/ . Embryo!, exp. Morph. Vol. 60, pp. 283-293, 1980
Printed in Great Britain © Company of Biologists Limited 1980
283
Correlative light and electron microscopy
of dissociated immature rat testicular cells
undergoing morphogenesis in vitro
By L. A. ERICKSON 1 , J. C. DAVIS*, P. R. BURTON AND
J. SNYDER 2
From the Department of Physiology and Cell Biology,
University of Kansas
(•Deceased, 27 August, 1979)
SUMMARY
Immature rat testicular cells undergo morphogenesis in primary culture (Davis, 1978).
Depending upon the number of dissociated testicular cells added to the culture dish, spherical
or tubular aggregates were formed. Spherical aggregates resulted from movement of cells
into centers of aggregation and the detachment of these cells from the substratum; on the
other hand, tubular aggregates resulted from detachment and retraction of the cell monolayer at certain points along its outer edge. In this investigation, the different methods of
formation of aggregates by immature rat testicular cells in primary culture were examined
with the scanning electron microscope (SEM). The cell types involved in such morphogenesis
and their associations within completely formed structures were examined by transmission
electron microscopy (TEM). In addition, the rates of formation of aggregates were established
by time-lapse cinemicroscopy. During formation of spherical aggregates, the rate of recruitment of cells into centers of aggregation (0-04± 0-006/*m/min; X±S.E.M., n = 78) was
much slower than the rate of cell detachment during formation of tubular aggregates
(11-7 ± 1-8 /tm/sec; JC + S.E.M., n = 110).
Although specific roles for each cell type in formation of aggregates have not been determined, the associations of cells within the two types of reformed aggregates appeared to be
similar. Myofibroblast cells were located in outer cell layers and Sertoli cells were observed to
underlie the layers of myofibroblasts in both types of aggregates. Germinal cells, however,
were found on the outer surface of spherical aggregates, but in tubular aggregates they were
located on the inner surface. Since spherical and tubular aggregates are formed by different
methods, this observation suggests that rearrangement of cells within the aggregates takes
place and contributes to the internal morphology of newly-formed aggregates.
INTRODUCTION
Dissociated, immature rat testicular cells undergo morphogenesis after 4 days
in culture to form spherical or tubular aggregates (Davis, 1978). The methods of
formation were different for the two types of structures; spherical aggregates
were formed preferentially with the addition of dissociated cells at less than
x
Author^s address: Department of Physiology and Cell Biology, 106B Snow Hall, University
of Kansas, Lawrence, Kansas 66045, U.S.A.
"[Author's address: Department of Biochemistry, Southwestern Medical School, Dallas,
Texas, U.S.A.
284
L. A. ERICKSON AND OTHERS
monolayer densities. In comparison, tubular aggregates appeared in greater
frequency than spherical aggregates with a nearly confluent monolayer present.
The organization of cellular aggregates from single cell preparations has been
studied by using rotation cultures (e.g. Moscona, 1962, 1965; Steinberg, 1963).
Also, reorganization of immature rat testicular cells has been examined with such
cultures (Ohno, Nagai, & Ciccarese, 1978; Zenzes, Wolf, Giinther, & Engel, 1978),
but the ultrastructure of reformed aggregates has not been examined. Studies
examining the formation of testicular cell aggregates by cells initially attached to
the substratum are lacking. The current study examines, by scanning electron
microscopy (SEM), the formation of aggregates from dissociated testicular cells
attached to the culture dish substratum, as well as the morphology of completely-formed aggregates using transmission electron microscopy (TEM). In
addition, the rates of cellular reassociation were analyzed by time-lapse cinemicroscopy. The results will show that (1) myofibroblast and Sertoli cells are
involved in formation of testicular cell aggregates, (2) the associations of cells
within reorganized spherical and tubular aggregates are similar except for the
association of germinal cells with each type of aggregate, and (3) the rates of cell
movement during formation of aggregates are different. The association of cells
within spherical and tubular aggregates and rates of cell movement during the
formation of each type of aggregate reflect the different methods of formation.
METHODS
/. Cell dissociation
Testes from eight 15-day-old Sprague-Dawley-derived rats (Sasco, Omaha,
Neb.) were removed, decapsulated, and placed in 2 ml NCTC-135 (Gibco).
Teased tubules were then dissociated in 10 ml of 0-1 % trypsin (Gibco) in Ca,
Mg-free phosphate-buffered saline (pH 7-40) in a 50 ml trypsinizing flask. The
tissue-free supernatant was collected at 10 min intervals, diluted with 8 % fetal
calf serum (Gibco), and filtered through a 28 jam Nitex cloth (TET-Kresslik)
(Davis & Schuetz, 1977). After 50 min, cells were collected by centrifugation at
1000 g for 10 min and resuspended in NCTC at a final concentration of 0-751-25 x 107 cells per ml. Cell number was determined with a Coulter ZBI Counter.
Only particles larger than red blood cells were counted.
II. Culture conditions
Dissociated rat testicular cells were added to NCTC-135 containing 10%
rat serum (Gibco) as follows: 7-5 xlO 5 cells/0-5 ml medium in 16 mm wells
(24-well culture plate, Falcon) or 4-8 xlO 6 cells/2.0 ml medium in 35 mm
dishes (Nunc). Cells were incubated at 32-5 °C under air:CO 2 (95 %:5%) with
penicillin 100 i.u./ml and streptomycin 100 /*g/ml. Examination of the cultures
immediately after the addition of cells indicated that this dissociation system
consistently produced single cell suspensions, with pre-attachment clusters
In vitro morphogenesis of rat testicular cells
285
consisting of 1-16 ±0-05 cells/cluster (jc± S.E.M., n == 4) (Davis, 1978). Fresh
medium was supplied initially on day 2 of culture and subsequently replenished
every 48 h. Care was taken not to remove or detach newly-formed aggregates.
IIT. Tissue preparation and examination
(A) Scanning electron microscopy (SEM)
To obtain cells for SEM examination, 12 mm circular glass coverslips (16 mm
wells) and 22 mm square glass coverslips (35 mm dishes) were flame-sterilized
and placed on the culture dish bottom prior to the addition of medium and cells.
At various times after the addition of cells, samples were fixed for 1 h in 2-5 %
glutaraldehyde in 0-1 M cacodylate buffer (pH 7-30) and then post-fixed 1 h in
1-0% osmium tetroxide buffered with cacodylate. Coverslips were rinsed in
buffer, dehydrated in an ethanol series, and passed through a Freon TF-ethanol
series in preparation for critical-point drying. After transfer into absolute
Freon TF, coverslips and adherent cells were placed in a specially designed
holder (capacity 10, 12 mm coverslips) and critical-point dried (Bomar SPC1500 instrument). Specimens were attached to aluminum stubs and coated with
gold-palladium for 2 min at 10 ma using a Technic Hummer II sputter coating
apparatus. Material was then examined with a Philips 501 scanning electron
microscope.
(B) Transmission electron microscopy (TEM)
Free-floating aggregates were fixed as outlined above. Following dehydration
in an acetone series, specimens were embedded in Araldite resin. Sections were
cut with a diamond knife and stained with aqueous uranyl acetate (30 min) and
lead citrate (10 min). Micrographs were taken with a Philips 300 transmission
electron microscope.
(C) Time-lapse cinemicroscopy
Movement of cells during formation of spherical and tubular aggregates was
analyzed by time-lapse cinemicroscopy using a Leitz phase-contrast microscope attached to a Bolex camera coupled to an Emdeco interval timer apparatus.
Only 35 mm dishes provided adequate optics for filming formation of aggregates. Rates of formation of spherical and tubular aggregates were calculated
from measurements obtained with the use of a film-editing device. Cells (1 x
106/dish or 6x 106/dish) were photographed on day 3 of culture following a
change of medium. Cells forming spherical aggregates (1 x 106/dish) were
photographed at two frames/min for 14 h; cells forming tubular aggregates
(6 x 106/dish) were photographed at 30 frames/min for 10 min.
19
EMB 60
286
L. A. ERICKSON AND OTHERS
Fig. 1. This micrograph shows foci of aggregation which are apparent soon after 10 %
rat serum is provided on day 2 of culture. Many rounded cells are located in central
regions of the foci (arrows). SEM, x 592 (scale line = 10/tm).
Fig. 2. A spherical aggregate is shown from day 4 of culture; by this time their
formation is complete. Such newly-formed aggregates are loosely attached to the
substratum. Rounded cells are apparent on the external surface of the aggregate
(arrow). SEM, x 387 (scale line = 10 /*m).
Fig. 3. The initial signs of formation of tubular aggregates are shown in this micrograph. About 2 h after fresh medium is provided on day 2 of culture, detachment
and retraction of peripheral regions of the monolayer occurs. Rounded cells are
often seen on the upper surface of the monolayer at the point of detachment. Cells
appear to be stretched in a direction perpendicular to the direction of cell movement
(arrow). SEM, x 802 (scale line = 10/tm).
Fig. 4. The tubular aggregate displayed here has undergone complete formation.
The process of formation of tubular aggregates was usually completed by 24 h after
the medium was changed on day 2 of culture. An area which contains numerous
rounded cells is apparent (arrow). SEM, x 128 (scale line = 25 /*m).
RESULTS
I. Ultrastructural observations
(A) Aggregate formation
SEM examination of dissociated rat testicular cells (4 x 106/per 35 mm culture
dish) maintained in 10 % rat serum showed foci of aggregation at the time fresh
medium was provided on day 2 of culture (Fig. 1). Spherical cells which appeared
to be loosely attached to the underlying flattened cells were located characteris-
In vitro morphogenesis of rat testicular cells
287
tically in central regions of these foci. Further aggregation and subsequent
detachment of attached cells resulted in formation of spherical aggregates after
4 days of culture (Fig. 2). Spherical aggregates ranged in size from 50 to 300 /.cm
in diameter. At this time, few free cells were observed and essentially all of the
cells attached to the substratum were incorporated into aggregates. Detachment
and withdrawal of radial cytoplasmic processes from the lower perimeter of the
spherical aggregate accompanied release of the aggregate from the substratum.
Following detachment of cells from the substratum, many rounded cells remained associated with the outer layer of flattened cells composing the aggregate. During subsequent culture, intermediate stages were seen in certain
instances which suggested that the spherical cells flatten and become part of the
outer cell surface of the aggregate.
When greater than 6 x 106 cells were added to 35 mm culture dishes and rat
serum was provided on day 2 of culture, formation of tubular aggregates was
initiated at various points along the periphery of the cell monolayer. Adjoining
cells detached from the substratum and started to retract towards the center of
the culture dish (Fig. 3). In most cases, rounded cells were attached to the monolayer surface and became incorporated into the inner cell layers of tubular
aggregates. Completely-formed tubular aggregates were seen on day 3 of culture
(Fig. 4). Tubular aggregates ranged from 100 to 200 /cm in diameter and 150 to
500 /urn in length. For the most part, the outer surface of such aggregates lacked
spherical cells; but, in spceial cases, a concentration of spherical cells was
located in a single area on the under surface of tubular aggregates (see arrow in
Fig. 4).
(B) Cell types and associations
Since spherical and tubular aggregates resulted from different methods of
formation, the possibility existed that cellular associations within the two types
of reformed aggregates were different. Although rounded cells were observed by
SEM to be associated with spherical and tubular aggregates in different ways,
examination of completely-formed aggregates by TEM showed that the cellular
associations within both types of aggregates were similar and allowed for the
identification of cell types present in spherical and tubular aggregates.
The outer layers of both types of aggregates consisted primarily of flattened
cells (Fig. 5). This cell type was characterized by the presence of elongate nuclei
and numerous inclusion droplets. Almost without exception these cells formed
the outermost cell layer in completely-formed aggregates. The ultrastructural
characteristics of such flattened cells were similar to those observed in myofibroblasts of both immature and adult rat testes in vivo. The elongate cells
observed in reformed aggregates were, therefore, termed myofibroblasts.
In both types of aggregates, cells lying directly beneath the myofibroblast
layers were characterized by the presence of a lobulated nucleus, abundant
dilated endoplasmic reticulum, dense peripheral heterochromatin, and num19-2
288
L. A. ERICKSON AND OTHERS
In vitro morphogenesis of rat testicular cells
^
-
289
^
4
''
/
•
9:.
k '
•
^
.;
Si
/
• IT
I
• • * "
•
»
t»
r
y'
• : . . . - • - • 4
A
• \
/
>' A
v
* •
'
Fig. 9. A densely-stained cell within a reformed tubular aggregate. These cells are
scattered throughout the interior of both types of aggregates and are few in number.
TEM, x 7620 (scale line = 1 /tm).
Fig. 10. A darkly-stained cell type located in seminiferous tubules of 7-day-old rat
testes. These cells are located directly beneath the myofibroblast cell layer (MF)
within the epithelium of the tubule. TEM, x 6912 (scale line = 1 /im).
F IGURES
5-8
Fig. 5. This micrograph shows a portion of the outer cell layer of a newly-formed
tubular aggregate. These elongated myofibroblasts (MF) form the limiting cell layer
of both types of aggregates. Transmission electron micrograph (TEM), x 5184 (scale
line = I /tm).
Fig. 6. The Sertoli cell shown here is located in the central region of a reformed
tubular aggregate. Abundant dilated endoplasmic reticulum (ER) and a lobulated
nucleus (N) are characteristic of this cell type. TEM, x 69.12 (scale line = 1 /tm).
Fig. 7. The cell-cell junction (arrow) shown here connects two Sertoli cells (SC) located in the central region of a spherical aggregate. Such junctions are apparent
throughout the interior of reformed spherical and tubular aggregates. TEM, x
30238 (scale line = 01 /urn).
Fig. 8. This micrograph shows the two types of germinal cells (GC) which adhere to
the surface of spherical aggregates. Some of these cells become incorporated into the
outer cell layer of aggregates. The characteristic mitochondria and nuclei are evident. TEM, x 4147 (scale line = 1 /tin).
i: • ,
n
290
L. A. ERICKSON AND OTHERS
Fig. 11. These frames from a cinemicroscopic sequence show a spherical aggregate
undergoing detachment from the substratum. The times indicated represent the
number of hours after the medium was changed on day 2 of culture: (a) 0-5 h and (b)
170 h. x53 (scale line =
erous specialized junctions (Figs. 6 and 7). These criteria compared well with
those observed in Sertoli cells of 14-day-old rat testes in vivo. On the basis of
such intracellular criteria, these cells were identified as immature Sertoli cells. Tn
rare instances, Sertoli cells within tubular aggregates surrounded a lumen.
Two classes of rounded cells were observed to be associated with the outer
cell layer of spherical aggregates. These cells were characterized by either 0 )
the presence of moderately dense mitochondria with some dilated cristae and a
nucleus with indications of chromosome condensation or (2) large, spheroidal
mitochondria and a large, angular nucleus with uniformly dispersed heterochromatin (Fig. 8). These cells were assumed to be germinal cells on the basis of
such intracellular criteria and their spherical shape (which is typical of in vivo
germinal cells at a comparable stage of differentiation). Rounded cells located
within tubular aggregates underwent rapid degeneration and necrosis.
Although few in number, a cell type with osmophilic cytoplasm and a lobulated
nucleus was seen within both spherical and tubular aggregates (Fig. 9). These
cells were located randomly throughout the reformed aggregates and often
exhibited a 'dumbbell-like' configuration. A cell type with similar characteristics was observed in 7-day-old rat testes in vivo (Fig. 10). This cell type may be
termed a Sertoli precursor cell, in line with the observations of Tung, Dorrington
& Fritz (1975).
II. Time-lapse analysis
Movement of single cells into centers of aggregation was observed to occur
spasmodically. In several cases, a sudden movement of a group of adjoining
cells was seen (Fig. 11). Time-lapse analysis of formation of spherical aggregates
indicated a rate of cell movement of 0-04 ± 0-006/tm/min (X±S.E.M., n = 78),
calculated by measuring the average diameter of the forming aggregate. Analysis
In vitro morphogenesis
of rat testicular cells
291
of the rate of movement of cells during formation of tubular aggregates was
accomplished by measuring the distance moved by the monolayer edge between
frames taken at 2 sec intervals. The velocity of movement was constant (11-7 ±
1-8/mi/sec; X±S.E.M., n = 110) until the moving cell front, which initiated at
the dish edge, approached the center of the dish. At random points, cells in the
monolayer remained attached to the substratum and slowed the movement; this
was invariably followed by a rapid acceleration and a velocity exceeding the
mean rate for periods of 4-6 sec. Although examination of the cells involved in
the two types of reaggregation by cinemicroscopy has not been correlated with
ultrastructural analysis during the same time periods, this preliminary cinemicroscopic analysis serves to characterize further the different methods of
formation of aggregates.
DISCUSSION
The four major cell types observed within both spherical and tubular aggregates were myofibroblasts, immature Sertoli cells, Sertoli percursor cells, and
germinal cells. They were identified using criteria which compared them with
corresponding cell types found in vivo in immature rat testes. In general, spherical
and tubular aggregates exhibited similar associations of cells. Since examination
of the cells attached to the substratum suggested no organized pattern of
cellular association, myofibroblasts, almost without exception, composed the
outer cell layers of both types of aggregates, and Sertoli cells were located in
central regions of aggregates, rearrangement of cells appeared to occur within
newly-formed aggregates. However, the possibility that an undifferentiated cell
type gave rise to both the peripherally located myofibroblasts and centrally
located Sertoli cells cannot be ruled out. The major morphological difference
between the two types of aggregates was the association of germinal cells with
the outer cell layer of spherical aggregates, but with the inner cell layers of tubular
aggregates. Germinal cells were associated with the upper surface of attached
cells; thus, this difference seemed to result from the different methods of formation of aggregates rather than from extensive cellular sorting within reorganized
aggregates. With germinal cells located within reformed tubular aggregates, in a
number of instances the cellular arrangement and associations observed in
tubular aggregates resembled those seen in the intact seminiferous tubule.
Although cellular necrosis was evident within the inner regions of reformed
aggregates, a majority of cells, including germinal cells, associated with the outer
cell layer of spherical aggregates, appeared to have a normal ultrastructure.
Cellular degeneration within the inner regions of the aggregates was minimal
when the size of the aggregate was such that only a few cell layers composed the
outer region (100-200 /«n in diameter). This was in contrast to the high level of
cellular degeneration observed in the inner regions of more frequently formed
multi-layered aggregates. Thus, when the size of reformed aggregates was similar
to the diameter of in vivo seminiferous tubules (150-250 /«n), the level of
292
L. A. ERICKSON AND OTHERS
cellular degeneration was low as compared with aggregates larger than seminiferous tubules.
The current study describes the cellular associations present within testicular
cell aggregates formed by either the movement of cells into discrete centers of
aggregation and subsequent detachment of such cells or the detachment and
retraction of portions of the cell monolayer. The cell types and associations
observed in these reformed aggregates compare favorably with the observations
of Eddy & Kahri (1976). They examined the associations of immature rat testicular cells as these cells migrated out from explants of seminiferous tubules in
culture. They found that Sertoli cells and elongate cells (possibly myofibroblasts) surrounded the original explant in a characteristic manner. Sertoli cells
were found to be closely associated with the explant, whereas elongate cells were
more peripherally located. In addition, germinal cells were associated with the
upper surface of Sertoli cells attached to the culture dish substratum. The results
discussed here provide a description of the association of immature rat testicular
cells initially attached to the substratum subsequent to the morphogenesis of
such cells into discrete cell aggregates, whereas Eddy and Kahri have described the associations of cells moving from a cell aggregate (explant) onto the
substratum.
Formation of spherical testicular cell aggregates from dissociated cells
involved the movement of single cells or small groups of cells into centers of
aggregation; on the other hand, tubular aggregate formation was initiated by
detachment of portions of the cell monolayer from the substratum and subsequent retraction of the monolayer edge. Rate studies carried out by cinemicroscopy showed that formation of tubular aggregates proceeded at a speed
many times that of formation of spherical aggregates. These observations are in
agreement with those of Davis (1978) and suggest that the mechanisms of
detachment of cells during formation of the two types of aggregates may be
different. Current studies are examining in greater detail stages in the formation
of aggregates in attempting to reconstruct aggregates which more closely resemble in situ seminiferous tubules. This system may prove useful in studies
concerned with the differentiation of testicular cells in culture.
This study was supported by grants from the NICHHD (5R01 HD10953-03) (J.C.D.), the
March of Dimes (5-138) (J.C.D.), and the PHS (GM-24992) (P.R.B.).
REFERENCES
J. C. (1978). Morphogenesis by dissociated immature rat testicular cells in primary
culture. /. Embryol. exp. Morph. 44, 297-302.
DAVIS, J. C. & SCHUETZ, A. W. (1977). Purification o f colony-forming' cells from immature
rat testis. Expt Cell Res. 106, 253-260.
EDDY, E. M. & KAHRI, A. I. (1976). Cell associations and surface features in cultures of
juvenile rat seminiferous tubules. Anat. Rec. 185, 333-358.
MOSCONA, A. A. (1962). Analysis of cell recombinations in experimental synthesis of tissues
in vitro. J. Cell Physiol. 60, 65-80.
DAVIS,
In vitro morphogenesis of rat testicular cells
293
A. A. (1965;. Recombination of dissociated cells and the development of cell
aggregates. In Cells and Tissues in Culture, vol. 1 (ed. E. IS. Willmer), pp. 489-524. New
York and London: Academic Press, Inc.
OHNO, S., NAGAI, Y. & CICCARESE, S. (1978). Testicular cells lysostripped of H-Y antigen
organize ovarian follicle-like aggregates. Cytogenet. Cell Genet. 20, 351-364.
STEINBERG, M. S. (1963). Reconstruction of tissues by dissociated cells. Science 141, 401-408.
TUNG, P. S., DORRINGTON, J. H. & FRITZ, I. B. (1975). Structural changes induced by follicle
stimulating hormone or dibutyryl cyclic AMP of presumptive Sertoli cells in culture.
Proc. natn. Acad. Sci., U.S.A. 72, 1838-1842.
ZENZES, M. T., WOLF, U., GUNTHER, E. & ENGEL, W. (1978). Studies on the function of
H-Y antigen: Dissociation and reorganization experiments on rat gonadal tissue. Cytogenet. Cell Genet. 20, 365-372.
MOSCONA,
(Received 7 February 1980, revised 14 May 1980)