Fluorescence-Quenched Mycoloyl-Arabinofuranoside Analogs: Fluorogenic Probes for Mycobacterial Outer Membrane-Degrading Enzymes

Abstract

Mycobacteria, including the tuberculosis pathogen Mycobacterium tuberculosis, are enclosed by a highly complex cell envelope with an outer membrane, or mycomembrane, which provides extraordinary protection from antibiotics and other stresses. The inner leaflet of the mycomembrane consists of arabinogalactan-linked mycolate (AGM), which is an enormous glycoconjugate comprising mycolic acids esterified to terminal D-arabinofuranosyl residues of an underlying arabinogalactan–peptidoglycan complex, also referred to as the mycoloyl-arabinogalactan–peptidoglycan (mAGP) complex. Whereas AGM biosynthesis is comparatively well characterized, less is known about AGM degradation by endogenous or exogenous factors. To facilitate studies on AGM breakdown by hydrolytic enzymes, here we synthesized fluorescence resonance energy transfer (FRET)-based mono- and disaccharide probes that mimic fragments of AGM and are designed to fluoresce upon cleavage. We devised a synthetic route that established the glycolipid core with the desired regio- and stereochemistry and allowed late-stage selective functionalization of the core with a FRET pair. Our data show that the intact FRET-AGM probes exist in a fluorescence-quenched state, but when exposed to lysin B (LysB), an AGM-degrading mycobacteriophage hydrolase with therapeutic relevance, the probes were activated through lipid ester hydrolysis, thereby generating fluorescence signal. FRET-AGM probes were activated by known mycomembrane glycolipid hydrolases, but not by several other types of hydrolases, demonstrating specificity. FRET-AGM probes may be useful in the future for identifying novel AGM hydrolases and quantitatively monitoring the activity of AGM hydrolases, which could provide insights into mycomembrane degradative processes and aid in tuberculosis therapeutic development.

1 |. Introduction

Tuberculosis (TB), which is caused by Mycobacterium tuberculosis (Mtb), killed 1.3 million people in 2023 [1]. A major factor in the pathogenesis of Mtb and related mycobacterial pathogens is their distinctive and highly complex cell envelope, which comprises layers of plasma membrane (PM), peptidoglycan (PG), arabinogalactan (AG), and outer mycomembrane (MM) (Figure 1A) [2–4]. The latter three layers are part of a massive covalent complex: PG is covalently linked to the AG, which, in turn, is covalently decorated by long-chain, branched, β-hydroxy fatty acids called mycolic acids (MA) [5]. The AG-linked mycolates (AGM) form the MM’s inner leaflet, while its outer leaflet contains trehalose dimycolate (TDM) as well as other lipids and glycolipids (Figure 1B). AGM and TDM form a remarkably hydrophobic barrier that protects the cell from environmental stresses, including host immune pressure and antibiotic treatment. In Mtb, the biosynthetic pathway that produces AGM and TDM from the common precursor trehalose monomycolate (TMM) is essential, and its protein components, such as Pks13, MmpL3, and mycoloyltransferases (Ag85A-C), are considered promising targets for TB antibiotic development [6–9].

FIGURE 1 |.

(A) Schematic of the mycobacterial cell envelope. (B) Chemical structures of AGM and TDM. The fragment of AGM exploited for probe development in this study is highlighted in blue.

While MM remodeling and degradative processes are less well characterized than their biosynthetic counterparts, enzymes that hydrolyze MM glycolipids have been identified and have generated interest for their possible therapeutic value. Mycobacterial TDM hydrolase (Tdmh) hydrolyzes TDM to free MA during stress (Scheme 1A), which promotes biofilm formation in the model organism Mycobacterium smegmatis (Msmeg) [10], and balances defense and nutrient acquisition in Mtb [11]. When exogenously administered, Tdmh lyses mycobacterial cells, implying that mycobacteria carefully regulate endogenous expression of Tdmh to avoid autolysis [12]. To date, no AGM hydrolase endogenous to mycobacteria has been reported, although such activity presumably exists to facilitate cell envelope turnover. However, a serine hydrolase named lysin B (LysB) encoded by mycobacteriophages hydrolyzes AGM and, to a certain extent, TDM to release free MA (Scheme 1B) [13]. LysB complements the activity of a canonical endolysin LysA, which hydrolyzes PG, to complete the degradation of the mycobacterial cell envelope and the release of phages at the burst point [13]. Importantly, purified recombinant LysB from mycobacteriophage D29 rapidly and efficiently kills multiple species of mycobacteria in vitro and in vivo, including Mtb, Msmeg, and Mycobacterium abscessus [14, 15]. Tdmh and LysB likely exert their antimycobacterial activity through impairment of the MM, specifically through the hydrolysis of TDM and/or AGM, although mechanistic understanding of these processes remains limited. Continued efforts to identify and characterize hydrolytic enzymes that degrade AGM and TDM will improve our fundamental understanding of MM remodeling and breakdown. Furthermore, such efforts may lead to new targets for the development of conventional antibiotics and/or new therapeutic enzyme candidates.

SCHEME 1 |.

Enzymatic hydrolysis of (A) TDM by mycobacterial Tdmh and (B) AGM by mycobacteriophage lysin B. (C) Concept of FRET-based fluorogenic glycolipid probes. F, fluorophore; FRET, fluorescence resonance energy transfer; Myc, mycolate chain; Q, quencher.

Chemical probes have been developed to facilitate the investigation of MM components and processes, primarily for the purpose of probing glycolipid biosynthesis, dynamics, and protein interactions [16]. Various groups, including ours, have developed a toolbox of probes based on MM biosynthetic precursors (e.g., trehalose, TMM, or related scaffolds) that enable metabolic labeling of different MM glycolipids with bioorthogonal, fluorescent, ^19/18F, and photocrosslinking groups [17–29]. In addition, the Kiessling group introduced a FRET-based fluorogenic TMM analog, termed QTF (quencher–trehalose–fluorophore), which converts from a quenched to fluorescent state upon processing by MM biosynthetic enzymes, specifically Ag85 mycoloyltransferases [30]. Less research has focused on probe development for investigation of MM degradative processes. To allow monitoring of TDM degradation, we previously developed a FRET-based fluorogenic TDM analog, called FRET-TDM, which fluoresces upon hydrolysis by Tdmh [31]. FRET-TDM was used to probe activity of Tdmh in recombinant purified form and when expressed in live mycobacterial cells [31]. Thus, synthetic fluorescence-quenched analogs of mycobacterial glycolipids have been successfully developed as fluorogenic probes of MM biosynthesis and remodeling processes.

Although AGM is a foundational component of the mycobacterial cell envelope, to date there have been no chemical tools specifically designed to probe AGM breakdown. Here, we applied the FRET probe concept that was successfully employed in the development of QTF and FRET-TDM to create the first fluorescence-quenched AGM analogs as AGM degradation probes (Scheme 1C). We synthesized mono- and disaccharide AGM fragments bearing strategically placed FRET pairs, i.e., FRET-AGM probes, and demonstrate that they are activated by LysB, an AGM-hydrolyzing enzyme from mycobacteriophages. We anticipate that FRET-AGM probes will be useful for discovering and characterizing AGM-targeting hydrolases, which could advance our knowledge of MM remodeling by endogenous or exogenous factors, as well as developing new enzyme-based TB therapeutic candidates.

2 |. Results and Discussion

2.1 |. Probe Design and Synthetic Strategy

We designed the FRET-AGM probes based on the AGM structure shown in Figure 1B. According to the model from Mtb [32], the AG heteropolysaccharide consists of a PG-connected galactan chain made up of ~30 galactofuranosyl (Galf) residues, which is modified by 2–3 branched arabinan chains made up of ~23 arabinofuranosyl (Araf) residues each. At the reducing ends of the arabinan chains are branched Araf₅ pentasaccharide motifs, which are decorated with 5-O-mycoloyl groups at approximately two-thirds of the terminal Araf residues. The resulting AGM structure forms the inner leaflet of the MM. As noted above, AGM can be broken down through enzyme-catalyzed mycoloyl ester hydrolysis at the terminal mycoloyl-Araf units (Scheme 1B).

We reasoned that FRET probes mimicking mycoloyl-Araf unit(s) of AGM, modified by fluorophore and quencher moieties at appropriate positions, would undergo fluorescence activation upon enzymatic ester hydrolysis (Scheme 1C). Given the high complexity of the multiply mycoloylated Araf₅ pentasaccharide motif, in this study we designed simplified mono- and disaccharide analogs as first-generation probes. The target structures and retrosynthetic analysis of the probes are shown in Scheme 2. The FRET-AGM general structures I consist of (i) either a mono- or di-Araf core, (ii) a dabcyl quencher linked to the reducing end through an octyl chain, and (iii) a fluorescein fluorophore linked to the nonreducing end through a 5-O-mycoloyl-mimicking chain. These probes, termed FRET-AGM-Mono and FRET-AGM-Di, resemble the blue-highlighted AGM fragment shown in Figure 1B. We chose Araf units with α-configured linkages to facilitate stereoselective glycosidic bond formations via neighboring group participation. In place of the native mycoloyl chain, we chose to use a truncated linear fatty ester to facilitate synthesis and enhance the solubility of the probes, as we did previously for FRET-TDM [31].

SCHEME 2 |.

FRET-AGM target structures and retrosynthetic analysis.

As shown in Scheme 2, we envisioned that FRET-AGM-Mono and FRET-AGM-Di targets I could be accessed from their corresponding intermediates II, which contain orthogonal azido- and 9-fluorenylmethoxycarbonyl (Fmoc)-protected amino groups to allow for late-stage FRET pair installation at the appropriate sites. Intermediates II could be obtained from the common synthetic precursor III, an Araf trichloroacetimidate glycosyl donor with a 2-O-acetyl group to enable selective 1,2-trans glycosylation and an orthogonally protected 5-O-position to enable selective esterification of that site with an appropriate fatty acid mimic.

2.2 |. Synthesis of FRET-AGM-Mono

At equilibrium, D-arabinose (1) predominantly exists in the pyranose form in aqueous solution [33]. To trap D-arabinose in the furanose form, it was reacted with the bulky silylation reagent tert-butyldimethylsilyl (TBDMS) chloride, which preferentially reacts with the 5-position primary hydroxyl group exposed in D-arabinofuranose (Scheme 3) [34]. This reaction generated the intermediate 5-O-tert-butyldimethylsilyl D-arabinofuranose, which was directly peracetylated using acetic anhydride in the same pot to yield compound 2 in 59% yield over two steps. Conversion of compound 2 to hemiacetal 3 was accomplished in 54% yield through selective anomeric deacetylation with benzylamine, after which the corresponding trichloroacetimidate glycosyl donor 4 was produced by reaction with trichloroacetonitrile in the presence of catalytic 1,8-diazabicyclo [5.4.0] undec-7-ene (DBU). Schmidt glycoslyation [35] of donor 4 with 8-chlorooctanol using catalytic trimethylsilyl trifluoromethane-sulfonate (TMSOTf) delivered compound 5 in 58% yield over two steps. Owing to 2-O-acetyl group participation, the glycosylation proceeded with complete α-stereoselectivity, as evidenced by ¹³C NMR analysis showing the anomeric position absorption at δ 106 ppm, consistent with the established δ 100–110 ppm range for α-arabinofuranosides (β-arabinofuranosides exhibit C-1 chemical shifts in the δ 95–100 ppm range) [36]. The chloro leaving group of 5 was displaced by azide anion via S_N2 reaction, giving compound 6 in 88% yield. Thus, the Araf monosaccharide core was formed with an α-configured C-1 octyl linker containing a terminal azido group for selective late-stage quencher installation.

SCHEME 3 |.

Synthesis of FRET-AGM-Mono (14). Reagents and conditions: (a) TBDMSCl, pyridine, rt; (b) acetic anhydride, pyridine, rt (59% over two steps); (c) benzylamine, THF, rt (54%); (d) DBU, Cl₃CCN, CH₂Cl₂, 0°C; (e) 8-chlorooctanol, TMSOTf, CH₂Cl₂, 0°C → rt (58% over two steps); (f) NaN₃, DMF, 60°C (88%); (g) NaOMe, MeOH, rt (96%); (h) NaH, PMBCl, DMF, 0°C → rt (51%); (i) TBAF, THF, rt (97%); (j) N-Fmoc-8-aminooctanoic acid, DCC, DMAP, CH₂Cl₂, rt (93%); (k) DDQ, CH₂Cl₂/H₂O (10:1), rt (89%); (l) piperidine, DMF, rt; (m) FITC, K₂CO₃, DMF (42% over two steps); and (n) 13, CuSO₄·5H₂O, sodium ascorbate, THF/MeOH/H₂O (1:1:1), 40°C (54%).

To avoid later incompatibility issues between the O-acetyl protecting groups and a 5-O-mycoloyl-mimicking chain that would arise during later deacetylation, we swapped the protecting groups at the 2- and 3-positions. To provide a balance of stability and flexibility, we selected the para-methoxybenzyl (PMB) protecting group, which was predicted to withstand the reaction conditions of the remaining steps but permit removal under mild acidic or oxidative conditions that would leave other sensitive functional groups intact (e.g., ester, azido groups) [37, 38]. Thus, compound 6 was subjected to de-O-acetylation using catalytic sodium methoxide in 96% yield, and the resulting diol 7 was para-methoxybenzylated to give compound 8 in 51% yield. Selective removal of the 5-O-TBDMS group was achieved with tetrabutylammonium fluoride (TBAF), giving alcohol 9 in 97% yield and setting up subsequent addition of the fatty acyl group. Dicyclohexylcarbodiimide (DCC)-mediated esterification was used to couple compound 9 with N-Fmoc-8-aminooctanoic acid in 93% yield, which added the mycolate-mimicking 5-O-octanoyl group containing an Fmoc-protected latent amine. Evidence of 5-O-acylation was provided by ¹H NMR analysis, which indicated a downfield shift of the H-5 absorption upon addition of an electron-withdrawing acyl group (δ 3.81 → 4.27 ppm).

The final stage of the FRET-AGM-Mono synthesis involved stepwise site-selective installation of the FRET pair on the Araf scaffold. First, the PMB protecting groups on the 2-O- and 3-O-positions of compound 10 were removed under mild oxidative conditions, specifically use of 2,3-dichloro-5,6-dicyano-1, 4-benzoquinone (DDQ). The resulting intermediate 11 was subjected to piperidine-mediated deprotection of the Fmoc group to reveal an amino group. Subsequently, treatment of the intermediate with amine-reactive fluorescein isothiocyanate (FITC) installed the fluorophore on the end of the 5-O-fatty acyl chain. The resulting FITC-modified intermediate 12 was obtained in 42% yield over three steps starting from compound 10. Although our initial plan was to reduce the azido group of 12 to its corresponding amine, which would permit addition of the quenching unit through N-hydroxysuccinimide (NHS) ester coupling, attempts to reduce the azido group also reduced the fluorescein group. Thus, we turned to Cu(I)-catalyzed azide–alkyne cycloaddition (CuAAC) reaction to attach the quencher. Azide 12 was reacted with alkyne-labeled dabcyl reagent 13 [39] in the presence of CuSO₄ and sodium ascorbate. Purification by reverse-phase (C18) column chromatography yielded FRET-AGM-Mono in 54% yield. Interestingly, high-resolution electrospray ionization mass spectrometry analysis indicated that the obtained product was structure 14, suggesting that CuAAC conditions converted the thiourea to a urea linkage, which was unanticipated but considered inconsequential with respect to the probe design and function. Overall, this synthesis delivered FRET-AGM-Mono (14), a fluorescence-quenched AGM monosaccharide fragment bearing fluorophore and quencher units at appropriate locations to enable fluorescence activation through enzymatic cleavage of the 5-O-ester bond.

2.3 |. Synthesis of FRET-AGM-Di

The synthesis of FRET-AGM-Di is shown in Scheme 4. First, to establish the designed di-Araf core, a Schmidt glycosylation between glycosyl acceptor 9 and glycosyl donor trichloroacetimidate 4 was conducted in the presence of catalytic TMSOTf as Lewis acid. Owing to neighboring group participation of the 2-O-acetyl group of donor 4, the glycosylation proceeded with complete α-stereoselectivity, generating disaccharide 15 in 75% yield. As done for the above-described glycosylation of 4 by 8-chlorooctanol (Scheme 3), the α-configurations of C-1 positions of 15 were confirmed by their ¹³C NMR chemical shifts (δ 106.4, 105.7 ppm). With the di-Araf core formed, the remainder of the synthesis closely followed that described for FRET-AGM-Mono. Briefly, the acetyl groups of compound 15 were removed (giving 16 in 66% yield) and swapped for PMB groups (giving 17 in 60% yield), after which the Ara-II 5-O-position was desilylated (giving 18 in 83% yield) and esterified with N-Fmoc-8-aminooctanoic acid (giving 19 in 72% yield). Compound 19 was converted to fluorescein-modified 21 through sequential DDQ-mediated PMB removal (giving 20 in 94% yield), followed by piperidine-mediated Fmoc removal and reaction with FITC (giving 21 in 69% over two steps). Finally, CuAAC-mediated coupling of azide 21 with alkyne-dabcyl 13 furnished FRET-AGM-Di (22) in 41% yield, again with concomitant conversion of the thiourea linkage to a urea linkage. Thus, a similar strategy of late-stage FRET pair installation on an orthogonally bifunctionalized Araf core was used to deliver both FRET-AGM probes in this study.

SCHEME 4 |.

Synthesis of FRET-AGM-Di (22). Reagents and conditions: (a) TMSOTf, CH₂Cl₂, 0°C → rt (75%); (b) NaOMe, MeOH, rt (66%); (c) NaH, PMBCl, DMF, 0°C → rt (60%); (d) TBAF, THF, rt (83%); (e) N-Fmoc-8-aminooctanoic acid, DCC, DMAP, CH₂Cl₂/H₂O (10:1), rt (72%); (f) DDQ, CH₂Cl₂, rt (94%); (g) piperidine, DMF, rt; (h) FITC, Et₃N, DMF (69% over two steps); and (i) 13, CuSO₄·5H₂O, sodium ascorbate, THF/MeOH/H₂O (1:1:1), 40°C (41%).

2.4 |. FRET-AGM Probes Are Activated by Mycomembrane-Degrading Hydrolases

We first assessed the quenching efficiency of the FRET-AGM probes in their intact, fluorescence-quenched states. To enable comparison, we used the fluorescein-modified decanoic acid derivative FITC-FA (23) [31], which models the fluorescent product of esterase-catalyzed cleavage of both FRET-AGM probes. Fluorescence intensity measurement of FRET-AGM probes 14 and 22 in comparison with compound 23 across a range of concentrations potentially used in biological experiments (1 nM–10 μM) revealed that FRET-AGM-Mono and FRET-AGM-Di had quenching efficiencies of 98.9% and 93.9%, respectively (Figure 2A). The lower quenching efficiency of FRET-AGM-Di (93.9%) is likely due to a greater distance between the fluorophore and quencher units resulting from the larger di-Araf core. Thus, FRET-AGM-Di may have modestly reduced assay sensitivity compared to FRET-AGM-Mono. Regardless, these data show that both FRET-AGM probes are fluorogenic, exhibiting minimal fluorescence emission intensity in their intact form and higher intensity in their ester-cleaved form.

FIGURE 2 |.

FRET-AGM probes are activated by mycobacteriophage D29 LysB and mycobacterial Tdmh. (A) The quenching efficiency of intact FRET-AGM probes was determined by comparing the slopes of lines generated from the plot of fluorescence intensity versus concentration of FRET-AGM probes (14 or 22) and a model fluorescent activation product FITC-FA (23). (B) SDS-PAGE and Western blot analysis of purified recombinant LysB and LysB-S82A. (C) Plot of time-dependent activation of FRET-AGM probes (0.1 μM) by LysB (20 μg/mL) in Tris-HCl buffer at 37°C compared to probe-only controls. (D–F) Activation of FRET-AGM probes (0.1 μM) by LysB versus LysB-S82A (20 μg/mL) in Tris-HCl buffer at 37°C (D), heat-inactivated LysB (LysB HI) (E), and various other hydrolases (F) after 60 min incubation. See Figure S1 for time-course curves for panels (D–F). A.u., arbitrary units. Error bars represent the standard deviation of three replicates. Data are representative of three independent experiments.

Next, we sought to test whether an AGM-degrading enzyme could activate the FRET-AGM probes. The only enzyme identified to cleave the mycoloyl ester bonds in AGM is LysB, an esterase present in mycobacteriophages, i.e., bacteriophages that specifically infect mycobacterial host cells [13]. The AGM breakdown activity of LysB is thought to be required for mycobacteriophages to breach the MM and facilitate completion of host cell lysis. Because the structure and function of mycobacteriophage D29 LysB is relatively well characterized [13], we selected this enzyme for initial validation of our FRET-AGM probes. Therefore, recombinant His₆-tagged D29 LysB and its corresponding inactive mutant lacking the catalytic serine-82 residue, LysB-S82A, were overexpressed in Escherichia coli BL21 DE3 using reported expression constructs [13, 14]. The proteins were purified from E. coli lysate by sequential Ni²⁺ immobilized metal affinity chromatography and size exclusion chromatography. SDS-PAGE and Western blot analysis of the purified proteins confirmed their homogeneity, expected molecular weight (untagged LysB is 28.5 kDa), and His₆ tag presence (Figure 2B).

We tested the ability of LysB and LysB-S82A enzymes to hydrolyze the FRET-AGM probes and elicit a fluorescence response. After optimization (not shown), enzymatic reactions were conducted with 0.1 μM FRET-AGM probes and 20 μg/mL of purified enzyme in Tris-HCl buffer containing 1% dimethylsulfoxide (DMSO) to promote probe solubility. After mixing the reaction components, a microplate fluorescence reader (excitation/emission 485/535 nm) was used to continuously monitor fluorescence emission intensity. FRET-AGM probes exposed to wild-type LysB enzyme were efficiently activated in a time-dependent manner, whereas the no-enzyme controls showed little to no fluorescence increase over time (Figure 2C). Under these assay conditions, FRET-AGM-Mono had a signal-to-background ratio of approximately 13:1, whereas that of FRET-AGM-Di was 6:1. The lower sensitivity of the FRET-AGM-Di probe was likely due to its lower FRET quenching efficiency in the “off” state, as noted above (Figure 2A). On the other hand, FRET-AGM-Di appeared to be more rapidly activated by LysB than FRET-AGM-Mono, suggesting that LysB may prefer the substrate with an additional Araf residue that more closely resembles the native AGM structure. The catalytically inactive mutant LysB-S82A failed to activate the probes, as did heat-denatured wild-type LysB (Figures 2D,E and S1). Together, these data demonstrate that FRET-AGM analogs are fluorogenic probes that are turned on through LysB activity.

We assessed the specificity of FRET-AGM probes for MM-degrading enzymes by testing against a focused panel of hydrolases. The enzyme panel included LysB and Tdmh, as well as three commercially available lipases (porcine pancreas lipase, Candida rugosa lipase, Burkholderia cepacia lipase), a protease (trypsin), and a glycoside hydrolase (porcine trehalase). Both LysB and Tdmh activated FRET-AGM probes, whereas the other hydrolases did not exhibit significant activity with the exception of low activity by porcine lipase on FRET-AGM-Mono, indicating that the probes have good specificity for hydrolases that act on MM glycolipids (Figures 2F and S1). The observation that Tdmh hydrolyzes FRET-AGM probes is unexpected given the prior report of preferential hydrolysis of TDM by Tdmh, with relatively less activity on purified mAGP [10]. This inconsistency could possibly be due to simplification of the FRET probe structures relative to the native structure of AGM (Figure 1B), as well as differences in the assay design and/or sensitivity between the two studies. Further experimentation will be required to resolve the discrepancy in detail. Regardless, FRET-AGM probes are shown here to selectively report on MM-degrading hydrolases.

We next used liquid chromatography–mass spectrometry (LC–MS) to characterize the reaction associated with LysB-mediated FRET-AGM activation. FRET-AGM probes were incubated in the presence (or absence) of LysB or LysB-S82A, and then reactions were quenched and analyzed by LC–MS. Combined extracted-ion chromatograms (EIC) were prepared that detect multiple relevant ions, including those from the unreacted probes and expected reaction products shown in Figure 3A. In the no-enzyme control samples, the major peaks correspond to unreacted FRET-AGM probes, whereas the LysB-treated samples show reaction product peaks resulting from ester cleavage, along with reduced intensity of the FRET-AGM reactant peaks (Figure 3B,C). Furthermore, the samples containing FRET-AGM probes incubated with the LysB-S82A mutant resulted in EICs identical to the no-enzyme controls. Based on these data, we conclude that FRET-AGM activation proceeds through LysB-dependent hydrolysis of the 5-O-mycoloyl-mimicking chain.

FIGURE 3 |.

(A) LysB activatation of FRET-AGM probes via ester hydrolysis. LC–MS combined EICs are shown for 10 μM FRET-AGM-Mono (B) and FRET-AGM-Di (C) incubated in the absence (top) or presence of 20 μg/mL LysB (middle) or LysB-S82A (bottom) in Tris-HCl buffer at 37°C for 24 h.

2.5 |. Insights into LysB-Mediated Probe Activation by Structural Modeling

To provide insights into the structural basis of FRET-AGM activation by LysB, we performed biomolecular interaction modeling using Boltz-1, an open-source deep learning model for predicting 3D structures of biomolecular complexes [40]. The Boltz-1-predicted structure of LysB was identical to the reported structure of unliganded LysB determined by X-ray crystallography (PDB: 3HC7) [13]. To model the interactions between LysB and FRET-AGM probes, we used Boltz-1 to predict the structures of LysB bound to simplified AGM substructures resembling the blue-highlighted fragment in Figure 1B. AGM mono- and disaccharide fragments with 5-O-octanoyl groups (AGM-Mono-C8 and AGM-Di-C8, structures shown in Figure 4A) were both found in the LysB active site pocket with similar orientations (Figure 4B). The ligands’ ester functional groups were near the LysB Asp-His-Ser catalytic triad, with the electrophilic carbonyl carbon atoms poised for attack by the nucleophilic Ser-82 hydroxy group located ~2.5 Å away (Figure 4C). The sugar residue(s) of the AGM fragments were engaged in interactions with polar amino acid residues within the active site and the lipid tail region proximal to the ester was accommodated by a hydrophobic groove adjacent to the catalytic triad (Figure 4D). The overall substrate location and orientation within the active site is consistent with the Ser-82-dependent FRET-AGM probe ester cleavage and fluorescence activation observed in the experiments described above. Together, these molecular interaction predictions provide a simplified model to help understand how LysB binds and catalyzes hydrolysis of AGM and AGM-mimicking probes.

FIGURE 4 |.

Modeling of interactions between LysB and AGM fragments. (A) Structures of AGM fragments used in modeling experiments. (B) Structures of LysB–AGM fragment complexes predicted by Boltz-1. (C) Interactions between AGM fragments (purple) and Asp-His-Ser catalytic triad in the LysB active site, with distance between Ser OH and AGM carbonyl groups indicated. (D) Interactions between AGM fragments (purple) and nonpolar residues (gray), polar uncharged residues (green), and polar charged residues (blue).

3 |. Conclusions

In this study, we developed chemical probes designed to study enzymes that break down AGM, an abundant mycobacterial glycolipid that contributes to the high impermeability of the mycobacterial cell envelope. Based on the prior success of developing a FRET-based chemical probe to study TDM degradation, our approach here was to design fluorescence-quenched structural analogs of AGM fragments that would undergo fluorescence activation upon exposure to AGM-degrading esterases. We developed a synthetic strategy that first established an orthogonally bifunctionalized Araf mono- or disaccharide core and subsequently installed the fluorophore and quencher pair at appropriate positions. PMB hydroxyl group protection was used in the synthesis to allow mild deprotection conditions that maintained sensitive functional groups. The synthesized mono- and disaccharide FRET-AGM probes were evaluated in proof-of-concept experiments with LysB, an AGM-degrading hydrolase from mycobacteriophage D29. It was demonstrated that both probes underwent LysB-dependent hydrolysis of their 5-O-mycoloyl-mimicking acyl chain, which led to fluorescence turn-on.

The findings of this study and the availability of FRET-AGM probes open various new research avenues. First, we are interested in further exploring the specificity of probe activation, both from the probe and enzyme perspectives. As described above, AGM is an enormously complex macromolecule, so the design of AGM analog probes necessitated structural simplifications. Our first-generation probes are based on mono- and di-Araf scaffolds with relatively easy-to-forge α-configured and 1,5-linked glycosidic bonds, as well as linear/short-chain acyl groups rather than branched/long-chain mycoloyl groups. Incorporating broader structural diversity into the Araf core (e.g., the branched pentasaccharide motifs of the AG terminus) and the fatty acyl chain (e.g., longer chains, α- and β-functionalities) within a larger panel of FRET probes could be a fruitful—albeit synthetically challenging—way to enable robust structure–activity relationship (SAR) investigations of AGM-degrading enzymes like LysB. Such a panel of structurally diverse substrates will also offer new insights into the distinct substrate preferences of LysB and Tdmh. For example, recent research on clickable TMM and free MA analogs suggests that building native-like features into mycolate probes can enhance specificity [41, 42]. Robust chemistries to access MA and AG/AGM fragments can be used in the service of such efforts. On the enzyme side, our study was limited to testing FRET-AGM probes with mycobacteriophage D29 LysB and mycobacterial Tdmh, two known mycolate glycolipid-degrading hydrolases, along with a small set of additional hydrolases that were commercially sourced. Testing additional hydrolases will provide richer information about the probes’ specificities.

We conclude by commenting on some of the possible future applications of FRET-AGM probes. First, FRET-AGM may facilitate the discovery of novel AGM-degrading enzymes. Due to their fluorogenic nature, FRET-AGM probes can potentially be employed in screening of cell lysates, enzyme libraries, or bacterial mutant libraries to identify candidate proteins with AGM hydrolytic activity. Of particular interest is to identify endogenous mycobacterial AGM degraders, which could play roles in MM remodeling and/or turnover. Second, FRET-AGM can be used to characterize the activity of purified AGM hydrolases through fluorescence assays. As mentioned above, recently published data suggest that mycobacteriophage D29 LysB has therapeutic potential for treating mycobacterial infections. Thus, identifying or evolving novel LysB enzymes with higher AGM-degrading activity is of interest, and FRET probes would be valuable tools in this regard. Finally, FRET-AGM and FRET-TDM probes could be used to screen for and characterize small-molecule inhibitors of enzymes involved in MM turnover. More broadly, a versatile set of fluorescence-quenched AGM and TDM glycolipid analog probes is now available to investigate different aspects of MM remodeling and degradation.

4 |. Experimental

4.1 |. 1,2,3-Tri-O-acetyl-5-O-(tert-butyldimethylsilyl)-D-arabinofuranoside (2)

Compound 2 was synthesized from D-arabinose as previously reported, with matching NMR spectra [34].

4.2 |. 2,3-Di-O-acetyl-5-O-(tert-butyldimethylsilyl)-D-arabinofuranose (3)

To a solution of compound 2 (21.2 g, 54.3 mmol) in anhydrous tetrahydrofuran (THF) (120 mL) stirring under argon at room temperature was added benzylamine (8.0 mL, 73 mmol). The reaction mixture was stirred under argon at room temperature for 4 h, after which the reaction was quenched with 1 M aqueous HCl and solvent was removed via rotary evaporation. The residue was dissolved in dichloromethane, transferred to a separatory funnel, and washed twice with water. The organic layer was collected and dried over MgSO₄, filtered, and dried via rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 3 as a 1:2 α:β mixture (10.3 g, 54% yield) as a colorless syrup. TLC (hexanes/ethyl acetate 2:1) R_f = 0.41.¹H NMR (500 MHz, CDCl₃, peaks from major β-anomer are listed): δ 5.43 (dd, J = 5.0, 11.5 Hz, 1 H, H-1), 5.27 (dd, J = 3.5, 5.0 Hz, 1 H, H-3), 5.09–5.07 (m, 1 H, H-2), 4.65 (d, J = 11.5 Hz, 1 H, 1-OH), 4.04–4.02 (q, J = 3.0 Hz, 1 H, H-4), 3.95 (dd, J = 2.0, 11.0 Hz, 1 H, H-5), 3.74 (dd, J = 2.0, 10.5 Hz, 1 H, H-5), 2.08–2.06 (m, 6 H, acetyl CH₃), 0.94 (s, 9 H, TBDMS tBu), 0.17 (s, 3 H, TBDMS CH₃), 0.14 (s, 3 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃, peaks from major β-anomer are listed): δ 171.2, 170.3, 95.3, 83.3, 79.1, 77.0, 64.2, 26.0, 21.0, 20.8, 18.5. HRMS (ESI-TOF) m/z: [M + Na]⁺ Calcd for C₁₅H₂₈O₇SiNa 371.1502; Found 371.1513.

4.3 |. 2,3-Di-O-acetyl-5-(tert-butyldimethylsilyl)-1-O-(8-chlorooctyl)-α-D-arabinofuranoside (5)

To a solution of 3 (5.01 g, 14.4 mmol) in anhydrous dichloromethane (50 mL) stirring at 0°C under argon, trichloroacetonitrile (1.77 mL, 17.7 mmol) and catalytic 1,8-diazabicyclo [5.4.0] undec-7-ene (DBU) (430 μL, 2.87 mmol) were added. The reaction was stirred at 0°C under argon for 30 min, after which solvent was removed via rotary evaporation. The residue was purified by silica gel column chromatography (hexanes/ethyl acetate 9:1 containing 1% triethylamine) to obtain trichloroacetimidate intermediate 4, which was used immediately in the subsequent glycosylation reaction. Compound 4 was coevaporated with toluene 2 times to remove residual water, and then combined with 8-chlorooctanol (2.12 mL, 12.7 mmol) and 4 Å powdered molecular sieves and dried for 30 min under high vacuum. The mixture was reconstituted in anhydrous dichloromethane (40 mL) and stirred under argon at 0°C for 30 min, after which catalytic TMSOTf (260 μL, 1.43 mmol) was added. The reaction was stirred under argon at 0°C with gradual warming to room temperature overnight. The reaction was quenched by addition of excess triethylamine, filtered through Celite, and dried by rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 5 (4.10 g, 58% over two steps) as a colorless syrup. TLC (hexanes/ethyl acetate 2:1) R_f = 0.85. ¹H NMR (500 MHz, CDCl₃): δ 5.07 (dd, J = 1.0, 4.0 Hz, 1 H, H-3), 5.03 (d, J = 1.5 Hz, 1 H, H-2), 4.98 (s, 1 H, H-1), 4.07 (q, J = 5.0 Hz, 1 H, H-4), 3.83 (d, J = 4.5 Hz, 2 H, H-5), 3.67 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.52 (t, J = 7.0 Hz, 2 H, linker CH₂Cl), 3.43 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 2.07 (s, 6 H, acetyl CH₃), 1.76 (pent, J = 7.0 Hz, 2 H, linker CH₂CH₂Cl), 1.62–1.56 (m, 2 H, linker OCH₂CH₂), 1.45–1.39 (m, 2 H, linker CH₂CH₂CH₂Cl), 1.37–1.29 (m, 6 H, linker CH₂), 0.90 (s, 9 H, TBDMS tBu), 0.08 (s, 3 H, TBDMS CH₃), 0.07 (s, 3 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃): δ 170.5, 170.2, 105.8, 83.0, 82.3, 67.7, 63.0, 45.4, 32.9, 29.6, 29.5, 29.1, 27.1, 26.2, 26.1, 21.18, 21.16, 18.6, −5.00, −5.06. HRMS (ESI-TOF) m/z: [M + Na]⁺ Calcd for C₂₃H₄₃ClO₇SiClNa 517.2365; Found 517.2387.

4.4 |. 2,3-Di-O-acetyl-1-O-(8-azideoctyl)-5-(tert-butyldimethylsilyl)-α-D-arabinofuranoside (6)

To a solution of compound 5 (3.63 g, 7.33 mmol) stirring in anhydrous N,N-dimethylformamide (DMF) (30 mL) at room temperature under argon was added NaN₃ (2.38 g, 36.6 mmol). The reaction mixture was stirred at 60°C for 36 h, after which ¹H NMR confirmed complete conversion of starting material into a single major product. The reaction mixture was filtered and concentrated via rotary evaporation, and then resuspended in brine and extracted with ethyl acetate 2 times. The combined organic layers were concentrated via rotary evaporation, reconstituted in dichloromethane, and washed twice with water. The organic layer was collected and dried over MgSO₄, filtered, and dried via rotary evaporation to obtain compound 6 (3.24 g, 88%) as a pale-yellow syrup. TLC (hexanes/ethyl acetate 2:1) R_f = 0.85. ¹H NMR (500 MHz, CDCl₃): δ 5.07 (d, J = 5.0 Hz, 1 H, H-3), 5.03 (s, 1 H, H-2), 4.97 (s, 1 H, H-1), 4.07 (q, J = 4.0 Hz, 1 H, H-4), 3.83 (d, J = 4.0 Hz, 2 H, H-5), 3.66 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.43 (dt, J = 6.0, 10.0 Hz, 1 H, linker OCH₂), 3.25 (t, J = 6.5 Hz, 2 H, linker CH₂N₃), 2.07 (s, 6 H, acetyl CH₃), 1.61–1.56 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.37–1.29 (m, 8 H, linker CH₂), 0.90 (s, 9 H, TBDMS tBu), 0.08 (s, 3 H, TBDMS CH₃), 0.07 (s, 3 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃): δ 170.5, 170.2, 105.8, 83.0, 82.3, 77.5, 67.7, 63.0, 51.7, 29.6, 29.5, 29.4, 29.1, 26.9, 26.2, 26.1, 21.1, 18.6, −5.00, −5.07. HRMS (ESI-TOF) m/z: [M + Na]⁺ Calcd for C₂₃H₄₃N₃O₇SiNa 524.2768; Found 524.2801.

4.5 |. 1-O-(8-Azideoctyl)-5-(tert-butyldimethylsilyl)-α-D-arabinofuranoside (7)

To a solution of compound 6 (1.60 g, 3.19 mmol) in anhydrous methanol (10 mL) stirring at room temperature under argon was added a freshly prepared solution of NaOMe in methanol (10 mL, 0.06 M). The reaction mixture was stirred at room temperature for 45 min, after which the reaction was neutralized with Dowex 50WX8–400 H⁺ ion-exchange resin. Following filtration of the resin, the mixture was concentrated via rotary evaporation, resuspended in water, and extracted twice with dichloromethane. The combined organic layers were dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was passed through a silica gel plug (methanol eluent) to obtain compound 7 (1.26 g, 96%) as a pale-yellow syrup. TLC (hexanes/ethyl acetate 2:1) R_f = 0.42. ¹H NMR (500 MHz, CDCl₃): δ 4.98 (s, 1 H, H-1), 4.19 (d, J = 12.0 Hz, 1 H, 3-OH), 4.15 (d, J = 1.5 Hz, 1 H, H-4), 3.97 (d, J = 12.0 Hz, 1 H, H-2), 3.93 (d, J = 11.5 Hz, 1 H, H-3), 3.85 (dd, J = 2.0, 11.0 Hz, 1 H, H-5), 3.82 (dd, J = 1.5, 11.0 Hz, 1 H, H-5), 3.72 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.43 (dt, J = 6.0, 10.0 Hz, 1 H, linker OCH₂), 3.25 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 3.03 (d, J = 12.0 Hz, 1 H, 2-OH), 1.61–1.54 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.29 (m, 8 H, linker CH₂), 0.90 (s, 9 H, TBDMS tBu), 0.12 (s, 3 H, TBDMS CH₃), 0.11 (s, 3 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃): δ 108.6, 88.1, 78.3, 78.2, 67.8, 53.6, 51.7, 29.7, 29.4, 29.3, 29.1, 26.9, 26.3, 26.1, 18.7, −5.2, −5.3. HRMS (ESI-TOF) m/z: [M + Na]⁺ Calcd for C₁₉H₃₉N₃O₅SiNa 440.2557; Found 440.2556.

4.6 |. 1-O-(8-Azideoctyl)-5-(tert-butyldimethylsilyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (8)

To a solution of compound 7 (0.866 g, 2.07 mmol) in anhydrous DMF (30 mL) stirring at 0°C under argon was added dropwise para-methoxybenzyl chloride (PMBCl) (0.70 mL, 5.2 mmol), followed by addition of NaH (60% dispersion in mineral oil, 0.185 g, 4.62 mmol). The reaction mixture was stirred for 6 h, initially at 0°C with gradual warming to room temperature, after which the mixture was diluted with ethyl acetate, washed with brine two times, dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 9:1) to obtain compound 8 (697 mg, 51%) as a colorless syrup. TLC (hexanes/ethyl acetate 9:1) R_f = 0.40. ¹H NMR (500 MHz, CDCl₃): δ 7.25 (d, J = 9.0 Hz, 2 H, PMB ArH), 7.22 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.88 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.85 (d, J = 9.0 Hz, 2 H, PMB ArH), 4.97 (s, 1 H, H-1), 4.50 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.49 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.48 (d, J = 11.0 Hz, 1 H, PMB CH₂), 4.43 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.03 (q, J = 4.5 Hz, 1 H, H-4), 3.98 (d, J = 3.0 Hz, 1 H, H-2), 3.88 (dd, J = 3.0, 6.5 Hz, 1 H, H-3), 3.80 (s, 3 H, PMB OCH₃), 3.79 (s, 3 H, PMB OCH₃), 3.74–3.71 (m, 2 H, H-5), 3.69 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.39 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 1.62–1.55 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.30 (m, 8 H, linker CH₂), 0.89 (s, 9 H, TBDMS tBu), 0.05 (s, 6 H, TBDMS CH₃s). ¹³C NMR (125.7 MHz, CDCl₃): δ 159.6, 159.5, 130.5, 130.1, 129.8, 129.7, 114.1, 114.0, 106.4, 88.6, 83.1, 82.3, 71.9, 71.8, 67.8, 63.5, 55.6, 55.5, 51.7, 29.8, 29.5, 29.4, 29.1, 27.0, 26.3, 26.2, 18.7, −4.90, −4.96. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₃₅H₅₅N₃O₇SiNH₄ 675.4153; Found 675.4163.

4.7 |. 1-O-(8-Azideoctyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (9)

To a solution of 8 (837 mg, 1.27 mmol) in anhydrous THF (12 mL) stirring at room temperature under argon was added tetrabutylammonium fluoride (TBAF, 1.0 M in THF) (1.45 mL, 1.45 mmol). After stirring at room temperature for 90 min, the reaction mixture was diluted with water and extracted twice with dichloromethane. The combined organic layers dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 9 (673 mg, 97%) as a colorless syrup. TLC (hexanes/ethyl acetate 3:1) R_f = 0.20. ¹H NMR (500 MHz, CDCl₃): δ 7.25 (d, J = 8.5 Hz, 2 H, PMB ArH), 7.21 (d, J = 8.0 Hz, 2 H, PMB ArH), 6.88 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.86 (d, J = 8.5 Hz, 2 H, PMB ArH), 4.99 (s, 1 H, H-1), 4.52 (d, J = 12.0 Hz, 1 H, PMB CH₂), 4.49 (d, J = 12.0 Hz, 1 H, PMB CH₂), 4.44 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.43 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.10–4.07 (m, 1 H, H-4), 3.99 (d, J = 3.0 Hz, 1 H, H-2), 3.94 (dd, J = 3.0, 6.5 Hz, 1 H, H-3), 3.83–3.79 (m, 1 H, H-5), 3.81 (s, 3 H, PMB OCH₃), 3.79 (s, 3 H, PMB OCH₃), 3.68 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.65–3.60 (m, 1 H, H-5), 3.39 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 1.85 (dd, J = 4.5, 7.0 Hz, 1 H, 5-OH), 1.61–1.56 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.29 (m, 8 H, linker CH₂). ¹³C NMR (125.7 MHz, CDCl₃): δ 159.7, 159.6, 130.2, 129.9, 129.8, 129.7, 114.2, 114.1, 106.6, 88.1, 82.7, 82.0, 72.2, 71.9, 67.9, 62.6, 55.63, 55.61, 51.7, 29.8, 29.5, 29.4, 27.0, 26.3. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₂₉H₄₁N₃O₇NH₄ 561.3288; Found 561.3299.

4.8 |. 1-O-(8-Azideoctyl)-5-O-(N-Fmoc-8-aminooctanoyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (10)

To a solution of N-Fmoc-8-aminooctanoic acid (254 mg, 0.667 mmol) in anhydrous dichloromethane (10 mL) stirring at 0°C under argon was added N,N’-dicyclohexylcarbodiimide (DCC) (137 mg, 0.667 mmol) to form the corresponding activated ester intermediate. After stirring for 30 min at 0°C under argon, the activated ester intermediate solution was filtered and transferred to a solution of compound 9 (240 mg, 0.441 mmol) in anhydrous dichloromethane (10 mL) stirring at 0°C under argon. N,N’-Dimethylaminopyridine (DMAP) (12 mg, 0.98 mmol) was added, and the reaction mixture was stirred under argon overnight with gradual warming from 0°C to room temperature. After filtration to remove urea byproduct, the filtrate was concentrated via rotary evaporation and purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 10 (372 mg, 93%) as a colorless syrup. TLC (hexanes/ethyl acetate 3:1) R_f = 0.57. ¹H NMR (500 MHz, CDCl₃): δ 7.76 (d, J = 7.5 Hz, 2 H, Fmoc ArH), 7.59 (d, J = 7.5 Hz, 2 H, Fmoc ArH), 7.39 (t, J = 7.0 Hz, 2 H, Fmoc ArH), 7.31 (dt, J = 1.0, 7.5 Hz, 2 H, Fmoc ArH), 7.25 (d, J = 8.5 Hz, 2 H, PMB ArH), 7.20 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.88 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.85 (d, J = 8.5 Hz, 2 H, PMB ArH), 4.99 (s, 1 H, H-1), 4.78–4.74 (m, 1 H, carbamate NH), 4.51 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.50 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.43 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.43–4.38 (m, 3 H, PMB CH₂, Fmoc CH₂), 4.28–4.25 (m, 1 H, H-5), 4.21 (t, J = 7.0 Hz, 1 H, Fmoc CH), 4.18–4.13 (m, 2 H, H-4, H-5), 3.99 (dd, J = 1.0, 3.0 Hz, 1 H, H-2), 3.82–3.78 (m, 1 H, H-3), 3.80 (s, 3 H, PMB OCH₃), 3.78 (s, 3 H, PMB OCH₃), 3.69 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.39 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 3.16 (q, J = 6.5 Hz, 2 H, octanoyl CH₂NHFmoc), 2.29 (t, J = 7.5 Hz, 2 H, octanoyl α-CH₂), 1.61–1.56 (m, 6 H, linker OCH₂CH₂, linker CH₂CH₂N₃, octanoyl α-CH₂CH₂), 1.50–1.44 (m, 2 H, octanoyl CH₂CH₂NHFmoc), 1.38–1.25 (m, 14 H, linker and octanoyl CH₂). ¹³C NMR (125.7 MHz, CDCl₃): δ 173.8, 159.7, 159.6, 156.7, 144.3, 141.6, 130.0, 129.9, 129.8, 129.7, 127.9, 127.3, 125.3, 120.2, 114.19, 114.13, 106.5, 88.3, 83.3, 79.0, 72.2, 72.1, 68.0, 66.7, 63.8, 55.6, 55.5, 51.7, 47.6, 41.3, 34.3, 30.2, 29.8, 29.5, 29.4, 29.3, 29.2, 29.1, 27.0, 26.8, 26.3, 25.0. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₅₂H₆₆N₄O₁₀NH₄ 924.5123; Found 924.5128.

4.9 |. 1-O-(8-Azideoctyl)-5-O-(8-[(fluorescein-5-yl)thioureido]octanoyl)-α-D-arabinofuranoside (12)

To a solution of compound 10 (112 mg, 0.123 mmol) in anhydrous dichloromethane (5 mL) stirring at room temperature was added 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) (42 mg, 0.165 mmol) and water (0.5 mL). The reaction mixture was stirred for 3 h at room temperature, after which it was diluted with ethyl acetate and washed successively with saturated sodium bicarbonate and brine, dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (dichloromethane/methanol 98:2) to obtain compound 11 (73 mg, 89%) as a colorless syrup. To a solution of compound 11 (102 mg, 0.153 mmol) in anhydrous DMF (10 mL) stirring at room temperature under argon was added piperidine (15 μL, 0.156 mmol). After stirring at room temperature for 4 h, the reaction mixture was concentrated via rotary evaporation. The crude intermediate was dissolved in anhydrous DMF (10 mL), and then potassium carbonate (37 mg, 0.267 mmol) and fluorescein isothiocyanate (FITC) (64 mg, 0.164 mmol) were added sequentially. After stirring at room temperature overnight, the reaction mixture was diluted with brine and extracted twice with ethyl acetate. The combined organic layers were dried over MgSO₄, filtered, and concentrated by rotary evaporation. The crude material was purified by silica gel column chromatography (dichloromethane/methanol 9:1) to obtain compound 12 (27 mg, 42% over two steps) as an orange-yellow syrup. TLC (dichloromethane/methanol 9:1) R_f = 0.40. ¹H NMR (500 MHz, CD₃OD with 10% CDCl₃): δ 8.17 (s, 1 H, FITC ArH), 7.82 (dd, J = 1.0, 8.0 Hz, 1 H, FITC ArH), 7.15 (d, J = 8.5 Hz, 1 H, FITC ArH), 6.73–6.69 (m, 4 H, FITC ArH), 6.57–6.54 (m, 2 H, FITC ArH), 4.88 (d, J = 0.5 Hz, 1 H, H-1), 4.34 (dd, J = 3.0, 11.5 Hz, 1 H, H-5), 4.19 (dd, J = 6.0, 11.5 Hz, 1 H, H-5), 4.09–4.06 (m, 1 H, H-4), 4.01 (dd, J = 1.5, 3.5 Hz, 1 H, H-2), 3.83 (dd, J = 4.0, 6.5 Hz, 1 H, H-3), 3.71 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.65–3.57 (s, 2 H, octanoyl CH₂-FITC), 3.44 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.27 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 2.39 (t, J = 7.5 Hz, 2 H, octanoyl α-CH₂), 1.72–1.63 (m, 4 H, octanoyl α-CH₂CH₂, linker CH₂CH₂-FITC), 1.63–1.56 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.44–1.28 (m, 14 H, linker and octanoyl CH₂). ¹³C NMR (125.7 MHz, CD₃OD with 10% CDCl₃): δ 175.9, 171.9, 154.9, 142.8, 131.8, 130.9, 126.4, 120.6, 114.6, 112.2, 109.8, 104.2, 83.8, 82.6, 79.6, 69.6, 65.7, 53.0, 46.2, 35.5, 31.1, 30.8, 30.7, 30.5, 30.44, 30.41, 28.3, 28.2, 27.6, 26.4. HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₄₂H₅₂N₅O₁₁S 834.3384; Found 834.3398.

4.10 |. FRET-AGM-Mono (14)

To a solution of compound 12 (56.4 mg, 0.0676 mmol) in THF/methanol/water 1:1:1 (6 mL) stirring at room temperature was added alkyne dabcyl reagent 13 [39] (20.8 mg, 0.0678 mmol), CuSO₄··5H₂O (32.0 mg, 0.128 mmol), and sodium ascorbate (23.9 mg, 0.120 mmol). The reaction mixture was stirred at 40°C for 48 h and then concentrated via rotary evaporation. The crude material was purified on a Biotage Isolera One automated flash chromatography system (10 g C18 column; 5% CH₃CN in H₂O → 95% CH₃CN in H₂O) to give FRET-AGM-Mono 13 (41.8 mg, 54%) as a reddish-orange solid. TLC (dichloromethane/methanol 19:1) R_f = 0.50. ¹H NMR (500 MHz, CD₃OD with 10% CDCl₃): δ 8.00 (dd, J = 1.0, 7.0 Hz, 1 H, FITC ArH), 7.95 (d, J = 8.5 Hz, 2 H, dabcyl ArH), 7.86 (d, J = 9.0 Hz, 2 H, dabcyl ArH), 7.84 (d, J = 9.0 Hz, 2 H, dabcyl ArH), 7.80–7.76 (m, 2 H, FITC ArH, triazole C5-CH), 7.40 (t, J = 7.5 Hz, 1 H, FITC ArH), 7.07 (d, J = 8.0 Hz, 1 H, FITC ArH), 6.84–6.76 (m, 3 H, FITC ArH and dabcyl ArH), 6.69 (d, J = 2.0 Hz, 2 H, FITC ArH), 6.56 (dd, J = 1.5, 8.5 Hz, 2 H, FITC ArH), 4.89 (d, J = 1.5 Hz, 1 H, H-1), 4.65 (s, 2 H, triazole C4-CH₂-NH), 4.35 (t, J = 7.0 Hz, 2 H, octanoyl CH₂-FITC), 4.33 (dd, J = 3.0, 9.5 Hz, 1 H, H-5), 4.18 (dd, J = 6.0, 11.5 Hz, 1 H, H-5), 4.09 (dt, J = 3.0, 6.0 Hz, 1 H, H-4), 4.01 (dd, J = 2.0, 4.0 Hz, 1 H, H-2), 3.83 (dd, J = 4.0, 6.5 Hz, 1 H, H-3), 3.70 (dt, J = 6.5, 10.0 Hz, 1 H, linker OCH₂), 3.42 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.23 (t, J = 7.0 Hz, 2 H, linker CH₂-triazole N1), 3.12 (s, 6 H, dabcyl N-(CH₃)₂), 2.37 (t, J = 7.5 Hz, 2 H, octanoyl α-CH₂), 1.90 (pent, J = 6.5 Hz, 2 H, octanoyl CH₂CH₂-FITC), 1.64 (pent, J = 6.5 Hz, 2 H, octa-noyl α-CH₂CH₂), 1.60–1.55 (m, 4 H, linker CH₂CH₂-triazole N1 and linker OCH₂CH₂), 1.41–1.28 (m, 14 H, linker and octanoyl CH₂s). ¹³C NMR (125.7 MHz, CD₃OD with 10% CDCl₃): δ 175.7, 169.6, 168.4, 157.8, 156.6, 154.6, 144.9, 143.0, 135.3, 134.5, 131.2, 130.9, 129.7, 129.6, 126.8, 124.5, 123.3, 112.9, 109.4, 104.1, 83.3, 82.4, 69.4, 51.8, 41.29, 41.13, 36.5, 35.4, 33.3, 31.5, 31.2, 31.0, 30.7, 30.4, 30.3, 30.2, 28.0, 27.6, 27.2, 26.1. HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₆₀H₇₀N₉O₁₃ 1124.5093; Found 1124.5034.

4.11 |. 2,3-Di-O-acetyl-5-O-(tert-butyldimethylsilyl)-α-D-arabinofuranosyl-(1→5)-1-O-(8-azideoctyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (15)

To a solution of 3 (1.50 g, 4.31 mmol) in anhydrous dichloromethane (20 mL) stirring at 0°C under argon, trichloroacetonitrile (0.53 mL, 5.3 mmol) and catalytic DBU (130 μL, 0.88 mmol) were added. The reaction was stirred at 0°C under argon for 30 min, after which solvent was removed via rotary evaporation. The residue was purified by silica gel column chromatography (hexanes/ethyl acetate 9:1 containing 1% triethylamine) to obtain a trichloroacetimidate intermediate 4, which was used immediately in the subsequent glycosylation reaction. Compound 4 was coevaporated with toluene two times to remove residual water, and then combined with compound 9 (1.05 g, 1.93 mmol) and 3 Å powdered molecular sieves and dried for 30 min under high vacuum. The mixture reconstituted in anhydrous dichloromethane (40 mL) and stirred under argon at 0°C for 30 min, after which catalytic TMSOTf (69 μL, 0.38 mmol) was added. The reaction was stirred under argon at 0°C with gradual warming to room temperature overnight. The reaction was quenched by addition of excess triethylamine, filtered through Celite, and dried by rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 15 (1.27 g, 75%) as a colorless syrup. TLC (hexanes/ethyl acetate 2:1) R_f = 0.75. ¹H NMR (500 MHz, CDCl₃): δ 7.25 (d, J = 8.0 Hz, 2 H, PMB ArH), 7.20 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.87 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.84 (d, J = 8.5 Hz, 2 H, PMB ArH), 5.12 (s, 1 H, Araf-II H-3), 5.07–5.05 (m, 2 H, Araf-II H-1, Araf-II H-2), 4.99 (s, 1 H, Araf-I H-1), 4.507 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.501 (d, J = 11.0 Hz, 1 H, PMB CH₂), 4.43 (d, J = 11.0 Hz, 1 H, PMB CH₂), 4.42 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.14–4.11 (m, 1 H, Araf-I H-4), 4.00–3.97 (m, 2 H, Araf-I H-2, Araf-II H-4), 3.94 (dd, J = 3.5, 7.0 Hz, 1 H, Araf-I H-3), 3.83 (dd, J = 4.5, 11.0 Hz, 1 H, Araf-I H-5), 3.81–3.78 (m, 8 H, PMB OCH₃, Araf-II H-5), 3.69 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.62 (dd, J = 3.5, 11.0 Hz, 1 H, Araf-I H-5), 3.39 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 2.06 (s, 3 H, acetyl CH₃), 1.98 (s, 3 H, acetyl CH₃), 1.62–1.55 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.29 (m, 8 H, linker CH₂), 0.89 (s, 9 H, TBDMS tBu), 0.07 (s, 3 H, TBDMS CH₃), 0.06 (s, 3 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃): δ 170.4, 169.9, 159.7, 159.5, 130.3, 130.0, 129.9, 129.6, 114.14, 114.10, 106.4, 105.7, 88.6, 83.4, 83.2, 81.9, 80.0, 72.1, 72.0, 67.9, 66.2, 62.8, 55.6, 55.5, 51.7, 29.8, 29.5, 29.4, 29.1, 27.0, 26.3, 26.1, 21.1, 21.0, 18.6, −5.00, −5.08. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₄₄H₆₇N₃O₁₃SiNH₄ 891.4787; Found 891.4771.

4.12 |. 5-O-(tert-butyldimethylsilyl)-α-D-arabinofuranosyl-(1→5)-1-O-(8-azideoctyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (16)

To a solution of compound 15 (1.15 g, 1.32 mmol) in anhydrous methanol (30 mL) stirring at room temperature under argon was added a freshly prepared solution of NaOMe in methanol (32 mL, 0.06 M). The reaction mixture was stirred at room temperature for 30 min, after which the reaction was neutralized with Dowex 50WX8–400 H⁺ ion-exchange resin. Following filtration of the resin, the mixture was concentrated via rotary evaporation, resuspended in water, and extracted twice with dichloromethane. The combined organic layers were dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 2:1) to obtain compound 16 (0.690 g, 66%) as a colorless syrup. TLC (hexanes/ethyl acetate 2:1) R_f = 0.20. ¹H NMR (500 MHz, CDCl₃): δ 7.27 (d, J = 8.5 Hz, 2 H, PMB ArH), 7.18 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.89 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.85 (d, J = 8.5 Hz, 2 H, PMB ArH), 4.977 (s, 1 H, Araf-I H-1), 4.972 (s, 1 H, Araf-II H-1), 4.52 (d, J = 12.0 Hz, 1 H, PMB CH₂), 4.49 (d, J = 12.0 Hz, 1 H, PMB CH₂), 4.40 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.29 (d, J = 12.0 Hz, 1 H, PMB CH₂), 4.13–4.10 (m, 1 H, Araf-I H-4), 4.04 (d, J = 12.0 Hz, 1 H, Araf-II 3-OH), 3.95 (d, J = 12.0 Hz, 1 H, Araf-II H-3), 3.94 (s, 1 H, Araf-I H-2), 3.91 (d, J = 12.5 Hz, Araf-II H-2), 3.82–3.75 (m, 8 H, PMB OCH₃, Araf-II H-4), 3.77 (dd, J = 5.0, 10.0 Hz, 1 H, Araf-I H-5), 3.75–3.73 (m, 2 H, Araf-II H-5), 3.71 (dd, J = 2.0, 6.0 Hz, 1 H, Araf-I H-3), 3.66–3.60 (m, 2 H, linker OCH₂, Araf-I H-5), 3.37 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.32 (d, J = 12.5 Hz, 1 H, Araf-II 2-OH), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 1.61–1.53 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.29 (m, 8 H, linker CH₂), 0.90 (s, 9 H, TBDMS tBu), 0.114 (s, 3 H, TBDMS CH₃), 0.110 (s, 3 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃): δ 159.8, 159.6, 130.1, 129.9, 129.8, 129.7, 114.2, 114.1, 107.8, 106.3, 87.9, 87.2, 83.6, 80.6, 78.3, 78.1, 71.9, 71.7, 67.8, 66.0, 63.5, 55.6, 55.5, 51.7, 29.7, 29.5, 29.4, 29.1, 27.0, 26.3, 26.1, 18.7, −5.2, −5.3. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₄₀H₆₃N₃O₁₁SiNa 812.4130; Found 812.4129.

4.13 |. 5-O-(tert-Butyldimethylsilyl)–2,3-di-(p-methoxybenzyl)-α-D-arabinofuranosyl-(1→5)-1-O-(8-azideoctyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (17)

To a solution of compound 16 (0.513 g, 0.649 mmol) in anhydrous DMF (10 mL) stirring at 0°C under argon was added dropwise PMBCl (263 μL, 1.95 mmol), followed by addition of NaH (60% dispersion in mineral oil, 65 mg, 1.62 mmol). The reaction mixture was stirred for 6 h, initially at 0°C with gradual warming to room temperature, after which the mixture was concentrated by rotary evaporation. The crude material was resuspended in water and extracted twice with dichloromethane. The combined organic layers were dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 9:1) to obtain compound 17 (0.401 mg, 60%) as a colorless syrup. TLC (hexanes/ethyl acetate 9:1) R_f = 0.37. ¹H NMR (500 MHz, CDCl₃): δ 7.24–7.22 (m, 4 H, PMB ArH), 7.19 (d, J = 8.5 Hz, 2 H, PMB ArH), 7.19 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.85–6.81 (m, 8H, PMB ArH), 5.11 (d, J = 1.0 Hz, 1 H, Araf-II H-1), 4.98 (d, J = 1.0 Hz, 1 H, Araf-I H-1), 4.50–4.40 (m, 7 H, PMB CH₂), 4.37 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.15–4.10 (m, 1 H, Araf-I H-4), 4.04–4.00 (m, 3 H, Araf-II H-2, Araf-I H-3, Araf-II H-4), 3.98 (dd, J = 1.5, 4.0 Hz, 1 H, Araf-I H-2), 3.90 (dd, J = 3.5, 6.5 Hz, 1 H, Araf-II H-3), 3.85 (dd, J = 4.0, 11.5 Hz, 1 H, Araf-I H-5), 3.78 (s, 3 H, PMB OCH₃), 3.773, (s, 3 H, PMB OCH₃), 3.772 (s, 3 H, PMB OCH₃), 3.76 (s, 3 H, PMB OCH₃), 3.72 (dd, J = 2.5, 4.0 Hz, 2 H, Araf-II H-5), 3.70–3.66 (m, 2 H, Araf-I H-5, linker OCH₂), 3.37 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 1.61–1.54 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.29 (m, 8 H, linker CH₂), 0.88 (s, 9 H, TBDMS tBu), 0.04 (s, 6 H, TBDMS CH₃). ¹³C NMR (125.7 MHz, CDCl₃): δ 159.6, 159.59, 159.52, 159.4, 130.59, 130.53, 130.17, 130.14, 130.0, 129.9, 129.8, 129.7, 129.6, 128.6, 114.1, 114.09, 114.06, 114.03, 106.7, 106.4, 83.2, 83.1, 82.5, 80.3, 72.3, 72.0, 71.9, 71.7, 67.9, 66.2, 63.3, 60.7, 55.59, 55.56, 51.7, 30.3, 29.8, 29.5, 29.4, 29.1, 27.0, 26.3, 26.2, 18.7, −4.90, −4.97. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₅₆H₇₉N₃O₁₃SiNH₄ 1047.5726; Found 1047.5696.

4.14 |. 2,3-di-(p-methoxybenzyl)-α-D-arabinofuranosyl-(1→5)-1-O-(8-azideoctyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (18)

To a solution of 17 (350 mg, 0.339 mmol) in anhydrous THF (5 mL) stirring at room temperature under argon was added TBAF (1.0 M in THF) (408 μL, 0.408 mmol). After stirring at room temperature for 3 h, the reaction mixture was diluted with water and extracted twice with dichloromethane. The combined organic layers dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 18 (260 mg, 83%) as a colorless syrup. TLC (hexanes/ethyl acetate 3:1) R_f = 0.30. ¹H NMR (500 MHz, CDCl₃): δ 7.23 (d, J = 8.0 Hz, 4 H, PMB ArH), 7.19 (d, J = 9.0 Hz, 2 H, PMB ArH), 7.18 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.86–6.82 (m, 8 H, PMB ArH), 5.12 (s, 1 H, Araf-II H-1), 4.99 (s, 1 H, Araf-I H-1), 4.50–4.47 (m, 4 H, PMB CH₂), 4.44 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.43 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.39 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.38 (d, J = 11.5 Hz, 1 H, PMB CH₂), 4.12 (pent, J = 4.0 Hz, 1 H, Araf-I H-4), 4.08–4.04 (m, 2 H, Araf-II H-4, Araf-II H-2), 4.02 (dd, J = 3.5, 7.0 Hz, Araf-I H-3) 3.99 (dd, J = 1.0, 3.5 Hz, 1 H, Araf-I H-2), 3.93 (dd, J = 3.0, 6.5 Hz, 1 H, Araf-II H-3), 3.83 (dd, J = 4.0, 11.5 Hz, 1 H, Araf-I H-5), 3.80–3.78 (m, 1 H, Araf-II H-5), 3.80–3.76 (m, 12 H, OPMB OCH₃), 3.72–3.67 (m, 2 H, Araf-I H-5, linker OCH₂), 3.63–3.58 (m, 1 H, Araf-II H-5), 3.37 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.24 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 1.86 (s, 1 H, Araf-II 5-OH), 1.61–1.54 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.38–1.28 (m, 8 H, linker CH₂). ¹³C NMR (125.7 MHz, CDCl₃): δ 159.7, 159.6, 159.59, 159.54, 130.4, 130.2, 130.06, 130.01, 129.9, 129.8, 129.74, 129.70, 106.8, 106.4, 88.7, 87.7, 83.1, 82.7, 82.2, 80.3, 72.28, 72.22, 72.0, 71.8, 67.9, 66.1, 62.5, 55.5, 51.7, 29.8, 29.5, 29.4, 29.1, 26.9, 26.3. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₅₀H₆₅N₃O₁₃NH₄ 933.4861; Found 933.4858.

4.15 |. 5-O-(N-Fmoc-8-aminooctanoyl)–2,3-di-(p-methoxybenzyl)-α-D-arabinofuranosyl-(1→5)-1-O-(8-azideoctyl)–2,3-di-O-(p-methoxybenzyl)-α-D-arabinofuranoside (19)

To a solution of N-Fmoc-8-aminooctanoic acid (126 mg, 0.330 mmol) in anhydrous dichloromethane (10 mL) stirring at 0°C under argon was added N,N-dicyclohexylcarbodiimide (DCC) (68 mg, 0.33 mmol) to form the corresponding activated ester intermediate. After stirring for 30 min at 0°C under argon, the activated ester intermediate solution was filtered and transferred to a solution of compound 18 (201 mg, 0.219 mmol) in anhydrous dichloromethane (10 mL) stirring at 0°C under argon. N,N-Dimethylaminopyridine (DMAP) (8.9 mg, 0.072 mmol) was added, and the reaction mixture was stirred under argon overnight with gradual warming from 0°C to room temperature. After filtration to remove urea byproduct, the filtrate was concentrated via rotary evaporation and purified by silica gel column chromatography (hexanes/ethyl acetate 3:1) to obtain compound 19 (302 mg, 72%) as a colorless syrup. TLC (hexanes/ethyl acetate 1:1) R_f = 0.55. ¹H NMR (500 MHz, CDCl₃): δ 7.76 (d, J = 7.5 Hz, 2 H, Fmoc ArH), 7.59 (d, J = 7.0 Hz, 2 H, Fmoc ArH), 7.39 (t, J = 7.0 Hz, 2 H, Fmoc ArH), 7.31 (dt, J = 1.0, 7.5 Hz, 2 H, Fmoc ArH), 7.24–7.21 (m, 4 H, PMB ArH), 7.20 (d, J = 8.5 Hz, 2 H, PMB ArH), 7.17 (d, J = 8.5 Hz, 2 H, PMB ArH), 6.86–6.82 (m, 8 H, PMB ArH), 5.13 (s, 1 H, Araf-II H-1), 4.98 (s, 1 H, Araf-I H-1), 4.79–4.76 (m, 1 H, carbamate NH), 4.52–4.35 (m, 10 H, PMB CH₂ and Fmoc CH₂), 4.26–4.20 (m, 2 H, Araf-II H-5, Fmoc CH), 4.17–4.10 (m, 3 H, Araf-II H-5, Araf-II H-4, Araf-I H-4), 4.04 (d, J = 3.0 Hz, 1 H, Araf-II H-2), 4.02–3.99 (m, 2 H, Araf-I H-2, Araf-I H-3), 3.85 (dd, J = 4.0, 12.0 Hz, 1 H, Araf-I H-5), 3.80–3.78 (m, 1 H, Araf-II H-3), 3.78 (s, 3 H, PMB OCH₃), 3.77 (s, 6 H, PMB OCH₃), 3.76 (s, 3 H, PMB OCH₃), 3.71–3.66 (m, 2 H, Araf-I H-5, linker OCH₂), 3.37 (dt, J = 7.0, 9.5 Hz, 1 H, linker OCH₂), 3.23 (t, J = 7.0 Hz, 2 H, linker CH₂N₃), 3.16 (q, J = 6.5 Hz, 2 H, octanoyl CH₂NHFmoc), 2.25 (t, J = 7.5 Hz, 2 H, octanoyl α-CH₂), 1.60–1.55 (m, 6 H, linker OCH₂CH₂, linker CH₂CH₂N₃, and octanoyl α-CH₂CH₂), 1.48–1.45 (m, 2 H, octanoyl CH₂CH₂NHFmoc), 1.32–1.25 (m, 14 H, linker and octanoyl CH₂). ¹³C NMR (125.7 MHz, CDCl₃): δ 173.8, 159.69, 159.67, 159.61, 159.5, 156.7, 144.3, 141.6, 130.4, 130.05, 130.02, 129.9, 129.7, 127.9, 127.3, 125.3, 120.2, 114.15, 114.12, 114.10, 114.0, 106.9, 106.4, 88.79, 87.9, 83.3, 83.2, 80.2, 79.2, 72.3, 72.1, 72.05, 72.00, 67.9, 66.7, 66.3, 63.8, 60.7, 55.5, 51.7, 47.6, 41.3, 34.3, 32.2, 30.2, 30.0, 29.9, 29.8, 29.6, 29.5, 29.29, 29.22, 29.15, 27.0, 26.8, 26.3, 25.0. HRMS (ESI-TOF) m/z: [M + NH₄]⁺ Calcd for C₇₃H₉₀N₄O₁₆NH₄ 1296.6696; Found 1296.6643.

4.16 |. 5-O-(8-[(fluorescein-5-yl)thioureido]octanoyl)-α-D-arabinofuranosyl-(1 → 5)-1-O-(8-azideoctyl)-α-D-arabinofuranoside (21)

To a solution of compound 19 (301 mg, 0.235 mmol) in anhydrous dichloromethane (10 mL) stirring at room temperature, 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) (133 mg, 0.588 mmol) and water (2.0 mL) were added. The reaction mixture was stirred for 3 h at room temperature, after which it was diluted with ethyl acetate and washed successively with saturated sodium bicarbonate and brine, dried over MgSO₄, filtered, and concentrated via rotary evaporation. The crude material was purified by silica gel column chromatography (dichloromethane/methanol 19:1) to obtain compound 20 (177 mg, 94%) as a colorless syrup. To a solution of compound 20 (33 mg, 0.041 mmol) in anhydrous DMF (5 mL) stirring at room temperature under argon was added piperidine (4.2 μL, 0.042 mmol). After stirring at room temperature for 4 h, the reaction mixture was concentrated via rotary evaporation. The crude intermediate was dissolved in anhydrous DMF (5 mL), and then triethylamine (10 μL, 0.074 mmol) and FITC (17 mg, 0.044 mmol) were added sequentially. After stirring at room temperature overnight, the reaction mixture was diluted with brine and extracted twice with ethyl acetate. The combined organic layers were dried over MgSO₄, filtered, and concentrated by rotary evaporation. The crude material was purified by silica gel column chromatography (dichloromethane/methanol 9:1) to obtain compound 21 (27 mg, 69% over two steps) as an orange-yellow syrup. TLC (dichloromethane/methanol 9:1) R_f = 0.40. ¹H NMR (500 MHz, CD₃OD): δ 8.17 (s, 1 H, FITC ArH), 7.79 (d, J = 8.0 Hz, 1 H, FITC ArH), 7.18 (d, J = 8.0 Hz, 1 H, FITC ArH), 6.72–6.68 (m, 4 H, FITC ArH), 6.58–6.56 (m, 2 H, FITC ArH), 4.97 (s, 1 H, Araf-II H-1), 4.87 (s, 1 H, Araf-I H-1, observed in COSY spectrum overlapping with DHO peak), 4.34 (dd, J = 3.5, 12.0 Hz, 1 H, Araf-II H-5), 4.20 (dd, J = 6.5, 12.0 Hz, 1 H, Araf-II H-5), 4.14–4.11 (m, 1 H, Araf-II H-4), 4.05–4.02 (m, 2 H, Araf-II H-2, Ara-I H-4), 3.99–3.97 (m, 1 H, Araf-I H-2), 3.91 (dd, J = 4.0, 6.0 Hz, 1 H, Ara-I H-3), 3.87–3.84 (m, 2 H, Araf-II H-3, Araf-I H-5), 3.74–3.61 (m, 4 H, Araf-I H-5, linker OCH₂, octanoyl CH₂-FITC), 3.44 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.29 (t, J = 6.5 Hz, 2 H, linker CH₂N₃), 2.40 (t, J = 7.5 Hz, 2 H, octanoyl α-CH₂), 1.72–1.66 (m, 4 H, octanoyl α-CH₂CH₂, linker CH₂CH₂-FITC), 1.63–1.57 (m, 4 H, linker OCH₂CH₂, linker CH₂CH₂N₃), 1.45–1.33 (m, 14 H, linker and octanoyl CH₂). ¹³C NMR (125.7 MHz, CD₃OD): δ 175.9, 131.2, 110.5, 110.3, 104.3, 84.4, 84.3, 84.0, 83.6, 80.05, 80.01, 69.7, 69.1, 65.8, 53.3, 35.7, 35.6, 31.5, 31.2, 31.0, 30.8, 30.7, 28.68, 28.62, 27.9, 27.6. Additional ¹³C peaks observed in HSQC-TOCSY spectrum: δ 132.5, 126.5, 120.7, 114.6, 46.3. HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₄₇H₆₀N₅O₁₅S 966.3807; Found 966.3805.

4.17 |. FRET-AGM-Di (22)

To a solution of compound 21 (46.3 mg, 0.0476 mmol) in THF/methanol/water 1:1:1 (6 mL) stirring at room temperature was added alkyne dabcyl reagent 13 (14.5 mg, 0.0476 mmol), CuSO₄··5H₂O (23.0 mg, 0.0921 mmol), and sodium ascorbate (19.4 mg, 0.0950 mmol). The reaction mixture was stirred at 40°C for 48 h and then concentrated via rotary evaporation. The crude material was purified on a Biotage Isolera One automated flash chromatography system (10 g C18 column; 5% CH₃CN in H₂O → 95% CH₃CN in H₂O) to give FRET-AGM-Di 22 (24.2 mg, 41%) as a reddish-orange solid. TLC (dichloromethane/methanol 8.5:1.5) R_f = 0.37. ¹H NMR (500 MHz, CD₃OD): 8.00 (s, 1 H, FITC H), 7.96 (d, J = 8.0 Hz, 2 H, dabcyl H), 7.86 (d, J = 9.0 Hz, 2 H, dabcyl H), 7.84 (d, J = 8.0 Hz, 2 H, dabcyl H), 7.76 (d, J = 8.0 Hz, 1 H, FITC H), 7.64 (s, 1 H, triazole C5-CH), 7.08 (d, J = 8.0 Hz, 1 H, FITC H), 6.79 (d, J = 9.0 Hz, 2 H, dabcyl H), 6.76–6.67 (m, 4 H, FITC H), 6.55 (d, J = 8.0 Hz, 2 H, FITC H), 4.99 (s, 1 H, Araf-II H-1), 4.88 (s, 1 H, Araf-I H-1), 4.66 (s, 2 H, triazole C4-CH₂-NH), 4.36 (t, J = 6.0 Hz, 2 H, octanoyl CH₂-FITC), 4.32 (dd, J = 3.5, 11.5 Hz, 1 H, Araf-II H-5), 4.19 (dd, J = 6.5, 12.0 Hz, 1 H, Araf-II H-5), 4.14 (q, J = 5.5 Hz, 1 H, Araf-II H-4), 4.07–4.04 (m, 2 H, Ara-II H-2, Ara-I H-4), 3.99 (s, 1 H, H-2), 3.92 (dd, J = 3.0, 5.5 Hz, 1 H, Ara-I H-3), 3.89 (dd, J = 5.0, 11.0 Hz, 1 H, Ara-I H-5), 3.84 (dd, J = 3.0, 6.0 Hz, 1 H, Ara-II H-3), 3.71–3.65 (m, 2 H, Ara-I H-5, linker OCH₂), 3.40 (dt, J = 6.5, 9.5 Hz, 1 H, linker OCH₂), 3.23 (t, J = 7.0 Hz, 2 H, linker CH₂-triazole N1), 3.12 (s, 6 H, dabcyl N-(CH₃)₂), 2.37 (t, J = 7.5 Hz, 2 H, octanoyl α-CH₂), 1.93–1.87 (m, 2 H, octanoyl CH₂CH₂-FITC), 1.65 (pent, J = 7.0 Hz, 2 H, octanoyl α-CH₂CH₂), 1.60–1.53 (m, 4 H, linker CH₂CH₂-triazole N1 and linker OCH₂CH₂), 1.41–1.29 (m, 14 H, linker and octanoyl CH₂s). ¹³C NMR (125.7 MHz, CD₃OD with 10% CDCl₃): δ 175.3, 169.3, 157.5, 156.2, 154.8, 154.3, 144.5, 135.0, 130.5, 129.2, 126.4, 122.9, 112.5, 109.4, 109.2, 103.7, 83.6, 83.0, 82.9, 78.94, 78.90, 70.2, 68.9, 68.8, 68.1, 68.0, 65.1, 51.4, 40.7, 36.2, 34.9, 31.1, 30.9, 30.4, 30.0, 29.96, 29.94, 29.8, 27.6, 27.3, 26.8, 25.7. HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₆₅H₇₈N₉O₁₇ 1256.5515; Found 1256.5465.

4.18 |. FRET-AGM General Storage and Use

Synthetic FRET-AGM probes were typically stored as a 1 mM stock solution in DMSO at −20°C. Enzyme assays utilized FRET-AGM at final concentrations of 0.1 μM, with the final concentration of DMSO being 1% in all samples and controls.

4.19 |. Protein Expression and Purification

LysB fused to a C-terminal 6XHis tag in a pET21b vector (reported in ref [14].) or LysB-S82A mutant fused to a C-terminal 6XHis tag in a pET21a vector (reported in ref. [13]) was transformed into E. coli BL21 DE3. Transformed E. coli were plated on Luria broth (LB) agar containing 100 μg/mL ampicillin (for LysB) or carbenicillin (for LysB-S82A) and grown overnight in standing incubator (37°C). A single colony was picked to inoculate 5 mL of LB liquid medium containing 100 μg/mL ampicillin (for LysB) or carbenicillin (for LysB-S82A) and incubated overnight (37°C, 200 rpm). The following day, the overnight culture was transferred to 1 L of Terrific broth (100 μg/mL ampicillin) and grown in a shaking incubator (37°C, 200 rpm) to mid-log phase (OD₆₀₀ ~ 0.5) before inducing expression of LysB or LysB-S82A with isopropyl β-D-thiogalactopyranoside (IPTG, final concentration 1 mM). The induced culture was incubated for 16 h (16°C, 200 rpm) followed by centrifugation (4000 × g, 15 min) to harvest bacteria from the culture and stored as pellets in a −80°C freezer. LysB or LysB-S82A was recovered from harvested E. coli cells after resuspension in wash buffer (500 mM NaCl, 50 mM NaH₂PO₄, 25 mM imidazole, pH 7.4) and lysis by sonication, using a GE Healthcare ÄKTA pure fast protein liquid chromatography (FPLC) system equipped with a 5 mL nickel affinity chromatography column (Cytiva HisTrap HP). Following column equilibration with wash buffer, lysate was loaded onto the column with a flow rate of 1 mL/min. Elution of bound LysB or LysB-S82A was achieved over a linear gradient of elution buffer (500 mM, 50 mM NaH₂PO₄, 250 mM imidazole, pH 7.4) from 1% to 100% over 24 column volumes at a flow rate of 2 mL/min. Eluted fractions were analyzed by SDS-PAGE and Western blot as described below and fractions containing target protein were pooled and subjected to FPLC size exclusion column chromatography (HiPrep 16/60 Sephacryl S-200 HR, Cytiva). Protein was loaded onto the column and eluted over two column volumes of size exclusion buffer (300 mM NaCl, 50 mM Tris-HCl, pH 7.2) at a flow rate of 0.25 mL/min. Fractions containing pure LysB or LysB-S82A were identified via SDS-PAGE and Western Blot and pooled together for storage in glycerol buffer (20% glycerol, 300 mM NaCl, 50 mM Tris-HCl, pH 7.2). Tdmh and Tdmh-S124A were expressed and purified as reported [10, 12, 31].

4.20 |. SDS-PAGE and Western Blot Analysis

For SDS-PAGE, 3 μg of protein was analyzed by 4%–20% polyacrylamide gel (BioRad) in a Tris-Glycine-SDS running buffer. The gel was fixed for 15 min (40% ethanol, 10% acetic acid in Milli-Q water), rinsed once with Milli-Q water and stained overnight with gentle agitation in QC Colloidal Coomassie stain (Bio-Rad). The gel was destained for 2 h (50% methanol, 10% acetic acid, in Milli-Q water) followed by rinsing with Milli-Q water until the background was clear and imaged using a ChemiDoc Touch Imaging System (Bio-Rad) and processed by Image Lab software (Bio-Rad). For Western blotting, 3 μg of protein were resolved by 4%–20% SDS-PAGE and transferred onto an Immuno-Blot PVDF membrane (Thermo Fisher) using the Trans-Blot Turbo Transfer System (Bio-Rad) at 25 V for 7 min. The PVDF membrane was blocked overnight at 4°C in 5% dry nonfat milk prepared in Tris-buffered saline containing 0.01% Tween 20 (TBST; 50 mM Tris, 0.5 M NaCl, 0.01% Tween 20, pH 7.4). The PVDF membrane was rinsed with TBST before incubation for 1 h at room temperature with constant shaking in 2% dry nonfat milk prepared in TBST with mouse anti-His primary antibodies (1:1000 dilution; Antibodies.com). After washing thrice with TBST, the PVDF membrane was incubated with goat anti-mouse secondary antibodies (1:10,000 dilution; Novus) for 1 h at room temperature. The PVDF membrane was then washed thrice with TBST and visualized via chemiluminescence using SuperSigna West Pico PLUS Chemiluminescent Substrate (Thermo Scientific) with a ChemiDoc Touch Imaging System (Bio-Rad) and processed by Image Lab software (Bio-Rad).

4.21 |. Fluorescence Assays

FRET-AGM quenching efficiency and activation by purified LysB were evaluated by fluorescence assays using a Tecan F200 or M200 multimodal plate reader. Excitation and emission wavelengths of 485 and 535 nm were used, respectively. Fluorescence measurements were performed in black 96-well plates. The gain setting was optimized to wells containing the fluorescent standard FITC-FA and mean fluorescence intensities are reported in arbitrary units (a.u.).

For the quenching efficiency experiments, varying concentrations of FRET-AGM (0.001–10 μM) and the fluorescent standard FITC-FA were prepared in Tris-HCl buffer (50 mM Tris-HCl, 300 mM NaCl, pH 8.0) with a final DMSO concentration of 1% in a black 96-well plate. Fluorescence measurements were taken as described above and fluorescence versus probe concentration plots were used to calculate quenching efficiency as described previously.

For evaluation of FRET-AGM activation by recombinant or commercial enzyme, 20 μg/mL of enzyme (or heat-inactivated enzyme) was incubated in the presence of FRET-AGM probe (0.1 μM) in Tris-HCl buffer with a final DMSO concentration of 1%, at 37°C with shaking (200 rpm) in black 96-well plates. Reactions were initiated by addition of the enzyme to the buffered probe solution, both preheated to 37°C, with measurements of fluorescent intensity every 5 min for 2.5 h.

4.22 |. LC–MS Analysis

FRET-AGM-Mono or FRET-AGM-Di probe (10 μM) were incubated in the presence or absence of LysB or LysB-S82A (20 μg/mL) in Tris buffer at 37°C for 24 h. The sample and control were treated with an equal volume of cold acetone and then centrifuged at 12,000 rpm for 20 min. All samples were diluted to 1 ng/μL with LC–MS grade water with 0.1% formic acid and analyzed on an Agilent 1290 Infinity II LC connected to an Agilent 6545XT UPLC-QTOF with a Dual JetStream source. For all runs, 10 ng of each sample was separated on a Zorbax Eclipse Plus C18 reverse phase column. The gradient started at 98% A (LC–MS grade water with 0.1% formic acid) and 2% B (LC–MS grade acetonitrile with 0.1% formic acid). Eluent B was linearly increased to 30% in 1 min, then to 98% in 3 min, and maintained at 98% for an additional 2 min at a flow rate of 0.4 mL/min. Post-time was used for 2 min to re-equilibrate the column to original conditions. The source parameters were: positive ion mode (3500 V), gas temp (320°C), drying gas (8 L/min), nebulizer (35 psig), sheath gas temp (350°C), sheath gas flow (11 L/min), and fragmentor (125 V). MS data were analyzed by Agilent MassHunter Qualitative Analysis software version 10.0 using the find by formula workflow. Exact masses of FRET-AGM-Mono, FRET-AGM-Di, and their corresponding predicted cleavage products were inserted into a database and extracted ion chromatograms (EIC) were generated for each target within ±5 ppm. Combined EICs for each sample were exported.

4.23 |. Biomolecular Interaction Modeling

Boltz-1 was used to predict interactions of LysB and AGM fragments. The sequence of LysB was downloaded from the Protein Data Bank (PDB ID 3HC7) and the structures of the AGM fragments (AGM-Mono-C8 and AGM-Di-C8) were prepared and converted to SMILES format using ChemDraw 23.1.1. Biomolecular interaction predictions were conducted in a Google Colaboratory Notebook running the Boltz-1 application. CIF files generated from the Boltz-1 application were analyzed and visualized using PyMOL 2.5.7.

Supplementary Material

Additional supporting information can be found online in the Supporting Information section. ¹H and ¹³C NMR spectra. Supporting Fig. S1: Time-dependent activation of FRET-AGM probes (0.1 μM) by LysB versus LysB-S82A (20 μg/mL) in Tris-HCl buffer at 37 °C (A), heat-inactivated LysB (LysB HI) (B), and various other hydrolases (C). A.u., arbitrary units. Error bars represent the standard deviation of three replicates.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.