Imaging enzymes at work: metabolic mapping by enzyme histochemistry

Abstract

For the understanding of functions of proteins in biological and pathological processes, reporter molecules such as fluorescent proteins have become indispensable tools for visualizing the location of these proteins in intact animals, tissues, and cells. For enzymes, imaging their activity also provides information on their function or functions, which does not necessarily correlate with their location. Metabolic mapping enables imaging of activity of enzymes. The enzyme under study forms a reaction product that is fluorescent or colored by conversion of either a fluorogenic or chromogenic substrate or a fluorescent substrate with different spectral characteristics. Most chromogenic staining methods were developed in the latter half of the twentieth century but still find new applications in modern cell biology and pathology. Fluorescence methods have rapidly evolved during the last decade. This review critically evaluates the methods that are available at present for metabolic mapping in living animals, unfixed cryostat sections of tissues, and living cells, and refers to protocols of the methods of choice.

Figure 1

Localization of the protein (A) and activity (B) of the proteinase cathepsin B in adjacent cryostat sections of human colon. (A) Immunohistochemical localization of cathepsin B protein using an immunogold method and silver enhancement (Koehler et al. 2000). Cathepsin B protein is represented by white dots visualized by reflection contrast microscopy. (B) Activity of cathepsin B as localized by metabolic mapping using N-Cbz-Ala-Arg-Arg-4-methoxy-2-naphthylamide as fluorogenic substrate and 5-nitrosalicylaldehyde as coupling reagent (Van Noorden and Frederiks 1992,2002). Yellow fluorescence represents cathepsin B activity, whereas green fluorescence is nonspecific nitrosalicylaldehyde staining. Epithelial cells (e) contain low levels of cathepsin B protein that is active only in the older epithelial cells of the mucosal surface (ms), where apoptosis occurs, and not in the crypts (c). Stroma of the lamina propria (lp) contains most cathepsin B protein, which shows activity in distinct areas only (asterisk). Bar = 50 μm.

Figure 2

In situ zymography of gelatinase (MMP-2 and MMP-9) activity using DQ-gelatin as substrate in human brain tumor tissue (glioblastoma multiforme). Imaging in time was performed using a Nikon C1 Eclipse confocal microscope (Nikon Instruments; Amstelveen, The Netherlands) equipped with Hamamatsu E717-63 photomultipliers (Hamamatsu; Hamamatsu City, Japan) and homemade controlled light exposure microscopy (CLEM) electronics. A Nikon Plan Apo 40×/1.0 objective and EZ-C1 3.90 software from Nikon were used. An acoustic optical modulator (Isomet; Springfield, VA) was placed in the optical path of a 488-nm argon-ion laser at 20 mW. Images were sampled at 288 × 288 × 7 (X, Y, Z) pixels using a 12-bit analog digital converter and a voxel size of 0.104 × 0.104 × 0.4 μm. The time between images was 45 sec. One hundred images were collected without (A) and with (B) the use of CLEM. The images shown are maximum-intensity projections. The fluorescent product of gelatinase activity is fluorescein, which fades due to excitation light exposure. Note the reduction in fading when using CLEM. Highest activity was found in a macrophage detected by immunohistochemistry, and low activity was found in glioma cells. (C) Quantification as relative fluorescence in time (Images) in the macrophage (black line) and glioma cells (green line) with CLEM and in the macrophage (red line) and glioma cells (blue line) without CLEM. Bar = 10 μm.

Figure 3

Metabolic mapping of NADP⁺-dependent isocitrate dehydrogenase (IDH) activity in time in human brain tumor (glioblastoma multiforme) with (A) and without (B) IDH1 mutation. Metabolic mapping was performed at 37C with a chromogenic method using nitro BT tetrazolium salt, isocitrate as substrate, and NADP as coenzyme. The formation of colored reaction product in time (sec) represents NADP-dependent IDH activity. Images are shown of the test reaction in the presence of isocitrate at 600, 1200, and 1800 sec of incubation and the control reaction in the absence of substrate at 1800 sec of incubation in non-mutated (A) and mutated (B) glioblastoma. (C) Quantification of the formation of colored reaction product (formazan) in time in non-mutated [wild-type (WT), test (T), and control (C)] and mutated [(M) T and C] glioblastoma was performed using image analysis and monochromatic light at 585 nm (Chieco et al. 1994,2001). Note the decreased activity in the IDH1-mutated sample (B,C) (Bleeker et al. in press). Bar = 100 μm.

Figure 4

Erythrocytes of a person with heterozygous glucose-6-phosphate dehydrogenase (G6PD) deficiency after metabolic mapping of G6PD activity showing a mixed population of stained G6PD-containing erythrocytes (arrows) and unstained G6PD-deficient erythrocytes (arrowheads) (Peters and Van Noorden 2009). Bar = 50 μm.

Figure 5

Images of fluorescence of [Ala-Pro]₂-cresyl violet, substrate of dipeptidyl peptidase IV (A), and cresyl violet, product of dipeptidyl peptidase IV activity (B), and a combination of images A and B in living human T-cells (C). Excitation was at 494 nm (A) and 594 nm (B), and emission at 550–580 nm (A) and at >620 nm (B), respectively. Substrate fluorescence was recorded in green (A,C) and product fluorescence in red (B) and yellow (C). Bar = 20 μm. With permission, Boonacker et al. J Histochem Cytochem 51:959–968, 2003.

Figure 6

Non-invasive imaging of brain tumors in mice at 2 weeks after stereotactic injection of human glioma cells (U87) into the brain that were transfected with the luciferase gene according to Kemper et al. (2006). Bioluminescence is shown in pseudocolors in the region of interest (ROI).

Figure 7

Fluorescent resorufin in a living rat hepatocyte generated by activity of ethoxyresorufin-O-deethylase representing P450-dependent monooxygenases (Taira et al. 2007). Accumulation of resorufin occurs mainly adjacent to the nuclei, suggesting localization in the endoplasmic reticulum. Bar = 10 μm. With permission, Taira et al. Cell Biol Toxicol 23:143–151, 2007.

Figure 8

Localization of phosphatase activity using ELF97 phosphate substrate in cultured human fibroblasts. The yellow fluorescent ELF97 product is present in granular intracellular form, probably representing acid phosphatase activity in lysosomes, and diffusely in cytoplasm and nuclei, probably representing unspecified phosphatases. Bar = 20 μm.

Figure 9

Gallery of confocal images of fluorescence in time of CD26/DPPIV–transfected intact living Jurkat cells using 20 μm (Ala-Pro)₂-cresyl violet as substrate to localize DPPIV activity. The gallery consists of images captured every 15 sec from 30 sec up to 165 sec after starting the incubation. Fluorescence of the reaction product, cresyl violet, is represented by pseudocolors, with black representing no fluorescence, yellow strongest fluorescence, and red in-between fluorescence. Bar = 15 μm.

Figure 10

Measuring peptide degradation in living cells. A peptide substrate with a quencher (Q) and a fluorescein (F) group attached to two different amino acids (A) is micro-injected into cells. The subsequent degradation by peptidases is visualized by an increase in fluorescence due to separation of the quencher from the fluorophore (B). Relative fluorescence can be measured per cell in time using image analysis (C). Red and green lines represent fluorescence in the two fluorescent cells in B. Bar = 10 μm.

Figure 11

Glucose-6-phosphatase (A,B) and thiamine pyrophosphatase (C) activity in rat liver at the light microscopical level (A) and electron microscopical level (B,C). Unfixed cryostat sections were incubated using a semipermeable membrane as tissue protectant in between gelled incubation medium and cryostat section (Song et al. 1996). The brown (A) electron-dense (B,C) reaction product is lead phosphate that is produced by phosphatase activity in the presence of glucose-6-phosphate (A,B) or thiamine phosphate (C) and lead ions. Glucose-6-phosphatase is more active in periportal zones (pp) than in pericentral zones (pc) of liver lobuli (A) and localized in the lumen of endoplasmic reticulum (er), where the enzyme is known to be present (B). When glucose-6-phosphate is replaced by thiamine phosphate, the electron-dense reaction product lead phosphate is present in the lumen of the Golgi apparatus (ga), where the enzyme is known to be localized. m, mitochondrion; n, nucleus; nu, nucleolus; g, granule. Bars: A = 200 μm; B,C = 1 μm.