Norcyanine-Carbamates Are Versatile Near-Infrared Fluorogenic Probes

Abstract

Fluorogenic probes in the near-infrared (NIR) region have the potential to provide stimuli-dependent information in living organisms. Here, we describe a new class of fluorogenic probes based on the heptamethine cyanine scaffold, the most broadly used NIR chromophore. These compounds result from modification of heptamethine norcyanines with stimuli-responsive carbamate linkers. The resulting cyanine carbamates (CyBams) exhibit exceptional turn-ON ratios (~170×) due to dual requirements for NIR emission: carbamate cleavage through 1,6-elimination and chromophore protonation. Illustrating their utility in complex in vivo settings, a γ-glutamate substituted CyBam was applied to imaging γ-glutamyl transpeptidase (GGT) activity in a metastatic model of ovarian cancer. Overall, CyBams have significant potential to extend the reach of fluorogenic strategies to intact tissue and live animal imaging applications.

Selectively activating a fluorescent signal is a powerful approach to interrogate biological processes. A common tactic uses multichromophore systems that rely on quenching through fluorescence resonance energy transfer (FRET) and related mechanisms.¹ Another approach uses fluorogenic probes, where the change in signal results from chemical transformations to the chromophore itself.^2–6 The latter often provides improved turn-ON ratios and benefits from requiring only a single chromophore. The most broadly used fluorogenic chemistry is based on derivatization of coumarin, rhodamine, and hybrid cyanine scaffolds, which absorb and emit light in the visible to far-red range.^4,5,7–10 To carry out such experiments in living organisms, it is desirable to have turn-ON probes that absorb and emit long-wavelength near-infrared (NIR) light (>700 nm), which is less attenuated by tissue. However, only systems based on FRET pairs or self-quenching have routinely been applied in this range (Figure 1A).^11–14

Figure 1.

(a) Previous NIR turn-ON approaches and (b) fluorogenic CyBams reported here.

Indocyanine dyes are exceptionally useful fluorescent probes. Heptamethine cyanines such as indocyanine green (ICG) and IR-800CW have been the subject of extensive preclinical and clinical in vivo imaging efforts.^15,16 Fluorogenic probes built on the heptamethine cyanine scaffold would benefit from the extensive infrastructure built for their use, as well as previous efforts to optimize these molecules.^13,14 We hypothesized that such probes could be created by carbamate derivatization of norcyanines; a class of molecules originally reported by Miltsov characterized by secondary, not tertiary, indolenine nitrogen atoms attached to the polymethine chromophore.^17–19 These dyes differ from conventional cyanines in that their NIR fluorescent signal requires nitrogen protonation (pK_a ~ 5). Achilefu and co-workers have demonstrated the exceptional utility of water-soluble variants of these dyes for in vivo tumor imaging.^20,21

Here, we report the synthesis, validation, and application of a series of fluorogenic cyanine carbamates, or CyBams, that can be activated by a range of stimuli (Figure 1B). We demonstrate that sulfonated heptamethine norcyanines can be derivatized with a range of cleavable benzyl carbamates to provide the stable CyBam probes (λ_abs = 430 nm). These molecules then undergo efficient triggered 1,6-elimination to provide the pH-responsive heptamethine norcyanines (protonated form λ_abs = 755 nm). Unlike existing far-red fluorogenic probes, the cationic chromophore is only formed through carbamate cleavage and nitrogen protonation, leading to an exceptional turn-ON ratio (>150×). A glutamate substituted CyBam can be activated by γ-glutamyl transpeptidase (GGT) with high selectivity in vitro and in a metastatic model of ovarian cancer. Overall, these studies provide a new class of versatile NIR fluorogenic probes with significant potential to extend turn-ON strategies to in vivo applications.

To test this approach, we first prepared a disulfonated heptamethine norcyanine, NorCy7, a decarboxylated form of a compound previously reported by Achilefu.²¹ In line with prior reports, the dye exhibits two forms in biologically relevant conditions, an unprotonated quenched form, NorCy7 (λ_abs = 520 nm @ pH 7.4), and a protonated fluorescent form, NorCy7-[H⁺] (λ_abs = 755 nm @ pH 4.5), with a pK_a of 5.2 (Figure S1a,b). After examining several conditions, we were delighted to find that exposure of NorCy7 and 4-nitrophenyl carbonates to NaH or Cs₂CO₃ in DMF afforded the corresponding carbamate products. Using this approach, we prepared CyBam-N₃ from NorCy7 and 1 in reasonable yield following purification by reversed-phase chromatography (Scheme 1).

Scheme 1. Synthesis of CyBam-N₃.

CyBam-N₃ allowed us to examine the turn-ON chemistry using a chemical trigger. Azide reduction, involving aza-ylide formation and hydrolysis, was hypothesized to initiate 1,6-elimination and carbamic acid hydrolysis to result in unmasking of the pH-sensitive norcyanine (Figure 2A).^22–24 CyBam-N₃ exhibits minimal absorbance and emission in the NIR range at either neutral (pH 7.2) or acidic (pH 4.5) conditions (Figure 2B, Figure S2). As anticipated, examining the absorbance profile of CyBam-N₃ and NorCy7 at both neutral and acidic pH revealed three readily distinguishable species (Figure 2B,C). We investigated the stability of CyBam-N₃ at physiological pHs and in serum and observed little degradation (<5%) over 24 h (Figure S4). Next, we examined the fluorogenic response of CyBam-N₃ by incubating with PPh₃ at pH 5.2 (Figure 2D). We observed rapid conversion to NorCy7-[H⁺] with a dramatic 170-fold increase in the fluorescence signal (Figure 2E). This reaction could also be carried out at neutral pH to provide the neutral form NorCy7 (Figure S3). Lastly, we compared the magnitude of the turn-ON response of CyBams to xanthene cyanines, which are far-red probes that have been broadly employed for in vivo fluorogenic imaging.^10,25–28 We prepared and tested a sulfonated, acetylated variant and found turn-ON ratios of 1.5 and 15 (with 640 nm ex.) at pH 4.5 and 7.2, respectively (Figure S5). The critical distinction is the absorption profile in the OFF state. The acetylated xanthene cyanines exhibit a substantial long-wavelength absorption band, which leads to significant emission from the quenched state (Figure S5). By contrast, CyBams exhibit minimal absorption in the NIR region at either neutral or acidic pHs.

Figure 2.

(a) The turn-ON mechanism of CyBam-N₃ is a two-step process requiring (i) cleavage of the carbamate linker and (ii) indolenine protonation. (b) Absorbance spectra and (c) images of 10 μM of solutions CyBam-N₃ (blue) and NorCy7 (green) in PBS, pH 7.4, and NorCy7-[H⁺] (red) in acetate buffer, pH 5.2. (d) CyBam-N₃ (10 μM; 1 equiv) and PPh₃ (100 μM; 10 equiv) in PBS:MeOH (1:1), pH 5.2, were monitored at 5 min intervals (n = 3). Complete conversion to NorCy7-[H⁺] occurred within 60 min of PPh₃ addition. (e) Conversion of CyBam-N₃ (10 μM) to NorCy7-[H⁺] (100 μM PPh₃, pH 4.5) resulted in a 170× enhancement of the NIR fluorescent signal (ex. 710 nm).

To investigate the utility of CyBams for cellular and in vivo imaging, we used a validated fluorogenic trigger with significant translational potential. GGT is a cell-surface-bound enzyme involved in maintaining cellular glutathione (GSH) and cysteine homeostasis.^29–32 Additionally, it has been used as a biomarker of several malignant tumors (including liver, cervical, and ovarian), and overexpression of GGT has been correlated with metastases.^33–36 The key cleavable glutamate was installed on CyBam-γ-Glu (Figure 3A) using a similar procedure to that described above (see the Supporting Information). After confirming the stability of CyBam-γ-Glu (Figure S6), we examined the probe in enzymatic assays. We observed a GGT-dependent increase in fluorescent signal (Figure 3B) with a Michaelis constant (K_M =16 μM) similar to those obtained with other GGT probes (Figure S7).³⁷ We also established that CyBam-γ-Glu is specifically activated by GGT and did not show any significant signal when incubated with representative proteases and esterases (Figure 3C). Lastly, we determined the selectivity of CyBam-γ-Glu using established GGT inhibitors.^38,39 We observed a 60% and 80% decrease in fluorescent signal in the presence of DON and GGsTop, respectively, confirming the selectivity of the probe (Figure S8).

Figure 3.

(a) Structures of CyBam-γ-Glu and CyBam-N.C. (b) Increase in fluorescent intensity of CyBam-γ-Glu (20 μM) after incubation with increasing concentration of GGT (0–400 U/L). (c) Activation of CyBam-γ-Glu after incubation with GGT (100 U/L), leucine aminopeptidase (LAP; 800 U/L), and pig-liver esterase (PLE; 800 U/L) at 37 °C for 30 min in PBS pH 7.4 followed by pH adjustment to pH 5.2. Cathepsin B (CatB; 2.5 μg) was used in acetate buffer pH 5.2 for 30 min at 37 °C. (d) Confocal images of activation of CyBam-γ-Glu (20 μM) and CyBam-N.C. (20 μM) in SHIN-3 cells. The fluorescent signal from the probe and nucleus (Hoechst) is shown in red and blue, respectively. Images were taken using a 63× oil immersed lens (numerical aperture, N.A. 1.4). (e) Quantification of fluorescent signal after incubation of CyBam-γ-Glu (20 μM) in the presence of GGT inhibitors (DON, GGsTop) and CyBam-N.C. in SHIN-3 cells using flow cytometry. The geometric mean fluorescent intensity (±SD) of the fluorescent signal in the cells is shown (n = 6 independent experiments). (f) In vivo imaging of the SHIN-3-ZsGreen metastatic tumor model at 3 h. Green and red pseudocolors are used to represent the signal from the GFP and Cy7 channels, respectively. Fluorescent line graph showing a correlation between the fluorescent signal from GFP and Cy7 channel across the metastatic tumor in two different regions (A and B). Data points are displayed as mean ± SD, and the p-values were evaluated by the Student’s t-test (*** p-value ≤0.001).

We then examined CyBam-γ-Glu, and the corresponding noncleavable variant CyBam-N.C., in cellular assays and in vivo imaging experiments (Figure 3A). We used a SHIN-3 ovarian cancer cell line that has been previously shown to overexpress GGT.⁴⁰ We first established that CyBam-γ-Glu exhibited minimal toxicity (Figure S9). We also established that NorCy7 exhibited significant cellular uptake in SHIN-3 cells. The majority of the fluorescent signal was observed in lysosomes, where the acidic microenvironment is likely responsible for formation of NorCy7-[H⁺] (Figure S10).²¹ Next, we evaluated the cellular activation and selectivity of CyBam-γ-Glu with and without incubation with GGT inhibitors. Using confocal microscopy and flow cytometry, we observed a strong fluorescent signal in cells treated with CyBam-γ-Glu. By contrast, minimal fluorescence signal was observed in cells that were either treated with CyBam-N.C. or preincubated with GGT inhibitors (Figure 3D,E and Figures S11 and S12). Encouraged by these results, we tested CyBam-γ-Glu in a metastatic tumor model of ovarian cancer. This model entails intraperitoneal injection of SHIN-3-ZsGreen cells, resulting in formation of a significant primary tumor in the greater omentum and locally disseminated metastases.⁴⁰ CyBam-γ-Glu (30 nmol) was injected intraperitoneally in mice, which were euthanized after 1, 3, and 6 h, and both the primary tumor and local metastases were imaged. Excellent colocalization between CyBam-γ-Glu and the ZsGreen signal suggests that the probe was activated and taken up selectively by tumor cells, with a significant signal at all three time points (Figure 3F and Figures S13 and S14). We also confirmed that these probes can be used in live mice using a conventional in vivo imaging system (IVIS). To do this, we compared CyBam-γ-Glu and CyBam-N.C. in an MDA-MB-468 xenograft, which is a triple-negative breast cancer cell line with modest GGT expression.^41,42 The CyBam-γ-Glu can be readily visualized with significant differences between the two agents in both tumor and liver signals, as well as tumor-to-background ratios (Figure S15).

Finally, to examine the versatility of this approach, we prepared and initially characterized the utility of CyBam probes to visualize common reactive oxygen species (ROS) with well-validated ROS-responsive triggers: boronic acid (H₂O₂) and phosphine oxide (O₂⁻).^7,43,44 We tested the induction of ROS in PC-3 cells, a prostate cancer cell line, by doxorubicin. Notably, doxorubicin absorbs in the visible region (λ_max = 480 nm) which can hamper the utility of conventional visible light-absorbing fluorogenic probes.^45,46 As expected, we observed the generation of both superoxide and hydrogen peroxide, with no interference from the addition of doxorubicin (Figures S16 and S17). These results suggest that these probes may have significant utility in the exploration of ROS biology.

Fluorogenic probes are powerful tools with the potential to noninvasively monitor enzymatic processes and other stimuli in real-time in living organisms. Here, we report CyBams, the first enzyme- or analyte-responsive fluorogenic probes based on the heptamethine cyanine scaffold. These readily water-soluble probes result from modification of the norcyanine scaffold with a cleavable carbamate linker that is activated through 1,6-elimination and chromophore protonation. This combination results in turn-ON ratios that dramatically exceed those found with existing far-red fluorogenic probes, particularly in acidic conditions. The results presented above suggest that CyBams have significant potential for use as activatable probes for in vivo imaging. We hypothesize their application may include optically guided surgical procedures and note that the extensive optical instrumentation in place for heptamethine cyanines makes this prospect more enticing.⁴⁷ Going forward, as CyBams are more emissive upon protonation in the lysosome, it is possible that efforts to improve their lysosomal targeting may serve to increase their signal intensity. Additionally, as cyanine fluorophores have historically been most useful as bioconjugatable probes, we anticipate that the utility of CyBams may be enhanced when combined with active targeting. In this scenario, targetable CyBams create the possibility to report on enzymatic activity at only a specific cell type or location of interest. Efforts toward these goals are ongoing and will be reported in due course.

ABBREVIATIONS

NIR

near-infrared

FRET

fluorescence resonance energy transfer

CyBam

cyanine carbamate

ICG

indocyanine green

LAP

leucine aminopeptidase

PLE

pig-liver esterase

CatB

cathepsin B

PBS

phosphate buffer saline

GGT

γ-glutamyl transpeptidase

γ-Glu

γ-glutamate

DON

6-diazo-5-oxo-l-norleucine

GGsTop

2-amino-4[3-(carboxymethyl)-phenyl](methyl)phosphono-butanoic acid

GSH

glutathione

GFP

green fluorescent protein

Cy7

heptamethine

DMF

dimethylformamide

NaH

sodium hydride

Cs₂CO₃

cesium carbonate

abs

absorbance

ex

excitation

em

emission

N.A.

numerical aperture