[CANCER RESEARCH 39, 332-344, February 1979
0008-5472/79/0039-0000$02.00
NuclearUltrastructureof EpithelialCell LinesDerivedfromHuman
Carcinomasand NonmalignantTissues1
Helene S. Smith,2 E. Louise Springer, and Adeline J. Hackett
Peralta Cancer Research Institute, Oakland, California 94609, and Donner Laboratory, Lawrence Berkeley Laboratory, University of California,
Berkeley,California94720
ABSTRACT
were not observed in any nonmalignant lines. Nuclear
consistent marker of epithelial cell transformation seemed
to be growth in agar suspension (1, 18); however, Katsuta
and Takaoka (14) noted that there was not a complete
correlation between tumorigenicity and growth in agar. All
of the reports agreed that there were subtle differences in
morphology after carcinogen treatment but often found it
difficult to quantify the changes. Some of the morphologi
cal changes observed after neoplastic transformation in
cluded pleomorphism in cell size and shape (14, 18, 19, 20,
30, 32) and nuclei (18, 20, 22), increased nuclear:cytoplas
mic ratio (18, 19, 32), prominent, enlarged, or numerous
nucleoli (18, 20, 22, 32), elongation of cells (32), basophilic
cytoplasm (18, 22), abnormal mitoses (14, 18, 22), reduced
intercellular cohesiveness (14, 32), and altered colony mor
phology(14,19,28,32).
envelope dilation was seen in all lines derived from cancer
ous organs as well as from malignant tissues but not in any
in vitro parameters
The nuclear ultrastructure of sixteen human epithelial
cell lines has been characterized in detail by transmission
electron microscopy. The cell lines were derived from
normal tissues, nonmalignant tissues of cancerous organs,
primary carcinomas, and metastatic carcinomas. Every cell
section on a grid containing a clearly defined nucleus and
nucleolus was scored blindly utilizing a checklist of
markers. The goal of these studies was to determine
whether any ultrastructural markers consistently distin
guished the different stages of malignant progression rep
resented among the lines. Nuclear bodies and perichroma
tin granules were found in all lines derived from cancer and
lines derived from normal tissue. Margination of chromatin,
irregularity of nuclear outline, redistribution
of nucleolar
components, and margination were expressed slightly by
the normal lines, to variable degrees by the lines derived
from cancerous organs, and to a much greater extent by all
lines derived from malignant tissues. No differences were
found between lines derived from primary carcinomas and
those derived from metastatic specimens. There were no
ultrastructural differences comparing subconfluent and
confluent cells or cells at different passage levels. In addi
tion, the nuclear ultrastructure of a malignant line in culture
was similar to that of a tumor induced by those cells in an
immunosuppressedmouse.
INTRODUCTION
Since most human cancers are carcinomas, that is, tu
mors of epithelial origin, it is important to define transfor
mation criteria using epithelial cells. There are a number of
reports characterizing the in vitro properties of normal and
carcinogen-treated rodent epithelial cells (11, 29, 30). In
many cases, treatment with the chemical gave rise to
cultures which produced carcinomas on inoculation into
animals but showed no distinctive features in vitro. Many of
the properties found to correlate with fibroblastic transfor
We have approached the problem of determining which
correlate
with human cancer by devel
oping cell lines3 of various human carcinomas and nonma
lignant epithelial tissues and then comparing the properties
of the nonmalignant cell lines with those derived from
carcinomas (10, 21 , 24, 25). The cell lines were derived
from tissues representing some of the stages of malignant
progression including normal tissue, nonmalignant tissue
of cancerous organs, primary carcinomas, and metastatic
carcinomas. There were many similarities between the
human carcinoma-derived lines and transformed rodent
epithelial cells. For the human cells, there was no correla
tion between cancer and growth to high-saturation density
(21) or increasedactivation of plasminogen.4The nonmalig
nant lines were consistently negative for growth in metho
cel, on density-inhibited monolayers, and in immunosup
pressed mice while the carcinoma-derived lines each ex
pressed a combination of these abnormalities with some
lines being much more abnormal than others.
Like the rodent epithelial lines, the most consistent differ
ences between tumor and nonmalignant cells were mor
phological (10, 21 , 24, 25). In particular, alterations of
nuclear and mitochondrial ultrastructure were noted in all
of the tumor-derived lines by preliminary studies using the
transmission electron microscope (25). Therefore, we initi
ated a detailed evaluation of the cell lines for a variety of
ultrastructural markers possibly associated with cancer (9,
mation such as refractivity, piling up of the cells, decreased 23) as well as markers associated with metabolic activity.
serum requirement, and increased plasminogen activator This paper details our observations on nuclear ultrastruc
(18) did not apply to epithelial cell systems. The most ture of the human epithelial lines.
I
This
work
was
supported
by
Contract
CP-70510
from
the
National
Cancer Institute and by Contract W7405-ENC48 from the Department of
Energy.
2 To
whom
requests
for
reprints
should
be
addressed.
Received June 5, 1978; accepted October 24, 1978.
332
3 Nomenclature
used
conforms
with
work
by
S.
Fedoroff
(7)
which
states
that a “cell
line―
arises from a primary culture at the first subculture while an
“established
cell line―is one which has demonstrated the potential to be
subcultured indefinitely.
4 H.
S.
S.,
unpublished
observations.
CANCERRESEARCHVOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
Ultrastructure of Human Epithelia! Cells
MATERIALS AND METHODS
Cell Culture. The growth medium used was Dulbecco's
modification of Eagle's medium (Grand Island Biological
Co., Grand Island, N. V. ; No. 196G) containing glucose (4.5
g/Iiter) supplemented with 10% fetal calf serum and insulin
(10 @tg/ml)(Calbiochem, San Diego, Calif.). Unless other
wise indicated, cells were harvested at confluence. All of
the lines, both tumor and those derived from nonmalignant
tissue, grew slowly in culture (21). There was no difference
in growth rate as a function of nuclear ultrastructure.
Electron Microscopy. The cell lines cultured in Falcon
flasks were fixed in situ with 2.5% glutaraldehyde in 0.1 M
sodium cacodylate buffer (pH 7.3) at room temperature for
3 hr, rinsed with buffer, and fixed with 2% buffered osmium
tetroxide for 2 hr at room temperature. En bloc staining
with ethanolic uranyl acetatefollowed stepwisedehydration
ence or absenceor degree of expression of various nuclear
ultrastructural characteristics for everycell section showing
a nucleus with a nucleolus. Although most of the scoring
was done directly at the transmission electron microscope
console, some photographs were taken of each specimen.
Similar results were obtained when the photomicrographs
were scored blindly. To be certain that a cell was scored
only once, we surveyed the serial sections to identify each
cell profile present. Then we chose different cell profiles for
evaluation.
RESULTS
The sources of all the cell lines in this study are described
in Table 1. The normal cell lines were derived from fetal
intestines and an adult bladder mucosa from a patient with
prostate cancer. Three lines were derived from noninvolved
embedment cut perpendicular to the plane of growth were areas of kidneys removed for carcinoma of that organ.
with ethanol. Ultrathin sections of the Spurr (26) or Epon
poststained with aqueous uranyl acetate and lead citrate.
The sections were collected on copper mesh grids so that
the sections were oriented parallel to each other. This
orientation permitted identification of the cell profiles,
thereby preventing duplicate scoring. The coded specimens
were received by the operator at the transmission electron
microscope console without knowledge of their tissue of
origin. Scoring was accomplished by identifying the pres
Since there is evidence suggesting that atypical hyperpla
sias are often found in noninvolved areas peripheral to
carcinomas (6, 31), these lines were categorized separately.
The malignant cell lines were derived from primary carci
nomas of the rectum, colon, breast, and transitional cells
of the urethra and kidney, as well as metastatic lesions of
pancreas, stomach, and 2 lines derived from metastatic
tissue with primary lesions of uncertain origin.
Table 1
specimensDesignationHistory of patients and biopsy
(carcinomas)Carcinomas785T
Sex
Age (yr)RaceDiagnosis
58Colon
carcinoma675TM
M
InformationColon
carcinomaunknown761T
M
carcinoma-kidney769T
carcinoma-urethra578T
F
breast766T F
M
lymphnode746T
M
cle700T
M
65C°Transitional
59NTransitional
cell
cell
64CPancreas
of
carcinoma metastaticto
74CStomach
carcinomametastaticto mus
74CCarcinosarcoma
(probablycolon, intestine,
or pancreas)metastaticto
430Adenocarcinoma
(probably colon, in
testine, or pancreas)metastaticto
61CCarcinoma
hip696T
M
sacrumPeripheral
tissue of carcinomatous
or
gans699K
M
60CNormal
kidney tissue from patient with
renal cell
M
65CNormal
kidney tissue from Patient with
transitional cell carcinoma kidney
carcinoma761
K
761T)71
(samepatient as
5K
M
carcinomaNonmalignant
tissues6771nt
disorder6801nt
76Normal
kidney tissue from patient with
renal
3- to 4-mo. fetusFamilial
3- to 4-mo. fetusFamilial
drome741
Int
3- to 4-mo. fetusTherapeutic
formalities767BL
M
45CNormal
history-immune
history of Wiskott-Aldrich syn
abortion, no known ab
bladder mucosa from patient
with prostate
a C, Caucasian;
N, Negro;
0,
carcinoma
Oriental.
FEBRUARY1979
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
333
H. S. Smith et a!.
To perform the ultrastructural studies, coded cell speci
mens were evaluated by E. Louise Springer who was una
ware of specimen origin. Every cell section on a grid
containing a clearly defined nucleus and nucleolus was
scored for presence or absence of each marker, and where
lines derived from peripheral tissue of cancerous organs
also had some sections showing dilated nuclear envelopes.
Chart 2 summarizes the data for 3 other nuclear markers,
cytoplasmic invaginations, chromatin margination, and nu
clear outline. For these 3 markers, all of the nonmalignant
appropriate, it was scored for degree of expression of that lines showed someexpression; however,the tumor-derived
marker.
lines exhibited these properties to a more marked degree.
Fig. 1 illustrates2 nuclearmarkersscored, nuclear bodies The same 3 cell lines derived from peripheral tissue of
and perichromatin granules. The types of nuclear bodies
were graded according to Bouteille's classification (3). The
cancerous organs which displayed nuclear envelope dila
tion (Chart 1) also exhibited more chromatin margination
simplest type of nuclear body (corresponding to Bouteille's and more irregular nuclear outlines than did the lines
derived from normal tissue.
Fig. 3 illustrates varying degrees of fibrillar center devel
opment which is thought to result from the redistribution of
ribonucleoprotein structures (nucleolonemas) (4). The nor
ing to Bouteille's type 3), had a distinct electron-dense
fibrillar capsule enclosing a granular center (Fig. lb). The mal nucleolus (Fig. 3a) is made up of evenly distributed
most complex classification of nuclear body (correspond nucleolar material. Fig. 3, b to d, shows increasing degrees
ing to Bouteille's type 4) had a well-formed fibrillar capsule of fibrillar center formation, which appear as light areas
enclosing a distinctly beaded matrix (Fig. ic). Fig. lb (inset) surrounded by more electron-dense fibrillar components.
illustrates perichromatin granules, electron-dense solitary The fibrillar centers were classified according to degree of
spherical granules (300 to 350 A in diameter) separated contrast observed between the electron-lucent centers and
from the adjacent chromatin by an electron-lucent halo. the electron-dense surrounding fibrillar material. Within a
This property was scored according to an estimation of single section, varying degrees of fibrillar center formation
quantity of granules present (approximately 5 to 10 per were noted. Classification was determined by the most
field, approximately 11 to 20 per field, or more than 20 per extreme form observed in the section. The nucleoli illus
field.)
trated in Fig. 3 are unattached to the nuclear membrane.
Three other nuclear properties which were scored ac Fig. 4 illustrates increasing association of the nucleolus
cording to degree of expression (nuclear outline, nuclear with the nuclear perimeter. This margination can be absent
envelope invaginations, and chromatin margination) are (Fig. 4a), slight (Fig. 4b), moderate (Fig. ‘Ic),
or extensive
(Fig. 4d).
illustrated in Fig. 2. Fig. 2a illustrates a smooth-margined
nucleus which lacked invaginations and had finely dis
Chart 3 summarizes the data on ultrastructure of the
persedeuchromatin without margination. Fig. 2d illustrates nucleolus. Slight fibrillary center development and margin
a nucleus characterized by a contorted, lobulated outline ation were seen in most of the nonmalignant cell lines,
with numerous fingerlike invaginations of cytoplasm into while extensive expression of these markers was typical of
the nucleoplasm. The chromatin tended to accumulate at the tumor-derived lines. Only 2 of the 3 cell lines derived
the nuclear margin in heavyirregular deposits. Fig. 2, b and from peripheral areas of cancerous organs which showed
C , illustrates
intermediate
degrees
of these
characteristics.
increased nuclear aberrancies (761K and 715K) also had the
In Fig. 2, the 3 properties are depicted at similar stages of more extensive fibrillar center formations and margination
abnormality; however, they were observed to vary inde of the nucleolus.
A normal nucleolus generally occupied less than 20% of
pendently (i.e. , extensive invagination might be seen in a
the nuclear volume. Nucleoli were scored as ‘
‘giant―
if they
cell section showing only moderate chromatin margina
tion). One additional property, morphology of the nuclear occupied approximately 50% or more of the nuclear vol
envelope, is illustrated in Fig. 2, e and f. A normal nuclear ume. This nucleolar hypertrophy was found in at least some
envelope with closely apposed bilayers is illustrated in Fig. cell sections of most of the tumor-derived lines; one periph
eral tissue cell line, 715K, also had enlarged nucleoli.
2e while an irregular dilated envelope is shown in Fig. 2/.
Charts 1 and 2 summarize the data accumulated on the Multiple nucleoli (usually 2 or 3 per section) which were
nuclear ultrastructure. A dot indicates the evaluation of a less frequently observed in the tumor-derived lines, were
single cell section for a given property. Each line was tested not observed in any of the nonmalignant lines (Chart 3).
To determine whether any of the ultrastructural differ
at least once between passages 4 and 20. A few lines were
tested at more than one passage, and in every case, the ences between tumor and nonmalignant cells varied as a
data at different passages were similar, including one cell function of the growth state, one normal line, 6801nt, and
line, 700T, which retained its ultrastructural features for as one tumor line, 766T, were compared under confluent and
subconfluent culture conditions. Subconfluent and con
long as 69 passages. Chart 1 shows that 2 nuclear proper
fluent cells proved to be similar (Charts 4 and 5) suggesting
ties, nuclear bodies and perichromatin granules, were ob
served in all of the tumor-derived lines and in none of the that the differences between tumor and normal cells were
nonmalignant lines. The perichromatin granules were not related to growth states after confluence. In addition, a
found in most of the tumor cell sections observed while tumor arising in an immunosuppressed mouse after inoc
nuclear bodies were found at least once in each tumor cell ulation of 766T cells was also studied. The tumor cells
which were analyzed directly were similar to those grown in
line but usually only in a minority of the sections observed.
In contrast, dilation of the nuclear envelope was observed culture, suggesting that the ultrastructural features de
in almost every tumor section examined. In addition, 3 cell scribed were not artifacts of in vitro culture conditions.
types 1 and 2) was composed of fibrils approximately 50 to
70 A in diameter which are arranged in concentric forma
tions (Fig. la). More complex nuclear bodies (correspond
334
CANCERRESEARCHVOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
@
Z@
Ultrastructure of Human Epithelial Cells
NUCLEAR
NUCLEAR
lz
LU
p-@
Source of Cells
(Designation) -- Passage
I
LU
@__J
;
a.
I-
a.
DILATiON
GRANULESENVELOPE
BODIESPERI-CHROMATIN
CM
@
ULTRASTRUCTURE
0!
ZU)
Z
I
10
0
l@
U
I
z
I
@o
I
I
I
u
‘U
I -@II
:;I
I
I
o
5 I
:@
4
a.
Derived From Carcinoma
Colon
(785T)
-9
Colon
(675T)
-4
—11
-13
C..
55,5
•.
S..
-9
-13
.•..
..
Trans. Cell
@
(761T)
Trans. Cell
(769T)
-12
Breast
(578T)
-9
Metastatic Pancreas (766T)
Metastatic Stomach (746T)
Metastatic *
(700T)
.
::::
••••
-9
—11
.
Derived
.•
-14
...
(us's
t ::@
...•
taii
5,::.
..
‘s's's
::ui
:un.
.....
•..
tsas.
“U:.
••..@ ..
.•
.
-33
•.....
:::r
•.
..
ssttU
From Peripheral
Tissue of Carcinomolous
Organ::@:
Kidney
(699K)
-5
Kidney
(761K)
-7
....Adult
Adult
-13Derived
Kidney
(715K)
Adult
•....,
fl's'
U?
.
u::
-6
(696T)
usgs,
•
C.
C...
-69
Metastatic ‘
....
...
:::::
505$
••••. •..•.
::::t:::::
::::t::s.
.•••.•
From Nonmalignant
Tissue
Fetal Intestine
(741nt)
-10
Fetal Intestine
(6771nt)
-12
Fetal Intestine
(6801nt)
-15
-26
Adult Bladder
(767B1)
-9SitU
@.
@Iih.
ur
:uu$flfl
@.
ifUll
,ts•
@%t$%$$
fII@i@
In.
u8
‘Primaryunknowis
Chart 1. Summary of observations on nuclear bodies, perichromatin granules, and nuclear envelope dilation.
DISCUSSION
In this paper, transmission electron microscopy was uti
lized to evaluate the nuclear ultrastructure of human epithe
hal cell lines derived from normal tissues, nonmalignant
tissues of cancerous organs, and carcinomas. Consistent
ultrastructural markers distinguished each of these cate
gories. Perichromatin granules and nuclear bodies were
found in all of the malignant lines. Dilation of the nuclear
envelope was found in all lines derived from cancerous
organs as well as in all malignant lines but not in any lines
derived from normal tissue. Margination of chromatin,
irregularity of nuclear outline, nucleolar margination, and
redistribution of nucleolar fibrillar components were ex
pressed slightly by normal lines, to variable degrees by the
lines derived from cancerous organs, and to a much in
creased extent by the malignant lines.
All of the markers present in the tumor cells have been
reported to be present in other physiological states (9). For
example, nuclear bodies are found in normal kidney (5),
thymus (12, 27), glia (17), and neural tissue (15), in various
viral diseases, as well as in a variety of tumors (9). The
nature and function of these bodies remain obscure. Pen
chromatin granules have also been found in normal tissues
and virally infected cells; however, their number has been
reported to increase in many tumors (9). The function of
penichromatin granules is also not clear, although it has
been suggested that RNA and DNA are present. Irregulani
ties of nuclear shape, nucleolar fibnillar redistribution, and
margination have also been observed in many tissues be
sides tumors (9).
In view of the fact that these markers are found in such a
FEBRUARY1979
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
335
@
I
0@
H. S. Smith et al.
NUCLEAR
CYTOPLASMIC
ULTRASTRUCTURE
MARGINATIONNUCLEAR
INVAGINATIONSCHROMATIN
OUTLINE
I>.
@
@
>
>
@
LU
@__J
@
@
Source of Cells
(Designation) —Passage
@
o@
a
Z@
>
LU
U.
U)
LU
,__J
>LU
@_J
LU
-3
I-
09
Zcn
@9
—@?
@>
U'
5
zLU
S
I
I >.@
I
LU
xLU
I
U,
l@wI
I
zLU
I-
QLU
0
x
iI
I
i-I-
4Z
>‘@
@LU
LUO
I
I@
Derived From Carcinoma
Colon
(785T)
-9
Colon
(675T)
-4
—11
-13
...
ii:
....
-9
-13
Trans. Cell
(769T)
-12
Breast
)578T)
-9
C...
Metastatsc Pancreas)766T)
-9
—11
...
Metastatic Stomach )746T)
-14
...
::s::....
Metastatic ‘
-6
-69
.
..
(lOOT)
4696T)
C...
@•
•
S is
)761T)
Derived
@
us•
Trans. Cell
Metastatic ‘
@
C
S's
in'
::u:
:8:::
.
...•
it:
....
-33
From Peripheral
Tissue of Carcinomalous
Organ::n:
Kidney
(699K)
-5
Adult Kidney
(761K)
-7
..Adult
Adult
-13Derived
Kidney
(715K)
::v
::::
‘S
•...@
•••@
S..
....
••@•
:::
.::::
From Nonmalignant
Tissue‘4@..
Intestine
@
@
@
)741nt)
-10
Fetal Intestine
(677lnt)
-12
Fetal Intestine
(S8Olnt)
-15
-26
.Fetal
Adult Bladder
)767B1)
-9
:::::
fflj%
:::
:::::::::.
::::
::n:
•.
•
•
:n:
::::
:::::.
U:
.....• .
....
•@
is:
•
:::::
‘Primaryunknown
Chart 2. Summary of observations on cytoplasmic invaginations, chromatin granules, and nuclear outline.
wide variety of physiological states, it is somewhat surpnis
ing that consistent differences were noted between the
malignant and nonmalignant cells in vitro. For this reason,
there was no difference between the transformed and
nontransformed 3T3 cells. However, 3T3 cells are clearly
not normal, and it has been suggested that they are poised
great canewas taken to ensure objectivity. To do this, each
in a premalignant
cell line was scored by a single operator (E. Louise Spnin
also find that human tUmor cells in vitro are ultrastructurally
similar to the tumor cells that we describe; however, in their
report, no normal cell lines were presented for comparison.
In view of the current findings and the paucity of systematic
comparisons between malignant and nonmalignant cells in
vivo (8), we suggest that further studies characterizing
transformed and nontransformed cells in vitro utilizing the
criteria and recording techniques described in this paper
are warranted.
Since this series of cell lines has been derived from a
variety of organ systems, it is difficult to distinguish organ
specific properties from tumor-specific properties. In par
ticular, the 3 lines derived from nonmalignant tissue of
cancerous organs were all of kidney origin. The normal
ger) unawareof the specimen origin, and every cell section
on the grid was evaluated by determining the degree of
marker expression. One explanation for why consistent
differences have not been reported in vivo (9) may relate to
the fact that tumors contain growing, nongrowing, and
necrotic cells while their normal counterparts are usually
nongrowing. Such physiological differences may have ob
scured differences relating to cancer while growth in cuk
tune may allow tumor and normal cells to be in a more
similar growth state. Consistent differences have also not
been seen after transformation of cells in vitro (8). While
McNutt et a!. (16) found that SV4O-transformed3T3 cells
were similar to the tumor cells described in this paper,
336
state
(2). Semen
and
Dmochowski
(23)
CANCERRESEARCHVOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
@
z@
Ultrastructure of Human Epithe!ia! Ce!!s
NUCLEOLAR
ULTRASTRUCTURE
ABRILLARCENTERFORMATiONMARGINAT1ONGIANTMULTiPLE
LU
>
@
LU
-3
@
@
S
0-
Source of @.Ils
(Dssignation) —Pass.g.
@
@
>
@
LU
C,)
I
c@
LU
a
z
LU
C,)
S
LU
I
I-
z@I
>1lu
I-
x
I
I
I4
I 81
Cl)
S
I
I
I
>
I-
I
I
Ii@j
I
z
S
Z
I
LU
,-
‘U
,,,
I
I
,,@
<
I
a.
xI
I
z
z
in
in
4
a.
Derived From carcinoma
colon
l785T)
-9
folon
(675T)
-4
—11
-13
[email protected]
S..
:55.
(761T) -9
(769T)
Br.ast
(578Tl
-12
C..
CC
.
C'.
Un
S
C.
C
..
r.z
C.'.
M*taststic Stomach l746T)
-14
•.
.4:55.
‘its
Mstastatic ‘
-6
-69
C..
-33
•.
5.•::•
Z2S@
•
:8:
•.
:::
.‘..
UU:::::
55'
“
U:.
.‘
Mstastatic'
(700T)
1696T)
us.
tSSZS$
•S
5@
Its's
15$..
.5
S.
C
-9
Pancreas (766T)
S..
15....
S...
-9
—11
Metastitic
uS...
‘U
US
-13
Trans. veIl
US'S.
. .•.
•5.
.5@
I 5...
.5.
—
5555
Us..
S
r@$. S
•
..
S
-S..
S....
S....
•5@•@
••.•.
:u
uu.
S..
5$'::
.
umu.@@
S....
mit
S
2255$
CSS••@
@•
mis
Derived From Peripheral Tissu.
of Corcinomotous Organ:::n
Kidny
1699K)
-5
Adult Kidney
1761K)
-7
Mult
-13Derived
Kidney
(715Kl
$$%@@Mult
•@
u:.: .
@‘•
SS.•5
@:s'::
.5..
.@..
@S55@:zr.t
From Nonmalignant
Tissue
@
F.tal lntestin@ (74lntl
-10
.....
F.tal Intutine
(6l7Intl
-12
•.
Fasal Intastine
l65Olnt)
-15
-26
@%It$$
.5.
Adult Bladdsr
l767B1)
-9.....
•....
..
u:::$s
;I%%:z.5
.55...
...4
sss..
s:t:S
5555
.5::
@/tt55
5••SS.5
5
UU5555
.5...
a:s:ss
@..
S.s@5S555
S
5$4•
‘Primaryunknown
Chart 3. Summary of observations on nucleolar ultrastructure.
intestine lines were of fetal origin, further complicating
interpretations. However, the one adult bladder line does
represent a similar cell type to the transition cell carcinoma
lines. In addition, preliminary results comparing primary
cultures of human breast epithelial cells from normal tissue,
tissue peripheral to carcinomas, and breast carcinomas
gave similar but not identical results. In that system, nuclear
bodies and penichromatin granules were found in both
carcinomas and in the cells peripheral to the tumor but not
in the normal cells. (M. A. Stampfer, personal communica
represent specific functions defining carcinomatous trans
formation. However, the sum total of all changes may
reflect an underlying metabolic state common to carcinoma
induction. There is some evidence at the light microscopic
level (6, 31) that the nonmalignant tissue peripheral to a
carcinomatous lesion contain many atypical hyperplasias,
suggesting that the entire organ is progressing toward
malignant transformation.
Hence, those ultrastructural
markers found to varying degrees in the nonmalignant
tissue peripheral to carcinomatous lesions and to a greater
tion).
extent in the carcinomatous lines may reflect metabolic
While we urge caution in interpreting these observations,
they do have immediate applications. First, within this
system, they offer potential criteria for monitoring transfor
mation in vitro and could be valuable for assaying the
effects of carcinogens on the normal human epithelial cells.
These studies may also provide insights into the mechanism
of malignant progression. Since all of the markers are
found in other physiological states, the changes cannot
changes underlying early precancer. Therefore, biochemi
cal studies on nucleoli and nuclear envelopes might be
warranted to elucidate the nature of the ultr@astructural
difference in these organelles.
Finally, these studies point out the need to develop more
human epithelial cell lines to determine which differences
observed among the lines are related to the tissue of origin.
The fact that we have successfully developed this series of
FEBRUARY1979
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
337
@
@
H1@
i h
@
i.@
•
;@
0
9
‘@
H. S. Smith et a!.
aA!sua@X3
S
C
S
0
@
•SS
aiqisiA
C
@!‘
••
•
•55S
___________________________________________‘C
__________
@
S.
S @•
5. :..
C
.2
@
aIqfss@
u:@
@
E Al1U!e@
0
.?
55.55
i&
•.•
•••
S
S
C.)
z.2
@
wasqy
.SS
Lu
S
S
@ON
::::
•5•.
S
I
S
.5
5.
S
S
••
•s
S
UO!P@5
S.
/o@<
,
@5
@
@
ci
@
.E
@
E
2
Cf
@
S.
.5
55
S.
‘C P@W@P@I
C
.—
AIa@eJepo@j
‘C
—
2'
.5S
•5
0 ______
•5S
5Ss
0
C
S
•555
‘-
.2
—
t
‘C
5
z
.@
/OL'S
aIqislA
@
@
:
2
S
••S
•
uo!p55
@
. S
••.
—
‘C
• .
A@dieq@
S
_____
@
@
—
uo!p85
/0@'0i
S
S
P°W@P'@I
C
___ S. I
@1UC@XJ@
A@1q6!@
S
S 55
S...
@ON
U)
t
S
S
S
S
•569
@
,;
.5
••
•SS
.555
@•S
a
S
p adAj
@
________
____
@
£.@adAi
aAlsua@x@
.
@—
S
________
—
a
:
S
0
S
S
0
C
Cl)
.2
•
Aue@
e.
.5
@
___
@
a
@
z@g
::
LedAI
i
•
S
S
.
5S
5.5
a,
E
.@
E
0
C)
t
‘C
‘C
C)
M9J
a.
.
—
S
2
...
S
S
C.)
@
eic,@sv@
S
@0N
:
.
•.5
•
55.5
•5•
55•
_______
S
aIq!SIA
S
tON
$55
•5
555
555
5*5
S
5555
5555
S
S
i:'
S
•.
z
0
@
@
@
‘C
@
@
@
p
0
‘0
N
a)
_J@@
@
p-
:9.
C
.E
@I
@
_
CD
(0
::
a
“C
@CDCD
5C@iCM
C
â€ẫ€˜
a'-'.
@r-,-
C0CD
C.)
.‘-
338
CANCERRESEARCHVOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
@
-@
.@
Ultrastructure of Human Epithelial Cells
VARIOUS GROWTH STATES
NUCLEOLAR ULTRASTRUCTURE
MARGINATIONGIANTa
FIBRILLAR CENTER FORMATION
a,
CELLS
@
Growth State.
Passage
@
a
.@
a
0.@
z>
.!‘
>
0
0.@
.9
a
.@
K
>5
LU.!
z>•-.!
.0
‘(
Metastatic Pancreas
(766T)Confluent
.11
Subconfluent .11
MouseTumorS5S$$.
•5SNormal
5
5
5•••• •.S
•5•.
..S••5.5SS
$fl555
•
S.
S
S.555
S$s:.
Its.S..
•S@S.
IN'!)Confluent
Fetal Intestine (680
Subconfluent
.26
@26•5SSS
:ut55
•5.••S•
:s:5
•u:.
us:.
Chart 5. Comparison of nucleolar ultrastructure as a function of growth state.
lines suggests that such an approach is feasible. Now that
pure epithelial cultures are available, it will also be possible
to perform experiments to optimize the growth rate by
varying media. Understanding the nutritional requirements
of these cells will also facilitate development of additional
lines.
ACKNOWLEDGMENTS
We wish to acknowledge the excellent technical assistance of Alan J.
Hiller and Marie Mizelle.
REFERENCES
1. Berkey, J. J., and Zolotar, L. Development and Characterization of Cell
Lines of Normal Mouse Bladder Epithelial Cells and 2-Acetylaminofluo
rene-induced Urothelial Carcinoma Cells Grown in Monolayer Tissue
Culture. In Vitro, 13: 63-75, 1977.
2. Boone, C. W. Malignant Hemangioendotheliomas Produced by Subcu
taneous Inoculation of BaIb/3T3 Cells Attached to Glass Beads. Science,
Res., 27: 330-343, 1969.
13. Katsuta, H., and Takaoka, T. Carcinogenesis in Tissue Culture XIV:
Malignant Transformation of Rat Liver Parenchymal Cells Treated with
4-Nitroquinoline 1-Oxide in Tissue Culture. J. NatI. Cancer Inst., 49:
1563—1576,
1972.
14. Katsuta, H., and Takaoka, T. Parameters for Malignant Transformation
of Mammalian Cells Treated with Chemical Carcinogens in Tissue
Culture. In: W. Nakahara, S. Takayama, T. Sugimura, and S. Odashima
(eds.), Topics in Chemical Carcinogensis, pp. 389—400.
Tokyo: University
of Tokyo Press, 1972.
15. Masurovsky, E. B., Benitez, H. H., Kim, S. U., and Murray, M. R. Origin
Development and Nature of Intranuclear Rodlets and Associated Bodies
in Chicken Sympathetic Neurons. J. Cell Biol., 44: 172-191 , 1970.
16. McNutt, N. S., CuIp, L. A., and Black, P. A. Contact-inhibited Revertant
Cell Lines Isolated from SV4O Transformed Cells II: Ultrastructural
Studies. J. Cell Biol., 50: 691-708, 1971.
17. Misrabi, M. Intranuclear Inclusions in Neurons and Glial Cells in the
Spinal Cord of Foetal, Neonatal and Adult Rats. J. Anat., 104: 588—589,
1969.
18. Montesano, R., Drevon, C., Kuroki, T., Saint Vincent, L., Handleman,
S.. Sanford, K. K., Defeo, D., and Weinstein, L. B. Test for Malignant
Transformation of Rat Liver Cells in Culture: Cytology, Growth in Soft
Agar and Production of Plasminogen Activator. J . NatI. Cancer Inst. , 59:
1651-1658,
1977.
188:68—70,
1975.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
19. Montesano, R., Saint Vincent, L., Drevon, C. , and Tomatis, L. Produc
tion of Epithelial and Mesenchymal Tumors with Rat Liver Cells Trans
Bouteille, M., Kalefat, S. R., and DeLarue, J. Ultrastructural Variations
formed In Vitro. Intern. J. Cancer, 16: 550-558, 1975.
of Nuclear Bodies in Human Diseases. J. Ultrastruct. Res. 19: 474—486,
20. Namba, M., Masuji, H., and Sato, J. Carcinogenesis in Tissue Culture IX:
1967.
Malignant Transformation of Cultured Rat Cells Treated with 4-Nitroqui
Busch, H., and Smetana, K. The Cytochemistry and Ultrastructure of the
noline 1-Oxide. Japan. J. Exptl. Med., 39: 253-265, 1969.
Nucleolus. In: The Nucleolus, p. 73. New York: Academic Press, Inc.,
1970.
21. Owens,R.,Smith,H.,Nelson-Rees,
W.,andSpringer,E.EpithelialCell
Cultures from Normal and Cancerous Human Tissues. J. NatI. Cancer
Farquhar, M. G., and Palade, G. E. Functional Evidence for the Exist
Inst., 56: 843—849,
1976.
ance of a Third Cell Type in the Renal Glomerulas. Phagocytosis of
Filtration Residues by a Distinctive “Third―
cell. J. Cell Biol., 13: 55—87, 22. Sato, J., Namba, M., Usui, K., and Nagano, D. Carcinogenesis in Tissue
1962.
Culture VIII: Spontaneous Malignant Transformation of Rat Liver Cells in
Long-term Culture. Japan. J. Exptl. Med., 38: 105-118, 1968.
Farrow, G. M., Utz, D. C., and Rife, D. D. Morphological and Clinical
23. Seman, G., and Dmochowski, L. Ultrastructural Characteristics of Hu
Observations of Patients with Early Bladder Cancer Treated with Total
man Tumor Cells In Vitro. In: J. Fogh (ed.), Human Tumor Cells In Vitro.
Cystectomy. Cancer Res., 36: 2495-2501 , 1976.
pp. 395-486. New York: Plenum Press, 1975.
Fedoroff, S. Proper Usage of Animal Tissue Culture Terms. J. NatI.
24. Smith, H. 5. In Wtro Properties of Epithelial Cell Lines Established from
Cancer Inst., 38: 607-61 1, 1967.
Human Carcinomas and Nonmalignant Tissue. J. NatI. Cancer Inst., in
Franks,
L. M. , and Wilson,
P. D. Origin and Ultrastructure
of Cells In
press, 1979.
vitro. Intern. Rev. Cytol., 48: 55-139, 1977.
25. Smith, H., Owens, R., Nelson-Rees, W., Springer, E., Dollbaum, C., and
Ghadially, F. N. Ultrastructural Pathology of the Cell. Boston: Butter
Hackett, A. Epithelial Cell Cultures from Human Carcinomas. In: H. E.
worths, 1975.
Hackett, A. J., Smith, H. S., Springer, E. L., Owens, R., Nelson-Rees, W.
Neiburgs (ad.), Prevention and Detection of Cancer, Part 1, Vol. 2, pp.
1465-1479. New York: Marcel Dekker, 1978.
A., Riggs, J. L., and Gardner, M. B. Two Syngeneic Cell Lines from
Human Breast Tissue: The Aneuploid Mammary Epithelial (Hs578T) and 26. Spurr, A. A Low-Viscosity Epoxy Resin Embedding Medium for Electron
the Diploid Myoepithelial (Hs578Bst) Cell Lines. J. NatI. Cancer Inst., 58:
Microscopy. J. Ultrastruct. Res., 26: 31-43, 1969.
1795—1806,
1977.
27. Weber, A. F., and Frommes, S. P. Nuclear Bodies: Their Prevalence,
Heidelberger, C. Chemical Carcinogenesis. Ann. Rev. Biochem., 44: 79Location and Ultrastructure in the Calf. Science, 141: 912-913, 1963.
119, 1975.
28. Weinstein, I. B., Orenstein, J. M., Gebert, R., Kaighn, M., and Stadler,
Henry, K., and Petts, V. Nuclear Bodies in Human Thymus. J. Ultrastruct.
U. Growth and Structural Properties of Epithelial Cell Cultures Estab
FEBRUARY1979
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
339
H. S. Smith et al.
lished from Normal Rat Liver and Chemically Induced Hepatomas.
Cancer Res., 35: 253-263, 1975.
29. Weinstein, I. B., Wigler, M., and Stadler, U. Analysis of the Mechanism
of Chemical Carcinogenesis in Epithelial Cell Cultures in Screening
Tests in Chemical Carcinogenesis. In: R. Montesano, H. Bartsch, and L.
Tomatis (ads.), IARC Publications No. 12, pp. 355-381 . Lyon, France:
International Agency for Research on Cancer, 1976.
30. Weinstein, I. B., Yamaguchi, N., Orenstein, M., Gebert, R., and Kaighn,
M. E. Mechanism of Chemical CarcinogenesisAnalyzed in Rat Liver and
Hepatoma Cell Cultures. In: L. E. Gerschenson and E. B. Thompson
(ads.), pp.
@1-459.Gene Expression and Carcinogenesis in Cultured
Liver. New York: Academic Press, Inc., 1975.
31. Wellings, S. R., Jensen, H. M., and Marcum, R. G. An Atlas of Subgross
Pathology of the Human Breast with Special Reference to Possible
Precancerous Lesions. J. NatI. Cancer Inst. , 55: 231-273, 1975.
32. Williams, G. M., Weisburger, E. K., Weisburger, J. H. Isolation and Long
term Culture of Epithelial-like Cells from Rat Liver. Exptl. Cell Res., 69:
106—112,
1971.
Fig. I . Arrows, nuclei illustrating various types of nuclear bodies; inset, perichromatin granules. a, type 1 to 2 nuclear body; b, type 3 nuclear body. Inset,
two
penichromatin
granules
surrounded
by electron
lucent
halos;
c, type
4 nuclear
body.
Marker,
1 .un; inset
magnification
equals
that
of b. The tumor
cell
lines Illustrated are 746T (a and b) and 700T (C). Uranyl acetate:lead citrate.
Fig. 2. Nuclei illustrating degrees of expression of 3 nuclear characteristics “nuclear
outline,―“nuclear
envelope invagination,―and “chromatin
margination,' as well as presence or absence of nuclear envelope dilation. a , a “smooth―
nuclear outline, “not
visible―status for nuclear envelope
invaginatlon and not visible status for chromatin margination. Normal fetal intestine cell line 68OInt; b, a “slightly
indented―nuclear outline, “few―
nuclear
envelope invaginations, “faint―
chromatin margination. Cell line 761K derived from tissue peripheral to a kidney carcinoma; c, a “moderately
indented―
nuclear outline, “many―
nuclear envelope invaginations (arrow) and “visible―
chromatin margination tumor-derived cell line 746T; d, a “sharply
indented―
nuclear outline and “extensive―
degree of nuclear envelope invagination (arrow) and chromatin margination. Tumor-derived cell line 746T; a, undiluted
nuclear envelope (arrow), normal fetal intestine cell line 68OInt; f, dilated nuclear envelope (arrow) in tumor-derived line 746T. The marker for a through d is
1 @an
and 0.5 ian, e and f. Uranyl acetate:Iead citrate.
Fig. 3. Nucleoli illustrating varying degrees of fibrillar center formation. a; evenly distributed nucleolar material, normal fetal intestine cell line 68Olnt; b,
slight center formation, cell line 761K; c, moderate center formation, tumor-derived line 700T; d, extensive center formation, tumor-derived line 7461. Panels
b, c, and d, a typical electron-lucent center (C) surrounded by an electron-dense fibrillar area (F). Increasing degree of boundary delineation between lucent
and dense components. Marker = 1 aim. Uranyl acetate:Iead citrate.
Fig. 4. Nucleoli (N) illustrating varying degrees of association with the nuclear margin. a, “not
associated,―normal fetal intestine line 6801nt; b, “slight―
margination, normal fetal intestine line 74lnt. c, “moderately―
marginated, tumor-derived cell line 746; d, “extensively
marginated,―tumor-derived cell line
766T. Marker = 1 sun. Uranyl acetate:lead citrate.
340
CANCERRESEARCHVOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
@
b
@
,a@&4
@
,
@
.‘-@.
I
‘
: .@
.‘--
:@,
.
-
:r@‘@,
.‘
S.
F
,‘
@
.‘@‘,
-‘.@
,.ivq@.'.@
::..4
@
@-
@
I :
@
@
“S
‘.
. ..*
@
(‘%.“(
I'
@,,
S
.
,_..;
@::
i:@
@
@
‘4
,@.
. . .
@4•@ ,‘
. .
“A
,.@
@
.@ @:-t ?‘:
@s.
@
@
:
,s@
@t.,‘ .@,
.c;,L
.@
4
.-@
@;,t
@:
:
‘z@
:-@,fp.
@@‘L••
:“@@@4t.
:@
.rz;:'•
FEBRUARY1979
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
341
@
@
@
@
@.
#
i@
..
.
H. S. Smith et al.
@
,;@
@
.@@@r:i@Il:
@
f@
@
@
@
:4,
..
•
.. .
)I!@@t
/@
.
@,
@/“@
@ :
‘
:@
@
@
‘,
..
. .
.
I
‘
.@,
- .$.
A
#@r
L.
Ir
,t.
,)
@0
/
i@
i'..
@
â€Ã
@
.
s@
@,
.- - â€:̃.@:.
@
..‘
@
.
.,
;,J@.
‘ @.e ‘@,@
.
@‘
@
@
I
:@‘
.
@‘‘
.
,,
:
.
@
@
,
@et
?
4.@
@
@. , . ...
@,
I
,@1 C
,
.3
‘
4@
@
.
‘
‘v'.@
@
c','1_@
?
.@. :‘ ‘ . , - I
@
@
4.-
S
‘5
C
t@
.
,,.
i'.;'-
@:
,@
—
“@i'@
•‘@(
:@
-,
‘@:@‘
@
.
.
.....-
-.@
t@t.:
@
S f@@jjg@Liø:;/
342
.
@.;
:“
-
,@
@•.-
,
. ‘@.‘
:
‘
,-.
1k @i
CANCER
RESEARCH
VOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
@
,1
Ultrastructure of Human Epithelial Cells
@
:
@
@
‘S
@
“$)#@‘
,@;‘
•:@‘@@
-
@
.
£@r_,@,@―\
:-.‘*@ •.i
\
:@ ‘
@.t.
i;@,,.
@1
3'S,
‘@
.
“
‘.
‘@;_
:
.@,
•
. , .)‘@
.
.q.
@4-
@..
..@
@
-,t'
@
‘;‘_
-..
J,,@
. -
r@kd@;
@
‘ .‘
.‘,_.@
@
..‘
@
‘S.
@
.@—,‘
.
@;
@1j
@
@
@
.
1)
,@
:JO
,‘,
U-::..
,,‘
:t.-@.
.,..@..
,@.. [email protected]
.@
:@.•
,@
.i:i.@;
S
y-.'•'-
, ,
‘@.@-,,
@
:
‘@‘
..
@
@
.
..-
.
.‘
..@,..
;
•‘d@―@
@-
.
.
‘‘.@f
.â€t̃:@
‘..._:
..
@.
‘.
“F
¼
:4: :‘
@‘
:.
@
-,
@1
.@:‘
, ‘5,'
;:@. @.
-....
@;:,S.-w_@_
•
‘
@
@s
‘4....
...@
.‘@do
•‘
‘___@___
;@
3d
@
@,
•:@
‘..‘@1,
FEBRUARY1979
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
343
@
@
@
@:
:
,@:1
.,@
-.@‘.
@
-@.,
.@
@
1.;;:
£‘@@ ‘@
@
.
@
@‘ .— ‘@‘,
.@
t&
‘@
.
@‘,‘
:—‘@
“
@%.,
@
@L:.
.‘
‘,.‘)_
@
,
a
‘C.,
.:-
‘-@@‘
.$ 4.,
:
q@..
@
-
@‘,@a
@
.-.
@
‘C
@
‘
‘C
-
:;
•“@@
A
,._.1s
@
.;‘,
.
.
r―
“:@@
‘‘
‘a'.
,.,.
.,
@
4 b
@y
.@
__________
:@!.!r'\:'.
.
.@,-:
4a
@
@
@
@
@
.
.
9
:@iL@@z@'-
.‘
:,
!..,?@
•@ *c@V@
: ,
,: ,,
:
. ,@
L.@.
@.
.
:@.
@
•1
@
•
.
•
-@
.
@4:
.
.-,@ 1'
@..
-
S
@
-.@
• .,@
..,‘
2.
,
..
•
“;@‘,cr―@.
;d@
.
@
:1―
@tj4flVP'
.
V
i@i@'
•@.....
.@.,
@
.
.
•@‘,‘t
@
.
. - .%.F(@4@ۍ
A
@
..
@
.
.
p
.@,,,_.,.‘.,c,
:‘
@I&@fç,
e@è
•A@ ‘
‘
.
•..,
.
•.
“•
:
*1 @:‘@
,
@
@
,
.‘.
@
,
C,'.
@
.
‘ ‘.
. -.
-:‘@f-.,,
-:-
@;
: ‘:‘@@
@
.
@
@---‘@:
@,
..@. .
.
@
4
c
@. \
--,---..
.1
-
.‘..
5@,fr'F:5
......I,
.
..@.
‘.‘j'.'._
j
.::@
,
@--@
@..-‘
@-
:
.@
@.e
. -
@
.@.
@J
! “ ‘
.,.
,-@
@
@‘
‘
.@
.
—
t.
@.
4d
. ‘,‘;.
344
CANCERRESEARCHVOL. 39
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.
Nuclear Ultrastructure of Epithelial Cell Lines Derived from
Human Carcinomas and Nonmalignant Tissues
Helene S. Smith, E. Louise Springer and Adeline J. Hackett
Cancer Res 1979;39:332-344.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/39/2_Part_1/332
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on April 14, 2017. © 1979 American Association for Cancer Research.