Oxygen Insensitivity of the Histochemical Assay of Glucose-6-Phosphate Dehydrogenase Activity for the Discrimination Between Nonmalignant and Malignant Cells

Abstract

We review here the oxygen insensitivity of the histochemical assay of glucose-6-phosphate dehydrogenase (G6PDH) activity to detect cancer cells. This inexpensive and rapid assay can be performed within half an hour. Discrimination between cancerous and noncancerous cells is based on a combination of elevated G6PDH activity, decreased super-oxide dismutase (SOD) activity, and decreased lipid peroxidation in cancer cells. The test discriminates between adenomas and carcinomas of the colon with a certainty of >80% and has a high prognostic value for survival of colon cancer patients. Pancreatitis and pancreatic cancer are discriminated with a certainty of 100%. Therefore, the test can be applied by pathologists to provide additional information in difficult cases of diagnosis of cancer and for prognosis.

Keywords

cancer malignancy neoplasm diagnosis prognosis glucose-6-phosphate dehydrogenase enzyme histochemistry quantitation

Almost three decades ago, Altman (1970) discovered that oxygen had different effects on tetrazolium salt reduction in normal and cancer cells. They localized activity of glucose-6-phosphate dehydrogenase (G6PDH), the regulatory enzyme of the pentose phosphate pathway, with the tetrazolium salt neotetrazolium chloride (NT) in carcinomas of the breast. When the reaction was performed in an atmosphere of nitrogen, they observed activity both in normal epithelium and in cancer cells. However, when oxygen was present in the incubation medium, cancer cells still showed activity, whereas normal cells did not. This phenomenon is known as oxygen insensitivity; an example is shown in Figure 1. The oxygen insensitivity assay has been successful in specifically detecting cancer cells of the breast (Altman 1970; Petersen et al. 1985a,b; Smyth et al. 1987; Barrón et al. 1991), bronchus (Butcher 1979,1982), stomach (Ibrahim et al. 1983), colon (Ibrahim et al. 1983; Best et al. 1990; Griffini et al. 1994; Van Driel et al. 1997a), and pancreas (Van Driel 1998). Van Driel (1998) demonstrated that the test can also be used for objective discrimination between adenomas and carcinomas of the colon with a certainty of >80% and between pancreatitis and pancreatic adenocarcinomas with a certainty of 100%. The test also has a predictive value of survival of colon cancer patients, when combined with clinical parameters, that is far better than that obtained on the basis of clinical parameters alone. The test can be easily performed within 15-30 min (see Protocol I) and is inexpensive. We conclude that the test has great potential to obtain additional diagnostic information when discrimination between nonmalignancy and malignancy is difficult, such as in the case of pancreatitis and pancreatic adenocarcinoma.

Figure 1

An example^- of the oxygen sensitivity of normal cells and oxygen insensitivity of cancer cells in serial sections of a biopsy of the bronchus. (A) Hematoxylin-eosin-stained section. ca, carcinoma of the bronchus; e, epithelium of the bronchus. (B) Activity of G6PDH dehydrogenase as localized with the histochemical assay in the absence of oxygen. Formazan production is a measure of activity. Absorbance is measured with image analysis in encircled areas and is taken as reference (100%). (C) The same assay as in B, but now performed in the presence of oxygen. Absorbance in the encircled areas is 4% in normal epithelium and 83% in the cancer cells of that in the section in B. This is the residual activity. Bar = 20 μm.

Chemical Backgrounds of Oxygen Insensitivity

Tetrazolium salts have been applied successfully for decades for the histochemical localization of the activity of dehydrogenases (for review see Stoward and Pearse 1991). The principle of the method is that when a specific substrate (and co-enzyme when necessary) for a dehydrogenase is (are) present in an incubation medium, that dehydrogenase produces electrons by oxidation of its substrate. These electrons can be picked up by a water-soluble colorless tetrazolium salt that is also added to the medium. Reduction of the tetrazolium salt generates a water-insoluble, intensely colored formazan that precipitates at the site of dehydrogenase activity (Figure 2; for details of tetrazolium salt-based histochemical methods, see handbooks such as Lojda et al. 1979; Stoward and Pearse 1991; Chayen et al. 1991; Van Noorden and Frederiks 1992). For localization of dehydrogenase activity at the LM level, tetranitro BT or nitro BT, and at the EM level, BPST are recommended for at least two reasons. The first is that tetranitro BT, nitro BT, and BPST are reduced easily and give a high spatial resolution because formazan crystals are fine (Van Noorden and Butcher 1984; Seidler 1991). BPST is recommended for use at the EM level. Although it is expensive, it is the only tetrazolium salt that is not by itself osmiophilic but its formazan is, which has a great advantage for ultrastructural localization of dehydrogenase activity (Altman 1976; Frederiks et al. unpublished results). The second reason is that interference by oxygen in the tetrazolium reduction (oxygen sensitivity) can be blocked by using a high tetrazolium concentration in the medium. When 5 mM of either of the three tetrazolium salts is used, the effect of oxygen is negligible (Altman 1970; Butcher 1978; Van Noorden 1988; Van Noorden and Butcher 1989; Stoward and Pearse 1991). However, this is not the case for NT. Irrespective of the concentration of NT in the incubation medium, oxygen preferentially takes up electrons, leaving NT unreduced in normal tissue (Figure 3) (Butcher 1978; Van Noorden 1988; Van Noorden and Butcher 1989). When incubations are performed in media that are saturated with N₂ to remove all O₂, NT formazan production is a direct measure of dehydrogenase activity (Butcher and Van Noorden 1985).

Figure 2

Principle of the tetrazolium salt method to histochemically detect G6PDH activity. G6P, glucose-6-phosphate; PG, phosphogluconate; mPMS, 1-methoxyphenazine methosulfate; mPMS.H₂, reduced mPMS; NBT, nitro BT; TNBT, tetranitro BT; BPST, 2-(2-ben-zothiazolyl)-3-(4-phthalhydrazidyl)-5-styryl-tetrazolium; NT, neotetrazolium.

Figure 3

Principle of the interference of oxygen in neotetrazolium (NT) reduction in the histochemical assay of G6PDH activity in normal cells. Reduced 1-methoxyphenazine methosulfate (mPMS.H₂) reduces NT in one-electron steps at the time to produce a ·NT radical. This radical is reoxidized to NT by molecular oxygen (O₂) which, in turn, becomes reduced to an oxygen radical (O₂ ^·-). Superoxide dismutase (SOD) present in cells converts O₂ ^·- rapidly in hydrogen peroxide (H₂O₂) that can generate a hydroxyl radical (·OH). This radical causes lipid peroxidation (LPO) production in the cell. LPO rapidly inactivates G6PDH during the assay. Because G6PDH is elevated and SOD and LPO are decreased in cancer cells, formazan (NT.H₂) can be produced here.

Oxygen interference is strongly reduced in malignant tissues (oxygen insensitivity). It has always been assumed that an unknown factor is involved in NT reduction that is increased or decreased in cancer cells in comparison with normal cells. Because tetrazolium salts are reduced in steps of one electron at a time (Raap 1983; Seidler 1991) to produce unstable intermediates, tetrazolinyl radicals (·NT; Seidler and Van Noorden 1994), it was always believed that oxygen radicals (O₂ ^·-) were involved. ·NT can donate its unpared electron to O₂ to form O₂ ^- and NT. NT formazan is formed when NT. picks up a second electron. This takes place in normal cells in the absence of oxygen and in cancer cells both in the absence and presence of oxygen.

The involvement of O₂ ^·- in the oxygen sensitivity of normal cells vs oxygen insensitivity of cancer cells is indicated by the following facts. First, superoxide dismutase (SOD), which uses O₂ ^·- as substrate, inhibits tetrazolium reduction in vitro (Nishikimi et al. 1972; Ponti et al. 1978; Raap 1983; Van Noorden and Butcher 1989; Liochev et al. 1995). Second, SOD levels are decreased in cancer cells (Dionisi et al. 1975; Oberley and Buettner 1979; Sun 1990; Oberley et al. 1994; Van Driel et al. 1997b). Finally, addition of exogenous SOD to the incubation medium containing O₂ and NT to histochemically detect G6PDH activity in sections of colon cancer made the reaction in cancer cells oxygen sensitive as though the cancer cells had become normal cells (Best et al. 1990). At that stage, it was concluded that diminished levels of SOD in cancer cells fail to convert all O₂ ^·- generated during incubation to detect G6PDH activity and O₂ ^·- then returns its unpaired electron to a NT. radical to generate NT formazan (Best et al. 1990). The unknown factor involved in oxygen sensitivity of nonmalignant cells would therefore be high levels of SOD to quickly convert all O₂ ^·- generated and to prevent NT formazan production, whereas the reduced levels of SOD in cancer cells would not be capable of doing.

This hypothesis was tested further by replacing substrate and co-enzyme for G6PDH by those for lactate dehydrogenase (LDH). It was argued that when SOD is the unknown factor, it does not make a difference which dehydrogenase generates the electrons (Griffini et al. 1994). However, these authors clearly demonstrated that LDH was not suitable for the discrimination between normal and cancer cells because the reaction in normal cells was not O₂-sensitive. Therefore, it was concluded that the chemical backgrounds of the oxygen insensitivity test are more complex than the difference between normal and cancer cells in the content of one compound (being SOD). In a series of experiments, it was demonstrated that the chemical background for the discrimination between normal and cancer cells on the basis of the histochemical assay of G6PDH activity is multifactorial (Figure 4) (Van Driel 1998). First, a number of dehydrogenases in addition to G6PDH and LDH were tested for their usefulness in the test. It appeared that G6PDH, NAD-dependent aldehyde dehydrogenase, and glyceraldehyde-3-phosphate dehydrogenase, but not other dehydrogenases, were oxygen-sensitive in normal tissue. Only these three dehydrogenases and not the other dehydrogenases tested are rapidly inactivated by lipid peroxidation products (Kanazawa and Ashida 1991; Szweda and Stadtman 1993; Szweda et al. 1993; Uchida and Stadtman 1993). Therefore, we assumed that lipid peroxidation products are involved in the oxygen sensitivity of the histochemical G6PDH activity assay (see below).

Figure 4

Schematic representation of the chemical backgrounds of oxygen sensitivity of normal cells and oxygen insensitivity of cancer cells in the histochemical assay of G6PDH activity. For details see text. Continuous arrows, flow of electrons; dashed arrows, inhibitory effects. The thickness of the lines indicates the relative amounts of electrons or inhibition. For abbreviations, see legends to Figures 2 and 3.

Because G6PDH activity is often elevated in cancer cells (Bannasch et al. 1982; Van Noorden 1984; Stumpf and Bannasch 1994) and the two other lipid peroxidation-sensitive dehydrogenases are expressed at low levels in cancer cells, the histochemical assay of G6PDH activity is the assay of choice for the oxygen insensitivity test.

Second, the test is SOD-dependent. When exogenous SOD is added to the oxygen-containing medium, cancer cells start to behave like normal cells (Best et al. 1990). When an O₂ ^- scavenger, ascorbate (Nishikimi et al. 1972), is added to the oxygen-containing medium, normal cells start to behave like cancer cells. We concluded from these experiments that not only did rapid scavenging of O₂ ^.- by SOD play a role but also the formation of H₂O₂ as follows:

\begin{matrix} SOD \\ 2 O_{2}^{. -} + 2 H^{+} \to H_{2} O_{2} \end{matrix}

This was supported by performing the histochemical assay of G6PDH activity in the absence of oxygen but in the presence of the product of SOD activity, H₂O₂ (Van Driel 1998). H₂O₂ mimicked the reaction in the presence of oxygen in a dose-dependent manner: normal cells did not exhibit reaction product, whereas cancer cells did. It was found that levels of both CuZn-SOD and Mn-SOD were decreased significantly in cancer cells (Van Driel et al. 1997b) in the same colon carcinoma samples that were used for these experiments.

Third, the test is dependent on lipid perodixation capacity. H₂O₂ is the substrate for ·OH formation which, in turn, is the most important substrate for lipid peroxidation (Michiels et al. 1994). It was proved that the lipid peroxidation capacity, as determined by quantitative histochemical means (Thomas et al. 1994), was decreased colon carcinoma biopsy specimens (Van Driel 1998). Furthermore, when the content of polyunsaturated fatty acids (PUFAs) in normal tissue was manipulated by diet (Van Noorden 1995), its oxygen sensitivity was manipulated as well. PUFAs are very sensitive to lipid peroxidation (Michiels et al. 1994), and normal rat liver tissue with high concentrations of PUFAs was extremely sensitive to oxygen or H₂O₂, whereas liver tissue containing low amounts of PUFAs was less sensitive to oxygen or H₂O₂.

Fourth, the test is NT-dependent. The test is NT-dependent not only because of oxygen interference in NT reduction via O₂ ^·- formation (Figure 4) but also because in vitro experiments showed that NT in combination with oxygen or H₂O₂ has an inhibiting effect on G6PDH. This inhibition did not occur when NT was replaced by nitro BT or G6PDH by LDH. We concluded on the basis of these experiments that the histochemical assay of G6PDH activity with NT as tetrazolium salt in the presence of oxygen does not produce formazan in normal cells because O₂ ^·- radicals are generated via the intermediate NT. radical. Then the O₂ ^·- radicals are rapidly converted by SOD to produce H₂O₂ and subsequently lipid peroxidation products that inactivate G6PDH during the assay. Moreover, G6PDH is directly inhibited by NT in combination with oxygen. In cancer cells, formazan is produced in the presence of oxygen because G6PDH activity is elevated and SOD activity and the lipid peroxidation capacity are reduced (Figure 4).

Diagnostic Value of the Oxygen Insensitivity Test in Cancer

Until recently, the stage in which carcinogenetic oxygen insensitivity starts to occur was not known. Because colorectal cancer evolves through a series of morphologically distinct stages, known as the adenoma-carcinoma sequence (Fearon and Vogelstein 1990), it provides a unique model to establish the diagnostic and prognostic value of the oxygen insensitivity assay. Van Driel et al. (1997a) applied the test to normal colon mucosa, hyperplastic mucosa, dysplastic mucosa, and colorectal carcinomas from humans and mice to determine the stage in which oxygen insensitivity appears. In both human and murine colorectal carcinogenesis, residual activity was always <20% in non-neoplastic tissues. Residual activity is defined as the ratio of the amounts of formazan produced in the presence and in the absence of oxygen × 100 (percentage). Residual activity is determined by incubating serial sections to demonstrate G6PDH activity in the presence and absence of oxygen. Formazan production in the same areas of those serial sections is determined as absorbance using image analysis (Figure 1). The ratio is determined and multiplied by 100 to yield the residual activity. When the residual activity is <20%, cells are considered to be nonmalignant, and they are considered to be malignant when the percentage is >20% (Griffini et al. 1994; Van Driel et al. 1997a). Oxygen insensitivity was first observed in adenomas; 50% of all primary human adenomas were oxygen-insensitive, but the residual activity never exceeded 35%. All carcinomas were oxygen-insensitive in both studies. Ninety-five percent of colorectal carcinomas showed a residual activity >35%. Therefore, when the residual activity is >35%, cells can be considered to be malignant.

In contrast to colon, one cannot always rely on morphology to diagnose cancer when chronic pancreatitis is also present. Both chronic pancreatitis and pancreatic carcinoma are characterized by fibrosis, atrophy of acini and prominence of islets of Langerhans (Stirling 1991). The oxygen insensitivity test was applied by Van Driel (1998) for differential diagnosis between the two diseases. In general, epithelial cells in biopsies and, more important, in brush cytology specimens of normal pancreas and chronic pancreatitis were oxygen-sensitive (residual activity <20%), whereas 94% of the pancreatic carcinomas showed residual activity >35%, which is completely in agreement with the findings in colorectal carcinomas. Therefore, the oxygen insensitivity test is considered a helpful additional tool for the pathologist to differentially diagnose pancreatitis and pancreatic cancer.

Prognostic Value of the Oxygen Insensitivity Test

An attempt was made by Van Driel (1998) to improve estimation of survival time of colorectal cancer patients by means of the oxygen insensitivity test and a standardized and self-learning classification system, CLASSIF1 (Valet et al. 1993, Valet and Höffkes 1997, Valet et al., in press). Classification was based on multiparameter analysis of 24 quantitative histochemical parameters which were directly or indirectly related to the oxygen insensitivity test, and on five clinical parameters. The prognostic value was highest when the classification system selected six parameters: oxygen insensitivity of G6PDH activity, SOD levels, lipid peroxidation capacity, lymph node metastasis, Dukes' stage (Dukes 1932), and age. Fatal outcome in 64% of the diseased patients and favorable outcome in 100% of the surviving patients were correctly predicted (overall correct classification 82%), which was distinctly better than classification based on Dukes' staging alone (overall recognition 64%), which is considered thus far to be the most accurate prognostic tool. This study indicated that oxygen insensitivity of G6PDH is related to malignancy and that it provides useful information for diagnosis and prognosis.

Concluding Remarks

Attempts have been made to elucidate the metabolic backgrounds that are responsible for the phenomenon of oxygen sensitivity of normal tissues and oxygen insensitivity of carcinomas, because a tumor marker as diagnostic or prognostic tool must have a solid basis. Biochemical experiments demonstrated that the histochemical oxygen insensitivity test is dependent on G6PDH activity, decreased SOD activity, and decreased lipid peroxidation capacity in cancer cells compared with their normal counterparts. The discriminatory power of the oxygen insensitivity test for normal and malignant epithelial cells is very useful in differential diagnosis of chronic pancreatitis and pancreatic carcinoma. Furthermore, the oxygen insensitivity of the histochemical assay of G6PDH activity proved to be a strong prognosticator for survival of patients with colorectal cancer, which is far better than Dukes' classification as single prognosticator. At present, it is probably the safest objective discriminator between normal and cancer cells and has great potential for use in the pathology laboratory to provide diagnostic and prognostic information. Because the assay is easy and rapid to perform (15-30 min) and because of its low cost, we recommend its application very strongly. Finally, we give below a detailed protocol on how to perform the assay (Protocol I; Van Noorden and Frederiks 1992).

Protocol I

The oxygen-sensitive tetrazolium salt method to detect glucose-6-phosphate dehydrogenase activity for the discrimination of normal and malignant cells (Van Noorden and Frederiks 1992) is presented below.

Dissolve 18 g polyvinyl alcohol (PVA; Sigma, St Louis, MO); number average mol wt 40,000 in 100 ml phosphate buffer (pH 7.45) under continuous stirring and heating in a water bath (80C) until a clear solution is obtained. (This type of PVA from Sigma has a number average molecular weight of 40,000 and a weight average molecular weight of 70,000-100,000 (Dr. A. O'Connor; Sigma, personal communication). In the past, the number average molecular weight was used to express the molecular weights of polymers, whereas now the weight average molecular weight is preferred.)

Store the clear PVA-containing solution at 60C in air-tight vials.

Cool the desired volume of the solution to 37C before incubation.

To 1 ml PVA-containing buffer, add 10 μl glucose-6-phosphate (300 mg/ml distilled water; final concentration 10 mM) and 10 μl NADP (63 mg/ml distilled water; final concentration 0.8 mM). Then add 10 μl sodium azide (33 mg/ml distilled water; final concentration 5 mM), 10 μl 1-methoxyphenazine methosulfate (15 mg/ml distilled water; final concentration 0.45 mM), 10 μl magnesium chloride (102 mg/ml distilled water; final concentration 5 mM; G6PDH activity is elevated in the presence of Mg²⁺ ions), and 40 μl neotetrazolium chloride (3 mg/ml dissolved in 20 μl ethanol and 20 μl dimethylformamide by gentle heating; final concentration 5 mM; Polysciences, Northampton, UK). When neotetrazolium is used from other suppliers, it may be contaminated with impurities and may need purification (Altman 1976).

Divide the medium into two equal parts and saturate one part with 100% oxygen and the other with 100% nitrogen. Because of the viscosity of PVA-containing media, use a tonometer for gassing the media (Butcher 1978). When a tonometer is not available, the media must stand in air-tight vials for 15-30 min after the gas has been bubbled through, to eliminate bubbles from the viscous medium. Gas bubbles in the medium tend to stick to the sections during incubation, thus interfering with the staining reaction.

Mix the compounds thoroughly with the buffer solution, using a spatula after each addition because of the viscosity of the medium.

Use unfixed cells or serial cryostat sections of the tissue to be diagnosed (8-10 μm thick).

Use plastic rings around the sections and stick them to the object glass with grease.

Pour the media in the rings onto the cells or sections, put a coverslip on top of the rings to prevent the gases from escaping from the incubation media, and incubate for 10 min at 37C.

10.

Rinse off media with 100 mM phosphate buffer (pH 5.3) at 60C.

11.

Mount cells or sections in glycerin-gelatin.

12.

Control incubations should be carried out in the absence of substrate or in the absence of substrate and co-enzyme.

13.

Analyze the cells or sections: malignant cells contain substantial amounts of formazan after incubation in oxygen, whereas normal cells do not. Select cells or areas in the sections to be measured in the specimens incubated both in the presence and in the absence of oxygen and determine by image analysis the residual activity as the ratio of the absorbance of the cells incubated in the presence and in the absence of oxygen X 100. An example is shown in Figure 1.

References

Altman

(1970) On the oxygen-sensitivity of various tetrazolium salts. Histochemie 22:256–261

Altman

(1976) Tetrazolium salts and formazans. Prog Histochem Cytochem 9/3:1–56

Altman

Bitensky

Butcher

Chayen

(1976) Integrated cellular chemistry applied to malignant cells. In Evans

DMD

, ed. Cytology Automation. Edinburgh, Livingstone, 82–97

Bannasch

Moore

Klimek

Zerban

(1982) Biological markers of preneoplastic foci and neoplastic nodules in rodent liver. Toxicol Pathol 10:19–34

Barron

Smyth

PPA

McDermott

Tobbia

Higgins

NJO

(1991) Quantitative cytochemistry of glucose-6-phosphate dehydrogenase in benign and malignant breast tumours. Eur J Cancer 27:985–989

Best

Das

Patel

HRA

Van Noorden

CJF

(1990) Quantitative cytochemical detection of malignant and potentially malignant cells in the colon. Cancer Res 50:5112–5118

Butcher

(1978) Oxygen and the production of formazan from neotetrazolium chloride. Histochemistry 56:329–340

Butcher

(1979) The oxygen insensitivity phenomenon as a diagnostic aid in carcinoma of the bronchus. In Pattison

Bitensky

Chayen

, eds. Quantitative Cytochemistry and Its Applications. New York, Academic Press, 241–251

Butcher

(1982) The microsomal NADPH oxidation system in lung cancer. In Cumming

Bonsignore

, eds. Cellular Biology of the Lung. New York, Plenum Press, 439–456

10.

Butcher

Van Noorden

CJF

(1985) Reaction rate studies of glucose-6-phosphate dehydrogenase activity in sections of rat liver using four tetrazolium salts. Histochem J 17:171–190

11.

Chayen

Bitenski

Butcher

(1991) Practical Histochemistry. New York, John Wiley & Sons

12.

Dionisi

Galeotti

Terranova

Azzi

(1975) Superoxide radicals and hydrogen peroxide formation in mitochondria from normal and neoplastic tissues. Biochim Biophys Acta 403:292–300

13.

Dukes

(1932) The classification of cancer of the rectum. J Pathol 35:323–332

14.

Fearon

Vogelstein

(1990) A genetic model for colorectal carcinogenesis. Cell 61:759–767

15.

Griffini

Vigorelli

Jonges

Van Noorden

CJF

(1994) The histochemical G6PDH reaction but not the LDH reaction with neotetrazolium is suitable for the oxygen sensitivity test to detect cancer cells. J Histochem Cytochem 42:1355–1363

16.

Ibrahim

Husain

Bitensky

Chayen

(1983) A modified tetrazolium reaction for identifying malignant cells from gastric and colonic cancer. J Clin Pathol 36:133–136

17.

Kanazawa

Ashida

(1991) Target enzymes on hepatic dysfunction caused by dietary products of lipid peroxidation. Arch Biochem Biophys 288:71–78

18.

Liochev

Batinic-Haberle

Fridovich

(1995) The effects of detergents on the reduction of tetrazolium salts. Arch Biochem Biophys 324:48–52

19.

Lojda

Gossrau

Schiebler

(1979) Enzyme Histochemistry. A Laboratory Manual. New York, Springler Verlag

20.

Michiels

Raes

Toussaint

Remacle

(1994) Importance of SE-glutathione peroxidase, catalase, and Cu/Zn-Sod for cell survival against oxidative stress. Free Rad Biol Med 17:235–248

21.

Nishikimi

Appaji

Yagi

(1972) The occurrence of superoxide anion in the reaction of reduced phenazine methosulfate and molecular oxygen. Biochem Biophys Res Commun 46:849–854

22.

Oberley

Buettner

(1979) Role of superoxide dismutase in cancer: a review. Cancer Res 39:1141–1149

23.

Oberley

Schultz

Oberley

(1994) In vitro modulation of antioxidant enzyme levels in normal hamster kidney and estrogen-induced hamster kidney tumor. Free Rad Biol Med 16:741–751

24.

Petersen

Hoyer

Hilgers

Briand

Van Deurs

(1985a) Characterization of epithelial cell islets in primary monolayer cultures of human breast carcinomas by the tetrazolium reaction for glucose-6-phosphate dehydrogenase. Virchows Arch [B] 50:27–42

25.

Petersen

Hoyer

Van Deurs

(1985b) Effect of oxygen on the tetrazolium reaction for glucose-6-phosphate dehydrogenase in cryosections of human breast carcinoma, fibrocystic disease and normal breast tissue. Virchows Arch [B] 50:13–25

26.

Ponti

Dianzani

Cheeseman

Slater

(1978) Studies on the reduction of nitroblue tetrazolium chloride mediated through the action of NADH and phenazine methosulfate. Chem Biol Interact 23:281–291

27.

Raap

(1983) Studies on the phenazine methosulfate-tetrazolium salt capture reaction in NAD(P) + -dependent dehydrogenase cytochemistry. III. The role of superoxide in tetrazolium reduction. Histochem J 15:977–986

28.

Seidler

(1991) The tetrazolium-formazan system: design and histochemistry. Prog Histochem Cytochem 24/1:1–86

29.

Seidler

Van Noorden

CJF

(1994) On the mechanism of the multistep reduction of tetrazolium salts with special reference to the involvement of tetrazolium radicals. Acta Histochem 96:43–49

30.

Smyth

PPA

Barron

Tobbia

O'Higgins

(1987) Cytochemical investigation of glucose-6-phosphate dehydrogenase activity in rat mammary tissue. Br J Exp Pathol 68:45–52

31.

Stirling

(1991) The exocrine pancreas: neoplasms in liver, biliary tract and pancreas. In Wight

DGD

, ed. Systemic Pathology. Vol 11. Edinburgh, Churchill Livingstone, 665–710

32.

Stoward

Pearse

AGE

(1991) Histochemistry. Theoretical and Applied. Vol 3. 4th ed. New York, Churchill Livingstone

33.

Stumpf

Bannasch

(1994) Overexpression of glucose-6-phosphate dehydrogenase in rat hepatic preneoplasia and neoplasia. Int J Oncol 5:1255–1260

34.

Sun

(1990) Free radicals, antioxidant enzymes, and carcinogenesis. Free Rad Biol Med 8:583–599

35.

Szweda

Uchida

Tsai

Stadtman

(1993) Inactivation of glucose-6-phosphate dehydrogenase by 4-hydroxy-2-nonenal. J Biol Chem 268:3342–3347

36.

Szweda

Stadtman

(1993) Oxidative modification of glucose-6-phosphate dehydrogenase from Leuconostoc mesenteroides by an iron(II)-citrate complex. Arch Biochem Biophys 301:391–395

37.

Thomas

Frederiks

Van Noorden

CJF

Bosch

Pompella

(1994) NADPH-dependent lipid peroxidation capacity in unfixed tissue sections: characterization of the pro-oxidizing conditions and optimization of the histochemical detection. Histochem J 26:189–196

38.

Uchida

Stadtman

(1993) Covalent attachment of 4-hydroxynonenal to glyceraldehyde-3-phosphate dehydrogenase. A possible involvement of intra- and intermolecular cross-linking reaction. J Biol Chem 268:6388–6393

39.

Valet

Höffkes

H-G

(1997) Automated classification of patients with chronic lymphocytic leukemia and immunocytoma from flow cytometric three-color immunophenotypes. Cytometry 30:275–288

40.

Valet

Rothe

Kellermann

(in press) Risk assessment for intensive care patients by automated classification of flow cytometric oxidative burst, serine and cysteine proteinase activity measurements using CLASSIF1 triple matrix analysis. In Robinson

Babcock

, eds. Cytometric Cellular Analysis. New York, Wiley-Liss

41.

Valet

Tschöpe

Gabriel

Rothe

Kellermann

Kahle

(1993) White cell and thrombocyte disorders: standardized, self-learning flow cytometric list mode data classification with the CLASSIF1 program system. Ann NY Acad Sci 677:233–251

42.

Van Driel

BEM

(1998) The oxygen insensitivity assay of G6PDH activity in human cancer pathology. Clinical and chemical aspects. Thesis, University of Amsterdam, Amsterdam, The Netherlands

43.

Van Driel

BEM

De Goeij

AFPM

Song

De Bruine

Van Noorden

CJF

(1997a) Development of oxygen insensitivity of the quantitative histochemical assay of G6PDH activity during colorectal carcinogenesis. J Pathol 182:398–403

44.

Van Driel

BEM

Lyon

Hoogenraad

DCJ

Anten

Hansen

Van Noorden

CJF

(1997b) Expression of CuZn- and Mn-super-oxide dismutase in human colorectal neoplasms. Free Rad Biol Med 23:435–444

45.

Van Noorden

CJF

(1984) Histochemistry and cytochemistry of glucose-6-phosphate dehydrogenase. Prog Histochem Cytochem 15/4:1–85

46.

Van Noorden

CJF

(1988) On the role of oxygen in dehydrogenase reactions using tetrazolium salts. Histochem J 20:587–593

47.

Van Noorden

CJF

(1995) Effects of n-3 and n-6 polyunsaturated fatty acids-enriched diets on lipid metabolism in periportal and pericentral compartments of female rat liver lobules and its consequences for cell proliferation after partial hepatectomy. J Lipid Res 36:1708–1720

48.

Van Noorden

CJF

Butcher

(1984) Histochemical localization of NADP-dependent dehydrogenase activity with four different tetrazolium salts. J Histochem Cytochem 32:998–1004

49.

Van Noorden

CJF

Butcher

(1989) The involvement of superoxide anions in the nitroblue tetrazolium chloride reduction mediated by NADH and phenazine metosulfate. Anal Biochem 176:170–174

50.

Van Noorden

CJF

Frederiks

(1992) Enzyme Histochemistry: A Laboratory Manual of Current Methods. Oxford, BIOS