Cells from normal human breast epithelium were cloned in monolayer culture and the clones were stained with monoclonal antibodies. Tissue was from reduction mammoplasty operations. Cloning efficiences were 5-30%.

Two types of clone were identified: 10 to 30% were of relatively spread cells whose boundaries were often difficult to see by phase-contrast microscopy but where they were visible they appeared as dark lines. The edges of the clones usually appeared to be under tension. These clones were stained by two monoclonal antibodies, LICR-LON-M8 and M24, that stain luminal epithelial cells in the intact tissue, but not myoepithelial or stromal cells. Within a clone the cells showed a full range of antigenic phenotypes. This was confirmed for clones grown from single cells that had been isolated manually. The second type of clone was more compact with little evidence of tension at the edges, and cell boundaries were clearly visible and bright under phase contrast. These clones were not stained by antibodies M8 or M24. Both types of clone stained with a third monoclonal antibody that is specific for luminal epithelial cells in the intact tissue, LICR-LON-M18, but the distribution of staining was different in the different types of clone.

The simplest interpretation of the two types of clone is that luminal epithelial cells give rise to the spread type of clone while the myoepithelial cells give rise to the more abundant and vigorous compact clones. Alternatively, the compact clones may be from luminal epithelial cells that have lost differentiated characteristics.

Adult human mammary epithelium is composed principally of two types of cell, the luminal epithelial cell and the myoepithelial cell. The luminal epithelial cells form a continuous layer surrounding the lumen while the myoepithelial cells are wrapped around them forming a continuous concentric layer in the ducts and a discontinuous layer in the lobuloalveolar units.

Mammary epithelial cells have been cultured in several laboratories from humans or rodents by plating fragments of epithelium, called ‘organoids’, obtained by collagenase digestion of chopped gland (reviewed by Janss, Hillman, Malan-Shibley & Ben, 1980; Yang et al. 1980; see also Easty et al. 1980; Stampfer, Hallowes & Hackett, 1980; Yang, Kube, Park & Furmanski, 1981). These organoids contain the various cell types making up the ductal and lobuloalveolar epithelium (Janss et al. 1980; Stampfer et al. 1980) and can also be contaminated with small numbers of cells from stroma and vessels. Identification of the cells in the cultures has been difficult because epithelial cells show diverse morphology that is not necessarily related to cell types seen in vivo and some of the ultrastructural characteristics used to identify them in the intact tissue tend to be lost rapidly in monolayer culture (Easty et al. 1980). In recent work from this laboratory (Edwards, Brooks & Monaghan, 1984) a series of monoclonal antibodies raised to antigens characteristic of luminal epithelial cells, and which do not bind to myoepithelial or stromal cells (Foster, Edwards, Dinsdale & Neville, 1982; Edwards & Brooks, 1984), were used to stain the surface of short-term primary cultures of normal resting human breast. Two of these antibodies (LICR-LON-M8 and M24, see Table 1) stained patches or sheets of cells arranged in an epithelium-like pavement, which sat on top of the bulk of the cells that spread out from the original tissue fragments. These ‘patch’ cells were difficult to discern by phase-contrast microscopy because they were very flattened with shallow boundaries, and they have escaped attention in most previous studies (Edwards et al. 1984). They are of particular interest because they represent at least one major mammary cell phenotype, i.e. cells expressing luminal antigens.

Table 1.

Monoclonal antibodies used

Monoclonal antibodies used
Monoclonal antibodies used

Recently, methods have been established for cloning human breast cells with high efficiency (Smith et al. 1981; Stoker, Perryman & Eeles, 1982). We report here an analysis of clones grown from normal ‘resting’ human mammary epithelium, using our monoclonal antibodies to detect antigens characteristic of epithelial cells.

Unless otherwise noted, materials for culture were from GIBCO (Paisley, Scotland).

Tissue preparation

Tissue from reduction mammoplasty (cosmetic) operations was prepared for culture by a slight modification of the methods described by Easty et al. (1980). Diced tissue was incubated with stirring for 18-26h at 37 °C in collagenase-containing ‘digestion medium’ consisting of Dulbecco’s Modified Eagle’s Medium (DMEM) with 10mM-HEPES, 50μ;gml−1 penicillin, 50μ;gml−1 streptomycin, 2·5μ;gml−1 amphotericin B (Fungizone, Squibb, Morton, Cheshire) and Iμ;gm l−1 minocyclin (Lederle Laboratories, Gosport, Hants.). Collagenase was type IA (Sigma Chemical Co., St Louis, Mo.) at 0·5mgml−1. Variations between breasts in the consistency of the stroma required some variation in the times of digestion. After digestion the epithelial tissue fragments were recovered by diluting the collagenase-containing medium with fresh DMEM (1:1) and centrifuging at 1000 g for 5 min. The pellet was resuspended in fresh medium and the fragments allowed to settle out at 1g for 18 h at 4°C. They were then treated with digestion medium containing 2·5mgml−1 collagenase type IA for a further 2h to remove residual stromal elements from the epithelium. The organoids, i.e. fragments of ducts and lobuloalveolar units (Stampfer et al. 1980), were then recovered by repeated sedimentation for 15-30 min at 1 g at 4°C, and were essentially free of fibroblasts and vascular cells at this stage. Organoids were then either plated out directly into primary culture, or disaggregated further to yield single cell preparations.

Culture and cloning

Clones and primary cultures of intact organoids were maintained with RPMI 1640 medium containing 10% foetal calf serum, antibiotics as in digestion medium (see above) and lμ;gml−1 hydrocortisone, 1μ;gml−1 insulin and 50ngml−1 ′ cholera toxin (all from Sigma Chemical Corp.) to stimulate proliferation of the epithelial cells (Taylor-Papadimitriou, Shearer & Stoker, 1977a; Taylor-Papadimitriou, Purkis & Fentiman, 1980; Stoker et al. 1982). For non-clonal primary cultures between 100 and 200 organoid fragments were plated per 25 cm2 culture surface. Clones were obtained both from such proliferating primary cultures 4—7 days after plating and from freshly disaggregated organoids. Cell suspensions were prepared by incubation of 25-cm2 cultures, or 100—200 organoids with 5ml of 0·5mgm l−1 bovine pancreatic trypsin (type III S, Sigma) in phoβphate-buffered saline with 0·02% (w/v) ethylene diamine tetraacetic acid (EDTA) for 5 min at 37°C followed by centrifugation and resuspension of cells, and plating. In all preparations used viability was >90% by Trypan Blue exclusion and >80% of the objects were single cells (as distinct from doublets, triplets or larger fragments). Cells were then seeded at 200 or 500 cells/dish into either 5 cm polystyrene dishes or, for fluorescence staining, into 5 cm collagen-coated (Whitescarver, 1974; Edwards et al. 1984) Petriperm dishes (Heraeus, Stockport, U.K.), both containing a feeder layer of 3T6 mouse fibroblast cells (Todaro & Green, 1963). Feeder layers were prepared by seeding dishes with 2× 105 cells each, from washed, trypsinized cultures of logphase 3T6 cells that had been treated for 18-24h with 3 μ;gml−1 ′ mitomycin C (Sigma) in serum-free DMEM (Taylor-Papadimitriou et al. 1977a). Cultures were usually examined after approximately 10 days growth, when breast clones were 0·1-2 mm in diameter, and the majority of the 3T6 feeder cells had detached. ‘Break-through’ of the mitomycin-C-treated 3T6 cells was encountered only rarely and resulting colonies were readily distinguishable from the breast-derived epithelial clones. Clonal morphology was established by phase-contrast observation of living cultures; cloning efficiency was determined at 7—10 days after Giemsa staining of formalin-fixed cultures.

Cloning of single cells by micropipette

In some experiments individual cells were aspirated into a micropipette under phase-contrast observation as described by Zagury, Morgan & Fouchard (1981) and plated individually in the wells of an 8-well no. 4808 ‘Lab-Tek’ plate (Miles Labs, Naperville, Illinois, U.S.A.) under conditions described above. Micropipetteβ were made by drawing out commercial Pasteur pipettes to a tip size of the order of 50-100 μ;m diameter.

Keratin staining

Cultures fixed in cold (4°C) methanol were stained immunocytochemically with a polyclonal rabbit antiserum to human callus keratins, the properties of which have been described elsewhere (Gustersonet al. 1982). It was kindly provided by Dr M. J. Warburton of this Institute. Antibody binding at a 1:100 dilution of the antiserum was visualized by an indirect procedure using alkaline phosphatase-conjugated anti-rabbit second antibody, prepared as described by Avrameas (1969), with naphthol AS:Bl phsophoric acid (sodium salt) as the chromogenic substrate.

Monoclonal antibodies

The production of monoclonal antibodies to human breast epithelial cells, from mice immunized with human milk fat globule membrane, has been described (Edwards, 1981; Foster et al. 1982). Their properties are summarized in Table 1. Unless otherwise stated, ascites fluid was the source of antibody, and it was used at a dilution of 1:100.

Immunofluorescence

Cultures were stained, unfixed, by indirect immunofluorescence as described previously (Edwards et al. 1984) except that the cells were photographed without mounting in PVA-glycerol. IgG1 and IgM-specific fluorochrome-conjugated second antibodies were used for double immunofluorescence where appropriate (Table 1). Two-colour staining with fluorescein fluorescence for antibodies M8 and M24 together, and rhodamine fluorescence for antibody M18 was achieved by a four-step procedure: incubating first with a mixture of antibodies M18 and M8, then with a mixture of an IgG1-specific fluorescein conjugate and an IgM-βpecific rhodamine conjugate, then with antibody M18 again, and finally with directly fluorescein-conjugated antibody M24 (Edwards & Brooks, 1984).

When short-term primary cultures of human breast were trypsinized cloning efficiency was between 20 and 35%, as obtained by others (Smith et al. 1981; Stoker et al. 1982). Direct cloning from freshly isolated organoids gave more variable results, with efficiencies between 5 and 25% for 12 breast samples.

Morphology of clones

Two morphological types of clone were consistently seen by phase-contrast microscopy of living cultures. Representative examples are shown in Figs 1 and 2. In the first type most of the cells are spread out in an attenuated fashion with the individual cell boundaries difficult to discern, and when visible they are seen as dark lines (see especially Figs 1A, 2c). After fixation and Giemsa staining the cells sometimes seem to be ‘elongated’ but in living cultures most of the cells are more or less isometric (see Figs 1A, 2C). The clones have a continuous and sharply demarcated boundary in which individual cells are again difficult to discern. Some cells at the periphery are elongated and circumferentially orientated with the impression of being under tension, and in older clones (10 days +) ‘cable-like’ forms can occasionally be seen.

Fig. 1.

Immunofluorescence staining of two clones, one stained by antibodies M8 and M24, the other negative. The stained clone is a typical spread clone, the other a typical compact clone. A. Phase contrast; note that in the left-hand, positive clone many of the cell boundaries appear as black lines. B. Same field, immunofluorescence staining by antibodies M8 and M24 together, both fluorescein fluorescence, c. Immunofluorescence staining by antibody M18 alone in rhodamine fluorescence; note the slight but characteristic difference in the texture of the staining between the clones: in the left-hand clone fluorescence is even and finely textured, with cell boundaries often being negative and never prominent; in the right-hand clone fluorescence is more spotty and cell boundaries are emphasized. ×300; bar, 100μ;m.

Fig. 1.

Immunofluorescence staining of two clones, one stained by antibodies M8 and M24, the other negative. The stained clone is a typical spread clone, the other a typical compact clone. A. Phase contrast; note that in the left-hand, positive clone many of the cell boundaries appear as black lines. B. Same field, immunofluorescence staining by antibodies M8 and M24 together, both fluorescein fluorescence, c. Immunofluorescence staining by antibody M18 alone in rhodamine fluorescence; note the slight but characteristic difference in the texture of the staining between the clones: in the left-hand clone fluorescence is even and finely textured, with cell boundaries often being negative and never prominent; in the right-hand clone fluorescence is more spotty and cell boundaries are emphasized. ×300; bar, 100μ;m.

Fig. 2.

Contrast in morphology between the clones stained by antibodies M8 and M24, and negative clones. The difference coincides with our morphological classification as spread or compact clones. A-C. Positive clones (spread morphology); D—F, negative clones (compact morphology); c is at higher magnification (× 281) than the others (×110) to show the characteristic appearance of the cell boundaries (arrowed). Bars, 100μ;m.

Fig. 2.

Contrast in morphology between the clones stained by antibodies M8 and M24, and negative clones. The difference coincides with our morphological classification as spread or compact clones. A-C. Positive clones (spread morphology); D—F, negative clones (compact morphology); c is at higher magnification (× 281) than the others (×110) to show the characteristic appearance of the cell boundaries (arrowed). Bars, 100μ;m.

The second type of clone is composed of more tightly packed cells distinguished by their brightly refractile boundaries, which makes the individual cells more distinct than in spread clones. Within these ‘compact’ clones there are considerable individual variations in the appearance of the epithelial cells, particularly at early stages of their growth. Some form regular pavement-like arrays, others contain vacuoles, and some are more fan-shaped or triangular. The latter also have a tendency to move away from the edges of the colonies, sometimes becoming completely detached, and imparting a more ragged appearance to the periphery, in contrast to the spread clones. Some of these latter cells could be the ‘open’ phenotype that has been described in human milk cultures (Stoker et al. 1982; Stoker & Perryman, 1984), although under phasecontrast it is clear that most retain some contacts with neighbouring cells. Despite extensive searching we have not seen any individual colonies of > 5 cells composed exclusively of these open cells.

The compact clones always predominated, ranging from 70% to 90% in the 12 breast samples, the remainder being spread. Over 2000 clones were counted, so the frequency of open clones in our cultures must be <0·1%. By observing representative clones from day 3 to day 10 we confirmed that both spread and compact colonies can arise by clonal growth. Both show mitoses and neither arises exclusively by the attachment and spreading, without division, of the few clumps in the cell suspension. At least the majority of the spread clones do not arise by conversion of compact clones, and vice versa. No significant differences have been noted between clones obtained directly from the organoids and the clones obtained by trypsinization of short-term cultures of intact organoids.

Staining of clones with antiserum to cytokeratins

To provide evidence that essentially all the clones were from the breast epithelium rather than from vessels or stroma, the clones were stained immunocytochemically with a polyclonal antiserum to cytokeratins described by Gusterson et al. (1982). On sections of breast this stains the cytoplasm of only luminal epithelial and myoepithelial cells. All the clones, of both morphological types, stained quite clearly, although to varying degrees, while residual 3T6 feeder cells were completely negative.

Staining of clones with monoclonal antibodies

A total of 10-30% of the clones stained with the monoclonal antibodies M8 and M24. Most of these clones were stained by both antibodies. In the clones that were stained, between about 25% and 100% of the cells were stained by one or other or both of the antibodies. Fig. 1 shows two clones that touch, one positive to antibodies M8 and M24, the other completely negative. The positive clones corresponded to the ‘spread’ morphological type, the negative clones were the compact type (Figs 1,2). (Our morphological classification was developed before, and so independently of, our staining with antibodies.) The staining resembled the staining of the sheets of cells in non-clonal cultures plated from intact organoids: even, finely textured fluorescence with cell boundaries usually negative (Fig. 1B). Many of the positive cells are directly attached to the plastic surface, i.e. they are not a surface layer of multilayered culture as they usually are in a bulk culture. Fig. 2 shows several more examples of the phase-contrast appearance of positive (on the left) and negative (on the right) clones (fluorescence not shown). The positive clones seem always to be of spread morphology, the negative clones of compact morphology.

Monoclonal antibody LICR-LON-M18 stains the clones that are stained by antibodies M8 and M24 in a very similar way. It also stains the clones that are negative to antibodies M8 and M24, but the appearance of the staining is distinctly different (Fig. 1c). The staining is generally weaker (Fig. 1c shows a relatively strong example), has a stippled texture, and instead of the cell boundaries being usually negative, they are often more strongly stained than the rest of the cell surface.

Heterogeneity of antigen expression within the clones

In the clones that stained with antibodies M8, M18 and M24 some cells were positive and some were negative to each antibody (Figs 1, 3); in fact the cells of a clone were as heterogeneous in their antigenic phenotypes as cells of the intact tissue (Edwards & Brooks, 1984; Edwards, 1985). To confirm that the clones showing this heterogeneity of antigen expression came from single cells, single cells were isolated with a micropipette. Fig. 3 shows two-colour immunofluorescence staining of a clone derived from such a cell: it contains cells of all antigenic phenotypes. This is in agreement with previous studies of expression of similar antigens in clones of mammary cells (Stoker et al. 1982; Peterson, Ceriani, Blank & Osvaldo, 1983).

Fig. 3.

Two-colour immunofluorescence of a typical positive clone of proven single-cell origin, showing the heterogeneity of antigen expression in a clone. The clone was grown from a cell isolated with a micropipette. A. Immunofluorescence staining by monoclonal antibody LICR-LON-M8, in fluorescein fluorescence; B, immunofluorescence staining of same field by monoclonal antibody LICR-LON-M24, in rhodamine fluorescence. Arrow in A marks an example of a cell that is negative in B, and vice versa. Asterisk marks a cell negative with both antibodies. ×270; bar, 100μ;m.

Fig. 3.

Two-colour immunofluorescence of a typical positive clone of proven single-cell origin, showing the heterogeneity of antigen expression in a clone. The clone was grown from a cell isolated with a micropipette. A. Immunofluorescence staining by monoclonal antibody LICR-LON-M8, in fluorescein fluorescence; B, immunofluorescence staining of same field by monoclonal antibody LICR-LON-M24, in rhodamine fluorescence. Arrow in A marks an example of a cell that is negative in B, and vice versa. Asterisk marks a cell negative with both antibodies. ×270; bar, 100μ;m.

Our main finding was that two types of clone could be distinguished in platings of cells from human mammary epithelium. One type, around 10—30% of the clones, had what we term spread morphology and stained with the monoclonal antibodies LICR-LON-M8 and LICR-LON-M24; the other type had compact morphology and was never stained by these two antibodies. The clones were all stained by a polyclonal antiserum to cytokeratins, confirming that they came from the epithelium rather than the stroma or vessels.

The properties of the clones suggest a simple interpretation of the non-clonal cultures obtained by plating intact organoids (Edwards et al. 1984). The bulk of the cells in the organoid cultures correspond to the compact clones: they are negative to antibodies M8 and M24, and are rather similar in phase-contrast appearance to the primary′ cultures. The superficial sheets of cells stained by antibodies M8 and M24 in the organoid cultures correspond to the positive clones: in both cases the cells are relatively flattened, the texture of the fluorescence staining is similar and the cell boundaries appear by phase contrast as black lines (Fig. 1A). Thus it appears that we have cloned out the two major types of cell that we distinguished in the primary organoid cultures.

It is difficult to relate our clone types to previous classifications of cultured mammary cells, as descriptions are somewhat subjective, the types of cells present in other cultures are different, and the types of colony obtained depend on the medium and feeders used (e.g. see Stoker, 1984). However, two types of colony, E and E1, were identified by Kirkland, Yang, Jorgensen, Longley & Furmanski (1979) and they look rather like our spread and compact types of clone, respectively. Clones of cells from milk have been classified as open and closed, open clones having few contacts between the cells (Stoker et al. 1982; Taylor-Papadimitriou, Shearer & Tilly, 1977b). Closed clones were further classified as cuboidal or elongated (types S and D, respectively, described by Taylor-Papadimitriou et al. 1977b). We did not find open clones. The elongated type seem to be spread in our classification but whether the cuboidal type corresponds to our spread type or is compact is not clear.

Both closed types of clone stain (Stoker et al. 1982; Chang, Keen, Lane & Taylor-Papadimitriou, 1982) with monoclonal antibody HMFG1 (Arklie et al 1981), which is similar (Ormerod, Steele, Edwards & Taylor-Papadimitriou, 1984) to our antibody LICR-LON-M8, suggesting that they may both correspond to our spread type.

Which cells in the intact tissue give rise to these clones? The spread clones clearly come from luminal epithelial cells. What is difficult to decide is where compact clones come from. If they are from luminal epithelial cells, where do the myoepithelial cells go? If they are from myoepithelial cells, why are they stained by the antibody M18? M18 recognizes a simple carbohydrate backbone structure present on most human cells but usually masked by terminal residues (Table 1; Gooi et al. 1983), so it might arise on myoepithelial cell clones as an artefact of culture or of the enzyme treatment of the tissue. Our tentative conclusion is that our spread clones, the E clones of Hallowes et al. (1977), and probably the closed colonies from milk (Taylor-Papadimitriou et al. 1977b; Stoker et al. 1982) arise from luminal epithelial cells. Our compact colonies and perhaps the E1 colonies of (Kirkland et al. 1979), being abundant in mammoplasty samples and post-weaning milks, may well arise from myoepithelial cells, but they could be epithelial cells that have lost expression of their normal range of antigens.

We thank Professor A. M. Neville for advice and support and the surgeons and their staff who kindly provided normal breast tissue.

Arklie
,
J.
,
Taylor-Papadimitriou
,
J.
,
Bodmer
,
W.
,
Egan
,
M.
&
Millis
,
R.
(
1981
).
Differentiation antigens expressed by epithelial cells in the lactating breast are also detectable in breast cancers
.
Int.J. Cancer
23
,
23
29
.
Avrameas
,
S.
(
1969
).
Coupling of enzymes to proteins with glutaraldehyde. Use of the conjugates for the detection of antigens and antibodies
.
Immunochemistry
6
,
43
52
.
Chang
,
S. E.
,
Keen
,
J.
,
Lane
,
E. B.
&
Taylor-Papadimitriou
,
J.
(
1982
).
Establishment and characterization of SV4O-transformed human breast epithelial cell lines
.
Cancer Res
.
42
,
2040
2053
.
Easty
,
G. C.
,
Easty
,
D. M.
,
Monaghan
,
P.
,
Ormerod
,
M. G.
&
Neville
,
A. M.
(
1980
).
Preparation and identification of human breast epithelial cells in culture
.
Int. J. Cancer
26
,
577
584
.
Edwards
,
P. A. W.
(
1981
).
Some properties and applications of monoclonal antibodies
.
Biochem. J
.
200
,
1
10
.
Edwards
,
P. A. W.
(
1985
).
Heterogeneous expression of cell-surface antigens in normal epithelia and their tumours, revealed by monoclonal antibodies
.
Br.J. Cancer
51
,
149
160
.
Edwards
,
P. A. W.
&
Brooks
,
I. M.
(
1984
).
Antigenic subsets of human breast epithelial cells distinguished by monoclonal antibodies. J
.
Histochem. Cytochem
.
32
,
531
537
.
Edwards
,
P. A. W.
,
Brooks
,
I. M.
&
Monaghan
,
P.
(
1984
).
Expression of epithelial antigens in primary cultures of normal human breast analysed with monoclonal antibodies
.
Differentiation
25
,
247
258
.
Foster
,
C. S.
,
Edwards
,
P. A. W.
,
Dindsale
,
E. A.
&
Neville
,
A. M.
(
1982
).
Monoclonal antibodies to the human mammary gland: I. Distribution of determinants in non-neoplastic mammary and extra mammary tissues
.
Virchows Arch. path. Anat
.
394
,
279
293
.
Gooi
,
H. C.
,
Uemura
,
K.-L
,
Edwards
,
P. A. W.
,
Foster
,
C. S.
,
Pickering
,
N.
&
Feizi
,
T.
(
1983
).
Two mouse hybridoma antibodies against human milk-fat globules recognise the I(MA) antigenic determinant β-D-Galp-(l → 4)-β-D-GlcNAc-(l → 6)
.
Carbohyd. Res
.
120
,
293
302
.
Gusterson
,
B. A.
,
Warburton
,
M. J.
,
Mitchell
,
D.
,
Ellison
,
M.
,
Neville
,
A. M.
&
Rudland
,
P. S.
(
1982
).
Distribution of myoepithelial cells and basement membrane proteins in the normal breast and in benign and malignant breast diseases
.
Cancer Res
.
42
,
4763
4770
.
Hallowes, R. C,
Millis
,
R.
,
Pigott
,
D.
,
Shearer
,
M.
,
Stoker
,
M. G. P.
&
Taylor-Papadimitriou
,
J.
(
1977
).
Results on a pilot study of cultures of human lacteal secretions and benign and malignant breast tumours
.
Clin. Oncol
.
3
,
81
90
.
Janss
,
D. H.
,
Hillman
,
E. A.
, Malan-S2
bley
,
L. B.
&
Ben
,
T. L.
(
1980
).
Methods for the isolation and culture of normal human breast epithelial cells
.
In Methods in Cell Biology
, vol.
21
(ed.
C. C.
Harris
,
B. F.
Trump
,
G. D.
Stoner
), pp.
107
134
.
New York, London
:
Academic Press
.
Kirkland
,
W. L.
,
Yang
,
N.-S.
,
Jorgensen
,
T.
,
Longley
,
C.
&
Furmanski
,
P.
(
1979
).
Growth of normal and malignant human mammary epithelial cells in culture. J
.
natn. Cancer Inst
.
63
,
29
41
.
McIlhinney
,
R. A. J.
,
Patel
,
S.
&
Gore
,
M. E.
(
1985
).
Monoclonal antibodies recognising epitopes carried on both glycolipids and glycoproteins of the human milk fat globule membrane
.
Biochem. J
.
227
,
155
162
.
Ormerod
,
M. G.
,
Steele
,
K.
,
Edwards
,
P. A. W.
&
Taylor-Papadimitriou
,
J.
(
1984
).
Monoclonal antibodies which react with epithelial membrane antigens. J
.
exp. Path
.
1
,
263
271
.
Peterson
,
J. A.
,
Ceriani
,
R. L.
,
Blank
,
E. W.
&
Osvaldo
,
L.
(
1983
).
Comparison of rates of phenotypic variability in surface antigen expression in normal and cancerous human breast epithelial cells
.
Cancer Res
.
43
,
4291
4296
.
Smith
,
H. S.
,
Lan
,
S.
,
Ceriani
,
R.
,
Hackett
,
A. J.
&
Stampfer
,
M. R.
(
1981
).
Clonal proliferation of cultured nonmalignant and malignant human breast epithelia
.
Cancer Res
.
41
,
4637
4643
.
Stampfer
,
M.
,
Hallowes
,
R. C.
&
Hackett
,
A. J.
(
1980
).
Growth of normal human mammary cells in culture
.
In Vitro
16
,
415
425
.
Stoker
,
M.
(
1984
).
Junctional competence in clones of mammary epithelial cells, and modulation by conditioned medium
.
J. cell. Physiol
.
121
,
174
183
.
Stoker
,
M.
&
Perryman
,
M.
(
1984
).
Cultures of exfoliated mammary epithelial cells: variation between donors
.
Breast Cancer Res. Treatment
4
,
11
18
.
Stoker
,
M.
,
Perryman
,
M.
&
Eeles
,
R.
(
1982
).
Clonal analysis of morphological phenotype in cultured mammary epithelial cells from human milk
.
Proc. R. Soc. Lond. B
,
215
,
231
240
.
Taylor-Papadimitriou
,
J.
,
Purkis
,
P.
&
Fentiman
,
I. S.
(
1980
).
Cholera toxin and analogues of cyclic AMP stimulate the growth of cultured human mammary epithelial cells. J
.
cell. Physiol
.
102
,
317
321
.
Taylor-Papadimitriou
,
J.
,
Shearer
,
M.
&
Stoker
,
M. G. P.
(
1977a
).
Growth requirements of human mammary epithelial cells in culture
.
Int. J. Cancer
20
,
903
908
.
Taylor-Papadimitriou
,
J.
,
Shearer
,
M.
&
Tilly
,
R.
(
1977b
).
Some properties of cells cultured from early-lactation human milk
.
J. natn. Cancer Inst
.
58
,
1563
1571
.
Todaro
,
G. J.
&
Green
,
H.
(
1963
).
Quantitative studies of the growth of mouse embryo cells in culture and their development into established lines
.
J. Cell Biol
.
17
,
299
313
.
Whtγescarver
,
J.
(
1974
).
Problems of in vitro culture of human mammary tumor cells
.
J. invest. Derm
.
63
,
58
64
.
Yang
,
J.
,
Richards
,
J.
,
Guzman
,
R.
,
Imagawa
,
W.
&
Nandi
,
S.
(
1980
).
Sustained growth in primary culture of normal mammary epithelial cells embedded in collagen gels
.
Proc. natn. Acad. Sci. U.SA
.
77
,
2088
2092
.
Yang
,
N. S.
,
Kube
,
D.
,
Park
,
C.
&
Furmanski
,
P.
(
1981
).
Growth of human mammary epithelial cells on collagen gel surfaces
.
Cancer Res
.
41
,
4093
4100
.
Zagury
,
D.
,
Morgan
,
D. A.
&
Fouchard
,
M.
(
1981
).
Production of human T-lymphocyte clones. I. Monoclonal cultures and functional cytotoxic maturation
.
J. imtnun. Meth
.
43
,
67
70
.