Quantification of hepatocytic proliferation in the laboratory mouse

Exp Toxic Pathol1994; 46: 95-100 Gustav Fischer Verlag lena

Sandoz Phanna Ltd., Drug Safety Assessment/Toxicology, Basle, Switzerland

Quantification of hepatocytic proliferation in the laboratory mouse A comparative study using immunohistochemical detection of bromodeoxyuridine (BrdU) incorporation and proliferating cell nuclear antigen (PCNA) expression PILAR LARDELLI, ELIAS PERENTES, GABRIELE MEIER, NIEVES NAVARRO and ROBERT A. ETTLIN With 2 figures and 2 tables Received: August 16, 1993; Accepted: September 8, 1993 Address for correspondence: ELIAS PERENTES, M. D., Sandoz Pharma Ltd., Building 8811403, CH-4002 Basle, Switzerland. Phone: +41 61 4695866; Fax: +41 61 4696565. Key words: Liver, mouse; Cell kinetics; Bromodeoxyuridine; Proliferating cell nuclear antigen.

Abstract The proliferation rate in livers of 120 mice (60 males and 60 females) was analyzed by immunohistochemical detection of bromodeoxyuridine (BrdU) incorporation and proliferating cell nuclear antigen (PCNA) expression on ethanol-fixed/paraffin-embedded specimens. Mice were divided into three groups, with 20 males and 20 females in each group: mice in the first group served as controls, while mice in the second and third groups were treated with a low and a high dose, respectively, of a non-genotoxic drug candidate for 2 weeks. A dose-related increase of the proliferating hepatocyte fraction was disclosed by both immunohistochemical methods, reaching statistical significance already in the low-dose male group for BrdU incorporation and in both male and female low-dose groups for PCNA expression. A good correlation between the degree of BrdU and PCNA labeling was observed and, as expected, the percentage of PCNA expressing cells was generally higher than the percentage of BrdU-positive cells. We concluded that the detection of PCNA expression represents a reliable method for the quantification of the hepatocytic proliferating fraction in rodents and allows the use of archival material for cell kinetic investigations in toxicologic pathology.

Introduction Long term administration of non-genotoxic drugs may increase liver cell proliferation and promote neoplasia (i.e. hepatocellular adenomas and carcinomas) by a pathogenetic mechanism involving direct cytotoxicity and cell regeneration (4, 8). Therefore, the use of accurate and reproducible methods for the evaluation of cell proliferation is of major importance for the assessment of putative epigenetic tumorigenicity of newly synthesized compounds.

Bromodeoxyuridine (BrdU) incorporation has been established as a reliable method for the analysis of cell kinetics in fresh and fixed tissues (15, 24), and has replaced former methods such as mitotic count and 3H thymidine incorporation (5, 12, 16, 17,20,23,28,36). However, the use of BrdU-anti-BrdU immunohistochemical methods requires in vivo or in vitro administration of BrdU, and retrospective studies with archival material are, therefore, not possible. Recently developed monoclonal antibodies (Mab) against the proliferating cell nuclear antigen (PCNA) (26, 35), an auxiliary protein of DNA polymerase 8, have proved to be valuable immunohistochemical markers for the identification of proliferating cells in fresh and routinely processed tissues (11, 13, 14,22, 30, 37). PCNA detection has been extensively used for cell kinetic analysis of different types of tumors in relation to histologic grading and prognosis (1, 9, 14, 18, 19,29,34) and has also been recently introduced for studying cell proliferation in toxicological assays (8, 10,27,33). In the present study, we investigated the liver cell proliferation in mice treated for two weeks with a new compound which had previously been found to induce hepatotoxic alterations. Immunohistochemistry using both anti-BrdU and anti-PCNA Mabs was performed in parallel, in order to compare the reliability of the above methods and to determine the value ofPCNA in the analysis of cell proliferation in toxicologic pathology.

Material and methods Sixty male and 60 female mice (SPF, Hanlbm: NMRI), 12 week-old, were used in the present study. Three groups of 20 males and 20 females were dosed as follows: control animals did not receive any treatment, mice in the low dose Exp Toxic Patho146 (1994) 2 95

group were treated with the same high dose as the one used in a previous mouse carcinogenicity study with the same compound, and animals in the high dose group received a three times higher dose. Ten males and 10 females from each of the three groups were necropsied after 2 days of treatment and the remaining mice (i.e. 10 males and 10 females per group) were killed on day 15, after 14 days oftreatment. Fifty mg/kg of body weight of 5' bromo 2' deoxyuridine (Sigma, St Louis, MO, USA) in 0.9 % NaCl was administered intraperitoneally to each mouse 90 min before sacrifice. Liver necropsy specimens from all mice were fixed in 70 % chilled ethanol for 24 h at 4 DC, embedded in paraffin wax, and 4 to 5 11m-thick sections were stained with hematoxylin and eosin (H&E). Immunohistochemical reactions were carried out on non-successive sections by using the streptavidinlperoxidase method with an anti-BrdU Mab (Becton Dickinson, CA, USA, lot N1011) and the PCIO Mab (Dako Co. CA, USA, lot 071), an anti-PCNA Mab. Sections were deparaffinized in xylene and rehydrated through gradet ethanol to water and the endogenous peroxidase activity was blocked by a 30 min incubation in methanol containing 0.5 % HP2' Sections for BrdU were then incubated for 30 min in 2N HCI (Merck, Darmstadt, Germany), at room temperature (RT), in order to denature DNA, thoroughly washed in bidistilled water, and neutralized with O.1M N~BP7 (Borax) (Sigma) for 3 min. After saturation with 10 % normal goat serum (Dako, lot X907) in 0.05M Tris-buffered saline (pH 7.6) for 20 min, alternate sections from all livers were incubated for 22h at 4 DC with the anti-BrdU Mab (diluted 1:900) and the PC 10 Mab (diluted 1:5000), respectively. Primary antibodies were dissolved in Tris-buffered saline containing 1 % normal goat serum. Specimens were washed and serially incubated at RT with biotinylated goat anti-mouse immunoglobulin (Dako, lots 012 and 120) diluted 1:100 in Tris-buffered saline and with streptavidin ABC complexlHRP (Dako, lots

012 and 120), prepared according to the manufacturer's instructions. Immunohistochemical reactions were developed with 3,3' diaminobenzidine tetrahydrochloride (Amersham Labs, Amersham, UK), 5 mg in 6.5 rn1 of 0.05M Tris buffer containing 0.015 % HP2' and sections were counterstained with hematoxylin. Necropsy specimens of small intestine from the corresponding animals were stained in parallel and served as positive controls. Negative controls were obtained by omitting the primary antibodies. Evaluation and quantification methods of nuclear immuno staining were identical for both Mabs: for each mouse liver, two non-consecutive sections were examined, and a total of 3,500 cell nuclei were counted from 15 to 25 randomly selected fields (final magnification 200x). All immunoreactive cell nuclei were recorded as being positive for the respective marker, regardless of the intensity of the immunostaining. The percentage of immunostained cell nuclei (labeling index, LI) was calculated for every animal, and the LI mean value and standard deviation were determined for every group of mice, according to sex and treatment (dose; period). Statistical analyses were performed using the Kruskal-Wallis test and the Dunn test (7,21). In addition, the distribution of LI' s in the above groups of animals was graphically represented.

Results Routine histopathological evaluation of the livers, on H&E stained sections, revealed minimal to marked hypertrophy of the centrilobular hepatocytes in treated mice. After 14 days of treatment, the severity and the incidence of this change were increased in a dose-related way, in both males and females (see also fig. lA). In the 2-day-treated mice, only females showed a no dose-related increase in the incidence of hepatocellular hypertro-

Fig. 1. Liver specimens, from the high dose groups of mice treated for 14 days, immunostained for bromodeoxyuridine (BrdU: A) and proliferating cell nuclear antigen (PCNA: B) expression, lightly counterstained with hematoxylin; x 400. A: BrdU-positive hepatocytic nuclei in the hypertrophic centrilobular hepatocytes in a female (BrdU LI: 3.48). Normal sized hepatocytes are present in the periportal area (left). B: Several hepatocytic nuclei expressing PCNA positivity in a male (PCNA LI: 4.64). 96

Exp Toxic Patho146 (1994) 2

Table 1. Bromodeoxyuridine (BrdU) and proliferating cell nuclear antigen (PCNA) labeling index (LI) mean values (± standard deviations) in mice treated for two days. Groups of mice

Males

Females

Control Low dose High dose

BrdULI 0.046 ± 0.052 0.121 ± 0.082 0.145 ± 0.163

0.056 ± 0.064 0.098 ± 0.075 0.185 ± 0.203

Control Low dose High dose

PCNALI 0.072 ± 0.055 0.200 ± 0.211 0.137 ± 0.087

0.186 ± 0.130 0.181 ± 0.132 0.296 ± 0.197

Table 2. Bromodeoxyuridine (BrdU) and proliferating cell nuclear antigen (PCNA) labeling index (LI) mean values (± standard deviations) in mice treated for 14 days. Groups of mice

Males

Females

Control Low dose High dose

BrdULI 0.019 ± 0.018 0.281 ± 0.160* 0.640 ± 0.190**

0.115 ± 0.095 0.173 ± 0.079 0.413 ± 0.323**

Control Low dose High dose

PCNALI 0.068 ± 0.052 0.483 ± 0.310** 2.186 ± 1.331 **

0.180 ± 0.117 0.441 ± 0.167* 1.705 ± 0.850**

Kruskal Wallis and Dunn tests: *2p < 0.05, **2p < O.oI

phy. In addition, a subtle increase of background liver alterations, including foci of inflammation/necrosis and single hepatocytic necrosis, were observed in treated animals. BrdU and PCNA immunolabeled hepatocytic nuclei were easily and clearly identifiable (fig. lA, B). With both markers, reactivity was localized exclusively within the nuclei and the intensity of the staining exhibited some variability, especially in the case of PCNA expression. Although positive cells were generally homogeneously distributed throughout the tissue sections, a higher number of immunoreactive cell nuclei was noticed in the subcapsular and centrilobular areas (fig. 1A). Similarly, a higher density of immunostained cell nuclei was also observed adjacent to inflammatory and necrotic foci; these areas were systematically excluded from cell quantification. Slight background cytoplasmic staining was present with both Mabs, being more pronounced in BrdU stained sections. Tables 1 and 2 show the proliferating fractions, in relation to sex and treatment, as determined with BrdU incorporation and PCNA expression, respectively. BrdU LI increased in a dose related manner after 2 and 14 days of

treatment. At this latter time point, the increase of the BrdU LI was statistically significant in the low dose group of male mice, as well as in both the male and female high dose groups. Similarly, the percentage of PCNA positive cells also increased in a dose-related manner and already became statistically significant, in comparison to the controls, at 14 days of treatment in the low dose group, in both male and female mice. In the control group, the percentage of immunostained cell nuclei was higher in females than in males with both markers. This difference was less marked in the 2-day-treated mice and, in the 14-day-treated animals; male mice presented a larger number of positive cell nuclei than females, although statistical significance was not achieved (data not shown). As can be appreciated in figure 2A-D, in all groups, except in the 2-day high dose male mice, the PCNA LI was higher than the BrdU LI. This finding was especially evident in the 14-day-treated mice, where an almost linear correlation was observed between BrdU labeling and PCNA expression for both male (R=0.823) and female (R=0.746) mice (fig. 2E and F).

Discussion Results of the present study further confirm the value of PCNA immunolabeling for the detection of proliferating cells in routinely processed tissues. Previous reports have documented the applicability of PC 10 Mab in formalin fixed material, as well as its limitations (13,19,27, 30, 37). Moreover, it has been reported that microwave incubation of formalin-fixed/paraffin-embedded material enhances the immunohistochemical staining of PCNA in archival rodent tissues (13). In the present investigation, PCI0 was applied in ethanol-fixed material in order to compare the PCNA immunoreactivity with the immunohistochemical detection of BrdU incorporation, the latter being optimally preserved in ethanol-fixed tissues (12). In agreement with previous studies recommending .the use of PC 10 in tissues fixed in ethanol or ethanol-based fixatives (11, 30, 36), we found excellent PCNA immunoreactivity in the positive controls and in the liver specimens with a Mab dilution of 1:5000. Using both BrdU and PCNA methods, less than 0.08 % and 0.20 % of proliferating hepatocytes were detected in the control male and female mice, respectively. These values increased in a dose-related way, and a trend to a higher proliferating rate was already noted in the 2-daytreated mice, although statistical significance was not achieved. Following 14 days of treatment, PCNA and BrdU LI values became statistically significant in all treated mice, with the exception of the low dose female group, where a statistically significant increase in the hepatocyte proliferation rate was detected by the PCNA LI only. As expected, the PCNA LI was higher than the BrdU LI in almost all groups of mice. The present findings are in agreement with previously reported data by others, who observed that the fraction of PCNA-positive cells Exp Toxic Pathol 46 (1994) 2 97

Females (2-Day Treatment)

Males (2-Day Treatment)

c:JBRDU

c:JBROU VDia PCNA

IigI Dose

A

VDia PCNA

Low Dose

B

Females ( 14-Day Treatment)

Males (1 4-Day Treatment)

c:JBROU VDia PCNA

c:JBROU VDia PCNA

0.4

c

Control GrOl4)

Low Dose

o

High Dose

Control GrOl4)

Males 14 Oays

5.0

PCNA vs BROU. R =0.823

High Dose

Females 14 Days

6.0

PCNA vs BROU, R =0.746

o o

Low Dose

5.0

4.0

4.0 3.0

g

~

4(

~

2.

to

1.0

0.1

E

3.0

2.

02

0.3

0.4

0.5

BROU(%)

0.6

0.7

0.8

0.9

o

0.0~-~~t4----+--t---_-.....----t

to

-02

F

0.0

02

0.4

0.6

0.8

to

12

BROU(%)

Fig. 2. A-D: Graphic representations ofbromodeoxyuridine (BrdU) and proliferating cell nuclear antigen (PCNA) labeling indices (LI' s) in relation to sex and treatment duration. Histograms represent mean labeling index values (± standard deviations). The percentage of PCNA expressing cells is higher in all groups, with the exception of the 2day high dose male group. E, F: Linear correlation between the number of BrdU- and PCNA-positive nuclei in male (E) and female (F) mice treated for 14 days with the high dose of the compound.

98

Exp Toxic Patho146 (1994) 2

was consistently higher than the fraction of BrdU labeled cells in immunostained sections (6, 36, 37). Some elements related to the biology of PCNA protein may help to better understand the reasons responsible for the presence of a higher number of PCNA immunoreactive nuclei (versus BrdU immunoreactive nuclei) in the same tissues. PCNA, also known as cycIin, is a 36 kd non-histonic nuclear protein. Its expression is closely linked to the cell cycle and it acts as a co-factor of DNA polymerase 5 (2, 25, 32). PCNA appears in the nucleus at the end of the GI-phase, its concentration becomes maximal in the S-phase, and decreases again during the Gz-phase and the M-phase (29). Fluctuations ofPCNA concentration in the different phases of the cell cycle may be responsible for the variation in the intensity of the staining observed with anti-PCNA Mabs. Since PCNA is present in the nucleus during the G1-, Gz- and M-phases, in addition to the S-phase, the PCNA LI is expected to be higher than the BrdU LI in the same tissue. Moreover, considering that the half-life of the protein has been estimated to be in the order of 20h (3), it is likely that PCNA can be immunohistochemically detectable in the cells after the Mphase. It thus appears that, while BrdU incorporation is limited exclusively to cells undergoing DNA synthesis, PCNA expressing cells correspond more strictly to the growth fraction than to the S-phase cells (11, 36). Overexpression of PCNA has been reported in different types of tumors (14,19,31,36), and has been considered to be consequent on several phenomena intrinsic to neoplastic growth. Hall and co-workers (14) reported that increased PCNA expression in a number of tumors, and in the adjacent normal tissue, could be related to PCNA gene deregulation. It was thus suggested that autocrine and paracrine growth factor-mediated regulation of PCNA expression might be responsible for the excess of PCNA immunoreactive cells in a number of neoplasms (14). In addition, overexpression of PCNA can result from the accumulation of cells arrested in the course of mitosis (6, 14). Therefore, PCNA expression does not permit discrimination between cells undergoing normal replication and slowly growing or arrested cells in the late G I, S-phase or the early G2-phase. It thus appears that, in cases of altered growth, PCNA expression might not be the marker of choice for cell proliferation (36), and a lack of correlation between PCN A expression, histologic grading and prognosis may exist in several types of cancer (9, 18, 19,22,36). Nevertheless, in a recent retrospective immunohistochemical study of normal and neoplastic rat livers, we found that the PCNA LI clearly separated liver malignancies from benign liver tumors, as well as from normal non-neoplastic liver tissues (27). In the present study, since an excellent correlation was found between the PCNA LI and the BrdU LI, especially in the 14-day-treated animals, it was assumed that the population of PCNA immunolabeled cells corresponded to the cycling fraction of hepatocytes in mouse livers. Our findings corroborate previous studies in which the PC10 Mab was found to be a reliable marker for asses-

sing cell proliferation in conditions of unperturbed growth (36). Collectively, our results indicate that PCNA immunostaining constitutes a reliable and easily reproducible method for the detection of proliferating cells in rodent liver. The major advantage of anti-PCNA antibodies, in toxicologic pathology, is that their immunohistochemical application does not require in vivo or in vitro manipulations, and retrospective studies using archival material are, therefore, possible.

References 1. ALLEGRANZA A, GIRLANDO S, ARRIGONI GL, et al.: Proliferating cell nuclear antigen expression in central nervous system neoplasms. Virchows Arch A Pathol Anat Histopathol1991; 419: 417-423. 2. BRAvoR, FRANKR, BLUNDELLPA, etal.: CyclinlPCNA is the auxiliary protein of DNA polymerase-c. Nature 1987;326: 515-517. 3. BRAVO R, MACDONALD-BRAVO H: Existence of two populations of cyclin/proliferating cell nuclear antigen during the cell cycle: association with DNA replication sites. J Cell Bioi 1987; 105: 1549-1554. 4. CLAYSON DB: International Commission for Protection against Environmental Mutagens and Carcinogens. ICPEMC publication No. 17. Can a mechanistic rationale be provided for non-genotoxic carcinogens identified in rodent bioassays? Mutat Res 1989; 221: 53-67. 5. VAN DIERENDONCK JH, KEIJZER R, VAN DE VELDE CJH, et al.: Subdivision of S-phase by analysis of nuclear 5bromodeoxyuridine staining patterns. Cytometry 1989; 10: 143-150. 6. VAN DIERENDONCK JH, WIJSMAN JH, KEIJZER R, et al.: Cell-cycle-related staining patterns of anti-proliferating cell nuclear antigen monoclonal antibodies. Comparison with BrdUrd labeling and Ki-67 staining. Am J Patho11991; 138: 1165-1172. 7. DUNNETT CW: A multiple comparison procedure for comparing several treatments with a control. J Am Stat Ass 1955; 50: 1096-1121. 8. ETTLIN RA, PERENTES E, KOLOPP M, et al.: Overview of quantitative methods in toxicologic pathology. In Vivo 1993; 7: 315-324. 9. EVANS AT, BLESSING K, ORELL JM, et al.: Mitotic indices, anti-PCNA immunostaining, and AgNORs in thick cutaneous melanomas displaying paradoxical behaviour. J Pathol 1992; 168: 15-22. 10. FOLEY J, TON T, MARONPOT RR, et al.: Proliferating cell nuclear antigen (PCNA): Comparison to tritiated thymidine ([3H]-Tdr) as a marker of proliferating hepatocytes in rats. Env Health Persp 1993; (in press). 11. GARCIA RL, COLTRERA MD, GOWN AM: Analysis of proliferative grade using anti-PCNNcyclin monoclonal antibodies in fixed, embedded tissues. Comparison with flow cytometric analysis. Am J Pathol 1989; 134: 733-739. 12. GONCHOROFF NJ, KATZMANN JA, CURRIE RM, et al.: Sphase detection with an antibody to bromodeoxyuridine. Role of DNase pretreatment. J Immunol Methods 1986; 93: 97-101. 13. GREENWELL A, FOLEY JF, MARONPOT RR: An enhanceExp Toxic Patho146 (1994) 2 99

ment method for immunohistochemical staining of proliferating cell nuclear antigen in archival rodent tissues. Cancer Lett 1991; 59: 251-256. 14. HALL PA, LEVISON DA, WOODS AL, et al.: Proliferating cell nuclear antigen (PCNA) immunolocalization in paraffin sections: an index of cell proliferation with evidence of deregulated expression in some neoplasms. J Patho11990; 162: 285-294. 15. HOSHINO T, NAGASHIMA T, MUROVIC JA, et al.: Cell kinetic studies of in situ human brain tumors with bromodeoxyuridine. Cytometry 1985; 6: 627-632. 16. HOSHINO T, NAGASHIMA T, CHO KG, et al.: S-phase fraction of human brain tumors in situ measured by uptake of bromodeoxyuridine. Int J Cancer 1986; 38: 369-374. 17. HOSHINO T, PRADOS M, WILSON CB, et al.: Prognostic implications of the bromodeoxyuridine labeling index of human gliomas. J Neurosurg 1989; 71: 335-341. 18. JAIN S, FILIPE MI, HALL PA, et al.: Prognostic value of proliferating cell nuclear antigen in gastric carcinoma. J Clin Pathol1991; 44: 655-659. 19. KARAMITOPOULOU E, PERENTES E, MELACHRINOU M, et al.: Proliferating cell nuclear antigen (PCNA) immunoreactivity in human central nervous system neoplasms. Acta Neuropathol1993; 85: 316-322. 20. KHOCHBIN S, CHABANAS A, ALBERT P, et al.: Application of bromodeoxyuridine incorporation to the determination of cell distribution within the S-phase of the cell cycle. Cytometry 1988; 9: 499-503. 21. KRUSKAL WH, WALLIS WA: Use of ranks in one-criterion analysis of variance. J Amer Stat Ass 1952; 47: 583-621. 22. MANN JM, JENNISON SH, Moss E, et al.: Assessment of rejection in orthotopic human heart transplantation using proliferating cell nuclear antigen (PCNA) as an index of cell proliferation. J Patho11992; 167: 385-389. 23. MORSTYN G, PYKE K, GARDNER J, et al.: Immunohistochemical identification of proliferating cells in organ culture using bromodeoxyuridine and a monoclonal antibody. J Histochem Cytochem 1986; 34: 697-701. 24. NAKAGAVA Y, PERENTES E: Determination by anti-bromodeoxyuridine monoclonal antibody of cells in Sphase in the human Y 79 retinoblastoma line maintained in an organ culture system: effects of butyric acid. Ophthalmic Res 1987; 19: 205-212. 25. NISHIDA C, REINHARD P, LINN S: DNA repair synthesis in human fibroblasts requires DNA polymerase 8. J BioI Chern 1988; 263: 501-510. 26. OGATA K, KURKI P, CELIS JE, et al.: Monoclonal anti-

100

Exp Toxic Patho146 (1994) 2

27.

28.

29.

30.

31.

32. 33.

34.

35.

36.

37.

bodies to a nuclear protein (PCNAIcyclin) associated with DNA replication. Exp Cell Res 1987; 168: 475-486. PERENTES E, ARNOLD J, MEIER G, et al.: Proliferating cell nuclear antigen expression in Wistar rat livers. A retrospective study of normal and neoplastic livers. Exp Toxic Pathol1994; 46: 105-110. RICCARDI A, DANOVA M, WILSON G, et al.: Cell kinetics in human malignancies studied with in vivo administration ofbromodeoxyuridine and flow cytometry. Cancer Res 1988;48:6238-6245. ROBBINS BA, DE LA VEGA D, OGATA K, et al.: Immunohistochemical detection of proliferating cell nuclear antigen in solid human malignancies. Arch Pathol Lab Med 1987; 111: 841-845. ROWLANDS DC, BROWN HE, BARBER PC, et al.: The effect of tissue fixation on immunostaining for proliferating cell nuclear antigen with the monoclonal antibody PC10. J Patho11991; 165: 356-357. SCOTT RJ, HALL PA, HALDANE JS, et al.: A comparison of immunohistochemical markers of cell proliferation with experimentally determined growth fraction. J Patho11991; 165: 173-178. So AG, DOWNEY KM: Mammalian DNA polymerase ex and 8: current status in DNA replication. Biochemistry 1988;27:4591-4595. THAETE LG, AHNEN DJ, MALKINSON AM: Proliferating cell nuclear antigen (PCNAICyclin) immunocytochemistry as a labeling index in mouse lung tissues. Cell Tissue Res 1989; 256: 167-173. TSUJI T, SHRESTHA P, YAMADA K: Proliferating cell nuclear antigen in malignant and premalignant lesions of epithelial origin in the oral cavity and the skin: an immunohistochemical study. Virchows Arch A Pathol Anat Histopathol1992; 420: 377-383. WASEEM NH, LANE DP: Monoclonal antibody analysis of the proliferating cell nuclear antigen (PCNA). Structural conservation and the detection of a nucleolar form. J Cell Sci 1990; 96: 121-129. WnSMAN JH, VAN DIERENDONCK JH, KEnzER R, et al.: Immunoreactivity of proliferating cell nuclear antigen compared with bromodeoxyuridine incorporation in normal and neoplastic rat tissue. J Pathol 1992; 168: 75-83. ZEYMER U, FISHBEIN MC, FORRESTER JS, et al.: Proliferating cell nuclear antigen immunohistochemistry in rat aorta after balloon denudation. Comparison with thymidine and bromodeoxyuridine labeling. Am J Pathol 1992; 141:685-690.