Male germ-line stem cell potential is predicted by
morphology of cells in neonatal rat testes
Kyle E. Orwig, Buom-Yong Ryu, Mary R. Avarbock, and Ralph L. Brinster*
Department of Animal Biology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 19104
Contributed by Ralph L. Brinster, July 11, 2002
Gonocytes are a transient population of male germ-line stem cells
that are derived from primordial germ cells in the embryo and give
rise to spermatogonial stem cells, which establish and maintain
spermatogenesis in the postnatal testis. In contrast to spermato-
gonial stem cells, gonocytes can be identified easily in neonatal rat
testis cell suspensions based on their large size and distinct mor-
transgenic rats demonstrated that gonocytes are the only cells that
express a lacZ reporter transgene. Two gonocyte subpopulations,
designated pseudopod and round, were identified and isolated
from neonatal (0–4 days postpartum) rat testis cell suspensions.
Male germ-line stem cells, identified by their ability to produce and
maintain colonies of spermatogenesis upon transplantation into
infertile recipient testes, were present almost exclusively in the
pseudopod gonocyte subpopulation. In contrast, annexin V stain-
ing indicated that the majority of round gonocytes undergo apo-
ptosis. These results indicate that a nearly pure population of male
germ-line stem cells can be prospectively identified in neonatal rat
testis cell suspensions by morphological criteria. Together, the
the establishment of spermatogenesis in the postnatal testis.
grate to the seminiferous tubule basement membrane, and give
rise to stem cell spermatogonia. Maintenance of spermatogen-
esis and other self-renewing systems in postnatal animals de-
pends on the activity of resident stem cells that have the capacity
to both self-renew and produce progenitors that give rise to the
specified cell lineage. Male germ-line stem cells are unique
among self-renewing systems of postnatal animals because they
can pass genes, through the germ line, to subsequent genera-
tions. Despite their critical position in mammalian physiology,
there is a paucity of information regarding the molecular and
biochemical characteristics of male germ-line stem cells. This
results, in part, from the fact that these cells are extremely rare,
comprising 1 in 3,333 cells of the adult mouse testis (1) and 1 in
500 cells of the adult rat testis (2, 3). Establishment of pure or
significantly enriched populations of male germ-line stem cells is
a critical first step that will facilitate biological and molecular
The development of a functional transplantation assay for
hematopoietic stem cells over 40 years ago (4, 5) enabled
investigators to identify biochemical markers (6), develop en-
richment strategies, and eventually purify hematopoietic stem
cells (7, 8). Spermatogenesis is the only other self-renewing
system for which a stem cell functional assay is available, and
male germ-line stem cells are defined by their ability to generate
and maintain colonies of spermatogenesis upon transplantation
into infertile recipient testes (9–11). Using this assay as a
functional endpoint, we have developed methods for enriching
spermatogonial stem cells from adult testis cell populations by
taking advantage of physical binding properties (3, 12, 13),
surgical manipulation (13), and immunoselection (14). In the
best case, a 166-fold enrichment of spermatogonial stem cells
was achieved by using fluorescence-activated cell sorting
permatogenesis is initiated in the testes of postnatal mam-
mals when quiescent gonocytes resume proliferation, mi-
(FACS) to isolate a subpopulation of cells from adult mouse
cryptorchid testes (14). This testis cell population, characterized
as cryptorchid?side scatterlo(SSClo)??6-integrinhi??v-inte-
grin(?), has a stem cell concentration of about 1 in 30 and
provides a valuable resource for further characterization of
spermatogonial stem cells. Continued development of enrich-
population of spermatogonial stem cells.
During development, germ-line stem cells first can be iden-
tified as a distinct population of primordial germ cells (PGCs)
that arises from the embryonic ectoderm (15). PGCs proliferate
and migrate to the genital ridge where they associate with
somatic cells of the presumptive gonad (16). In females, PGCs
form oocytes, which stop dividing and enter meiosis. In males,
the germ cells are enclosed in the sex cords, become gonocytes,
and cease dividing until after birth (17). The differentiation of
PGCs into gonocytes marks the transition from a cell with
multiple potentials, because pluripotent embryonic germ cells
can be derived from PGCs (18, 19), to one that has the restricted
potential to develop the male germ cell lineage (17, 20). There-
fore, gonocytes are the first stem cells committed to male
germ-line development and the only germ cells in the neonatal
testis. Shortly after birth, some gonocytes resume proliferation,
migrate to the basement membrane of seminiferous tubules
(21–25), presumably differentiate into spermatogonial stem
cells, and initiate spermatogenesis (21, 22, 26, 27). However,
gonocytes of the immature testis are a complex population from
which only a portion are destined to become stem cells (17). A
significant number of gonocytes (30–75%) degenerate (22, 23,
28, 29), whereas some appear to differentiate directly into type
A1 spermatogonia (17). The molecular and biochemical char-
cell, differentiate, or die are not known. Histological and in vitro
studies demonstrate that gonocytes can be readily identified in
immature testes and testis cell cultures based on distinct mor-
phological characteristics (27), they can be isolated to homoge-
neity by micromanipulation (30, 31), and they can be maintained
in culture (24, 25, 30–34).
We took advantage of the morphological characteristics of
gonocytes, their purity as the only germ cells in the neonatal
testis, and their unique expression of a transgene to identify two
subpopulations of gonocytes (pseudopod and round) from dis-
persed neonatal rat testis cell suspensions. These two popula-
tions have distinct developmental potential, which is already
established at the time of birth; pseudopod cells become stem
pure subpopulations of gonocytes with specific destinies has
enormous potential for identification of their biological and
Donor Rats and Cell Collection. Donor testis cells were obtained
from neonatal Sprague–Dawley rats of 0–4 days postpartum
Abbreviations: 6-CFDA, 6-carboxyfluorescein diacetate; dpp, days postpartum; MT, metal-
lothionein I; PGC, primordial germ cell; X-Gal, 5-bromo-4-chloro-3-indolyl ?-D-galactoside.
*To whom reprint requests should be addressed.
September 3, 2002 ?
vol. 99 ?
(dpp, day of birth is 0 dpp) carrying a fusion transgene composed
of the mouse metallothionein I (MT) promoter and the Esche-
richia coli lacZ (lacZ) structural gene encoding a nuclear local-
ized ?-galactosidase protein (35, 36). The lacZ transgene is
expressed in germ cells derived from MT-lacZ donor rats (36),
facilitating the identification of donor-derived spermatogenesis
after transplantation into recipient seminiferous tubules by
staining with the ?-galactosidase substrate, 5-bromo-4-chloro-
3-indolyl ?-D-galactoside (X-Gal, ref. 37). Testes of MT-lacZ
neonatal rats were decapsulated and cell suspensions were
prepared by digestion of seminiferous tubules in Hank’s bal-
anced salt solution containing trypsin (0.17%) and DNase I (2.3
mg?ml) for 5 min at 37°C followed by strong pipeting. Enzymatic
aggregates were removed by straining through a 40-?m nylon
mesh. Dispersed cells were suspended in DMEM contain-
ing 10% FBS and plated in 35-mm Petri dishes at a density
(0.5 ? 106cells?dish) that allowed visualization of single cell
Selection of Gonocytes by Micromanipulation. Two subpopulations
of gonocytes (pseudopod and round) were identified in dis-
persed cell suspensions from neonatal rat testes (0–4 dpp) by
using a phase-contrast microscope, based on size (12–15 ?m)
and distinct morphological characteristics (low intracellular
complexity, prominent round nuclei, and in some cases pseudo-
pods) that have been described previously by others (25).
Individual gonocytes with smooth round or pseudopod mor-
phologies were selected with a micromanipulator and a glass
capillary pipette (tip diameter ?20 ?m), during the first 1–2 hr
after testis cell collection, and deposited into a second 35-mm
Petri dish or on poly-L-lysine-coated slides for transplantation or
staining, respectively. To assess ?-galactosidase activity, uns-
elected and selected populations were fixed in 0.5% glutaralde-
hyde for 5 min at 22°C, washed three times with PBS, and stained
Annexin V Staining and Detection of Apoptosis. Annexin V staining
(Apoptosis Detection Kit, Annexin V-CY3, Sigma) was used to
identify cells in early stages of apoptosis and exhibiting loss of
plasma membrane asymmetry. Selected round or pseudopod
gonocyte subpopulations (from 0 to 4 dpp rat testes) were placed
in drops (50 ?l) of PBS on poly-L-lysine-coated slides and
allowed to adhere to the slide by incubating for 10 min at 22°C.
Cells were then washed three times in binding buffer (10 mM
Hepes, pH 7.5?140 mM NaCl?2.5 mM CaCl2) and double-
stained with CY3-labeled annexin V (1 ?g?ml) and 6-carboxy-
min at 22°C. After the incubation, cells were washed five times
with binding buffer and observed immediately with an epifluo-
Recipient Mice and Transplantation Procedure. Unselected donor
testis cell populations (107cells per ml) and selected gonocytes
(from 0 to 4 dpp rat testes) were transplanted into immunode-
ficient NCr nude (nu?nu, Taconic Farms) recipient mice that
were treated with busulfan (44 mg?kg, Sigma) at 4–6 weeks of
age (36, 38). Busulfan treatment destroys the majority of en-
dogenous germ cells and creates space for donor-derived sper-
matogenesis. Selected rat gonocytes (12–40 gonocytes?testis)
were injected with carrier cells (107cells per ml) obtained from
the testes of Wv?W54or W?Wvmutant mice that are incapable
of producing spermatogenesis because of a mutation in the c-kit
receptor tyrosine kinase (39). Approximately 10 ?l of donor cell
suspension was introduced through the efferent ducts of each
recipient testis, about 6 weeks after busulfan treatment (40).
Recipient mice were anesthetized by Avertin injection (640
Analysis of Recipient Testes. Three months after transplantation,
recipient mouse testes were collected and stained with X-Gal to
visualize donor-derived spermatogenesis (37). Donor spermato-
gonial stem cells are defined by their ability to produce and
maintain colonies of spermatogenesis in recipient testes, and
each colony is thought to be clonally derived from a single
spermatogonial stem cell (37, 41). Expression of the lacZ
transgene allows unequivocal identification of donor stem cell-
derived colonies because differentiated germ cells cannot pro-
duce and maintain colonies of spermatogenesis and endogenous
germ cells do not express the transgene.
Statistical Analyses. ?2analyses were used to compare the relative
frequencies, stem cell potential, and apoptotic activity of pseu-
dopod and round gonocytes in neonatal rat testes (SPSS, Chi-
cago). Pearson and Mantel–Haenszel ?2analyses were used to
test for linear trends in cell frequencies, stem cell potential, and
apoptotic activity across 0, 1, 2, 3, and 4 dpp.
Quantification and Morphological Characterization of Gonocytes in
Neonatal Rat Testis Cell Suspensions. Histological examination and
staining of neonatal MT-lacZ rat testes demonstrated that lacZ
expression was in large cells located in the seminiferous tubule
lumen, but no staining was observed in other cells (Fig. 1).
Classical morphological studies have established clearly that
gonocytes are a distinct population of relatively large cells
located in the seminiferous tubule lumen of newborn rats (27).
Therefore, blue staining correlates well with gonocyte morphol-
ogy, and this feature provided a rapid assay to identify gonocytes
neonatal MT-lacZ rat testes and stained with X-Gal, 1.41%
(571?40,428) were blue (Fig. 2, Table 1). Two distinct gonocyte
morphologies, designated pseudopod (Fig. 2C) and round (Fig.
2D), were evident when stained testis cell suspensions were
examined at high magnification, and these subpopulations were
lacZ-encoded ?-galactosidase activity. (A) Macroscopic image showing that
magnification image of an individual seminiferous tubule from the testis
shown in A demonstrating that lacZ expressing cells are located in the center
of the seminiferous tubule. (C) Histological analysis confirms that lacZ-
expressing cells are located in the seminiferous tubule lumen, the character-
istic location of gonocytes in neonatal testes. (D) Higher magnification of the
seminiferous tubule boxed in C reveals the large size of gonocytes relative to
other cells in the neonatal rat testis. Counterstain (C and D), hematoxylin and
eosin. [Bars ? 0.5 mm (A), 40 ?m (B), 100 ?m (C), and 30 ?m (D).]
MT-lacZ neonatal rat testis (2 dpp) stained with X-Gal to detect
Orwig et al.
September 3, 2002 ?
vol. 99 ?
no. 18 ?
equally represented in the neonatal rat testis (P ? 0.78). From
571 total gonocytes (blue cells) counted, 45.0% were classified
as pseudopod (257?571) and 42.6% were classified as round
(243?571). The remaining 12.4% of blue cells (71?571) could not
relative numbers of pseudopod, round, and other gonocytes was
consistent for all replicates analyzed from 0 to 4 dpp (P ? 0.54).
Isolation of Gonocyte Subpopulations from Neonatal Rat Testis Cell
Suspensions. To determine whether viable pseudopod and round
gonocytes could be identified and isolated at high purity from
unstained neonatal rat testis cell suspensions, individual cell
populations were selected based on size and morphological
characteristics, using phase-contrast microscopy and a micro-
manipulator. Relative to other cells in neonate rat testis cell
suspensions, gonocytes can be recognized readily by their large
size (12–15 ?m), prominent nuclei, and low intracellular com-
plexity (few organelles are revealed by light diffraction). In
addition, round gonocytes have an unusually smooth surface,
whereas pseudopod gonocytes have distinct cytoplasmic exten-
sions (Fig. 2 C and D). Cells selected by using these character-
istics were deposited on poly-L-lysine-coated slides, fixed, and
stained with X-Gal to assess the accuracy of gonocyte selection.
Blue staining indicates that nearly homogeneous populations of
pseudopod (?99%) and round (?90%) gonocytes can be iso-
lated routinely from neonatal rat testis cell suspensions (Fig. 2
E and F). Occasionally (?1%), a gonocyte that appears round
will develop a pseudopod after selection. The gonocytes, char-
acterized as other, were morphologically heterogeneous and
difficult to identify and select prospectively from unstained testis
Transplantation and Functional Analysis of Stem Cell Activity in
Selected Gonocyte Subpopulations.Invitroandinvivomorphologic
studies suggest that gonocytes in neonatal rat testes are a
functionally heterogeneous population from which some will
develop into spermatogonial stem cells (17), some will differ-
entiate (17), and some will degenerate (22, 23, 28, 29). Results
described in the present study indicate that newly isolated
neonatal rat gonocytes are also phenotypically heterogeneous.
To ascertain whether cell phenotype predicts function, homo-
geneous populations of pseudopod and round gonocytes were
transplanted into nude mouse recipient testes to determine their
spermatogonial stem cell activity. Recipient animals were ana-
lyzed 3 months after transplantation, and the results demon-
strate that the majority of spermatogonial stem cell activity was
found in the pseudopod subpopulation of gonocytes (Table 2,
produced and maintained colonies of spermatogenesis in recip-
ient seminiferous tubules (Table 2 and Fig. 3). In contrast, round
gonocytes produced only one colony of spermatogenesis from
579 transplanted cells (Table 2) and, as mentioned above, it is
possible that this one colony resulted from the selection of a
pseudopod gonocyte that was mistakenly identified as round.
Transplantation results were consistent for all replicates ana-
lyzed from 0 to 4 dpp (P ? 0.2).
(arrow) gonocytes can be recognized by their large size and distinct morphol-
ogy. (B) Stained, unselected cell suspension. Pseudopod (arrowhead) and
round (arrow) gonocytes express the lacZ transgene and stain blue in the
presence of X-Gal. The majority of gonocytes (blue cells) in MT-lacZ neonatal
rat testes can be categorized as pseudopod or round and these two pheno-
magnification pictures of pseudopod (C) and round (D) gonocytes demon-
strating large size and unique morphology. Distinct pseudopod extensions
because ?-galactosidase activity is localized in the nucleus, pseudopod exten-
sions do not stain blue. Homogeneous populations of pseudopod (E) and
round (F) gonocytes were selected from unstained, live testis cell suspensions,
placed in a second Petri dish, and stained with X-Gal to assess selection
accuracy. [Bars ? 200 ?m (A and B) and 50 ?m (C–F).]
Dispersed cell suspensions from MT-lacZ neonatal rat testes. (A)
Table 1. Classification of gonocytes in neonatal MT-lacZ rat
testis cell populations
No. of cells‡
Percent of total§
*Single cell suspensions were prepared from the testes of neonatal MT-lacZ
transgenic rats (0–4 dpp) and stained with the ?-galactosidase substrate,
X-Gal. Blue cells (gonocytes) were morphologically classified as pseudopod,
round, or other.
†Sum of pseudopod, round, and other gonocytes.
§Percent of total cells counted (40,428).
Table 2. Colonization of recipient testes after transplantation of
pseudopod, round, or unselected neonatal rat testis cells
6.4 ? 106
*Single cell suspensions were prepared from the testes of neonatal MT-lacZ
transgenic rats (0–4 dpp), and pseudopod, round, or unselected cells were
transplanted into immunodeficient nude mouse recipient testes.
†Results are combined from 5–8 replicate experiments, encompassing 0–4
‡Percent of cells in the transplanted rat donor population that produced and
maintained colonies of spermatogenesis in nude mouse recipient testes.
www.pnas.org?cgi?doi?10.1073?pnas.182412099Orwig et al.
Analysis of Apoptotic Activity in Selected Gonocyte Subpopulations
by Annexin V Staining. Because the pseudopod phenotype indi-
cates germ-line stem cell potential, we hypothesized that the
round phenotype might represent the gonocyte subpopulation
that is destined to undergo apoptosis and degenerate. To test this
hypothesis, homogeneous populations of pseudopod or round
gonocytes were stained with annexin V. Viable cells maintain an
asymmetric arrangement of phospholipids in the plasma mem-
brane such that phosphatidylserine (PS) normally is located on
the inner leaflet, oriented toward the cell cytoplasm. Annexin V
is one of a family of proteins that binds with high affinity to
anionic phospholipids, such as PS, in the presence of calcium.
One of the early cellular changes associated with apoptotic cell
death is the loss of plasma membrane asymmetry without loss of
membrane integrity (reviewed in refs. 42 and 43). Consequently,
PS is exposed on the cell surface of apoptotic and necrotic cells
and is available for binding by annexin V-CY3 that gives red
fluorescence (42). Apoptosis can be distinguished from necrosis
when the nonfluorescent compound 6-CFDA is hydrolyzed to
produce the green fluorescent compound, 6-carboxyfluorescein
(6-CF) in viable cells (44). Three results are possible with this
double-staining procedure: (i) live cells stain only with 6-CFDA
(green), (ii) necrotic cells stain only with annexin V-CY3 (red),
and (iii) cells in the early stages of apoptosis stain with 6-CFDA
(green) and annexin V-CY3 (red). The results in Fig. 4A
demonstrate that few pseudopod gonocytes exhibit annexin V
displayed red fluorescence (Lower Center). Green fluorescence
demonstrated that most cells (?92%) in both subpopulations
were alive (Fig. 4A Right), although round gonocytes contained
significantly more necrotic cells (red only) than pseudopod
gonocytes (7.9% vs. 1.0%, P ? 0.01, data not shown). Overlap-
ping red and green in the majority of round gonocytes produced
yellow fluorescence, demonstrating apoptotic activity (Fig. 4A
Lower Right). Fig. 4B provides a quantitative analysis of the
results shown in Fig. 4A; 2.2% of viable pseudopod gonocytes
were undergoing apoptosis, compared with 66.7% of viable
round gonocytes (P ? 0.01). Necrotic cells were not included in
this analysis. Although apoptosis staining in pseudopod cells was
consistently low for all replicates analyzed (P ? 0.9, 0–4 dpp),
the apoptotic activity of round cells tended to increase from
approximately 55% to 75% between 0 and 4 dpp (P ? 0.01).
Gonocytes represent a transient population of male germ-line
stem cells that bridge the developmental gap between primordial
germ cells in the embryo and spermatogonial stem cells in the
postnatal testis. At the time of birth in rats, quiescent gonocytes
appear uniformly round and located in the center of the semi-
niferous tubules (ref. 27, Fig. 1). Within a few days (?2–5 dpp)
important changes in the gonocyte population occur that are
essential for the establishment of spermatogenesis in the post-
natal testis. Histological observations demonstrate that some
gonocytes develop cytoplasmic extensions, which may presage
migration to the seminiferous tubule basement membrane (23–
25) to a position important for establishing the germ line. This
relocation event is critical because gonocytes that fail to relocate
and remain centrally located in the seminiferous tubule even-
tually degenerate (21, 23). Whether gonocytes at birth are a
homogeneous population, whose developmental fate will be
determined subsequently by location in the seminiferous tubule
at a critical time postpartum, or a heterogeneous population, in
which cells are intrinsically programmed for migration or de-
generation, cannot be determined histologically.
from neonatal MT-lacZ rats and transplanted into nude mouse recipient
testes. (A) Macroscopic picture of nude mouse recipient testis 3 months after
transplantation showing a typical colony of donor-derived spermatogenesis.
(B) Histological analysis demonstrates that donor pseudopod gonocytes pro-
duced normal colonies of spermatogenesis with multiple germ cell layers and
spermatozoa in the tubular lumen. Blue staining in differentiated germ cells
in the testes of MT-lacZ rats is variable and may reflect stage of the spermat-
ogenic cycle. Counterstain: eosin (B). [Bars ? 2 mm (A) and 20 ?m (B).]
gonocyte subpopulations by annexin V staining. (A) Selected pseudopod
(Upper) and round (Lower) gonocytes were double-stained with CY3-labeled
annexin V (red) and 6-CFDA (green) to identify apoptotic and viable cells,
respectively. Phase-contrast microscopic images show homogeneous popula-
tions of pseudopod (A Upper Left) and round (A Lower Left) gonocytes.
Pseudopod gonocytes rarely exhibit annexin V staining (A Upper Center), but
many round gonocytes bind annexin V (A Lower Center), indicating that they
are dying. Staining with 6-CFDA demonstrates that the majority of selected
pseudopod (99%, A Upper Right) and round (92% A Lower Right) are viable.
Overlapping red and green produces yellow fluorescence in most round
gonocytes (A Lower Right), demonstrating that the annexin V staining can be
attributed to apoptosis (a regulated event) rather than necrosis. The fluores-
cence results (A Center and Right) were obtained by using an epifluorescent
nm) and H3 (Right, green ? red: excitation, 420–490 nm; emission, ?515 nm)
filter cubes. (Bar ? 100 ?m.) (B) Quantitative analysis of annexin V staining
results indicate that only 2.2 ? 0.9% of viable pseudopod gonocytes are
Necrotic cells were excluded from this analysis. Values (means ? SEM) were
from eight replicate experiments encompassing 0–4 dpp. An average of
42.6 ? 4.0 cells was evaluated for each replicate.
Evaluation of apoptotic activity in selected pseudopod and round
Orwig et al.
September 3, 2002 ?
vol. 99 ?
no. 18 ?
Gonocytes have unique characteristics (e.g., large size) that
allow them to be identified in and isolated from heterogeneous
testis cell populations (refs. 20, 32, and 45, current study). They
can be maintained in gonocyte?Sertoli cell cocultures and can,
in some measure, recapitulate their in vivo activities, including
the development of cytoplasmic extensions and migratory ac-
tivity (24, 32–34). However, these activities were not observed
results from these complex culture systems reflect gonocyte
biology in vivo is not known. In the current study, the demon-
rat testis that express the lacZ transgene (Figs. 1 and 2) was
critical to confirm the accuracy of gonocyte selection from
dispersed testis cell suspensions. This feature allowed us un-
equivocally to identify and characterize gonocytes and provided
assurance that large cells selected from unstained viable cell
suspensions were gonocytes and not large somatic cells. These
cells were isolated at a time (0–4 dpp) preceding and concurrent
with the observed onset of gonocyte migration and degenera-
of freshly collected neonatal rat testis cell suspensions (Table 1),
which is in excellent agreement with previous reports from van
Dissel-Emiliani et al. (1%, ref. 20) and Li et al. (1–2%, ref. 45)
and lower than the estimates of Orth and Boehm (32), who used
older rat pups (3–6%, 6 dpp). Examination of isolated gonocytes
during the first 1–2 hr after collection indicated that the majority
(87.6%, 500?571, Table 1) could be assigned to one of two
distinct morphological categories, pseudopod or round. It is
possible that these two subpopulations are analogous to the
migrating and degenerating gonocytes described by others (22–
gonocytes is difficult to assess in histological section, but in vitro
studies found that whereas most gonocytes were round at the
beginning of culture, 2.3–22% of gonocytes developed cytoplas-
mic extensions by 2–4 days (24, 32, 33). In contrast, our analysis
of fresh neonatal rat testis cell suspensions indicates that pseu-
dopod and round gonocytes are equally represented (45% vs.
42.6% of total gonocytes) and this relationship did not change
during the first 4 days of postnatal life. The discrepancy in the
proportion of pseudopod to round gonocytes between this and
previous in vitro studies might be attributed to differences
between fresh and cultured testis cells. For example, identifica-
tion and selection, particularly of pseudopod gonocytes, be-
comes increasingly difficult after 1–2 hr in vitro. Furthermore,
the presence of feeder cells and the culture environment in
previous studies likely affected cell morphology. Finally, approx-
imately 70% of rat pup male germ-line stem cells are lost during
the first 12 hr of culture (46).
Using the spermatogonial transplantation system to evaluate
germ-line stem cell activity, we found that stem cells reside
almost entirely in the pseudopod gonocyte population. Approx-
imately 5% of pseudopod gonocytes generated colonies of
spermatogenesis after transplantation, suggesting that this pop-
ulation of cells is identical or very similar functionally to adult
spermatogonial stem cells. Based on morphological estimates,
there are approximately 35,000 stem cells in an adult mouse
testis, which constitutes about 0.03% of total germ cells in the
testis, or one stem cell in 3,333 total cells (1). When adult mouse
testis cells are transplanted, about 19 colonies of spermatogen-
esis are generated per 106cells transplanted, or one colony in
52,632 total cells (3, 37, 41). Therefore, about one colony of
spermatogenesis is produced for every 16 stem cells (3,333?
52,632), resulting in a transplantation efficiency of 6.3%. A
similar efficiency was observed (8.5%) when adult rat testis cells
were transplanted into nude mouse recipient testes (3). Thus, the
production of spermatogenic colonies by 5% of transplanted
pseudopod gonocytes in the present study suggests that these
cells represent a nearly homogeneous population of male germ-
line stem cells. Furthermore, to produce colonies of spermato-
genesis in the transplantation assay, pseudopod gonocytes had to
migrate from the lumen of the recipient seminiferous tubule to
establish residence on the basement membrane. These results
strongly suggest that pseudopod gonocytes represent the migrat-
ing population, described in histological studies, which are
Considering these results, if 0.64% of cells in the neonatal rat
testis are stem cells (pseudopod gonocytes, Table 1) and 5% of
these produce colonies in the transplantation assay, we would
expect 0.032% (0.64% ? 5%) of unselected neonatal rat testis
cells to produce and maintain colonies of spermatogenesis upon
on our gonocyte data are nearly 3-fold higher than actually
observed when unselected neonatal rat testis cells were trans-
planted into nude mouse recipient tubules (0.032% vs. 0.012%,
Table 2). Thus, the selected gonocyte population appears to
colonize more efficiently than the unselected population. Per-
haps the W mutant mouse carrier cells, which contain many
mouse Sertoli cells, enhanced the colonizing efficiency of se-
lected rat pseudopod gonocytes in nude mouse recipient semi-
niferous tubules. The colonization performance of pseudopod
gonocytes and the absence of stem cell activity in round gono-
cytes suggest that the pseudopod gonocyte population can
account for the entire male germ-line stem cell activity found in
neonatal rat testes.
development and disrupting the balance of survival vs. death
signals in genetic models results in spermatogenic dysfunction
(47–49). Morphological studies revealed that between 30% and
75% of gonocytes are lost during the first few postnatal days (22,
23, 28), which can be attributed in large part to apoptotic cell
death starting on 2 dpp in the rat (50). Our results are in
agreement with these observations because annexin V staining
demonstrated that the majority of round gonocytes are apoptotic
(Fig. 4), consistent with their failure to produce colonies of
spermatogenesis, and these cells comprise 42.6% of the total
gonocyte population (243?571, Table 1). Our ability to detect
apoptotic activity in round gonocytes starting on the day of birth,
at least 2 days earlier than previous in situ observations (50), may
be related to the use of annexin V staining, which reflects very
early stages of apoptosis (42). Therefore, the round gonocytes
identified in fresh neonatal rat testis cell suspensions probably
represent the cells that fail to migrate, remain centrally located
in the seminiferous tubule and degenerate in vivo. Interestingly,
Wang et al. (29) found that in the early postnatal period in the
features of necrotic rather than apoptotic cell death. Although
we observed more necrosis in the round than pseudopod pop-
ulation of testis cells, the majority of round gonocytes were alive
(?92%, 6-CFDA staining) and exhibited early signs of apoptosis
(annexin V staining, Fig. 4). These results may reflect a differ-
ence between the rat and mouse during the time when spermat-
ogenesis is established in the postnatal testis.
The spermatogonial transplantation system provides an es-
sential biological assay to retrospectively characterize stem cell
activity in donor testis cell populations, although there is an
inherent 2- to 3-month delay between transplantation and
analysis. In the current study, we have demonstrated that phe-
notype predicts function in the heterogeneous gonocyte popu-
lation of neonatal rat testes. Based on these results it is now
possible to prospectively identify and isolate nearly homoge-
neous populations of male germ-line stem cells (pseudopod
gonocytes) and cells destined for programmed cell death (round
gonocytes) using morphological criteria. This progress was pos-
sible because (i) our analyses were restricted to testis cell isolates
during the first 1–2 hr after donor cell collection, (ii) lacZ
www.pnas.org?cgi?doi?10.1073?pnas.182412099 Orwig et al.
staining provided a rapid assay to evaluate strategies for gono-
cyte selection, and (iii) the spermatogonial transplantation
system allowed definitive identification?confirmation of sper-
matogonial stem cell activity in selected gonocyte populations.
Although gonocyte migration and degeneration are not ob-
served until after postnatal day 2 in the rat (23, 25, 28, 50), the
present investigation suggests that developmental fate is already
determined at the time of birth. This conclusion follows from the
consistent relationship between the number and biological ac-
tivity of pseudopod and round gonocytes from 0 to 4 dpp. It
seems unlikely that significant numbers of cells changed mor-
phology?function from pseudopod to round, or the reverse,
because we rarely observed a round cell develop a pseudopod
(?1%), round cells almost never made colonies, and the fraction
of pseudopod cells that produced colonies (5%) remained
constant. The disparate developmental destinies (stem cell vs.
death) are undoubtedly orchestrated by divergent genetic pro-
grams. Thus, the pseudopod and round gonocyte cell popula-
tions will provide valuable tools for identifying molecular mech-
anisms controlling cell fate and the establishment of
spermatogenesis in the postnatal testis. In addition, examination
of these cell populations may provide general insight into stem
cell biology and apoptotic processes that are critical for the
development and maintenance of a variety of self-renewing
We thank Drs. R. Behringer, H. Kubota, and E. Sandgren for critical
evaluation of the manuscript and helpful comments. We appreciate the
assistance of C. Freeman and R. Naroznowski with animal maintenance
and experimentation, C. Brensinger for statistical assistance, and J.
Hayden for help with photography. The MT-lacZ transgenic rat line was
a gift from R. Hammer. Microscopic sections were produced in the
Institute for Human Gene Therapy, Cellular Morphology Core, Uni-
versity of Pennsylvania (5-P30-DK-47747-07). Financial support for the
research was from the National Institutes of Health Institute of Child
Health and Human Development Grant 36504; the Commonwealth and
General Assembly of Pennsylvania; and the Robert J. Kleberg, Jr., and
Helen C. Kleberg Foundation.
1. Tegelenbosch, R. A. J. & de Rooij, D. G. (1993) Mutat. Res. 290, 193–200.
2. Huckins, C. (1971) Anat. Rec. 169, 533–557.
3. Orwig, K. E., Shinohara, T., Avarbock, M. R. & Brinster, R. L. (2002) Biol.
Reprod. 66, 944–949.
4. Till, J. E. & McCulloch, E. A. (1961) Radiat. Res. 14, 213–222.
5. Harrison, D. E. (1980) Blood 55, 77–81.
6. Spangrude, G. J. (1989) Immunol. Today 10, 344–350.
7. Smith, L. G., Weissman, I. L. & Heimfeld, S. (1991) Proc. Natl. Acad. Sci. USA
8. Osawa, M., Hanada, K., Hamada, H. & Nakauchi, H. (1996) Science 273,
9. Brinster, R. L. & Avarbock, M. R. (1994) Proc. Natl. Acad. Sci. USA 91,
10. Brinster, R. L. & Zimmermann, J. W. (1994) Proc. Natl. Acad. Sci. USA 91,
11. Brinster, R. L. (2002) Science 296, 2174–2176.
12. Shinohara, T., Avarbock, M. R. & Brinster, R. L. (1999) Proc. Natl. Acad. Sci.
USA 96, 5504–5509.
13. Shinohara, T., Avarbock, M. R. & Brinster, R. L. (2000) Dev. Biol. 220,
14. Shinohara, T., Orwig, K. E., Avarbock, M. R. & Brinster, R. L. (2000) Proc.
Natl. Acad. Sci. USA 97, 8346–8351.
15. Ginsburg, M., Snow, M. H. & McLaren, A. (1990) Development (Cambridge,
U.K.) 110, 521–528.
16. Tam, P. P. L. & Snow, M. H. L. (1981) J. Embryol. Exp. Morph. 64, 133–147.
17. de Rooij, D. G. (1998) Int. J. Exp. Pathol. 79, 67–80.
18. Resnick, J. L., Bixler, L. S., Cheng, L. & Donovan, P. J. (1992) Nature (London)
19. Matsui, Y., Zsebo, K. & Hogan, B. L. (1992) Cell 70, 841–847.
20. van Dissel-Emiliani, F. M. F., de Rooij, D. G. & Meistrich, M. L. (1989) J.
Reprod. Fertil. 86, 759–766.
21. Clermont, Y. & Perey, B. (1957) Am. J. Anat. 100, 241–260.
22. Beaumont, H. M. & Mandl, A. M. (1963) J. Embryol. Exp. Morph. 11, 715–740.
23. Roosen-Runge, E. C. & Leik, J. (1968) Am. J. Anat. 122, 275–299.
24. McGuinness, M. P. & Orth, J. M. (1992) Eur. J. Cell Biol. 59, 196–210.
25. McGuinness, M. P. & Orth, J. M. (1992) Anat. Rec. 233, 527–537.
26. Sapsford, C. S. (1962) Aust. J. Zool. 10, 178–192.
27. Orth, J. M. (1993) in Cell and Molecular Biology of the Testis, eds. Desjardins,
C. & Ewing, L. L. (Oxford Univ. Press, New York), pp. 3–42.
28. Huckins, C. & Clermont, Y. (1968) Arch. Anat. Histol. Embryol. 51, 341–354.
29. Wang, R. A., Nakane, P. K. & Koji, T. (1998) Biol. Reprod. 58, 1250–1256.
30. Hasthorpe, S., Barbic, S., Farmer, P. J. & Hutson, J. M. (1999) J. Reprod. Fertil.
31. Hasthorpe, S., Barbic, S., Farmer, P. J. & Hutson, J. M. (2000) J. Reprod. Fertil.
32. Orth, J. M. & Boehm, R. (1990) Endocrinology 127, 2812–2820.
33. van Dissel-Emiliani, F. M. F., de Boer-Brouwer, M., Spek, E. R., van der Donk,
J. A. & de Rooij, D. G. (1993) Cell Tissue Res. 273, 141–147.
34. Orth, J. M., Qiu, J., Jester, W. F., Jr., & Pilder, S. (1997) Biol. Reprod. 57,
35. Rhim, J. A., Sandgren, E. P., Degen, J. L., Palmiter, R. D. & Brinster, R. L.
(1994) Science 263, 1149–1152.
36. Clouthier, D. E., Avarbock, M. R., Maika, S. D., Hammer, R. E. & Brinster,
R. L. (1996) Nature (London) 381, 418–421.
37. Nagano, M., Avarbock, M. R. & Brinster, R. L. (1999) Biol. Reprod. 60,
38. Russell, L. D. & Brinster, R. L. (1996) J. Androl. 17, 615–627.
39. Geissler, E. N., Ryan, M. A. & Housman, D. E. (1988) Cell 55, 185–192.
40. Ogawa, T., Are ´ghaga, J. M., Avarbock, M. R. & Brinster, R. L. (1997) Int. J.
Dev. Biol. 41, 111–122.
Dev. 53, 142–148.
42. van Engeland, M., Nieland, L. J., Ramaekers, F. C., Schutte, B. & Reuteling-
sperger, C. P. (1998) Cytometry 31, 1–9.
43. Bedner, E., Li, X., Gorczyca, W., Melamed, M. R. & Darzynkiewicz, Z. (1999)
Cytometry 35, 181–195.
44. Rotman, B. & Papermaster, B. W. (1966) Proc. Natl. Acad. Sci. USA 55,
45. Li, H., Papadopoulos, V., Vidic, B., Dym, M. & Culty, M. (1997) Endocrinology
46. Orwig, K. E., Avarbock, M. R. & Brinster, R. L. (2002) Biol. Reprod., in press.
47. Knudson, C. M., Tung, K. S., Tourtellotte, W. G., Brown, G. A. & Korsmeyer,
S. J. (1995) Science 270, 96–99.
48. Furuchi, T., Masuko, K., Nishimune, Y., Obinata, M. & Matsui, Y. (1996)
Development (Cambridge, U.K.) 122, 1703–1709.
49. Rucker, E. B., III, Dierisseau, P., Wagner, K. U., Garrett, L., Wynshaw-Boris,
A., Flaws, J. A. & Hennighausen, L. (2000) Mol. Endocrinol. 14, 1038–1052.
50. Boulogne, B., Olaso, R., Levacher, C., Durand, P. & Habert, R. (1999) Int. J.
Androl. 22, 356–365.
Orwig et al.
September 3, 2002 ?
vol. 99 ?
no. 18 ?