromphilic amination reaction requires formaldehyde, an application of deuterated formaldehyde would result in the development of a not expensive isotopic tag. , which correspond to the Hoffman elimination, homolytic cleavage, and charge remote fragmentation, respectively, as in the case of 5a. Analogously, we have observed the change of the peak pattern with increasing collision energy. Furthermore, we identified the *b 3 y 3 , *c 3 y 3 , and *b 3 x 3 ions, whose structures are schematically given in Figure 7b).
We have also performed an MS/MS experiment for the deuterated analogue: parent ion-643.2, CE = 15 eV (Figure 7c). Similarly, the same fragmentation pattern as for the non-deuterated analogue was observed, which confirms that the exchangeable hydrogens are not located on ethyl groups of Safirinium P system. In the MS/MS spectrum, there were also present signals that correspond to the deuterated analogues of daughter ions, as indicated previously in the spectrum presented in panel (b). Hence, shifts of 5 Da for the *c 3 y 3 -(d 5 ) and 3 Da for the *b 3 y 3 -(d 3 ) and *b 3 x 3 -(d 3 ) ions, with respect to the number of labile protons in the daughter deuterated ions, were observed in the spectrum, respectively. indicate the unreacted 5b ester (m/z calc. 515.359) and compound 3b formed by hydrolysis of 5c, respectively. We have also identified the * · c 3 y 3 radical ion, which is the product of the fragmentation of 7b, i.e., the loss of one octyl group. Panel b represents the MS/MS spectrum of 7b: parent ion: 805.5, CE = 30 eV. The CID fragmentation of 7b results in the peak pattern (signals 691-693 m/z), which corresponds to the loss of 113, 114, and 115 Da. As in the cases of compounds 7a and 5a, the observed pattern results from Hoffman elimination, homolytic cleavage, and charge remote fragmentation, respectively. The change of peak pattern with respect to increasing collision energy, from 20 to 30 eV, has been analyzed. Similarly to the previous experiments, the most abundant peaks shifted to lower m/z values; however, according to the MS/MS spectrum of 7b, other fragmentation products occur in both MS and MS/MS spectra. For example, the structure of one of the low mass fragmentation products (276.193) is shown in Figure 8b.
The MS/MS spectrum of the unreacted probe 5b obtained with a collision energy of 25 eV shows signals at 515. 3 and 288.195, which correspond to the parent and daughter ions as depicted in Figure 8c. The characteristic Safirinium P system fragmentation pattern leading to the ions with m/z from 401 to 403, as well as shift to lower m/z with respect to rising collision energy, have been observed. However, it should be pointed out that lower CE are needed for the fragmentation of 5b compared to the conjugate 7b.
Finally, an MS/MS experiment of ion m/z = 418.336, which corresponds to compound 3a, has been performed. The spectrum at collision energy of 20 eV is shown in Figure 8d. In case of this spectrum, the characteristic fragmentation leading to the formation of ions with m/z from 304 to 306 has been evidenced. Similarly to other Safirinium P systems presented above, a shift to lower m/z values with rising collision energy has been recognized.
Summing up, a 5b ionization tag provides less information regarding the analyzed peptide than probe 5a, since the diethyl tag gives a single daughter ion, and its fragmentation requires higher CE when compared to the dioctyl counterpart.

Analysis of Conjugate 7c, i.e., Peptide Ac-AKF-NH 2 Tagged with Pyrrolidine Derivative 5c
Finally, a cyclic Safirinium P analogue 7c has been examined. This conjugate was expected to give different MS/MS results than the Safirinium P analogues described above, since it incorporates an azoniaspiro system. As indicated earlier by Setner et al. [52], the azoniaspiro systems do not undergo the Hoffman elimination. The MS spectrum of crude compound 7c is presented in Figure 9b. The most abundant signal in the spectrum at m/z 635.366 corresponds to the parent ion 7c (m/z calc. 635.367), while the low abundant signal at m/z = 471.273 indicates its daughter ion *b 3 . There is also a signal at m/z 345.158, which corresponds to the unreacted probe 5c. Subsequent MS/MS spectrum of the peptide conjugate was measured for the parent ion 635.4 (CE = 30 eV), as shown in Figure 9c. In the case of azoniaspiro Safirinium P analogue 7c, neither Hoffman elimination nor charge remote fragmentation, which are characteristic for other Safirinium compounds, has been observed. Along with the daughter ion *b, observed also in the MS spectrum, the MS/MS experiment was revealed *b 3 y 3 and *b 3 x 3 ions. Additionally, spectrum analysis proved fragmentation mechanisms that lead to homolytic cleavage in the azoniaspiro system with the formation of daughter radical ions * · c 3 y 3 , * · c 3 x 3 , and the ion with a given molecular formula: C 24 H 38 N 6 O +· . The structures of the daughter ions are shown in panel (a). It is probable that the homolytic cleavage occurs during CID fragmentation of azoniaspiro Safirinium P conjugates, which results in electron donation from Safirinium P system to the peptide chain. Hence, further fragmentations lead to c and x type ions, similarly to electron transfer dissociation processes described in the literature [53].
Similarly to other Safirinium P derivatives and peptide conjugates described above, we analyzed how peak patterns change with rising collision energies ( Figure 10). Hence, CE were set from 20 to 30 eV, since effective fragmentations required higher collision energies. In general, more sophisticated fragmentation patterns have been observed compared to previously analyzed Safirinium P conjugates, which implies more complicated fragmentation mechanism. However, CE increase did not affect the characteristic signal pattern as much as in previous analyses.  Finally, MS/MS analysis of the parent ion for ester 5c (345.1, CE = 20 eV) was performed. When compared to the remaining Safirinium P NHS esters, the characteristic fragmentations were not observed, and analysis required higher CE. At 20 eV, the ester starts to fragment, giving the daughter ion at m/z 315.098. Its probable structure is shown in the spectrum (Figure 9d).
In comparison to peptide conjugates 7a and 7b, unwanted Hoffman elimination and charge remote fragmentation do not occur in analyses of 7c; however, homolytic cleavage leads to rather complicated spectra. Similarly to an analysis of conjugate 7a, in the spectra of compound 7c, only signals of daughter ions containing the tagged lysine residue were enhanced, while the remaining ions containing alanine or phenylalanine residues were not visible; hence, peptide sequencing would not be possible. However, the Safirinium P system creates characteristic c and x type ions that usually are common for electron transfer dissociations (ETD).  Figure 11. Panel (a) includes structures that illustrate fragmentation pattern. Panel (b) shows MS spectrum of 8, i.e., the molecular ion (m/z calc. 633.3719, found 633.3674). The spectrum presents also signals that correspond to synthesis co-products, including acetylated and non-acetylated peptide intermediates with deletions of alanines or even with its additional insertion (704.406). The MS/MS spectrum of the parent ion 633.4 obtained with collision energy of 15 eV is shown in panel (c). The signal pattern from 603 to 605 Da indicates losses of 30, 29, and 28 Da, respectively. The same fragmentation patterns, i.e., Hoffman elimination, homolytic cleavage, and charge remote fragmentation were observed for probe 5a and peptide conjugate 7a. Similarly to other Safirinium P derivatives, a rise in collision energy resulted in abundance change, indicating that the main fragmentation product moves from homolytic cleavage product (m/z 604.333) to charge remote fragmentation product (m/z 603.325).
The MS/MS spectrum (Figure 11c) reveals an *x 4 ion and peak that resulted from methyl group loss, i.e., ion C 27 H 43 N 8 O 6 + (m/z 575.286), whose structure is shown in panel (a).
Furthermore, non-typical series of daughter ions with neutral losses of subsequent alanine residues were observed, which are analogues of *z ions and * · z radical ions without side chains, i.e., alanine methyl groups. Finally, the spectrum presents a highly abundant signal at m/z 414.230, which based on mass spectrum simulations corresponds to a radical ion with the chemical formula: Similarly, to other peptide-Safirnium P conjugated systems, the label enhances signal intensities; however, under MS/MS conditions, also radical reactions and additional fragmentations occur, which results in atypical fragmentation patterns. In the case of the peptide conjugate 8, peptide sequencing based on neutral losses was sophisticated albeit possible.

Derivatization of Ubiquitin Hydrolysate with Reagent 5a
The ubiquitin pepsin hydrolysate was tagged with diethyl derivative 5a. The reaction scheme along with ubiquitin hydrolysis procedure is shown in Scheme 4. Ubiquitin hydrolysate and its analogue tagged by 5a have been analyzed by MS and LC-MS. The mass spectrum of ubiquitin hydrolysate is shown in Figure 12. The most abundant signals correspond to the products identified previously by Jaremko et al. [54].  Successively, the ubiquitin hydrolysate tagged with Safirnium P 5a has been analyzed. The orange and the colored lines in the LC-MS analysis ( Figure 14) indicate total ion current and the extracted ion chromatograms, respectively. The black line indicate TIC of non-tagged ubiquitin hydrolysate. Hence, TIC of tagged peptides as well as the extracted ion chromatograms display significantly higher intensities when compared to the EIC intensities of underivatized ubiquitin hydrolysis products. Successively, the ubiquitin hydrolysate tagged with Safirnium P 5a has been analyzed. The orange and the colored lines in the LC-MS analysis ( Figure 14) indicate total ion current and the extracted ion chromatograms, respectively. The black line indicate TIC of non-tagged ubiquitin hydrolysate. Hence, TIC of tagged peptides as well as the extracted ion chromatograms display significantly higher intensities when compared to the EIC intensities of underivatized ubiquitin hydrolysis products.
The calculated m/z values for all identified signals of ubiquitin proteolysis products with different charges and different numbers of conjugated 5a tags are presented in Table 1 Figure 14. In general, the majority of peptide conjugates eluted in 14-25 min, which indicates that the Safirinium P system slightly increases the hydrophobicity of the analytes. The most abundant signals in the chromatogram were identified. Ester 5a reacts mainly with peptides that contain lysine residues but also with N-terminal free amine groups. The most abundant signals correspond to tagged peptides: H-VKTLTGKTITL-OH and H-VKTLTGKT-OH, which are rich in lysine residues and both conjugated to three Safirinium P tags, and H-AGKQLEDGRTLSD-OH, which conjugated to two Safirinium P tags. The intensities of these signals are in the range from 7000 to 8000; thus, they are approximately four times higher than the intensities observed for the corresponding non-tagged peptides. In some cases, the observed sensitivity enhancements amount up to eight times based on intensity comparison of EIC signals for free peptides and their tagged analogues. The calculated m/z values for all identified signals of ubiquitin proteolysis products with different charges and different numbers of conjugated 5a tags are presented in Table 1  Finally, the signals of 5a, its fragmentation products, and hydrolysed ester 3a were not observed, which indicates that the active ester 5a reacts efficiently with the ubiquitin hydrolysate.

Fluorescence Microscopy
Since NHS esters of Safirinum form covalent bonds with amine groups, we have tested their usefulness for fast, one-step non-specific labeling of E. coli bacteria. After 1 h of incubation of E. coli-pET30 bacteria suspended in phosphate-buffered saline (PBS) with Safirinum probe 5a at a concentration of 1 mg/mL the obtained cells pellet indicated fluorescence similar to that observed for strains: E. coli-GFP and E. coli-DsRed2 labeled with GFP (green fluorescence) and DsRed2 (red fluorescence) proteins, respectively. A lack of fluorescence was evidenced for the control pellet of E. coli-pET30 bacteria not incubated with Safirinum reagent ( Figure S1). In order to perform a better comparison, we have utilized PBS mixture of these three strains to visualize free floating bacteria by means of fluorescence microscopy (Figure 15a-d). Moreover, a long time continuous fluorescence microscopy experiment, i.e., 40 s continuous time lapse video (Video S1) was recorded to estimate the usefulness of Safirinum probes for cells labeling. Finally, we have analyzed the biofilm formed by all three bacterial strains after 6 h of incubation at 37 • C by fluorescence microscopy (Figure 15e-h). In all of the above experiments, Safirinum derivatization reagent acts as an efficient fluorescence label of bacterial cells with low background emission when analyzed with excitation wavelengths specific for GFP and TRITC (DsRed2) fluorescence ( Figure S2). In addition, the bacteria showed the ability to form a biofilm dependent on the presence of the Ag43 surface protein at the comparable level to that of the unstained strain. Hence, this experiment proved that the functionality of Ag43 protein after Safirinium staining is not compromised.

General Information
All chemicals were used as received from commercial sources (Acros Organics-Geel, Belgium, Alfa-Aesar-Ward Hill, MA, USA, or Sigma-Aldrich-Darmstadt, Germany) and used without further purification. All nuclear magnetic resonance (NMR) spectra were recorded on a Varian Mercury-VX 300 MHz spectrometer at 25 • C. Chemical shifts (δ) are given in parts per million (ppm) and internally referenced to solvent signals. Coupling constants (J) are reported in Hertz (Hz). Splitting patterns are designated as s (singlet), bs (broad singlet), d (doublet), t (triplet), or m (multiplet). The IR (KBr pellets) spectra were recorded on a Thermo Scientific Nicolet 380 FT-IR spectrometer. Melting points were determined on an X-4 melting point apparatus with a microscope and were uncorrected. Analytical reverse-phase high-performance liquid chromatography with mass spectrometry detection (RP-HPLC-MS) experiments were performed on Agilent 1200 chromatograph, equipped with the UV detector-210 and 280 nm and coupled with micrOTOF-Q mass spectrometer in positive electrospray ionization (ESI) mode (described above) and SHIMADZU (Kyoto, Japan) LCMS-2020 EV mass spectrometer. The separation was carried on a reversed phase column: Aeris Peptide, Phenomenex XB-C18 (50 × 2.1 mm, 100 Å, 3. For FT-ICR, the specified resolution was 1,000,000 FWHM, although the set one was 100,000 FWHM. For MS/MS, the parent ions were fragmented by collision-induced dissociation (CID) with argon as the collision gas. The collision energy was optimized for the best fragmentation efficiency. The mass spectra analyses were performed using Compass DataAnalysis 4.0 (Bruker, Bremen, Germany) software. Samples for lyophilization were frozen in liquid nitrogen and lyophilized on a Speedvac or Freeze-dryer from Labconto (Kansas City, MO, USA). The absorption and emission spectra of Safirinium dyes and their derivatives have been presented previously [42,43,48].

Synthesis of Peptide Conjugates 7a-c
Reactive ester (5a-c) (0.1 mmol) and lysine-containing peptide Ac-AKF-NH 2 (0.040 g, 0.1 mmol) were dissolved in 3 mL of DMF. The reaction mixture was stirred for 16 h at room temperature. The progress of the reaction was monitored by LC-MS. When the reaction was finished, the mixture was evaporated and washed with acetone (3 × 1 mL). Compound 7a has been reported previously [42]. The peptide conjugate 7a was subjected to hydrogen-deuterium exchange. To the 10 µL of 2 mM solution of 7a in water 10 µL of triethylamine (TEA) in 980 µL of D 2 O was added and subsequently incubated for 24 h at room temperature. Next, upon lyophilization, the sample was re-dissolved in water and incubated for up to 5 days for back-exchange. All the samples of both non-deuterated and deuterated analogues were analyzed by ESI-FT-ICR-MS and MS/MS.

Synthesis of the Derivatized Peptide 8
A model Ac-AAK(Mtt) sequence for on-resin tagging was synthetized manually using standard Fmoc-solid-phase-peptide-synthesis (Fmoc-SPPS) procedure on Fmoc-Lys(Mtt)-Wang resin (loading 0.58 mmol/g). Orthogonal protection of the lysine side chain by the methyltrityl (Mtt) group was applied, and it was deprotected by 1% trifuoroacetic acid (TFA)/dichloromethane (DCM). Ac-AAAK-Wang was treated on-resin with 5a. For this purpose, 500 mg of the peptydylresin was added to a polypropylene syringe with a sinter (Intavis) and swelled for 30 min in DMF. Then, 198 mg of 5a (0.58 mol-2 eq.) dissolved in 4 mL of DMF was added, and the mixture was stirred for 24 h at the rotator. When the reaction was complete, as confirmed by negative Kaiser test, the resin was washed by the following mixtures:

Fluorescence Microscopy of Bacteria
The E. coli Top10 bacteria (Invitrogen, Carlsbad, CA, USA) were transformed with the following plasmids: pSFOXB20-daGFP, pDsRed2 and pET30, giving strains: E. coli-GFP, E. coli-DsRed2, and E. coli-pET30. pSFOXB20-daGFP (Oxford Genetics, Oxford, UK) is a vector encoding GFP (Green Fluorescent Protein) controlled by the OXB20 constitutive promoter. The plasmid possesses an origin of replication derived from pBR322 and a kanamycin resistance gene. It was used to visualize the bacterial cells as green fluorescent using fluorescence microscopy. pDsRed2 (Takara, Saint-Germain-en-Laye, France) is a plasmid encoding DsRed2 (red fluorescent protein) controlled by the lac promoter. The plasmid is the derivative of pUC19 and encodes an ampicillin resistance gene. It was used to visualize the bacterial cells as red fluorescent. pET30 (Novagen-Merck, Darmstadt, Germany) plasmid possesses an origin of replication derived from pBR322 and a kanamycin resistance gene. This strain was labeled with Safirinum dye 5a and visualized by means of blue fluorescent staining. All used bacterial strains were cultivated overnight with agitation of 160 rpm at 37 • C in 50 mL of Luria broth (LB) supplemented with the appropriate antibiotics (100 µg ampicillin mL −1 and 20 µg kanamycin mL −1 ; Sigma-Aldrich). Overnight cultures were centrifuged and suspended in phosphate-buffered saline (PBS) to an optical density of OD 600 = 0.5. For fluorescence microscopy, the E. coli-pET30 strain suspended in PBS was incubated with dye 5a at the final concentration of 1 mg/mL for 60 min at room temperature. Then, the non-bound dye was removed by a five-fold exchange of PBS buffer by consecutive centrifugation and suspension steps. Finally, the Safirinum labeled E. coli-pET30 bacteria were adjusted with PBS to OD 600 = 0.5.
In the first type of experiments, equal volumes of all three bacterial strains were mixed. Then, 3 mL of the suspension was placed in 35 mm polystyrene dishes (Corning, New York, NY, USA), and free floating bacteria were visualized by fluorescence microscopy. In the second type of experiments, 3 mL of each bacterial strain was inoculated in a separate dish and incubated at 37 • C for 6 h. After this time, the formation of bacterial biofilm at the bottom of the dish was visualized using fluorescence microscopy. Each type of experiment was repeated three times.
The free floating or biofilm-forming bacteria emitting appropriate fluorescence were recorded with the Olympus IX73 inverted fluorescence microscope equipped with the UCPlanFL N 20×/0.70 or LUCPlanFL N 40×/0.60 objectives, a Olympus U-HGLGPS fluorescence light source, and a Hamamatsu Orcaflesh 2.8 CMOS digital camera. For blue, green, and red fluorescence, the following filter sets were used, respectively: DAPI HC BP, GFP-1828A, and TRITC-B-000 (all from Semrock, IDEX Corporation (Lake Forest, IL, USA). Single frame images and time-lapse digital videos were recorded using the Olympus cellSens Dimension 1.18 software. The time-lapse videos were converted to the mp4 format using Kdenlive.

Conclusions
Safirinium NHS active esters have been found useful in bacteria bioimaging applications and enhancement of peptide signals in ESI-MS analyses. The preparation of the enhancers is inexpensive and simple. The developed Safirinium P tags can be applied in solid phase peptide synthesis, which is now the most common and efficient way for the preparation of peptides, and the resulting labeled product is stable in the course of peptide cleavage using 95% TFA. The Safirinium labels undergo transformations during CID experiments, including Hofmann elimination, homolytic cleavage, and charge remote fragmentation, which limits their application and results in a characteristic peak pattern. Although the described compounds do not allow for neat peptide sequencing, the presented data proved that they react efficiently with both ε-amine groups of lysine and N-terminal amine groups of peptides. Thus, the introduction of these fluorescent tags [42], that bear quaternary ammonium salt (QAS) moiety, enables the effective detection of such labeled peptides by means of both fluorescence and mass spectrometry. When compared to the analysis of ubiquitin hydrolysate performed by Jaremko et al. [54], the protein sequence coverage is very good and reaches 80%. The Safirinium system enhances the intensity of MS signals up to eight times, which confirms its potential applications in ionization tagging and sensitive detection of analyte signals, especially of lysine-rich peptides. This may lead to the discovery of new biomarkers based on proteins of low abundance; therefore, the examined systems could find application as ionization enhancers. Furthermore, cell labeling experiments on E. coli Top10 bacteria revealed that the studied probes can stand the challenging environment of biological systems such as living cell cultures, and the fluorescent visualization of free-floating bacteria cells with Safirinium P dyes gives comparable results to those obtained for bacteria labeled with GFP (green fluorescence) and DsRed2 (red fluorescence) proteins.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.