[CANCER RESEARCH 43,4571-4576, September 1983]

In Vivo and in Vitro Biological Activities of Two Human Cell LinesDerived from Anaplastic Lung Cancers1

Richard S. Bookman,2 Anne Bellin, Mary A. Repo, Noreen J. Hickok, and Toru Kameya

Laboratory of Calcium Metabolism, Memorial Sloan-Kettering Cancer Center, New York, New York 10021 [R. S. B.. A. B., M. A. P.. N. J. H.J, and The National Cancer

Center Research Institute, Tokyo, Japan [T. K.]

ABSTRACT

Cell lines Lu-65 and SK-Luci-6 were established from twopatients with anaplastia (non-oat cell) lung cancers. These cell

lines showed in vivo and in vitro functional activities that couldexplain the paraneoplastic syndromes which were clinically manifested. In both patients, elevated white blood cell counts occurred in the absence of any evidence of sepsis. Tumor fragments taken directly from one patient and transplanted to nudemice produced a progressive leukocytosis in the mice. Tissueculture-derived cells from both cell lines enhanced white bloodcell numbers following heterotransplantation to nude mice. Cell-free extracts from both cell lines were found to enhance granu-locyte-macrophage colony formation in soft agar. Greater colonyformation was consistently found with the cell line (SK-Luci-6)

that was derived from the patient manifesting the more markedleukocytosis. These data suggest that the tumor cells releasecolony-stimulating activities. Coincidently, one cell line (Lu-65)

synthesized and released large amounts of prostaglandin E2withlittle or no other prostaglandin product; the other cell line produced no prostaglandins. When the tumor cell lines were cocul-

tured with explanted fetal rat bones, enhanced bone résorptionwith excessive calcium release occurred. Bone-resorbing activity

correlated with tumor prostaglandin synthesis for the cell linereleasing prostaglandin E2. An osteolytic factor that was neitherprostaglandin nor parathyroid hormone was released by the SK-Luci-6 cell line. Hypercalcémiewas a persistent feature only in

the patient from whom the latter tumor line was derived.

INTRODUCTION

Paraneoplastic syndromes associated with ectopie hormonesecretion are well-described occurrences with lung cancer (2, 7,14). Difficulty in establishing permanent cell lines from undiffer-

entiated lung carcinomas has impeded the elucidation and identification of tumor-secreted factors which could account for the

observed clinical syndromes. Recently, 2 cell lines have beenestablished into permanent cell culture; these cell lines werederived from patients with anaplastic giant-cell lung cancers (1).

As part of an effort to understand the mechanisms for thepresenting clinical syndromes, in vitro functional activities of thecell lines have been examined. Each cell line demonstrated theability to produce factors with biological activities sufficient toexplain the clinical syndromes seen in the respective patient.Clinically, both patients had marked leukocytosis in the absence

of any evidence of sepsis. A similar leukocytosis could be inducedin the nude mouse model to which the patients' tumors were

transplanted. Both cell lines could be shown to augment granu-locyte-macrophage colony formation in soft-agar cultures of bone

marrow cells. In one patient in whom hypercalcemia occurred,an osteolytic activity was released by the cultured cells. Tumorsynthesis of PGs3 could not account for the osteolytic activity

released by these cells, nor could significant levels of ¡mmunc-

reactive PTH be detected.

MATERIALS AND METHODS

Cell Unes

Lu-65. The Lu-65 culture was derived from a 64-year-old Japanesemale (Patient 1) with an anaplastic lung carcinoma. During the patient's

clinical course, elevated leukocyte counts consisting primarily of matureneutrophils were persistent and occurred in the absence of any evidenceof sepsis. In an effort to obtain a permanent cell culture, a finely mincedportion of the initial biopsy specimen was established in nude mice4;

subsequently, the tumor was adapted to tissue culture (1). In culture,the cells appeared pleomorphic with 10% giant cells. Cytogenetic studiesdemonstrated at least 2 cell populations, one diploid and the otherpolyploid. Only human chromosomes were noted, and polymorphic enzyme analyses showed human enzymic activities that were unique tothis cell line (1).

SK-Luci-6. The SK-Luci-6 cell line was derived from a 41-year-old

white female (Patient 2) with an anaplastic lung carcinoma. During herclinical course, the patient was noted to have persistent leukocytosis aswell as hypercalcemia (range, 11.0 to 17.1 mg/dl). To establish a tumorcell line, minced tumor fragments from a biopsy specimen of the lunglesion were overlaid onto a feeder layer of human fibroblasts (1). Thecells established in culture showed a relatively uniform cell size and werenoted by cytogenetic analyses to be hyperdiploid; only human chromosomes were seen. The morphological, histochemical, cytokinetic, andgenetic properties of the cell lines have been described previously (1).For both cell lines, the morphology of the tumors grown in the nudemouse model and the cell lines in tissue culture were found to beessentially identical to that of the original biopsy specimens. Both celllines were shown to be free of Mycoplasme contamination by culture aswell as by electron microscopic examination. Both cell lines showedanchorage-independent growth by growing continuously while sus

pended in 0.3% agarose gel.

Heterotransplantation

Heterotransplantation of tumor cells was carried out by injecting 1 to3 x 106 viable tumor cells or tumor fragments s.c. into 6-week-old

athymic nude (BALB/c) mice. Tumor engraftment and growth in the nudemice occurred routinely for both cell lines. The passaged tumors could

1This work was supported in part by grants from the American Cancer Society(PDT-60) and the USPHS through National Cancer Institute Grants CA 3072, CA06882, and CA 08748.

2To whom requests for reprints should be addressed, at Memorial Sloan-

Kettering Cancer Center, 1275 York Ave., New York, N. Y. 10021.Received July 21.1982; accepted June 8,1983.

3The abbreviations used are: PG, prostaglandin; PTH. parathyroid hormone;CSA, colony-stimulating activity; CFlk:, colony-forming unit-culture; BMC, bonemarrow cells; TLC, thin-layer chromatography; PGE, prostaglandin E; PGE,, prostaglandin Ei; PGE2, prostaglandin E2, PGFa,, prostaglandin F;....

4T. Kameya, unpublished observations.

SEPTEMBER 1983 4511

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

R. S. Bookman et al.

be shown to have the same morphological pattern as did the originalcultures and biopsy specimens (1). At various times during tumor growth,blood samples were obtained. Serum samples were assayed for CSA asdescribed below, and differential cell counts were determined by cytolog-

ical smears of peripheral blood.

Granulocyte-Macrophage Colony Formation

The ability of tumor cell-derived factors to initiate the proliferation anddifferentiation of committed granutocyte-macrophage precursor cells(CFu-c) was tested in a soft-agar assay system which has been describedpreviously (10,11). Briefly, 10s normal human or mouse (C57BL/6; The

Jackson Laboratory, Bar Harbor, Maine) BMC were suspended inMcCoy's Medium 5A supplemented with 10% fetal calf serum, 5% pooled

human serum, essential and nonessential amino acids, and vitamins(Grand Island Biological Co., Grand Island, N. Y.) to which 0.3% agar(Bacto-Agar; Difco Laboratories, Detroit, Mich.) was added. This cellsuspension was overlaid onto a 1- to 2-mi feeder layer containing 0.5%agar. Aliquots of serum, particle-free cell sonicates, and dialyzed ornondialyzed cell-free conditioned media were added to the feeder layer

prior to addition of the BMC suspension. The plates were placed in a37°humidified atmosphere of 10% CO2. After 14 days of culture, the

plates were scored for the number of CFU-c colonies containing more

than 50 cells/colony. Clusters are defined as grouped cells containingless than 50 cells/group.

PG Synthesis

PG synthesis by the tumor cell lines was tested by measuring theability of the cells to convert labeled precursor into PG and by radtoim-

munoassay.Radiola bel Conversion. Briefly, 106 tumor cells were incubated with

|-'H|arachidonic acid ([5,6,8,9,11,12,14,15-3H]arachtdonic acid; New

England Nuclear, Boston, Mass.). Arter 1 to 2 hr the amount of r>l]-

arachidonic acid incorporated was determined, and then the cells werewashed free of nonincorporated label. The cells were placed in freshmedia, and the release of radiolabeled products was followed as described previously (3, 5). LJpids in cell-conditioned media and cells were

extracted by the method of Folch (5). The products of PG synthetasewere identified by 2-dimensional TLC on silica gel plates (Analtech, Inc.,Newark, Del.) and by high-performance liquid chromatography. The

distribution of arachidonic acid in cellular lipids was determined by one-

dimensional TLC.Radioimmunoassay. PGE levels of cell-conditioned media were de

termined as described previously (3, 9). The assay does not distinguishbetween PGE, and PGE, and shows less than a 5% cross-reactivity with

the prime metabolites of PGE or with PGFa„.

Bone Resorption

The bone-resorbing activities of the tumor lines were tested in an

explanted fetal rat bone system (6). Fetal rat bones were labeled in uteroby treating the mother at 18 days of gestation with an injection of 0.2 to0.5 mCi of 4SCaCI2.The fetal forelimb bones were removed at 20 daysand placed in culture on stainless steel rafts bathed in Bigger's medium

(Fritton-Jackson modification) containing 5% fetal calf serum (GrandIsland Biological Co.). After an 18-hr equilibration period, the supema-

tants were removed, and fresh media with or without test substanceswere added to paired bones. After an additional 48 hr of culture at 37°

in a 5% humidified CO2 atmosphere, the supematants were removed tomeasure ^Ca release. The data are expressed as the ratio of cpm

released from experimental bones (£)divided by the cpm released frompaired control bones (C); each reported value represents the mean E/Cof at least 4 pairs of explanted bones ±S.E.

Statistical AnalysisStudent's f test to compare means was carried out using the signifi

cance limits of a 2-tailed test.

RESULTS

Clinical Course. There was rapid progression of disease inboth patients, this being marked in Patient 2. Neither patientshowed significant response to therapy. Both patients manifested a persistent leukocytosis in the absence of any evidenceof intercurrent infection (Chart 1); persistently higher WBC wereseen in Patient 2. The serum calcium level in Patient 1 remainednormal during the course of his disease. Further, he was withoutany evidence of flushing, asthma, or diarrhea. Patient 2 waspersistently hypercalcémie, and the serum calcium level wascontrolled by repeated administration of mithramycin.

Heterotransplantation. Engraftment of Lu-65 and SK-Luci-6

1A

WBC

(«10")

5 6

Months

1B

WBCt»10«)

4 9Months

1B'

100

50

- Mo

Sag

456Months

Chart 1. WBC (A, B) and differentials (A'. B') during clinical course for Patient

1, from whom Lu-65 was derived (A), and Patient 2, from whom SK-Luci-6 wasderived (B). Mo, monocytes; Ly, lymphocytes; Band, band neutrophils; Seg, segmented neutrophils.

4512 CANCER RESEARCH VOL. 43

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Biological Activities in Human Cell Lines

in nude mice was followed weekly. Beginning at the fourth weekposttransplant, 3 of 3 animals bearing Lu-65 tumors showed a

progressive increase in measurable tumor size. The increase intumor size was associated with a progressive increase in WBCthat represented an absolute increase in polymorphonuclear cells

2A

x 10swee

10,000

Tumorsize(mmS)

234567

Weeks

2A'

/Oinoi\j\j50MoLym_Seg1/f.Õ•:•¡!;;—

'ÌSYti¡:;,--'1u.Ij:iÎI::i1"~

~ ~-11i;:I••'s\1I•:•

*li i ii34567

ControlsWeeks

2BivrvruWBCU

IO»)(

ÕIOOinf

• ~*

/////;//

j*•-/——̂_——*^^tlililílUtUVWTumorsize

(mm»)(

j1000inn

2345

Weeks

2B'

100

60

20

Mo

Lym

Seg

•—f1

I4

1 Controls5 6

WeeksChart2. WBC (A, B), tumor volume (A, B), and differentials (A', B') for nude

mice bearing tumors derived from Lu-65 (A) and SK-Luci-6 (B) cell lines. Mo,

monocytes; Lym, lymphocytes; Seg, segmented neutrophils.

Table 1Colony-forming activity" in serum6 of tumor-bearing nude mice

Lu-65SK-Lua-6Aliquot

GJ|)20SO2050Clusters31+3°63±10d137

±8"228±14"Colonies3101648+

1±1±

3?±5'N66g6

Control plates (noserum)

Control serum 2050

25 ±7

36±1039±4

0.7 ±17 ±2

* Numbers of granulocyte-macrophage colonies containing 50 cells measuredat 14 days in soft-agar suspension of 2 x 10s BMC from C57BL/6 mice in 1.0 ml0.3% agar containing McCoy's Medium 5A supplemented with 10% fetal calf serum

and 5% human serum.Serum from tumor-bearing nude mice sacrificed 5 to 7 weeks following tumor

transplantation.0 Mean ±S.E." Two-tailed Student's f test of means compared to control serum, p < 0.001.e Two-tailed Student's f test of means compared to control serum, p < 0.05.

Two-tailed Student's f test of means compared to control serum, p < 0.01.

(Chart 2A), Tumor excision in 2 of 2 animals with tumor volumesof 10,000 cu mm was associated with a rapid fall in WBC from270,000 to 10,000 cells/cu mm. Tumor growth was rapid in theanimals successfully engrafted with the SK-Luci-6 cell line. Within

3 weeks, large tumors were noted in 3 of 3 animals. Persistentleukocytosis was noted during tumor engraftment characterizedby a progressive absolute monocytosis (Chart 28). Abscessformation was not evidenced on the tumors examined at excisionor postmortem. Thus, tumor engraftment in the nude micerecapitulated the paraneoplastic syndrome seen in the patients;in addition, SK-Luci-6 appeared to be the faster-growing tumor,

as was evidenced by the more malignant course of Patient 2.CFU-c. Serum samples from tumor-bearing animals 5 to 7

weeks posttransplantation were tested for CSA in the soft-agarculture system. Serum samples from tumor-bearing nude mice

were found to increase the number of cell clusters and coloniesarising from committed granulocyte precursors in normal mouse(C57BL/6) BMC when compared to control sera (Table 1). Thedifferences were significant at all doses tested for the SK-Luci-6

samples but only for cluster formation and at the higher serum(50-iil) dose for Lu-65. It should be noted that greater CSA wasdemonstrated in the serum of mice bearing SK-Luci-6 tumors

and correlates with the higher WBC noted in the patient (Patient2) from whom the latter cell line was derived.

To determine if the tumor cells released CSA, lysed tumorcells and media conditioned by the tumor cell lines were testedin the soft-agar culture system. Approximately 2 to 4 x 105

tumor cells were lysed by ultrasonic disruption. Cell debris wassedimented by centrifugation, and the supernatant media weretested for CSA. Tumor cell lysates showed significant colony-forming activity (Table 2). The SK-Luci-6 cell line proved to havemore CSA when compared to Lu-65.

CSA could be shown to be released into the media by testingcell-conditioned media in the soft-agar culture system (Table 3).SK-Luci-6 cells appeared to release more CSA than did Lu-65;

however, it was noted that the levels of PGE, a known inhibitorof CFU-c, were significantly greater in the Lu-65 cell cultures(Table 3). Dialysis of the cell-conditioned media reduced measurable PGE levels. Unexpectedly, cluster- and colony-formingactivity in Lu-65-conditioned media was reduced following dialysis; little effect was noted for SK-Luci-6. Dialyzed conditioned

SEPTEMBER 1983 4513

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

R. S. Bookman et al.

media from Lu-65 were tested in the soft-agar system usingnormal human BMC. The addition of Lu-65-conditioned media tothe human BMC increased CFU-c numbers [5 ±1 (S.E.) colonies

and 126 ±7 clusters] when compared to control samples whichproduced no colonies or clusters in the absence of CSA. Condi-

Tabte2Colony-forming" activity of tumor cell sonicates1'

TumorLu-65SK-Luci-6SK-RC-39'SK-RC-/SK-RC-41'Control

platesNo.

of cells(X105)62424444Clusters0272

±14'444

±61427±23491

±44063±

9019

± 3Colonies''14±

724±567±986

±130001

± 0.8N25333339* Numbers of granulocyte-macrophage colonies containing 50 cells measured

at 14 days in soft-agar suspension of 2 x 10s BMC from C57BL/6 mice in 1.0 ml0.3% agarcontaining McCoy's Medium 5A supplementedwith 10%fetal calf serum

and 5% human serum.'' Supematants of centrifugea sonicates of 2 to 4 x 10stumor cells.0 Two-tailed Student's t test of means of cluster formation compared to control,

p < 0.001, for all doses tested." Two-tailed Student's t test of means for colonies compared to control, p <

0.001, for all doses tested.* Mean ±S.E.' Human tumor cell line, derived from patient with renal cell carcinoma and not

associated with neutrophilia.

Table 3Colony-forming" activity in tumor cell-conditioned media''

Nondialyzed"Lu-65SK-Luci-6Control

platesDialyzed"Lu-65SK-Luci-6Control

platesNo.

ofcells

(X105)24242424N736610566610Clusters070

±25'123+19141

±21100±3114

±343

±1883±4186

±27128±2314±

3Colonies6±25±38±218±201

±11±0.53±111

±30p"0.020.050.050.001NS"(0.2)(0.2)0.001PGE(ng)e1.733.460.120.240.300.600.040.08

* Numbers of granuiocyte-macrophage colonies containing 50 cells measuredat 14 days in soft-agar suspension of 2 x 10s BMC from C57BL/6 mice in 1.0 ml0.3% agarcontaining McCoy's Medium 5A supplementedwith 10% fetal calf serum

and 5% human serum.'' Cell-freemedia that had been conditioned for 24 hr by tumor cells; mediaweredialyzedagainst phosphate-bufferedsalineusing membraneswith a M, 8000 cutoff.

Two-tailed Student's f test of meansof cluster comparedto control,p < 0.001."Analyzed by 2-taiied Student's ( test of means for colonies compared to

control.' ' ng of PGEreleasedinto conditioned mediaas measuredby radioimmunoassay.' Mean ±S.E.9 NS, not significant.

tioned media from a variety of tumor cell lines not producingCSA do not cause colony formation when tested in this assaysystem (12). Based on these findings, it would appear that thereis a circulating CSA in the sera of mice bearing the Lu-65 andSK-Luci-6 tumors. The tumor cells contain a CSA and sponta

neously release the activity into the media during cell culture.These data suggest that tumor cell production of CSA wasresponsible for the leukocytosis manifested by the patients andby the nude mice.

PG Synthesis. Both cell lines could be shown to rapidlyincorporate [3H]arachidonic acid into membrane phospholipids.

However, only the Lu-65 cells were capable of converting the| 'H larachidonic acid into PG, and remarkably there appeared to

be a single class of PG produced, PGE, (Table 4). Essentially,background levels of other tritiated PG except for PGE2 werereleased into the media after 4 hr of culture of the prelabeled Lu-65 cells. Cell homogenization of Lu-65 increased total cpm

release as PG, but 5 to 10 times the amount of tritiated PGE2was released. SK-Luci-6 released levels of tritiated PG products

that could not be distinguished from background. Greater than97% of the counts recovered from lysed SK-Luci-6 cells wereidentified as arachidonic acid. The data from 2-dimensional TLCand high-performance liquid chromatography would suggest that

a single PG synthetase complex with a unique isomerase waspresent in the Lu-65 cells. These observations were quantifiablyconfirmed by radioimmunoassay where a dose-dependent increase of PGE2 was measured with increasing Lu-65 cell numberplaced into culture; no changes in PGE.. levels above backgroundlevel contained in the fetal calf serum were noted with the SK-Luci-6 cell line (Table 5).

Bone Resorption. Fetal rat bones cultured with tumor cells ortumor cell-conditioned media from both cell lines caused amarked augmentation in 4~Carelease compared to controls. Forthe Lu-65 cell line, increased !5Ca release was correlated withtumor cell PG synthesis. Coculture of Lu-65 cells (106 cells/culture) with fetal rat bones augmented 4SCarelease (E/C. 2.46

±0.09), and this effect could be abrogated by the addition ofindomethacin (2.5 um) (17) to the cocultures (E/C, 1.36 ±0.02),the difference being highly significant (p < 0.001). Consistentwith these findings, dialysis of Lu-65 media significantly reducedPGE levels in the media and resulted in a significant loss of bone-

resorbing activity (E/C, 1.44 ±0.09, falling to 1.17 ±0.02 afterdialysis; p < 0.025). Media conditioned by SK-Luci-6 causedaugmented 45Ca release from the bone expiants (E/C, 1.37 ±

0.09; n = 6; p < 0.001 compared to control). Bone résorptionby SK-Luci-6 was not due to tumor cell production of PG, since

insufficient PGE levels were measured by radioimmunoassayand little conversion of tritiated arachidonate could be demon-

Table 4

Tritiated PGs released from tumor cell lines

cpm

Tumor*Lu-65Lu-65"

SK-Luci-66-K

PGE,„"89

±18°

58±16PGFa,60

±530 ±28 ±17

1454TXBj20

± 2242 ±21

21 ± 2PGE618

±2141±22

±2142

764e

3PGD235±

4716±283

31 ± 5PGAî206

±146

±1216Arachidonic

acid3791

4691±545±3828Tumor cells (10*) coincubated with 1 /<Ci|:<Hlarachidonicacid for 2 hr; unincorporated label removed by washing; tritiated PG release given as cpm at 4 hr into the

culture media; counts were separated by 2-dimensionalTLC and identified by their comigratJonwith authentic standards." 6-K PGF,.,.6-ketoprostaglandm F,.,; PGF2...prostaglandmF,..; TXBr, thromboxane B,; PGD?,prostaglandmD2;PGA:, prostaglandmA?.c Mean ±S.E.; n = 4; background, 30 cpm.0 ["H]PG production by homogenates of 10* cells; background, 200 cpm.8 More than 95% of the cpm in the PGE zone could be converted to PGA: after acidification.

4514 CANCER RESEARCH VOL. 43

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Biological Activities in Human Cell Lines

TablesPGElevels in tumor cell-conditioned media

No.of cells(x 10s)8 ngPGEreteased/4hr

Lu-65

SK-Luci-6

25

10

25

10

3.697.16

15.90

0.920.911.06

8 Number of cells used to condition media for 24 hr at 37°.

strated (Tables 4 and 5). Media that had been conditioned by 2x 106 SK-Luci-6 cells caused a marked increase in PGE2 from

test bones compared to control bones (27.6 ±3.6 versus 1.5 ±0.2 ng/bone; p < 0.001). The ability of the SK-Luci-6-conditionedmedia to induce endogenous bone PG synthesis resembles theaction of lymphokine preparations but not that of PTH (6).Measurement of PTH levels in the conditioned media and cellhomogenates using a commercially available COOH-terminalassay failed to demonstrate significant levels of immunoassay-

able hormone. Examination of cell homogenates with clonedPTH-specific DNA probe in a sensitive hybridization assay failedto demonstrate PTH mRNA in either cell line.5 Such data suggestthat SK-Luci-6 releases an osteolytic factor that is neither PGE2

nor PTH. The biochemical characterization of this factor is currently being pursued.

DISCUSSION

Paraneoplastic syndromes suggestive of ectopie hormone production have been reported with undifferentiated lung cancers(2, 7, 14). A description of 2 newly established cell lines fromgiant-cell tumors of the lung has recently been reported (1). We

now present information on the functional activities associatedwith these cell lines that may provide insight into the pathophys-iology responsible for the observed clinical syndromes. In bothcases, heterotransplantation of tumor cells to the nude mousemimicked the persistent leukocytosis characteristic of the patients' clinical course. Surgical extirpation of the tumor from the

nude mouse rapidly abrogated the neutrophilia associated withtumor growth. Neutrophilia is a rare occurrence in nude micebearing human tumors (12). Using an in vitro model, it waspossible to show that the plain and dialyzed cell-conditioned

media as well as filtered cell lysates of the tumor cells inducedcommitted granulocyte precursor cell proliferation and differentiation. These data suggest that both tumor lines produced andreleased factor(s) with CSA-like activity. A growth factor isnecessary for induction of CFU-c proliferation; therefore, it isunlikely that the removal of PGE, a known inhibitor of CFU-c

proliferation (10,11 ), would be sufficient to explain the augmentation of colony formation which was observed. Furthermore,increased colony formation was noted to be induced by the SK-Luci-6 line, a cell line which does not produce PGE. Finally,

neither dialysis nor the addition of indomethacin caused anenhancement of colony formation in either cell line. Therefore, itis more probable that the tumors produce a positive stimulus tocell growth. Recent studies have shown that only the CFU-c

committed to the macrophage lineage are inhibited by PGEadministration (13). Thus, it is possible that the CSA released bythe Lu-65 cell line may be specific for granulocyte precursors

since removal of PGE2 did not augment colony formation.Both cell lines rapidly incorporated [3H]arachidonic acid into

membrane phospholipids, and the label was found to distributeamong the major classes of membrane phospholipids. Only theLu-65 cell line was able to convert arachidonate to PG. Surpris

ingly, a single class of PG product was synthesized, PGE2. Sucha finding strongly suggests that a single species of PG synthe-tase exists, one that is associated with a unique PG isomeraseactivity. Since little is known about PG ¡somerases and thecellular controls of isomerase activity, the Lu-65 cell line and its

single PG synthetase provide a rare opportunity to examine theenzyme system and its regulation. To our knowledge, this represents the first report of a human cell line that produces ngamounts of PGE2 per 10* cells.

Both cell lines demonstrated the ability to secrete an osteolyticfactor when studied in the explanted fetal rat bone culturesystem. The activity from the Lu-65 appeared to be closely

correlated to the ability of the cells to release PGE2, sinceinhibition of PG synthetase abrogated the osteolysis induced bythe tumor-conditioned media. PGE2 is known to be a potent

osteolytic agent in the in vitro rat system utilized (4, 8, 16).Therefore, it seems probable that PG production is the primemechanism by which osteolysis was induced by the the Lu-65cell line. In vivo, PGE2 is rapidly converted by the lungs, liver,and kidneys to the prime circulating metabolite 13,14-dihydro-15-ketoprostaglandin E2. We have demonstrated previously thatthe 15-ketoprostaglandin E2 derivative has no osteolytic activity

(4), and this is one possible explanation for the inability of tumorPGE production to cause cancer hypercalcemia in the patient.Unfortunately, neither 15-ketoprostaglandin E2 in the serum nor

the prime urinary metabolite of PGE2 was measured in the patientfrom whom the Lu-65 line was derived. The osteolytic activityreleased by the SK-Luci-6 cell line clearly was not related to

tumor PGE synthesis since the latter cell line could not be shownto convert arachidonate to PG. It is unlikely that ectopie PTHsecretion would account for the hypercalcemia seen in thepatient. Radioimmunoassay of the patient's serum and of tumor

cell homogenates and conditioned media demonstrated barelydetectable levels of immunoreactive material. No PTH mRNAcould be detected in lysed tumor cell preparations.5 Further

characterization of the osteolytic factor derived from SK-Luci-6is now in progress. It is of interest to note that Lu-65 releaseslarge amounts of material that cross-reacts in parallel fashionwith specific antibody to the a-subunit of the pituitary glycopro-teins and human chorionic gonadotropi.6

In 2 recent reports, patients with squamous cell carcinomasassociated with leukocytosis and hypercalcemia were described(12,15); at present, no explanation exists for the coincidence ofthese paraneoplastic syndromes. In vitro studies with the Lu-65and SK-Luci-6 may provide some insight as to whether a unique

factor or several factors mediate the pathophysiological changesseen in these patients.

In summary, 2 new cell lines derived from human anaplasticlarge-cell lung cancers have been established. These cell lines

demonstrate several functional activities when tested in vitro andin vivo. The latter activities could explain the paraneoplasticclinical syndromes described in the patients from whom thetumors were derived and thus provide unique models for study.

5E. Simpson, G. R. Mundy, R. S. Bookman er al., manuscript in preparation. * R. S. Bockman er al., unpublished observations.

SEPTEMBER 1983 4515

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

R. S. Bookman et al.

ACKNOWLEDGMENTS

The technical assistance of R. Ferguson is acknowledged with gratitude.

REFERENCES

1. Anger, B., Bockman, R S . Andrée«,M., Ertandson, R., Jhanwar. S., Kameya,T., Saigo, P., Wright, W., Beattie, E. J., Oettgen, H. F., and Old, L. J.Characterization of two newly established human cell lines from patients withlarge cell anaplastia lung carcinoma. Cancer (Phila.), 50:1518-1529,1982.

2. Asano, S., Urabe, A., Okabe, T., Sato, N Kondo, Y., Ueyama, Y., Chiba, S.,Ohsawa, N., and Kosaka, K. Demonstration of granulopoietic factors in plasmaof nude mice transplanted with a human lung cancer and in the tumor tissue.Blood. 49: 845-852, 1977.

3. Bockman, R. S. Stage-dependent reduction in T-cokxiy formation in Hodgkin's

disease. J. Clin. Invest., 66: 523-531,1980.4. Bockman, R. S. Hypercalcenraa in malignancy. Clin. Endocrino!. Metab.. 9:

317-333,1980.5. Bockman, R. S. Prostaglandin production by blood monocytes and mouse

peritoneal macrophages: synthesis dependent on in vitro culture conditions.Prostaglandins, 21:9-31,1981.

6. Bockman, R. S . and Repo, M. A. Lymphokine-mediated bone résorptionrequires endogenous prostaglandin synthesis. J. Exp. Med., 754: 529-534,

1981.7. Fusco, F. 0., and Rosen, S. W. Gonadotropi producing anapiasiic large cell

carcinomas of the lung. N. Engl. J. Med., 275: 507-515,1966.8. Klein, D. C., and Raisz, L. C. Prostaglandins: stimulation of bone résorptionin

tissue culture. Endocrinology, 86:1436-1440,1970.9. Kurland, J. I., and Bockman, R. S. Prostaglandin E production by human blood

monocytes and mouse peritoneal macrophages. J. Exp. Med., 147:952-957,

1978.10. Kurland, J. I., Bockman, R. S., Broxmeyer, H., and Moore, M. A. S. Limitation

of excessive myeiopoiesis by the extrinsic modulation of macrophage derivedprostaglandin E. Science (Wash. O. C.), J99: 552-555,1978.

11. Kurland, J. I., Broxmeyer, H., Bockman, R. S., Pelus, L, and Moore, M. A. S.Role for monocyte-macrophage derived colony stimulating factor and prosta

glandin E in positive and negative feedback control of myeioid stem cellproliferation. Blood, 52:388-407,1978.

12. Ohsawa, N., Ueyama, Y., Monta, K., and Kondo, Y. Heterotransplamation ofhuman functioning tumors to nude mice. In: T. Nomura. N. Ohsawa, N.Tamaoki, and K. Fujieana (eds.), Proceedings of the Second InternationalWorkshop on Nude Mice, pp. 395-404. Tokyo: University of Tokyo Press,1977.

13. Pelus, L. M., Broxmeyer, H. E., Kurland, J. I., and Moore, M. A. S. Regulationof macrophage and granulocyte proliferation: specificities of prostaglandin Eand lactoferrin. J. Exp. Med., 750:277-292,1979.

14. Primack, A., Broder, L. E., and Diasio, R. B. Production of markers bybronchogenic carcinoma: a review, in: M. J. Strauss (ed.), Lung Cancer, p. 35.New York: Gruñeand Stratton, 1977.

15. Saito, K., Kuratomi, Y., Yamamoto, K., Saito, T., Kuzuya. T.. Yoshida. S.,Moriyama. S., and Takahashi, A. Primary squamous eel carcinoma of thethyroid associated with marked leukocytosis and hypercalcémie. Cancer(Phila.), 48: 2080-2083,1981.

16. Tashjian, A. H., Voeikei, E. F., Levine, L, and Goldhaber, B. Evidence thatbone resorption-stimuiating factor produced by mouse fibrosarcoma cells isprostaglandin E,. J. Exp. Med., 736:1329-1343,1972.

17. Vane, J. R., Flower, R. J., and Salmon, J. A. Inhibitors of aracnidonic acidmetabolism, with especial reference to the aspirin-Ike drugs. In: T. J. Powies,R. S. Bockman, K. V. Honn, and P. RamweR (eds.), Prostaglandins and CancerFirst International Conference, pp. 21-46. New York: Alan R. Liss, Inc., 1982.

4516 CANCER RESEARCH VOL. 43

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1983;43:4511-4516. Cancer Res   Richard S. Bockman, Anne Bellin, Mary A. Repo, et al.   Lines Derived from Anaplastic Lung Cancers

Biological Activities of Two Human Cellin Vitro and In Vivo

  Updated version

  http://cancerres.aacrjournals.org/content/43/9/4511

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/43/9/4511To request permission to re-use all or part of this article, use this link

on August 6, 2021. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from