13
Somatic Cell Genetics, Vol.8, No. 1, 1982, pp. 1-13 Carcinoembryonic Antigen (CEA) Expression in Somatic Cell Hybrids 1 Denise Sheer, 2 Ruth M. Brown, 3 and Martin Bobrow 4 Genetics Laboratory, Departmentof Biochemistry, South Parks Road, Oxford, OX1 3QU, England Received 22 May 1981--Final 14 September 1981 Abstract--Five hybrids (LSB) were formed between LS174T, a human CEA-producing colonic tumor cell line, and BU25.CAPR, a HeLa derivative which does not produce CEA. All five hybrids produce CEA, but less per cell than LS174 T. Approximately 10 % of the chromosomes have been lost from these hybrids. In an attempt to map the gene(s) coding for the protein moiety of CEA, 7 LSPG and 28 LSR hybrids were formed between LS174T and PG19, a mouse melanoma cell line, and LS174T and RAG, a mouse kidney adenocarcinoma cell line, respectively. These hybrids retain between 4 and 21 human chromosomes, and each human chromosome is represented in at least seven hybrids. Two hybrids appeared to produce trace amounts of CEA. These results might represent repression by the mouse genome of CEA production or the production of a structurally abnormal CEA molecule. INTRODUCTION The genetic control of malignancy has been studied by examining the tumorigenicity of somatic cell hybrids. These experiments have shown that malignancy is suppressed in some hybrids (1), but not in others (2), and have been interpreted as evidence for the recessive or dominant nature of malig- tSubmitted by D.S. in partial fulfillment of the requirements for the D. Phil degree, Oxford University. 2present address: Imperial Cancer Research Fund, P.O. Box 123, Lincoln's Inn Fields, London, WC2A 3PX, England. 3Present address: Department of Paediatrics, Royal Children's Hospital, Parkville, Victoria, Australia 3052. *Present address: Institute of Human Genetics, University of Amsterdam, Sarphatistraat 217, Amsterdam 1018 BX, The Netherlands. 1 0098-0366/82/0100-0001503.00/0 1982 Plenum Publishing Corporation

Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

Embed Size (px)

Citation preview

Page 1: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

Somatic Cell Genetics, Vol. 8, No. 1, 1982, pp. 1-13

Carcinoembryonic Antigen (CEA) Expression in Somatic Cell Hybrids 1

Denise Sheer, 2 Ruth M. Brown, 3 and Martin Bobrow 4

Genetics Laboratory, Department of Biochemistry, South Parks Road, Oxford, OX1 3QU, England

Received 22 May 1981--Final 14 September 1981

Abstract--Five hybrids (LSB) were formed between LS174T, a human CEA-producing colonic tumor cell line, and BU25.CAP R, a HeLa derivative which does not produce CEA. All five hybrids produce CEA, but less per cell than LS174 T. Approximately 10 % of the chromosomes have been lost from these hybrids. In an attempt to map the gene(s) coding for the protein moiety of CEA, 7 LSPG and 28 LSR hybrids were formed between LS174T and PG19, a mouse melanoma cell line, and LS174T and RAG, a mouse kidney adenocarcinoma cell line, respectively. These hybrids retain between 4 and 21 human chromosomes, and each human chromosome is represented in at least seven hybrids. Two hybrids appeared to produce trace amounts of CEA. These results might represent repression by the mouse genome of CEA production or the production of a structurally abnormal CEA molecule.

I N T R O D U C T I O N

The genetic control of malignancy has been studied by examining the tumorigenicity of somatic cell hybrids. These experiments have shown that mal ignancy is suppressed in some hybrids (1), but not in others (2), and have been interpreted as evidence for the recessive or dominant nature of malig-

tSubmitted by D.S. in partial fulfillment of the requirements for the D. Phil degree, Oxford University.

2present address: Imperial Cancer Research Fund, P.O. Box 123, Lincoln's Inn Fields, London, WC2A 3PX, England.

3Present address: Department of Paediatrics, Royal Children's Hospital, Parkville, Victoria, Australia 3052.

*Present address: Institute of Human Genetics, University of Amsterdam, Sarphatistraat 217, Amsterdam 1018 BX, The Netherlands.

1

0098-0366/82/0100-0001503.00/0 �9 1982 Plenum Publishing Corporation

Page 2: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

2 S h ~re ta l .

nancy. Klinger et al (3) and Klinger and Shows (4) have demonstrated that no human chromosome in single copy is able to suppress tumorigenicity in their hybrids between nontumorigenic human diploid fibroblasts and tumo- rigenic CHO cells but that certain combinations of human chromosomes are able to do so.

Experiments concerning the genetic control of malignancy are often difficult to evaluate since most of the properties studied, such as tumorigenic- ity in the nude mouse (5), probably represent the effects and regulation of many genes. We therefore wished to investigate, using somatic cell hybridiza- tion, a marker of malignancy which would be likely to represent a primary gene product.

Carcinoembryonic antigen (CEA) is a glycoprotein of molecular weight 200,000 which is produced in large amounts by human colonic tumors and fetal colon (6, 7) but only in trace amounts by normal adult colon (8). It is often produced in moderate amounts by other tumors such as breast and urogenital tumors (9).

We report the study of CEA expression in intraspecific somatic cell hybrids between a human CEA-producing cell line and a HeLa derivative which does not produce CEA. We have also formed interspecific hybrids between the human CEA-producing line and mouse cell lines in an attempt to map the gene(s) coding for the protein moiety of CEA.

MATERIALS AND METHODS

LS174T. This epithelial-like cell line was derived from a moderately well-differentiated colonic adenocarcinoma (10, 11). As measured by radioimmunoassay (RIA), 106 confluent LS174T cells secrete approximately 80 ng CEA/ml /day (see Table 3). Chromosome analysis demonstrated a modal chromosome number of 48 with a range of 46-53 in 10 cells analyzed. Every cell is trisomic for chromosome 7, and some cells have an additional copy of chromosome 1, 13, and 15. No abnormal chromosomes are present. LS 174T expresses the B form of G6PD and type 2 of PGM-3, measured by starch gel electrophoresis by Dr. S. Povey, who carried out all the enzyme analyses.

BU25.CAP R. This epithelial-like cell line was produced by Ms. E. Munro from a fusion of BU25, a TK-deficient derivative of the HeLa $3 cell line (12), with cytoplasts of MC63, a chloramphenicol-resistant derivative of the HeLa B cell line (13). BU25.CAP R is resistant to 30 #g BrdU/ml and 50 #g chloramphenicol/ml. It possesses several HeLa marker chromosomes (Fig. 1) and has a modal chromosome number of 59, with a range of 57-62 in 10 cells analyzed. It expresses the A form of G6PD and type 1 of PGM-3, as measured by starch gel electrophoresis. CEA was not detected in 3- to 7-day

Page 3: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

CEA in Hybrids 3

Fig. 1. G-banded karyotype of BU25.CAP R. Marker chromosomes which appear in Fig. 2 are labeled M1-M5.

spent medium samples, nor in cell sonicates of BU25.CAP R in radioimmu- noassays with a CEA detection limit of 2 ng/ml (Table 4).

PG19. This is an HPRT-deficient mouse cell line derived from a transplantable melanoma in the C57/6J strain (14). CEA was not detected in 6-day spent medium from PG19 in four assays with a CEA detection limit of 2 ng/ml.

RAG. This is an HPRT-deficient mouse cell line derived from a renal adenocarcinoma in the BALB/cd strain (15). CEA was not detected in 6-day spent medium from RAG in four assays with a CEA detection limit of 2 ng/ml.

Tissue Culture. Cells were routinely grown in RPMI 1640 supple- mented with 10 or 20% fetal calf serum (FCS).

Page 4: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

.SB

16

*.

I 2

3 4

5

; 7

8 9

l0

. 2

(

3 14

15

16

17

8

9 2

0

21

22

nam

er

:hm

mso

mes

~ ml

M2

M3

m

M5

"ira

2.

G-b

ande

d ka

ryot

ype

of L

SB

16.

Mar

ker

chro

mso

mes

whi

ch a

ppea

r in

Fig

. 1

are

labe

led

M l

-M5

. 0 m

Page 5: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

CEA in Hybrids 5

Fusions. Fusions were carried out in suspension using 50% polyethylene glycol (PEG) 1500 according to the method of Davidson et al. (16). Distinct and well-separated colonies were picked approximately 2 weeks after fusion using bent Pasteur pipets. Intraspecific hybrids were selected in medium containing hypoxanthine (10 -4 M), methotrexate (10 -5 M), and thymidine (1.6 x 10 -5 M) (HMT medium) and chloramphenicol (50/~g/ml) starting 4 days after fusions. Control cultures of LS174T and BU25.CAP R were set up in selective medium simultaneously with each fusion. Interspecific hybrids were selected in HMT medium supplemented with 100 ~M ouabain for 4 weeks after the fusions and maintained thereafter in HMT medium.

Chromosomes. The hybrid status of clones derived from the intraspecific fusions was confirmed using G (Fig. 2) and Q banding (17, 18). Human chromsomes in selected interspecific hybrids were identified by the sequential use of the G-11 (19) and Q-banding techniques.

Enzymes. The hybrid status of clones derived from the intraspecific fusions was also confirmed by enzyme analyses.

CEA Assay. Spent medium samples and cell sonicates were subjected to radioimmunoassay (RIA) (20). In this assay, the CEA being measured

Bound 100, j BU25.CAP R (7days)

9080 """ ~ ~ LSB 46 (3days)

70 ~ / 6 (3days) 60 50 40 LSB 44 (lOdays) 30 " ~ LS174T (201~urs)

LSB 20 ~ -,looa,s, 1 .A lOdays)

I I I I I 5 10 50 1 0 0 - 500

CEA (ng/ml) Fig. 3. Radioimmunoassay with 6G6 antiserum of LSB hybrids, LS174T, and BU25.CAP R. LSB hybrids were assayed at approximately 3 x 105 cr LS174T and BU25.CAP R were assayed at approximately 8 x 105 cells/ml. The CEA concentration in each sample was determined from the intersection of the "percent binding" of the sample with the standard c u r v e .

Page 6: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

6 S h ~ r e t a i .

competes with a fixed amount of [~25I]CEA for binding to a limiting constant dilution of anti-CEA antiserum. Free [125I]CEA is separated from [~25I]CEA bound to antibody by precipitating the antigen-antibody complexes with PEG, and the precipitate counted in a gamma counter. The amount of CEA in the sample is then determined from a standard curve set up in the same assay with increasing concentrations of unlabeled CEA (Fig. 3). Medium samples and cell sonicates were assayed in triplicate. All hybrids were assayed with two anti-CEA antisera (see below), raised in goats, at a dilution of 1:2000 to 1:6000.

The 6G6 antiserum was obtained from the 6th bleed of an immunized goat. Antiserum obtained from the 3rd bleed of the same goat has been reported to react with the carbohydrate moiety of CEA and also with the related nonspecific cross-reacting antigen (NCA) (21, 22).

Ace (pool 63-66) antiserum reacts mainly, if not completely, with the protein moiety of CEA but not with NCA (21, 23).

Preparation of Medium Samples. Cultures were grown to confluence and the medium replaced to give an approximate cell concentration of 3 x 105 and 2 x 106 cells/ml for the intraspecific and interspecific hybrids, respec- tively. Then 3-15 days later, spent medium was centrifuged in a Beckman microfuge for 4 min to remove any cellular debris, and aliquots were frozen until assay. Cultures were assayed when confluent as both LS 174T and LoVo cell lines have been shown to synthesize more CEA during the stationary growth phase than during active cell proliferation (24, 25).

Preparation of Cell Sonicates. To 50 mg of pelleted cells 400 ~1 R P M I 1640 and 10% FCS was added. The cells were then sonicated for 2-3 min with a Fisons sonicator and the suspension centrifuged in a Beckman microfuge for 4 min. The supernatants were collected and diluted 1:5 to 1:125 for assay.

RESULTS

Intraspecific Fusions. Ninety-one colonies were picked from three fusions but most died within 4 weeks of fusion. Five viable clones, called LSB

Table 1. Enzyme Analyses of LS174T, BU25.CAP R, and LSB Hybrids ~

Cells G6PD type PGM-3 type

LS 174T B 2 BU25.CAP R A 1 LSB 16 A/B 2-1 LSB 2O A/B 2-1 LSB 44 A/B 2-1 LSB 45 A/B 2-1 LSB 46 ?A 2-1

~ 16, 20, 44, and 45 each expressed the A and B and hybrid A/B forms of G6PD.

Page 7: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

CEA in Hybrids 7

Table 2. Chromosome Analyses a of LS174T, BU25.CAP R, and LSB Hybrids

Modal Number of chromosome chromosomes

Cells number Range lost b

LS174T 48 l 46-53 l total = 107 103-115

BU25.CAP R 59J 57~i2J LSB 16 90 88-91 17 LSB 20 94 85-96 13 LSB 44 102 94-107 5 LSB 45 104 97-109 3 LSB 46 95 93-99 12

~ 10 cells of each of the parental lines and the hybrid clones were counted. bFrom total of modal chromosome numbers of LS174T and BU25.CAP R. The parental origin of

the segregated chromosomes could not be determined.

16, 20, 44, 45, and 46, were obtained. LSB 16 and 20 were obtained from two separate fusions. LSB 44, 45, and 46 were obtained from one flask of the third fusion, The morphological characteristics of LSB 44 and 45 are very similar: both are epithelial-like and fairly adherent. LSB 46 is a less firmly attached fibroblast-like line. The cells in the five LSB clones were large, and no smaller cells were seen in these cultures.

Enzymes and Chromosomes of lntraspecific Hybrids. Enzyme analyses are summarized in Table 1. These show that the five LSB clones are hybrids between LS174T and BU25.CAP R.

Chromosome analyses of the LSB clones are shown in Table 2. BU25.CAP R marker chromosomes were present in all five LSB clones (e.g., Fig. 2), confirming that they are hybrids. No near-diploid cells were seen in the chromosome preparations; at least 50 metaphases of each LSB clone were examined. All cells examined contained BU25.CAP R marker chromosomes, excluding the possibility of major contamination with LS 174T hybrids. Thus, at least three and possibly five independent hybrids were obtained in this series.

RIA oflntraspecific Hybrids. Table 3 shows the CEA values obtained: (1) for the spent medium samples standardized to ng/ml/106 cells/day; and (2) for the cell sonicates standardized to a 1:25 dilution of the original

sonicates and on a "per cell" basis by assuming that, considering the relative chromosome numbers and cell size, the LSB sonicates contain approximately half as many cells as the LS174T sonicates. The CEA values for the LSB sonicates have therefore been doubled. These results show that all five LSB hybrids produce substantial

quantities of CEA. lnterspecific Fusions. Seven hybrid clones called LSPG were isolated

from one fusion between LS174T and PG19. Twenty-eight hybrid clones called LSR were isolated from two fusions between LS174T and RAG.

Page 8: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

8 Sheer et al.

Table 3. RIA of Intraspecific Hybrids

Medium samples ~ (CEA ng/ml/106 cells/day)

Cell sonicates b (CEA ng/ml adjusted

to 1:25 dilution)

6G6 ACE ACE 6G6 ACE ACE Assay no. I c 2 3 4 5 6

Cells LSB 16 20 17 230 140 152

13 LSB 20 17 23 30 320 256 LSB 44 17 3 90 70 40 LSB 45 33 26 140 200 480 LSB 46 t 7 17 40 100 240 500

26 LSI74T d 80 _+ 10 86 _+ 8 350 500 608 BU25.CAP R <2 <2 5 <2 <2 <5

5

QThe same medium sample was used in assays 1 and 2. A different medium sample was used in assay 3.

bThe same sonicate of LSB 20 was used in assays 4, 5, and 6. For the other clones, one sonicate was used in assays 4 and 5, and a different sonicate was used in assay 6.

cSee Fig. 3; the numbers are assay numbers. d Results for medium samples of LS 174T (mean +_ standard deviation) are taken from four assays with 6G6 antiserum and four assays with ACE antiserum.

Chromosomes of lnterspecific Fusions. The h u m a n chromosomes pres- ent in LSPG and LSR hybrids examined are summarized in Table 4.

RIA of Interspecific Hybrids. Results of assays of spent medium samples are summarized in Table 5. Al though the results obtained with 6G6 ant iserum for the older cultures of LSPG 1, LSR 9, and LS R 39 were close to the CEA detection limit, which was conservatively estimated, they were consistently well above those for the other hybrids. Subcloning and the use of a fluorescence-activated cell sorter were unable to define subpopulations of CEA-producing cells in these clones. The borderl ine positive results are therefore due either to a part ial ly cross-reacting product or to very small amounts of CEA.

Several hybrid clones, PG19, and R A G were reassayed for CEA by Prof. S. von Kleist, Freiburg University, using a highly sensitive enzyme-l inked immunoadsorbent assay kit (EIA; Abbot t Laboratories). The results are shown in Table 6. They suggest that LSPG 1 and LS R 9 produce a substance which is not CEA, but which cross-reacts weakly with our 6G6 ant iserum. Subcloning did not define subpopulat ions of CEA-producing cells in LS R 8

and 39. Cell sonicates of L S P G 1, and LSR 4, 7, 8, 31, 34, and 39 were assayed

at 1:5, 1:10, and 1:125 dilutions with 6G6 and A C E antisera to investigate the possibility of C E A being produced in reasonable amounts but not being

Page 9: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

W"

Tab

le 4

. C

hrom

osom

e A

naly

ses

of I

nter

spec

ific

Hyb

rids

Cel

ls c

onta

inin

g h

um

an c

hrom

osom

es (%

) T

rans

lo-

No.

of

cati

ons

cell

s C

lone

1

2 3

4 5

6 7

8 9

10

11

12

13

14

15

16

17

18

19

20

21

22

X

Y

etc.

an

alyz

ed

Bo

LS

R4

81

10

0 90

72

54

10

0 45

45

54

81

81

72

27

72

72

63

36

72

63

L

SR

6

83

100

100

? 66

92

66

10

0 L

SR

7

50

50

20

80

70

60

80

50

10

80

100

LS

R8

56

89

89

67

33

56

33

8

94

4

89

22

89

89

1!

100

56

33

22

78

100

78

LS

R9

64

50

57

57

7

78

100

43

50

36

14

43

57

85

92

29

92

78

78

LS

R1

0

71

71

71

56

71

LS

RI1

10

0 83

66

33

66

66

75

25

75

33

41

91

91

50

75

10

0 83

91

25

10

0 66

91

L

SR

16

30

10

0 60

70

50

20

70

40

90

L

SR

18

91

91

91

91

54

18

27

63

54

36

54

63

72

54

100

82

91

LS

R2

1

89

67

89

100

44

67

56

56

67

56

67

56

67

89

56

44

33

56

78

89

89

78

LS

R 2

5 27

36

64

45

10

0 82

L

SR

30

55

89

55

100

LS

R3

1

8 38

69

31

15

85

38

46

7

7 15

7

7 7

7 85

L

SR

34

91

32

10

0 91

82

55

37

73

55

64

10

0 36

82

64

73

55

82

10

0 L

SR

38

78

67

11

11

44

L

SR

39

90

10

90

10

70

90

80

20

60

80

90

80

50

90

60

90

40

40

70

20

50

10

0 L

SR

40

10

0 70

10

10

0 10

50

70

80

10

20

90

60

50

30

60

10

0 60

L

SP

G 1

10

0 75

8

75

25

84

58

75

LS

PG

2

16

9 55

55

9

45

27

73

64

45

55

16

55

36

55

LS

PG

4

? 42

42

57

85

57

71

71

28

57

?

LS

PG

8

31

7 39

69

62

69

39

7

44

54

39

15

15

62

+ m

arke

rs

+ m

arke

rs

+ m

arke

rs

11

12

10 9 14 7 12

i0

11 9 11 9 13

11 9 10

10

12

11 7 13

Page 10: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

10 Sheer et al.

Table 5. RIA of Spent Medium from Interspecific Hybrids

6G6 ACE

C E A CEA Days in detection CEA detection CEA

Cells culture limit (ng/ml) (ng/ml) limit (ng/ml) (ng/ml)

LSPG 1 12 <12 3 12 <12 5 <5

12 <12 3 10 <10 6 10 12 5 <5

5 5 8 2 4

Other 12 < 12 LSPG 3 12 <12 5 <5 clones 12 < 12

6 10 <10 5 <5 LSR 9 3 12 <12

8 2 2 11 4 4 5 <5 11 2 2 2 <2

LSR 39 3 12 <12 10 2 <2 10 2 <2 14 2 5 5 <5

4 4 Other 3 12 < 12

LSR 6 2 <2 2 <2 clones 15 2 <2 2 <2

PG19 6 12 <12 2 <2 2 <2 2 <2 2 <2

RAG 6 12 <12 2 <2 2 <2 2 <2 2 <2

Table 6. RIA of Spent Medium from Interspecific Hybrids

9-day spent medium 5-day spent medium (lyophilized and concentrated a)

Cells CEA (ng /ml) CEA (ng/ml)

RAG <0.1 <0.1 PG19 <0.1 <0.1 LSPG 1 <0.1 <0.1 LSR 4 <0.1 <0.1 LSR 8 0.2 0.6 LSR 9 <0.1 <0.1 LSR 34 <0.1 <0.1 LSR 39 0.1 10.4

aSpent medium was lyophilized, dialyzed, and redissolved in RPMI 1640 + 10% FCS to give a concentration of 75 mg dry weight/ml.

Page 11: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

CEA in Hybrids 11

secreted. CEA was not detected in any of the hybrids in radioimmunoassays with a CEA detection limit of 4 ng/ml.

DISCUSSION

The poor survival rate of the clones picked in the intraspecific fusions may reflect their sensitivity to chloramphenicol. Several groups (26, 27) have reported difficulty in transferring chloramphenicol resistance between human cells of different origin.

All five intraspecific LSB hybrids produce substantial amounts of CEA. CEA production therefore behaves as a dominant function in this situation. Although it is difficult to quantitatively compare CEA production in LS 174T and the LSB hybrids, standardized CEA production (per 106 cells/day) was 80-86 ng/ml for LS174T and 3-40 ng/ml for the LSB hybrids. This suggests a possible partial repression of CEA production in these intraspecific hybrids.

Most of the interspecific human-mouse hybrids show no detectable CEA production. Two hybrids, LSR 8 and 39, possibly produce CEA but at a level below the confidence limits of the assay. Several explanations have been considered (see below) for the apparent massive reduction in CEA production in the interspecific hybrids: repression, production of an abnormal molecule, effect of cell fusion, presence of the incorrect homolog, necessity for more than one human chromosome.

Since each human chromosome is present in at least seven interspecific hybrids, the most likely explanation is that the mouse genome has repressed CEA production. Repression of differentiated functions is a common occur- rence in hybrids between different cell types (reviewed in ref. 28), and CEA might therefore behave as a differentiated function in this regard.

It is also possible that an abnormally glycosylated CEA molecule is produced by some of the interspecific hybrids.

The fusion process itself does not repress CEA production in LS174T since all five intraspecific LSB hybrids continue to produce CEA.

In the intraspecific LSB hybrids CEA production is dominant in the sense that an inactive gene does not repress an active one. CEA production in LS174T may therefore be based on only one active homolog. It is therefore theoretically possible that this active homolog was not retained in any of the interspecific hybrids. However, the extensive representation of each human chromosome in these hybrids argues against the inactive homolog being retained in each case.

It is possible that more than one human chromosome is necessary for CEA production by the hybrids. If this is so, it seems unlikely that none of our

Page 12: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

12 Sheer etal .

hybrids has the correct combination of chromosomes since several hybrids, e.g., LSR 21, contain most of the human chromosomes.

If the low positive results obtained for LSR 8 and 39 are not artifactual, they might reflect incomplete repression of CEA production. Alternately, these two hybrids might produce an abnormally glycosylated CEA molecule which cross-reacts weakly with the anti-CEA antisera used.

Production of an embryonic antigen by a tumor could reflect derivation of the tumor from a stem cell in which the relevant gene(s) remained unrepressed, or from a differentiated cell in which the gene(s) have been derepressed. Dominance of CEA expression in the intraspecific LSB hybrids might therefore suggest that CEA production in LS174T is refractory to repression. Our finding that CEA production has probably been extinguished in the interspecific LSPG and LSR hybrids suggests that the relevant gene(s) in LS 174T are still able to be repressed.

Alternatively, BU25.CAP R might specifically be permissive for CEA production. CEA is produced in 63% of cervical carcinomas (29). Thus absence of CEA production by BU25.CAP R itself might suggest an alteration in or lack of the relevant gene(s) in this line.

The control of CEA expression in LS174T would be elucidated by further crosses between LS174T and other human cells, particularly diploid cells. We have so far been unable to obtain a selectable marker in LS174T which is necessary for these experiments.

If CEA production in the interspecific LSPG and LSR hybrids is repressed by the mouse genome, fusion of LS174T with a permissive mouse cell line may enable the production of hybrids which do produce CEA. Although no selectable mouse intestinal tumor cell line was available, a mouse teratocarcinoma cell line might fulfill this function since CEA is an embryonic as well as a tumor product. We have therefore made a further 35 interspecific hybrids, called LPC hybrids, between LS 174T and PCC4.aza 1. Several LPC hybrids gave positive results in radioimmunoassays with 6G6 and ACE antisera. However, although CEA has not been found in any species besides man, under certain experimental conditions PCC4.aza 1 also appears to produce small amounts of CEA-like material. Provided the human and mouse CEA products can be distinguished, it should be possible to use the LPC hybrids to map the gene(s) for CEA.

ACKNOWLEDGMENTS

We are indebted to Dr. E. Solomon for constructive discussions and advice. We wish to thank Dr. S. Povey for carrying out the enzyme analyses, and Dr. W.F. Bodmer, Dr. P.N. Goodfellow, Prof. S. von Kleist, Dr. M.L. Ellison, Dr. A.M. Neville, and Ms. J.E. Heritage for helpful comments, We

Page 13: Carcinoembryonic antigen (CEA) expression in somatic cell hybrids

CEA in Hybrids 13

are grateful to Dr. B.H. Tom, Ms. E. Munro , Prof. H. Harris, and Dr. P.N. GoodfeUow for LS174T, BU25 .CAPR, PG19, and PCC4.aza 1 cell lines, respectively. We also thank Dr. D.J.R. Laurence and Dr. M.L. Ellison for 6G6 ant iserum and CEA, and Dr. C.W. Todd for A C E ant iserum.

L I T E R A T U R E C I T E D

1. Jonasson, J., and Harris, H. (1977). J. Cell. Sci. 24:255-264. 2. Koprowski, H., and Croee, C.M. (1977). Proc. Natl. Acad. Sci. U.S.A. 74:1142-1146. 3. Klinger, H.P., Bairn, A.S., Eun, C.K., Shows, T.B., and Ruddle, F.H. (1978). Cytogenet.

Cell Genet. 22:245-249. 4. Klinger, H.P., and Shows, T.B. (1979). Cytogenet. Cell Genet. 25:172-173. 5. Freedman, V.H., and Shin, S. (1974). Cell 3:355-359. 6. Gold, P., and Freedman, S.O. (1965). J. Exp. Med. 121:439-462. 7. Gold, P., and Freedman, S.O. (1965). J. Exp. Med. 122:467-481. 8. Fritsche, R., and Mach, J.P. (1977). lmmunochemistry 14:119-127. 9. Hansen, H.J., Snyder, J.J., Miller, E., Vandervoorde, J.P., Moiler, D.N., Hines, L.R., and

Burns, J.J, (1974). Hum. Pathol. 5:139-147. 10. Tom, B.H., Rutzky, L.P., Jakstys, M.M., Oyasu, R., Kaye, C.I., and Kahan, B.D. (1976).

In Vitro 12:180-191. 11. Kahan, B.D., Rutzky, L., Berlin, B., Tomita, J., Wiseman, F., LeGrue, S.J., Noll, H., and

Tom, B.H. (1976). Cancer Res. 36:3526-3534. 12. Kit, S., Dubbs, D.R., and Frearson, P.M. (1966). Int. J. Cancer 1:19-30. 13. Siegel, R.L., Jeffreys, A., Sly, W., and Craig, I.W. (1976). Exp. Cell Res. 102:298-310. 14. Jonasson, J., Povey, S., and Harris, H. (1977). J. Cell. Sci. 24:217-254. 15. Klebe, R.J., Chen, T.R., and Ruddle, F.H. (1970). J. Cell Biol. 45:74-82. 16. Davidson, R.L., O'Malley, K.A., and Wheeler, T.B. (1976). Somat. Cell Genet. 2:271-

280. 17. Seabright, M. (1971). Lancet 2:971-972. 18. Pearson, P.L., Bobrow, M., and Vosa, C.G. (1970). Nature 226:78-80. 19. Bobrow, M., and Cross, J. (1974). Nature 251:77-79. 20. Laurence, D.J.R., Stevens, U., Darcy, D.A., Turberville, C., and Neville, A.M. (1974). Int

A.E.A. 2:275-297. 21. Ormerod, M.G. (1978). Scand. J. lmmunol. 8(Suppl. 8): 433-438. 22. yon Kleist, S., Chavanel, G., and Burtin, P. (1978). Proc. Natl. Acad. Sci. U.S.A.

69:2492-2494. 23. Westwood, J.H., and Thomas, P. (1975). Br. J. Cancer 32:708-719. 24. Goldenberg, D.M., Pavia, R.A., Sharkey, M., and Bennet, S. (1978). In Tumour Markers,

(eds.) Grifliths, K., Neville, A.M., and Pierrepoint, C.G. (Alpha Omega, Cardiff), pp. 29-40.

25. Drewinko, B., and Yang, L.Y. (1976). Exp. Cell Res. 101:414-416. 26. Wallace, D.C., Bunn, C.L., and Eisenstadt, J.M. (1977). Somat. Cell Genet. 3:93-119. 27. Munro, E., Siegel, R.L., Craig, I.W., and Sly, W.S. (1978). Proc. R. Soc. Lond. Ser. B.

201:73-85. 28. Ringertz, N., and Savage, R.E. (1976). Cell Hybrids. (Academic Press, New York). 29. van Nagell, J.R., Jr., Donaldson, E.S., Gay, E.C., Hudson, S., Sharkey, R.M., Primus, F.J.,

Powell, D.F., and Goldenbcrg, D.M. (1979). Cancer 44:944-948.