Gadolinium-Conjugated Lipoic Acid Hydrogels for Magnetic Resonance Imaging

Abstract

A gadolinium-based contrast agent (GBCA) disulfide homodimer (Gd³⁺SS) has been incorporated into a lipoic acid (LA)-based hydrogel (Gd³⁺Gel) for enhanced magnetic resonance imaging (MRI). This study evaluates the magnetic properties and in vitro behavior of Gd³⁺Gel for potential applications in tracking internal injuries. Results indicate a direct shortening in the relaxation rate with greater magnitude of LA polymerization and a 2.8-fold enhancement in relaxivity (r₁) at 1.4 T when Gd³⁺SS was conjugated into the hydrogel. This effect is attributed to a significant increase in the rotational correlation time (τ_r) from 0.22 ± 0.05 ns (Gd³⁺SS) to 6 ± 1 ns (Gd³⁺Gel). Retention studies confirm that Gd³⁺SS remains covalently within the hydrogel, with retention of 64.7 ± 1.9% for Gd³⁺SS and 14.0 ± 1.4% for noncovalent binding gadoterate. The hydrogel relaxation rate (1/T₁) increases from 1.1 to 3.5 s⁻¹ at 7 T from blank gel to Gd³⁺Gel (0.24 mM Gd³⁺). Cell studies show that PC3-PIP and RAW 264.7 cells maintain high viability with Gd³⁺SS but exhibit reduced viability with Gd³⁺Gel, consistent with known lipoic acid effects on immortalized cell lines. Cellular uptake studies using ICP-MS and confocal fluorescence microscopy confirm that monomeric Gd³⁺SS is readily internalized, whereas Gd³⁺Gel significantly limits diffusion and uptake. Rheology was conducted to determine the zero-shear viscosity of the LA hydrogel at various concentrations of LA. These findings suggest that the LA hydrogel scaffold enhances MRI contrast, minimizes leaching, and is easily injectable. Gd³⁺Gel is a promising tool for potential targeted imaging and controlled uptake during healing processes.

Graphical Abstract

INTRODUCTION

Magnetic resonance imaging (MRI) is a powerful tool in clinical diagnostics with excellent spatial resolution, lack of invasiveness, and nonionizing radiation.¹ Gadolinium-based contrast agents (GBCAs) allow for improved enhancement of T₁-weighted MRI.^2–5 Current advances in GBCAs have focused on improving MRI sensitivity due to the lack thereof compared to other imaging modalities.^1,2,6 Several approaches have been developed to enhance the relaxivity (r₁) of GBCAs, increase the accumulation of the GBCAs ([Gd³⁺]), or a combination of both according to eq 1:

$$\frac{1}{T_{1\text{,obs}}}=r_1[{\text{Gd}}^{\text{3+}}]+\frac{1}{T_{\text{1,d}}}$$ (1)

where 1/T_1,obs is the observed ¹H relaxation time, [Gd³⁺] is the local concentration of GBCA, and 1/T_1,d is the diamagnetic background. In this study, we investigate a GBCA conjugated to a lipoic acid (LA)-based hydrogel for potential fate-mapping. Hydrogel conjugation also increases the r₁ of the GBCA probe through a slowing of the rotational correlation time (τ_r). The GBCA concentration in our design is easily tunable during the LA hydrogel synthesis.

Hydrogels are porous polymers containing channels of water that have shown promising results in biomedical applications,^7–9 and LA is a naturally occurring metabolite, essential for aerobic metabolism and redox homeostasis.¹⁰ LA hydrogels are therefore highly biocompatible and offer a combination of therapeutic effects including excellent external wound healing, antibacterial activity, and anti-inflammatory properties.^8,11,12 Hydrogels have been used for controlled release of a drug or an imaging agent.^13–15 However, there has been minimal progress in investigating the hydrogels for monitoring in vivo injuries. Several hydrogels have been used to investigate various MRI properties,^16–20 with few studies of Gd³⁺-conjugated hydrogels for T₁-weighted imaging^9,21–25 and none looking at LA-based hydrogels. LA hydrogels use a simple linking of disulfide bonds from the favorable ring opening of a five-membered 1,2-dithiolane ring to form linear chains with hydrophilic carboxylate side chains interacting with water and Tris base counterions. This allows for facile hydrogel derivatization and tunability by the use of substances containing a thiol or disulfide.

Disulfide and thiol functional GBCAs have previously been investigated for their relaxivity and cell uptake/compatibility.^26,27 Along with wound healing, the LA hydrogel and the disulfide GBCAs take advantage of common thiol-mediated uptake when reacting to exofacial protein thiols (EPTs). EPTs are essential for several metabolic activities, including cell signaling and cellular uptake.²⁸ There has been extensive investigation into the passive and active thiol-mediated uptake.^29–31 These combined traits can be used to fate-map the healing process over time, imaging cells as they remove the hydrogel. In this work, we develop a disulfide GBCA by synthesizing a homodimer (Gd³⁺SS) and incorporating it into the LA hydrogel (Gd³⁺Gel) to investigate the relaxivity and in vitro properties (Scheme 1). To investigate the tailoring properties of the LA hydrogel further, we also developed a fluorescein isothiocyanate disulfide (FITC-SS and FITC-Gel) derivative for confocal microscopy.

Scheme 1. Synthesis of the Gd3+Gel Using Base-Catalyzed Polymerization^a.

a The inset is a 1.38 M LA hydrogel upside down in a relaxivity tube to show the high viscosity that holds the gel in place.

EXPERIMENTAL SECTION

Synthesis.

The amino-DO3A and FITC-SS were synthesized according to previous literature studies.^32,33 The below is in reference to Scheme S1.

Tri-tert-butyl 2,2′,2″-(10-(2-(3-(Tritylthio)propanamido)ethyl)-1,4,7,10-tetraazacyclododecane-1,4,7-triyl)triacetate (1).

The amino-DO3A (1.0480 g, 1.882 mmol), S-trityl mercaptopropionic acid (0.6635 g, 1.904 mmol), EDAC (0.7345 g, 3.846 mmol), and HOBt (0.5094 g, 3.801 mmol) were dissolved in DCM (20 mL) at 0 °C. DIPEA (1 mL, 5.806 mmol) was added. After 30 min, the reaction was taken to room temperature and allowed to stir overnight. The reaction was poured into a separation funnel with H₂O (50 mL) and DCM (50 mL). The organic layer was removed, and the aqueous layer was washed once more with DCM (50 mL). The combined organic layers were dried over Na₂SO₄ and concentrated. The crude oil was purified by flash column chromatography with a gradient of 3 to 5 to 7.5% MeOH in DCM (7.5% MeOH in DCM, r_f = 0.25) to give a viscous liquid (0.5559 g, 33% yield). ¹H NMR (500 MHz, CDCl₃) δ: 7.38 (d, 6 H), 7.26 (t, 6H), 7.19 (t, 3H), 3.64–2.24 (m, 31 H), 1.45 (m, 27 H). ¹³C NMR (125 MHz, CDCl₃) δ: 173.4, 172.5, 171.8, 169.9, 144.9, 129.6, 127.8, 126.5, 82.9, 82.5, 81.9, 66.5, 56.5, 55.6, 52.8, 50.4, 35.6, 34.9, 28.1, 28.1, 27.9, 27.7. HRMS (ESI-MS) m/z: [M + H]⁺ calcd for C₅₀H₇₃N₅O₇S 888.530; found 888.536.i

Gd³⁺SS.

1 (0.5559 g, 0.6267 mmol) was dissolved in a solution of TFA (10 mL), DCM (10 mL), and TIPS (0.39 mL, 1.91 mmol) at room temperature and stirred for 12 h whereupon DMSO (1 mL) was added. After 7 h, the reaction was concentrated down. The crude solid was resuspended in water (25 mL) and centrifuged down (15,000 rcf, 10 min, rt). The supernatant was decanted and filtered. The pH was adjusted to 6, and GdCl₃·6H₂O (0.2297 g, 0.626 mmol) was added. The pH was maintained at 6 until it stabilized. After 24 h, the pH was adjusted to 10+, filtered, and lyophilized. The crude was redissolved in Milli Q water and purified by HPLC method 1, with a retention time of 14.6 min. The solution was lyophilized to give a white powder (0.1533 g, 39% yield). HRMS (ESI-MS) m/z: [M + H]⁺ calcd for C₃₈H₆₂Gd₂N₁₀O₁₄S₂ 1263.244; found 1263.246.

Formation of Hydrogels.

Tris base (250 mg) was dissolved in water (1 mL) with heat (90 °C), and the mixture was vortexed vigorously. Gd³⁺SS, gadoterate meglumine, or FITC-SS was added from stock solutions in water that were sterilized by passage through 0.2 μm sterile filters. LA was added (50, 125, 250, 500, or 750 mg for concentrations of 0.19, 0.44, 0.81, 1.38, or 1.82 M, respectively) while the solution was hot and aggressively vortexed. While hot, the hydrogel was molded into a syringe, relaxivity tube, or falcon tube. Small aliquots were taken to determine the final concentration of Gd³⁺ by ICP-MS for each hydrogel. The mold was frozen at – 20 °C overnight, thawed, and immediately tested. All relaxivity at 1.4 T was measured at 37 °C.

Retention.

Redundant 0.2 mM Gd³⁺, 1.38 M LA gels were made with Gd³⁺SS or gadoterate meglumine and molded in 15 mL falcon tubes. After formation, the hydrogels were swollen with an additional 5 mL of water using heat and vortexing. The hydrogels were centrifuged (10,000 rcf, 10 min, rt). The supernatant was filtered (0.2 μm), and the Gd³⁺ concentration was measured by ICP-MS. The experimental concentration was divided by the expected concentration from the dilution and subtracted from 100 to determine the percent of Gd³⁺ retention.

Relaxivity Overtime.

Nine redundant 0.2 mM Gd³⁺, 1.38 M LA hydrogels were made with Gd³⁺SS and molded into relaxivity tubes in three triplicate sets. The initial relaxivity of each set was measured after a single freeze–thaw cycle. Thereafter, each set was incubated at variable temperatures. One set was immediately refrozen upon measurement at −20 °C relaxivity and only thawed daily to measure. The other two sets were kept at either rt (21 °C) or 37 °C. T₁ was measured for each set every 24 h for 10 days.

Rheology.

Rheometric measurements were performed using an Anton Paar Modular Compact Rheometer (MCR 302). Variable shear rate oscillatory experiments were performed at 37 °C from 1 × 10^–3 Hz to 1 × 10³ Hz using a 50 mm and 1° cone–plate geometry. A double gap fixture was used to perform shear experiments on a 0.19 M hydrogel and a single gap PP25 fixture for all other gels. Approximately 400 μL of gel was deposited onto the plate and surrounded by a water reservoir for humidity control. Four oscillatory experiments were conducted for each gel. Viscosities are reported as the average of four zero-shear viscosity values obtained for each gel. The measurement system was cleaned thoroughly with water before the gels. Acetone was also used for cleaning higher viscosity gels.

Cell Lines and Cultures.

PC3-PIP cells were provided by Professor James Basillion (Case Western Reserve University), and RAW 264.7 was purchased from the Northwestern Developmental Therapeutic Core, Evanston, IL. PC3-PIP cells were grown in RPMI 1640 medium and RAW 264.7 cells in DMEM supplemented with 10% fetal bovine serum (FBS), l-glutamine, penicillin, and streptomycin. All cell lines were grown in a humidified incubator at 37 °C and 5% CO₂ and harvested using 0.25% TrypLE. Scraping was needed for RAW 264.7 harvesting. Solutions were filtered through 0.2 μm sterile filters before use. For seeding, cells were counted by a hemocytometer.

Cell Uptake.

Cells were seeded at 3 × 10⁵ cells per well in a six-well plate. For the time dependence, the cells were incubated overnight before dosing and cells were harvested at the designated time points over 48 h. For the dose dependence, the cells were immediately dosed at the time of seeding and incubated for 24 h before harvesting. Cells were dosed for Gd³⁺SS at 5, 20, or 200 μM Gd³⁺ or Gd³⁺Gel at 5 μM Gd³⁺/1.4 mM LA. The standard harvesting procedure is as follows: the medium was aspirated, and cells were washed with 1 mL of PBS and incubated in another 1 mL of PBS for 10 min. The PBS was aspirated, and the cells were incubated in 1 mL of 0.25% TrypLE for 10 min. The cells were resuspended in media and spun down (10 min, 250 rcf, 4 °C). The supernatant was aspirated, and the cells were resuspended in approximately 200 μL of media. A total of 40 μL was used for cell counting with a Gauva Easy Cyte Mini according to the manual. The remaining cells were used for ICP-MS to determine the concentration of Gd³⁺. All dilutions for ICP-MS were determined from the mass.

Cell Viability Assay.

Cells were seeded at 5000 cells per well in a 96-well plate. Triplicate treatments were added such that the well concentrations were as follows: Gd³⁺-SS (5, 10, 24, 50, 100, 200, 457.5, 915, or 1830 μM Gd³⁺) and Gd³⁺Gel (Gd³⁺ conc. of 3.8, 7.6, 15.2, 38.0, 76.0, 152, or 304 μM with the corresponding LA conc. of 1.1, 2.2, 4.4, 11, 22, 44, or 88 mM). A modified hydrogel was made to control the pH where 1 equiv of Tris base to LA was used to make a stock of 0.44 M LA/1.51 mM Gd³⁺ hydrogel. This stock and dilutions of this stock were added directly into the wells with the cells. The total working volume was 100 μL, and 200 μL of PBS was used on the outside wells to keep the working volume from evaporating. The cells were incubated at 37 °C and 5% CO₂ for 24 h. The cell medium was aspirated, cells were washed with PBS, and fresh media was added. 20 μL of CellTiter-Glo 2.0 Reagent (Promega, Madison, WI) was added and the cells were incubated for 3.5 h before reading the absorbance at 490 nm. A background control without cells was subtracted from the sample absorbance, and values were normalized to the negative control wells.

In Vitro Fluorescence Imaging.

A total of 100,000 cells were plated on a 35 mm FluoroDish (World Precision Instruments, Sarasota, FL, USA) and incubated overnight (37 °C, 5% CO₂). Cells were dosed with the FITC-SS (10 μM in PBS) or with a pH-controlled FITC-Gel (5 μM FITC, 11 mM LA). The hydrogel dish was gently swirled to try and disperse the hydrogel. After 15 min of incubation (37 °C, 5% CO₂), the medium was removed and the cells were washed thrice with PBS. FluoroBrite DMEM (Gibco, Waltham, MA) was added as the media. Live cells were imaged using a Nikon SoRa Spinning Disk confocal at a laser wavelength of 488 nm (intensity 15% for FITC-SS and 80% for FITC-Gel), objective of 60× (NA – 1.42) oil immersion, and using a Hamamatsu ORCA-Fusion Digital CMOS Camera. Z-stacking was used to enhance the images.

RESULTS AND DISCUSSION

Synthesis of the Hydrogel.

The gadolinium agent was synthesized from a common ^tBuDO3A chelator³⁴ with an amino arm³² for peptide coupling a trityl-protected mercaptopropionic acid. A deprotection with TFA, DCM, TIPS, and DMSO³⁵ allowed for the formation of a disulfide bond, and a metalation gave Gd³⁺SS (Scheme S1). The disulfide form was chosen over the reduced thiol due to the enhanced stability in storage.

Gd³⁺SS was investigated by covalently conjugating into a lipoic acid-based hydrogel (Scheme 1). By using the proper chelator and covalently attaching the Gd³⁺SS, we further decrease its bioavailability to potential off targets compared to other hydrogels that attempted to chelate the Gd³⁺ with the hydrogel itslef.³⁶ While several methods are currently in the literature for formation of the lipoic acid hydrogel,^37–40 we adapted a method using Tris base.⁸ Tris base would be dissolved in water at either a molar equivalence or an excess of LA. A range of LA concentrations were used to measure their effects on physical, magnetic, and cellular properties. These gels were easily molded into a syringe, NMR tube, or falcon tube.

Relaxivity.

Water proton relaxation rates were measured at 1.4 T and 37 °C, conditions chosen to reflect common clinical MRI parameters. In the presence of lipoic acid (LA), a linear relationship (R² = 0.9758) was observed between the longitudinal relaxation rate (1/T₁) and the concentration of LA (Figure 1a). This result suggests that the extent of LA polymerization (i.e., “linking”) can be inferred from the relaxation behavior of water. As the LA concentration increases, longer polymer chains and higher polymer density likely enhance water–polymer interactions, raising the relaxation rate. An increase in LA concentration subsequently increases the macroscopic viscosity, giving rise to a similarly linear relationship with r₁. This may indicate an accompanying rise in microscopic viscosity, further elevating the relaxation rate and prolonging the rotational correlation time (τ_r) of water molecules.⁴¹

Figure 1.

(a) Relaxation rates (1/T₁ and 1/T₂) of LA hydrogels without Gd³⁺SS and increasing concentrations of LA. (b) T₁ and T₂ of Gd³⁺Gels stored at room temperature, 37 °C, or −20 °C over time. The gels are 1.38 M LA. (c) Relaxivity of the Gd³⁺Gels with increasing concentrations of LA. (d) Percent retained from swelling and dilution of the LA hydrogels with gadoterate and Gd³⁺SS (**** = p < 0.001, Student’s t test). (e) r₁ of Gd³⁺SS at different pH levels adjusted using NaOH and HCl. (f) Nuclear magnetic resonance dispersion (NMRD) of Gd³⁺SS and 1.38 M LA Gd³⁺Gel. All relativities are per Gd³⁺ concentration.

Several synthesis parameters were examined to determine whether they influenced the degree of LA linking specifically, extended stirring times, and immediate heating after LA addition. Neither procedure altered the final extent of cross-linking (see the Supporting Information). However, storing Gd³⁺-labeled hydrogels (Gd³⁺Gels) under different conditions—room temperature, 37 °C, and repeated freeze–thaw cycles at −20 °C—revealed that elevated temperatures promote greater cross-linking (Figure 1b). All T₁ measurements were conducted at 37 °C, with frozen samples thawed just long enough for measurement before being refrozen. Over time, a slow decrease in T₁ was observed for frozen gels, whereas samples kept at room temperature or 37 °C reached a near-plateau in T₁ after about 1 day. Although heat-induced linking of LA-based materials is well-documented,³⁹ its direct demonstration via T₁ relaxation in LA hydrogels has not been previously reported. These findings highlight the sensitivity of T₁ to polymerization and the potential for the temperature to accelerate hydrogel formation.

Despite hydrogels being highly attractive for in vivo applications—owing to their ability to carry aqueous cargo and capacity for mobility—there are comparatively few reports of Gd³⁺-conjugated hydrogels designed specifically for T₁-weighted MRI.^9,21–24 Across all tested samples, Gd³⁺Gels consistently exhibited a T₁/T₂ ratio of 1.49 ± 0.04, characteristic of GBCAs.

All subsequent r₁ measurements were performed immediately after the first thaw to maintain consistency across the samples. The monomeric Gd³⁺SS exhibited an r₁ of 7.3 ± 0.2 mM⁻¹ s⁻¹ at 1.4 T and 37 °C, which is slightly higher than a similar monomeric analogue.⁴² This modest increase in r₁ is attributed to the larger molecular mass of Gd³⁺SS compared to the simpler monomer.²⁴ When Gd³⁺SS was incorporated into the LA hydrogel (Gd³⁺Gel), the r₁ more than doubled, reaching 20.2 ± 0.2 mM⁻¹ s⁻¹ at 1.4 T and 37 °C for the hydrogel containing 1.38 M LA. Increasing the LA content in the hydrogel continued to elevate r₁, presumably due to partial immobilization of Gd³⁺SS by the denser polymer network (Figure 1c). However, the correlation between r₁ and LA concentration in Gd³⁺Gels was somewhat lower (R² = 0.8455) than the near-linear relationship seen with 1/T₁ versus LA concentration alone.

To verify whether this enhancement in r₁ was an inherent feature of covalent Gd³⁺ attachment, a control experiment was conducted with commercially available gadoterate meglumine. When both gadoterate and Gd³⁺SS were incorporated into hydrogels at the same LA concentration (1.38 M), they showed comparable r₁ values (19.6 ± 0.5 and 20.2 ± 0.2 mM⁻¹ s⁻¹, respectively; see Table S2). Interestingly, although gadoterate and Gd³⁺SS differ in their free-solution r₁ (gadoterate r₁ = 3.32 ± 0.13 mM⁻¹ s⁻¹),⁴³ both converge to similar relaxivities once embedded in the hydrogel matrix.

A swelling–dilution experiment provided further insights into the covalent binding of Gd³⁺SS. Under these conditions, 64.7 ± 1.9% of Gd³⁺SS was retained compared to only 14.0 ± 1.4% of noncovalently incorporated gadoterate (Figure 1d), a statistically significant difference (p < 0.001). These data confirm that Gd³⁺SS is more strongly retained within the LA-based hydrogel network, consistent with covalent conjugation. Despite the difference in binding modes, however, both gadoterate and Gd³⁺SS displayed similar r₁ values when entrapped in the hydrogel at equivalent LA concentrations, indicating that the presence or absence of covalent bonds does not significantly affect the relaxivity in these particular formulations.

The pH sensitivity of Gd³⁺SS was evaluated. The standard Gd³⁺Gel dilutions kept the concentration of Tris base constant while varying the LA concentration. This would make the lower concentrations of LA more basic, which could potentially impact the relaxation rate of the Gd³⁺Gel. The pH titration showed no significant shift in the basic region of the Tris buffer (Figure 1e).

Nuclear magnetic resonance dispersion (NMRD) was used to better understand the relaxation mechanisms in Gd³⁺Gel (Figure 1f and Table S1). While the relaxivity profile for the Gd³⁺SS is characterized by a single dispersion (common for fast reorienting small Gd³⁺ complexes), the appearance of a relaxivity peak around 20 MHz in the Gd³⁺Gel clearly points to a τ_r enhancement effect near the clinical field strength. Fitting of the NMRD curves with the Solomon–Bloembergen–Morgan (SBM) model shows that the correlation time modulating relaxation can be ascribed for 10–25% to a longer τ_r, which increases from 0.22 ± 0.05 to 6 ± 1 ns on passing from Gd³⁺SS to Gd³⁺Gel. Additionally, there is a faster reorientation time (τ_l), which passes from 23 ± 3 to 40 ± 5 ps for the remaining 90–75%. This latter time can be ascribed to the large residual internal mobility of the DO3A-derivative allowed by the flexible linker. The obtained values agree with what is expected from the structure of the Gd³⁺SS monomer and with the slowing due to gel formation. In this analysis, for both systems, two water molecules that are coordinated⁴² to the Gd³⁺ ion have been considered in fast exchange, in agreement with the temperature dependence of the profiles. Remarkably, from this analysis, it results that the relaxivity peak observed for the Gd³⁺Gel is only determined by the small fraction modulated by the longer reorientation time. Therefore, we predict that significantly higher relaxivity values could be obtained by reducing the fast internal mobility through formation of gels that can further restrict the motion.

7 T phantoms were taken to determine the enhancement effects of the gel at higher field strengths (Table 1). The gel without Gd³⁺SS already observes a four-fold rate enhancement of the T₁ over the blank PBS. When adding Gd³⁺SS (0.24 mM Gd³⁺) to the 1.38 M LA gel, there is a 10.5-fold rate enhancement. This Gd³⁺Gel is shorter in T₁ than equivalent concentrations of Gd³⁺SS, demonstrating improvements in other magnetic properties besides the τ_r boost. Overall, Gd³⁺Gel sees a broad relaxivity enhancement.

Table 1.

7 T Phantoms (500 ms Delay for Representation) of Gd³⁺SS and Hydrogels with and without Inclusion of Gd³⁺SS at Room Temperature

	Water	Hydrogel	Gd3+SS	Gd3+Gel
Phantoms (500 ms)
Gd3+ (mM)	0	0	0.26	0.24
LA (M)	0	1.38	0	1.38
T1 (ms)	2706 ± 15	662 ± 17	328.4 ± 1.4	258 ± 4
T2 (ms)	240 ± 18	84 ± 2	130 ± 3	41.0 ± 0.5

Rheology.

While the Gd³⁺SS is a terminal group of the linear disulfide polymer, we saw no significant difference in rheology at the highest ratio of Gd³⁺SS to LA concentration (0.002). The gels showed non-Newtonian properties, by having dynamic viscosities over a range of shear rates. The LA hydrogels demonstrated constant viscosity at low shear rates (<1 rad/s) and an inversely linear viscosity at high shear rates (>10 radians/s) when plotted logarithmically, except for the most diluted 0.19 M LA hydrogel (Figure 2a). Zero-shear viscosity values were determined by averaging the viscosity values at a low shear rate for each gel except the 0.19 M LA (Table S4). The observed decrease in viscosity with an increasing shear rate is a characteristic shear-thinning behavior.⁴⁴ This makes these gels amenable to in vivo applications, as they can be tailored to extrude through small gauge needles for injections. LA hydrogels could be a good alternative to extracellular matrix hydrogels, which are being investigated to accelerate healing for ischemic strokes and hemorrhage.^45,46

Figure 2.

(a) Rheology of the LA hydrogels with different LA concentrations and an increasing shear rate. Cell viability with (b) Gd³⁺SS and (c) Gd³⁺Gels. (d) Dose-dependent cell uptake via ICP-MS when incubated after 24 h and (e) time dependence when incubated in 200 μM Gd³⁺SS. (f) Cell uptake of 5 μM Gd³⁺ of Gd³⁺SS and Gd³⁺Gels (1.4 mM LA) incubated after 24 h (**** = p < 0.001, ns = not significant, Student’s t test). All dosages are per Gd³⁺ concentration, and “No Gd³⁺” represents the limit of detection.

Uniquely, the relaxivity had a linear relationship with the logarithm of the macroscopic viscosity, which increases exponentially with LA concentration (r₁, R² = 0.9959, Figure S1). A previous study was conducted to measure the viscosity using MRI empirically but did not utilize a GBCA,⁴⁷ despite the role of viscosity in MRI contrast.^48,49 Future studies could be conducted to measure the microscopic viscosity using magnetic resonance.

Cell Viability.

The cell viability of the Gd³⁺SS and Gd³⁺Gel was then explored using the CellTiter-Glo luminescence assay. PC3-PIP was selected as an epithelial cell line and RAW 264.7 macrophages for their active uptake of antigens. Both cell lines are known to express EPTs.^50,51 Similar to previous Gd³⁺ disulfide and thiol-containing MR agents,²⁷ Gd³⁺SS shows no significant cell death (Figure 2b). The LA hydrogel however shows cell death, which is common for immortalized and tumorigenic cell lines (Figure 2c).⁵² LA is not toxic in normal cell lines and is even used as a nutritional supplement with several preliminary studies showing high safety profiles.¹⁰ A similar lipoic acid hydrogel has shown promise with accelerated healing when replacing an excised tumor.³⁸

Modified Gd³⁺Gels were used to control for the pH by using equimolar Tris base to LA and all dilutions done in PBS. Even though the Gd³⁺Gel constrains much of the lipoic acid and Gd³⁺SS, the linking is not complete as the yellow color indicates that five-membered, 1,2-dithiolane rings remain.³⁹

This means some dosages of the surrounding cells of the gel are still exposed to lower concentrations of free LA and likewise any other thiol or disulfide monomer.

Cell Uptake.

This property is observed by ICP-MS. The overall LA was kept at 1.4 mM or lower where each cell line would still be viable enough to measure after 24 h. When both cell lines are dosed with Gd³⁺SS, there is significant cell uptake, like previous sulfur uptake probes.^26,27 However, previous probes did not test a wide range of concentrations, specifically lower micromolar concentrations where there is a clear nonlinear dosage dependence with Gd³⁺SS (Figure 2d). Uniquely, lower concentrations saw higher cell uptake than those with higher dosages. Several peptides with thiols have been investigated for cell uptake but they lack in dosages at higher ranges,^30,31 whereas previous Gd³⁺ disulfide agents only dosed concentrations higher than 1 mM.²⁷ The previous study did cell uptake with increasing concentrations of a reductant, tris(2-carboxyethyl) phosphine (TCEP), which showed that the highest concentration of TCEP had a lower Gd³⁺ uptake.²⁶ This nonlinear trend could be due to the sensitive regulatory processes that ETC’s have on cellular signals, as this is a main mode of transportation for viruses,⁵³ and should be taken into account when designing future agents.

When a time dependence curve is performed, the 20 μM concentration sees a higher overall uptake but no clear trend (Figure S2), whereas the larger 200 μM dose sees a lower yet steady linear increase (Figure 2e). This characteristic is reflected when dosed with 5 μM Gd³⁺ in Gd³⁺SS and Gd³⁺Gels. RAW 264.7 and PC3-PIP see 2.3 ± 0.8 and 1.6 ± 0.5 fmol of Gd³⁺ per cell with the Gd³⁺SS, respectively, whereas the Gd³⁺Gel uptake is not different from negative controls (Figure 2f). This is again likely due to Gd³⁺ being constrained to the Gd³⁺Gel.

Confocal fluorescence microscopy was used to investigate uptake, where a similar FITC disulfide (FITC-SS) was synthesized and conjugated into the LA hydrogel (FITC-Gel). The cells were incubated for 15 min with FITC-SS or FITC-Gel (10 μM FITC). The imaging further confirmed the uptake results from the ICP-MS experiments where FITC-SS saw intracellular accumulation compared to the FITC-Gel (Figure 3). The morphologies of the FITC-Gel exposed cells exhibit cells under stress or apoptotic as they become rounder and/or bleb. There is additional evidence of FITC-Gel debris showing the limiting diffusion that the hydrogel produces.

Figure 3.

Bright field and confocal fluorescence microscopy of PC3-PIP and RAW 264.7 incubated in 10 μM FITC in either FITC-SS or FITC-Gel (1.4 mM LA) after 15 min. Note that the debris in the FITC-Gel cells has more round morphologies of stressed cells, whereas the FITC-SS cells exhibit more typical morphology and brighter fluorescence.

CONCLUSIONS

In summary, the GBCA was covalently incorporated into a lipoic acid (LA) hydrogel to explore its potential for mapping the healing process of internal injuries. When the LA hydrogel and the Gd³⁺SS compound were combined, they synergistically improve water relaxation, evident by a direct increase in the relaxation rate with greater linking of the LA polymer. In cell studies, PC3-PIP and RAW 264.7 cells remained highly viable when exposed to the monomer (Gd³⁺SS) but not with Gd³⁺Gel, a response typical of immortalized cell lines in the presence of lipoic acid.

The minimal leaching observed in LA-based hydrogels confers an advantage for targeted imaging and localized therapy rather than presenting a limitation. By retaining the imaging agent (e.g., Gd³⁺ complexes) within the hydrogel, only cells in direct contact with the matrix take up the agent, allowing for precise, site-specific visualization of tissue–hydrogel interactions. These preliminary findings highlight the potential of LA-based hydrogels as multifunctional platforms that present robust T₁-weighted contrast, offering promising avenues for future biomedical applications.

Supplementary Material

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsabm.5c00584.

Materials, further methods, and characterizations (ICP-MS, hydrogel methods, relaxivity, rheology, cell studies, synthesis, and small molecule characterization) (PDF)