Comparison between pimonidazole binding, oxygen electrode measurements, and expression of endogenous hypoxia markers in cancer of the uterine cervix

BACKGROUND

Although tumor hypoxia has been associated with a more aggressive phenotype and lower cure rate, there is no consensus as to the method best suited for routine measurement. Binding of the chemical hypoxia marker, pimonidazole, and expression of the endogenous hypoxia markers HIF-1alpha and CAIX were compared for their ability to detect hypoxia in tumor biopsies from 67 patients with advanced carcinoma of the cervix.

METHODS

Two biopsies were taken one day after administration of pimonidazole and were analyzed for pimonidazole binding using flow cytometry or immunohistochemistry. CAIX and HIF-1alpha expression and degree of colocalization were measured in sequential antibody-stained sections. Patient subsets were examined for tumor oxygen tension using an Eppendorf electrode, S phase DNA content, or change in HIF-1alpha expression over the course of treatment.

RESULTS

Approximately 6% of the tumor area stained positive for pimonidazole, HIF-1alpha, or CAIX. The CAIX positive fraction correlated with the pimonidazole positive fraction (r = 0.60). Weaker but significant correlations were observed between pimonidazole and HIF-1alpha (r = 0.31) and CAIX and HIF-1alpha (r = 0.41). Taking the extent of marker colocalization into consideration increased the confidence that all markers were identifying hypoxic regions. Over 65% of stained areas showed a high degree of colocalization with the other markers. Oxygen microelectrode measurements and S phase fraction were not correlated with the hypoxic fraction measured using the three hypoxia markers. HIF-1alpha levels tended to decrease with time after the start of therapy.

CONCLUSIONS

Endogenous hypoxia marker binding shows reasonable agreement, in extent and location, with binding of pimonidazole. CAIX staining pattern is a better match to the pimonidazole staining pattern than is HIF-1alpha, and high CAIX expression in the absence (or low levels) of HIF-1alpha may indicate a different biology.

Evidence that involucrin, a marker for differentiation, is oxygen regulated in human squamous cell carcinomas

The majority of hypoxic cells in squamous cell carcinomas of the head and neck and cervix express involucrin, a molecular marker for differentiation. This raises the question of whether involucrin is an oxygen-regulated protein and, if so, whether it could serve as an endogenous marker for tumour hypoxia. Consistent with oxygen regulation, involucrin protein was found to increase with increasing hypoxia in confluent cultures of moderately differentiated human SCC9 cells. Cells harvested at the point of confluence and exposed to graded concentrations of oxygen revealed a K_m of approximately 15 mmHg for involucrin induction. This is similar to K_ms for HIF-1α, CAIX and VEGF. Involucrin induction showed a steep dependence on pO₂ with a transition from minimum to maximum expression occurring over less than an order of magnitude change in pO₂. In contrast to SCC9 cells, involucrin was not induced by hypoxia in poorly differentiated SCC4 cells. It is concluded that involucrin is an oxygen-regulated protein, but that differentiation modulates its transcription status with respect to hypoxia induction.

Hypoxia and perfusion measurements in human tumors--initial experience with pimonidazole and IUdR

We describe our preliminary studies on the development of methods to measure hypoxia in standard paraffin sections of human tumors. Three parameters were investigated. First, image analysis of tumor vascularity yielded the parameter diffusion limited fraction (DLF), which is the amount of tumor tissue greater than a fixed distance from the nearest blood vessel. Secondly, the amount of tumor tissue stained with antibodies against bound reduced products of the bioreductive marker pimonidazole was assessed. Finally, the fraction of blood vessels showing no surrounding tumor tissue labeled with lUdR, a cell kinetic marker, was measured. DLF and pimonidazole monitor primarily chronic hypoxia, while it is hypothesized that the IUdR-negative fraction monitors acute hypoxia. Feasibility was demonstrated in a series of 10 esophageal and 10 rectal tumors (no drug administration), 10 cervix tumors (pimonidazole) and 14 head and neck tumors (pimonidazole and lUdR). Significant differences between tumors were found for all parameters. DLF correlated significantly with the pimonidazole fraction when all images of all tumors were included, although mean values per tumor showed no correlation. The IUdR-negative fraction did not correlate with either of the other two parameters. We conclude that it is feasible to measure hypoxia-related, and possibly perfusion-related, parameters on paraffin sections for predictive purposes, although each method needs further validation. Each parameter will be correlated with outcome in a larger study on head and neck tumors treated with surgery with or without postoperative radiotherapy.

Carbonic anhydrase 9 as an endogenous marker for hypoxic cells in cervical cancer

The presence of radiation-resistant hypoxic cells in some solid tumors is known to predict for relapse after radiotherapy. Use of an endogenous marker of hypoxia would be a convenient alternative to current methods that measure tumor oxygenation, provided the marker could be shown to reliably identify viable, radiation-resistant, hypoxic cells. Carbonic anhydrase 9 (CA9) is a transmembrane protein overexpressed in a wide variety of tumor types and induced by hypoxia. Using a monoclonal antibody and cell sorting, CA9-positive cells in SiHa cervical carcinoma xenografts growing in immunodeficient mice were found to be clonogenic, resistant to killing by ionizing radiation, and preferentially able to bind the hypoxia marker pimonidazole. CA9 and pimonidazole immunostaining were compared in formalin-fixed sections from tumors of 18 patients undergoing treatment for cancer of the cervix. Excellent colocalization was observed, although the area of the tumor section that bound anti-CA9 antibodies represented double the number of cells that bound anti-pimonidazole antibodies. Occasional regions staining with pimonidazole but not CA9 could be indicative of transient changes in tumor perfusion. Results support the hypothesis that CA9 is a useful endogenous marker of tumor hypoxia.

Tumor vasculature is targeted by the combination of combretastatin A-4 and hyperthermia

BACKGROUND AND PURPOSE

Combretastatin A-4 disodium phosphate (CA-4) enhances thermal damage in s.c. BT(4)An rat gliomas. We currently investigated how CA-4 and hyperthermia affect the tumor microenvironment and neovasculature to disclose how the two treatment modalities interact to produce tumor response.

METHODS

By confocal microscopy and immunostaining for von Willebrand factor, we examined the extent of vascular damage subsequent to CA-4 (50 mg/kg) and hyperthermia (waterbath 44 degrees C, 60 min). The influence on tumor oxygenation was assessed using interstitial pO(2)-probes (Licox system) and by immunostaining for pimonidazole. We examined the direct effect of CA-4 on the tumor cell population by flow cytometry (cell cycle distribution) and immunostaining for beta-tubulin (cytoskeletal damage).

RESULTS

Whereas slight vascular damage was produced by CA-4 in the BT(4)An tumors, local hyperthermia exhibited moderate anti-vascular activity. In tumors exposed to CA-4 3 h before hyperthermia, massive vascular damage ensued. CA-4 reduced the pO(2) from 36.1 to 17.6 mmHg (P=0.01) in the tumor base, and tumor hypoxia increased slightly in the tumor center (pimonidazole staining). Extensive tumor hypoxia developed subsequent to hyperthermia or combination therapy. Despite a profound influence on beta-tubulin organization in vitro, CA-4 had no significant effect on the cell cycle distribution in vivo.

CONCLUSION

Our results indicate that the anti-vascular activity exhibited by local hyperthermia can be augmented by previous exposure to CA-4.

Cyclosporin A causes a hypermetabolic state and hypoxia in the liver: prevention by dietary glycine

Acute cyclosporin A (CsA) treatment inhibits mitochondrial respiration, yet effects of chronic treatment remain unclear. Accordingly, the effects of chronic CsA on oxygen metabolism in perfused rat liver and isolated mitochondria were investigated. Basal rates of oxygen uptake of around 120 micromol/g/h in isolated perfused livers from vehicle-treated controls were elevated about 1.6-fold by chronic CsA treatment. In the presence of ammonium chloride, a substrate for urea synthesis, oxygen uptake was about 150 micromol/g/h and was increased about 1.7-fold by CsA, indicating that chronic CsA treatment causes a robust hypermetabolic state in the liver. In isolated mitochondria, state 3 rates of oxygen uptake were increased about 1.6-fold by chronic CsA treatment. Since significant increases in oxygen consumption could cause hypoxia, the hypoxia marker pimonidazole was given. Pimonidazole binding in the liver was increased about 3-fold by chronic CsA. Moreover, intracellular calcium in Kupffer cells isolated from vehicle-treated rats was not altered by CsA addition; however, in cells isolated from chronic CsA-treated rats, CsA increased intracellular calcium about 15-fold and prostaglandin E(2) (PGE(2)) production 3.5-fold. Importantly, dietary glycine (5%) largely blocked chronic CsA-induced activation of Kupffer cells, blunted production of PGE(2), prevented the hypermetabolic state, and minimized tissue hypoxia. Taken together, it is concluded that chronic CsA treatment causes a hypermetabolic state leading to hypoxia and injury to the liver. It is hypothesized that CsA activates Kupffer cells and increases production of PGE(2), which alters mitochondria leading to a hypermetabolic state. Glycine inhibits activation of Kupffer cells thus preventing liver injury.

Optimizing the use of combined radioimmunotherapy and hypoxic cytotoxin therapy as a function of tumor hypoxia

Combined radioimmunotherapy (RAIT) and hypoxic cytotoxin therapy (SR4233 or NLCQ-1) have been evaluated with both modalities administered on the same day with only moderate improvement compared with the effects of RAIT alone. In a series of studies using oxygen electrodes, immunohistochemistry and radiotracers, we have demonstrated that RAIT induces a prolonged state of hypoxia in most tumors, without affecting the pO(2) levels in normal tissues. Using serial microelectrode measurements through subcutaneous (s.c.) GW-39 human colonic xenografts, we established that the median pO(2) was unrelated to the initial size of the tumor, over a range of sizes from 1.0 to 4.0 cm. Fourteen days after mice were given a 240-microCi dose of (131)I-MN-14 anti-carcinoembryonic antigen immunoglobulin G, their median pO(2) declined from 26.1 +/- 9.6 mmHg to 9.8 +/- 3.9 mmHg (p < 0.001). Using the radiotracer (3)H-MISO that accumulates in hypoxic regions, uptake in GW-39, LoVo and LS174T s.c. human colonic tumors increased 3.0- to 4.2-fold from day 14 through day 28 post-RAIT, but uptake of (3)H-MISO in CALU-3 tumors remained unchanged after RAIT. Normal tissue (liver, kidney, lung) uptake of (3)H-MISO did not exhibit significant changes. The increase in tumor hypoxia was also demonstrated visually using anti-PIMO staining of tumor sections. We postulated that sequential delivery of the 2 therapeutic agents, with the hypoxic cytotoxin given 2 weeks after RAIT when tumor pO(2) levels were at their nadir, would improve the therapeutic response above either modality alone or above the 2 agents delivered on the same day. Tumor growth was compared in mice given either RAIT or cytotoxin alone, the combined treatment on the same day or with the cytotoxin delivered 14 days after RAIT. Tumor size on day 35 for RAIT-treated and SR4233-treated GW-39 were 3.56 +/- 0.40 and 7.98 +/- 2.50 cm(3). When RAIT + SR4233 were delivered on the same day, tumor size dropped to 2.78 +/- 0.80 cm(3). If RAIT was given on day 0 and SR4233 on day 14, size further declined further to 1.74 +/- 0.32 cm(3) (p < 0.05 compared with same day delivery). For LS174T, tumor size on day 28 for RAIT-treated and SR4233-treated tumors were 1.14 +/- 0.36 cm(3) and 3.65 +/- 0.78 cm(3), respectively. When RAIT + SR4233 were delivered on the same day, size was 0.51 +/- 0.174 cm(3). If RAIT was dosed on day 0 and SR4233 was given on day 14, tumor size was 0.13 +/- 0.07 cm(3) (p < 0.05). Similar results were obtained for LoVo, but not for CALU-3 tumors. Another hypoxic cytotoxin, NLCQ-1, was also more efficacious 2 weeks after RAIT, compared with same-day dosing. Thus, information on tumor hypoxia after radioantibody therapy could be important for ascertaining a window of opportunity when the surviving tumor regions are most responsive to hypoxic cytotoxins.

Hypokalemia induces renal injury and alterations in vasoactive mediators that favor salt sensitivity

We investigated the hypothesis that hypokalemia might induce renal injury via a mechanism that involves subtle renal injury and alterations in local vasoactive mediators that would favor sodium retention. To test this hypothesis, we conducted studies in rats with diet-induced K+ deficiency. We also determined whether rats with hypokalemic nephropathy show salt sensitivity. Twelve weeks of hypokalemia resulted in a decrease in creatinine clearance, tubulointerstitial injury with macrophage infiltration, interstitial collagen type III deposition, and an increase in osteopontin expression (a tubular marker of injury). The renal injury was greatest in the outer medulla with radiation into the cortex, suggestive of an ischemic etiology. Consistent with this hypothesis, we found an increased uptake of a hypoxia marker, pimonidazole, in the cortex. The intrarenal injury was associated with increased cortical angiontensin-converting enzyme (ACE) expression and continued cortical angiotensin II generation despite systemic suppression of the renin-angiotensin system, an increase in renal endothelin-1, a decrease in renal kallikrein, and a decrease in urinary nitrite/nitrates and prostaglandin E(2) excretion. At 12 wk, hypokalemic rats were placed on a normal-K+ diet with either high (4%)- or low (0.01%)-NaCl content. Despite correction of hypokalemia and normalization of renal function, previously hypokalemic rats showed an elevated blood pressure in response to a high-salt diet compared with normokalemic controls. Hypokalemia is associated with alterations in vasoactive mediators that favor intrarenal vasoconstriction and an ischemic pattern of renal injury. These alterations may predispose the animals to salt-sensitive hypertension that manifests despite normalization of the serum K+.

Activated Kupffer cells cause a hypermetabolic state after gentle in situ manipulation of liver in rats

Harvesting trauma to the graft dramatically decreases survival after liver transplantation. Since activated Kupffer cells play a role in primary nonfunction, the purpose of this study was to test the hypothesis that organ manipulation activates Kupffer cells. To mimic what occurs with donor hepatectomy, livers from Sprague-Dawley rats underwent dissection with or without gentle organ manipulation in a standardized manner in situ. Perfused livers exhibited normal values for O(2) uptake (105 +/- 5 micromol. g(-1). h(-1)) measured polarigraphically; however, 2 h after organ manipulation, values increased significantly to 160 +/- 8 micromol. g(-1). h(-1) and binding of pimonidazole, a hypoxia marker, increased about threefold (P < 0.05). Moreover, Kupffer cells from manipulated livers produced three- to fourfold more tumor necrosis factor-alpha and PGE(2), whereas intracellular calcium concentration increased twofold after lipopolysaccharide compared with unmanipulated controls (P < 0.05). Gadolinium chloride and glycine prevented both activation of Kupffer cells and effects of organ manipulation. Furthermore, indomethacin given 1 h before manipulation prevented the hypermetabolic state, hypoxia, depletion of glycogen, and release of PGE(2) from Kupffer cells. These data indicate that gentle organ manipulation during surgery activates Kupffer cells, leading to metabolic changes dependent on PGE(2) from Kupffer cells, which most likely impairs liver function. Thus modulation of Kupffer cell function before organ harvest could be beneficial in human liver transplantation and surgery.

Autonomic nervous system and gut-derived endotoxin: involvement in activation of Kupffer cells after in situ organ manipulation

Gentle in situ organ manipulation rapidly causes disturbances in the hepatic microcirculation, hypoxia, and activation of Kupffer cells. Because the mechanisms of Kupffer cell activation after organ manipulation remain unclear, the possible role of the autonomic nervous system and gut-derived endotoxin were assessed. To mimic what occurs with major abdominal surgery, livers from female Sprague-Dawley rats (200-230 g) underwent minimal dissection for 12 minutes and were manipulated gently or were left alone for 13 subsequent minutes. Kupffer cells were activated 2 hours after manipulation, reflected by a significant increase in intracellular calcium ([Ca2+]i) from about 90 nM in unmanipulated controls to more than 180 nM in response to lipopolysaccharide (LPS 100 ng/ml). Furthermore, Kupffer cells from manipulated rats produced about threefold more tumor necrosis factor-alpha after LPS (100 ng/ml) than did the unmanipulated controls. Moreover, O2 uptake of ex situ perfused liver was increased from about 110 micromol/g/hr in unmanipulated controls to more than 160 micromol/g/hr 2 hours after organ manipulation. Binding of pimonidazole (120 mg/kg IV), a 2-nitroimidazole hypoxia marker given 2 hours after manipulation, increased about 2.5-fold, and hepatic glycogen was depleted. Two hours after organ manipulation gut permeability to horseradish peroxidase was elevated and endotoxin in the portal venous blood was increased twofold. Microsurgical hepatic denervation, ganglionic blockade, adrenalectomy, and antibiotics to sterilize the gut before manipulation prevented activation of Kupffer cells by organ manipulation. Hexamethonium and adrenalectomy prevented increases in gut permeability caused by manipulation. Although antibiotics blunted the increase in portal venous endotoxin significantly, there was no effect on gut permeability. These data indicate for the first time that both the autonomic nervous system and gut-derived endotoxin are involved in activation of Kupffer cells after organ manipulation.

Determination of hypoxic region by hypoxia marker in developing mouse embryos in vivo: a possible signal for vessel development

Hypoxia is a well-known signal for angiogenesis, but the recent proposal that hypoxia exists in developing embryonic tissues and that it induces vascular development remains to be proven. In the present study, we demonstrate the presence of hypoxia in normal developing embryos by means of a hypoxia marker, pimonidazole, and its associated antibody. Our data clearly show that hypoxia marker immunoreactivity was highly detected in developing neural tubes, heart, and intersomitic mesenchyme at an early stage of organogenesis, suggesting that hypoxia may exist in the early stages of embryo development. We also found that hypoxia inducible factor-1alpha (HIF-1alpha) and vascular endothelial growth factor (VEGF) were spatiotemporally co-localized with possible hypoxic regions in embryos. Investigation of platelet endothelial cell adhesion molecule (PECAM) expression provides evidence that endothelial cells proliferate and form the vessels in the hypoxic region in developing organs. Furthermore, we found that hypoxia induced both HIF-1alpha and VEGF in F9 embryonic stem and differentiated cells. Thus, we suggest that hypoxia may exist widely in developing embryonic tissues and that it may act as a signal for embryonic blood vessel formation in vivo.

Semiquantitative immunohistochemical analysis for hypoxia in human tumors

OBJECTIVE

The goal of this study was to develop a semiquantitative scoring system for measuring hypoxia in human tumors by an immunohistochemical marker approach.

METHODS AND MATERIALS

Eighteen patients diagnosed with squamous cell carcinoma of the uterine cervix or head and neck were infused intravenously with a solution of pimonidazole hydrochloride at a dose of 0.5 gm/m2. Twenty-four hours later, four biopsies on average from each tumor were fixed in formalin, processed into paraffin blocks, and sectioned. Tissue sections were immunostained for the presence of pimonidazole adducts. Microscopic images (x200) of immunostaining were captured and quantitated by standard image analysis. Images with known amounts of hypoxia spanning ranges of > 0% to 5%, > 5% to 15%, > 15% to 30%, and >30% were assigned scores of +1, +2, +3, and +4, respectively. Three observers then used this calibrated scoring system to analyze hypoxia in tumor sections in a blinded fashion.

RESULTS

Excellent interobserver reproducibility was obtained with the calibrated, semiquantitative, immunohistochemical assay for hypoxia in squamous cell carcinomas.

CONCLUSION

The calibrated, semiquantitative assay shows promise as an approach to simplifying the quantitation of human tumor hypoxia by immunohistochemical techniques.

Invasive oxygen measurements and pimonidazole labeling in human cervix carcinoma

PURPOSE

This study was designed to compare tumor hypoxia assessed by invasive O2 sensitive electrodes and pimonidazole labeling in primary human cervix carcinomas.

METHODS AND MATERIALS

Twenty-eight patients with primary cervix carcinomas (FIGO Stage Ib-IVa) were investigated. Both invasive pO2 measurements and pimonidazole labeling were obtained in all patients. Before treatment, patients were given pimonidazole as a single injection (0.5 g/m2 i.v.). Ten to 24 h later, oxygenation measurements were done by Eppendorf histography, and after this procedure biopsies were taken for pimonidazole-binding analysis. Tumor oxygen partial pressure (pO2) was evaluated as the median tumor pO2 and the fraction of pO2 values < or = 10 mmHg (HF10). Biopsies were formalin fixed and paraffin embedded, and hypoxia was detected by immunohistochemistry using monoclonal antibodies directed against reductively activated pimonidazole. Pimonidazole binding was evaluated by a semiquantitative scoring system.

RESULTS

Both Eppendorf measurements and pimonidazole binding showed large intra-and intertumor variability. A comparison between pimonidazole binding expressed as the fraction of fields at the highest score and HF10 showed a trend for the most well-oxygenated tumors having a low fraction of fields; however, the correlation did not reach statistical significance (p = 0.43, r = 0.165; Spearman's rank correlation test).

CONCLUSION

Hypoxia measured in human uterine cervix carcinomas is heterogeneously expressed both within and between tumors when assessed by either invasive pO2 measurements or pimonidazole binding. Despite a trend that tumors with high pO2 values expressed less pimonidazole binding, no correlation was seen between the two assays in this preliminary report.

Comparison between the comet assay and pimonidazole binding for measuring tumour hypoxia

Pimonidazole is finding increasing use in histochemical analyses of hypoxia in tumours. Whether it can identify every hypoxic cell in a tumour, and whether the usual subjective criteria used to define 'positive' cells are optimal, are less certain. Therefore, our aim was to develop an objective flow cytometry procedure for quantifying pimonidazole binding in tumours, and to validate this method by using a more direct indicator of radiobiologic hypoxia, the comet assay. SCCVII tumours in C3H mice were analysed for pimonidazole binding using flow cytometry and an iterative curve-fitting procedure, and the results were compared to the comet assay for the same cell suspensions. On average, cells defined as anoxic by flow analysis (n = 43 tumours) bound 10.8 +/- 0.95 times more antibody than aerobic cells. In samples containing known mixtures of aerobic and anoxic cells, hypoxic fractions as low as 0.5% could easily be detected. To assess the flow cytometry assay under a wider range of tumour oxygen contents, mice were injected with hydralazine to reduce tumour blood flow, or allowed to breathe various gas mixtures during the 90 min exposure to pimonidazole. Hypoxic fraction estimated by the pimonidazole binding method agreed well with the hypoxic fraction measured using the comet assay in SCCVII tumours (r2 = 0.87, slope = 0.98), with similar results in human U87 glioma cells and SiHa cervical carcinoma xenografts. We therefore conclude that this objective analysis of pimonidazole labelling by flow cytometry gives a convenient and accurate estimate of radiobiological hypoxia. Preliminary analyses of biopsies from 3 patients given 0.5 g m-2 pimonidazole also suggest the suitability of this approach for human tumours.

Changes in tumor hypoxia measured with a double hypoxic marker technique

PURPOSE

Development of a double hypoxic cell marker assay, using the bioreductive nitroimidazole derivatives CCI-103F and pimonidazole, to study changes in tumor hypoxia after treatments that modify tumor oxygenation.

METHODS AND MATERIALS

Both hypoxic markers were visualized by immunohistochemical techniques to detect changes in hypoxic fraction induced by carbogen breathing (95% O(2) and 5% CO(2)) or hydralazine injection. The protocol was tested in a human laryngeal squamous cell carcinoma xenograft line. Quantitative measurements were derived from consecutive tissue sections that were analyzed by a semiautomatic image analysis system. Qualitative analysis was obtained by double staining of the two hypoxic markers on the same tissue section.

RESULTS

A significant correlation between the hypoxic fractions for the two markers, CCI-103F and pimonidazole, was found in air breathing animals. After carbogen breathing, the hypoxic fraction decreased significantly from 0.07 to 0.03, and after hydralazine treatment, the hypoxic fraction increased significantly. Reduction of hypoxia after carbogen breathing was most pronounced close to well-perfused tumor regions.

CONCLUSIONS

With this method, employing two consecutively injected bioreductive markers, changes in tumor hypoxia can be studied. A significant reduction in hypoxia after carbogen breathing and a significant increase in hypoxia after hydralazine administration was demonstrated.

Vascular architecture and hypoxic profiles in human head and neck squamous cell carcinomas

Tumour oxygenation and vasculature are determinants for radiation treatment outcome and prognosis in patients with squamous cell carcinomas of the head and neck. In this study we visualized and quantified these factors which may provide a predictive tool for new treatments. Twenty-one patients with stage III-IV squamous cell carcinomas of the head and neck were intravenously injected with pimonidazole, a bioreductive hypoxic marker. Tumour biopsies were taken 2 h later. Frozen tissue sections were stained for vessels and hypoxia by fluorescent immunohistochemistry. Twenty-two sections of biopsies of different head and neck sites were scanned and analysed with a computerized image analysis system. The hypoxic fractions varied from 0.02 to 0.29 and were independent from T- and N-classification, localization and differentiation grade. No significant correlation between hypoxic fraction and vascular density was observed. As a first attempt to categorize tumours based on their hypoxic profile, three different hypoxia patterns are described. The first category comprised tumours with large hypoxic, but viable, areas at distances even greater than 200 micrometer from the vessels. The second category showed a typical band-like distribution of hypoxia at an intermediate distance (50-200 micrometer) from the vessels with necrosis at greater distances. The third category demonstrated hypoxia already within 50 micrometer from the vessels, suggestive for acute hypoxia. This method of multiparameter analysis proved to be clinically feasible. The information on architectural patterns and the differences that exist between tumours can improve our understanding of the tumour micro-environment and may in the future be of assistance with the selection of (oxygenation modifying) treatment strategies.

Hypoxia in a human intracerebral glioma model

OBJECT

The development of hypoxia in human gliomas is closely related to functional vasculature and the presence of hypoxia has important biological and therapeutic consequences. Assessment of hypoxia is necessary to understand its role in treatment response and to evaluate treatment strategies to improve tumor oxygenation. In this study, the authors report findings of their analysis of the degree of hypoxia in relation to other vascular parameters in a human intracerebral glioma xenograft.

METHODS

In sections of tumor, hypoxic regions were identified immunohistochemically by using the hypoxic marker pimonidazole. The S-phase marker bromodeoxyuridine was used to detect cell proliferation, and the perfusion marker Hoechst 33342 was used to delineate perfused vessels. Vascular structures were stained with an endothelial marker. Hypoxic tumor regions were clearly present in this human intracerebral glioma model. Hypoxic areas were usually found in nonperfused regions, whereas tumor cell proliferation was especially marked in perfused tumor areas. Furthermore, by using in situ hybridization the authors identified infiltrating tumor cells in the normal brain. This feature is often observed in gliomas in patients.

CONCLUSIONS

This model is a representative human glioma model that provides the researcher with the opportunity to analyze the relationship between the degree of hypoxia and vascular parameters, as well as to examine the effects of treatments aimed at modification of the oxygenation status of a tumor.

Spatial relationship between hypoxia and the (perfused) vascular network in a human glioma xenograft: a quantitative multi-parameter analysis

PURPOSE

To quantitatively study the spatial distribution of tumor hypoxia in relation to the perfused vasculature.

METHODS AND MATERIALS

Using a human glioma xenograft model, nude mice were administered two different hypoxia markers (NITP or pimonidazole) and the perfusion marker Hoechst 33342. Frozen tumor sections were sequentially scanned for perfusion, hypoxia, and vasculature, respectively, to quantitate perfusion, vasculature, and hypoxia parameters in the same section.

RESULTS

All tumors showed incomplete perfusion. Both NITP and pimonidazole stained the same hypoxic tumor areas. No statistically significant differences between the two markers were observed. The density of the perfused vessels was inversely related to the hypoxic fraction. At critical distances from perfused vessels, hypoxia occurred. These data suggest that predominantly diffusion-limited hypoxia was detected, based on the spatial distribution of nearby vessels. Also, the proportion of hypoxia distributed over arbitrary zones of 50 microm around perfused vessels was calculated. The largest proportion of hypoxia was found at distances beyond 100 microm from perfused vessels.

CONCLUSION

With the multiple staining and functional microscopic imaging technique described here, the spatial relationship between perfused vessels and hypoxia was quantified in whole tumor cross-sections. The usefulness of this histologically-based method to quantitate morphological and physiological aspects of the tumor microenvironment was evaluated.

Changes in blood perfusion and hypoxia after irradiation of a human squamous cell carcinoma xenograft tumor line

The effect of irradiation depends on the oxygenation status of the tissue, while irradiation itself also changes the oxygenation and perfusion status of tissues. A better understanding of the changes in tumor oxygenation and perfusion over time after irradiation will allow a better planning of fractionated radiotherapy in combination with modifiers of blood flow and oxygenation. Vascular architecture (endothelial marker), perfusion (Hoechst 33342) and oxygenation (pimonidazole) were studied in a human laryngeal squamous cell carcinoma tumor line grown as xenografts in nude mice. The effect of a single dose of 10 Gy X rays on these parameters was evaluated from 2 h to 11 days after irradiation. Shortly after irradiation, there was an 8% increase in perfused blood vessels (from 57% to 65%) followed by a significant decrease, with a minimum value of 42% at 26 h after irradiation, and a subsequent increase to control levels at 7 to 11 days after irradiation. The hypoxic fraction showed a decrease at 7 h after treatment from 13% to 5% with an increase to 19% at 11 days after irradiation. These experiments show that irradiation causes rapid changes in oxygenation and perfusion which may have consequences for the optimal timing of radiotherapy schedules employing multiple fractions per day and the introduction of oxygenation- and perfusion-modifying drugs.

A clinical study of hypoxia and metallothionein protein expression in squamous cell carcinomas

The objective was to discover whether the oxygen-regulated protein, metallothionein, is expressed in the hypoxic cells of squamous cell carcinomas. Twenty patients with squamous cell carcinoma of the uterine cervix or head and neck were infused with a solution of the hypoxia marker, pimonidazole hydrochloride, at a dose of 0.5 g/m2. The following day, biopsies were collected, formalin fixed, paraffin embedded, and sectioned at 4 microm. Sections from each biopsy were immunostained for pimonidazole binding, metallothioneins I and II, involucrin, and proliferating cell nuclear antigen. A total of 84 biopsies were analyzed. Sixty-four of 84 biopsy sections contained hypoxia. Of the hypoxia-containing sections, 43 of 64 or 67% showed no microregional overlap between hypoxia and metallothionein; 7 of 64 showed overlap; and 14 of 64 showed a combination of overlap and no overlap. On a tumor-by-tumor basis, 5 of 7 head and neck and 7 of 13 cervix tumors showed no overlap between metallothionein and hypoxia at the microregional level. Ranges for the percentage of the area of hypoxia in head and neck (<0.9 to 17%) and cervix (<0.1 to 14%) tumors were similar. In the hypoxia-containing sections, immunostaining for involucrin, a molecular marker for differentiation, overlapped with that for hypoxia in 82% of the cases. The majority of hypoxic cells in squamous cell carcinomas do not express metallothionein protein, although metallothionein is induced by hypoxia in human tumor cells in vitro. Hypoxic cells in human tumors tend to be in regions immunostaining for involucrin, and it seems possible that differentiation of hypoxic cells in squamous cell carcinomas might affect metallothionein I and II expression.

Early wound healing exhibits cytokine surge without evidence of hypoxia

OBJECTIVE

To ascertain the spatial and temporal relation of wound hypoxia to the cell types involved, expression of selected angiogenic cytokines, the proliferative status of cells in the wound site, and angiogenesis.

SUMMARY BACKGROUND DATA

Hypoxia is considered to drive the angiogenic response by upregulating angiogenic cytokines observed during wound healing. But this correlation has not been shown on a cell-to-cell basis in vivo because of limitations in measuring tissue PO2 at the cellular level.

METHODS

Using punch biopsy wounds in rats as a wound healing model, the distributions of vascular endothelial growth factor, transforming growth factor-beta, tumor necrosis factor-alpha, and pimonidazole adducts (as a hypoxia marker) were followed immunohistochemically during the healing process.

RESULTS

Hypoxia was absent on day 1 after wounding, even though angiogenesis and maximal expression of cytokines were observed in the wounds. Hypoxia peaked in the granulation tissue stage at day 4 and correlated with increased cellularity and cellular proliferation. Hypoxia started to decrease after day 4 and was limited to the remnant blood vessels and epithelial layer in the scar tissue.

CONCLUSIONS

Induction of angiogenic cytokines early during wound healing may be due to triggering mechanisms other than hypoxia. Alternatively, the unique pattern of development and decline of cellular hypoxia as wound cellularity and proliferation regress suggest its involvement in initiating vascular regression during the later stages of healing.

Gentle organ manipulation during harvest as a key determinant of survival of fatty livers after transplantation in the rat

Both in situ organ manipulation during harvest and steatosis reduce survival after liver transplantation via mechanisms involving Kupffer cells; thus, their effect on survival was compared here. Moderate steatosis was induced by a single dose of ethanol to Lewis rats, while long-term administration of ethanol yielded severe steatosis in donor animals. After minimal dissection during the first 12 min, livers were either manipulated gently or left alone for 13 min subsequently. Orthotopic liver transplantation was performed after 1 h of cold storage in UW solution. Ethanol increased hepatic lipid content to a level of moderate or severe steatosis that reduced survival after transplantation from 100% to approximately 70% (P < 0.05). However, gentle manipulation decreased survival to approximately 30% (P < 0.05) in livers from normal, saline-treated rats and in livers from rats fed a high-fat control diet. Moreover, after short- or long-term ethanol administration, manipulation of fatty livers decreased survival from 70% to approximately 13% (P < 0.05). Further, manipulation elevated serum transaminases, total bilirubin, and necrosis significantly about 2- to 20-fold in fatty grafts after transplantation. At the end of harvest, trypan blue distribution time and hypoxia assessed from 2-nitroimidazole binding were elevated significantly about two- to fourfold by manipulation of fatty grafts. Gadolinium chloride, a Kupffer cell toxicant, blocked the detrimental effects of manipulation. These data demonstrate for the first time that, while steatosis is detrimental for survival, organ manipulation plays a much greater role than fat in mechanisms of primary nonfunction.

Osteocyte hypoxia: a novel mechanotransduction pathway

Bone is a unique tissue in which to examine mechanotransduction due to its essential role in weight bearing. Within bone, the osteocyte is an ideal cellular mechanotransducer candidate. Because osteocytes reside distant from the blood supply, their metabolic needs are met by a combination of passive diffusion and enhanced diffusion, arising when the tissue is loaded during functional activity. Therefore, we hypothesized that depriving a bone of mechanical loading (and thus eliminating diffusion enhanced by loading) would rapidly induce osteocyte hypoxia. Using the avian ulna model of disuse osteopenia, we found that 24 h of unloading results in significant osteocyte hypoxia (8.4 +/- 1.8%) compared with control levels (1.1 +/- 0.5%; P = 0.03). Additionally, we present preliminary data suggesting that a brief loading regimen is sufficient to rescue osteocytes from this fate. The rapid onset of the observed osteocyte hypoxia, the inhibition of hypoxia by brief loading, and the cellular consequences of oxygen deprivation are suggestive of a novel mechanotransduction pathway with implications across organ systems.

Binge drinking disturbs hepatic microcirculation after transplantation: prevention with free radical scavengers

Disturbances in hepatic microcirculation increase graft injury and failure; therefore, this study evaluates the effects of ethanol on microcirculation after liver transplantation. Donor rats were given one dose of ethanol (5 g/kg) by gavage 20 h before explantation, and grafts were stored in University of Wisconsin solution for 24 h before implantation. Acute ethanol treatment decreased 7-day survival of grafts from about 90 to 30%, increased transaminase release nearly 4-fold, and decreased bile production by 60%. Moreover, portal pressure increased significantly and liver surface oxygen tension decreased about 50%, indicating that ethanol disturbs hepatic microcirculation. Pimonidazole, a 2-nitroimidazole hypoxia marker, was given i.v. to recipients 30 min after implantation, and grafts were harvested 1 h later. Ethanol increased hepatic pimonidazole binding about 3-fold, indicating that ethanol led to hypoxia in fatty grafts. Ethanol also significantly increased free radicals in bile. Catechin (30 mg/kg i.v. upon reperfusion), a free radical scavenger, and Carolina Rinse solution, which contains several agents that inhibit free radical formation, minimized disturbances in microcirculation and prevented pimonidazole adduct formation significantly. These treatments also blunted increases in transaminase release and improved survival of fatty grafts. Destruction of Kupffer cells with GdCl(3) (20 mg/kg i.v. 24 h before explantation) or inhibition of formation of leukotrienes with MK-886 (50 microM in University of Wisconsin or rinse solution) also minimized hypoxia and improved survival after transplantation. Taken together, these results demonstrate that ethanol disturbs hepatic microcirculation, leading to graft hypoxia after transplantation, most likely by activating Kupffer cells and increasing free radical production.

Mechanisms of alcohol-induced hepatotoxicity: studies in rats

Alcohol treatment results in increases in the release of endotoxin from gut bacteria and membrane permeability of the gut to endotoxin, or both. Females are more sensitive to these changes. Elevated levels of endotoxin activate Kupffer cells to release substances such as eicosanoids, TNF-alpha and free radicals. Prostaglandins increase oxygen uptake and most likely are responsible for the hypermetabolic state in the liver. The increase in oxygen demand leads to hypoxia in the liver, and on reperfusion, alpha-hydroxyethyl free radicals are formed which lead to tissue damage in oxygen-poor pericentral regions of the liver lobule.

Reperfusion injury in livers due to gentle in situ organ manipulation during harvest involves hypoxia and free radicals

Kupffer cell-dependent injury in livers gently manipulated during harvest develops upon transplantation; however, underlying mechanisms remain unknown. Thus, the purpose of this study was to identify factors involved in mechanisms of injury. Livers from female Sprague-Dawley rats (200-230 g) were cold stored for 24 h in University of Wisconsin solution. Subsequently, livers were perfused at 37 degrees C with oxygen-saturated Krebs-Henseleit buffer containing fluorescein-dextran to assess microcirculation. Cell death was assessed by uptake of trypan blue, a vital dye. Minimal dissection during harvest had no effects on sinusoidal lining cells; however, gentle organ manipulation dramatically increased trypan blue uptake about 5-fold (p <.05). In contrast, perfusion with N2-saturated buffer after cold storage totally prevented cell death due to manipulation. At harvest, portal venous pressure was increased significantly by 70% due to manipulation. Furthermore, vascular space and microcirculation were decreased by more than 50% (p <.05), reflecting the rate of entry and exit of fluorescein-dextran. Pimonidazole, a 2-nitroimidazole marker, was given to rats before harvest to detect hypoxia in liver. Pimonidazole adduct binding was increased significantly about 2-fold by manipulation. To detect free radical adducts by electron spin resonance (ESR) spectroscopy in bile, C-phenyl-N-tert-butylnitrone was given as spin trapping reagent to the donor before operation. Free radical formation was increased about 3-fold by organ manipulation (p <.05). Donors given gadolinium chloride, a selective Kupffer cell toxicant, or dietary glycine, which prevents activation of Kupffer cells, significantly blunted microcirculatory disturbances, hypoxia, and death of endothelial lining cells. These data indicate for the first time that gentle organ manipulation during harvest causes oxygen-dependent reperfusion injury to endothelial lining cells via mechanisms involving hepatic microcirculation, hypoxia, and Kupffer cells.

Graft survival is improved by hepatic denervation before organ harvesting

BACKGROUND

In a recent study, disturbances of hepatic microcirculation at harvesting caused by in situ organ manipulation dramatically reduced survival after a liver transplant. Because hepatic innervation is involved in the regulation of liver hemodynamics, the effect of denervation before harvesting was assessed here.

METHODS

The livers were harvested from female Lewis rats (200-230 g) within 25 min. Briefly, after minimal dissection during the first 12 min, the livers were either manipulated gently or left alone for 13 min. Subsequently, an orthotopic liver transplant was performed after 1 hr of storage in cold UW solution. Some donors livers underwent microsurgical denervation before harvesting or rats were given hexamethonium (10 mg/kg, i.v.), a ganglionic blocking agent.

RESULTS

In the nonmanipulated group, survival was 100% after the transplant; however, gentle manipulation decreased survival by about 50%. Furthermore, manipulation elevated serum transaminases and bilirubin 6- to 8-fold 8 hr after the transplant and caused necrosis of about 25% of hepatocytes. After organ harvesting, the rate of entry and exit of fluorescein dextran, a dye confined to the vascular space, was decreased 2- to 4-fold, and the maximal increase of surface fluorescence was blunted about 2-fold. Pimonidazole binding, which reflects tissue hypoxia, was increased 2-fold by manipulation. Denervation of the liver before organ harvesting or treatment with hexamethonium prevented the effects of organ manipulation on all parameters studied.

CONCLUSION

These data indicate for the first time that hepatic denervation before organ harvesting prevents detrimental effects of brief, gentle manipulation of the liver during harvesting on survival after the transplant. This is consistent with the hypothesis that organ manipulation disturbs the hepatic microcirculation and causes hypoxia at harvesting using mechanisms dependent on innervation.

Comparisons among pimonidazole binding, oxygen electrode measurements, and radiation response in C3H mouse tumors

Pimonidazole binding was compared with oxygen electrode measurements and with measurements of the radiobiologically hypoxic fraction in C3H mammary tumors in which oxygenation was manipulated by means of subjecting tumor-bearing CDF1 mice to air breathing, carbogen breathing, oxygen breathing, hydralazine injection or tumor clamping. Hypoxia measured by pimonidazole binding could be correlated with both pO2 (r2 = 0.81) and radiobiologically hypoxic fraction (r2 = 0.85) in this system. The scope and limitation of pimonidazole as an immunohistochemical marker for tumor hypoxia is discussed.

Intravenous glycine improves survival in rat liver transplantation

In situ manipulation by touching, retracting, and moving liver lobes gently during harvest dramatically reduces survival after transplantation (P. Schemmer, R. Schoonhoven, J. A. Swenberg, H. Bunzendahl, and R. G. Thurman. Transplantation 65: 1015-1020, 1998). The development of harvest-dependent graft injury upon reperfusion can be prevented with GdCl3, a rare earth metal and Kupffer cell toxicant, but it cannot be used in clinical liver transplantation because of its potential toxicity. Thus the effect of glycine, which prevents activation of Kupffer cells, was assessed here. Minimal dissection of the liver for 12 min plus 13 min without manipulation had no effect on survival (100%). However, gentle manipulation decreased survival to 46% in the control group. Furthermore, serum transaminases and liver necrosis were elevated 4- to 12-fold 8 h after transplantation. After organ harvest, the rate of entry and exit of fluorescein dextran, a dye confined to the vascular space, was decreased about twofold, indicating disturbances in the hepatic microcirculation. Pimonidazole binding, which detects hypoxia, increased about twofold after organ manipulation, and Kupffer cells isolated from manipulated livers produced threefold more tumor necrosis factor-alpha after lipopolysaccharide than controls. Glycine given intravenously to the donor increased the serum glycine concentration about sevenfold and largely prevented the effect of gentle organ manipulation on all parameters studied. These data indicate for the first time that pretreatment of donors with intravenous glycine minimizes reperfusion injury due to organ manipulation during harvest and after liver transplantation.

Oxidative stress occurs in perfused rat liver at low oxygen tension by mechanisms involving peroxynitrite

Ethanol increases free radical formation; however, it was recently demonstrated that it also causes extensive hypoxia in rat liver in vivo. To address this issue, it was hypothesized that peroxynitrite formed in normoxic periportal regions of the liver lobule has its reactivity enhanced in hypoxic pericentral regions where the pH is lower. Via this pathway, peroxynitrite could lead to free radical formation in the absence of oxygen. Livers from fed rats were perfused at low flow rates for 75 min. Under these conditions, periportal regions were well oxygenated but pericentral areas became hypoxic. Low-flow perfusion caused a significant 6-fold increase in nitrotyrosine accumulation in pericentral regions. During the last 20 min of perfusion, the spin-trap alpha-(4-pyridyl-1-oxide)-N-tert-butylnitrone was infused and adducts were collected for electron-spin resonance analysis. A six-line radical adduct signal was detected in perfusate. Direct infusion of peroxynitrite produced a radical adduct with identical coupling constants, and a similar pattern of nitrotyrosine accumulation was observed. Retrograde perfusion at low rates resulted in accumulation of nitrotyrosine in periportal regions. Although the magnitude of the radical in perfusate was increased by ethanol, it was not derived directly from it. Both nitrotyrosine accumulation and radical formation were reduced by inhibition of nitric oxide synthase with N-nitro-L-arginine methyl ester, but not with the inactive D-isomer. Radical formation was decreased nearly completely by superoxide dismutase and N-nitro-L-arginine methyl ester, consistent with the hypothesis that the final prooxidant is a derivative from both NO. and superoxide (i.e., peroxynitrite). These results support the hypothesis that oxidative stress occurs in hypoxic regions of the liver lobule by mechanisms involving peroxynitrite.

Relationship of hypoxia to metallothionein expression in murine tumors

PURPOSE

To investigate if metallothionein, an endogenous chemo- and radioprotectant, is expressed in hypoxic cells in mouse C3H mammary carcinomas and if that expression responds to acute changes in tumor hypoxia.

METHODS AND MATERIALS

C3H mammary tumors were established in the hind legs of female CDF1 mice. The mice were then subjected to air breathing (chronic hypoxia), carbogen breathing (acute decrease in hypoxia), or hydralazine injection (acute increase in hypoxia). Ninety minutes after the start of the experiment, tumors were excised, fixed in formalin, and sectioned. Hypoxic cells and metallothionein-containing cells were quantitated by image analysis. Pimonidazole hydrochloride and an IgG1 mouse monoclonal antibody were used to detect hypoxia, and a mouse antimetallothionein monoclonal antibody (DAKO) was used to detect Type I and II metallothionein in sets of contiguous tissue sections.

RESULTS

The distribution of immunostaining intensity for metallothionein was the same in all three groups-heavy in hypoxic cells and light in other regions of the tumors. The acute increase in hypoxia caused by hydralazine injection was accompanied by an increase in metallothionein expression (p = 0.04). Carbogen breathing largely eliminated pimonidazole binding, but metallothionein expression persisted in the tumors of carbogen-breathing mice.

CONCLUSIONS

Hypoxic cells in C3H mammary carcinomas strongly express metallothionein. Metallothionein expression is responsive to acute increases in hypoxia brought about by hydralazine injection. The effectiveness of hydralazine in enhancing the activation of bioreductive cytotoxins might be offset by the increased expression of metallothionein. The persistence of metallothionein in tumors of carbogen-breathing mice might contribute to a residual radioresistance in the tumors.

Pimonidazole: a novel hypoxia marker for complementary study of tumor hypoxia and cell proliferation in cervical carcinoma

BACKGROUND

Tumor hypoxia may be associated with treatment resistance, cell proliferation, and metastatic potential, which contribute to poor prognosis. Complementary techniques for detecting hypoxia, cell growth, and metastases are required to study these relationships.

OBJECTIVES

The purpose of this study was to demonstrate the clinical feasibility of quantitative hypoxia detection with pimonidazole, a novel hypoxia marker, and to correlate hypoxia with S-phase markers of tumor proliferation.

METHODS

Pimonidazole binds to thiol-containing proteins specifically in hypoxic cells. Ten patients with cervical carcinoma received 0.5 g/m2 pimonidazole intravenously followed by biopsy of the cervical carcinoma the next day. Hypoxic cells were recognized by immunohistochemical detection of pimonidazole using a mouse monoclonal antibody. Cell proliferation was detected with a commercially available monoclonal antibody for proliferating cell nuclear antigen (PCNA). Assessment of hypoxia and cell proliferation was made qualitatively with light microscopy and quantitatively using point counting and image analysis software methods.

RESULTS

No clinical toxic effects were associated with pimonidazole administration. Immunostaining with pimonidazole antibody was observed in 9 of 10 tumors, suggesting that hypoxia is a common occurrence in cervical carcinoma. Quantitatively, tumors that had large numbers of hypoxic cells had the greatest percentage of S-phase cells, but some tumors with smaller amounts of hypoxia also had substantial numbers of S-phase cells.

CONCLUSION

Pimonidazole can be used for qualitative and quantitative assessment of tumor hypoxia.

Cyclosporin A increases hypoxia and free radical production in rat kidneys: prevention by dietary glycine

The major side effect of cyclosporin A is severe nephrotoxicity. It is likely that cyclosporin A causes vasoconstriction leading to hypoxia-reperfusion injury; therefore, these experiments were designed to attempt to obtain physical evidence for hypoxia and free radical production in kidney following cyclosporin A. Rats were treated daily with cyclosporin A (25 mg/kg ig) for 5 days, and pimonidazole, a hypoxia marker, was injected 2 h after the last dose of cyclosporin A. A dose of alpha-(4-pyridyl-1-oxide)-N-tert-butylnitrone (4-POBN) was injected 3 h after cyclosporin A to trap free radicals. Cyclosporin A doubled serum creatinine and decreased glomerular filtration rates by 65% as expected. Pimonidazole adduct binding in the kidney was increased nearly threefold by cyclosporin A, providing physical evidence for tissue hypoxia. Moreover, cyclosporin A increased 4-POBN/radical adducts nearly sixfold in the urine but did not alter levels in the serum. Glycine, which causes vasodilatation and prevents cyclosporin A toxicity, minimized hypoxia and blocked free radical production; however, it did not alter cyclosporin A blood levels. These results demonstrate for the first time that cyclosporin A causes hypoxia and increases production of a new free radical species exclusively in the kidney. Therefore, it is concluded that cyclosporin A causes renal injury by mechanisms involving hypoxia-reoxygenation, effects which can be prevented effectively by dietary glycine.

Hypoxia and vascular endothelial growth factor expression in human squamous cell carcinomas using pimonidazole as a hypoxia marker

Hypoxia in human tumors is associated with poor prognosis, but the molecular mechanisms underlying this association are poorly understood. One possibility is that hypoxia is linked to malignant progression through vascular endothelial growth factor (VEGF) induction and the associated angiogenesis and metastasis. The present clinical study measures hypoxia and VEGF expression on a cell-by-cell basis in human squamous cell carcinomas to test the hypothesis that hypoxia and VEGF protein expression are coupled in human tumors. Eighteen patients with invasive squamous cell carcinoma of the uterine cervix and head and neck have been investigated by a quantitative image analysis of immunostained sections from their tumors. The hypoxia marker pimonidazole was used to measure tumor hypoxia, and a commercially available antibody was used to measure VEGF protein expression. A quantitative immunohistochemical comparison of hypoxia and VEGF protein expression revealed no correlation between the two factors.

The role of gut-derived bacterial toxins and free radicals in alcohol-induced liver injury

Previous research from this laboratory using a continuous enteral ethanol (EtOH) administration model demonstrated that Kupffer cells are pivotal in the development of EtOH-induced liver injury. When Kupffer cells were destroyed using gadolinium chloride (GdCl3) or the gut was sterilized with polymyxin B and neomycin, early inflammation due to EtOH was blocked. Anti-tumour necrosis factor (TNF)-alpha antibody markedly decreased EtOH-induced liver injury and increased TNF-mRNA. These findings led to the hypothesis that EtOH-induced liver injury involves increases in circulating endotoxin leading to activation of Kupffer cells. Pimonidazole, a nitro-imidazole marker, was used to detect hypoxia in downstream pericentral regions of the lobule. Following one large dose of EtOH or chronic enteral EtOH for 1 month, pimonidazole binding was increased significantly in pericentral regions of the liver lobule, which was diminished with GdCl3. Enteral EtOH increased free radical generation detected with electron spin resonance (ESR). These radical species had coupling constants matching alpha-hydroxyethyl radical and were shown conclusively to arise from EtOH based on a doubling of the ESR lines when 13C-EtOH was given. Alpha-hydroxyethyl radical production was also blocked by the destruction of Kupffer cells with GdCl3. It is known that females develop more severe EtOH-induced liver injury more rapidly and with less EtOH than males. Female rats on the enteral protocol exhibited more rapid injury and more widespread fatty changes over a larger portion of the liver lobule than males. Plasma endotoxin, ICAM-1, free radical adducts, infiltrating neutrophils and transcription factor NFkappaB were approximately two-fold greater in livers from females than males after 4 weeks of enteral EtOH treatment. Furthermore, oestrogen treatment increased the sensitivity of Kupffer cells to endotoxin. These data are consistent with the hypothesis that Kupffer cells participate in important gender differences in liver injury caused by ethanol.

Identification of nonproliferating but viable hypoxic tumor cells in vivo

We have used the combination of pimonidazole labeling of hypoxic cells, bromodeoxyuridine labeling of proliferating cells, and cell sorting based on Hoechst 33342 perfusion to directly study hypoxia and proliferation in human tumor xenografts and transplantable murine tumors in vivo. Hypoxia was largely confined to cells in regions with the least perfusion, although in tumors exhibiting transient blood flow, hypoxic cells were not as highly localized. Similarly, proliferation and hypoxia were mutually exclusive except in areas of a tumor subjected to transient changes in perfusion. By determining the clonogenic potential, pimonidazole labeling intensity, and radiosensitivity of sorted tumor cell subpopulations, we have provided direct evidence that pimonidazole identifies hypoxic tumor cells of therapeutic relevance in vivo. Given that pimonidazole exhibits few diffusion or delivery problems and no apparent cytotoxicity, it appears to be a versatile and useful label for hypoxic cells in solid tumors.

Reductive metabolism of the hypoxia marker pimonidazole is regulated by oxygen tension independent of the pyridine nucleotide redox state

2-Nitroimidazoles, such as pimonidazole, are reduced in cells with low oxygen tension and are, therefore, used as hypoxia markers. However, the effect of the pyridine nucleotide redox state on pimonidazole reduction is not known. Therefore, livers from fed or fasted rats were perfused with oxygen-saturated buffer containing pimonidazole (400 microM) in the presence and absence of an inhibitor of the mitochondrial respiratory chain, potassium cyanide; these treatments were used to modulate the mitochondrial and cytosolic pyridine nucleotide redox states. Pimonidazole-induced increases in oxygen uptake over basal values were as follows: fed, 15.1 +/- 2.4; fasted, 4.2 +/- 0.8; fed + KCN, 32.1 +/- 0.9; fasted + KCN, 0.2 +/- 0.2 micromol x g(-1) x h(-1). However, if NADPH was added in excess, microsomal oxygen uptake due to oxidative metabolism of pimonidazole was independent of treatment. These results indicate that pimonidazole-stimulated O2 uptake, due predominantly to N-oxidation and glucuronidation, is dependent on the NADPH redox state. In contrast, reduced pimonidazole adducts, detected immunochemically, accumulated in pericentral regions in liver. Increasing the NADH redox state by inhibiting the mitochondrial respiratory chain with KCN decreased protein-bound pimonidazole adducts. Concomitantly, the average O2 tension of the liver was increased at least 30%. However, KCN had no effect on total pimonidazole adducts detected by ELISA, although both cytosolic (lactate/pyruvate) and mitochondrial (3-hydroxybutyrate/acetoacetate) NADH redox states were elevated by at least a factor of eight. These results indicate that, unlike oxidative metabolism, the pyridine nucleotide redox state does not determine the rate of reductive metabolism of pimonidazole. Instead, the cellular oxygen tension regulates this process. Therefore, even in cases where the supply of reducing equivalents is increased (e.g., ethanol metabolism), accumulation of the reduced bound product of pimonidazole is oxygen dependent in liver.

Development and characterization of a new model of tacrine-induced hepatotoxicity: role of the sympathetic nervous system and hypoxia-reoxygenation

Tacrine is an acetylcholinesterase inhibitor approved for the treatment of Alzheimer's disease. Unfortunately, reversible hepatotoxicity in about 30% of patients at therapeutic doses limits clinical use. The purpose of this study was to develop and characterize a model of tacrine hepatotoxicity to begin to understand the mechanisms of injury. Rats were given tacrine (10-50 mg/kg, intragatrically) and killed 24 hr later. An increase in serum aspartate aminotransferase was observed up to 35 mg/kg and histology revealed pericentral necrosis and fatty changes. Aspartate aminotransferase was increased from 12 to 24 hr and returned to control values by 32 hr. Livers were perfused in a nonrecirculating system to measure oxygen uptake and trypan blue was infused at the end of each experiment to evaluate tissue perfusion. Time for trypan blue to distribute evenly throughout the liver 3 hr after tacrine treatment was significantly increased (6.9 +/- 1.3 min) compared to controls (1.0 +/- 0.3 min) reflecting decreased tissue perfusion. Tacrine also significantly increased the binding of a hypoxia marker, pimonidazole, in pericentral regions almost 3-fold, and increased portal pressure in vivo significantly. It is hypothesized that tacrine, by inhibiting acetylcholine breakdown in the celiac ganglion, increases sympathetic activity in the liver leading to vascular constriction, hypoxia and liver injury. To test this hypothesis, the hepatic nerve was severed and animals were allowed to recover before tacrine treatment. This procedure significantly reduced serum aspartate aminotransferase, time of dye distribution, pimonidazole binding and portal pressure. Furthermore, a free radical adduct was detected with spin trapping and electron spin resonance spectroscopy 8 hr after tacrine treatment, providing evidence for reoxygenation. When catechin (100 mg/kg, i.p.), a free radical scavenger, was given before tacrine, injury was decreased by about 45%. Furthermore, feeding 5% arginine in the diet significantly reduced portal pressure and time of dye distribution. These data are consistent with the hypothesis that tacrine hepatotoxicity is a hypoxia-reoxygenation injury mediated through the sympathetic nervous system.

Hypoxia marker labeling in tumor biopsies: quantification of labeling variation and criteria for biopsy sectioning

BACKGROUND AND PURPOSE

The error associated with using biopsy-based methods for assessing parameters reflective of the tumor microenvironment depends on the variability in distribution of the parameter throughout the tumor and the biopsy sample. Some attention has been given to intratumoral distribution of parameters, but little attention has been given to their intrabiopsy distribution. We evaluated the intrabiopsy distribution of CCI-103F, a 2-nitroimidazole hypoxia marker.

MATERIALS AND METHODS

The hypoxia marker CCI-103F was studied in dogs bearing spontaneous solid tumors. Two biopsies were taken from each of seven tumors, for a total of 14 biopsies. Biopsies were serially sectioned and four to six contiguous slides from each 100-150 microm of the biopsy were used to formulate the best estimate of CCI-103F labeled area throughout the biopsy sample. One, two or four slides were then randomly selected from each biopsy and the labeled area, based on this limited sample, was compared to the estimate obtained from counting all available slides. Random sampling of slides was repeated 1000 times for each biopsy sample.

RESULTS

CCI-103F labeling variance throughout the biopsy decreased as the estimated overall labeled area in the biopsy decreased. The error associated with estimating the overall labeled area in a biopsy from a randomly selected subset of slides decreased as the number of slides increased, and as the overall labeled area in the biopsy decreased. No minimally labeled biopsy was classified as unlabeled based on limited sampling.

CONCLUSION

With regard to CCI-103F labeling, quantification of the labeled area in four randomly selected slides from a biopsy can provide, in most biopsies, an estimate of the labeled area in the biopsy within an absolute range of +/-0.05.

Longevity of pimonidazole adducts in spontaneous canine tumors as an estimate of hypoxic cell lifetime

The longevity of pimonidazole adducts in tumors was quantified as an estimate of the lifetime of hypoxic cells. Pimonidazole was given before irradiation to 12 dogs bearing spontaneous tumors, and tumors were biopsied 24, 48 and 72 h later. Pimonidazole antigen was quantified in the biopsies using ELISA and immunohistochemistry. Pimonidazole antigen was detectable in the initial biopsy in all dogs. In 5 dogs the amount of detectable antigen decreased to less than 50% of the initial amount, in 5 other dogs the amount of detectable antigen decreased to an amount between 50 and 100% of the initial amount, and in 2 dogs the amount of antigen appeared to increase relative to the initial amount. Tumors with high initial adduct concentration were characterized by greater decreases in adduct concentration than tumors with low initial adduct concentration. Immunohistochemically, labeled cells were present in 11 of 12 tumors. The geographic area in tumor biopsies labeled immunohistochemically with pimonidazole adducts (labeled area fraction) tended to decrease over time in 6 dogs, remain stable in 4 dogs and seemingly increase in 1 dog. There was no relationship in individual tumors between the relative change in antigen concentration and the relative change in labeled area fraction. Hypoxic cells which bind pimonidazole may persist for days during fractionated radiation therapy, and the potential exists for them to exert a negative effect on the host.

Role of Kupffer cells, endotoxin and free radicals in hepatotoxicity due to prolonged alcohol consumption: studies in female and male rats

Alcohol ingestion results in increases in the release of endotoxin from gut bacteria or membrane permeability of the gut to endotoxin, or both. Female rats are more sensitive to these changes. Elevated levels of endotoxin activate Kupffer cells to release substances such as eicosanoids, tumor necrosis factor-alpha and free radicals. Prostaglandins increase oxygen uptake and most likely are responsible for the hypermetabolic state in the liver. The increase in oxygen demand leads to hypoxia in the liver, and on reperfusion, alpha-hydroxyethyl free radicals are formed that lead to tissue damage in oxygen-poor pericentral regions of the liver lobule.

Chronic enteral ethanol treatment causes hypoxia in rat liver tissue in vivo

It is known that ethanol increases oxygen consumption in in vitro liver models, which could lead to hypoxia. Although it was shown recently that one large dose of ethanol caused hypoxia in rat liver in vivo, whether ethanol produces hypoxia in a clinically relevant chronic model remains unclear. In the present study, therefore, the effect of chronic ethanol on hypoxia was investigated in vivo using the 2-nitroimidazole hypoxia marker, pimonidazole. Male Wistar rats (300-325 g) were exposed to enteral ethanol continuously for 4 weeks. In this model, rats develop steatosis, inflammation, and necrosis characteristic of early stages of clinical alcoholic liver disease in humans. One hour before they were killed, rats were injected with pimonidazole (120 mg/kg intravenously), and livers were surgically isolated, removed, and fixed. Protein-bound pimonidazole adducts were identified on formalin-fixed, paraffin-embedded tissue with immunohistochemistry. Ethanol administration for 4 weeks significantly increased serum aspartate transaminase levels and hepatic pathology scores for steatosis, inflammation, and necrosis, as expected. Ethanol treatment significantly increased both the extent and number of cells that stained positive for pimonidazole compared with control animals given an enteral diet without ethanol. Quantitative image-analysis of pimonidazole binding showed that 4 weeks of ethanol administration nearly doubled the pimonidazole-positive area in tissue. Ethanol also increased pimonidazole binding significantly at 7 days, long before inflammation and necrosis could be detected. These results indicate that chronic ethanol causes hypoxia at the cellular level in rat liver in vivo and lend support to the hypothesis that hypoxia is involved in mechanisms of early alcoholic liver injury.

Proliferation and hypoxia in human squamous cell carcinoma of the cervix: first report of combined immunohistochemical assays

PURPOSE

To characterize the distribution of hypoxia and proliferation in human squamous cell carcinoma of the cervix via an immunohistochemical approach prior to initiation of therapy.

METHODS AND MATERIALS

Patients with primary squamous cell carcinoma of the cervix uteri received a single infusion of the 2-nitroimidazole, pimonidazole (0.5 g/m2 i.v.), and 24 h later punch biopsies of the primary tumor were taken. Tissue was formalin fixed, paraffin embedded, and sectioned for immunohistochemistry. Hypoxia was detected by monoclonal antibody binding to adducts of reductively activated pimonidazole in malignant cells. Staining for endogenous MIB-1 and PCNA was detected in tumor cells via commercially available monoclonal antibodies. Point counting was used to quantitate the fraction of tumor cells immunostained for MIB-1, PCNA, and hypoxia marker binding.

RESULTS

Immunostaining for pimonidazole binding was distant from blood vessels. There was no staining in necrotic regions, and only minimal nonspecific staining, mostly in keratin. In general, cells immunostaining for MIB-1 and PCNA did not immunostain for pimonidazole binding. Cells immunostaining for MIB-1 and PCNA showed no obvious geographic predilection such as proximity to vasculature. Quantitative comparison showed an inverse relationship between hypoxia marker binding and proliferation.

CONCLUSIONS

Immunohistochemical staining for pimonidazole binding is consistent with the presence of hypoxic cells in human tumors and may be useful for estimating tumor hypoxia prior to radiation therapy. Immunostaining for pimonidazole binding is an ideal complement to immunohistochemical assays for endogenous proliferation markers allowing for comparisons of tumor hypoxia with other physiological parameters. These parameters might be used to select patients for radiation protocols specifically designed to offset the negative impact of hypoxia and/or proliferation on therapy. The inverse relationship between pimonidazole binding and proliferation markers is a preliminary result requiring verification.

Quantification of CCI-103F labeling heterogeneity in canine solid tumors

PURPOSE

The purposes of this study were to assess sources of variation in the distribution of nitroimidazole-labeled hypoxic cells in canine tumors and to quantify the reliability of estimating overall nitroimidazole-labeled area fraction from biopsies.

METHODS AND MATERIALS

Hypoxic cells were labeled in 24 canine tumors by immunostaining of the nitroimidazole hypoxia marker CCI-103F. In tumors with a volume < 100 cm3, each cubic centimeter of tumor was examined; in larger tumors 100 randomly selected 1 cm3 samples were examined. These data were used to estimate the overall CCI-103F-labeled area fraction in the tumor. A variance components model was used to quantify intertumoral, intratumoral, and within slide (residual) sources of variation. The ability to estimate intratumoral CCI-103F-labeled area fraction based on information obtained from biopsies was assessed by randomly selecting two or four samples from the dataset for each tumor and comparing the mean CCI-103F-labeled area fraction from this limited sample to the labeled area fraction based on each cubic centimeter; this simulation process was repeated 1000 times.

RESULTS

Intratumoral (27% of total) and intertumoral (30% of total) variation in CCI-103F-labeled area fraction were similar. Residual variation (variation at the microscopic level) accounted for 43% of total variation in CCI-103F labeling. Intratumoral variation in labeling decreased as the intratumoral CCI-103F mean labeled area fraction decreased. The accuracy of estimating the intratumoral CCI-103F-labeled area fraction in a tumor from limited sampling increased as the number of samples increased or the intratumoral labeled area fraction decreased. When four random samples were used to estimate overall CCI-103F-labeled area fraction in the tumor, estimates from approximately 90% of the 1000 simulations were within 0.10 of the intratumoral CCI-103F-labeled area fraction. Classifying a minimally labeled tumor as unlabeled based on limited sampling was unlikely.

CONCLUSION

Despite intratumor variation, acceptable estimates of nitroimidazole-labeled cells in a tumor may be obtained from a clinically feasible number of biopsies.

Acute alcohol produces hypoxia directly in rat liver tissue in vivo: role of Kupffer cells

Previous studies using liver slices and isolated perfused rat liver have suggested that ethanol causes hypoxia by increasing oxygen consumption. However, ethanol also increases blood flow to the liver, a phenomenon that may counteract the effects of hypermetabolism by increasing oxygen delivery. Thus whether ethanol causes hypoxia in vivo remains unclear. To clarify this important point, female Sprague-Dawley rats (100-125 g) simultaneously received pimonidazole (120 mg/kg ip), a 2-nitroimidazole hypoxia marker, and one large dose of ethanol (5 g/kg ig), which increase hepatic oxygen uptake dramatically and elevate ethanol metabolism (swift increase in alcohol metabolism) in 2-3 h. After 2 h, ethanol significantly increased the accumulation of bound pimonidazole in pericentral regions of the liver lobule. Treatment of animals with the Kupffer cell-specific toxicant, GdCl3 (10 mg/kg iv, 24 h before experiment), blocked ethanol-induced increases in pimonidazole binding. It is concluded that one large dose of ethanol causes pericentral hypoxia in rat liver tissue in vivo and that Kupffer cells are involved.

Marking hypoxic cells for complement and cytotoxic T lymphocyte-mediated lysis: using pimonidazole

Artificial antigens are created when 2-nitroimidazoles bind to hypoxic cells. These antigens have been used in the immunodetection of tumour hypoxia but they might also serve to stimulate immune lysis of hypoxic tumour cells by complement- and cell-mediated processes. In order to test this hypothesis, lymphocytes isolated from the spleens of C3H/HeN mice that had been immunised with pimonidazole-labelled 3152-PRO cells were subcultured and tested for their ability to lyse chromium-51 loaded, pimonidazole-labelled 3152-PRO cells in an in vitro assay. In a parallel study, commercially available, rabbit complement was tested for its ability to lyse pimonidazole-labelled V79-4 cells in the presence of monoclonal antibodies which recognise protein adducts of reductively activated pimonidazole. Complement-mediated cell lysis was measured by means of an MTT assay. Complement-mediated and cell-mediated lysis was observed at pimonidazole concentrations which, in themselves, do not produce cell killing.

Evidence that hypoxia markers detect oxygen gradients in liver: pimonidazole and retrograde perfusion of rat liver

Nitroimidazole markers of tumour hypoxia bind to normoxic liver and the question has been raised whether this is due to low oxygen concentration or microregional activity of specialised nitroreductases. To answer this question, the binding patterns of the 2-nitroimidazole, pimonidazole, were compared following perfusion of surgically isolated rat livers in anterograde and retrograde directions. Perfusion at low flow rates in anterograde or retrograde directions can be used intentionally to alter oxygen gradients without altering enzyme distributions. Perfusion by means of the portal vein (anterograde direction) produced pimonidazole binding in the pericentral region of liver similar to that observed for pimonidazole binding in vivo. A complete reversal of this binding pattern occurred when the isolated liver was perfused by way of the central vein (retrograde direction). In this case, pimonidazole binding occurred in the periportal region. The extent and intensity of binding in the periportal region during perfusion in the retrograde direction was similar to that in the pericentral region during perfusion in the anterograde direction. It is concluded that low oxygen concentration rather than the non-homogeneous distribution of nitroreductase activity is the primary determinant of 2-nitroimidazole binding in liver.

Distribution of hypoxia and proliferation associated markers in spontaneous canine tumors

The therapeutic response of malignant tumors depends on a number of factors associated with tumor microenvironments including the possibility that these microenvironments change during treatment. Two factors, tumor hypoxia and cell proliferation, have been examined in spontaneous canine tumors undergoing multifraction radiation therapy. The approach utilizes immunohistochemical analyses of hypoxia (CCI-103F) and proliferation associated (PCNA) antigens in biopsy samples taken before and after 5 daily fractions of 3 Gy (total dose 15 Gy). The tissue samples were formalin-fixed and paraffin-embedded for the immunohistochemical study. Immunostaining of the sections for PCNA and hypoxia marker reveals little or no overlap when the analysis is made prior to irradiation. An increased degree of overlap seems to occur after 15 Gy but the situation is complicated by a change towards more diffuse PCNA immunostaining in the cells of the irradiated tissues.

Distribution of the hypoxia marker CCI-103F in canine tumors

PURPOSE

The purpose of this study was to identify the prevalence and distribution of hypoxic tumor cells in spontaneous canine tumors, and to relate these parameters to various tumor and patient characteristics, such as tumor volume, tumor type, or tumor location.

METHODS AND MATERIALS

Hypoxic tumor cells were labeled in vivo in 32 primary malignant canine tumors by bioreductive binding of the nitroimidazole hypoxia marker CCI-103F. CCI-103F was given at 40 mg/kg i.v. Tumors were completely excised, and CCI-103F adducts were detected in histologic sections (mean, 138 sections-per-tumor) by peroxidase-antiperoxidase immunostaining. Area fraction (area labeled/total area examined) of labeled regions was measured via computer assisted image analysis. In tumors with a volume < 100 cm3, each cubic centimeter of tumor was examined; in larger tumors 100 randomly selected 1 cm3 samples were examined.

RESULTS

There were 13 soft-tissue sarcomas, 11 mast-cell tumors, five carcinomas, two lymphosarcomas, and one melanoma. Tumors varied from < .001 to > 2000 cm3. Labeled cells were present in 31 of 32 canine tumors examined, and varied between 0 and 35%. Mean (+/- SD) % label was 12.2% +/- 16.7%; 13 of the 32 dogs had % labeled area < 5.0%. The area fraction was not related to tumor site, tumor type, tumor volume, presence and degree of necrosis or tumor grade. Dog characteristics such as sex, age, and body size did not affect the degree of labeling of tumors. CCI-103F adducts were randomly distributed grossly, and at the microscopic level were not found near blood vessels or regions containing mitoses. Labeling was seen in a variety of normal tissues; not all binding in normal tissues could be attributed to hypoxia.

CONCLUSION

CCI-103F labeling of hypoxic regions in tumors provides a nonradioactive method of detecting nitroimidazole adducts at the cellular level, and allows concurrent histologic examination. The pattern of labeling is consistent with detection of hypoxic tumor cells arising from oxygen diffusion limitations. This method may have clinical applicability in the detection of tumor hypoxia.