Assessment of regional cytochrome P450 activities in rat liver slices using resorufin substrates and fluorescence confocal laser cytometry.

Characterizing constitutive activities and inducibility of various cytochrome P450 isozymes is important for elucidating species and individual differences in susceptibility to many toxicants. Although expression of certain P450s has been studied in homogenized tissues, the ability to assess functional enzyme activity without tissue disruption would further our understanding of interactive factors that modulate P450 activities. We used precision-cut, viable rat liver slices and confocal laser cytometry to determine the regional enzyme activities of P450 isozymes in situ. Livers from control and beta-naphthoflavone (beta NF)-treated rats were sectioned with a Krumdieck tissue slicer into 250-microns thick sections. A slice perfusion chamber that mounts on the cytometer stage was developed to allow for successive measurement of region-specific P450-dependent O-dealkylation of 7-ethoxy-, 7-pentoxy-, and 7-benzyloxyresorufin (EROD, PROD, and BROD activity, respectively) in the same liver slice. Images of the accumulated fluorescent resorufin product within the tissue were acquired using a confocal laser cytometer in confocal mode. As expected, slices isolated from beta NF-treated rats showed high levels of centrilobular EROD activity compared to slices from control rats, whereas PROD and BROD activities remained at control levels. These techniques should allow for the accurate quantification of regional and cell-specific P450 enzyme activity and, with subsequent analysis of the same slice, the ability to correlate specific P450 mRNAs or other factors with enzymatic activity. Moreover, these techniques should be amenable to examination of similar phenomena in other tissues such as lung and kidney, where marked heterogeneity in cellular P450 expression patterns is also known to occur. ImagesFigure 1.Figure 2.Figure 3.Figure 3.Figure 4.Figure 4.Figure 5.Figure 6.

should be amenable to =vamnation of similarp in other tissues suh as lung andd kidney, where mar heo ity in cdlular P450 epon patterns is also known to occur. Key words: cytochrome P450, hepatic zones, laser cytometry, liver ic, reonal atiity.
Environ HBpl -P 104: 536-543 (196) Cytochrome P450 monooxygenases (P450s) are a supergene family of enzymes involved in the biotransformation of a wide range of both endogenous and exogenous compounds. P450s play important roles in the metabolism of many drugs and in the activation of a variety of chemical toxicants and carcinogens in both humans and animals (1,2). High levels of P450s exist principally in the liver (1), and studies using immunohistochemical techniques (3)(4)(5)(6) and in situ hybridization (7) have demonstrated differences in the regional distribution and region-specific induction of several P450s within the liver.
Constitutive activities and inducibility of different P450 isozymes are important in dictating species and individual differences in susceptibility to toxicants. The ability to assess gene expression and resulting enzyme activity without tissue disruption on an individual cell basis would further our ability to identify linkages between P450 expression and the zonal distribution of hepatic lesions caused by different toxicants. Recently, we reported on the use of alkoxyresorufin homologues in combination with noninvasive scanning laser cytometry as a method for directly determining CYPlAl functional activity in hepatoma cell lines (8). In the present study, we used precision-cut, viable rat liver slices and confocal laser cytometry to determine the regional activities of P450 isozymes in situ. A slice perfusion chamber that mounts on the cytometer stage was developed to allow for successive measurement of region-specific, P450-dependent 0-dealkylation of 7-ethoxy-, 7-pentoxy-, and 7-benzyloxyresorufin (EROD, PROD, and BROD activity, respectively) in the same liver slice. EROD activity in rat liver has been shown to be specific to CYPIA isozymes and is inducible by ,-naphthoflavone (,NF) and other polycyclic aromatic hydrocarbon-type inducing agents (8,9). BROD and PROD activities have been shown to be specific to CYP2B and CYP3A isozymes and are inducible by phenobarbital (9,10). The perfusion system ensures a constant concentration of substrate and limits the extracellular accumulation of the fluorescent product, resorufin (9). In addition, the slice perfusion chamber allows for residual substrate and product to be washed away between successive perfusion assays.

Materials and Methods
Chemicals. Earle's balanced salt solution (EBSS) and Williams' E medium were purchased from Gibco BRL (Grand Island, New York). fiNF, factor IIand factor VIIdeficient rat plasma, rat thrombin, and dicumarol were purchased from Sigma (St. Louis, Missouri). 7-Ethoxy-, 7-pentoxy-, and 7-benzyoloxyresorufin (ER, PR, and BR, respectively) were purchased from Boehringer Mannheim Corp. (Indianapolis, Indiana), and resorufin was purchased from Aldrich Chemical Co. (Milwaukee, Wisconsin). Animals and slice preparation. Adult, male Sprague-Dawley rats (150-200 g) were purchased from Simenson Labs (Gilroy, California) and were fed Wayne Lablox and given water ad libitum. PNFtreated rats were administered 65 mg/kg body weight ,3NF in corn oil by intraperitoneal injection for 2 consecutive days before sacrifice. Control rats were untreated. The perfusion technique and methods for measurement of alkoxyresorufin 0dealkylase activities by scanning confocal laser cytometry were developed during a number of preliminary experiments. Immunocytochemical, microsomal, and dynamic organ culture (DOC) data were obtained from the caudate lobes of a control and a ,NF-treated rat. Fluorescence images and slice perfusion data were obtained from the caudate lobes of separate control and ,NF-treated rats.
Precision-cut, viable liver slices used in laser cytometry experiments and in DOC experiments were prepared as described by Krumdieck et al. (11) using the Krumdieck tissue slicer (K & F Research Co., Birmingham, Alabama). Briefly, animals were killed by ether anesthetization/cervical dislocation. Livers were excised and immediately placed in Williams' E medium containing 10 mM HEPES (pH 7.4; referred to as WEM from this point). An 8-mm metal biopsy corer was used to prepare tissue cores from the liver lobes. Cores were then placed in the Krumdieck slicer while submerged in WEM. Optimal slice thickness was previously shown to be 250 pim (11). The first and last slices containing the liver capsule were discarded. Liver slices were then placed in DOC vials (2 slices/vial) containing 2 ml of WEM as described by Smith et al. (12). Slices were allowed to preincubate for at least 2 hr before use in subsequent perfusion assays or DOC EROD assays.
Slice perfusion. The slice perfusion apparatus shown in Figure 1 was used to successively measure region-specific EROD, PROD, and BROD activities in precision-cut liver slices. A plasma clot similar to that described by Gahwiler (13) was formed by evenly distributing 25 pl of reconstituted rat plasma (50 mg/ml PBS, pH 7.4) on a 22-mm square glass cover slip. We then added 10 p1 of reconstituted thrombin (117 U/ml in 0.15 M NaCl and 0.05 M sodium citrate, pH 6.5) to the center of the plasma-coated cover slip. The liver slice was then placed on the rapidly formed plasma clot. A cover glass perfusion chamber similar to that described by Forscher et al. (14) was formed by suspending the cover slip holding the liver slice over a microscope slide on two 720pm thick plastic spacers. The spacers and cover slip were held in place by coating spacers with a thin film of silicone grease.
The microscope slide portion of the perfusion chamber was fixed to the cytometer stage by a custom-designed stage mount. The mount positioned the slide over the microscope objective and held reagent supply and vacuum lines in position. Reagent was supplied via 0.58-mm ID polyethylene tubing attached to a 25ml reservoir positioned approximately 30 cm above the perfusion chamber. A triangular piece of filter paper attached to the vacuum line was positioned on the outlet side of the chamber, as shown in Figure 1, to wick away reagent and provide for a laminar flow of reagent through the chamber. Vacuum and reservoir height were adjusted to establish a flow rate of 0.5 ml/min. Slice perfusion experiments were carried out at room temperature. EBSS (phenol-red-free, supplemented with 10 mM HEPES, 5 mM MgCl2 and 2.5 mM CaCI2, pH 7.5), gassed with 95% 02/5% CO2 and containing 25 pM dicumarol, was used as the perfusion buffer. Dicumarol is a specific inhibitor of quinone oxidoreductase (DT-diaphorase) (15), a cytosolic enzyme involved in the reduction of resorufin to a nonfluorescent product. Dicumarol is therefore required in reaction buffers to measure maximum alkoxyresorufin O-dealkylase activities in hepatic subcellular fractions and hepatocyte homogenates (16). We used 25 pM dicumarol in our assays with liver slices, as this concentration was previously found to be optimal for measuring CYPlAl functional activity in hepatoma cell lines by scanning laser cytometry (8). Sulfate anion was omitted from the reaction buffer to minimize conjugation via sulfation (17), as previously described (8).
We determined region-specific EROD, BROD, and PROD activities by adding the appropriate alkoxyresorufin substrate to the perfusion buffer to a final concentration of 5 piM. EBSS without substrate was used to remove residual substrate and product between successive perfusions. Slices were washed with EBSS until fluorescence returned to baseline levels and was monitored by continuously scanning the liver slice.
Laser cytometry. Alkoxyresorufin 0dealkylase activity was assessed directly in liver slices using an ACAS (adherent cell analysis and sorting) Ultima Laser Cytometer (Meridian Instruments, Okemos, Michigan) equipped with an argon ion laser and image analysis software. Initial focusing was carried out by examining the slice with normal transmitted light using a lOx objective. The same objective was used for laser illumination of the slice in confocal mode (pinhole setting of 100 pm). Maximum fluorescence response occurred within the first two to three cell layers from the surface of the slice and could be optimized by scanning in the Z. direction upon addition of substrate. Fluorescence due to resorufin formation was monitored using an excitation wavelength of 514 nm and an emission wavelength of >570 nm (using a long-pass dichroic filter). The photomultiplier tube voltage was held constant and data were collected using the kinetics software package supplied with the instrument. Data were analyzed on a DASY 9000 workstation (Meridian Instruments). EROD, BROD, and PROD activities were measured as increases in fluorescence intensity (arbitrary scale) within user-defined polygon regions of the liver slice exhibiting maximum EROD activity. Representative fluorescence pseudocolor images were saved as tagged image file format (TIFF) files and imported into a MacIntosh IIci personal computer. These images were then printed using PowerPoint software (Microsoft Corp., Redmond, Washington) and a Tektronix Phaser II SDX dye sublimation printer (Tektronix Corp., Beaverton, Oregon).
In vitro assays. We measured EROD activity in liver slices isolated from control and ,NF-treated rats in DOC as follows. Slices were preincubated for 2 hr in WEM as described above. Slices were then rinsed with EBSS, then incubated in 2 ml EBSS containing 5 pM ethoxyresorufin (ER) and 25 pM dicumarol. After the specified incubation time, the concentration of fluorescent resorufin product in the culture medium was determined by fluorescence spectrometry. We determined resorufin concentrations by comparing fluorescence intensity with a standard curve prepared from resorufin standards. All fluorescence measurements were made on a Perkin-Elmer LS-50 Luminescence Spectrofluorometer (Beaconsfield, UK) using an excitation wavelength of 530 nm and an emission wavelength of 580 nm.
Liver microsome samples were prepared by standard procedures. Microsomal protein concentrations were determined by the bicinchoninic acid method described by Smith et al. (18). Microsomal (alkoxyresorufin-O-dealkylase activity assays were performed essentially as described by Burke and Mayer (19) and Lubet et al. (20). Microsomes were allowed to preincubate for 2 min in 2 ml of 100 mM Tris-HCL (pH 7.5) containing 5 mM MgCl2 and 5 pM alkoxyresorufin substrate. Reactions were initiated by adding 10 pl of 100 mM NADPH (final concentration of 500 pM NADPH) and were carried out at 370C. Maximum EROD activities were obtained using 148 pig of control microsomal protein and 14 pg of PNF-treated microsomal protein. Maximum BROD and PROD activities were obtained using 148 pg of control microsomal protein and 140 pg of 3NF-treated microsomal protein. We monitored reaction kinetics on the luminescence spectrofluorometer described above using the kinetics software package supplied with the instrument. Slice viability. We used control rat liver slices to determine the effect of culture method (DOC versus perfusion) and the effect of treatment (EBSS control versus EBSS containing 5 pM ER and 25 pM dicumarol) on slice viability. Intracellular K+ concentrations were measured in three slices before preincubation began (0 hr), in three slices that had been preincubated 2 hr in WEM, and in three slices from each of the culture and treatment groups at two time points (30 and 60 min). Intracellular K+ determinations were performed as described by Fisher et al. (21).
Immunocytochemistry. Immunocytochemical staining of rat liver sections was performed by the method of Farin et al. (22) using a Vectastain ABC kit (Vector Laboratories, Burlingame, California). Briefly, acetone-fixed, 5-pm thick, paraffin sections of control and PNF-treated rat liver on poly-L-lysine coated slides were deparaffinized and rehydrated. Tissue sections were sequentially exposed to normal goat serum for 30 min to decrease nonspecific binding, previously prepared anti-CYPlAl primary antibody for 90 min, biotinylated goat-anti-rabbit IgG secondary antibody for 60 min, avidin-conjugated horseradish peroxidase complex for 60 min, 3,3'-diaminobenzidine in Trissaline containing H202 for 10 min, and 1% osmium tetroxide for 2 min. Between each of these components, the tissue sections were washed extensively with phosphate-buffered saline. The tissue sections were counterstained with 2% methyl green, and dehydrated, and cover slips were placed on top of the tissue sections with a drop of mounting fluid. We confirmed the specificity of the CYPlAI antibody by Western blot analysis of control and ,BNFtreated HepG2 cells and rat liver (23). The antibody bound only to a single microsomal protein isolated from PNF-treated HepG2 cells, indicating that the antibody does not cross-react with CYP1A2 [see Farin et al. (22) for details of antibody preparation and specificity].
Statistical analysis. Intracellular K+ concentrations for preincubated slices and slices that were perfused or incubated by DOC both with and without substrate at 30 and 60 min were compared by analysis of variance. EROD, PROD, and BROD activities measured by slice perfusion were Articles v Imaging P450 activity in rat liver slices compared by regression analysis. We used the StatView statistics program (Brain-Power Inc., Calabasas, California) for statistical analyses. Scheffe's F-test was used to establish significant differences.

Results
Liver slice intracellular K+ concentrations increased from 35.4 ± 5.0 pmol K+/g liver (mean ± SD) to 56.6 ± 5.0 pmol K+/g liver during the 2-hr preincubation period (data not shown). Recovery of intracellular K+ during the preincubation period is characteristic of viable liver slices (12). Preincubated slices that were then incubat-ed with or without ER for 30 and 60 min by DOC at 370C maintained 94-103% of the intracellular K+ measured in preincubated slices. Intracellular K+ concentrations in liver slices perfused at room temperature appeared to decrease after a 30-min perfusion (42.9 ± 7.8 and 42.4 ± 8.1 in control and ER-treated slices, respectively), then increase after a 60-min perfusion (50.0 ± 5.8 and 44.9 ± 4.1 in control and ERODtreated slices, respectively). However, when the mean intracellular K+ values for all treatment groups (DOC versus perfusion and EBSS control versus EBSS containing 5 pM ER and 25 pM dicumarol) and preincubated slices were compared, the differences were not significant (p = 0.05).
The fluorescence images in Figure 2 were collected over a period of minutes in a typical experiment in which liver slices isolated from control and PNF-treated rats were perfused on the laser cytometer stage with EBSS containing 5 pM ER. Areas of increased fluorescence are due to the formation of the fluorescent resorufin product and are indicative of high EROD activity. Whole-slice scans (data not shown) revealed a heterogeneous pattern of EROD activity, suggesting differential induction of CYP IAl activity across regions of the liver lobule. The time elapsed during whole-slice scans made it impractical for kinetic analyses of entire slices, as fluorescence intensity increased substantially during the time required to stage-scan an entire slice (10 mm x 10 mm). Based on these kinetic considerations, we chose to scan smaller fields (1 mm x 1 mm, requiring <30 sec/scan using the fast-scanning mirror option on the ACAS). These fields contained vascular structures that appeared to be central veins. These structures were visible by light microscopy and could be targeted during initial focusing by manually positioning the slice with the computer-controlled X,Y stage. Areas of the slice where vascular structures appeared to be circular, indicating they were cut perpendicular to the slice surface, were suitable fields for resolution of the liver lobule. The size of the areas scanned by confocal laser cytometry roughly corresponds to the size of areas of distinct zonal immunocytochemical staining of CYPlAI shown in the photomicrograph of the liver section isolated from a ,BNF-pretreated rat (Fig. 3).
The fluorescence images in Figure 2 show areas of intense fluorescence response in a liver slice isolated from a fNF-treated rat. The fluorescence level attained after 1275 sec of scanning in these areas is approximately 600-800 fluorescence intensity uniis (FIU), which is at least twofold higher than that observed in less-intense adjacent areas. Immunocytochemical data showing high levels of CYPlAI in centrilobular areas suggest that these images depict high EROD activity in centrilobular regions of the liver slice containing what appear to be small central veins.
Perfusion of control rat liver slices with ER resulted in a low fluorescence response and reflects the relatively low EROD activity in uninduced rat liver (9,10,16,24). The control rat liver section in Figure 3  rats with selected alkoxyresorufin substrates (5 pM). Slices were sequentially perfused on the laser cytometer stage as described in Materials and Methods. ER , BR, and PR represent time point at which ethoxy-, benzyloxy-, or pentoxyresorufin were added to the perfusion buffer, respectively. WO (washout) represents time point at which respective substrate was removed from the perfusion buffer to wash out residual resorufin product and return fluorescence to background levels before subsequent addition of next substrate. Values represent the mean ± SD (dotted lines) of seven circular areas containing the maximum fluorescence response in ONF-treated liver slice. Control slice values were measured in one large, circular area containing maximum fluorescence response.
,NF-treated rat liver section shows a marked increase in CYPlAl expression, but only in centrilobular regions.
To further substantiate the suggested association between high centrilobular expression of CYPlAl and the heterogeneous fluorescence response of the ,BNFtreated rat liver slice perfused with ER, we sequentially perfused liver slices from control and PNF-treated rats with ER, benzy-loxyresorufin (BR), and pentoxyresorufin (PR). Figure 4 demonstrates that the fluorescence response of control rat liver slices toward BR and PR is comparable to the fluorescence response observed when control slices are perfused with ER. Increases in fluorescence intensity were below the limits of quantification ( 30, or 60 min as described in Materials and Methods. EROD activity in liver slices isolated from OiNF-treated rats was linear for periods of up to 60 min, and the rate of resorufin formation was nearly 60-fold higher in treated slices than in control slices. slices have low, constitutive levels of EROD, BROD, and PROD activity, as has been shown previously in microsomes from untreated rats (9,24). In contrast to the equivalent response of control slices toward ER, BR, and PR, slices isolated from ,NF-treated rats show a highly differentiated response toward these three substrates. Figure 4 shows the increase in fluorescence intensity of seven defined areas containing cells and/or groups of cells exhibiting maximum fluorescence response. The relative maximum fluorescence response of liver slices from OiNF-treated rats toward ER is approximately 20-fold higher than the response toward BR and PR. The maximum rate of increase in fluorescence intensity over an approximate 3-min time period for each of the substrates in 3NF-treated slices was 307, 18.5, and 16.1 FIU/min for ER, BR, and PR, respectively. These results indicate that liver slices from ,NF-treated rats have high EROD activity relative to BROD and PROD activity and that EROD, BROD, and PROD activities are at least 30-, 1.8-, and 1.6-fold greater, respectively, in ,NFtreated slices than in control slices.
To ensure that the lack of fluorescence response of liver slices from ,NF-treated rats perfused with BR and PR was not due to possible depletion of cofactors resulting from the preceding ER perfusion, we reperfused slices with ER following BR and PR perfusion. The fluorescence response of the liver slice during the second ER perfusion approached the fluorescence response observed during the initial perfusion (Fig.  4). The maximum rate of increase in fluorescence intensity for the second ER perfu- sion was 193 FIU/min. Although this represents only 63% of the activity observed in the initial ER perfusion (significant at p <0.05), EROD activity measured during the second ER perfusion is substantially greater than the preceding BROD and PROD activities. Taken together, these findings suggest that only a small depletion of cofactors that support 0-dealkylase activity, such as NADPH, occurred during the 60 min preceding the second ER perfusion. The magnitude of the second ER response indicates that the low BROD and PROD activities shown in Figure 4 are more likely due to low substrate-specific isozyme expression rather than depletion of cofactors. Figure 5 shows the regression lines for the amount of the fluorescent resorufin product formed over time by liver slices isolated from both control and ONF-treated rats incubated with 5 pM ER in DOC. The correlation coefficient for the regression line that describes the rate of resorufin formation (EROD activity) in ,NFinduced liver slices indicates a high degree of linearity (R2 = 0.949). This suggests, as does the second ER perfusion in Figure 4, that cofactors supporting 0-dealkylase activity are adequately maintained in liver slices from ONF-treated rats for periods of at least 60 min. In addition, the slope of the PINF-treated regression line is nearly 60-fold greater than the slope of the control regression line. Figure 6 further demonstrates the selective induction of EROD activity by ONF, as previously shown in whole-liver slice perfusion experiments (Fig. 4). Microsomes isolated from ,BNF-treated rats had greater than 20-fold higher EROD activity than control rat liver microsomes. This is consistent with previous studies that compare microsomal EROD activity in control and PNF-treated rats (9,24). It is interesting to note, however, that the profile of 0dealkylase activities in microsomes does not parallel the activity profile seen in wholeliver slice perfusion experiments. The activity profile for liver slices isolated from 3NF-treated rats could be described as EROD>>>BROD=PROD, whereas in microsomes from PNF-treated rats, the activity profile is EROD>BROD>> PROD. These differences may be due to the intact nature of the liver slice and may reflect differences in isozyme kinetics as a result of the more biologically relevant cofactor and substrate concentrations achieved in whole-cell systems.

Discussion
The fluorescence images in Figure 2 show that perfusion of liver slices isolated from 3NF-treated rats with EBSS containing 5 pM ER results in an intense fluorescence response localized to specific regions of the liver lobule. Immunocytochemical staining of liver sections isolated from control and INF-treated rats (Fig. 3) demonstrates that ,BNF treatment results in centrilobular induction of CYPlAl. Precision-cut liver slices incubated with ER in DOC show that slices isolated from ONF-treated rats ( Fig. 5) have nearly 60-fold higher EROD activity than slices isolated from control rats. In addition, analysis of a series of alkoxyresorufin O-dealkylase activities in precision-cut liver slices by confocal laser cytometry (Fig. 4) shows that ONF treatment results in increased EROD activity to a much greater extent than either PROD or BROD activity. A similar response is observed in liver microsomes isolated from control and jNF-treated rats (Fig. 6), except that BROD activity was also induced to a greater extent in microsomes than in intact liver slices. Based on these findings, we conclude that the fluorescence images in Figure 2 accurately reflect high levels of centrilobular EROD activity due to region-specific induction of CYPlAl by iNF in the rat.
fNF and 3-methylcholanthrene (MC) are prototypic inducing agents representative of a broad range of compounds such as polyaromatic hydrocarbons, coplanar polychlorinated biphenyls, and 2,3,7,8-tetrachlorodibenzo-p-dioxin. Treatment with these compounds results in increased expression of the CYPIA gene family (2). Baron et al. (25) showed by immunocytochemical techniques that treatment of male Holtzman rats with ,BNF (40 mg/kg) or MC (25 mg/kg) results in increased expression of CYP1A isozymes that is nearly homogeneous across the liver lobule. In addition, the relative amounts and distribution of these isozymes are the same in the right, median, left, and caudate lobes of the liver (25). More recently, van Sliedregt and van Bezooijen (6) found that the induction pattern in Brown Norway rats was dependent on dose and that at low doses of MC (2.5, 5, 7.5 and 10 mg/kg) the highest expression levels were concentrated around the central vein. Higher doses of MC (25 mg/kg) resulted in a homogeneous pattern of induction.
Based on these findings and the comparatively high dose of fNF (65 mg/kg) used in our experiments, we would expect to see a homogenous pattern of CYPlAI induction. However, we found that sections isolated from PNF-treated Sprague-Dawley rats (Fig. 3) show very dark staining for CYPlAI in centrilobular areas only, a pattern identical to that observed by van Sliedregt et al. (6) at low doses of MC.
Recent work using periportal and perivenous hepatocytes isolated by zonerestricted digitonin treatment during in situ perfusion have demonstrated regional differences in the constitutive expression and inducibility of a number of P450s (4,26). In addition, Buhler et al. (4) noted that the inducible forms of P450 they studied (2B1/2, 2E1, and 3A1) were generally induced in hepatocytes of the same zonal origin in which they were constitutively expressed. These findings led them to suggest that phenotypic differences exist between centrilobular and periportal hepatocytes with respect to factors that control P450 expression, causing hepatocytes of different zonal origins to respond differently to exogenous and endogenous signals.
Comparative studies investigating the phenobarbital responsiveness of rat hepatic hyperplastic nodules (HHN) have provided further evidence for phenotypic differences between centrilobular and periportal hepatocytes in factors that control P450 isozyme expression. Phenobarbital induces CYP3A1 in centrilobular hepatocytes (23) and CYP2B1/2 in centrilobular and midzonal hepatocytes (23,27). Chen et al. (23,28) have demonstrated that HHNs induced by the aflatoxin BI(AFB,) administration protocol (5) arise through clonal expansion of centrilobular hepatocytes only. In contrast, HHNs induced by the Solt-Farber induction protocol (5) are derived from centriobular, midzonal, and periportal hepatocytes (28). When HHN-bearing rats were administered phenobarbital, all AFB1induced HHNs expressed increased levels of CYP3A1 and CYP2B1/2 (28). In contrast, only a portion of Solt-Farber-induced HHNs were shown to respond tO pheno-Environmental Health Perspectives * Volume 104, Number 5, May 1996 barbital as measured by increased expression of these isozymes (23,28). These findings led Chen and Eaton (28) to hypothesize that phenobarbital responsiveness is determined primarily by the zonal origin of precursor hepatocytes and provide indirect evidence for distinct, heritable phenotypic differences between hepatocytes of different zonal origin that control P450 expression.
At the present time, relatively little is known about factors controlling region-specific constitutive and inducible P450 isozyme activity. Anundi et al. (26) and Lindros et al. (29) have described linkages between zone-specific P450 isozyme expression and zone-specific hepatic lesions. In addition, the effects of P450 expression by HHNs on tumor promotion and progression have been well documented (30). These findings emphasize the importance of developing in vitro assays that could be used to colocalize functional P450 isozyme activity with factors governing P450 expression. Sidhu et al. (8) have demonstrated the ability of noninvasive scanning laser cytometry to measure CYPlAl functional activity in hepatoma cell lines. More recently, this same group demonstrated large individual cell differences in the functional activity and the immunoreactive protein content of primary rat hepatocytes exposed to prototypic P450 inducers in vitro (8,31). The ability of scanning laser cytometry to measure functional activity in a whole-cell system led us to hypothesize that laser cytometry could be used to measure functional P450 isozyme activity in viable precisioncut liver slices and that, because the architecture of the liver lobule in liver slices remains intact, marked differences in regional activity could be observed.
In conclusion, using a well-characterized induction protocol and P450 isozymespecific alkoxyresorufin substrates, we were able to show that confocal laser cytometry of precision-cut liver slices can be used to assess P450 isozyme-specific induction in intact liver tissue. More importantly, the method allows for the analysis of regionspecific induction of P450 activities, which are difficult to determine by conventional methods. These techniques should allow for the accurate quantification of P450 enzyme activity in situ and, with subsequent analysis of the same slice after fixation and processing, the ability to correlate specific P450 isozyme mRNA, specific P450 isozyme protein content, or other factors, with enzyme activity on an individual cell basis. These techniques should also be amenable to examination of similar phenomena in other tissues such as lung and kidney, where heterogeneity in cellular P450 expression is also known to occur.