rat hepatocellular carcinoma, but not biliary or paren- chymal cells from normal liver, contain an isozyme of alkaline phosphatase (ALKP) similar to the placental.
Oncodevelopmental Expression A
ka line
of
Rat Placental
Phosphatase
Detection in Oval Cells During Liver Carcinogenesis
PAUL YASWEN, PhD, NANCY L. THOMPSON, MS, and NELSON FAUSID, MD
From the Department of Pathology and Iaboratory Medicine, Brown University, Providence, Rhode Island
Oval cells isolated from livers of rats fed a choline-deficient diet containing 0.1% DL-ethionine (CDE) have an alkaline phosphatase (ALKP) isozyme which can be distinguished by its electrophoretic mobility from the enzyme present in parenchymal cells isolated from normal liver or livers of rats fed the CDE diet for 4 weeks. The oval cell ALKP has the same electrophoretic mobility as the enzyme from fetal rat liver and placenta. ALKPs from oval cells, parenchymal cells, and placenta all differ from the intestinal enzyme by their electrophoretic mobility, isoelectric focusing, and the patterns of amino acid inhi-
bition of enzyme activity. Oval cells in preneoplastic livers, fetal hepatocytes, and tumor cells of a primary hepatocellular carcinoma induced by CDE feeding stained with a monoclonal antibody directed against rat placental ALKP. Hepatocytes (in normal or preneoplastic livers) and bile duct cells in normal liver did not stain with the same antibody. Placental ALKP may thus be a useful marker in tracing the origin and fate ofoval cells during hepatocarcinogenesis. (Am J Pathol 1985, 121:505-513)
THE PROLIFERATION of small, oval-shaped cells with scant cytoplasm is a common event in liver carcinogenesis in experimental animals. Although morphologically similar to biliary epithelial cells, the origin of oval cells and their role in carcinogenesis, if any, are not clearly understood. The morphologic identification of these cells by light and electron microscopy has been a recurrent theme in studies of hepatic carcinogenesis for more than 25 years.`3 The similarity in the oval-cell response during liver carcinogenesis induced in rats by ethionine, 2-acetylaminofluorene, and 3'-methyl-4-dimethylaminoazobenzene has been pointed out.4 Since the original proposal by Wilson and Leduc5 that cells from bile ductules (cholangioles) might be "considered reserve cells of a primitive type capable of differentiating into either parenchymal or bile duct cells," there has been much interest in characterizing oval cells, comparing them with cells of the biliary epithelium and hepatocytes, and determining their developmental potential.6 These studies have been greatly aided by the discovery that oval cells (or subpopulations) contain a-fetoprotein (AFP) and albumin,7'10 by the development of techniques for the isolation of non-
parenchymal cell populations from preneoplastic livers,"1-"3 and by the availability of suitable culture methods. 14-15 Oval cells isolated in our laboratory by centrifugal elutriation from livers of rats fed a choline-deficient diet containing 0.17o ethionine9 have albumin and AFP mRNAs, and 30% and 600/o of the cells are positive for AFP and albumin, respectively, by immunoperoxidase staining. 12 These cells contain the fetal liver isozymes aldolase A and pyruvate kinase K in addition to fetal-adult hybrids for each enzyme. 16 In this paper, we demonstrate that oval cells and cells of a primary rat hepatocellular carcinoma, but not biliary or parenchymal cells from normal liver, contain an isozyme of alkaline phosphatase (ALKP) similar to the placental enzyme. Various ALKP (EC 3.1.3.1) isozymes and variants Supported by NIH Grant CA 23226 from the National Cancer Institute. Accepted for publication July 17, 1985. Address reprint requests to Nelson Fausto, MD, Professor and Chairman, Department of Pathology and Laboratory Medicine, Brown University, Providence, RI 02912.
505
506
YASWEN ET AL
have been described in normal and cancerous tissues, both in rats and in humans, and are valuable markers for neoplasia.1718 The different isozymes are distinguishable by differences in electrophoretic mobility, sensitivity to inhibition by amino acids, thermal stability, digestion of carbohydrate moieties by neuraminidase, pH optima, as well as immunologic identity. In the human, ALKP isozymes of placental type are often expressed in tumors at various sites.19 In the rat, both intestinal and placental ALKP have been detected by immunohistochemistry and electrophoresis in hepatomas induced by 3'-methyl-4-dimethylaminoazobenzene.2.21 Although ALKP activity has been demonstrated in oval and biliary duct cells by biochemical and histochemical methods,11'3 no studies of the ALKP isozyme composition of these cells during hepatocarcinogenesis have been reported.
Materials and Methods Liver Cell Isolation The following cell fractions were isolated from the livers of male Sprague-Dawley rats (130-150 g, Charles River Breeding Laboratories, Wilmington, Mass): 1) parenchymal cells and bile-duct-enriched cells from normal livers; 2) parencymal cells and/or oval cells from livers of rats fed for 4-16 weeks a choline-deficient diet containing 0.1I7o ethionine9 (CDE diet); 3) bile-duct-enriched cells from the livers of rats 2-6 weeks after ligation of the common bile duct. The methods used for these separations, which include ezymatic cell dissociation and centrifugal elutriation, have been described in detail previously.12 The following minor modifications were introduced: 1) to maximize the yields of oval cells from preneoplastic livers and of bile duct cells from bile-duct-ligated livers, we pooled the elutriation fractions most enriched in these cell types (3, 4, and 5) and used them for enzyme assays and electrophoretic separation of isozymes; 2) to increase the purity of hepatocytes obtained from preneoplastic livers, we maintained the elutriation rotor speed at 1000 rpm and collected cell fractions at counterflow rates of 15, 20, 25, 30, and 40 ml/min. Fractions collected at 20-40 ml/min were pooled. Cell suspensions were washed in 0.9qo NaCl solution for removal of any residual calf serum and were centrifuged at 300g for 10 minutes at 4 C. The cell pellets were frozen at -70 C prior to butanol extraction. For purposes of isozyme comparison, the following tissues were obtained from freshly killed rats, immediately frozen in liquid nitrogen, and stored at -70 C: adult rat liver and intestine, 17-day fetal liver, and term placenta.
AJP * December 1985
Enzyme Activity Determination and Isozyme Separation
Cells or tissues were homogenized in 5 volumes (wt/vol) of 0.25 M sucrose followed by extraction with 1.7 volumes of butanol.22 These mixtures were centrifuged at 30,000g for 15 minutes at 2 C, after which the aqueous phases were aspirated from under the butanol layers and dialyzed for 2 days at 4 C in 0.01 M Tris-HCl buffer, pH 7.5, containing 0.2 mM MgCl2. ALKP activity was assayed in the supernatants by adding 0.1 ml of each appropriately diluted supernatant to 0.5 ml of a standard assay solution (8 mM pnitrophenyl phosphate [Sigma Chemical Co., St. Louis, Mo], 0.1 mM MgCl2, 0.5 M 2-amino-2-methyl-1,3,propanediol [Sigma] adjusted to pH 10.2) and measuring at 37 C the change in absorbance at 205 nm in a continuously recording spectrophotometer.23 One unit of activity was defined as the amount causing the release of 1 Mmol of p-nitrophenol from p-nitrophenyl phosphate per minute at 37 C. Ten millimolar Lphenylalanine (National Biochemicals Co., Cleveland, OH), 10 mM L-leucine (National Biochemicals Co.) or 10 mM L-homoarginine (Sigma) was added to the standard assay solution for inhibition studies. The protein concentration of enzyme extracts was determined by the method of Bradford.24 Supernatants containing 0.001-0.01 unit of ALKP activity were analyzed by nondenaturing polyacrylamide gel electrophoresis in either 5% or 7% vertical slab gels.25 The sample loading buffer and the gels contained 0.05 o Triton X-100 (Sigma) to facilitate the migration of ALKP enzymes. The positions of the ALKP isozymes were localized in gels using a-naphthyl phosphate (Sigma) as the substrate and 4-aminodiphenylamine diazonium sulfate (Sigma) as the coupling reagent.26 Some supernatants were treated with 1.0 U/ml neuraminidase (Type V, Sigma) at 30 C for 18 hours prior to electrophoresis.27 Isoelectric focusing of ALKP isozymes was performed overnight at 400 volts and 2 C in polyacrylamide gel discs (4% acrylamide, 27o ampholyte, pH 3-10 [Pharmacia, Piscataway, NJ], 5% glycerol, 0.05% Triton X-100, 0.14%o TEMED [Bio-Rad, Richmond, Calif], 0.02% ammonium persulfate [Bio-Rad]).28 The gels were extruded into staining tubes and stained for ALKP activity as described above. Cell Localization of Placental ALKP Placental ALKP was localized by the indirect immunoperoxidase technique in formalin-fixed, paraffinembedded tissue sections.29 A mouse IgM monoclonal
PLACENTAL ALKALINE PHOSPHATASE IN OVAL CELLS
Vol. 121 * No. 3
antibody (3A2) against rat placental ALKP was kindly provided by Dr. Y. Yoshida.21 This antibody was pretested and titrated on sections of rat placenta. For each slide, a corresponding tissue section was incubated with nonimmune mouse serum at the same dilution as the antibody for assessment of the degree of background staining with the peroxidase technique. The tissue sections used for study were normal adult rat liver, 17-day fetal liver, livers of rats fed the CDE diet for 6 weeks, and liver of an animal which developed hepatocytic tumors after receiving the CDE diet for 35 weeks. The tumor nodules were approximately 1-2 cm in diameter and consisted of differentiated hepatocytic cells encircled on the periphery of the nodule by oval cells.
Results Electrophoretic Mobility and Isoelectric Focusing Butanol extracts of whole adult and fetal rat liver and of isolated liver cells were analyzed by electrophoresis in polyacrylamide slab gels followed by staining for ALKP activity (Figure 1). For comparison of the electrophoretic mobility of various ALKP isozymes, extracts of adult rat intestine and term placenta were loaded on the same gels. Figure la shows that extracts of normal adult liver (NL) and of hepatocytes isolated from these livers (PCN) contained one major, sharply defined, ALKP band. Occasionally, one or two rapidly migrating minor bands were also detected in these ex-
507
tracts. The major hepatocyte band migrated more rap-
idly than the placental isozyme (PL) and did not coincide with either of the two bands detected in intestinal extracts (I). The ALKP isozyme pattern of normal hepatocytes was unchanged in hepatocytes isolated from livers of rats fed for 2 (PC, 2 CDE) or 4 weeks (PC, 4 CDE) a choline-deficient diet containing 0.1 % ethionine. The major ALKP band in extracts of oval cells isolated at 4 (OC, 4 CDE), 8 (OC, 8 CDE), and 16 weeks (OC, 16 CDE) of the CDE diet migrated slightly slower than the hepatocyte band and coincided with the ALKP band detected in extracts of fetal liver (FL, 17 days of gestation) and placenta (PL, 20 days of gestation). Surprisingly, the extract from a bile-duct-cell-enriched fraction prepared from livers of rats 2 weeks after ligation of the common bile duct, exhibited a band of ALKP activity with slower mobility than either the adult hepatocyte or oval-cell enzyme. The difference in mobility between adult hepatocyte and oval-cell ALKP is small but reproducible and is not due to variations in the preparation of extracts, as demonstrated in Figure lb. An oval-cell extract combined with a hepatocyte preparation at equivalent amounts of enzyme activity (OC + PC) exhibited ALKP mobilities corresponding to each individual cell type (OC and PC). On isoelectric focusing, however, extracts of placenta, normal hepatocytes, and oval cells and hepatocytes isolated from livers of CDE-fed rats shared a common pattern of bands of ALKP activity (Figure 2). This common electrofocusing pattern differs
a
b I
Pl
NL
PC N
PC
PC
OC
OC
OC
2CDE 4CDE 4CDE 8CI 16CDE
FL
BDC 2BWL
OC
OC
+PC
PC
w __
_ j
U Vl
_i
_. _
Ei,
......
;
Mi.
Figure 1 -Polyacrylamide slab gel electrophoresis of alkaline phosphatase in butanol extracts of rat tissues and isolated liver cell populations. aNormal adult intestine homogenate (/), 20-day placenta homogenate (PL), normal adult liver homogenate (NL), isolated parenchymal cells from normal adult liver (PC-N), isolated parenchymal cells from CDE liver at 2 weeks (PC-2CDE) and 4 weeks (PC-4CDE) of feeding, isolated oval cells from CDE liver at 4 (OC-4CDE), 8 (OC-8CDE), and 16 (OC-16CDE) weeks of feeding, 17-day fetal liver homogenate (FL), and isolated bile duct cells from 2-week bile-duct-ligated liver (BDC-2BDL). b-Isolated oval cells from CDE liver at 4 weeks of feeding (OC), isolated parenchymal cells from CDE liver at 2 weeks of feeding (PC), and combined extracts (OC + PC). ALKP activity was localized by histochemical staining in the gels after electrophoresis.
508
YASWEN ET AL
Pl OC PC PC I CDE CDE N
Figure 2-Isoelectric focusing of alkaline phosphatase in polyacrylamide disc gels. Butanol extracts of 20.day placenta homogenate (PL), oval cells isolated from CDE livers at 4 weeks (OC-CDE) of feeding, parenchymal cells isolated from CDE liver at 4 weeks of feeding (PC-CDE), parenchymal cells from normal adult liver (PC-N), and normal adult intestine homogenate (/) were electrophoresed overnight in 4% polyacrylamide disc gels containing 2% ampholyte (pH 3-10) and 0.05% Triton X-100 as described. ALKP activity was localized by histochemical staining in the gels after electrophoresis.
from that of the intestinal enzyme in that the ALKP doublets detected in liver and placental extracts are not detected in the intestinal preparations. Neuraminidase Treatment and Amino Acid Inhibition of ALKP Activity Treatment of rat serum or tissues with neuraminidase causes a change in the electrophoretic migration
AJP * December 1985
of bone, liver, kidney, and serum ALKP but does not affect the intestinal enzyme.30 Figure 3 shows that neuraminidase treatment slowed the anodal electrophoretic mobility of ALKP of normal parenchymal cells (PCN), parenchymal (PC, 4CDE) and oval cells (OC, 4CDE) isolated at 4 weeks of feeding the CDE diet, and the placental enzyme (PL), but did not alter the mobility of intestinal ALKP (I). Although neuraminidase treatment under these conditions did not appear to abolish the relative differences in the electrophoretic pattern of oval cell and parenchymal cell enzymes, no firm conclusions can be reached on this point because oval-cell extracts treated with neuraminidase generally display a very disperse banding pattern of ALKP activity in polyacrylamide gels. Different ALKP isozymes are inhibited to varying degrees by specific amino acids.30 Table 1 shows that L-phenylalanine reduced the intestinal ALKP activity by 50(o, but that L-homoarginine had no effect. In contrast, placental ALKP activity was markedly (87%o) inhibited by L-homoarginine but only moderately (25 07o inhibition) reduced by L-phenylalanine. Homoarginine markedly inhibited ALKP activity (approximately 80-90O7o) in placenta, fetal liver, cells isolated from 2and 4-week CDE liver, and bile duct cells. Approximately 67%o inhibition by homoarginine was noted in normal hepatocytes and in oval cells from 16-week CDE liver. Inhibition by phenylalanine in all isolated liver cell preparations was similar (21-33Vo) to that of the placental enzyme (2507o) but different from the intestinal form (5207o). Thus, the ALKP band detected in oval cells differs in many properties from the intestinal enzyme but is similar to the placental enzyme in electrophoretic mobility, neuraminidase sensitivity, and amino acid inhibition patterns. The ALKP from hepatocytes can be distinguished from the oval cell and placental enzyme only by its electrophoretic mobility but not by amino acid inhibition, sensitivity to neuraminidase, or banding by isoelectric focusing. Tissue Localization With Monoclonal Antibody To determine whether oval cell ALKP could be distinguished from the hepatocyte enzyme on the basis of staining with an anti-placental ALKP monoclonal antibody, we incubated formalin-fixed liver sections with a mouse IgM anti-rat placental ALKP monoclonal antibody (3A2)21 and stained them with peroxidaseconjugated anti-mouse IgM. Normal liver showed no positive reaction product except for an occasional cell lining a bile duct (Figure 4a). Hepatocytes in livers of 17-day fetal rats showed intense staining with mono-
PLACENTAL ALKALINE PHOSPHATASE IN OVAL CELLS
Vol. 121 * No. 3
C
PC N
PL
I N
N
C
c
N
PC 4CDE C .:*
Figure 3-Effect of neuraminidase on the electrophoretic mobility of alkaline phosphatase from rat tissues and isolated liver cells. Abbreviations are identical to those in Figure 2. Extracts were incubated at pH 5.6 for 18 hours at 30 C with (N) or without (C) neuraminidase (1.0 U/ml) before electrophoresis.
509
OC
N
4CDE C
N
Ilmom_oi
..
+
clonal antibody, in agreement with Yoshida et al21 (Figure 4b). In livers of rats maintained on the CDE diet for 4-6 weeks, many, but not all, oval cells and most bile ductules showed positive reaction product in their cytoplasm as well as their cell membranes (Figures 4c and 5b), but hepatocytes were negative. In the liver of a rat which developed hepatocellular carcinomas, most tumor cells stained with the 3A2 antibody, as did some cells lining atypical bile ducts in areas of cholangiofibrosis which surrounded the tumor nodules (Figure 4d). Thus, ALKP of oval cells could be discriminated from that of hepatocytes and bile ducts in normal livers and the enzyme of hepatocytes in preneoplastic livers by the staining with the 3A2 monoclonal antibody to rat placental ALKP. Yoshida et aV1 have noted that the staining with the 3A2 antibody is located mainly in the cytoplasm of fetal hepatocytes and hepatoma cells; whereas, with the conventional histochemical method, the enzyme reaction product is located in cannalicular membranes. They suggested that the monoclonal antibody is directed against a determinant present on catalytically inactive enzyme molecules which may be located in the cytoplasm. Sasaki and Fishman31 have shown that in human ovarian cancer cells the product of the histochemical reaction for ALKP is located in the cytoplasm if these cells contained the Regan isozyme (placental). In cases where the placental isozyme was not present, the product of the reaction was confined to the cell membrane. With the 3A2 antibody, we detected reaction product primarily in the cytoplasm of oval cells, tumor cells, and fetal hepatocytes.
Table 1-Amino Acid Inhibition of Alkaline Phosphatase Activity in Rat Tissue and Isolated Liver Cells % Inhibition* Amino Acid (10 mM L-Leu L-Phet L-Hom final concentration) Tissue Intestine (normal) (n = 3)t Placenta (term) (n = 4) Liver (normal) (n = 3) Liver (17 day fetal) (n = 4) Isolated liver cells Hepatocytes (normal) (n = 7) Hepatocytes (2CDE)t
(n
=
=
56.2 ± 1.9
6.8 ± 8.4
25.1 + 3.1
47.6 ± 7.9
86.6 ± 1.8
32.5 ± 4.5 24.3 ± 5.0
49.5 ± 0.8 45.8 ± 3.2
77.2 ± 2.6 91.6 ± 2.4
32.2 + 3.5
60.5 ± 13.1
66.6 ± 4.7
22.6 ± 3.3
45.9 ± 1.9
82.7 ± 1.6
33.3 ± 4.8
52.1 ± 3.2
78.3 + 3.6
22.1 ± 10.0
59.2 ± 11.6
83.6 ± 3.8
20.7 ± 6.5
45.2 ± 3.7
82.7 ± 3.8
32.2 ± 4.9
46.9 ± 6.7
67.4 ± 7.0
28.1 ± 1.5
46.1 ± 2.8
85.4 ± 4.4
3)
Hepatocytes (4CDE) (n = 4) Oval cells (4CDE)
(n
52.3 + 2.4
3)
Oval cells (8CDE) (n = 4) Oval cells (16CDE) (n = 4) Bile duct enriched cells§ (n = 4)
Values are expressed as the mean ± standard deviaton of enzyme activity inhibited compared with untreated controls. Alkaline phosphatasespecific activities of the preparations were as follows: intestine, 3.52; placenta, 0.54; normal liver, 0.03; fetal liver, 0.09; 2CDE hepatocytes, 0.04; 4CDE hepatocytes, 0.08; 4CDE oval cells, 0.18; 8CDE oval cells, 0.43; 16CDE oval cells, 0.14; bile duct enriched cells, 0.30. t Abbreviations: L-Phe, L-phenylalanine; L-Leu, L-leucine; L-Hom, Lhomoarginine; (x)CDE, cells isolated from livers of rats fed for x number of weeks a choline-deficient diet containing 0.1% DL-ethionine. t n, number of determinations. § Cells isolated from livers of rats with oval cell isolation protocol 2 weeks after bile duct ligation.
t-
Xcnu I7 00
_a. Co CL E CWC
0 CD CDC
C0'
>0
.5.3 co
E
3
c
0
n
W
_
"-.
W
co (D 0
Zv@ co60
.0 cc
E CL
o3 01 C
g3E Cm ac ) c%Jc CnI
.C co~~:
WoEm C a-., Z
c0$ I-"
*.A
0
in 2'.
.'
E0C C
E o.Cn E
V
C .t
-.r
.a
.e -
gX C
,
CD
>n7=, oE
'i
, 05eq
E
.
C
.-
co
ow;WL
*W U)
0 =
E. wUJ
;C u.ZD
E
Y C Co
PLACENTAL ALKALINE PHOSPHATASE IN OVAL CELLS
Vol. 121 * No. 3
511
Figure 5-Formalin-fixed sections of 6-week CDE liver stained with hematoxylin and eosin (a) and stained with mouse monoclonal antibody 3A2 against rat placental ALKP and lightly counterstained with methyl green (b). Hepatocytes are negative for placental ALKP (large arrows); clumps of oval cells are positive (small arrow). (Original magnification, x 500)
Discussion Although a classification of rat ALKP isozymes comparable to that of the human enzyme17 has not yet been established, it is clear that enzymes found in adult rat liver, intestine, and placenta differ from each other. The intestinal enzyme can be distinguished from the placental and adult rat liver enzymes by its electrophoretic mobility, lack of sensitivity to neuraminidase digestion, and inhibition by L-homoarginine. The enzyme present in adult liver parenchymal cells differs from the placental enzyme or the enzyme forms detected in fetal liver and oval cells by its electrophoretic mobility. The rat placental enzyme could not be distinguished from the ALKP forms present in fetal liver oval cells or tumors by any of these criteria. Yoshida et all' have prepared monoclonal antibodies against rat placental ALKP and shown by immunohistochemical staining that cells of fetal liver and azo-dye-induced liver tumors stained by the placental antibody. Azo-dye carcinogenesis (similarly to the CDE regimen) induces early, widespread proliferation of oval cells.7 Since oval cells are thought to give rise to hepatocytes during azo-dye carcinogenesis,32 it was of interest to determine 1) whether oval-cell populations might contain variant ALKP isozymes in preneoplas-
tic livers of rats fed the CDE diet and 2) whether similar forms would be found in the tumors induced by this diet. Electrophoretic analysis of ALKP in isolated oval cells and hepatocytes from livers of CDE-treated animals indicates that oval cells contain an ALKP form which has the same mobility as that of term placenta and 17-day fetal liver, whereas hepatocytes isolated up to 4 weeks of feeding the CDE diet retain the ALKP isozyme pattern of normal liver. Staining with a monoclonal antibody against the rat placental enzyme showed that a variable proportion of oval cells and most cells of a primary hepatocellular tumor induced by the CDE diet contain an ALKP form, recognizable by this antibody, which is absent from normal hepatocytes and normal bile duct cells. Very few hepatocytes were found to stain with the placental antibody in liver sections of rats fed the CDE diet for up to 6 weeks. However, we do not know whether hepatocytes would stain with this antibody at later stages of carcinogenesis. It is of interest that the monoclonal antibody against placental ALKP prepared by Yoshida et al stains oval cells and tumor cells but not the bile duct cells of normal liver (staining of a few individual cells lining bile ducts in normal liver was occasionally seen). We cannot exclude the possibility that a very small proportion of cells in the normal biliary epithelium, such as those
512
YASWEN ET AL
lining some cholangioles or some "connective tissue" cells located near bile ducts, stain for placental ALKP. Wootton et al33 have shown that ALKP from parenchymal cells migrated differently from that of extracts of biliary tract preparations. Preliminary results indicate that the 3A2 antibody also stains some cells lining bile ducts or ductules which proliferate after ligation of the common bile duct in rats. The detection of placental alkaline phosphatase in preneoplastic liver and a primary hepatocellular tumor is an interesting example of oncodevelopmental gene expression. It is of significance that when the placental enzyme found in fetal hepatocytes reemerges in preneoplastic liver, it is detected in oval cells but not in hepatocytes (at least up to 6 weeks from the start of the feeding regimen). It remains to be established whether the presence of placental alkaline phosphatase in oval cells is an indication that these cells differentiate into hepatocytes, are arrested in their differentiation, or differentiate into atypical bile duct cells. Tatematsu et al.,34 have shown recently that oval cells may undergo intestinal metaplasia and play a role in the development of cholangiofibrosis in the liver of rats fed 2-acetylaminofluorene and partially hepatectomized. We did not detect the presence of intestinal ALKP in preneoplastic liver by biochemical methods, but cell localization studies with a monoclonal antibody against the intestinal enzyme will be necessary to clarify this question. Yoshida et al20'21 had detected the presence of both placental and intestinal ALKP isozymes in rat hepatomas induced by 3'-methyl-4-dimethylaminoazobenzene, and it is quite possible that atypical bile ducts in areas of cholangiofibrosis in livers of animals fed the CDE diet may contain the intestinal enzyme. Although the biologic function of the placental ALKP in rat liver is unknown, it may be possible to track the origin and fate35'36 of a subset of cells with this oncodevelopmental marker during preneoplasia and establish whether such cells have the characteristics of transitional cells that may be precursors of normal or transformed hepatocytes.
References 1. Grisham JW, Hartroft WS: Morphologic identification by electron microscopy of "oval" cells in experimental hepatic degeneration. Lab Invest 1961, 10:317-332 2. Lombardi B: On the nature, properties and significance of oval cells, Recent Trends in Chemical Carcinogenesis. Vol 1. Edited by P Pani, F Feo, A Columbano. Cagliari, ESA, 1982, pp 37-56 3. Sell S, Leffert HL: An evaluation of cellular lineages in the pathogenesis of experimental hepatocellular carcinoma. Hepatology 1982, 2:77-86 4. Farber E: Similarities in the sequence of early histological changes induced in the liver of the rat by ethionine,
AJP * December 1985
5. 6.
7.
8.
9. 10.
11.
12.
13.
14. 15.
16.
17.
18. 19.
20.
21. 22.
23. 24.
2-acetylaminofluorene, and 3'-methyl-4-dimethylaminoazobenzene. Cancer Res 1956, 16:142-148 Wilson JW, Leduc EH: Role of cholangioles in restoration of the liver of the mouse after dietary injury. J Pathol Bacteriol 1958, 76:441-449 Grisham JW, Thal SB, Nagel A: Cellular derivation of continuosly cultured epithelial cells from normal rat liver. Gene Expression and Carcinogenesis in Cultured Liver, Edited by LE Gerchenson, EB Thompson. New York, Academic Press, 1975, pp 1-23 Dempo K, Chisaka N, Yoshida Y, Kaneko A, Onoe T: Immunofluorescent study of alpha-fetoprotein-producing cells in the early stage of 3'-methyl-4-dimethyl-aminoazobenzene. Cancer Res 1975, 35:1282-1287 Kuhlmann WD: Localization of alpha-fetoprotein and DNA synthesis in liver cell populations during experimental hepatocarcinogenesis in rats. Int J Cancer 1978, 21:368-380 Shinozuka H, Lombardi B, Sell S, Iammarino RM: Early histological and functional alterations of ethionine liver carcinogenesis in rats fed a choline-deficient diet. Cancer Res 1978, 38:1092-1098 Guillouzo A, Belanger L, Beaumont C, Valet J-P, Briggs R, Chiu J-F: Cellular and subcellular immunolocalization of alpha-fetoprotein and albumin in rat liver: Reevaluation of various experimental conditions. J Histochem Cytochem 1978, 26:948-959 Sells MA, Katyal SL, Shinozuka H, Estes LW, Sell S, Lombardi B: Isolation of oval cells and transitional cells from the livers of rats fed the carcinogen DL-ethionine. JNCI 1981, 66:355-362 Yaswen P, Hayner NT, Fausto N: Isolation of oval cells by centrifugal elutriation and comparison with other cell types purified from normal and preneoplastic livers. Cancer Res 1984, 44:324-331 Sirica AE, Cihla HP: Isolation and characterizations of oval and hyperplastic bile ductular cell-enriched populations from the livers of carcinogen and noncarcinogentreated rats. Cancer Res 1984, 44:3454-3466 Grisham JW: Cell types in long-term propagable cultures of rat liver. NY Acad Sci 1980, 349:128-137 Tsao M-S, Grisham JW, Nelson KG, Smith JD: Phenotypic and karyotypic changes induced in cultured rat hepatic epithelial cells that express the "oval" cell phenotype by exposure to N-methyl-N'-nitro-N-nitrosoguanidine. Am J Pathol 1985, 118:306-315 Hayner NT, Braun L, Yaswen P, Brooks M, Fausto N: Isozyme profiles of oval cells, parenchymal cells and biliary cells isolated by centrifugal elutriation from normal and preneoplastic livers. Cancer Res 1984, 44:332-338 Fishman WH: Perspectives on alkaline phosphatase isoenzymes. Am J Med 1974, 56:617-650 Moss DW: Alkaline phosphatase isoenzymes. Clin Chem 1982, 28:2007-2016 Ibsen KH, Fishman WH: Developmental gene expression in cancer. Biochem Biophys Acta 1979, 560:243-280 Yoshida Y, Kaneko A, Chisaka N, Onoe T: Appearance of intestinal type of tumor cells in hepatoma tissue induced by 3'-methyl-4-dimethylaminoazobenzene. Cancer Res 1978, 38:2753-2758 Yoshida Y, Hattori A, Kaneko A, Mori M: Monoclonal antibody against placental alkaline phosphatase. Tumor Res 1983, 18:S57-S66 Ohkubo A, Langerman N, Kaplan MM: Rat liver alkaline phosphatase. J Biol Chem 1974, 249:7174-7180 Nakasaki H, Matsushima T, Sato S, Kawachi T: Purification and properties of alkaline phosphatase from the mucosa of rat small intestine. J Biochem 1979, 86:1225-1231 Bradford MA: A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the
Vol. 121 * No. 3
25.
26.
27.
28. 29. 30. 31.
32.
PLACENTAL ALKALINE PHOSPHATASE IN OVAL CELLS
principle of protein-dye binding. Anal Biochem 1976, 72:248-254 Green S, Cantor F, Inglis NR, Fishman WH: Normal serum alkaline phosphatase isoenzymes examined by acrylamide and starch gel electrophoresis and by isoenzyme analysis using organ-specific inhibitors. Am J Clin Pathol 1972, 57:52-64 Fishman L: Acrylamide disc gel electrophoresis of alkaline phosphatase of human tissues, serum and ascites fluid using Triton X-100 in the sample and the gel matrix. Biochem Med 1974, 9:309-315 Righetti AB-B, Kaplan MM: The origin of the serum alkaline phosphatase in normal rats. Biochem Biophys Acta 1971, 230:504-509 Angellis D, Inglis NR, Fishman WH: Isoelectric focusing of alkaline phosphatase isoenzymes in polyacrylamide gels. Am J Clin Pathol 1976, 66:929-934 Sternberger LA: Immunohistochemistry. 2nd edition. New York, John Wiley & Sons, 1979, pp 104-169 Otani R, Okochi T, Higashino K, Ito F, Miyamoto M: Alkaline phosphatase isoenzyme of colonic carcinoma in Wistar-Furth rats. Oncodev Biol Med 1982, 3:255-268 Sasaki M, Fishman WH: Ultrastructural studies on Regan and non-Regan isoenzymes of alkaline phosphatase in human ovarian cancer cells. Cancer Res 1973, 33:3008-3018 Ogawa H, Minase T, Onoe T: Demonstration of glucose6-phosphatase activity in the oval cells of rat liver and
33. 34.
35.
36.
513
the significance of the oval cells in azo dye carcinogenesis. Cancer Res 1974, 34:3379-3386 Wootton AM, Neale G, Moss DW: Some properties of alkaline-phosphatases in parenchymal and biliary tract cells separated from rat liver. Clin Chim Acta 1975, 61:183-190 Tatematsu M, Kaku T, Medline A, Farber E: Intestinal metaplasia as a common option of oval cells in relation to cholangiofibrosis in liver of rats exposed to 2-acetylaminofluorene. Lab Invest 1985, 52:354-362 Tatematsu M, Ho RH, Kaku T, Ekem JK, Farber E: Studies on the proliferation and fate of oval cells in the livers of rats treated with 2-acetylaminofluorene and partial hepatectomy. Am J Pathol 1984, 114:418-430 Sell S, Salman J: Light-and electron-microscopic autoradiographic analysis of proliferating cells during the early stages of chemical hepatocarcinogenesis in the rat induced by feeding N-2-fluorenylacetamide in a choline-deficient diet. Am J Pathol 1984, 114:287-300
Acknowledgments We thank Dr. Y. Yoshida for his very generous gift of the 3A2 antibody, Ms. Mary Brooks for help with the early experiments, and Mrs. Anna-Louise Baxter for typing the manuscript.
