A Phenotypic Compound Screening Assay for Lysosomal Storage Diseases

Materials

Lysotracker-blue DND-22 (#L7525), Lysotracker-green DND-26 (#L7526), Lysotracker-yellow HCK-123 (#LL12491), Lysotracker-Red DND-99 dye (#L7528), Hoescht 33342 nuclear dye (# H3570), and CellMask Red (#H32712) were purchased from Invitrogen (Carlsbad, CA). DRAQ5 nuclear dye was obtained from Cell Signaling. Filipin dye (#F9765) and MβCD (#M7439) were obtained from Sigma-Aldrich. The 96-well black clear-bottom plates (#655090) were purchased from Greiner Bio-One (Monroe, NC).

Cells and Cell Culture

Patient-derived skin fibroblast cell lines and control cell line (wild type) were purchased from the Coriell Cell Repository (Camden, NJ). Cells were cultured in DMEM medium (Invitrogen, cat No. 11995-040) supplemented with 10% fetal bovine serum, 100 unit/mL penicillin, and 100 µg/mL streptomycin in a humidified incubator with 5% CO₂ at 37 °C. Cells were seeded at 1500 cells/well and 3000 cells/well in 100 µL medium in 96-well plates for the Lysotracker experiments using imaging and fluorescence intensity readout formats, respectively.

Lysotracker Staining Measured by Imaging Analysis

The assay was optimized to visualize the enlarged lysosomes by staining cells with appropriate concentration of the Lysotracker dye after overnight culture in 96-well plates. Briefly, cells were live stained with 100 µL/well 50 nM Lysotracker-red DND-99 dye (Invitrogen, #L-7528) in medium at 37 °C for 60 min. After the plate was washed twice with phosphate-buffered saline (PBS), 100 µL/well of 1 µg/mL Hoechst 33342 in 3.2% formaldehyde solution was added to fix the cells and stain the nuclei. After 30 min incubation at room temperature (RT), the plate was washed twice with PBS and stored at 4 °C until further imaging measurement. The image acquisition was carried out in an IN Cell Analyzer 2000 (GE Healthcare, Waukesha, WI) that uses a halide arc lamp as the light source. The DAPI (Ex = 350 ± 50, Em = 455 ± 50 nm) and TRITC (Em = 545 ± 20, Ex = 593 ± 20 nm) filter sets were used to visualize Hoechst nuclear staining and Lysotracker-red staining, respectively. Nine fields of images per well were usually recorded with a 20× objective and numerical aperture of 0.45. The exposure time was 0.05 s for Hoechst nuclear staining and 0.9 s for Lysotracker-red staining. Similarly, the images of Lysotracker-blue staining was taken by the DAPI filter set and Lysotracker-green and Lysotracker-yellow staining were taken using the FITC filter set (Ex = 490 ± 20, Em = 525 ± 36 nm). The Draq5 nuclear staining was co-stained with these three Lysotracker dyes and measured by a different filter set (Em = 645 ± 30, Ex = 705 ± 72 nm).

Lysotracker Fluorescence Intensity Assay Using Fluorescence Plate Reader

After an overnight culture, the medium was aspirated and discarded from 96-well assay plates followed by addition of 100 µL/well of 156 nM Lysotracker-red solution in medium (Supplementary Table S2). The dye incubation time was typically 1 h at 37 °C except where specifically defined. After washing twice with PBS, the assay plate was measured in a fluorescence intensity mode (Ex = 570 ± 10, Em = 600 ± 10 nm) on a fluorescence plate reader (Tecan, Männedorf, Switzerland). For the 384-well plate assay, cells were seeded at a density of 375 cells/well in 25 µL of medium and incubated with compounds or DMSO (final concentration of 0.1% as a solvent control) for 72 h at 37 °C. After aspirating the medium, the cells were co-stained with 625 nM Lysotracker-red and 1 µg/mL Hoechst for 1 h at 37 °C. Higher concentration of Lysotracker-red dye was used in a 384-well plate to obtain a better signal. Cells were washed twice with PBS, and the fluorescence intensity was measured in a fluorescence plate reader (Ex = 570 ± 10, Em = 600 ± 10 nm). The percentage increase of fluorescence intensity was calculated by using the fluorescence intensity in control cells as a basal response.

Filipin Staining

Filipin dye stains the unesterified cholesterol in cells at an appropriate concentration. Briefly, the cells after overnight incubation were washed twice with PBS and fixed with 100 µL/well of a 3.2% formaldehyde solution at RT for 30 min. After washing twice with PBS, the cells were stained with 100 µL/well of 50 µg/mL filipin solution (freshly dissolved in DMSO at 10 mg/mL and then diluted in PBS) at RT for 1 h followed by cell wash with PBS. The resulting assay plates were stored at 4 °C for further imaging analysis. On the day of imaging, cells were stained with 100 µL/well of Draq5 diluted 1:1000 in PBS at RT for 30 min followed by cell wash. The plates were imaged using an IN Cell Analyzer 2000. The DAPI and Cy5 filter sets were used to detect filipin and Draq5 staining, respectively. Nine fields of images per well were usually recorded with a 20× objective and numerical aperture of 0.45. The exposure time was 0.05 s dye and 0.025 s for filipin staining.

Data Analysis

Image analysis was conducted using IN Cell Analyzer 2000 software (GE Healthcare, version 3.7). The Multi-Target Analysis protocol was used for quantitation of Hoechst stained nuclei and Lysotracker-red stained lysosomes. Nuclei were segmented using the Top Hat segmentation method with a minimum area set at 150 µm and a sensitivity set at 50. Lysosomes were identified as “Organelles” within the analysis software and were segmented using the Multiscale top-hat algorithm. Settings for lysosome detection were identify granules ranging in size of 2 to 10 µm (3 to 13 pixels) and a sensitivity setting of 60. Total organelle intensity was calculated by multiplying the mean intensity per granule by the total area of the organelles. A single contrast setting for image analysis was applied using the imaging analysis software included in the IN Cell Analyzer 2000. Concentration-response curves were analyzed and EC₅₀ values (mean ± SD) calculated using Prism software (GraphPad, San Diego, CA). Results in figures are expressed as mean of triplicates ± SD unless they are specified.

Dye Incubation Time for Lysotracker Staining

The time course of staining with Lysotracker-red dye was carried out in NPC patient-derived fibroblast cells in comparison with healthy control cells. The fluorescence staining of Lysotracker-red dye reached a plateau after 1 h of incubation at 37 °C (Supplementary Fig. S1). Longer times for Lysotracker dye incubation did not further increase fluorescence staining. Thus, 1 h incubation of Lysotracker dye with NPC cells was selected as an optimal condition for subsequent experiments.

Optimal Lysotracker Probes and Concentration

There are several additional Lysotracker dyes with different fluorescence spectrum. We compared four Lysotracker dyes with emissions in the blue, green, yellow, and red fluorescence spectrum. Results indicated that these dyes have different sensitivities for staining enlarged lysosomes in NPC fibroblasts. The fluorescence signals from the Lysotracker-blue and Lysotracker-yellow dyes were relatively weak, where a 3.2 µM dye concentration was needed for the measurement ( Fig. 1A , B ). Both the Lysotracker-green and Lysotracker-red dyes stained enlarged lysosomes robustly at a substantially lower concentration ( Fig. 1C , D ). The optimal concentrations for these two dyes were 100 and 50 nM, respectively. Together, the data indicate that the Lysotracker-green and Lysotracker-red dyes are more sensitive for staining enlarged lysosomes in NPC fibroblasts. Because the red fluorescence readout is less prone to compound fluorescence interference in compound screening, the Lysotracker-red dye was selected for the further evaluation.

OPEN IN VIEWER

Figure 1.

Fluorescence staining of four types of Lysotracker dyes in skin fibroblasts derived from a Niemann Pick type C (NPC) patient in comparison with control cells. Lysotracker dye staining increased in NPC cells, indicating the enlarged lysosomes in these patient-derived cells. (A) A total of 3.2 µM of Lysotracker-blue dye was needed to visualize the difference between NPC and control cells. (B) A clear difference in Lysotracker-yellow staining between NPC and control cells was observed at 1.6 µM dye concentration. (C) Images of Lysotracker-green staining showed that 50 or 100 nM concentration of this dye was enough to distinguish the difference between NPC and control cells. (D) Lysotracker-red exhibited the strongest fluorescence staining in the NPC cells in comparison with the control cells. The concentration of 50 nM of this dye was optimal for the staining in these fibroblasts.

Lysotracker Assay Using Fluorescence Plate Reader

Although the imaging assay for Lysotracker staining visualizes the enlarged lysosomes in patient-derived cells, it requires a special imaging plate reader and has lower screening throughput compared with a regular fluorescence plate reader assay. Therefore, we used a fluorescence plate reader for measurement of fluorescence signals in the Lysotracker assay. We found that the optimal cell density and dye concentration detected by a fluorescence plate reader were similar to those used in the imaging assay. The fluorescence intensity reached a plateau between 156 and 625 nM Lysotracker-red dye (Supplementary Fig. S2). We found that the higher concentration of LysoTracker-red dye was required in the plate reader assay compared with that in the imaging assay. The signal-to-basal ratio was approximately 2- to 3-fold for this assay with small well-to-well variation. The percentage increase in fluorescence intensity detected from the patient-derived cells (relative to the control cells) was used for the data calculation. These results demonstrate that the Lysotracker assay in a fluorescence plate readout format is useful for compound screening for identification of active compounds that reduce the enlarged lysosomes in NPC cells.

Increased Lysotracker Staining in Fibroblasts from Other LSDs

Because lysosome enlargement is a common feature in many LSD patient-derived cells, we determined whether the Lysotracker staining assay developed in NPC cells was applicable to other LSD patient-derived cells. Sixteen additional LSD fibroblast cell lines (Supplementary Table S1) were tested in the Lysotracker assay together with NPC and control cells to compare the results from the imaging readout with these from fluorescence plate reader measurement. The fluorescence intensities of Lysotracker staining increased in 9 patient-derived cell lines (NPC, ML-III, ML-IV_8 months old, MPS-I, MPS-VI, MPS-VII, Wolman, Farber, and NPA) in comparison with the control cells, whereas eight other cell lines (α-MAN, MLD, ML-IV_2 years old, MPS-II, MPS-IIIB, Batten, Fabry, and Tay-Sachs) did not show any increase in Lysotracker staining ( Fig. 2A ). Differences in Lysotracker staining were observed in cells derived from two different ML-IV patients who displayed a significant increase in Lysotracker staining in cells derived from an eight-month-old patient and no increase in cells derived from a two-year-old patient. These results indicate that not all the LSD patient cells exhibit an increase in Lysotracker staining. Patient-to-patient variation was observed in the level of Lysotracker staining within the same disease. Thus, the variability in Lysotracker staining could stem from the differences in disease type, stage, and clinical severity.

OPEN IN VIEWER

Figure 2.

Comparison of results of Lysotracker staining in 17 patient-derived fibroblast lines measured by an IN Cell Analyzer 2000 with these determined by a fluorescence plate reader. (A) Increase in Lysotracker staining was observed in patient cells with NPC, ML-III, ML-IV (8 months old), MPS-I, MPS-VI, MPS-VII, Wolman, Farber, and NPA disease. But no significant increase in Lysotracker staining was found in patient cells with α-MAN, MLD, ML-IV (2 years old), MPS-II, MPS-IIIB, Batten, Fabry, and Tay-Sachs diseases. (B) Increase in fluorescence intensity of Lysotracker staining determined by a fluorescence plate reader was observed in all patient cells except those with MLD, ML-IV (2 years old), MPS-II, and Fabry disease. (C) Increase in fluorescence intensity of Lysotracker staining calculated from the images in (A).

The fluorescence intensity of Lysotracker staining in these 16 cells was also measured using a fluorescence plate reader. An increase in Lysotracker staining was observed in 13 patient cell lines, including 4 patient-derived cell lines ( Fig. 2B ; α-MAN, MPS-IIIB, Batten, and Tay-Sachs) that were not identified in the imaging assay ( Fig. 2C ). The discrepancy in the number of positive patient cell lines found in two detection methods may be associated with the different detection sensitivity in two detection methods, limitation of imaging software for data calculation, and day-to-day experimental variation. The irregularity in individual lysosome size and variable dye staining pattern in patient-derived cells could also contribute to the variable results from the imaging data calculation. In addition, the number of cells analyzed in image experiments could be varied and may affect the calculated results. Taken together, although the visual examination of Lysotracker staining is a useful application to confirm enlarged lysosomes in patient-derived cells, the plate reader–based measurement is more robust for the primary screen of compound collections.

A DMSO plate test for the Lysotracker assay was carried out using a fluorescence plate reader for detection. The signal-to-basal ratio of 3.8-fold, coefficient of variation (CV) of 12.0%, and Z factor of 0.41 were obtained using the NPC patient-derived fibroblast cells (Supplementary Fig. S3). The signal-to-basal ratio and CV data indicate a relatively good screening assay. These scores are indicative of an acceptable screening assay, and thus the Lysotracker assay in a fluorescence plate reader assay format can be used in the screening of large compound libraries.

Cyclodextrin Reduced Lysotracker Staining

We determined the comparability of two detection methods (e.g., plate reader and imaging) of this Lysotracker assay for the measurement of compound activity in NPC patient-derived cells. MβCD was chosen as a positive compound as it has been shown to reduce the enlarged lysosomes by decreasing lysosomal cholesterol accumulation in NPC fibroblasts.^10,15 The effect of MβCD measured by the Lysotracker assay was compared with a filipin staining assay and a biochemical Amplex-red cholesterol assay, whereby both assays measure cellular cholesterol levels. ¹¹ Filipin dye binds to unesterified cholesterol, which becomes fluorescent and can be detected by fluorescence imaging. We observed that the Lysotracker staining was significantly reduced in the NPC cells after treatment with 300 µM MβCD ( Fig. 3A ). A similar effect of MβCD on the NPC cells was observed in the filipin staining assay ( Fig. 3B ). In the Lysotracker assay using a fluorescence plate reader, MβCD treatment reduced the fluorescence intensity with an IC₅₀ value of 133 µM, indicating the reduction of enlarged lysosomes in NPC cells. Similarly, an IC₅₀ of 78.7 µM of MβCD on the reduction of cellular cholesterol was determined in the Amplex-red cholesterol assay ( Fig. 3C ). Together, the results demonstrate that the compound activity measured by the Lysotracker assay correlates with that determined in the Amplex-red cholesterol assay. Therefore, the Lysotracker assay is effective for measurement of compound activity against enlarged lysosomes in patient-derived cells.

OPEN IN VIEWER

Figure 3.

Effect of methyl-β-cyclodextrin (MβCD) on reduction of cholesterol accumulation and lysosome size in Niemann Pick type C (NPC) fibroblasts determined by Lysotracker-red staining, filipin cholesterol staining, and Amplex-red cholesterol assays. (A) Images of filipin cholesterol staining. NPC cells exhibit an increase in filipin staining compared with the control (wild-type) cells, indicating the accumulation of unesterified cholesterol in lysosomes. The filipin staining in NPC cells was reduced in a concentration-dependent manner after treatment with MβCD. (B) Images of Lysotracker-red dye staining. The fluorescence signals of Lysotracker dye staining dramatically increased in NPC cells compared with the control cells, indicative of enlarged lysosomes in the patient-derived cells. The treatment with MβCD concentration-dependently reduced Lysotracker dye staining. (C) Results of the fluorescence plate readout format of Lysotracker assay and Amplex-red cholesterol assay. MβCD reduced both lysosomal cholesterol accumulation (measured by the Amplex-red cholesterol assay) and enlarged lysosomes (determined by the Lysotracker-red staining assay using a fluorescence plate reader) in a concentration-dependent manner.

In contrast to target-based drug discovery, phenotypic screening is based on the disease phenotypes, which may not link to a specific target protein. A cell-based disease model using patient-derived cells is usually disease relevant. Unlike the NPC disease, specific dyes capable of detecting various lipids accumulated in lysosomes of other LSDs (Supplementary Table S1) are often unavailable for high-throughput screening assays. Although immunofluorescence assays using the lipid-specific antibodies could potentially be developed, it is usually not applicable for compound screening because of the nonspecific interaction of these antibodies and complexity of the immunofluorescence assay procedure. Our results demonstrate that the Lysotracker staining assay can be applied to patient cells derived from many LSDs for compound screening because it measures a common phenotypic feature of enlarged lysosomes. However, the application of this assay is limited only to those LSD cells in which strong Lysotracker staining is observed. A cell viability assay should be used for hit confirmation to eliminate cytotoxic compounds that could be the false-positives. The disease-relevant cell types such as neurons and hepatocytes should also be applied for the further confirmation of the activities of lead compounds, whereas the skin fibroblast cells are used in the primary screens.

In conclusion, we have developed and optimized a phenotypic Lysotracker staining assay that measures enlarged lysosomes in patient-derived cells. The Lysotracker assay can be detected by both imaging readout and fluorescence plate reader formats in compound screening to identify lead compounds for drug development. Although the plate reader format of the Lysotracker assay can be used for high-throughput screening, the imaging format of the Lysotracker assay is useful for the compound confirmation and follow-up studies. In addition, Lysotracker staining could potentially be developed into a flow cytometry assay for clinical disease diagnosis and for biomarker development using patient-derived lymphocytes.