Bioorthogonal Radiolabeling of Azide-Modified Bacteria Using [18F]FB-sulfo-DBCO

Purpose: This study was motivated by the need for better positron emission tomography (PET)-compatible tools to image bacterial infection. Our previous efforts have targeted bacteria-specific metabolism via assimilation of carbon-11 labeled d-amino acids into the bacterial cell wall. Since the chemical determinants of this incorporation are not fully understood, we sought a high-throughput method to label d-amino acid derived structures with fluorine-18. Our strategy employed a chemical biology approach, whereby an azide (-N3) bearing d-amino acid is incorporated into peptidoglycan muropeptides, with subsequent “click” cycloaddition with an 18F-labeled strained cyclooctyne partner. Procedures: A water-soluble, 18F-labeled and dibenzocyclooctyne (DBCO)-derived radiotracer ([18F]FB-sulfo-DBCO) was synthesized. This tracer was incubated with pathogenic bacteria treated with azide-bearing d-amino acids, and incorporated 18F was determined via gamma counting. In vitro uptake in bacteria previously treated with azide-modified d-amino acids was compared to that in cultures treated with amino acid controls. The biodistribution of [18F]FB-sulfo-DBCO was studied in a cohort of healthy mice with implications for future in vivo imaging. Results: The new strain-promoted azide–alkyne cycloaddition (SPAAC) radiotracer [18F]FB-sulfo-DBCO was synthesized with high radiochemical yield and purity via N-succinimidyl 4-[18F]fluorobenzoate ([18F]SFB). Accumulation of [18F]FB-sulfo-DBCO was significantly higher in several bacteria treated with azide-modified d-amino acids than in controls; for example, we observed 7 times greater [18F]FB-sulfo-DBCO ligation in Staphylococcus aureus cultures incubated with 3-azido-d-alanine versus those incubated with d-alanine. Conclusions: The SPAAC radiotracer [18F]FB-sulfo-DBCO was validated in vitro via metabolic labeling of azide-bearing peptidoglycan muropeptides. d-Amino acid-derived PET radiotracers may be more efficiently screened via [18F]FB-sulfo-DBCO modification.


■ INTRODUCTION
The diagnosis and management of bacterial infection in clinical practice frequently rely on imaging, including plain films (Xrays), computed tomography (CT), and magnetic resonance (MR).These tools primarily evaluate the structural changes that accompany infection, including the presence of abnormal tissue or fluid collections.The nuclear medicine techniques positron emission tomography (PET) and single photon emission tomography (SPECT) may also be applied, for example, indium-111 white blood cell scanning using a patient's own white blood cells to detect abnormal leukocyte accumulation. 1,2A major limitation of these methods is that they primarily detect the host immune response rather than the bacteria themselves, motivating the more recent development of PET radiotracers targeting bacteria-specific metabolism.This class of tracers uses metabolic pathways and structures that only bacteria possess, for example, the folate biosynthesis pathway and the special ability of bacteria to metabolize certain sugars and sugar alcohols including maltose, sorbitol, and mannitol.Following the groundbreaking report of Weinstein et al. validating the sorbitol derivative 2-deoxy-2-[ 18 F]fluoro-Dglucose ([ 18 F]FDS) as an Enterobacteriaceae-specific PET tracer, 3 many tracers were subsequently explored including folate-targeted probes ([ 11 C]trimethoprim, 4 α-[ 11 C]PABA, 5 2-[ 18 F]F-PABA 6,7 ) and sugars/sugar alcohols in addition to [ 18 F]FDS including [ 18 F] maltose derivatives 8−10 and 2-[ 18 F]fluoro-D-mannitol 11 ([ 18 F]FMtl).These radiotracers have been studied in numerous animal models of infection and more recently in patients with promising results. 12,13ur laboratory is particularly interested in musculoskeletal infections including vertebral discitis-osteomyelitis (VDO), 14 its typical nontuberculous causative pathogen Staphylococcus aureus, and PET radiotracers targeting the bacterial cell wall. 15ased on a report describing bacterial incorporation of exogenous D-methionine, we initially developed D-methyl-[ 11 C]-methionine as a PET tracer using a trivial modification of its common L-enantiomer radiosynthesis via [ 11 C]methyl iodide, 16 later developing an in-loop method using an acyclic homocysteine precursor 17 and studying the tracer in patients with suspected prosthetic joint infections. 18We also developed a radiosynthesis of D-[3-11 C]alanine, which showed remarkable sensitivity to bacterial infections in animal models of myositis, pneumonia, and VDO. 19For these carbon-11 tracers, the major limitation is the short half-life (t 1/2 = 20 min), making it potentially difficult to produce tracers on-demand in acute care/inpatient settings.In addition, this short half-life requires an on-site cyclotron, which would restrict access to these tools, which can only be generated in major academic centers.We have therefore focused on developing S. aureus-sensitive radiotracers using the longer t 1/2 radionuclide fluorine-18.−23 These studies have suggested that metabolic labeling of the bacterial cell wall using fluorine-18 is likely feasible, which motivates the current approach.
However, despite extensive and compelling data in the literature, the structural determinants of unnatural D-amino acid incorporation are not fully understood.For fluorine-18 tracer development, the serial design and radiosynthesis of various D-amino acid-derived PET tracers are remarkably costly and inefficient.In developing these radiopharmaceuticals, we sought a method whereby aspects of the D-amino acid scaffold could be more rapidly explored, including variable side-chain structure/length and C-terminal modification, as suggested by a recent report indicating both positive and negative effects of carboxylate substitution on fluorescent D- amino acid accumulation. 24Motivated by previous work in the Bertozzi lab, 25,26 we hypothesized that this search might be facilitated using a bioorthogonal chemistry approach, whereby an exogenous D-amino acid derived bearing a side-chain azide is incorporated into bacterial peptidoglycan and subsequently "discovered" via cycloaddition using a detectable strained cyclooctyne 27 (Figure 1).Analogous methods have targeted other components of the bacterial cell wall including its Nacetyl muramic acid 28,29 and 3-deoxy-D-manno-2-octulosonic acid (KDO) 30,31 content, using bioorthogonal chemistry.For both fluorescence and PET, these cell-wall specific structures may be used to visualize living bacteria. 15In this work, we developed a water-soluble and fluorine-18 labeled strained cyclooctyne prosthetic group ([ 18 F]FB-sulfo-DBCO) analogous to previously reported fluorescent dibenzocyclooctyne derivatives that could be used to generate pathogen-specific PET signals.We believe that this strategy could be used to establish a "pre-targeting" infection imaging method for which a patient-delivered D-amino acid would be followed by a PET tracer used to detect it.Understanding that clinical applications might require a bioorthogonal imaging approach using faster cycloaddition chemistry, 32 we chose strain-promoted azide− alkyne click chemistry (SPAAC) ligation for our proof-ofconcept study based on the fluorescence literature. 25Using a D-amino acid probe with a small side-chain (i.e., D-azido- alanine) would likely facilitate bacterial incorporation due to resemblance to D-alanine itself, whose robust bacterial incorporation is well-established. 33 Fura et al. showed an inverse relationship between side-chain size and bacterial accumulation for D-amino acid-derived fluorescent probes, 20 and our previous work demonstrated that D-[3-11 C]alanine incorporation was significantly higher than that of D- [methyl- 11 C]methionine by bacteria in vitro. 17,19This concept would also have the advantage of sensing only peptidoglycanincorporated D-amino acids in contrast to signals generated using D-[3-11 C]alanine, for example, which may be incorporated into bacteria and mammalian cells via other mechanisms.To date, most pretargeting approaches for PET have targeted cancer 34−36 and dementia 37 with numerous elegant studies published in the past decade.A recent report describing the detection of S. aureus using a radiolabeled antibody suggested the potential value of pretargeting, 38 partially motivating the current study.
In this paper, we describe the successful detection of azidemodified peptidoglycan in S. aureus using [ 18 F]FB-sulfo-DBCO and highlight its use in exploring modified D-amino acid incorporation.In the future, this approach may be used to detect living bacteria in vivo, which may require more reactive bioorthogonal cycloaddition pairs.

Fluorescence Detection of N 3 -Modified Bacteria Using Strain-Promoted Azide−Alkyne Cycloaddition.
In preparation for PET studies, we initially developed an efficient workflow to screen azide-modified bacteria using fluorescent dyes, including MB 543, Cyanine 5.5, and Alexa Fluor 488 (Figure 2A, Supporting Information Figure S1) analogous to previously reported cycloadditions for unnatural amino acid-containing muropeptide detection.Before attempting fluorescent modification of bacteria, we first tested AF 488 (DBCO) SPAAC ligation using commercially available azideand alkyne-modified agarose beads.As expected, AF 488 (DBCO) resulted in lower signals when incubated with the alkyne agarose compared to the agarose-azide control (P < 0.001) (Figure 2B).We also investigated the chemical stability of D-azido-alanine and showed using chiral stationary phase HPLC absence of racemization in mouse and human sera over 3 h at 37 °C (Supporting Information Figure S3).We subsequently incubated S. aureus with D-azido-alanine in several media: Phosphate Buffered Saline (PBS), M9 minimal Media (M9), Ham's F-12 (F-12), and Lysogeny Broth (LB) for 1 h.Following serial centrifugation and washing, the bacterial pellet was resuspended in media with MB 543 (DBCO) and Cy 5.5 (DBCO) for SPAAC ligation with facile visualization of ligation (Figure 2C) for all media tested.In contrast to the retained color for D-azido-alanine-treated bacteria, there was minimal SPAAC seen for D-alanine-treated S. aureus controls, a visual result further quantified for AF 488 (DBCO) ligation (Supporting Information Figure S3).These results formed the basis for subsequent studies using the cyclooctyne PET tracer [ 18 F]FB-sulfo-DBCO.
Radiosynthesis of [ 18 F]FB-sulfo-DBCO.Based on initial concerns that a hydrophobic cyclooctyne-derived PET probe might result in increased nonspecific binding, we sought a water-soluble PET tracer compatible for SPAAC ligation with robust and reproducible radiosynthesis.Based on the commercial availability of the strained cyclooctyne-containing and sulfated amine Sulfo DBCO amine, we envisioned rapid amide formation via the readily available activated ester Nsuccinimidyl 4-[ 18 F]fluorobenzoate ([ 18 F]SFB) 39 (Figure 3A), analogous to other reported fluorine-18 prosthetic groups. 40,41he amine precursor was reacted with [ 18 F]SFB for 20 min at 40 °C in the presence of Et 3 N.The reaction mixture was purified via semiprep HPLC using a Phenomenex Luna C18 column.The calculated radiochemical yield (RCY) for [ 18 F]FB-sulfo-DBCO starting from [ 18 F]fluoride was 50.8% ± 9.3 (decay corrected) with radiochemical purity of >99% and a calculated molar activity of 2.9 ± 0.6 GBq/μmol (Figure 3B).Analytical HPLC was used to confirm the identity of the new radiopharmaceutical [ 18 F]FB-sulfo-DBCO versus an NMR/HRMS-characterized cold standard (Figure 3C).
PET Labeling of Pathogenic Bacteria Including S. aureus Using [ 18 F]FB-sulfo-DBCO.We hypothesized that analogous to fluorescent SPAAC labeling, we could detect azide-modified bacteria using [ 18 F]FB-sulfo-DBCO (Figure 4; Supporting Information Figure S5).As in the fluorescent case, the ligation of [ 18 F]FB-sulfo-DBCO to azide-modified agarose beads was verified (Supporting Information Figure S4).We then tested the effect of D-azido-alanine concentration on [ 18 F]FB-sulfo-DBCO SPAAC ligation in S. aureus cultures through a series of dilutions determining that 5 mM D-azidoalanine produced the highest signals (Supporting Information Figure S6).We then compared [ 18 F]FB-sulfo-DBCO ligation in cultures treated with D-azido-alanine to those treated with 5 mM of L-azido-alanine (Figure 4B) showing significantly higher uptake for the D enantiomer (5.3-fold; P < 0.000001).The method was subsequently applied to longer side-chain lengths (Figure 4C), showing the highest SPAAC ligation for D-azido-alanine (N = 1), and to D-azido-alanine-OMe, which resulted in lower ligation versus the free carboxylate (5.6-fold; P = 0.000061) (Figure 4D).Finally, labeling experiments were conducted using the additional bacteria Listeria monocytogenes, Pseudomonas aeruginosa, Acinetobacter baumannii, and Escherichia coli with significantly more ligation observed in D-azido- alanine-treated cultures versus the D-alanine-treated controls for L. monocytogenes and A. baumannii (1.6-fold, P = 0.003 and 2.1-fold; P = 0.0003, respectively) (Figure 4E), with a similar pattern observed at lower D-amino acid concentrations (Supporting Information Figure S7).
Dynamic PET/CT Imaging Analysis of [ 18 F]FB-sulfo-DBCO in Normal Mice.The potential application of [ 18 F]FBsulfo-DBCO to in vivo imaging was analyzed in healthy mice.A cohort of healthy mice (N = 4) was studied using [ 18 F]FBsulfo-DBCO, with region of interest (ROI) analysis used to evaluate organ-specific tracer clearance (Figure 5) at early time points and ex vivo analysis (N = 5) at 90 min used to assess tracer retention in organs via harvesting and gamma counting.The initial peak uptake of [ 18 F]FB-sulfo-DBCO was slightly higher in the heart (C max = 15.8 ± 2.2% ID g −1 , T max = 0.2 ± 0.1 min) than in the lung (C max = 10.5 ± 3.7% ID g −1 , T max = 0.2 ± 0.1 min).Clearance of [ 18 F]FB-sulfo-DBCO from the lung (T 1/2 = 3.1 ± 2.3 min) was relatively slower than that from the heart (T 1/2 = 1.4 ± 1.0 min).After the initial uptake in the heart and lung, [ 18 F]FB-sulfo-DBCO was cleared through the liver and kidneys via both hepatobiliary and urinary excretion.In the kidneys, the peak uptake of [ 18 F]FBsulfo-DBCO (C max = 14.6 ± 2.5% ID g −1 ) was observed at the early time after intravenous injection (T 1/2 = 0.5 ± 0.4 min), followed by excretion through the bladder over the time of imaging.Meanwhile, [ 18 F]FB-sulfo-DBCO showed slightly higher uptake (C max = 17.1 ± 3.6% ID g −1 , T max = 4.8 ± 0.4 min) in the liver than in the kidneys, with slow washout (T 1/2 = 10.6 ± 3.3 min).[ 18 F]FB-sulfo-DBCO accumulated in the liver was excreted through the gallbladder, spleen, and gastrointestinal system, as shown in the time-course wholebody distribution images (Figure 5A), suggesting that [ 18    (1).(D) As an additional control, ligation using D-azido-alanine-treated S. aureus was compared to that using its corresponding methyl ester D-azido-alanine-OMe; this modification has been reported to decrease the incorporation of exogenous D-amino acid-derived structures.The SPAAC ligation using D-azido-alanine-treated S. aureus resulted in higher signal (5.6 fold, N = 3, P = 0.000061) vs methyl ester treatment and higher signal (6.2 fold, N = 3, P = 0.00006) vs treatment with D-alanine itself.(E) Using the D-azido-alanine [ 18 F]FB-sulfo-DBCO method to label a panel of clinically significant Gram-positive and Gram-negative pathogens (SA, S. aureus; LM, L. monocytogenes; PA.P. aeruginosa; AB, A. baumannii; and EC, E. coli).Significantly higher SPAAC labeling using D-azido-alanine vs the parent D-alanine was seen for S. aureus (6.2 fold, N = 3, P = 0.00006), L. monocytogenes (1.6 fold, N = 3, P = 0.003), and A. baumannii (2.1 fold, N = 3, P = 0.0003).several reports of imaging using pretargeted PET, 36,42 we attempted in vivo detection of azide-modified bacteria using [ 18 F]FB-sulfo-DBCO (Supporting Information Figure S8).An S. aureus culture was treated with D-azido-alanine as above and used to inoculate a murine cohort using previously reported methods. 19After 3 h, [ 18 F]FB-sulfo-DBCO was injected intravenously as above, and the mice were studied after 30 min via tissue harvesting and ex vivo biodistribution analysis.The calculated retention of fluorine-18 by infected muscle was ∼0.1% ID/g (Supporting Information Figure S9), which was significantly lower than that we observed for D-amino acid-derived 11 C tracers used successfully in vivo (∼7% ID/g for D-[3-11 C]alanine). 19

■ DISCUSSION
In this work, we explored a new technique for efficient screening of D-amino acid-derived fluorine-derived tracers using [ 18 F]FB-sulfo-DBCO, a new water-soluble alkyne compatible with SPAAC recognition of azide-bearing peptidoglycan muropeptides.This fluorine-18 prosthetic group was readily synthesized from [ 18 F]SFB, which is increasingly used in radiotracer synthesis.After validating our microbiology methods using fluorescent tools analogous to those previously reported, we showed that for several bacteria including S. aureus, [ 18 F]FB-sulfo-DBCO labeling of 3-azido-D-alaninetreated cultures showed higher PET signals than those treated with 3-azido-L-alanine, 3-azido-L-alanine-OMe, or longer sidechain controls.These results, and those previously reported, 20 have reinforced our thinking that for the incorporation of exogenous unnatural D-amino acids, resemblance to D-alanine itself is key.It is well-known that deviations from canonical amino acid structures (i.e., D-alanine, D-glutamate) impart changes to the biochemical behavior related to lipophilicity, size, and solvation.As expected, 3-azido-L-alanine-OMe showed poor [ 18 F] signals likely reflecting low muropeptide incorporation, with the intriguing hypothesis that certain Cmodified D-amino acids such as carboxamides actually increase bacterial uptake 24 to be explored in future work.In addition, other D-amino acid scaffolds may be pursued.Previous work has shown that numerous exogenous D-amino acids can be incorporated into peptidoglycan, including D-methionine, D- phenylalanine, D-tryptophan, and D-valine. 43In addition, another carbon-11 PET infection imaging strategy used D- [3-11 C]glutamine. 44,45Derivatives of these D-amino acids, therefore, represent additional routes for bioorthogonal bacterial detection using PET.
We also tested the new fluorine-18 SPAAC partner [ 18 F]FBsulfo-DBCO in vivo to predict how the reagent might be employed in future pathogen-specific imaging.Similar to previously reported aliphatic fluorine-18-labeled DBCO, [ 18 F]FB-sulfo-DBCO showed a dominant hepatobiliary excretion pathway in normal mice.Meanwhile, we found that [ 18 F]FB-sulfo-DBCO had lower background signals than the aliphatic fluorine-18-labeled DBCO likely due to the enhanced hydrophilicity of [ 18 F]FB-sulfo-DBCO by the insertion of a sulfonic acid group.In addition, aromatic fluorine-18-labeled [ 18 F]FB-sulfo-DBCO showed significantly lower bone uptake than that of the aliphatic fluorine-18-labeled DBCO. 46We found that [ 18 F]FB-sulfo-DBCO had weak background signals in normal mice.The accumulation of [ 18 F]FB-sulfo-DBCO was less than 1% ID/g in most organs, with the highest uptake seen in the stomach, liver, and spleen, which would potentially limit the potential detection of bacteria in those organs.We also observed weak signals in tissues infected with azidemodified S. aureus.For successful pretargeting, we will explore faster bioorthogonal cycloadditions for more sensitive bacterial detection. 32For example, larger rate constants are observed using inverse electron demand Diels−Alder (iEDDA) click chemistry (1−10 6 M −1 s −1 ), 47 which may enhance the in vivo feasibility of pretargeted PET imaging.In the future, we anticipate that in vivo bacterial detection will be possible using tetrazine-modified bacteria-specific substrates imaged using fluorine-18-labeled dienophiles.
To our knowledge, these are the first data supporting a pretargeted infection imaging approach using PET-compatible radiopharmaceuticals. Several analogous chemical biology approaches have been previously reported in the fluorescence literature, using cell wall-specific tools.In addition to the muropeptide-targeted azide−alkyne approach previously discussed, the alternating N-acetyl muramic acid/N-acetyl glucosamine (NAM/NAC) sugar backbone of peptidoglycan has also been explored by the Grimes group for detecting the bacterial cell membrane, via both azide and alkyne modification of the N-acetyl moiety of NAM. 28Several subsequent manuscripts reported additional NAM derivatives, potentially supporting modification of this scaffold for PET. 29ram-negative organisms have been labeled via azide-modified 3-deoxy-D-manno-octulosonic acid (KDO), a component of lipopolysaccharide (LPS) in the bacterial outer membrane. 31ensing LPS or "endotoxin" might have a significant impact on human health, based on both Gram-negative specific recognition and the interactions between LPS and the immune system resulting in fever and occasionally septic shock. 48The discovery of these chemical biology tools has also significantly accelerated our understanding of microbe behavior.
A major obstacle to PET tracer discovery is the low throughput of radiosynthesis and radiopharmaceutical validation, with exploration of a new tracer concept typically conducted with several steps in series (precursor synthesis, variable radionuclide incorporation, lengthy HPLC characterization and purification, stability testing, and biologic validation).Therefore, PET methods compatible with highthroughput analysis are critical for the efficient discovery of new pharmaceuticals.The fluorescence literature provides numerous clues as to bacteria-tolerated unnatural D-amino acids, but we wanted to develop a method that would more efficiently generate [ 18 F] radiopharmaceuticals compatible with PET imaging.In particular, modifications of the D-amino acid side chain and C-terminus might result in tracers with enhanced microbial sensitivity that are potentially compatible with a pretargeting imaging method.We anticipate that [ 18 F]FB-sulfo-DBCO will have further use in discovering bacteria-specific tools and new radiopharmaceuticals for other applications.

■ CONCLUSIONS
In conclusion, we have synthesized and tested the new watersoluble [ 18 F]SFB-derived strained cyclooctyne radiotracer [ 18 F]FB-sulfo-DBCO for SPAAC labeling of the bacterial cell wall.Bacteria pretreated with azide-bearing D-amino acids including 3-azido-D-alanine showed increased [ 18 F]FB-sulfo-DBCO labeling versus cultures treated with the cognate L- amino acid or a C-terminal modified derivative.In vivo, [ 18 F]FB-sulfo-DBCO showed rapid clearance and low background with no evidence of defluorination.Future efforts will focus on optimizing this pretargeting imaging approach and developing [ 18 F]-labeled tracers with minimal structural/steric alterations versus the D-alanine parent.
[ 18 F]FB-sulfo-DBCO was isolated in a 2−3 mL fraction.The fraction was diluted with H 2 O (30 mL) before being passed through a Sep-pak light C18 Cartridge at 5 mL/min.An additional 10 mL of H 2 O was used to wash the cartridge.After the cartridge was flushed with air, the product was eluted using EtOH (0.5 mL) for direct formulation prior to in vitro or in vivo use.Isolated [ 18 F]FB-sulfo-DBCO had RCY = 50.8%± 9.3 n.d.c (n = 12), RCP ≥ 99%, and A m = 2.9 ± 0.6 GBq/ μmol.The chemical purity of [ 18 F]FB-sulfo-DBCO was

Figure 1 .
Figure 1.Strain-promoted azide−alkyne cycloaddition (SPAAC) strategy for metabolic labeling of S. aureus.(A) Analogous to Bertozzi's work, an azide-bearing artificial D-amino acid is incorporated into the muropeptides of S. aureus peptidoglycan and subsequently labeled with a labeled strained cyclooctyne for detection.In this work the cyclooctyne was modified with a dye for fluorescent/visual detection or with 18 F via [ 18 F]SFBderived [ 18 F]FB-sulfo-DBCO.(B) Potential modifications of the amino acid scaffold for the metabolic labeling of pathogenic bacteria.We considered using D vs L amino acids (2-position stereochemistry), side-chain length (1, 2, 3, and 4 added methylene groups), and modification of the C terminus via a methyl ester control.Modification of the C-terminus has been shown to have a profound effect on the incorporation of exogenous D-amino acids with both decreased and increased uptake vs the carboxylate parent.

Figure 2 .
Figure 2. Fluorescence/visual detection of N 3 -modified bacteria using a reported strained cyclooctyne method.(A) Chemical structures of dibenzocyclooctyne (DBCO) conjugates of MB 543 (543 nm λ ex , 566 nm λ em ), Cy 5.5 (683 nm λ ex , 703 nm λ em ), and AF 488 (488 nm λ ex , 496 nm λ em ).The first two dye conjugates were chosen for easy visual identification of muropeptide-retained signal, while AF 488 was employed for improved quantification of fluorescent signals.(B) Quantification of signals generated by AF 488 ligation to azide-labeled agarose beads was measured to test the feasibility of SPAAC ligation.Agarose beads were separated from media to measure SPAAC fluorescence.Azide agarose was used as a positive control to approximate the upper limit of measurable fluorescence, while alkyne agarose was used as a negative control.Data are represented as a percentage of fluorescent SPAAC ligation to the agarose beads in comparison to its filtrate (N = 3).(C) A series of SPAAC reactions in S. aureus grown (left to right) in PBS, M9, F-12, and LB.D-Azido-alanine can be visualized via ligation with the alkyne-labeled fluorophore partners MB 543 (DBCO) and Cy 5.5 (DBCO).In contrast, the negative control D-Ala does not react with fluorophores bearing alkynes.

Figure 4 .
Figure 4. SPAAC radiolabeling of pathogenic bacteria using [ 18 F]FB-sulfo-DBCO.In all cases, incorporated radioactivity was determined using a gamma counter.All data are represented as a percentage of radiolabeled SPAAC ligation to the bacterial pellets in comparison to the filtrate, normalized by their final measured OD 600 (N = 3 per experiment).(A) Cycloaddition product of D-azido-alanine-modified S. aureus and [ 18 F]FBsulfo-DBCO.(B) SPAAC ligation using D-azido-alanine vs L-azido-alanine modified S. aureus, showing higher uptake for the D-amino acid N 3 partner (5.3 fold, N = 3, p < 0.000001).(C) The effect of increased N 3 D-amino acid side chain length via increasing the number of methylene additions (1−4) on SPAAC incorporation in S. aureus.The highest labeling was observed for D-azido-alanine(1).(D) As an additional control, ligation using D-azido-alanine-treated S. aureus was compared to that using its corresponding methyl ester D-azido-alanine-OMe; this modification has been reported to decrease the incorporation of exogenous D-amino acid-derived structures.The SPAAC ligation using D-azido-alanine-treated S. aureus resulted in higher signal (5.6 fold, N = 3, P = 0.000061) vs methyl ester treatment and higher signal (6.2 fold, N = 3, P = 0.00006) vs treatment with D-alanine itself.(E) Using the D-azido-alanine [ 18 F]FB-sulfo-DBCO method to label a panel of clinically significant Gram-positive and Gram-negative pathogens (SA, S. aureus; LM, L. monocytogenes; PA.P. aeruginosa; AB, A. baumannii; and EC, E. coli).Significantly higher SPAAC labeling using D-azido-alanine vs the parent D-alanine was seen for S. aureus (6.2 fold, N = 3, P = 0.00006), L. monocytogenes (1.6 fold, N = 3, P = 0.003), and A. baumannii (2.1 fold, N = 3, P = 0.0003).