Isolipoic acid-linked gold nanoparticles bearing the thomsen friedenreich tumor-associated carbohydrate antigen: Stability and in vitro studies

We have previously prepared gold nanoparticles (AuNPs) bearing the Thomsen-Friedenreich antigen disaccharide (TFag), a pan-carcinoma, Tumor-Associated Carbohydrate Antigen (TACA), as tools for various assays and biological applications. Conjugation to AuNPs typically involves the use of thiols due to the affinity of sulfur for the gold surface of the nanoparticle. While a use of a single thiol-containing ligand bound to the gold surface is standard practice, several studies have shown that ligands bearing multiple thiols can enhance the strength of the conjugation in a nearly linear fashion. (R)-(+)-α-Lipoic acid (LA), a naturally occurring disulfide-containing organic acid that is used as a cofactor in many enzymatic reactions, has been used as a linker to conjugate various molecules to AuNPs through its branched di-thiol system to enhance nanoparticle stability. We sought to use a similar system to increase nanoparticle stability that was devoid of the chiral center in (R)-(+)-α-lipoic acid. Isolipoic acid, an isomer of LA, where the exocyclic pentanoic acid chain is shifted by one carbon on the dithiolane ring to produce an achiral acid, was thought to act similarly as LA without the risk of any contaminating (L)-(−) isomer. We synthesized AuNPs with ligands of both serine and threonine glycoamino acids bearing the TFag linked to isolipoic acid and examined their stability under various conditions. In addition, these particles were shown to bind to Galectin-3 and inhibit the interaction of Galectin-3 with a protein displaying copies of the TFag. These agents should prove useful in the design of potential antimetastatic therapeutics that would benefit from achiral linkers that are geometrically linear and achiral.

24 h in the dark. Excess ligand was removed by centrifugation using Amicon spin filter at 5000 rpm for 15 mins and nanoparticles were resuspended in MiIli-Q water. The purification cycle was repeated 3x. The nanoparticles were lyophilized, and concentration was determined by using a UV-Vis method.

Characterization of citrate-capped and functionalized gold nanoparticles
All the nanoparticles were characterized using a combination of several techniques. Hydrodynamic size and zeta potentials were determined by a Zetasizer Nano ZS (Malvern Panalytical, UK) using disposable UV-Vis cuvette and disposable Malvern DTS 1060 zeta potential measurement cell respectively. Surface plasmon resonance (SPR) was recorded on a NanoPhotometer® NP80 (Implen GmbH, München, Germany) in the wavelength range of 300-700 nm at room temperature. Elemental gold (Au) content in citrate capped AuNPs was determined by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) at the NCL which was further confirmed by UV-Vis method. The core size, shape and morphology of the nanoparticles were determined via transmission electron microscopy (TEM) on a JEOL JEM1400 electron microscope after drying on a copper grid.

Confirmation of ligand functionalization on AuNPs
Confirmation of ligand loading on gold nanoparticles were confirmed by SPR, NMR and MALDI-TOF techniques. Surface plasmon of functionalized nanoparticles showed bathochromic shifts from 519 nm to 524 nm, indicative of effective functionalization. All the functionalized nanoparticles were resuspended in water and deuterium oxide (9:1) and their water suppression NMR spectra were obtained. Ligand signal of NMR spectra confirmed the surface functionalization of the AuNPs. For MALDI-TOF spectra, α-Cyano-4-hydroxycinnamic acid was used as a matrix. Both the ligands and matrix were dissolved in water and ethanol mixture (1:1) and spotted on a MALDI plate sequentially and allowed to dry for 10 mins. Subsequently, their MALDI-TOF spectra were recorded in the operating voltage of 80 eV using the reflectron mode.

Quantification of TF-ligand loading on gold nanoparticles
Quantification of TF-Ser/Thr ligands loading were determined by a well-known phenol-sulfuric acid method by comparing with a standard curve generated using β-lactose solutions having concentrations in the range of 0.025-0.5 mg/mL. TF-Ser/Thr-AuNPs solutions (50 µL ) were first treated with 5 µL of concentrated HCl and incubated in the dark overnight to cleave the sugars from the nanoparticle surface. The resulting hydrolysate/nanoparticles mixture were centrifuged at 14000 rpm for 10 mins and supernatants were taken transferred to glass vials which were treated with 150 µL of concentrated sulfuric acid. After 30 mins incubation period, 30 µL of 5% phenol was added to each vial and heated at 90ºC for 1 h. Once the color developed the reaction solutions were transferred to microwell plates and absorbance was measured at 490 nm using a microplate reader.

Stability of gold nanoparticles in High Salt Solutions
The stability of both the citrate-capped and functionalized nanoparticles in different concentration of NaCl was assessed following a literature procedure with modifications. [1] In a 96-well plate, 50 µL of AuNPs (100 µg/mL) were placed in each well. NaCl dissolved in Milli-Q (2 M) water was added to each well to obtain final concentrations of 0, 12.5, 25, 37.5, 50, 100, 125, 250, 500 and 1000 mM sodium chloride in each well. After 24 h incubation in the dark at room temperature, the UV-Vis spectra (surface plasmon) were recorded from a microplate reader in the wavelength range of 400-700 nm. Changes in color (SPR peak) of AuNPs indicated their aggregation behaviors and thus stabilities.

Stability of gold nanoparticles in human serum
We studied all the nanoparticles for their serum stability by following a. [2] Briefly, both citratecapped and functionalized nanoparticles (500 µL) were treated with 500 µL of human serum in centrifuge tubes and incubated at 37ºC. After 24 h incubation time, the AuNPs were centrifuged at 14,000 rpm for 10 mins, supernatants were discarded, and the nanoparticles were resuspended in Milli-Q water. The hydrodynamic size was determined by DLS in a Zetasizer Nano ZS.

Binding studies of AuNPs to Gal-3
Binding of functionalized AuNPs to Galectin-3 was reconfirmed by an ELISA assay. The nanoparticles (0.5 mg/mL, 50 µL) were immobilized in F16 Maxisorp NUNC-Immuno Modules (Thermo Scientific, Roskilde, Denmark) for 12 h at 4⁰C. After washing three times with PBST (PBS with 0.05% Tween 20), the wells were treated with BSA (3% w/v) to block the residual binding site. The wells were washed again three times with PBST and subsequently different concentration of Galectin-3 (2 µM to 4 µM, 50 µL) were added. After 1.5 h incubation at 37⁰C, un-bound galectin-3 was removed by washing three times with PBST and wells were treated with HRP conjugated mouse anti-galectin-3 antibody (1:2000 in PBS, 50 µL) for 1 h at 37⁰C. The wells were again washed with PBST three times and TMB One substrate solution was added to initiate the reaction with peroxidase. After 30 min, the reaction was stopped by adding 1 M hydrochloric acid (50 µL). The resulting signal was recorded at 450 nm by using a Biotek plate reader.

Inhibition studies of Gal-3 by AuNPs
Competitive inhibition of galectin-3 by AuNPs were assessed by an ELISA assay following a literature procedure with modifications. [3] N-Acetyl-D-Lactosamine was utilized as a control Galectin-3 inhibitor. Asialofetuin (0.1 µM in PBS, 50 µL), a well-known Galectin-3 binder, was first immoblized in F16 Maxisorp NUNC-Immuno Modules (Thermo Scientific, Roskilde, Denmark) for 12 h at 4⁰C. The microwells were then washed with PBST (PBS with 0.05% Tween 20) three times (3X 250 µL). Bovine serum albumin (3% w/v in PBS, 300 µL) was added to each well and incubated for 1 h at room temperature to block residual binding site. After washing with PBST ( 3X 300 µL), 25 μL of galectin-3 (1 μM in PBS) and 25 μL of AuNPs ( 5-10 nM) were added to each well simultaneously. After 1 h incubation at 37⁰C, un-bound Galectin-3 was removed by washing with PBST three times. Bound Galectin-3 was detected by utilizing HRP conjugated mouse anti-galectin-3 antibody and subsequent reaction with TMB substrate following above mentioned procedure. Binding signal was recorded at 450 nm by using a Biotek plate reader.

Aggregation Studies of AuNPs to Gal-3 by UV-Vis spectroscopy and DLS
Galectin-3 induced aggregation behavior of functionalized AuNPs were studied by following a literature procedure with modifications. [4] Functionalized AuNPs (10 nM in water, 100 µL) were added to each well in a 96-well plate. 100 µL of galectin-3 with varying concentration from 0-10000 nM were added to each well. After 30 min incubation at room temperature the absorption spectrum were recorded (wavelength range 400 to 750 nm,10 nm step) by a plate reader. The absorbance at 700 nm were plotted against concentration of galectin-3 and Kd value of aggregation was determined by fitting the curve to a Hill functions using GraphPad Prism software. Similar experiments were performed with hydrodynamic size measurements over time as determined by DLS. and hydrodynamic sizes were determined by DLS (after 30 min incubation period with Gal-3).

Aggregation kinetics of AuNPs to Gal-3
Aggregation kinetics of AuNPs to galectin-3 were studied by DLS and UV-Vis Spectroscopy by following literature procedures. [1] To 96 well microplates were added 100 µL of AuNPs (10 nM) followed by Gal-3 in a 1:1 ratio to obtain K D concentration of galectin-3 (116 nM for TF-Ser-AuNPs and 101 nM for TF-Thr-AuNPs) and 5 nM concentration of AuNPs. For Control-AuNPs, 150 nM concentration of galectin-3 was added to microwell plate. The absorbance at 700 nm was recorded at every minute for 30 min. Similarly, hydrodynamic size was determined at every minute with an acquisition time of 55s for 30 min using ultra-low volume cuvettes.

Synthesis of iso-Lipoic Acid
(Although published in 2010 by Tucker, et al., the synthetic methods we used are included here since it has not been reported anywhere else. Some slight modifications were made; see original publication for details)

Diethyl 2-(Hex-5-enyl)malonate
Diethyl malonate in THF (50 mL; 10.62 mL, 70.0 mmol) was very slowly added to a stirred suspension of sodium hydride in THF (150 mL, 2.80 g of a 60% w/w mixture in mineral oil, 70.0 mmol) over 45 min at 0°C. The resulting colorless reaction mixture was stirred for another 15 min at 0°C, then at room temperature for 45 min. The reaction mixture was again cooled down to 0°C, and a diluted solution of 6-bromo-1-hexene (8.50 mL, 63.6 mmol) in THF (50 mL) was added slowly over a period of 15 min. The reaction mixture was subsequently refluxed at 70°C for 18 h. The reaction was monitored by LC-MS, and upon completion was cooled to room temperature and quenched by the addition of 100 mL of cold water. Another 200 mL of water was added and the aqueous solution was extracted with ethyl acetate (3 × 100 mL). The organic layers were combined, washed three times with brine (3 × 100 mL), dried with MgSO4, and solvent was removed. The crude product was then purified by flash chromatography using hexane and ethyl acetate as mobile phase with a gradient of 0%-15% ethyl acetate for 27 min to afford the pure product as a colorless oil (10.

2-(Hex-5-enyl)propane-1,3-diol
In a two-neck round bottom flask, lithium aluminum hydride (3.29 g, 87 mmol) was suspended in THF (200 mL) and cooled to 0°C under nitrogen gas. A solution of the above-prepared diethyl ester (10.49 g, 43.3mmol) pre-dissolved in THF (75 mL) was slowly added to this suspension over 15 min. The reaction mixture was allowed to stir for 45 min at 0 °C followed by warming to room temperature and stirring for another 6 h. THF (100 ml) was added and the solution was cooled to 0 °C. Cold water (10 mL) was added slowly followed by addition of aqueous sodium hydroxide solution (2 M; 10 mL). The reaction mixture was then stirred vigorously at room temperature for 12 h. The resulting reaction mixture was then filtered through Celite, and washed with ethyl acetate (100 mL). The filtrate was concentrated under reduced pressure using yielding the crude product which was purified by flash chromatography using Hexane and ethyl acetate to give the title compound as a colorless oil (5.35 g, 78%).

7-((methylsulfonyl)oxy)-6-(((methylsulfonyl)oxy)methyl)heptanoic acid
Osmium tetroxide (1.24 mL of a 2.5% w/w solution in tBuOH, 0.099 mmol) was added to a stirred solution of the olefin (3.10 g, 9.86 mmol) in DMF (65 mL), and the resulting mixture was stirred for 5 min. Methanesulfonic acid 2-methanesulfonyloxymethyloct-7-enyl ester (2.088 g, 6.1 mmol) was dissolved in dry DMF (50 mL). Osmium tetroxide (0.766 mL of a 2.5% w/w solution in tBuOH, 0.061 mmol) was added to the reaction vessel and stirred for 5 min. After that oxone (15 g, 24.4 mmol) was added to the reaction mixture in three parts over 15 min and allowed to stir for 3.5 h at room temperature. Sodium sulfite (13.84 g) was then added in two portions and stirred vigoriously for another 2 h. The reaction was then quenched by adding water (100 mL), aqueous hydrochloric acid (

Procedure for the synthesis of Control ligand (LA-PEG-OH)
In a dry 20 mL vial, EDC•HCl (0.192 g, 1 mmol) and 1-hydroxybenzotriazole hydrate (0.153 g, 1 mmol) were dissolved in dry DCM (5 mL) at 0°C. After 10 min, Isolipoic acid (0.103 g, 0.5 mmol) dissolved in DCM (2 mL) was added in one batch. The reaction mixture was stirred for 45 min at 0°C. After that, Amino-PEG 7 -alcohol 10 (0.163 g, 0.5 mmol) was added to the reaction mixture followed by addition of DIPEA (35 μL, 0.2 mmol). The reaction was then stirred for 12 h at room temperature. The reaction was monitored by LC-MS for completion. The reaction mixture was then absorbed into silica gel and purified by flash chromatography using DCM and methanol with a gradient of up to 10% methanol for 30 min. The useful fractions were identified by LC-MS and combined and subjected to another purification step via HPLC using water and acetonitrile as the mobile phase to yield the title compound 11 as yellowish sticky solid (0.22 g, 86%). EDC•HCl (0.192 g, 1 mmol) and 1-hydroxybenzotriazole hydrate (0.153 g, 1 mmol) in dry DCM (5 mL) was stirred at 0°C in a 20 mL glass vial. After 5 min, Isolipoic acid (0.103 g, 0.5 mmol) in DCM (2 mL) was added to the reaction mixture and stirred 45 min at room temperature. t-Boc-N-amido-PEG6-amine (1) (0.212 g, 0.5 mmol) in dry DCM (2 mL) and DIPEA (45 μL, 0.32 mmol) were added sequentially to the reaction mixture. The reaction was stirred for 12 h and monitored by LC-MS for reaction completion. The crude reaction mixture was then incorporated into silica gel and subjected to purification by flash chromatography using DCM and methanol (9:1) as eluent for 35 min. The pure fractions were combined and purified again by HPLC to give the ILA-PEG 6 -Tboc (2) as sticky yellowish solid (0.275 g, 90%). 50% Trifluoroacetic acid in DCM (2 mL) was added to the ILA-PEG 6 -Tboc (0.275, 0.5 mmol) conjugate in a 20 mL glass vial. The reaction was stirred for 6 h at room temperature. TFA was removed under nitrogen gas flow, incorporated into silica gel and purified by flash chromatography using DCM and methanol as mobile phase with a gradient of up to 10% methanol. The pure fractions were identified, combined and re-purified by HPLC using water and acetonitrile as mobile phase to give the ILA-PEG 6 -NH 2 linker (3) as a yellowish sticky solid (0.21 g, 91%). To a stirring solution of TF-Ser/Thr amino acid (0.144 g, 0.75 mmol) in DCM (7 mL) were added EDC.HCl (0.144 g, 0.75 mmol) and 1-hydroxybenzotriazole hydrate (0.115 g, 0.75 mmol) at 0°C. After 45 mins, ILA-PEG 6 -NH 2 (3) (0.144 g, 0.75 mmol) in DCM (2 mL) and DIPEA (5 μL, 0.3 mmol) were added to the reaction mixture and stirred for 12 h at room temperature. The reaction was monitored by LC-MS, once completed the reaction mix was absorbed into silica and purified by flash chromatography using DCM and methanol (9:1) as mobile phase for 30 min. Pure fractions were identified, combined and purified again by HPLC using water and acetonitrile as eluent to yield the conjugate as white solid.