Characterization of a cell line derived from rhabdoid tumor of kidney

Characterization of a Cell Line Derived From Rhabdoid Tumor of Kidney TED GANSLER, MD, WILLIAM GERALD, MD, PHD, GAIL ANDERSON, BS, T. STOKES GRAMLING, BS, CINDY H. WILLIAMS, AS, HT (ASCP), DONALD SENS, PHD,AND A. JULIAN GARVIN, MD, PHD Rhabdoid tumor of kidney (RTK) is a rare, highly malignant childhood neoplasm of uncertain histogenesis. Several recent studies have described considerable histochemical heterogeneity among cases of RTK, with confusing combinations of epithelial, mesenchymal. myogenous, and neuroepithelial markers in some tumors. Thle present study characterizes the histology, ultrastructure, histochemistry, cytogenetics, and oncogene expression in a cell line derived from RTK. The surgical specimen, nude mouse xenograft, and cell cultures demonstrated characteristic intermediate filament whorls by electron microscopy and expressed vimentin (diffusely) and cytokeratin (focally, in hyaline cytoplasmic inclusions) without detectable desmin, Thy-l, or epithelial membrane antigen. S-100 protein was absent in the surgical specimen and hleterotransplant, and was seen very weakly and focally in the cell cultures. Light microscopic features of cultures were unchanged. by several compounds (tissue plasminogen activator, nerve growth factor, cyclic adenosine monophosphate) which induce differentiation of some other pediatric neoplasms. The growth factor requirements of RTK cultures indicate a cell with mesenchyrnal features. Insulin-like growth factor-2 mRNA was detected in the RTK and in three Wilms’ tumors also studied. Unlike moist Wilms’ tumors, RTK expresses the c-myc rather than the N-myc oncogene. HUM PATHOL 22:259-266. Copyright !% 1991 by W.B. Saunders Company

Rhabdoid tumor of the kidney (RTK) is a highl) aggressive malignant) which comprises about 2% of’ all childhood renal tumors. ’ The nomenclature ti)~ this tumor often leads to confusion with Wilms’ tumors (W’I‘) which contain a large amount of skeletal muscle, the so-called “fetal rhabdomyomatous” WT.’ .Uthougll previouslv characterized as a \,ariant of WT. RT K has many distinctive morphologic and clinical features and is no longer- included in the classific.ation of MY’. I.’ Rhabdoid tumor of the kidney usu,tlly occurs in children under 2 years of age. with half of the patients diagnosed before reaching 1 year of age. I.’ II is often accompanied bv a neural tumor, usuallv medulloblastoma. and extrarenal metastasis is

often present at diagnosis.-‘,’ Unlike most W’I‘, RTK is highly resistant to chenlotherapy.4 Histologically, RTK cells are characterized by large nuclei with large, single nucleoli and abundant, glassy eosinophilic cytoplasm with inclusions composed of whorls of intermediate filat~~ents.1-‘5 However, these intermediate filament aggregates are neither necessary nor sufficient for diagnosis of RTK.“J’ Rhabdoid tumor of the kidney may coexpress keratin, vimentin. desmin, and neurofilaments with various combinations of S- 100, epithelial membrane antigen (EMA). m~oglobin. neuron-specific enolase, and a,-antichymotrvpsin.‘.“-” These findings suggest either that RTK cells are pluripotential or that RTK is a histologic pattern rather than a homogeIIOUS pathologic entity. Information regarding the histogenesis of RTK has been limited bv a scarcity of’ fresh tissue for cellular and molecular- studies. For this reason, we have propagated R’I‘K cells as athymic mouse heterotransplants and cell cultures. These cells were used for cvtogenetic. ultrastructural. and histochemical anal!;&. for determination of growth factor requirements m vitro, and for analvsis of myc OIKOgene and insulin-like growth factor-2 (lGF12) gene expression. ‘l‘hese studies indicate that RTK has a phenotype which differs in some respects from that of‘ WT, and that these characteristics remain stable during serial passage in vitro and as nude mouse heterotransplants. CASE PRESENTATION h renal mass was noted at birth in an olllt.rwise health\ male intant. Nephrectomy revealed a medullary renal tumor with light microscopic findings typical ot‘ RTK. Despite negative surgical margins. the tumor recur-red within the abdomen, resulting in bowel obstruction in the sixth postoperative week. A computed tomographic :xan of the head rrvealed a Y-cm cerebellar mass. I%e infant developed increasing intracl-anial pressure and died 1 week later. Postmortem examination revealed metastatic R? K involving regional lymph nodes and the liver, and extensive intraabdominal spread to serosal surfaces forming a 291-g mass. A 2.5cm right cerebetlar medulloblastoma was present. and demonstrated focal astroqtic and ependvmal differentiation.

..~

__FIYNI~ IIre 1)tparlment ot Pathology and Laboratot \ Medicine, Lledical L:niverGt\ of South (Carolina, Charleston. SC; and the Departnwnt ot Biolog!, Furrnml Lniversity. (;rrcvvillt~. SC1. Accepted

for publication ,)uI\ .5. 1990. Suppwted I)) National Institutes of Health C;r;tn1 No. (:A 37X87 froru the National (:ancer Institute. Bethesda. MD. Dr Garrslet, is presently affiliated with the Departtnent of PaIhology awl La~)~lI-aror\ Xledic inc. Emory I-niversir\ Hospital, Alrlnta. GA. A’P~uwrclc: r-habdoid tunlor. kidney. cell wlture. Addrer\ correspondence and reprint requests to A. Julian (Marvin, MD, PhD. Department of Patholog, and Laborator) Sledtc-ine. Medic-al L’rriversit\ of South Car&la. Ii1 .hshle, Ave. (Zharleston. SC: L’!I11’.5. (:opyright k’, 1991 bv W.B. Saunders (Zotnpan\ OO/fi-8 I7i!‘l l/2?oJ-ooIo$.~.ooio

MATERIALS AND METHODS Rhabdoid Tumor Heterotransplantation Nude Mice 22 portiott finet)

259

minced

of’ tutnor from (L’-mm diamettr

the surgical 1.ragments)

Into specimen was in phosphate-

HUMAN PATHOLOGY

Volume 22, No. 3 (March 1991)

buffered saline. Tumor fragments were injected subcutaneously into the flank of male BALB/C athymic mice, 6 to 10 weeks of age, obtained from the National Cancer Institute. Approximately 10’ cells were injected per mouse. Subsequent tumor nodules were aseptically excised, minced, and injected subcutaneously into additional nude mice.‘” To determine the effect of passage in vitro on tumorigenicity, RTK cultures were trypsinized. resuspended in phosphate-buffered saline, and 10’ cells were injected into nude mice.

Cell Culture Subcutaneous tumor nodules were excised from sacrificed mice, finely minced in Dulbecco’s modified Eagle’s medium (DMEM) with penicillin and streptomycin, and vigorously pipetted to yield a cell suspension. The susperlsion was centrifuged (8OOg. 5 min), and the resulting pellet was resuspended in culture medium. Serum-supplemented medium (DMEM:Ham’s F-12 (1: I) with l-10% fetal bovine serum [FBS]), and serum-free medium (DMEM:Ham’s F12 [ 1: l] + 5 ng/mL selenium. 5 PgiinL insulin. 5 FgimL transferrin. 36 ng/mL hydrocortisone. 4 pg/mL triiodothyronine, and 10 ng/mL epidermal growth factor) were evaluated. Aliquots of the cell suspensions were pipetted into 75-cm’ flasks, with or without a bovine collagen tape 1 matrix. Cultures were maintained at 37°C in a humidified atmosphere of 5% carbon dioxide in air and were fed every 3 days. Adherence to the substrate, cell morphology, and degree of confluence were evaluated by phase-content microscopy.

Thymidine Incorporation Assay Cells of RTK were harvested from subconfluent cultures and plated in 2% FBS onto 96-well trays at a cell concentration producing monolayers of approximatel) SO%, confluence. On day 2, medium was changed to Basal medium (DMEMIF 12 + selenium. hydrocortisone, triiodothyronine). Growth factors in various combinations were added to triplicate wells on day 4, as indicated in the Results section. These growth factors were obtained from Collaborative Research (Bedford, MA) and included epidermal growth factor (EGF), insulin, platelet-derived growth factor (PDGF). and transforming growth factor-p (TGF-P). The nerve growth factor (NGF) was a gift from Dr Gary Landreth, Medical Universitv of South Carolina. Cells were exposed to tritiated thymidine ( 1.O ~Ciiwell) 6 hours later. After 12 hours of thymidine incorporation, cells were detached with trypsiniethylenediaminetetraacetic acid (EDTA). and the DNA-associated radioactivity was immobilized on glass fiber filters, using a semiautomatic cell harvester (Skatron Inc, Sterling, VA). Tritium was quantitated by liquid scintillation counting and the mean values of triplicate wells were expressed as percentage of control assays were also (basal medium only) values. I7 Identical performed using cultures of WT cells and human proximal tubular epithelial cells. The WT tumor cultures were obtained from tumors containing primarily, blastemal and epithelial components. The tumors contamed scant stroma with no evidence of heterologous differentiation, eg, skeletal muscle. Statistical significance of data was evaluated by Student’s t-test.

ter overnight attachment, the medium was replaced by DMEM:F- 12 (1: 1) containing various growth factor combinations. Wells were refed with the same medium and growth factor supplements again on day 4. On day 7. cells were fixed by adding 1110 volume of 11% glutaraldehyde. rinsed in water, and air dried. Cells were stained with 0.1% crystal violet for 30 minutes, rinsed 4 times in water, and ail dried. Stain was eluted by adding 10%) acetic acid, and the optical density at 650 nm was determined in a microplate Ix Optical densitv at 650 nm was found spectrophotometer. to increase linearly with cell numb&, as measured bv hemocytometer counts of replicate wells (data not shown). Cells grown under these conditions were found to reach saturation density by day 7 (data not shown). Data were analyzed using Student’s’t-test.

Cytogenetic Analysis Cells of RTK were treated with colchicine (0.0 1 kg/mL) for 20 minutes, detached by trypsinization, hypotonicallv treated with 0.075 mol/L potassium chloride, and fixed with methanol:acetic acid (3: 1). Chromosomes on air-dried slides were G-banded with 0.25% trypsin and stained with 0.5% Wright’s stain. Karyotypes were at approximately the 4OO-band level.

In Vitro Differentiation Cells of RTK were sparsely plated onto I:! well trays in DMEMiHam’s F-12/27( fetal calf serum. Sixteen hours later, the medium was removed and duplicate wells were refed with media containing NGF (10 PgimL), tissue plasminogen activator (2.5 mol/L), or dibutyryl cyclic adenosine monophosphate (2.5 PmoliL). Media were replaced as above on day 3. Differentiation was evaluated by phase contact microscopy on day 6 posttreatment.

lmmunohistochemistry Imnlunohistochelriical studies of the surgical specimen, nude mouse heterotransplant, and cell monolayers used antibodies against cvtokeratin (Triton Biosciences, Alameda, CA), desmin (dioGenex, San Ramon, CA). vimentin (BioGenex), EM,4 (BioGenex), S- 100 protein (BioGenex). and Thy-l antigen. These antibodies were localized by the avidin-biotin-peroxidase method (Vector, Burlingame, CA).”

Electron Microscopy Tissue specimens and cell cultures were tixed in 2.5% glutaraldehyde, postfixed in 1% osmium tetroxide, and embedded in Epon 8 12. Sections were stained with uranyl acetate and lead citrate, and examined on ;L JEOL lOUS microscope.

RNA Isolation and Analysis Total cellular RNA was prepared bv acid guanidinium thiocyanate-phenol-chloroform extra&on.“” Ten micrograms of total RNA were added to each lane of a 1.2% agarose glyoxal gel. ” Following electrophoresis, gels were stained with ethidium bromide, identifying the positions of the 1% and 2% ribosomal RNAs, and verifying that RNA was intact and that each lane contained the same amount of RNA. RNA was transferred to nylon filters by alkaline blotting in 10 mmol/L sodium hydroxide. Actin, IGF-2, N-rllyc and c-mvc mRNAs were detected by hybridization (1 mmoli

Cell Proliferation Assay Rhabdoid tumor of’kidney cells was plated onto 96-well trays (2,500 cells/well) in DMEM:F-I2 (1: 1) + 1% FBS. Af-

260

RHABDOID TUMOR OF KIDNEY (Gansler et al)

1. EIYl‘A. 1j.5 mol/ll sodium phosphate, 7’1 sodium dodecyl sulfate, (i!i”C) with DNA probes (American Tape (Iulture (:ollectic)rl. Kockville. MD) labeled with “>P by nick translation.” Filters were washed in 1 mmol/L EDNA, 40 mmol/L sodium phosphate, and 1% sodium dodecyl sulfate ;it 65°C: and l;ibeled bands were detected by autoradiograph\.

RESULTS mice

Subcutaneous inoculated

tumor nodules with suspensions

grew in four of five of’ RTK c-ells from

the primary tumor, reaching a size of 1.5 cm after 1 to L’ months of growth. These heterotransplants were histologically identical to the primary rumor. Passage of RTK I:ells for several generations in vitro did not alter their in viva tumorigenicity. The tumor is currently in its ninth serial passage in nude mice, without evidence of any morphologic changes. Cells of RTK attached readil?, to both the type-l collagen matrix and to the uncoated plastic flasks. The cells grew rapidly in m’edium containing FBS concentrations of 1 %p to ‘lo% (doubling time, approximately 4 days) but did not grow as bvell in the serum-free medium used in this laboratorv for growth of WT cells and normal human proximal tubular (HPT) cells (WTiHPT = DMEMiHam’s F-12 with selegrowth medium Ilium, h)drocortisone, triiodothyronine. transferrin, and LX~F).2”~2:i The requirements of RTK insulin, cells for serum and peptide growth factors were more quantitatively assessed by a thymidine incorporation assav and a cell prolifer&on assay. ‘The :‘H-thymidine incorporation assay revealed several d.ifferences in hormonal regulation among RTK, W’l, and HPT cells. These data (Fig 1, left) indicate that DIVA synthesis by RTK cells is most strongly stimulated by PDGF, TGF-P, and FBS. Cells of WT a.nd HPT are not significantly affected b)

PDGF or TGF-P. Thymidine incorporation by W’Iand HPT cells was most sensitive to insulin and EGF. and the combination of insulin, transferrin, and EGF was approximately as effective as FBS. Cells of RTK and WT showed significant (P < 0.05) differences in their responses to PDGF and TGF-P (alone or in addition to insulin, transferrin. and EGF). and to FBS. Although the cell proliferation assay and thymidine incorporation assay both support similar conclusions (see Discussion section), there were some differences which probably reflect the fact that different substances may regulate short-term DSA synthesis and longer-term proliferation over several cell generations (Fig I. right). Fetal bovine ser,Jrn was far more active than any serum-free formulation in promoting RTK cell proliferation (P c 0.001). The growth factors PDGF and TGF-P alone had little effect on RTK cells, but PDGF considerably potentiated (1’ < 0.001) the effect of other peptides (combination of insulin, transferrin, and EGF). For WT. serumsupplemented and serum-free media were equally effective. While addition of PDGF to the insulin/ transferrin/EGF combination increased WT growth (P < O.OOl), addition of TGF-P did not. Cells of HPT grew best in serum-free medium with insulin/ transferrin/EGF and proliferated less when that combination was supplemented by T(;F-p (f’ < 0.001). The tumor cells had a 46,XY karyotype without any detectable cytogenetic abnormalities. Immunohistochemical staining 01’ the surgical specimen, nude mouse heterotransplant, and cell cultures showed similar patterns of expression of intermediate filament proteins and epithelial and neural markers. Cells showed diffuse cytoplasmic reactivity for vimentin. while cytokeratin was predominantly restricted to cytoplasmic inclusions which were intensely positive. In vitro. occasional large, flattened

FBS

FElS

I/E/T + TGF-B

I/E/l + TGF-B

I/E/T + PDGF

I/E/T + PDGIF

IlE3-r

I/E/T

TGF-B Fix3

FiXiF

j-J

TRANS

TRANS

n

BJ

!33

HPT

EIw RTK

INS

INIS 200

400

600

800

0

1000

200

400 OD

CPM ‘H, % of Control

650

600

600

1000

1200

1400

% of Control

FIGURE 1. (Left] Thymidine incorporation assay of growth factor effects on normal human renal proximal tubular epitheliol cells (HPT), Wilms’ tumor cells m, and rhabdoid tumor of kidney cells (RTK). Data represent mean 2 SEM from two experiments using two HPT cell lines, three experiments with two classic and one anaplastic WT cell lines, and three experiments with one RTK cell line in consecutive in vitro passages. Each experiment was done in triplicate culture wells. Ins, 5 pg/mL insulin; EGF, 5 ng/mL epidermal growth factor; trans 10 pg/mL transferrin; PDGF, 5 ng/mL platelet-derived growth factor; TGF-p, 2 ng/mL transforming growth factor-p: IET, insulin + EGF + transferrin in concentrations shown above; FBS, 10% fetal bovine serum. [Right) Cell proliferation assay of growth factor effects on HPT, WT. and RTK. Data are mean ? SEM from triplicate values of a representative experiment using one HPT, WT. and RTK cell line. Growth factor concentrations are the same as for the left panel

261

HUMAN PATHOLOGY

Figure 2.

Volume 22, No. 3 [March 1991)

See page 263 for legend.

clinical specimen, mouse heterotransplant, and cell cultures revealed similar morphology. Cells of RTK had poorly developed Golgi apparatus, numerous free ribosomes but sparse rough endoplasmic reticulum, oval mitochondria, and rare intermediate filament aggregates. Neurosecretory granules, mi-

cells showed diffuse cytoplasmic keratin positivity but, in most cells, cytokeratin was limited to cytoplasmic inclusions. Desmin, Thy-l, and EMA were not detected. S- 100 antigen was questionably detected in occasional cells in vitro. Representative photomicrographs are shown in Fig 2. Electron microscopy of the

262

RHABDOID TUMOR OF KIDNEY

(Gansler et al)

FIGURE 2. lmmunohistochemistry of RTK. Tissue from the nephrectomy specimen [A C, and E) or RTK cell cultures (6, D, and F) was fixed in Carno\/s solution and immunohistochemical staining for cytokeratin (A and 8). vimentin (C and D), or desmin (E and F) was performed using the avidin-biotin-peroxidase method. Arrowheads indicate cytokeratin in cytoplasmic inclusions.

lial (cytokeratin, EMA), myogenous (desmin, myoglobin), or neural (neuron-specific enolase, S-100 antigen, neurofilament proteins) differentiation.“-‘” These observations have raised the question of whether RTK represents a multipotential neoplastic cell which may express (often simultaneously) several classes of markers, or whether several morphologically similar entities are being lumped together as RTK.” The present study of RTK used three strategies: (1) morphologic and histochemical characterization of a case with clinical and light microscopic features typical of RTK; (2) examination of phenotypic and molecular features (growth factor requirements for c-my, N-myc,and IGF-2 mRNA exin vitro culture; pression) of RTK not previously studied, which allow additional criteria for comparison with other normal and neoplastic cell types; and (3) similar experiments using RTK cells2”sY5 since several other poorly differentiated pediatric neoplasms can be induced to undergo differentiation in vitro (revealing clues regarding their histogenesis). Light microscopic and ultrastructural features of this case are typical of RTK, and remained unchanged during nine passages as subcutaneous heterotransplant in nude mice. Intermediate filament aggregates were present in the surgical specimen, heterotransplant, and in cultured RTK cells. Cytokeratin and vimentin filaments were detected immunohistochemically in tumor tissue specimens and in cultures, whereas desmin was absent. Coexpression of cytokeratin and vimentin has been noted in many

crovilli. and intercellular junctions were not apparent. Morphology of cell cultures (evaluated by phase contrast microscopy) was unaffected by exposure to NGF, cyclic adenosine monophosphate, or tissue plasminogen activator. Hybridization of Northern blots of total RNA revealed undetectable expression of c-rnyc mRNA in normal renal tissue, in WT surgical specimens, and in WT mouse xenografts (Fig 3). The surgical specimen and mouse xenograft of RTK showed prominent expression of c-mvr mRNA. In contrast, N-m~c expression was present in all surgical specimens and xenografts of WT, but was undetectable in normal kidney and IKTK samples. N-myc expression was much stronger in W’I‘ xenografts than in surgical specimens of the primary tumors they were derived from, possibly reflecting more rapid growth or a greater proportion of the blastemal component in the nude mouse tumors. The growth factor IGF-2 mRNA was not detected in normal kidney tissue, but was strongly expressed by surgical specimens and xenografts of WT and RTK. Actin mRNA expression was constant in all samples. DISCUSSION Despite several ultrastructural and histochemical studies (including the present one) of RTK, a clear phenotypic and histogenetic picture of this neoplasm remains an elusive goal. Previous reports have noted immunohistochemical markers which might be considered evidence of mesenchymal (vimentin), epithe263

HUMAN PATHOLOGY

Volume 22, No. 3 [March 1991)

FIGURE 3. Myc oncogene and IGF-2 gene expression. Ten micrograms of total RNA were fractionated by electrophoresis. transferred to nylon membranes, and hybridized with probes for IGF-2, N-myc, c-myc, and actin. Key: NK normal kidney tissue adjacent to tumor; WT. Wilms’ tumor; RTK. rhabdoid tumor of kidney; S, surgical pathology specimen; M, nude mouse tumor xenograft. Starting from the leff side of the figure, the first three groups of lanes are from three different cases of WT. The fourth group [two WT xenografts) are cases for which corresponding fresh surgical specimens were not available.

Actin

264

RHABDOID TUMOR OF KIDNEY

(,f‘

R’l-K

I.T.H.I:(.I

I

although some cases contain 0111~one t\‘pe of intermediate filament proteitt.4~7~x~1:i 111most reports, the cellular distribution of‘ intermediate filament staining is not specificallv discussed. Weeks et al reported several cases of K’fK with diffuse c\-toplasmic staining for vimentin and cytokeratin locali;/ed predominantly to cytoplasmic inclusions? This pattern was also observed in our surgical specimen.. nude tno~~se heterotransplants, and cell cttltures of R’I‘K. S- 1t)O antigen has been detected in occasional K’I‘K specimetls,x.‘:’ but WIS verv weakly and f’oc;tlh rrxtive in our KI‘K cultures, anti entit-el\ ;thsenr in rhr surgical specimen and mouse heterotransplant E:pir helial membrane antigen has been previousI!, reported to he present, 7~x.I:ibitt bxs absent in this cask. ‘l‘hv- 1 is an antigen originally detected on I! mphocyttbs and subsequently found in neural and rrnal tissttct. Recent studies in our laboratory havr rcvealcd J‘hv- 1 expression in the medullar!. s;rc,mal ~~11s of human fetal kidne\, antl in congenital mesoblastic nephrotii;~s.- “Ii ‘l‘he &sence of ‘l‘h\,- 1 in R’I‘K suggests that its phetiotvpe differs from t;ot-ma1 fetal renal stroma and f’rotii t;eoplastic stromal cells ofcongenital mrsoblastic nephroma. In summar\, histochemic-al :)tuti\ of‘ this C;ISCdrtected a combination of. market-s whic,lt ;IW frccluent1.y seen with R’l‘K but at-c’ bl no tnems diagnostic of thts neoplasm. It should be noted that S- 100 protein and c\ tokeratin expression i5 not limitecl IO nrur:tl or epi;helial tw~pl~~stns. resl)ec-timely, :ttitl thxt both m;iv he present in tttniors of ttresenchvmal origin. (;zllS from this tumor \vct-e successfully grown in \.itro. usiris media containing fetal calf st7utii at ;I ~‘OIl( elltt-atloll of at k1st 1’7 . dnct qrew much less rapidle in scruni-free medium. ‘l‘his is in c.ontt-ast to not‘n,;;l ttumtit rctt;~l tubular cpithelial cells and cells 1roni WI‘, which gr’nt’ as well or better in set-tin-free media.““.” ’ Fil~roblasts and other mrsenchvmal cells urt~;~llv t-ecluirr serum f’or growth in Ctro. (jtte of‘the mosf essential ~r(~~~Th factors for cultures 01 meseti’ I‘ransforming cliym;il trll5 ‘is Pl)(;F.” growth fi’<‘tc,r-f3 is also an important mesenchvmal tell mitoget’. \\,hich also inhibits proliferation oi rn;tny.types of nonneopl.r~ticepirhelial cells.” I‘hytttidine tncorpor;‘tion assavs (estimating DNA replication) showed that l’l)(;F an’1 ‘IGF-p are highly mitogenic li)r RTK cells. In cl )ntr;tst, these two growth factors have little ef‘t’&t on norrttal renal tubular epithelial cells or W71‘ ct*lls. ‘I‘hese results suggest that R’I‘K cells resemble manv t\-pes of mesenc-h~mal cells with respect to hortiion~l ;.ontrol of prolittration. Alth’btlgh l’D(;F and ‘I‘(;F-P markedlv stimulated th~mtdine ittcorporation in DNA, neither functioned as
265

(Gansler et al)

transferrin, and EGF combination inhibited HPT cell proliferation without effect on RTK or M”r cells. Human proximal tubular cells grew best itt serum-free medium, R?‘K cells grew much faster in serumsupplemented mediutn, and the two medium types were nearly equivalent for supporting WT cell growth. Based on the thymidine incorporation and cell proliferation assa,ys, we concluded that the growth requirements of HP?‘ cells are typical of nonneoplastic epithelium, RTK cells arc most consistent with mesenchymal cells, and WT cells appear to have an intermediate phenotvpe. Cytogenetic exan;inatioti revealed a 46,XE karyotype without any apparent abnormalities. Although most WT have normal karyotype, anaplastic W’l‘ which resemble RTK in their biologic aggressiveness and resistance to therapv are nearly always atleuploid.“‘,:“’ Previous flow cytometric studies of RTK have demonstrated diploid Dh’A content.l:‘,“” .I‘he 1 1p 13 deletion which may be seen in W’I‘ and is hiFhl!, specific for that neoplasm was not present in thts case or in any other reported karyotype of R’I‘K. I:’ ‘l‘risomv of chromosome 2 has recently been reported in ‘)I& extrarenal tumor histologically resemblit!g rhabdoid tumor or epitltelioid sarcoma. sttggesttng a similarity of that neoplasm to t-habdomvosarcotna. Ii However. this abnorrnalitv was absent in’the present case ~:-nr~~ ;tnd N-NIVCprotootlco~enes enc.‘& nuclear proteins- which at-; important m regulation of cell proliferation and differentiation. Levels of C-V~~C mRNA and protein are commonlv elevated in mai) types of neoplastns, whereas h’-~& overexpression is less common and is most consistently prt.sent in neuroblastomas and \4T.z”.1’:’ ‘I‘hus. the pattern of‘ myc gene expression in this case of RTK (increased c-&c) (liffers from that observed in most caseb of W’I‘ (increased N-trryc). Among those rumors c,urrently or previously thought to be related to R’I‘K. c-WLJCis frequently overcxpressect in renal cell carcinomas and rh~tt~clorri~osar~~~n~as.~‘~‘~~~i It&in-like growth f&tot--Z is a pepride mitogen, expressed at low levels in adult tissues (including kidne)~ and at higher levels in fetal kidney tissue and L1~‘l‘.:“1.z’7 Insulin-like growth factor-:! m RNA and/or protein has been also detected in several other pediatric :tnd adult neoplasms. including pheochromocytotna.-” t-hab(tom~osarcc)nta.‘“’ hepatot~lastoma,:S’~ colon carcinoma.“x and liposarcoma. ” If IGF-2 mRNA had been absent in RTK, this difference between KI‘K and U;I‘ would have been interesting. However, the presence of l<;F-L’ mRNA in R’I‘K does not necessarily imp,ly ,close similarit\, to WI‘. Since dtfferentiation of some other pediatric and adult neoplasms (neuroblastoma, E:win#‘s sarcoma. pheocht,otnocytc,ma, rhabdomyosarcoma) can be induced itt \.itro by treatment with phorbol esters, cyclic adenosine monophosphate, or N(;F.“‘.“i.‘“’ we exposed RTK cultures to these agents. without evidence of neurite outgrowth, myotube formation, or othet morphologic features of’ neural or muscular differentiation detected by phase contrast microscopy.

HUMAN PATHOLOGY

Volume 22, No. 3 (March 1991)

Although there remains no single morphologic, cytogenetic, histochemical, phenotypic, or molecular marker for RTK at the present time, the diagnosis of this neoplasm can be suggested by considering both clinical context and histologic appearance, and can be supported but not unequivocally confirmed by the ancillary studies noted above. While this and previous studies have had limited success in determining exactly what RTK is (probably mesenchymal). we have obtained evidence relevant to what RTK is not (a variant of Wilms’ tumor). These cellular and molecular studies support the clinical and histologic impression that RTK and WT are not closely related, and support previous histochemical and ultrastructural data classifying RTK as a mesenchymal neoplasm. Acknowledgment. ‘I‘he authors thank Sandy Nelson and Carol Hill for editorial assistance. We thank Dr Wayne Stanley for performing the cytogenetic analysis, and James Nicholson, Helen Braid, and Leslie Jensen for photographic assistance. The NGF used was the generous gift of’ Dr Gary Landreth, and the Thy-l antigen was generously provided by Dr An-Chuan Wang.

REFERENCES 1, Beckwith JB. Palmer NF: Histopathology and pi-ognosis of Wilms’ tumor. Results from the First National Wilms’ tumor study. Cancer 41:1937-1948, 1978 2. Wigger J: Fetal rhabdonlyomatous tlephroblastorna. HL~M 7:613-62.3. 1976 3. Haas.JE. Palmer NSF, Weinberg

PATHOL

A(;. et al: Ultrastructure of malignant rhabdoid tumor of the kidney: A distinctive renal tumor of children. HUM PATHOL 12:646-657. 19X I 4. Weeks DA, Beckwith JB. Mierau GW. et al: Rhabdoid tllmor of kidney. .4 report of I I I cases from the National Wilms’ study tumor pathology center. Am J Surg Pathol 13:439-45X, 1989 5. Bonnin JM. Rubinstein L,J, Palmer NF, et al: The association of embryonal tumors originating in the kidney and in the brain: A report of seven cases. Cancer 54:2137-2 146, I984 6. Weeks DA, Beckwith JB. Mierau GW: Rhabdoid tumor: .411 entity or a phenotype? .4rch Pathol Lab Met1 113:I 1% I 14. 1989 7. Tsokos M, Kouraklis G. Chandra RS. et al: Malignant rhabdoid tumor of the kidney and soft tissues. Evidence for a diverse morphological and immunocytochemical phenotype. Arch Pathol Lab Med I 13:115-120, 1989 8. Sea IS, Min KW, Brodhecker C. et al: Malignant renal rhabdoid tumor. Immunohistochemical and ultrastructural stuciies. Histopathology 13:657-666, 1988 9. Rousseau-Merck MF. Nogues C. Nezelof C. et al: Infantile renal tumors associated with hypercalcemia. Characterization of intermediate-filament clusters. Arch Pathol Lab Med 107:31 l-314, 1983 10. Rutledge J, Beckwith JB. Benjamin D, et al: Absence of immunoperoxidase staining for myoglobin in the malignant rhabdoid tumor of the kidnev. Pediatr Pathol 1:93-98, I983 11. Sotelo-Avila C. (Gonzalez-Crussi F, De Mello 1). et al: Renal and extrarenal rhabdoid tumor in children: .4 clinicopathologic study of I4 patients. Semin Diagn Pathol 3:15l-163. I986 12. Schmidt D, Harms D, Zieger G: Malignant rhabdoid tumor of the kidney. Histopathology, ultrastructure and comments on differential diagnosis. Virchows Arch [A] 398: IO I - 108. 1982 13. Schmidt D, Leuschner I. Harms D, et al: Malignant rhabdoid tumor. A morphological and tlow cytometric study. Pathol Res Pratt 184:202-212, 1989 14. Vogel AM, Gown AM, Caughlan J, et al: Rhabdoid tumors of the kidney contain mesenchymal specific and epithelial specific intermediate filament proteins. Lab Invest 50:232-238, 1984 15. Molenaar WM. DeJong B, Dam-Meiring A, et al: Epithelioid sarcoma or malignant rhabdoid tumor of soft tissue. HUM PATHOL ?0:347-35 1, 1989

266

16. Marvin AJ, Congleton L, Inabnett T. et al: Growth characteristics of human Wilms’ tumors in nude mice. Pediatr Pathol 8:599-615. 1989 17. Gansler T, Hsu WC, (iramling TS. et al: Growth facto1 binding and bioactivity in human kidney epithelial cell cultures. In Vitro Cell Dev Biol 26:285-290, 1990 1X. Kueng W. Silber E, Eppenberger U: Quantification of cells cultured in 96-well plates. Anal Biochem 182: 16-19. 1989 19. Hsu SM. Raine L. Fanger H: LJse of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:.577-80. I981 20. Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:l56-159, 1987 21. Maniatis T. Fritsch EF. Sambrook J: hlolecular Cloning. Cold Spring Harbor, NY. Cold Spring Harbor Laboratory. 1982, pp 200-201 22. Marvin A4J, Sullivan JL, Bennett DD, et al: I‘he lri r~tro growth, heterotransplantation, and immunohistochemical characterization of the blastemal component of Wilms’ tumor. Am ,J Pathol 129:353-363, 1987 23. Detrisac CJ, Sens MA, Marvin AJ, et al: Tissue culture of human kidney epithelial cells of proximal tubule origin. Kidney lnt 25:3X3-391), 1984 24. Cavazzana AO. Miser JS, Jefferson J, et al: Experimental evidence for a neural origin of Ewing’s sarcoma of bone. Am J Pathol 127:507-5 IX, 1987 25. lshikawa S: Differentiation of human neuroblastoma cells 1~ ~~ifro_Morphological changes induced by dibutyryl cyclic AMP. .4cta Pathol Jpn 27:697-711. 1977 26. Chao CC, Hazen-Martin DJ. Wang II’. et al: Immunohistochemical localization of thy- I in human kidney: A comparison of fetal and adult human kidney with selected renal neoplasms. J Histochem <:ytochem (in press) 27. Deuel TF. Pierce GF, Hsiu-.Jeng Y, et al: Platelet-derived growth factor/s in normal and neoplastic cell growth. J Cell Physiol [Suppl] 5:95-99. 19x7 28. Sporn MB, Robert AB, Wakefield LM. et al: Transforming growth factor-p: Biological function and chemical structure. Science 233:532-534, 19% 29. Schmidt D. Wiedemann B, Keil W, et al: Flow cytometric analysis of nephroblastomas and related neoplasms. Cancer 58:2494-2.500, 1986 30. Douglass EC, Look AT, Webber B. et al: Hy,perdiploidy and chromosomal rearrangements define the anaplastlc variant of Wilms’ tumor. J Clin Oncol 4:975-9X I, 1986 31. Arthur DC: Genetics and cytogenetics of pediatric cancers. Cancer .58:334-X0, 1986 32. Nisen PD. Zimmerman KA. Cotter SV, et al: Enhanced expression of the N-I~L~cin Wilms’ tumors. Cancer Res 46:62176222, I986 33. Shaw APW, Poirier V, Tyler S, et al: Expression of the N-n!yc oncogene in Wilms’ tumor and related tissues. Oncogene 3:143-149. 1988 34. Yao M, Shwin T. Misaki H. et al: Enhanced expression of c-niyc and epidermal growth factor receptor (C-&B-l) genes in primary human renal cancer. Cancer Res 48:6753-6757, 1988 35. Triche TJ. Cavazzana A<>. Navarro S, et al: N-myc protein expression in small round cell tumors, in Evans AE, D’Angio GJ, Seeger KC (eds): Advances in Neuroblastoma Research 2. New York, NY, Liss, 1988, pp 475-485 36. Scott .J, Cowelf ,J. Robertson ME, et al: Insulin-like growth factor-II gene expresslon in Wilms’ tumor and embryonic tissues. Nature 317:260-?63, 1985 37. Haselbacher C;K, Irminger J-C. Zapf J, et al: Insulin-like growth factor II human adrenal pheochromocytomas and Wilms’ tumors: Expression at the mRNA and protein level. Proc Nat1 Acad Sci USA 84:1104-l 106. 1987 38. ‘rricoli JV, Rall LB, Karakousis CP, et al: Enhanced levels of insulin-like growth factor messenger RNA in human colon carcinomas and liposarcomas. Cancer Res 46:6169-6173, 1986 39. Garvin AJ. Stanley WS. Bennett DD. et al: The in ?~tro growth. heterotransplantation. and differentiation of a human rhabdomyosarcoma cell line. Am J Pathol 125:208-2 17. 1986