Supporting Information Selective Fluorogenic Imaging of Hepatocellular H 2 S by a Galactosyl Azidonaphthalimide Probe

We have developed a galactosyl azidonaphthalimide probe for the selective fluorogenic imaging of hepatocellular H2S, an important gaseous transmitter produced in the liver.

Selective fluorogenic imaging of hepatocellular H 2 S by a galactosyl azidonaphthalimide probe † De-Tai Shi, a Dan Zhou, ab Yi Zang, b Jia Li,* b Guo-Rong Chen, a Tony D. James, c Xiao-Peng He* a and He Tian a We have developed a galactosyl azidonaphthalimide probe for the selective fluorogenic imaging of hepatocellular H 2 S, an important gaseous transmitter produced in the liver.
Despite its toxicity, hydrogen sulfide (H 2 S) has been identified as an important gaseous transmitter in many cellular signalling pathways. 1Cystathionine b-synthase (CBS) and cystathionine g-lyase (CSE) are the two major enzymes responsible for the production of H 2 S with L-(homo)cystein as the substrate.Because of the tissue-specific distribution of these enzymes, H 2 S is predominantly generated in the liver. 2 However, this metabolic organ is vulnerable to damage from the aberrant expression of redox-active species including H 2 S, leading to hepatic diseases such as liver cirrhosis. 3As a result, selective detection of hepatocellular H 2 S represents an important goal to aid not only cell biology research but also disease diagnosis.
Conventional techniques for H 2 S detection depend on chromatographic and electrochemical sensors, which are not applicable for live cell imaging. 4To address these issues, numerous fluorescence small-molecule probes for H 2 S have been developed over recent years.3][14] Nevertheless, although elegant H 2 S probes with cellular trapping, 15 lysosome 16 and mitochondria targeting 17 abilities have been developed, fluorogenic probes that address hepatocyte-selective imaging of the species are rare.
Recently, we [18][19][20][21] and others [22][23][24][25][26] determined that the introduction of a glycosyl moiety (as a targeting agent) to a fluorescence probe can greatly enhance its selective internalisation by cells derived from a certain tissue.This could be ascribed to the selective, sugar-receptor-mediated endocytosis of the glycoprobes by the cells.Here we show that a galatosyl azidonaphthalimide probe has the ability to image H 2 S selectively in hepatocytes among other cells, due to galactoside receptorpromoted endocytosis.
The desired DT-Gal was synthesised by a click-coupling of alkynyl naphthalimide 1 with azido galactoside 2, followed by a sequential azidation and deacylation (Fig. 1a).The presence of the azide weakens the fluorescence of naphthalimide by an ICT process, and a subsequent reduction by H 2 S produces aminonaphthalimide, enhancing the fluorescence. 14DT-OH which is a known H 2 S probe that lacks the galactosyl targeting agent was used as a control. 27ith these probes in hand, we tested their fluorescence response to H 2 S in an aqueous solution (PBS, pH 7.4).The presence of increasing H 2 S caused a gradual fluorescence increase of both DT-OH (Fig. 1b) and DT-Gal (Fig. 1c) with the latter being more sensitive.This might be a result of the better water solubility of the latter glyco-probe.Among a range of anions (Fig. 1d) and amino acids (Fig. S1, ESI †), both probes showed good selectivity.Mass spectroscopic analysis indicated that the aminonaphthalimide derivative of DT-Gal was produced upon reaction with H 2 S (Fig. S2, ESI †).
A kinetic investigation determined that DT-Gal showed rapid fluorogenic response to H 2 S (Fig. 1e), and good linearity was observed by plotting the fluorescence intensity of the probe as a function of increasing H 2 S (Fig. 1f).The limit of detection of DT-Gal for H 2 S was determined to be 0.78 mM (3s b /k).Additionally, we observed that the probe functioned well over a wide pH range (4-12, Fig. 1g), suggesting its applicability for cellular analysis.
Subsequently, we tested the ability of the probes to image H 2 S in live cells.We used the human hepatoma cell line (Hep-G2) with over-expressed asialoglycoprotein receptor (ASGPr) that recognises galactosides [18][19][20][21] as well as cancer cells without the receptor expression such as human colon cancer HCT116, cervix cancer HeLa and lung cancer A549. 18We observed that, the addition of DT-Gal to the cell lines pre-treated with H 2 S led to fluorescence enhancement for Hep-G2, but only a slight fluorescence increase for other cells (Fig. 2a and c).In sharp contrast, co-incubation of DT-OH with the cells led to slight, unselective fluorescence enhancement (Fig. 2b and d).9][20][21] Notably, the assumed endocytosis led to much stronger fluorescence generation of DT-Gal than the DT-OH without a targeting group.
To corroborate that the imaging was a result of selective ASGPrgalactose interaction, we used a Hep-G2 cell line with a reduced ASGPr expression level (sh-ASGPr) 21 to incubate with DT-Gal.We observed that addition of the galactosyl probe to sh-ASGPr resulted in a much weaker fluorescence compared to Hep-G2 (Fig. 3a and d).This correlates with the distinct ASGPr expression level of the two cell lines (Fig. 3c).Moreover, pre-incubation of increasing free galactose with Hep-G2 inhibited the fluorescence enhancement of DT-Gal in a concentration-dependent manner (Fig. 3b and e).These data support the assumption that the strong and selective fluorogenic signal produced by DT-Gal in raw Hep-G2 cells is facilitated by ASGPr-mediated endocytosis.We also tested the kinetics of DT-Gal for H 2 S imaging in the hepatocellular cell line.We observed a time-dependent fluorescence enhancement of the imaging experiments and equilibrium was achieved in about 30 min (Fig. 4a and b).Note that this cellular kinetics is slower than that in solution probably because of the complexity of the cellular environment.Interestingly, DT-Gal (Fig. 4c) was much less toxic towards Hep-G2 as well as a human kidney cell line (HEK293) than DT-OH (Fig. 4d).This suggests that, in addition to its targeting ability, the presence of the galactosyl moiety might also mitigate the intrinsic cytotoxicity of naphthalimide-based compounds. 28e note that a similar azidonaphthalimide probe for H 2 S has been previously reported. 29However, unlike our DT-Gal probe, that probe does not contain a cellular targeting unit, which is vital for target-specific imaging in vivo.Since azidonaphthalimide can be reduced to the fluorescent amino derivative within cells, 25 we carried out a time-dependent imaging assay with Hep-G2 and just DT-Gal (Fig. S3, ESI †).While the fluorescence increased slightly, the co-existence of H 2 S results in significantly enhanced fluorescence.This clearly demonstrates that the fluorescence enhancements observed here are due to H 2 S recognition within the cells.
In conclusion, a galactosyl azidonaphthalimide based fluorogenic probe for hepatocellular-selective imaging of H 2 S was developed, which sets a basis for the target-specific imaging of H 2 S in the liver, the main organ that produces this gaseous transmitter.This research is supported by the 973 project (2013CB733700), the National Science Fund for Distinguished Young Scholars (81125023), the National Natural Science Foundation of China (21176076, 21202045 and 91213303), the Program of Shanghai Subject Chief Scientist (13XD1404300), the Key Project of Shanghai Science and Technology Commission (13NM1400900) and the Fundamental Research Funds for Central Universities.The Catalysis And Sensing for our Environment (CASE) network is thanked for research exchange opportunities.

Fig. 1 2À 4 À,
Fig. 1 (a) Synthesis of DT-Gal and the structure of DT-OH.Reagents and conditions: (I) CuSO 4 ÁH 2 O, Na ascorbate in H 2 O-CH 2 Cl 2 .(II) NaN 3 in DMF, and then Et 3 N. Fluorescence titration of (b) DT-OH (10 mM) and (c) DT-Gal (10 mM) in the presence of increasing H 2 S (0-200 mM).(d) Fluorescence change of DT-Gal (10 mM) in the absence (blank) and presence of, from left to right, F À , Cl À , Br À , I À , CO 3 2À , HCO 3 À , SO 4 2À , HSO 4 À , S 2 O 3 À , PO 4 3À , HPO 4 2À , H 2 PO 4 À , Phe, Ser, Thr, Pro, Val, His, Gly, Cys, GSH and H 2 S (100 mM), where I f and I fo are the fluorescence intensity of the probe in the presence and absence of analyte, respectively (inset: photographed fluorescence change of DT-Gal in the presence of H 2 S).(e) Plotting the fluorescence change of DT-Gal (10 mM) with H 2 S (200 mM) as a function of time, where I fx and I f0 are the fluorescence intensity of the probe with H 2 S at different time points and that of the probe alone, respectively.(f) Plotting the I f of DT-Gal (10 mM) in the presence of increasing H 2 S. (g) Plotting the I f of DT-Gal (10 mM) without H 2 S (black) and with H 2 S (200 mM) (blue) at different pH.All fluorescence spectra were recorded in PBS (0.05 M, pH 7.4 or varied pH as indicated) with excitation at 426 nm.

Fig. 2
Fig. 2 Cell imaging (a for DT-Gal and b for DT-OH) and quantification of the fluorescence intensity (I f ) (c for DT-Gal and d for DT-OH) in the absence (À) and presence (+) of H 2 S (Hep-G2: human liver cancer; HCT-116: human colon cancer; HeLa: human cervix cancer; A549: human lung cancer).

Fig. 3
Fig. 3 Cell imaging (a) and fluorescence quantification (d) of DT-Gal for Hep-G2 and sh-ASGPr with reduced ASGPr expression.Cell imaging (b) and fluorescence quantification (e) of DT-Gal with Hep-G2 in the absence and presence of increasing competing free galactose.(c) ASGPr mRNA level of Hep-G2 and sh-ASGPr measured by PCR (***P o 0.001).