[CANCER RESEARCH 44,3752-3756, September 1984]

Nuclear Pores and DMA Ploidy in Human Bladder Carcinomas1

Bogdan Czerniak,2 Leopold G. Koss,3 and Andrew Sherman

Department of Pathology, Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York 10467

ABSTRACT

The number'of nuclear pores per sq /¿mwas determined on

the freeze-fractured nuclei of 20 human bladder tumors and five

control samples of normal bladder epithelium. Measurements ofthe nuclear surface and volume were also performed, and themean number of pores per nucleus and the ratio of pore tovolume were calculated. The DNA distribution pattern on thesame samples was determined by flow cytometry. All controlsamples and 12 tumors were diploid, and eight tumors wereaneuploid. The mean number of pores per sq urn and mean totalnumber of pores per nucleus in the control samples and in diploidtumors were similar. In the aneuploid tumors, both values weresignificantly higher. However, the ratio of pore to volume wasshown to be constant regardless of the DNA content. It wasfurther observed that, in aneuploid tumors, there are two populations of nuclei, one with density of pores similar to the diploidtumor and one with a higher pore density. Because aneuploidbladder tumors have been shown to have more aggressivebehavior than diploid tumors, increased density of nuclear poresor their total number per nucleus may be related to tumorbehavior. This view is supported by the observation that five ofeight tumors with increased density and total pore number wereinvasive, while all tumors with low pore number were noninva-

sive.

INTRODUCTION

Nuclear pores have been shown to play an important role innucleo-cytoplasmic exchange (11,18). The factors governing the

mechanisms of pore formation have been extensively investigated in experimental systems (19-22).

At this time, there is very limited information on nuclear poresin human cancer cells. It has been suggested in preliminarystudies that the density of nuclear pores may be increased inhuman cancer cells in effusions (36), in cells of malignant lym-

phoma (25), and in leukemia (9). It has also been suggested thata correlation may exist between the number of pores and biological properties of thyroid tumors (13).

In this paper, we report a quantitative study of nuclear poreson 20 human bladder tumors with appropriate controls. Thedensity of pores per sq urn of the nuclear surface, their totalnumber per nucleus, and the ratio of total number of pores tonuclear volume were established. These findings were correlatedwith 2 factors known to be of prognostic value, DNA ploidy,measured by flow cytometry, and histological grading.

MATERIALS AND METHODS

Tissue Samples. Twenty fresh, surgically removed samples of bladdertumors were the subject of this study. The tissue was obtained bytransurethral resection in 15 patients and in 5 cases from radical cystec-

tomy specimens. Samples of histologically normal urothelium from the 5resected bladders were used as controls. Using the WHO histologicalclassification (34), there were 5 tumors of Grade I, 10 tumors of GradeII, and 5 tumors of Grade III. Invasion of the bladder wall was noted in 4tumors classified as Grade III and in one tumor of Grade II.

Freeze Fracture. Small tissue fragments, approximately 1 cu mm,were fixed in 2.5% glutaraldehyde in phosphate buffer (pH 7.4) for 1 hrat 4°,washed 3 times in the same buffer, and impregnated in 20%

glycerol in the same buffer. Tissue slices were mounted on gold carriers,frozen in Freon 22, and placed in liquid nitrogen. The tissues werefractured in a Balzers Model 300 apparatus, shadowed at 45° with aplatinum-carbon electron gun, and reinforced by carbon at 90°. The

replicas were cleaned for 24 hr in 50% Clorox solution, rinsed in distilledwater, and mounted on Formovar-coated 200 mesh grids. The replicas

were examined in a JEM 10OXC electron microscope, and pictures weretaken at a constant magnification of X20.000. The nuclear pore countswere made on twice-enlarged prints (final magnification, x40,000) usinga computerized planimeter (Videoplan 2; Zeiss, West Germany4), and

the number of pores per sq ^m of nuclear surface (density of pores) wasestablished. Care was taken to eliminate areas of the nuclear membranewith obvious curvature to avoid an overestimation of the pores (3). Ineach sample, pores were counted on exposed surfaces of 20 to 30nuclei. The range of density of nuclear pores and their mean values werecalculated for each sample.

Isolation of Nuclei. The measurements of the nuclear surface andvolume and flow cytometric measurements of DNA were performed onnuclei isolated from aliquots of the same samples. The nuclei wereisolated with the use of a citrate buffer as described previously withminor modifications (35). In brief, the tissue samples were finely mincedwith scissors and incubated in a homogenization flask with 50 ml ofCASC5 at room temperature for 20 min with periodic agitation using a

vortex (Setting 4), (Fisher Scientific, Pittsburgh, PA). The homogenatewas filtered through 8 layers of gauze and centrifugea at 220 g for 10min. The crude sediment containing nuclei was resuspended in CASC,filtrated through 30-nm gauze (Nitex Nylon 8C3-30; Tetko, Inc., Elmsford,

NY), and was washed twice by resuspension in CASC, followed bycentrifugation. The resulting suspension of nuclei was kept at 4°until

used for either flow cytometry or measurements of surface and volume.Aliquots of this nuclear suspension were checked by microscopic examination with the use of phase-contrast and, after Papanicolaou staining, bright-field microscopy. Without exception, the preparation revealedwell-preserved nuclei with only occasional wisps of cytoplasm attached.

It was documented previously by biochemical methods that there is nosubstantial loss of DNA in nuclei isolated with CASC (35).

Estimation of Nuclear Surface and Volume. In order to ensurecompatibility with the freeze-fradure procedure, the isolated nuclei were

treated in the same manner as tissues subjected to freeze fracture. Thenuclei were washed 3 times in phosphate buffer (pH 7.4), fixed in 2.5%

1Supported in part by Grant 5 R01 CA32345 from the National Cancer Institute,

Department of Health and Human Services.2 Fogarty Fellow at Montefiore Medical Center, Bronx, NY.' To whom requests for reprints should be addressed, at 111 East 21Oth Street,

Bronx, NY 10467.Received January 25,1984; accepted May 24,1984.

4Videoplan 2 is a computerized planimeter with programs designed to calculate

the contours, surfaces, volumes, and particle density per unit area of microscopicobjects. For details, see Videoplan 2 Manual. Oberkochen, West Germany: Zeiss,1982.

* The abbreviations used are: CASC, citrate buffer containing 0.09 M citric acid

and 0.01 M sodium citrate (pH 2.6); PI. propidium iodide.

3752 CANCER RESEARCH VOL. 44

Research. on December 13, 2020. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from

glutaraldehyde ¡nthe same buffer for 1 hr at 4°, and subsequently

impregnated at room temperature in 20% glycerol in a phosphate buffer.A drop of the suspension of nuclei was placed on a glass slide andcoverslipped. Two hairs were placed between the coverslip and the slideto prevent squashing of the nuclei. The nuclear contours of 200 nucleiper sample were traced under a Zeiss phase-contrast microscope on

line with Videoplan 2 at a magnification of X1250, and the perimeter ofeach nucleus was computed. The nuclear surface and volume werecalculated from the nuclear perimeter approximated to a circle usingcorresponding equations for the sphere. The mean total number of poresper sample was calculated from the mean density of pores per sq firnand the mean nuclear surface. The ratio of pore to volume was calculatedfrom the mean total number of pores per nucleus and the mean nuclearvolume of the same sample.

In order to determine whether the isolation method of nuclei withCASC produced any change in the nuclear size, the mean nuclear surfaceand volume were additionally determined from tissue sections in 2samples of normal urothelium and 2 tumor samples in the followingmanner.

Fresh tissue fragments fixed in 2.5% glutaraldehyde in phosphatebuffer for 1 hr at 4° were dehydrated in a series of acetones and

embedded in Epon. Nuclear outlines were measured on 0.2-/im-thick

sections stained with toluidine blue, with the use of the light microscopeon line with Videoplan 2. In each sample, stereological reconstruction ofthe nuclear surface and volume was based on random sections through500 nuclei. The mean values thus obtained were compared with themean values of the sample obtained on isolated nuclei. The nuclearsurface areas calculated on tissue sections were from 2.6 to 3.8% higherthan those obtained on isolated nuclei. The differences between thenuclear volume obtained on tissue sections and on isolated nuclei wereless than 1%. These differences were statistically not significant.

Measurements of DNA. DNA histograms were generated by themeasurement of PI (Calbiochem-Behring Corp., La Jolla, CA) fluores

cence on isolated nuclei obtained as described above. For staining withPI, nuclei were resuspended in 1 ml of 0.1% sodium citrate containing0.05 mg PI and equilibrated for 15 min at 4°.DNA distributions were

determined by the measurement of red fluorescence at 600 nm madewith the Cytofluorograph Model 4892A (Ortho Instruments, Westwood,MA), interfaced with a MINC-11 /03 microcomputer (Digital EquipmentCorp., Maynard, MA) using a 12-bit differential analogue to digital con

verter. The reproducibility of the results was determined by generating3 histograms (each based on the measurement of 104 nuclei) for each

sample. For control of staining, instrument adjustment, and calibration,normal unstimulated human lymphocytes were used. The lymphocyteswere isolated from samples of peripheral blood from normal blood donorsby Ficoll-Hypaque gradient centrifugation (Ortho Diagnostics, Raritan,

NJ).In order to determine whether isolation of nuclei with CASC affected

the DNA distribution patterns, touch preparation smears of 2 samplesfrom the normal urothelium and 4 samples of bladder tumors werestained with the Feulgen method after fixation for 12 hr in 2% formaldehyde (12). The DNA content of 200 sequential nuclei from each of thesamples was determined by the quantitation of Feulgen absorption usinga high-resolution photometric system consisting of a Zeiss UniversalScanning Microphotometer on line with the PDP-12 computer as de

scribed previously (15). The histograms of DNA distribution obtained bythe quantitation of Feulgen absorption were compared with those generated by flow cytometry and were found to be nearly identical, as notedpreviously (35).

Determination of DNA Ploidy. The flow cytometric histograms of DNAdistribution were estimated by a computer program according to arbitrarydefinitions proposed by Tribukait et al. (33) for bladder tumors. A population of nuclei deviating less than 10% from the standard lymphocyteswas regarded as diploid. Aneuploid population of nuclei was consideredto be present when, in addition to the diploid peak, another distinct peakwas found deviating more than 10% from the diploid standard. An

Nuclear Pores and DNA Ploidy in Bladder Carcinoma

aneuploid population of nuclei was also considered to be present whenthe percentage of hyperdiploid nuclei exceeded the percentage of S +GÃŒnuclei found in normal urothelium by 3 S.D.s.

Statistical Evaluation Data. The differences between means weretested with the Student r test. The relations between the density ofpores, their total number per nucleus, ratio of pore to volume, andpercentage of nuclei beyond the 2C region of the DNA histogram weretested by linear regression analysis.

RESULTS

Classification of Bladder Tumors. According to DNA distribution patterns obtained by flow cytometry, the bladder tumorswere divided into 2 groups. In one group composed of 12samples, the histogram patterns were similar to those obtainedwith the 5 samples of normal bladder urothelium. In this groupof tumors, as in the benign samples, the main DNA peak was inthe 2C region. The proportion of nuclei with DNA content largerthan 2C was 10.5 ±2.3 (S.D.) for these tumors and 8.8 ±2.2for the normal urothelium. This difference was statistically notsignificant (p = 0.06), and thus, the tumors were classified as

diploid (Table 1). In the remaining 8 tumors, the percentage ofthe nuclei beyond the diploid region was much higher (mean,74.1 ±20.5%). The histograms of these tumors had variouspatterns of DNA distribution with additional prominent peaks inthe 3C and 4C regions, or even beyond (Chart 1). This group oftumors was classified as aneuploid (Table 2). The difference in

Talle 1Summaryof DNA measurementsin normal urothelium and diploid bladder tumors

Normal urotheliumDiploid tumorsNo.

of samples512eNuclei

in 2C region(%)91

.3 ±2.2o

89.5 ±2.3Nuclei

beyond 2Cregion(%f8.8

±2.210.5 ±2.3

a S + Gznuclei." Mean ±S.D.c Five histológica!Grade I and 7 histológica!Grade II. No invasion of bladder

wall was noted.

B

Z-,

2c Be

Relative DNA Value

Chart 1. Histograms of DNA distribution patterns based on 2 x 104Pi-stained

nuclei.A, histogram obtained from normalbladder urotheliumshowing a diploid cellpopulation. B, histogram obtained from bladder carcinoma, Grade I, showing asimilar DNA distribution pattern as in normal urothelium. The tumor was classifiedas diploid. C, histogram obtained from a Grade II bladder carcinoma with anadditionalprominentpeak in the 4C region.The tumor was classifiedas aneuploid.D, histogramof a Grade IIIcarcinomawith bladderwall invasion;2 prominentpeaksnearthe 3C and 5C regionsof the histogramare apparent.The tumor was classifiedas aneuploid.

SEPTEMBER 1984 3753

Research. on December 13, 2020. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from

8. Czemiak et al.

Table 2

Summary of DNA measurements in 8 bladder tumors classified as aneuploid

No. oí Main peaks of histogram (% of nuclei) N ^¡ h-»,-..., ->r Mainaneu-cam- MUG«! DeyOfKJ ¿(j .**! nnnula.•mri»2aS64«5*6aT68e2C62.739.229.616.93C30.538.325.54C

5C51.458.368.053.4

±7°41.2689.579.1region

(%)37.260.870.383.061.689.997.493.1|j*muspulci -

tion ofnucleiTriploidTetraploidTetraploidTetraploidTriploid,pentaploidTriploid,tetraploidTetraploidTetraploid

74.0 ±20.5

* Invasive tumors." Histological Grade II; all other tumors were Grade III.c Mean ±S.D.

proportion of the nuclei beyond the 2C region between the diploidand aneuploid groups of tumors was statistically significant (p <0.001).

All tumors classified historically as Grade I were diploid, andall Grade III tumors were aneuploid. Among the 10 Grade IItumors, 7 were diploid, and 3 were aneuploid. In the group ofaneuploid tumors, 5 of 8 were invasive (Table 2).

Density of Pores. Four various images of nuclear pores wereobserved on the nuclear membrane fracture faces, i.e., cylindricalprojections, raised rimes, cylindrical holes, and slightly raiseddomes (29). For the purpose of this study, all pores were countedtogether regardless of their various images.

The surface area of the fractured nuclei on which the poreswere counted varied from 3.7 to 12.2 sq u.m. The number ofpores counted on one fractured nucleus varied from 37 to 180.The density of pores on different nuclei varied from 3.7 to 11.3pores/sq u.m in normal urothelium and from 3.9 to 16.8 pores/sq urn in bladder tumors. The mean values of density of poresvaried from 5.9 ± 1.5 to 7.4 ± 1.7 pores/sq u.m in samplesof normal urothelium and between 5.7 ±1.9 and 12.2 ±3.2pores /sq ^m in tumor samples. When these values were correlated with the DNA ploidy, the mean densities of the pores ofnormal urothelium (7.0 ±1.8 pores/sq u.m) and diploid tumors(7.3 ±2.2 pores/sq u.m) were nearly identical (p = 0.6). On theother hand, the mean density of pores in aneuploid tumors wassignificantly higher (10 ±3.4 pores/sq u.m;p < 0.001 ).

When the density of pores was plotted as a function ofpercentage of nuclei in a given category, there was an obviousoverlap for the benign nuclei and the nuclei of diploid tumorswithin the range of 4 to 11 pores/sq u.m (Chart 2X1). In theaneuploid tumors, there were 2 populations of nuclei: one overlapping with the diploid tumors; and one with pore density from12 to 17 pores/sq /im (Chart 2B).

Total Number of Pores and Ratio of Pore to Volume. Theestimation of the total number of pores per nucleus and the ratioof pore to volume were based on the karyometric measurementsof the nuclear surface and volume on isolated nuclei. The meannuclear surface of diploid tumors (369 ±102 sq urn) was similarto the mean nuclear surface of normal urothelium (329 ±80 sqMm; p = 0.06). The mean nuclear surface of aneuploid tumors

was significantly higher (518 ± 150 sq um) (p < 0.001).The mean total number of pores in the normal urothelium was2446 ±398 pores/nucleus, and in the diploid tumors, 2692 ±

569 pores/nucleus. The difference between these values wasstatistically not significant (p = 0.1). The mean total number of

pores in aneuploid tumors (5465 ±713 pores/nucleus) washigher than in normal urothelium or diploid tumors (p < 0.001).In 2 cases of bladder tumors with predominantly tetraploidpopulations of nuclei (89.5 and 79.1% of nuclei in the 4C region;Table 2, Cases 7 and 8), the total number of pores was 5811and 5245 pores/nucleus, respectively, or approximately twicethe value for diploid tumors and normal urothelium. The densityof pores in these 2 cases was also elevated (9.7 and 9.3 pores/sq um, respectively).

The ratio of pore to volume was calculated from the totalnumber of pores per nucleus and the mean nuclear volumemeasured in the same samples. The mean nuclear volume ofaneuploid tumors (1292 ±495 cu u.m)was approximately twiceas high as that of diploid tumors (685 ±283 cu ^m) and normalurothelium (574 ±213 cu um). The ratio of pore to volume,however, was almost identical in control samples and 2 tumorgroups, i.e., 3.9 ±0.3 pores/cu urn in normal urothelium, 3.8 ±0.5 pores/cu ^m in diploid, and 4.5 ± 0.6 pores/cu pm inaneuploid tumors. The differences between these values werenot statistically significant (p = 0.6). The mean values of density

of pores, their total number per nucleus, and ratio of pore tovolume in various samples of bladder tumors and normal urothelium are shown in Chart 3. A summary of correlation between

30-,

20-

IO-

20-

IO-

4 8 I2 I6

pores/(jm2

Chart 2. Histograms of pore density distribution pattern A, normal urothelium( ) and diploid tumors ( ), based on 128 and 252 fractured nuclei,respectively. B, aneuploid bladder tumors, based on 189 fractured nuclei.

I4-I2-IO-M1

«l»l

6'4-2-7-,6-"€

5-K^

4-•S1

*•1

2-1-.:%

«:•••*°0°

*1ÜjfcfèS|z§2

$&Ao

PD PN PV

Chart 3. The mean values of pore density (PD), their total number per nucleus(PN), and pore to volume ratio (PV) in samples of normal urothelium (A) and tumorsclassified as diploid (O) or aneuploid (•).

3754 CANCER RESEARCH VOL. 44

Research. on December 13, 2020. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from

Nuclear Pores and DNA Ploidy in Bladder Carcinoma

Tabte3Comparison between nuclear pore observations and flow cytometric measurements of ONA ploidy in 20 bladder

carcinomas and 5 control samples of normal urothelium

Normal urothelium

Diploid tumors

Aneuploid tumorsSurface

ofnu-329

± 80"

369 ±102

518 ±150Volume

of nucleus (sqpm)574

±213

685 ±283

1292 ±495Density

ofpores(pores/sq7.0

±1.8

7.3 ±2.2

10.0 ±3.4Total

no. ofpores2446

±398

2692 ±569

5465 ±713to

volume(pores/cu¿¿m)3.9

±0.3

3.8 ±0.5

4.5 ±0.6a Mean ±S.D.

DNA ploidy level and nuclear pore findings is shown in Table 3.To answer the question of whether the number of pores in

bladder tumors was related to the degree of DNA ploidy, a linearregressional analysis was done between the percentage of thenuclei beyond the 2C region of the DNA histogram and thedensity of pores, their total number, and ratio of pore to volume.The density of pores and their total number per nucleus wererelated to DNA content with a statistically significant correlationcoefficient of 0.79 and 0.9, respectively. The ratio of pore tovolume remained similar regardless of the increased DNA contentof the nuclei. Since a constant ratio of pore to volume was found,additional regression analysis was done to compare the densityof pores and their total number with the mean nuclear volume.A statistically significant correlation coefficient (0.82) was obtained between nuclear volume and total number of pores pernucleus.

DISCUSSION

This study documented that the density of nuclear pores andtheir total number per nucleus were higher in aneuploid tumorsof the urinary bladder when compared with diploid tumors andnormal urothelium. No such differences were observed betweennormal urothelium and diploid tumors. As described, some of thetumors of histological Grade II were diploid, and some wereaneuploid. The pore density and total number were in keepingwith the ploidy level and not with histological grade. It was furtherobserved that, in aneuploid tumors, there are 2 populations ofnuclei: one with density of pores similar to the diploid tumorsand one with a higher pore density. Yet, the ratio of the totalpore number to nuclear volume was constant regardless of DNAploidy. Pauli ef al. (27) mention briefly that, in cells of humanbladder carcinoma, the density of nuclear pores per unit area ofnuclear membrane is generally greater than in normal urothelium.This observation, based on unpublished data, was not quanti-

tated and not correlated with the ploidy of the tumors. It is ofnote that the density of nuclear pores in normal human urotheliumreported by the same authors (8.70 ±0.67) is similar to thevalues observed by us in normal urothelium (7.0 ±1.8) and indiploid tumors (7.3 ±2.2).

It has long been postulated that nuclear pores are pathwaysfor nucleocytoplasmic exchange (11, 18). During the past 2decades, experimental evidence for the transport of materialacross the pores has been presented (1, 5-7, 26). It has been

shown that the particle size, rather than a selective filter mechanism, is the limiting factor for pore flow (7, 26). Other observations suggest, however, that changes in the permeability of porescould also play a role in the regulation of nucleocytoplasmicexchange (8).

The determinants regulating pore numbers are complex andappear to be associated with metabolic activity and differentiationof cells, or the sequence of events related to cell proliferation,I.e., the cell cycle and mitosis (for review, see Ref. 18). Inreference to metabolic activity, it has been shown that hormonalstimulation of the prostate and thyroid in experimental animalsresulted in an increase in the number of pores, while involuntion-

ary processes in these organs were associated with a decreasednumber of pores (2, 3, 17). A decreased number of pores wasalso found in isolated renal cortex nuclei from an ischemie kidney(4). In reference to cell cycle of HeLa cells, it was documentedthat the formation of pores is biphasic; pores are formed at thehighest rate during the first hr after mitosis. A second increaseoccurs shortly before the onset of S phase, and the number ofpores doubles by the end of the cell cycle, suggesting that, incycling cells, the number of pores is proportional to the DNAcontent (23, 24). An increase in the density of nuclear pores persq urn was also observed in virally transformed Syrian hamsterfibroblasts (28) and in a variety of experimental and human celllines (30). To the best of our knowledge, neither the density ofnuclear pores nor their total number was ever correlated withDNA ploidy level and biological significance of these factors inhuman tumors.

Our studies suggest that the increased density and totalnumber of pores per nucleus in aneuploid bladder tumors arerelated to DNA content. Since the ratio of pore to volume wasconstant, the volume of the nucleus may be yet another factorinfluencing the number of pores. Similar observations on theconstant ratio of pore to volume in the HeLa cell cycle wereinterpreted as suggestive that actively growing cells may requirea constant rate of transport of nuclear or cytoplasmic products(23). In the rat liver, Maul (18) observed 2 populations of nucleiwith higher and lower density of nuclear pores, correspondingto tetraploid and diploid DNA content. It is therefore reasonableto assume that the presence of 2 populations of nuclei in theaneuploid bladder tumors may also reflect higher and lower DNAcontent. This suggestion is further supported by the observationon 2 predominantly tetraploid tumors (Cases 7 and 8 in Table 2),wherein the density of pores was increased and the total numberof pores was almost the double of these values in diploid tumorsand in normal controls. Still, it must be noted that, in variousexperimental systems, the changes in the number of pores werenot thought to depend on DNA content and nuclear surface andvolume, but rather on transcriptional activity of the nucleus (20,21).

Prior studies of DNA ploidy of bladder tumors have documented that aneuploid tumors have a high proportion of cells inS phase and a more aggressive behavior than diploid tumors;the aneuploid tumors are more likely to invade and metastasize

SEPTEMBER 1984 3755

Research. on December 13, 2020. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from

B. Czerniak et al.

than diploid tumors (10, 16, 31-33).The observations recorded in our study would therefore sug

gest that increased density of nuclear pores or their total meannumber per nucleus may be related to tumor behavior. This viewis supported by the observation that 5 of 8 tumors with increaseddensity and total pore number were invasive, while all tumorswith low pore number were noninvasive.

The thesis that the number of pores can reflect some biologicalproperties of the malignant tumors receives added support fromthe study of papillary thyroid carcinomas. The low number ofnuclear pores in these tumors may reflect their slow growth andindolent biological behavior. It is of interest that these tumorsalso had a low percentage of cells in S phase and a lower DNAcontent than normal thyroid (13, 14).

Long-term clinical follow-up will be required to determinewhether the density or total pore number in bladder tumors is ofpredictive value in bladder tumors and whether or not it is directlyrelated to DNA content or constitutes yet another independentfactor of biological behavior.

REFERENCES

1. Bonner, W. M. Protein migration into nuclei. I. Frog oocyte nuclei in vivoaccumulate microinjected histories, allow entry to small proteins, and excludelarge proteins. J. Cell Bid., 64: 421-430, 1975.

2. Caramo-Fonseca, M. Testosterone-induced changes in nuclear pore complexnumber of prostatic nuclei from castrated rats. J. Utrastruct. Res., 80: 243-

251,1982.3. Caramo-Fonseca, M., and David-Ferreira, J. F. Quantitative study of nuclear

pore complexes in ventral prostate isolated nuclei from intact and castratedrats. J. Utrastruct. Res., 76: 202-214, 1981.

4. Coleman, S. E.. Duggan. J., and Hackett, R. L. Freeze-fracture study ofchanges in nuclei isolated from ischemie rat kidney. Tissue & Cell, 6: 521-

534, 1974.5. DeRobertis, E. M., Longthome, R. F., and Gurdon, Y. B. The intercellular

migration of nuclear proteins in Xenopus oocytes. Nature (Lond.), 272: 254-

256, 1978.6. Feldherr, C. M. The nuclear annuii as pathways for nucleocytoplasmic ex

change. J. Cell Bid., 14: 65-72, 1962.7. Fetóherr, C. M. A comparative study of nucleocytoplasmic interactions. J. Cell

Bid., 42: 841-845, 1969.8. Fetóherr, C. M. Evidence for changes in nuclear permeability during different

physiological states. Tissue & Cell, 3: 1-8, 1971.9. Foa, C., Foa, R., Bazin, H., Lazar, P., and Carcassonne. Y. Étudestatistique

des pores nucléairesdans tes cellules lymphoïdes. Nouv. Rev. Fr. Hematd.,14: 721 -726, 1974.

10. Fossa, S. D., Kaalhus, O., and Scott-Knudsen, 0. The clinical and histopath-ological significance of Feulgen DNA values in transitional cell carcinoma ofthe human urinary bladder. Eur. J. Cancer, 73: 1155-1162. 1977.

11. Franke, W. W. Structures, biochemistry, and functions of the nuclear envelope.Int. Rev. Cytd., 4 (Suppl.)- 71-236, 1974.

12. Gaub, Y . Auer. G . and Zetterberg, A. Quantitative cytochemical aspects of acombined Feukjen-naphthd yellow S staining procedure for the simultaneousdetermination of nuclear and cytodasmic proteins and DNA in mammaliancells. Exp. Cell Res., 92: 323-332, 1975.

13. Johannessen, J. V., Sobrinho-Simoes, M., Finseth, J., and Dilstrom, L. Papillarycarcinomas of the thyroid have pore-deficient nuclei. Int. J. Cancer, 30: 409-

411,1982.

14. Johannessen, J. V., Sobrinho-Simoes, M., Tangen, K. O., and ündmo,T. Aflow cytometric deoxyribonucleic acid analysis of papillary thyroid carcinoma.Lab. Invest., 45: 336-341,1981.

15. Koss, L. G., Bartels, P. H., Bibbo, M., Freed, S. Z., Taylor, J., and Wied, G. L.Computer discrimination between benign and malignant urothelial cells. ActaCytd., ?9: 378-391,1975.

16. Levi, P. E., Cooper, E. H., Anderson, D. K., and Williams, R. E. Analysis ofDNA content, nuclear size, and cell proliferation of transitional carcinomas inman. Cancer (Phila.), 23:1074-1085,1969.

17. Many, M. C., Denet, J. F., Cordier. A. C., and Haumont, S. Morphometry ofnuclear pore complexes in thyrdd cells during hyperplasia and involution. Mol.Cell. Endocr., 23:137-149,1981.

18. Maul, G. G. The nuclear and the cytopiasmic pore complex: structure; dynamics; distribution; and evdutkxi. Int. Rev. Cytd., 6 (Suppl.): 75-186,1977.

19. Maul, G. G. Nuclear pore complexes: elimination and reconstruction duringmitosis. J. Cell Bid., 74: 492-500,1977.

20. Maul, G. G., and Deaven, L. L. Quantitative determination of nuclear porecomplexes in cycling cells with different DNA content. J. Cell Bid., 73: 748-

760,1977.21. Maul, G. G., Deaven, L. L., Freed, J. J., Campbell, G., Le, M., and Becak, W.

Investigation of the determinants of nuclear pore number. Cytogenet. CeliGenet, 26:175-190,1980.

22. Maul, H. M., Hsu, B. Y., Borun, T. M., and Maul, G. G. Effect of metabolicinhibitors on nuclear pore formation during HeLa S:, cell cycle. J. Cell Bid., 58:669-676,1973.

23. Maul, G. G., Maul, H. M., Scogna, J. E., Lieberman. M. W., Stein, G. S., Hsu,B. Y., and Bonjn, T. W. Time sequence of nuclear pore formation in phytohe-magglutinin-stimulated lymphocytes and in HeLa cells during the cell cycte. J.Cell Bid., 55: 433-447,1972.

24. Maul, G. G., Price, J. W., and Lieberman M. W. Formation and distribution ofnuclear pore complexes in interphase. J. Cell Bid., 51: 405-418,1971.

25. Nakai, E., Aasamoto, H., Furuta, M., and Nishiura, M. Ultrastructure of immu-noblastic sarcoma as observed by freeze-fracturing. In: Proceedings of theNinth Congress of Electron Microscopy (Toronto), Vd. 2, pp. 316-317,1978.

26. Paine, P. L., Moore, L. C., and Horowitz, S. B. Nuclear envelope permeability.Nature (Lond.), 254; 109-114.1975.

27. Pauli, B. U., Alroy. J., and Weinstein, R. S. The ultrastructure and pathobiologyof urinary bladder cancer. In: G. T. Bryan and S. M. Cohen (eds.), The Pathologyof Bladder Cancer, Vd. 2, pp. 41-140. Boca Raton, FL: CRC Press, Inc.,

1983.28. Rejthar, A., and Blumajer, J. Difference in density of nuclear pores in normal

and malignant fibrodasts of Syrian hamster. Neoplasma (Bratisl.), 21: 479-

482,1974.29. Severs, N. Y., and Jordan, E. G. V. An interpretive freeze fracture study of

plant cell nuclear pore complex. J. Utrastruct. Res., 52: 85-88,1975.30. Svejda, J., Vrba, M., and Blumajer, J. A Ireeze-etch study of occurrence of

nuclear pores in normal and tumor cells. Neoplasma (Bratìsl.).22: 385-390,1975.

31. Tribukait, B. Flow cytometry in surgical pathology and cytology of tumors ofthe gentio-urmary tract. In: L. G. Koss and D. V. Coleman (eds.). Advances ¡nClinical Cytology. Vd. 2, pp. 163-189. New York: Masson Publishing, 1984.

32. Tribukait, B., Gustafson, H., and Esposti, P. L. Pioidy and proliferation inhuman bladder tumors as measured by flow cytotluorometnc DNA-analysisand its relation to histopathdogy and cytology. Cancer (Phila.), 43: 1742-

1751,1979.33. Tribukait, B., Gustafson, H., and Esposti, P. L. The significance of ploidy and

proliferation in the clinical and biological evaluation of bladder tumors: studyof 100 untreated cases. Br. J Urol.. 54: 130-135,1982.

34. WHO. International Histological Classification of Tumors, No. 10. Geneva:WHO, 1973.

35. Wolley. R. C., Schreiber, K., Koss, L. G.. Karas, M., and Sherman, A. DNAdistribution in human colon carcinomas and its relationship to clinical behavior.J. Nati. Cancer Inst., 69:15-22,1982.

36. Woyke. S., and Czerniak. B. The application of the freeze-fractured replicasfor the evaluation of the nuclear envelope of malignant cells from the humaneffusions. Acta Med. Pd., 20: 477-478,1979.

3756 CANCER RESEARCH VOL. 44

Research. on December 13, 2020. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from

1984;44:3752-3756. Cancer Res   Bogdan Czemiak, Leopold G. Koss and Andrew Sherman  Nuclear Pores and DNA Ploidy in Human Bladder Carcinomas

  Updated version

  http://cancerres.aacrjournals.org/content/44/9/3752

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/44/9/3752To request permission to re-use all or part of this article, use this link

Research. on December 13, 2020. © 1984 American Association for Cancercancerres.aacrjournals.org Downloaded from