Bimodal Imaging of Mouse Peripheral Nerves with Chlorin Tracers

Abstract

Almost 17 million Americans have a history of cancer, a number expected to reach over 22 million by 2030. Cancer patients often undergo chemotherapy in the form of antineoplastic agents such as cis-platin and paclitaxel. Though effective, these agents can induce debilitating side effects; the most common neurotoxic effect, chemotherapy-induced peripheral neuropathy (CIPN), can endure long after treatment ends. Despite the widespread and chronic nature of the dysfunction, no tools exist to quantitatively measure chemotherapy-induced peripheral neuropathy. Such a tool would not only benefit patients but their stratification could also save significant financial and social costs associated with neuropathic pain. In our first step toward addressing this unmet clinical need, we explored a novel dual approach to localize peripheral nerves: Cerenkov luminescence imaging (CLI) and fluorescence imaging (FI). Our approach revolves around the targeting and imaging of voltage-gated sodium channel subtype Na_V1.7, highly expressed in peripheral nerves from both harvested human and mouse tissues. For the first time, we show that Hsp1a, a radiolabeled Na_V1.7-selective peptide isolated from Homoeomma spec. Peru, can serve as a targeted vector for delivering a radioactive sensor to the peripheral nervous system. In situ, we observe high signal-to-noise ratios in the sciatic nerves of animals injected with fluorescently labeled Hsp1a and radiolabeled Hsp1a. Moreover, confocal microscopy on fresh nerve tissue shows the same high ratios of fluorescence, corroborating our in vivo results. This study indicates that fluorescently labeled and radiolabeled Hsp1a tracers could be used to identify and demarcate nerves in a clinical setting.

Graphical Abstract

INTRODUCTION

Almost 17 million Americans have a history of cancer, a number expected to reach over 22 million by 2030.¹ Of the available treatment options, chemotherapeutics in the form of antineoplastic agents are the gold standard of care for patients.^1,2 Although these drugs have been shown to be highly effective in arresting the progression of different cancers, they come with inherent drawbacks and side effects. The most common neurotoxic and therefore dose-limiting side effect of these agents is chemotherapy-induced peripheral neuropathy (CIPN), with up to 85% of patients self-reporting symptoms in the form of tingling, burning, cramps, numbness, and/or acute pain, which is the standard diagnosis for CIPN.³ For patients suffering from CIPN, treatments often continue to rely on subjective assessments of these symptoms, which can last for months or even years after treatment.^4–6

The early identification of nerves affected by chemotherapy could provide tremendous benefits, including the potential to stratify individuals early, before severe neuropathy develops. In response to this shortfall, researchers have sought to identify distinct molecular and cellular changes associated with specific pain pathologies, but this research is still challenging.^7,8

Even though a number of studies have investigated targets and methods to enhance visualization of the peripheral nervous system,^9–12 relatively few cases have successfully achieved this goal by means of attaining high specific uptake or high target-to-background ratios. In 2011, the groups of Frangioni and Nguyen reported the use of fluorescently tagged nerve tracers, namely, GE3082 and FAM-NP41, for the visualization of sciatic nerves.^13,14 In 2017, the group of Gibbs reported a dual fluorophore tissue staining method that reduced fluorescence background signals and aided in nerve-specific visualization.¹⁵ Parallel to this work, the Chin lab reported the use of [¹⁸F]FTC-146 for the visualization of nerve injuries in a neuropathic pain model via targeting of the sigma-1 receptor.¹⁶ In 2013, Du Bois and colleagues reported the use of ¹⁸F-labeled saxitoxin for the imaging of neuromas in a rat model of spared-nerve injury.¹⁷ While this represented the first small-molecule positron emission tomography (PET) imaging agent capable of targeting voltage-gated sodium channels, the reagent lacked specificity for a particular sodium channel subtype.

Sodium channels play an important role in the formation and conduction of action potentials in excitable tissues and are involved in a variety of physiological processes. In response to membrane depolarization, these ~260 kDa transmembrane proteins undergo conformational changes that allow Na⁺ ions to pass down an electrochemical gradient into the cell.^18,19 The scientific community has taken a growing interest in the voltage-gated sodium channel subtype Na_V1.7 as potential targets for analgesics.^20,21 Na_V1.7 is expressed on peripheral neurons and has shown abundant expression on both the transcription and protein levels, with increasing levels in patients suffering from peripheral neuropathy.²² This, coupled with numerous findings demonstrating the relationship between Na_V1.7 and peripheral pain, prompted us to postulate that Na_V1.7 represents a viable imaging target in patients with CIPN. Thus, and in an effort to find a medium through which Na_V1.7 could be targeted with high specificity, we turned our attention to labeled peptides as potential vectors for delivering probes for the noninvasive imaging of the peripheral nervous system.²³

In 2019, we reported the use of Hsp1a-FL, a fluorescently labeled cystine knot peptide, for the imaging of both mouse and human peripheral nerves.²⁴ This proof-of-concept work demonstrated that Hsp1a, a peptide isolated from the venom of the Peruvian tarantula Homoeomma spec., is an ideal anchor through which fluorescent sensors may be delivered to the peripheral nervous system. Inspired by these initial findings, we hypothesized that Hsp1a may also serve as a viable candidate for delivering radiolabeled imaging agents to the peripheral nervous system.

To this end, we demonstrate in this study for the first time the conjugation of Hsp1a peptide to a fluorescent synthetic chlorin,^25–27 while retaining specific binding to Na_V1.7 in in vivo mouse models. This retention in binding potency was validated in human peripheral nerve biospecimens. Conveniently, radiolabeling with copper-64 allows for the Cerenkov luminescence imaging of mouse sciatic nerves in vivo. In so doing, this work not only expands our growing portfolio of peptide-based imaging agents and also serves as a platform through which we may seek to explore noninvasive imaging of CIPN in vivo but it also adds a brand new attribute/application of synthetic chlorins in the radiochemical field. Concisely, we showed that spider peptides can target and deliver radio-labeled/fluorescent substances to peripheral nerves, which facilitates the imaging and differentiation of sciatic nerves.

EXPERIMENTAL SECTION

General.

Unless otherwise stated, all solvents and reagents were obtained from Sigma-Aldrich or Fisher Scientific and used without further purification. Chlorin-FL (ChL-FL) was prepared following our previous report.²⁷ Acetonitrile (ACN) and water (H₂O) were of high-performance liquid chromatography (HPLC) grade and of liquid chromatography mass spectroscopy (LCMS) grade, respectively. Phosphate-buffered saline (PBS) without Ca²⁺ or Mg²⁺ was obtained from the Media Preparation Facility at Memorial Sloan Kettering Cancer Center and used for all in vivo injections. Reverse-phase (RP) HPLC purifications were performed on a Shimadzu HPLC system equipped with a DGU-20A degasser, an SPD-M20A UV detector, an LC-20AB pump system, and a CBM-20A communication BUS module using RP-HPLC columns (Atlantis T3 C18, 5 μm, 4.6 × 250 mm, P/N:186003748). Electrospray ionization mass spectroscopy (ESI-MS) spectra were recorded with a Waters Aquity UPLC (Milford, CA) with an electrospray ionization SQ detector. ⁶⁴Cu was produced at the Mallinckrodt Institute of Radiology (Washington University) on a 19 MeV-beam energy cyclotron (Advanced Cyclotron Systems Inc. British Columbia, Canada) via the ⁶⁴Ni(p,n)⁶⁴Cu nuclear reaction to yield ⁶⁴Cu with a specific activity of 298–1830 mCi/μg (978mCi/μg avg.). For the radiochemical HPLC analysis, an RP-HPLC column (Atlantis T3 C18, 5 μm, 4.6 × 250 mm, P/N:186003748) was used with a gradient of 5–95% B over 20 min and then 95% B for 10 min [A: 0.1% trifluoroacetic acid (TFA), B: 99.9% ACN/0.1% TFA]. Radio-HPLC detection was carried out on an Agilent 1100 series HPLC system (Agilent Technologies, Stockport, U.K.) equipped with a g-RAM Model 3 γ-detector (IN/US Systems Inc., Florida), and Laura 3 software (LabLogic, Sheffield, U.K.) was used for processing all analytical HPLC chromatograms. Cherenkov imaging and epifluorescence imaging were performed on an IVIS Spectrum (PerkinElmer). The radioactivity of organs for biodistribution was counted with a WIZARD² automated γ-counter from PerkinElmer, and the histology slides of 10 μm of human vagus nerves were obtained as a donation from the Fusion Solutions Bioskills Laboratory autopsy program. Confocal microscopy images were captured using a Leica SP8 inverted-stand confocal microscope equipped with a tunable white light laser with a 470–670 nm range. The microscope was also equipped with a 405 nm diode for detection of Hoechst 33342, an argon laser (with 476, 488, 496, and 514 nm laser lines), and a 610 nm laser for NIR imaging coupled with avalanche photodiode detectors (APDs), which were used for detection of ChL-Hsp1a. A Lumar fluorescence stereoscope (SteREO Luma.V12, Zeiss, Jena, Germany) was used to obtain the fluorescence stereoscopic images.

Synthesis of Hsp1a.

As previously reported by our team, Hsp1a was synthesized and oxidized on a CEM Liberty Prime microwave peptide synthesizer (CEM corporation, NC) on rink-amide polystyrene resin to afford an amidated C-terminal.²⁴ The peptide was simultaneously released from the resin and the side-chain protecting groups were removed using trifluoroacetic acid (TFA)/triisopropyl-silane (TIPS)/water (48:1:1 v/v/v) for 2.5 h. After trituration in chilled diethyl ether (Et₂O) of the crude Hsp1a, the precipitated peptide was subjected to separation using solvent A/B (45% v/v acetonitrile, 0.05% v/v TFA), lyophilized, and then prepurified using C18 RP-HPLC. The pure peptide was eluted using a linear gradient of 10–60% solvent B (90% v/v ACN; 0.05% v/v TFA) over 50 min at a flow rate of 8 mL·min⁻¹. Oxidation of free cysteines took place at room temperature for 16 h in a buffer containing 2 M urea, 0.1 M Tris pH 8, 0.15 mM reduced glutathione, and 0.3 mM oxidized glutathione. A single sharp peak was obtained from the final analytical RP-HPLC purification and 96% purity was achieved as calculated from area under the curve. LC-ESI-MS (ES⁺), m/z calculated for the Hsp1a peptide, [C₁₄₈H₂₂₀N₄₀O₄₀S₆] 3389.52, [C₁₄₈H₂₂₀N₄₀O₄₀S₆ + 2H]²⁺ 1696.76, found [M + 2H]²⁺ 1696.00, [C₁₄₈H₂₂₀N₄₀O₄₀S₆ + 3H]³⁺ 1130.84, found [M + 3H]³⁺ 1131.05, [C₁₄₈H₂₂₀N₄₀O₄₀S₆ + 4H]⁴⁺ 848.38, found [M + 4H]⁴⁺ 848.55, [C₁₄₈H₂₂₀N₄₀O₄₀S₆ + 5H]⁵⁺ 678.90, found [M + 5H]⁵⁺ 679.05. LC-ESI-MS (ES⁺), m/z calculated for ChL-Hsp1a, [C₁₉₈H₂₃₆F₂₀N₄₆O₄₁S₆] 4485.58, [C₁₉₈H₂₃₆F₂₀N₄₆O₄₁S₆ + 3H]³⁺ 1496.19, found [M + 3H]³⁺ 1497.23, [C₁₉₈H₂₃₆F₂₀N₄₆O₄₁S₆ + 4H]⁴⁺ 1122.40, found [M + 4H]⁴⁺ 1123.03.

Synthesis of ChL-Hsp1a.

We chose to modify Hsp1a via nucleophilic substitution, analogous to our previous work with similarly sized bioactive peptides.^24,28 ChL-Hsp1a was synthesized by conjugating ChL-NHS ester to Hsp1a peptide (0.37 mM, 250 μg in 200 μL of ACN) in a solution of Na₂CO₃ (1 M, 40 μL). ChL-NHS (50 μL of a 0.82 mM solution) was dissolved in ACN and added dropwise to the Hsp1a in the reaction mixture, which was allowed to react for 10 min. The product, ChL-Hsp1a, was purified using RP-HPLC. Excess solvent was removed in vacuo, yielding ChL-Hsp1a as a green powder (80 μg, 24% isolated yield). The final analysis of the RP-HPLC purification showed 96% purity.

Tryptic Digestion of ChL-Hsp1a and Hsp1a.

A solution (10.5 μL) containing 4 μg of ChL-Hsp1a was added to a 0.5 mL amber microcentrifuge tube containing 15 μL of digestion buffer and 1.5 μL of reducing buffer. The final volume was adjusted to 27 μL with ultrapure water for the reaction. The sample was then incubated at 95 °C for 5 min. The sample was allowed to cool down to room temperature before 2 μL of activated trypsin was added, followed by incubation at 37 °C overnight. The sample was analyzed with LC-ESI-MS.

LC-ESI-MS (ES⁺), m/z, calculated for the fragmentation of ChL-Hsp1a was observed as follows: [C₁₆₂H₁₈₃F₂₀N₃₆O₃₅S₅] 3732.20, one calculated fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅ + 2H]²⁺ 1867.10, found [M + 2H]²⁺ 1699.05, and a second fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅ + 3H]³⁺ 1245.07, found [M + 3]³⁺ 1132.95. LC-ESI-MS (ES⁻), m/z, calculated for the fragmentation of ChL-Hsp1a was observed as follows: [C₁₆₂H₁₈₃F₂₀N₃₆O₃₅S₅] 3732.20, one calculated fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅] 3732.20, one calculated fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅ + 2H]²⁺ 1867.10, found [M + 2H]²⁺ 1699.05, and a second fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅ + 3H]³⁺ 1245.07, found [M + 3]³⁺ 1132.95. LC-ESI-MS (ES−), m/z, calculated for the fragmentation of ChL-Hsp1a was observed as follows: [C₁₆₂H₁₈₃F₂₀N₃₆O₃₅S₅] 3732.20, one calculated fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅-2H]²⁻ 1864.10, found [M − 2H]²⁻ 1697.40, and a second fragment [C₁₆₂H₁₈₁F₂₀N₃₆O₃₅S₅ + OH]³⁻ 1864.10, found [M + OH]³⁻ 1139.90.

Hsp1a was digested following the same tryptic digestion protocol and the sample was analyzed with LC-ESI-MS. LC-ESI-MS (ES⁻), m/z, calculated for the fragmentation of Hsp1a was observed as follows: one calculated fragment [C₉₀H₁₃₈N₂₄O₂₇S₄ + 2H]²⁺ 138 1058.45, found [M + 2H]²⁺ 1059.55, and a second fragment [C₉₀H₁₃₈N₂₄O₂₇S₄ + 3H]³⁺ 705.97, found [M + 3H]³⁺ 706.65.

Synthesis of [⁶⁴Cu]Cu-ChL-Hsp1a.

Radiolabeling of ChL-Hsp1a (0.56 mM, 50 μg in 20 μL of DMSO) was performed using 3.8 μL of ⁶⁴Cu (111 MBq) in 500 μL of ammonium acetate buffer for 30 min at 37 °C. The [⁶⁴Cu]Cu-ChL-Hsp1a was trapped on a preconditioned C18 cartridge (ethanol/water), washed with H₂O to remove free copper-64, and then eluted with ethanol to obtain pure [⁶⁴Cu]Cu-ChL-Hsp1a. The specific activity of [⁶⁴Cu]Cu-ChL-Hsp1a throughout the study was measured to be between 4.52 and 9.51 MBq/μg. [⁶⁴Cu]Cu-ChL-Hsp1a was formulated with PBS for in vivo studies.

Instant Thin-Layer Chromatography (iTLC) Quantification.

The crude reaction mixture and purified labeled Hsp1a peptide solution were analyzed by iTLC (glass microfiber chromatography paper impregnated with silica gel, 100 10 mm) using 50 mM ethylenediaminetetraacetic acid (EDTA) (pH 5.5) as the mobile phase. For the control, [⁶⁴Cu]CuCl₂ (approx. 0.074 MBq) was mixed with ammonium acetate (0.1 M, pH 6.0) and incubated for 30 min at 37 °C. The same mobile phase was used for separation. The different iTLC plates were analyzed by iTLC and Laura 3 software (LabLogic, Sheffield, U.K.). [⁶⁴Cu]CuCl₂ R_f = 0.67–0.78; [⁶⁴Cu]Cu-ChL-Hsp1a R_f = 0.25–0.31.

Blood Half-Life of ChL-Hsp1a.

Female mice (n = 3) were injected through the tail vein with ChL-Hsp1a (1 nmol, 10 μM ChL-Hsp1a in 100 μL of PBS). Blood from animals was collected via retro-orbital eye bleeds at different time points after the injection. Blood (20–25 μL) was drawn for each time point using a heparinized capillary tube (length: 75 mm, wall thickness: 0.20 mm) and immediately imaged for its fluorescence using the IVIS (excitation = 610/645 and emission = 680/700 nm).

Human Tissue.

Human vagus nerves (human nerves, n = 6) were donated by the Fusion Solutions Bioskills Laboratory, Long Island, NY. Nerves were sectioned at 10 μm thickness for staining with ChL-Hsp1a. In addition, the nerves were paraffin-embedded, formalin-fixed, and sectioned at 10 μm thickness for hematoxylin and eosin (H&E) histology and immunohistochemistry (IHC).

Animal Model.

Female athymic nude mice (4–10 weeks old, athymic nude (outbred) (Stock#:069); ENVIGO RMS, INC.) were allowed to acclimatize at the Memorial Sloan Kettering Cancer Center (MSK) vivarium for 1 week with food and water available ad libitum prior to the experimental procedure. For epifluorescence and Cherenkov imaging experiments, animals were sacrificed 30 min after tail vein injection of either ChL-Hsp1a, [⁶⁴Cu]Cu-ChL-Hsp1a, Hsp1a/[⁶⁴Cu]Cu-ChL-Hsp1a, or PBS. All animal experiments were performed in accordance with institutional guidelines and approved by the IACUC of MSK, following NIH guidelines for animal welfare.

Immunohistochemistry.

Nerves donated by the Fusion Solution Bioskills Laboratory autopsy program were used in immunohistochemistry experiments. Na_V1.7 in human vagus nerves was detected using immunohistochemical (IHC) staining techniques at the Molecular Cytology Core Facility of MSK using the Discovery XT processor (Ventana Medical System, Tucson, AZ). Anti-Na_V1.7 antibody [N68/6] (Abcam ab85015) specifically bound to both human and mouse Na_V1.7 (0.5 μg/mL). Paraffin-embedded formalin-fixed 10 μm sections were deparaffinized with EZPrep buffer. For IHC detection, a 3,3′-diaminobenzidine (DAB) detection kit (Ventana Medical Systems, Tucson, AZ) was used according to the manufacturer’s instructions. In addition, sections were counterstained with hematoxylin and eosin (H&E) staining techniques and coverslipped with Permount (Fisher Scientific, Pittsburgh, PA).

Confocal Microscopy.

Staining with ChL-Hsp1a was performed in both mouse sciatic nerves and human vagus nerves. For mice, 10 μm cryosections of OCT-embedded sciatic nerve tissues from mice previously injected with ChL-Hsp1a (1 nmol, 10 μM ChL-Hsp1a in 100 μL of PBS), Hsp1a/ChL-Hsp1a (ChL-Hsp1a, 10 μM, 1 nmol and Hsp1a, 204 μM, 21 nmol in 100 μL of PBS), or PBS were used. Tissues were counterstained with Hoechst 33342 (blue, 20 μM, 1 nmol in 50 μL of PBS) up to 90 min postmortem and placed directly on a microscope slide for imaging.

For human tissue, 10 μm cryosections of vagus nerve were immersed in ChL-Hsp1a (3 nmol, 30 μM ChL-Hsp1a in 100 μL of PBS), Hsp1a/ChL-Hsp1a (ChL-Hsp1a, 30 μM, 3 nmol and Hsp1a, 90 μM, 9 nmol in 100 μL of PBS) or PBS. Tissues were incubated with Hoechst 33342 (blue, 20 μM, 1 nmol in 50 μL of PBS) to counterstain nuclei. For human vagus nerve tissue, to obtain a fluorescence signal, the specimen was incubated for at least 2 min, followed by one PBS wash cycle. For the block experiment, nerve tissue was first immersed in Hsp1a solution for at least 5 min, followed by immersion in ChL-Hsp1a for at least 2 min, followed by one PBS wash cycle.

Epifluorescence Imaging.

Animals were intravenously injected with ChL-Hsp1a (1 nmol, 10 μM ChL-Hsp1a in 100 μL of PBS, n = 6). To assess the specificity of the ChL-Hsp1a accumulation, we injected a combination of Hsp1a and ChL-Hsp1a (ChL-Hsp1a, 10 μM, 1 nmol, and Hsp1a, 204 μM, 21 nmol, in 100 μL of PBS, n = 6) or PBS (n = 9). Animals were sacrificed 30 min postinjection and epifluorescence images were obtained. Epifluorescence images of right sciatic nerves (RSN) and left sciatic nerves (LSN) were obtained first in situ and then ex vivo, together with excised muscle, heart, kidney, liver, and brain tissue using an IVIS Spectrum (PerkinElmer) with a predefined filter set (excitation = 610–645 nm, emission = 650–680 nm). Autofluorescence was removed through spectral unmixing. Semiquantitative analysis of the ChL-Hsp1a signal was conducted by measuring the average radiant efficiency (in units of [p/s/cm²/sr]/[μW/cm²]) in regions of interest (ROIs) that were drawn on all resected organs under white light guidance.

Fluorescence Stereoscope Imaging.

The fluorescence stereoscopic signal in mouse peripheral nerves was also detected visualized using a fluorescence stereoscope approximately 30 min after an intravenous injection of ChL-Hsp1a (1 nmol, 10 μM ChL-Hsp1a in 100 μL of PBS) or PBS (n = 3/group). Fluorescence images were obtained using mice with surgically exposed nerves from the upper body and lower body. Imaging was performed in bright field and fluorescence modes, with a 640–650 nm laser excitation and a 680–700 emission filter, and an exposure time of 100–300 ms.

Cerenkov Luminescence Imaging.

Animals were intravenously injected with [⁶⁴Cu]Cu-ChL-Hsp1a (3.5–4.2 MBq in 200 μL of PBS, n = 9). To assess the specificity of the [⁶⁴Cu]Cu-ChL-Hsp1a accumulation, we injected a combination of Hsp1a, [⁶⁴Cu]Cu-ChL-Hsp1a (3.5–4.2 MBq), and Hsp1a (204 μM, 21 nmol in 200 μL of PBS, n = 6) or PBS (n = 9). Animals were sacrificed 30 min postinjection and CLI was performed. CLI of RSN and LSN was performed first in situ with a surgical cut to expose the sciatic nerves, and then CLI of the resected RSN and LSN was performed for comparison. CLI of the excised RSN, LSN, muscle, heart, kidney, liver, and brain was also performed with an IVIS Spectrum (PerkinElmer). Semiquantitative analysis of the [⁶⁴Cu]Cu-ChL-Hsp1a signal was conducted by measuring the radiant efficiency in regions of interest (ROIs) that were placed on all resected organs under white light guidance. Moreover, the radioactivity of organs for biodistribution was counted with a WIZARD² automated γ-counter from PerkinElmer.

Animals and Monitoring.

Animals were monitored before and after the intravenous injection of ChL-Hsp1a (3 nmol, 30 μM ChL-Hsp1a in 100 μL of PBS) or PBS in female athymic nude mice, n = 3. Three other mice were used as controls, n = 3. An intravenous catheter was placed in each mouse’s tail vein, and they were further anesthetized using Isoflurane (Novaplus, Telangana—India). The anesthesia plane was kept using a constant amount of isoflurane (1 to 1.5 L/min) and oxygen (2 L/min). Animals were further placed in a heated platform maintained at 39 °C and monitored using a rodent surgical monitor (Scintica, Instrumentations, Houston). Numeric data points were recorded before injection and at the following time points after injection: immediately after, 5, 10, and 15 min. The mouse’s body core temperature was measured by placing a rectal probe. The high-resolution electrocardiogram (EKG) was acquired by placing the four paws at the platform electrodes. Electrical contact was assured using a conducting gel (Electrode cream, Indus instruments, Huston). Heart rate was calculated from the R–R peaks derived from the EKG signal. The peripheral oxygen saturation (SpO2) was measured by placing a clip sensor on the animal’s right thigh.

Statistical Analyses.

Statistical analyses were performed using GraphPad Prism 8. Unless otherwise stated, data points represent mean values, and error bars represent standard deviations of biological replicates. P values were calculated using a Student’s unpaired t-test.

RESULTS

Design of the Fluorescent-Labeled Hsp1a Conjugate ChL-Hsp1a.

An NHS-ester modified chlorin (ChL) (ex_max = 610 nm, em_max = 660 nm) was conjugated to the spider-derived venom peptide Hsp1a²⁴ (Figure 1A). The reaction was performed under basic conditions in a solution of water and acetonitrile, yielding ChL-Hsp1a in 25% yield and 94% purity. The retention time (t_R) changed from 21.0 min for Hsp1a to 32.0 min for ChL-Hsp1a (Figures 1B and S1A). The major impurity was the partially reduced peptide, 11% (t_R 32.2 min), which was also present in the starting material (t_R 21.0 min, 90%, and r_t 21.2 min, 10%, for Hsp1a and reduced Hsp1a, respectively). LC–MS spectra for both Hsp1a and ChL-Hsp1a showed clean peak families, confirming the calculated masses of the peptides, 3389 and 4488 Da, for Hsp1a and ChL-Hsp1a, respectively (Figures 1C and S1B). The absorption and emission spectra of Hsp1a and ChL-Hsp1a were also determined, and the characteristic 650 nm peak for the chlorin family was used as a reference and it was observed for ChL-Hsp1a at 652 and 660 nm for absorption and emission, respectively; these bands overlapped with the peaks of the ChL-NHS ester, but not for the Hsp1a peptide alone (Figure 1D,E). In addition, more bands were detected at 505 and 598 nm, which confirmed the conjugation of the chlorin moiety to Hsp1a. Exclusive features of ChL-FL and ChL-Hsp1a comprising absorbance and emission maxima, individual gradients for purity, and obtained yields are presented in Figure 1F.

Figure 1.

Conjugation and characterization of ChL-Hsp1a. (A) Reaction scheme for the conjugation of the Hsp1a peptide with ChL-NHS ester dye. The ribbon model of ChL-Hsp1a shows disulfide bridges (in orange) and the covalent conjugation of the chlorin moiety. (B) RP-HPLC chromatograms of the Hsp1a peptide (black) and of the fluorescent tracer ChL-Hsp1a (orange) at 280 nm absorption. (C) LC–MS trace of ChL-Hsp1a, showing three major ion species that correspond to the calculated mass of the fluorescent Hsp1a venom. (D) Absorbance spectra of 0.1 μM ChL-FL (black), 0.1 μM ChL-Hsp1a (orange), and 0.1 μM Hsp1a (blue) observed from 400 to 800 nm, and where the classical 650 nm chlorin peak was observed. (E) Fluorescence spectra of 0.1 μM ChL-FL (black), 0.1 μM ChL-Hsp1a (orange), and 0.1 μM Hsp1a (blue) observed at a range of 625–700 nm, and an excitation wavelength of 610 nm. (F) Exclusive features of ChL-FL and ChL-Hsp1a comprising absorbance and emission maxima, individual gradients for purity, and obtained yields.

Chemical Affinity and Target Binding of the ChL-Hsp1a Conjugate.

To determine the location of the chlorin moiety conjugation to the Hsp1a peptide, a tryptic digestion experiment was performed. For the unmodified Hsp1a, we found ions that correspond to the fragment F5-R22 (2113.88 Da, Figure S1C). The aforementioned fragment was not seen for digested ChL-Hsp1a; instead, the conjugated peptide gave rise to the novel fragment Y1-R22 plus the mass of the chlorin scaffold minus two pentafluoro phenyls C₁₂F₁₀ (3398.22 Da, Figure S1D). Observation of this peak suggests that conjugation occurred at K4 of the Hsp1a. If the chlorin group were not conjugated at K4 of the Hsp1a, the peptide digestion would have generated fragment F5-R22 and the same ion species with mass 2113.88 would have been observed as in the digestion of the unmodified Hsp1a. In principle, Hsp1a could have been modified at three nucleophilic positions (the N-terminal amine, K4, and K26); in this case, we did not observe conjugation at K26 or at the N-terminal amine. We presume that the predominant modification of Hsp1a at K4 is likely electronically or sterically favored.

Hsp1a Conjugate [⁶⁴Cu]Cu-ChL-Hsp1a.

With ChL-Hsp1a in hand, chelation to copper-64 was carried out under acidic aqueous conditions (NH₄OAc, pH 5.5) at room temperature for about 30 min to yield [⁶⁴Cu]Cu-ChL-Hsp1a (Figure 2A). The reaction was carried out with a radiochemical conversion rate of more than 70%. The product was obtained in >95% radiochemical purity and in 40–50% isolated radiochemical yield (ndc), which corresponded to maximum molar activities of 0.1–0.2 GBq·μmol⁻¹. The reaction was monitored by instant thin-layer chromatography (iTLC) (Figure S1E) and purified by the C18 cartridge. The isolated product was confirmed by iTLC and RP-HPLC (Figure 2B). With a C18 column, the retention time for ChL-Hsp1a (observed at 280 nm) and [⁶⁴Cu]Cu-ChL-Hsp1a was 21 min in both cases. For the successful synthesis of [⁶⁴Cu]Cu-ChL-Hsp1a, two requirements had to be met: first, the ligand must have the ability to chelate ⁶⁴Cu, and second, the radiotracer as a whole must maintain the properties of the Hsp1a peptide that were also observed in subsequent experiments. Furthermore, the exclusive features of Hsp1a, ChL-Hsp1a, and [^natCu]Cu-ChL-Hsp1a such as emission and absorbance maxima, gradients for purity of peptides, modifications, and observed ions are presented in Figure 2C.

Figure 2.

Radiolabeling and characterization of [⁶⁴Cu]Cu-ChL-Hsp1a. (A) Radiochemical synthesis for the labeling of the ChL-Hsp1a peptide with ⁶⁴Cu. (B) Radio-HPLC chromatogram of a labeled [⁶⁴Cu]Cu-ChL-Hsp1a at 37 °C for 30 min with a corresponding HPLC chromatogram of the cold ChL-Hsp1a standard at 280 nm absorption. (C) Amino acid sequences, individual gradients, molecular weights, and observed ions for the unmodified Hsp1a peptide, the fluorescent tracer ChL-Hsp1a, and [^natCu]Cu-ChL-Hsp1a.

Pharmacokinetics of ChL-Hsp1a and [^natCu]Cu-ChL-Hsp1a.

Under mesoscopic imaging conditions, we observed rapid and selective accumulation of ChL-Hsp1a within the peripheral nerves of mice (Figures 3 and S2A). Mice were injected intravenously with ChL-Hsp1a (1 nmol, 10 μM ChL-Hsp1a in 100 μL of PBS) or ChL-Hsp1a in combination with an excess of the unmodified Hsp1a peptide, block formulation (ChL-Hsp1a, 10 μM, 1 nmol and Hsp1a, 21 nmol, 204 μM, 21 nmol in 100 μL of PBS), and sacrificed 30 min after tail vein injection. The right and left sciatic nerves were then exposed and epifluorescence imaging was performed using an IVIS Spectrum in vivo imaging system (ex: 600/20 nm; em: 650–670 nm). In mice receiving just the imaging agent, the sciatic nerves were clearly visible, whereas in mice receiving the imaging agent in combination with the unmodified peptide, uptake was significantly reduced (radiant efficiency: 0.7 ± 0.4 × 10⁷ and 0.3 ± 0.1 × 10⁷ for ChL-Hsp1a and co-injection (blocking), respectively, Student’s unpaired t-test, P < 0.05, Figures 3A,B and S2B,C).

Figure 3.

ChL-Hsp1a accumulation in mouse sciatic nerves. ChL-Hsp1a and [⁶⁴Cu]Cu-ChL-Hsp1a. (A) Epifluorescence images of animals injected with 100 μL of PBS, ChL-Hsp1a (1 nmol, 10 μM ChL-Hsp1a in 100 μL of PBS), or a ChL-Hsp1a/Hsp1a formulation (ChL-Hsp1a, 10 μM, 1 nmol and Hsp1a, 204 μM, 21 nmol in 100 μL of PBS). Images were taken 30 min after tail vein injection. (B) Fluorescence intensity quantification of panel (A). (C) Epifluorescence images of resected right and left mouse sciatic nerves that were injected with ChL-Hsp1a or [⁶⁴Cu]Cu-ChL-Hsp1a. High fluorescence intensities (due to dye accumulation) were only observed in sciatic nerves injected with ChL-Hsp1a alone. No fluorescence was observed after 30 min in mice injected with [⁶⁴Cu]Cu-ChL-Hsp1a. (D) Fluorescence quantification comparison of animals injected with ChL-Hsp1a and [⁶⁴Cu]Cu-ChL-Hsp1a. Statistics were calculated with a nonparametric Student’s t-test. *P < 0.05; **P < 0.01; ***P < 0.001; and ****P < 0.0001.

This was also corroborated under ex vivo imaging conditions, where we found high fluorescence uptake for the sciatic nerves of animals injected with ChL-Hsp1a (Figure S2D). We found a mean radiant efficiency of 0.5 ± 0.1 × 10⁷, whereas in animals receiving the imaging agent in combination with the unmodified peptide (21 nmol, 204 μM in 100 μL of PBS), a statistically significant 93-fold reduction in radiant efficiency was observed (0.005 ± 0.001 × 10⁷ radiant efficiency, P < 0.005) (Figure S2E). We observed a trend toward higher fluorescence intensities in the livers of animals injected with the imaging agent only (radiant efficiency: 0.92 ± 0.14 × 10⁷ and 0.72 ± 0.19 × 10⁷ for livers injected with ChL-Hsp1a and Hsp1a/ChL-Hsp1a, respectively). A similar trend was observed for the kidneys (radiant efficiency: 0.2 ± 0.02 × 10⁷ and 0.7 ± 0.2 × 10⁷ for injected with ChL-Hsp1a and Hsp1a/ChL-Hsp1a, respectively). We did not observe any significant fluorescence in any other organs, including the muscle, heart, spleen, and brain, when comparing animals injected with ChL-Hsp1a and PBS (Figure S2E).

It is known from the literature that the fluorescence of several dyes can be quenched by paramagnetic Cu(II) ions.²⁹ To evaluate whether this is the case for ChL-Hsp1a, the fluorescence emission profiles of ChL-Hsp1a and [^natCu]Cu-ChL-Hsp1a were compared. Figure S3A demonstrates a clear dependence of emission upon Cu(II) complex formation, resulting in near-complete quenching after chelation. No significant fluorescence emission was detected when animals were injected with the radiolabeled Hsp1a peptide (mean radiant efficiency ex vivo: 0.0008 ± 0.0001 × 10⁷) (Figure 3C,D). Fluorescence quenching is likely a product of an energy-transfer process (intersystem crossing) from the excited electrons in the chlorin part to the Cu(II) metal center, for instance, Cu(II) metal, similar to what has been shown before.^30,31

Histological Correlation and Validation of ChL-Hsp1a.

In the past, we have shown that fluorescently labeled Hsp1a imaging agents highlight nerve structures on a mesoscopic level and at cellular resolution. Consistent with this, our experiments have confirmed physiological similarities of human vagus (Figures 4A,B, S4, and S5) and mouse sciatic nerves. For topical staining of human vagus nerves, the tissues were immersed in a solution of either ChL-Hsp1a (3 nmol, 30 μM ChL-Hsp1a in 100 μL of PBS), Hsp1a/ChL-Hsp1a (ChL-Hsp1a, 30 μM, 3 nmol and Hsp1a, 90 μM, 9 nmol in 100 μL of PBS), or PBS. Additionally, tissues were incubated with Hoechst 33342 (blue, 20 μM, 1 nmol in 50 μL of PBS) to counterstain nuclei of Schwann cells (Figure 4C). The fluorescence signal, after topical staining of human vagus nerves with ChL-Hsp1a, showed linear patterns and structures when nerves were sliced longitudinally, consistent with nerve physiology and likely representing axons. These patterns were not observed when ChL-Hsp1a was blocked by previous incubation with Hsp1a. Moreover, we performed a histological analysis on mouse sciatic nerves, which were injected with ChL-Hsp1a (1 nmol, 10 μM in 100 μL of PBS) or block formulation (ChL-Hsp1a, 10 μM, 1 nmol and Hsp1a, 204 μM, 21 nmol in 100 μL of PBS) before their sciatic nerves were removed and sectioned for imaging. Mouse nerves showed similar staining patterns, corroborating the specificity of ChL-Hsp1a when staining human nerves (Figure S6). Co-injection of Hsp1a prevented uptake of ChL-Hsp1a in sciatic nerves, suggesting that the probe was specific and its target was saturable.

Figure 4.

Anatomic representation of a human vagus nerve, Na_V1.7 expression in human vagus nerves, schematic structure of the skeleton of a human vagus nerve, and the histological validation of ChL-Hsp1a specificity in human vagus nerves after topical application. (A) Schematic drawing of the anatomic location of the human vagus nerve (yellow) within the head and neck region and a white light photograph of a surgically resected human vagus nerve autopsy specimen. Anti-Na_V1.7 (left bottom) and H&E (right bottom) immunohistochemical staining of a vagus nerve autopsy specimen, which shows high Na_V1.7 expression within the human peripheral nerve, showing specific staining and adjacent H&E staining sections. (B) Diagram of the structure of a human vagus nerve showing different topological layers of the nerve structure, from the epineurium to the endoneurium, from the superior part to the middle part of the nerve. (C) Human vagus nerve cryosections were stained with the fluorescent Hsp1a peptide (first and second columns, 3 nmol, 30 μM ChL-Hsp1a in 100 μL of PBS) or with a mixture of Hsp1a and ChL-Hsp1a (third column, ChL-Hsp1a, 30 μM, 3 nmol and Hsp1a, 90 μM, 9 nmol in 100 μL of PBS). Human vagus nerves were counterstained with Hoechst 33342 (blue, 20 μM, 1 nmol in 50 μL of PBS). The sectioning of the nerves shows epineurium to endoneurium slicing that can be better observed in composite images.

Pharmacokinetics of [⁶⁴Cu]Cu-ChL-Hsp1a.

For [⁶⁴Cu]Cu-ChL-Hsp1a, the uptake specificity in sciatic nerves was striking (Figure 5). In mice receiving the imaging agent alone (3.5–4.2 MBq in 200 μL of PBS), the sciatic nerves were clearly visible, with a mean radiant efficiency of 6830 ± 2434 p/s/cm²/sr (Figures 5A,B and S7A), whereas mice receiving the imaging agent in combination with the unmodified peptide (21 nmol, 204 μM in 100 μL of PBS) had a statistically significant value (Student’s unpaired t-test, P < 0.0001, 48-fold reduced 143 ± 58 radiant efficiency in vivo). While it should be noted that quantitative assessment of Cerenkov luminescence (CL) across organ systems is not feasible, the only site of off-target uptake both in vivo and ex vivo appears to be the liver (Figures 5C,D and S7B (bottom)). No notable uptake was observed in other organs, including the muscle, kidney, heart, and brain, for animals injected with the combination of [⁶⁴Cu]Cu-ChL-Hsp1a and unmodified Hsp1a (Figure S5B (middle)). Following this, acute biodistribution studies were performed and radioactivity associated with each organ was expressed as a percentage of injected dose per gram of tissue (% ID/g). This quantitative measure showed uptake in the sciatic nerves, liver, and spleen (Figure 5D). We speculate that the contrasting results of uptake in the spleen between CLI and the γ counter may be explained by the swallowing of the CL in dark, highly absorbing tissue.

Figure 5.

CL of [⁶⁴Cu]Cu-ChL-Hsp1a in mice, biodistribution, and quantification. (A) CL images of mice injected with Hsp1a and [⁶⁴Cu]Cu-ChL-Hsp1a (left) or [⁶⁴Cu]Cu-ChL-Hsp1a alone (right). High radiance is observed after 30 min in mice injected with [⁶⁴Cu]Cu-ChL-Hsp1a alone. (B) CL quantification of panel (A). (C) Representative luminescence image of sciatic nerves, muscle, spleen, heart, kidney, liver, and brain of mice injected with PBS and [⁶⁴Cu]Cu-ChL-Hsp1a. (D) Radiance quantification of all sciatic nerves injected with [⁶⁴Cu]Cu-ChL-Hsp1a or Hsp1a and [⁶⁴Cu]Cu-ChL-Hsp1a. Statistics were calculated with a parametric Student’s t-test. *P < 0.05; **P < 0.01; ***P < 0.001; and ****P < 0.0001.

In Vivo Circulation of ChL-Hsp1a.

To observe the circulation of the fluorescent peptide in vivo, we have determined blood half-life of ChL-Hsp1a using an in vivo imaging system. After withdrawing blood from animals injected with ChL-Hsp1a, the blood half-life was determined to be approximately 6 min (Figure S8). After about 2 h, the ChL-Hsp1a appeared to wash out from the blood of mice slowly.

Surgical Microscopy Imaging of ChL-Hsp1a.

To corroborate the feasibility of ChL-Hsp1a fluorescent imaging in a more clinically relevant setting, we chose to assess our contrast agent with a Lumar surgical fluorescence stereoscope (SteREO Lumar.V12, Zeiss, Jena, Germany). Analogous to our previous work, mice were injected with ChL-Hsp1a (1 nmol, 10 μM in 100 μL of PBS) or 100 μL of PBS. For this experiment, after 30 min, mice were sacrificed and their peripheral nerves were exposed for imaging. Figure S9 shows images of the upper and lower body nerves of animals obtained under white light imaging conditions (top row) and ChL-Hsp1a fluorescence (bottom row) for both mice injected with ChL-Hsp1a and PBS. Conveniently, the combination of ChL-Hsp1a and the Lumar stereoscope could thus potentially be used clinically for concomitant mapping and intraoperative procedures.

Toxicity of ChL-Hsp1a.

To assess the toxicity of our ChL-Hsp1a, we chose to monitor the vital signs of mice after the intravenous injection of ChL-Hsp1a (3 nmol, 30 μM ChL-Hsp1a in 100 μL of PBS) or PBS with a rodent surgical monitor (Scintica, Instrumentations, Houston). For this experiment, mice’s oxygen levels, heart rates, and body temperatures were monitored for at least 15 min. Figure S10 shows no abnormal changes and behavior in the animals injected with PBS and ChL-Hsp1a; this confirms that there are no safety concerns observed in these animals after the injection of the fluorescent agent.

DISCUSSION

In 2019, our group reported the imaging agent Hsp1a-FL, a fluorescently labeled congener of the peptide Hsp1a. By targeting the sodium channel subtype, Na_V1.7, we demonstrated Hsp1a-FL to be a viable sensor for the imaging of the peripheral nervous system.²⁴ Within the context of intraoperative imaging, such an agent would be highly useful, given the current clinical shortfalls. However, the fluorescent BODIPY tag used in this report presents limitations, not allowing penetration due to visible light absorption.³² Within the context of CIPN, such an agent would not be ideal.³³ We therefore set about developing an Hsp1a imaging agent that would allow for noninvasive imaging.

Of the imaging modalities available for noninvasive imaging, positron emission tomography (PET) has established itself as one of the most powerful tools in this field. To date, several studies have reported the potential of ¹⁸F-FDG as a tool for imaging peripheral and chronic pain.^34,35 However, for cancer patients, such a strategy is particularly challenging, given that ¹⁸F-FDG often serves as the gold standard for whole-body imaging of many cancers, limiting its effectiveness in visualizing peripheral nerves.

In light of this, and given the high targeting specificity of Hsp1a for Na_V1.7, we anticipated that a radiolabeled derivative of Hsp1a could serve as a powerful tool for the selective targeting and imaging of the peripheral nervous system, especially for patients with CIPN. The use of Hsp1a-FL to image sciatic nerves Hsp1a-FL to image sciatic nerves in a mouse model and in human vagus nerves²⁴ represented an ideal model to pursue. However, we encountered a significant drawback in the use of PET: given the size of sciatic nerves in mice, the partial volume effect and rapid static formation of nerves would prevent precise quantification of the uptake of radiolabeled Hsp1a within the nerves.³⁶ To circumvent this problem, we selected a positron-emitting radionuclide that would be amenable to CLI.^37,38 This would allow us to maintain the same mouse model for both in vivo imaging and ex vivo biodistribution studies. To this end, and given the plethora of chelators available, we selected ⁶⁴Cu as a suitable radionuclide for initial investigations.³⁹

In line with our previous experience with Hsp1a, we set about devising a strategy that would proceed via modification of a lysine residue, which ultimately would take the form of amide bond formation.²⁴ For this, we selected a ChL-NHS ester derivative for conjugation. Chlorin is not only amenable for chelation to copper but is also a known fluorophore⁴⁰ and as such would allow for in vivo quantification to be performed prior to radiolabeling to observe any possible loss of specificity for the peripheral nervous system.

Synthetic chlorins have multiple applications; in this work, we show the insertion (chelation) of copper-64 into the reaction center of a chlorin. Conjugation of ChL-NHS (ex_max = 610 nm, em_max = 660 nm) to Hsp1a was successfully carried out, affording ChL-Hsp1a in 24% isolated yield. Both absorbance spectra and fluorescence spectra (Figure 1D,E) of ChL-Hsp1a and ChL-NHS demonstrated that conjugation did not interfere with the photochemical properties of both the peptide and the chlorin. In addition, we show that the conjugation (insertion of chlorin) occurs at the lysine (K4) position of the Hsp1a peptide, demonstrating chemical affinity. ChL-Hsp1a’s toxicity was monitored by observation of vital signs of mice injected with the contrast agent. When injected into nude mice, ChL-Hsp1a showed high uptake in the sciatic nerves after 30 min, as confirmed by IVIS imaging and ex vivo biodistribution. The aforementioned finding was corroborated with stereoscopic images that showed ChL-Hsp1a fluorescent signals at the upper and lower body nerves of mice. Specific uptake of ChL-Hsp1a was confirmed using an excess of the unlabeled Hsp1a. In so doing, this confirmed that the chlorin tether did not significantly perturb the binding of Hsp1a to its target, Na_V1.7. In line with our previous studies, histological evaluation of ChL-Hsp1a in human vagus nerves was carried out (Figure 4). Not only did these results confirm specific uptake in ChL-Hsp1a, but they also substantiated our previous claim that mouse models may serve as suitable foundations for the future development of Na_V1.7 targeting agents.

Following this work, we proceeded to radiolabel ChL-Hsp1a and evaluate it in vivo. Using [⁶⁴Cu]CuCl₂, chelation to ChL-Hsp1a was carried out in a slightly acidic buffer (Figure 2A). After monitoring by iTLC, the identity of the product was confirmed by reverse-phase HPLC (Figure 2B). Similar to ChL-Hsp1a, [⁶⁴Cu]Cu-ChL-Hsp1a showed high specific uptake in the sciatic nerves of mice. In vivo, uptake was quantified through the use of CLI. This was supported by ex vivo analysis of the key organs. Biodistribution studies were carried out using a γ counter and the resulting % ID/g were correlated to radiance values from ex vivo CLI (Figure 5D). The correlation observed here demonstrates that CLI may serve as a powerful alternative to whole-body in vivo imaging of Na_V1.7 agents in mouse models.

In summary, we report the synthesis of two novel agents for the targeting and imaging of Na_V1.7. Using a chlorin-derived ligand, both fluorescent and CLI imaging agents show high specific binding in in vivo mouse models. Taken together, these results suggest that [⁶⁴Cu]Cu-ChL-Hsp1a and ChL-Hsp1a may be further pursued for the imaging of peripheral nerves. Future research will seek to expand this work into larger animal models, thus allowing for in vivo quantification to be carried out with PET. [⁶⁴Cu]Cu-ChL-Hsp1a will be evaluated for its ability to act as a noninvasive imaging agent for CIPN in response to antineoplastic agents including cis-platin and paclitaxel.

CONCLUSIONS

The data presented herein shows that ChL-Hsp1a and [⁶⁴Cu]Cu-ChL-Hsp1a peptides accumulated and highlighted peripheral nerves in mice in vivo. The statistically significant fluorescence/Cerenkov signal observed in peripheral nerves in mice, as a result of using fluorescent/PET active Hsp1a peptides, is clear evidence of peripheral nerve localization. This fluorescent/PET approach could be especially beneficial to stratify patients, as treatments often continue to rely on subjective assessments of these symptoms by physicians after patients’ self-reporting. ChL-Hsp1a and [⁶⁴Cu]Cu-ChL-Hsp1a peptides are promising tools that could potentially help physicians to rapidly identify peripheral neuropathy in patients.

ABBREVIATIONS

CLI

Cerenkov luminescence imaging

FI

fluorescence imaging

CIPN

chemotherapy-induced peripheral neuropathy

Na_V1.7

voltage-gated sodium channel subtype 1.7

Hsp1a

Homoeomma spec. Peru

IHC

immunohistochemistry

H&E

hematoxylin and eosin stain

HPLC

high-performance liquid chromatography

ChL-Hsp1a

fluorescence Hsp1a-chlorin

[⁶⁴Cu]Cu-ChL-Hsp1a

radioactive Hsp1a

DMSO

dimethyl sulfoxide