Senescence Detection Using Reflected Light in Adipose Stromal Vascular Fraction

Summary

This protocol presents reflected light confocal microscopy for objective, single-cell detection of senescence-associated beta-galactosidase (SA-β-gal) activity in the stromal vascular fraction isolated from human adipose tissue. The method is compatible with immunocytochemistry, incorporates pH controls for quantitative assessment, and additionally provides a more sensitive and less subjective alternative to conventional SA-β-gal staining.

Abstract

Cellular senescence is a stress-induced state characterized by permanent cell-cycle arrest and the development of a distinctive secretory profile that impacts tissue function and contributes to aging and metabolic disease. Senescence-associated β-galactosidase (SA-β-gal) activity is widely used as a marker of senescent cells; however, conventional SA-β-gal assays often rely on subjective visual assessment and provide limited quantitative information. These limitations are particularly evident in primary human cell populations such as the stromal vascular fraction (SVF) derived from adipose tissue, which contains a heterogeneous mixture of preadipocytes, immune cells, and endothelial cells. Here, we present an optimized reflected light confocal microscopy approach for high-resolution, quantitative detection of SA-β-gal activity in human SVF cells. This protocol enables objective single-cell analysis of SA-β-gal activity, allows simultaneous immunocytochemistry for multiplexed detection of additional senescence or lineage markers, and incorporates pH-matched controls. By combining these features, this method provides a sensitive, reproducible, and quantitative approach to studying cellular senescence in heterogeneous primary human cell populations. It offers an improved alternative to conventional SA-β-gal staining.

Introduction

Cellular senescence is a multifaceted response that can arise from many types of stress, including DNA damage, telomere attrition, and oncogene activation. It leads to a stable arrest of the cell cycle together with increased resistance to apoptosis, driven in part by the induction of anti-apoptotic proteins^1,2. Senescence has been linked to improved cell survival in contexts where preservation of tissue integrity is essential, such as repair and maintenance^3,4. Beyond its role in tissue homeostasis, senescence also contributes to normal development by shaping embryonic patterning^5,6and as a barrier to tumor formation by promoting the clearance of damaged or potentially malignant cells^7,8,9. Although senescence can serve beneficial functions, multiple studies have highlighted a contrasting aspect in which senescent cells acquire a pro-inflammatory secretory profile that can disturb local metabolism, damage surrounding tissue, and even support tumor progression^10,11,12,13.

Senescence has been detected in diverse cell types across multiple tissues and is often associated with aging, disease, tissue dysfunction, or cancer. In the liver, senescence of hepatic stellate cells has been shown to promote tumor progression by altering metabolic homeostasis through secreted factors^14,15. Studies examining hepatocytes¹⁶ or whole liver biopsies¹⁷ indicate that senescence in these cells can contribute to metabolic comorbidities such as metabolic dysfunction-associated steatotic liver disease and dysfunction-associated steatohepatitis. Furthermore, the relationship between type 2 diabetes and senescence has gained significant attention over the past decade, with evidence linking senescence to diabetes-related complications in pancreatic β-cells, hepatocytes¹⁶, and adipocytes^18,19,20,21. These findings highlight the importance of accurately identifying senescent cells in metabolic tissues, where their accumulation and heterogeneity contribute to disease progression and therapeutic targeting²².

The signaling and protein expression changes associated with senescence are often cell- and tissue-specific; however, several features are broadly characteristic of senescent cells. One of the most widely used markers is the increased enzymatic activity of β-galactosidase at pH 6.0, referred to as senescence-associated β-galactosidase (SA-β-gal)²³. β-galactosidase is a lysosomal hydrolase present in all cells and is normally active at acidic pH (~4.5–5.0), where it contributes to the degradation of glycosylated substrates. The defining feature of SA-β-gal is the retention of β-galactosidase activity under suboptimal pH conditions. In the classical assay, cells are incubated with the synthetic substrate 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) at pH 6.0, where enzymatic activity is reduced in most cells but remains detectable in senescent cells. This leads to the formation of an insoluble blue precipitate that can be visualised microscopically²⁴.

While this assay has been widely adopted, detection of the X-gal precipitate has traditionally relied on brightfield microscopy. Brightfield detection relies on visual identification of blue precipitate and therefore has limited sensitivity for low-level signals and early senescence states. In contrast, reflected light imaging captures the optical density of the X-gal precipitate, enabling quantitative detection across a broader dynamic range. More recently, fluorogenic substrates have been developed to improve sensitivity²⁵. However, these approaches typically measure β-galactosidase activity under conditions that do not preserve the classical pH-dependent definition of SA-β-gal. As a result, they may increase signal intensity but reduce specificity for senescence. In contrast, the inherent opacity of the X-gal precipitate enables its detection using confocal reflected light imaging, providing a more sensitive and quantitative readout while maintaining the original enzymatic definition of SA-β-gal²⁶.

Over the years, additional hallmarks of senescence have been identified, including increased expression of the cyclin-dependent kinase inhibitors P16 and P21, as well as the transcription factor P53, which regulates cell-cycle arrest and other senescence programs^{27,28,29,30,31,32,33}. These advances have led to a more robust definition of senescence, in which cells are typically classified based on the combined presence of multiple markers. In practice, however, these markers are often measured using different modalities: nuclear markers, such as P16 and P21, are typically detected using fluorescence-based techniques, whereas SA-β-gal activity has traditionally been assessed by brightfield microscopy. This separation has limited the ability to evaluate multiple senescence features within the same cell. Reflected light imaging overcomes this limitation by enabling direct integration of SA-β-gal detection with fluorescence-based markers within a single imaging workflow, facilitating a more robust and biologically meaningful assessment of senescence in heterogeneous cell populations.

While Dedic and colleagues address several limitations of traditional SA-β-gal imaging, including improved sensitivity, the use of pH-matched controls for objective thresholding, and compatibility with immunocytochemistry, their reflected-light confocal method was optimized specifically for isolated mature adipocytes²⁶. However, senescence in adipose tissue is not restricted to adipocytes^34,35, and additional insight can be gained by examining the stromal vascular fraction (SVF). The SVF comprises preadipocytes, endothelial and fibroblast populations, and diverse immune cells, all of which may undergo senescence and contribute to adipose remodeling, inflammation, and altered metabolic signaling. Characterizing senescence within this heterogeneous compartment is therefore important for linking cell-type-specific processes to tissue-level and systemic outcomes.

Here, we extend reflected light SA-β-gal detection to SVF cells, enabling its application across multiple adipose tissue cell types. By combining this approach with immunocytochemistry, the method enables integrated single-cell analysis of SA-β-gal activity alongside lineage and functional markers in primary human samples.

Protocol

All experiments were performed in accordance with the statutes of the Declaration of Helsinki. The study was approved by the regional ethics committee in Stockholm, DNR 2012/50-31.

1. Cell preparation

1. Use stromal vascular fraction (SVF) cells isolated from human adipose tissue by collagenase digestion.
   ​NOTE: Cell yield varies depending on biopsy size and tissue composition; however, approximately 1.0 × 10⁶ SVF cells are typically obtained per 10 g of adipose tissue.

2. X-gal staining and slide preparations

1. Fix cells in 1 mL of 2% paraformaldehyde (PFA) in a 1.5 mL microcentrifuge tube for 20 min at room temperature (RT) on a rocker.
2. Centrifuge at 600 × g for 10 min and remove the supernatant.
3. Prepare the X-gal staining solution using the senescence β-Galactosidase staining kit according to the manufacturer’s instructions.
4. Divide the staining solution into 3 tubes (1 mL each) and adjust the pH to 5, 6, and 8 with 1 M HCl and 1 M NaOH. Measure pH using a calibrated pH meter.
5. Add X-gal solution to 0.5 × 10⁶ cells for condition and incubate for 24 h at 37 °C without CO₂.
6. Centrifuge at 600 × g for 10 min and remove supernatant.
7. Wash cells with PBS (pH 7), centrifuge at 600 × g for 10 min, and remove the supernatant.
8. Prepare primary antibody solution (CD31 [1:200] or IgG control [1:333]) in 5% normal donkey serum (NDS). Incubate cells overnight at 4 °C.
9. Wash cells 3 times in PBST (0.1%) for 10 min at RT on a rocker.
10. Prepare secondary antibody solution (anti-mouse Alexa Fluor 555 [1:500]) and wheat germ agglutinin (WGA [1:200]) in 5% NDS. Incubate for 1 h at RT on a rocker.
11. Centrifuge at 600 × g for 10 min and remove secondary antibody staining solution.
12. Wash the cells 3 times in PBST (0.1%) for 10 min on a rocker at RT.
13. Centrifuge at 600 × g for 10 min and remove the supernatant.
14. Resuspend cells in the antifade mountant.
15. Pipette the cell suspension onto microscope slides.
16. Allow the mounting medium to cure overnight at RT to immobilize the cells.
17. Add 100 µL of 100% glycerol, place a coverslip on top, and seal with nail polish.

3. Microscopy

NOTE: Image cells using a confocal microscope configured for reflected light detection. Reflected light signal should be collected at or near the excitation wavelength using a manufacturer-provided beam splitter or dichroic mirror configuration. Use a 30× objective with silicone oil (numerical aperture 1.05) for all experiments.

1. Select the T90/R10 beam splitter setting within the microscope software.
2. Use 633 nm excitation and collect reflected light between 635–645 nm.
3. Acquire Hoechst signal using 405 nm excitation.
4. Acquire WGA signal using 488 nm excitation.
5. Acquire CD31 signal using 555 nm excitation.
6. Set the Z step size to the system optimum (0.74 µm) and define a Z stack based on the membrane signal to capture the full cell volume.
7. Acquire images using identical acquisition settings for all samples. Save files in the native microscope format (e.g., .oir) without compression.

4. Image analysis

NOTE: Analyze images using ImageJ and CellProfiler.

1. Split channels in ImageJ to generate Z-stacks for each channel (405, 488, 555, 633).
2. Generate summed intensity projections using the Z Project function and export as 16-bit TIFF files.
3. Import images into CellProfiler (version 4.2.8) for automated cell segmentation and quantification of reflected light intensity per cell (Supplementary Figure 1 and Supplementary Figure 2).
4. Export intensity measurements to a spreadsheet for downstream analysis.

5. SA-β-gal scoring

1. Define the threshold for SA-β-gal positivity as the mean reflected light intensity of the pH 8 condition plus three standard deviations.
2. Classify cells in the pH 6 condition with intensity values above this threshold as SA-β-gal positive.

Results

Reflected light SA-β-gal staining of human SVF cells produced clear and reproducible patterns across the three pH-controlled conditions. CD31 stain was included to identify endothelial cells within the SVF, allowing these cells to be distinguished from other populations. Although only a small number of CD31-positive cells were present, the marker confirmed the presence of endothelial cells and demonstrated the compatibility of the method with heterogeneous primary samples.

The pH 5 condition generated a strong signal in most cells, consistent with high endogenous beta galactosidase activity at acidic pH (Figure 1Aa–Ac). At pH 6, the condition used to assess SA-β-gal activity, staining intensity was reduced compared to pH 5, and only a subset of cells displayed signal above background (Figure 1Ba–Bc). This shift in both signal intensity and frequency is consistent with the expected pH dependence of the assay and enables identification of cells with retained SA-β-galactosidase activity under suboptimal conditions. The pH 8 condition produced minimal reflected light signal, with most cells near or at background levels (Figure 1Ca–Cc). This condition was used to define the threshold for SA-β-gal positivity, indicated by the red dotted line in Figure 1Ac,Bc,Cc.

For quantitative analysis, the threshold for SA-β-gal activity was defined as the mean reflected light intensity of the pH 8 condition plus three standard deviations. Cells in the pH 6 condition with intensity values exceeding this threshold were classified as SA-β-gal positive. This approach provides an objective, data-driven cut-off that accounts for background signal and minimizes false-positive classification.

To assess detection sensitivity and compatibility with immunocytochemistry, reflected light and brightfield imaging were compared across all pH conditions. At pH 5 (Figure 2Aa–Ad), reflected light revealed strong X-gal precipitate, while brightfield detected only the most prominent deposits. At pH 6 (Figure 2Ba–Bd), reflected light identified SA-β-gal positive cells that were not reliably detectable by brightfield, where the signal appeared weak or diffuse. At pH 8 (Figure 2Ca–Cd), reflected light detected minimal signal, consistent with low enzymatic activity, whereas brightfield did not reveal detectable precipitate. These observations indicate that reflected light imaging provides improved detection of X-gal precipitate, particularly at intermediate and low activity levels.

The reflected light SA-β-gal assay was also performed in a 96-well plate (Supplementary Figure 3). This setup resulted in increased background signal due to reflection from the plastic surface, reducing contrast between positive and negative cells. Surface imperfections, such as scratches, further increased background reflection (Supplementary Figure 3, black arrows). As a result, quantitative analysis was less reliable compared to imaging on glass coverslips. If a plate format is required, glass-bottomed plates are recommended for optimal signal detection.

Together, these results show a clear separation between the strongly positive pH 5 condition, the intermediate signal at pH 6, and the minimal background at pH 8. This pattern is consistent with the expected behavior of the assay and supports the use of reflected light imaging for the detection of SA-β-gal activity in heterogeneous primary cell populations.

Figure 1: pH-dependent reflected light SA-β-gal detection in SVF cells.
WGA (Aa, Ba, Ca) and reflected light imaging (Ab, Bb, Cb) detect X-gal precipitate across pH 5, 6, and 8 conditions. Signal is strongest at pH 5, reduced at pH 6, and minimal at pH 8. Quantification of reflected light intensity (Ac, Bc, Cc) shows the distribution of β-galactosidase activity under each condition. Each dot represents a single cell, and the red dotted line indicates the threshold used to classify SA-β-gal-positive cells. Please click here to view a larger version of this figure.

Figure 2: Comparison of reflected light and brightfield SA-β-gal detection in SVF cells. Zoomed-in images of X-gal-stained SVF cells at pH 5, 6, and 8 show WGA staining (Aa, Ba, Ca), CD31 staining (Ab, Bb, Cb) to identify endothelial cells, reflected light imaging (Ac, Bc, Cc) to detect X-gal precipitate, and corresponding brightfield images (Ad, Bd, Cd). Reflected light imaging detects X-gal precipitate across conditions, whereas brightfield primarily reveals higher-intensity deposits and provides limited detection at lower signal levels. CD31 staining is observed in both pH 5 and pH 6 conditions, enabling identification of endothelial cells. Please click here to view a larger version of this figure.

Supplementary Figure 1: Overview of CellProfiler pipeline. Please click here to download this file.

Supplementary Figure 2: Workflow for nuclei and cell segmentation. Representative images of (A) Hoechst-stained nuclei and (B) corresponding nuclei segmentation. (C) Cell segmentation based on the membrane stain. (D) Overlay of nuclei and cell boundaries.Please click here to download this file.

Supplementary Figure 3: Reflected light SA-β-gal imaging in a plastic-bottom 96-well plate. (A) WGA staining and (B) reflected light imaging of X-gal-stained SVF cells in a standard plastic-bottom 96-well plate. An increased background signal is observed compared to imaging on glass, consistent with light reflection from the plastic surface, which reduces contrast for SA-β-gal detection. Surface scratches (black arrows) further increase the background signal (B). Please click here to download this file.

Discussion

Senescence-associated β-galactosidase is one of the most widely used markers for identifying senescent cells; however, traditional brightfield detection methods are limited by low sensitivity and a high degree of subjectivity. Reflected light imaging of X-gal-stained SVF cells provides a more sensitive and reproducible approach to measure SA-β-gal activity at the single cell level, while remaining compatible with fluorescent antibody staining. By optimizing this method for SVF, SA-β-gal activity can be detected alongside cell-specific markers such as CD31, enabling analysis of heterogeneous primary cell populations without the need for cell sorting.

Several technical parameters are critical for reliable implementation. Accurate pH calibration of the X-gal staining solution is essential, as small deviations can alter enzymatic activity and affect signal detection. The pH meter should be calibrated immediately prior to use with standard buffers (e.g., pH 4.0 and pH 7.0), and the pH of the staining solution should be verified after preparation. In addition, Z-stack acquisition must capture the full cell volume, as incomplete coverage leads to underestimation of reflected light intensity. Imaging parameters, including laser power and detector gain, should be optimized to maximize signal without saturation and kept consistent across samples.

Compared to alternative senescence assays, reflected light imaging offers several advantages. Brightfield SA-β-gal detection relies on visual identification of precipitate and therefore has limited sensitivity and is not readily compatible with multiplexed analysis. In contrast, reflected light imaging detects the optical density of the X-gal precipitate while preserving the pH-dependent enzymatic definition of SA-β-gal. This enables simultaneous detection of SA-β-gal activity alongside fluorescence-based markers, allowing senescence to be assessed using multiple features within individual cells.

Despite these advantages, the method has limitations. Reflected light imaging detects the optical density of the X-gal precipitate but cannot definitively distinguish this signal from other dense intracellular structures, making pH-controlled conditions essential. The pH 8 condition provides a baseline for background reflected light and supports discrimination between specific and nonspecific signals. The approach also requires a confocal microscope with appropriate reflected light detection capabilities, which may not be available on all systems. In addition, imaging surface properties can affect signal quality: plastic-bottom plates increase background due to reflection and reduce contrast, whereas glass surfaces provide more reliable detection.

This method builds on the work of Dedic et al., who established reflected light imaging for SA-β-gal detection in mature adipocytes, and extends the approach to a broader range of cell types²⁶. Application to freshly isolated SVF demonstrates that reflected light detection of X-gal precipitate is compatible with diverse primary human cell populations. The ability to combine SA-β-gal detection with lineage markers such as CD31 further enables analysis of defined subpopulations within heterogeneous tissues.

Although variability between donors is expected in primary human samples, the aim of this study is to establish the technical implementation of the method rather than to characterize biological variability. Future applications across multiple biological samples or experimental conditions should incorporate appropriate statistical analyses. In addition, validation in established senescence induction models and integration with orthogonal senescence markers will further refine the interpretation of SA-β-gal activity in complex cell populations.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1.5 mL tubes	BioSphere	72.706.200	Biosphere safeseal tube 1.5 mL, 250 PC
5 mL tubes	Eppendorf	30119401	Eppendorf tubes 5.0 mL microtube
96-well plate	Corning	4680	Corning 96-well high content ccreening microplates with film bottom
CD31 primary antibody	Abcam	Ab9498	Anti-CD31 antibody clone JC/70A
Confocal microscope with reflected light capabilities	Evident	FV-4000
Coverslips	Epredia	BB02400500A113MNZO	Coverslips 24 mm x 50 mm, #1.5 thickness (0.13-0.16 mm)
DPBS	Gibco	14190-136	DPBS, no calcium, no magnesium
Fixation buffer	Sigma Aldrich	1.00496.5000	Formaldehyde solution 4%, buffered, pH 6.9
HCl	Sigma-Aldrich	320331-500mL	Hydrochloric acid (37%)
Incubator	Stuart	S130H	Hybridization oven/shaker
Microscopy slides	Avantor VWR	631-1560	Single-frosted slides, coloured ends
Mounting media	Invitrogen	P36984	ProLong glass antifade mountant excitation/emission max 361/497 nm
Mouse IgG1 isotype control	R&D Systems	MAB002
NaOH	Sigma-Aldrich	S8045-500G	Sodium hydroxide ≥ 98%
Normal donkey serum	Jackson ImmunoResearch Labs	017-000-121
pH meter	MeterLab	R21M002	PHM210 standard pH meter
Pipette tips (1000 µL)	Biosphere	70.1186.210	1250µL filtered pipette tips
Pipette tips (200 µL)	Corning	4823	200µL filtered pipette tips
SABG Kit	Cell signaling	9860S	Senescence beta-galactosidase staining kit
Secondary antibody, Alexa Fluor 555	Invitrogen	A31570	Donkey anti-mouse IgG (H+L) highly cross-adsorbed excitation/emission max 553/568 nm
Tween-20	Fisher Scientific	10419000	Pure, Fisher chemical
Water bath	Rowe Scientific	SWB20D	Standard water bath
WGA	Invitrogen	W11261	Wheat Germ Agglutinin (WGA) Alexa Fluor 488 conjugate: Excitation/emission max 495/519 nm