Constitutive expression of cyclooxygenase-2
in rat vas deferens
JAMES A. MCKANNA,1 MING-ZHI ZHANG,1 JUN-LING WANG,2
H.-F. CHENG,2 AND RAYMOND C. HARRIS2
Departments of 1Cell Biology and 2Medicine, George M. O’Brien Center for Kidney and Urologic
Diseases, Vanderbilt University School of Medicine, Nashville, Tennessee 37232
male; reproduction; erection; prostaglandins
PROSTAGLANDINS WERE discovered in human semen as
acidic lipoid factors that modify the contractility of
uterine smooth muscle (5). The name has persisted
despite consensus that prostaglandins in semen are not
of prostatic origin and demonstration of their important roles in many other organ systems. Eliasson (4)
analyzed the semen and various accessory male glands
of man, stallion, ox, bull, boar, ram, goat, rabbit, and
rat; high concentrations in ram vesicular glands led to
the generalization that seminal vesicles are the primary sources of male prostanoids, but extensive studies of rodents also implicated the vas deferens (7).
Despite more than 66,000 publications on prostaglandins since 1965, the roles of prostanoids in male
reproductive functions remain largely unknown (11)
and controversies regarding their sites of synthesis
have not been resolved (13).
The initial step in production of prostanoids is a
two-step conversion of arachidonic acid first to prostaglandin G2 and then to prostaglandin H2 by the ratelimiting enzyme prostaglandin G2/H2 synthase, also
known as cyclooxygenase (18). Subsequent metabolism
to different prostanoid species (e.g., PGE, PGD, etc.) is
accomplished by specific synthases that define the
distinct prostanoid profiles of various tissues.
Two cyclooxygenase genes are now recognized. COX-1
encodes a 2.7- to 2.9-kb transcript found in many
mammalian tissues and generally is referred to as
‘‘constitutive.’’ COX-2 encodes a 4.0- to 4.4-kb transcript originally detected in Swiss 3T3 cells treated
with phorbol ester (12); COX-2 generally is called the
‘‘inducible’’ form. Although the translation products of
both COX-1 and COX-2 are of similar size (,73 kDa)
and possess similar cyclooxygenase activities, they
share only ,66% homology in amino acid sequence
(19).
The literature has emphasized that COX-2 expression is low in normal adult animals and is increased in
response to growth factors or inflammatory stimuli (15,
17, 24). However, after performing extensive tests to
validate COX-2 immunolocalization, our laboratory
generated evidence that sustained COX-2 expression
was observed in the kidney cortex of normal adult rats
(9). The sustained COX-2 expression in specific epithelial cell populations could be upregulated by physiological variables, including low-salt diet and adrenalectomy, and in normal development (27; McKanna and
Harris, unpublished).
In light of our evidence that COX-2 expression at
certain loci was not transient and that it was sensitive
to steroids, we were fascinated by a report that COX-2
expression is normally high in human prostate (20). We
postulated that localization of COX-2 in male accessory
glands would be of interest, but because reliable immunolocalization depended on adequate fixation, an experimental animal was required. With the use of immunohistochemistry supplemented by Western and Northern
analysis, the present studies have determined that the
predominant cyclooxygenase species in male rat reproductive organs is COX-2 and that the primary locus of
COX-2 expression is the epithelium of the distal vas
deferens. Interestingly, the intense COX-2 activity is
juxtaposed to a submucosal vascular plexus that communicates with the penile corpora cavernosa. Thus
prostaglandins from the vas deferens may be instrumental in penile erection.
METHODS
Specimens. Male rats of Sprague-Dawley and Long-Evans
strains and hybrids of these strains were studied. No differences in results were noted among the various genetic
backgrounds. Surgical procedures were performed under
sterile conditions using Nembutal (50 mg/kg) anesthesia.
Northern analysis. Total RNA was isolated from vas deferens homogenates by the acid guanidinium thiocyanate-phenol-
0363-6119/98 $5.00 Copyright r 1998 the American Physiological Society
R227
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
McKanna, James A., Ming-Zhi Zhang, Jun-Ling Wang,
H.-F. Cheng, and Raymond C. Harris. Constitutive expression of cyclooxygenase-2 in rat vas deferens. Am. J. Physiol.
275 (Regulatory Integrative Comp. Physiol. 44): R227–R233,
1998.—Prostaglandins, lipoid substances discovered in human semen as modulators of uterine muscle contractility, are
known to play significant roles in virtually all mammalian
organ systems, but their male reproductive functions are
unclear. Cyclooxygenase, the rate-limiting enzyme in prostaglandin synthesis, occurs in two isoforms distinguished on
the basis of constitutive (COX-1) or inducible (COX-2) expression patterns in mammalian tissues. However, in the adult
rat male reproductive system, immunohistochemistry and
Western and Northern analysis showed that COX-2 is the
predominant isoform and is heavily localized to the epithelium of the distal vas deferens, where constitutive expression
is manyfold greater than in any other organs of the body.
COX-2 is not detected in the proximal one-half of the vas nor
in the testis, epididymis, seminal vesicles, or prostate. Elimination of luminal sperm by vasectomy does not affect COX-2
levels, whereas castration severely depletes COX-2 and androgen replacement after castration restores COX-2, indicating
that COX-2 expression in the vas is androgen dependent.
Because the distal vas also comprises an extensive submucosal venous plexus connected to the penile corpora cavernosa,
prostaglandins from the vas may play a role in erection.
R228
COX-2 IN RAT VAS DEFERENS
diaminobenzidine as chromogen, followed by a light counterstain with toluidine blue.
Micrography. Bright-field images from a Leitz Orthoplan
microscope with Optronics DEI750 three-chip RGB video
camera were digitized by the BIOQUANT TrueColorWin95
system (R&M Biometrics, Nashville, TN) and saved as computer files. Contrast and color level adjustments (Adobe
Photoshop) were performed for the entire image, i.e., no
region- or object-specific editing or enhancements were performed.
RESULTS
Control specimens. After fixation in situ, representative regions of the reproductive organs were combined
in single blocks and sectioned. The inclusion of multiple
specimens in a single section ensures that they receive
identical treatment in the immunohistological procedures. At low magnification (Fig. 1A), the dark brown
diaminobenzidine (DAB) reaction product identifying
COX-2-ir was absent from testis, epididymis, and seminal vesicles but was observed in some sections of the
vas deferens and at highly restricted sites in the
prostate. At higher magnification of the prostate (Fig.
1B), it was apparent that the COX-2-ir was not actually
situated in prostatic glandular elements but rather was
restricted to ducts penetrating the gland. Based on
their fluted epithelial lumen and muscular tunics,
these ducts were identified as the prostatic segments of
the vasa deferentia; the prostatic tissue itself was void
of COX-2-ir.
The vasa deferentia, bilateral tubes that convey
sperm from the cauda epididymis at each testis to the
ejaculatory duct buried in the prostate gland, are 5–6
cm long in the adult rat. Initial examinations of COX2-ir in sections of the vas deferens were puzzling
because some showed intense epithelial staining,
whereas others showed none (Fig. 1A). To assess the
distribution of COX-2-ir along the vas, the ducts were
Fig. 1. Histological sections of adult male rat tissues perfusion fixed with glutaraldehyde-periodate-acid-saline
(GPAS) and immunostained for cyclooxygenase-2 immunoreactivity (COX-2-ir; except E and H stained for lipocortin
1). A: low magnification survey of male reproductive organs. Intense brown reaction product indicating COX-2-ir is
observed in ducts in the prostate (p) and in 4 sections of the distal vas deferens (dv). It is absent from the testis (t),
epididymis (e), seminal vesicles (s), and from 4 segments of the proximal vas deferens (pv) (33.7). B: duct entering
the prostate is surrounded by venous plexus (v) and smooth muscle (sm); it exhibits strong COX-2-ir at the surface
surrounding the lumen (p). The rest of its wall and the adjacent glandular tissues (g) are COX-2 negative (330). C-E:
proximal vas deferens at 330 (C) appears to have a 250-µm diameter lumen (p) surrounded by a muscular wall (sm)
of similar thickness. At 3120 (D), the mucosal epithelium and lamina propria are visible, but no COX-2-ir is
apparent. In a similar section at 3480 (E) and stained for lipocortin 1 to reveal the basal cells (arrowheads), both the
epithelium and lamina propria are ,20 µm thick. F-H: distal vas deferens at 330 (F) is more than twice the
diameter of the proximal and exhibits intense COX-2-ir surrounding the lumen (p). At 3120 (G), the lamina propria
(lam p) and venous plexus are apparent, as well as the uniform intensely stained epithelium. In a similar section at
3480 (H) with basal cells (arrowheads) stained for lipocortin 1, the increased height of the epithelium appears due
to amplification of the lacy infranuclear compartment (inc). I: after 4 wk vasectomy, the distal vas wall remains
robust and the epithelium continues to display intense COX-2-ir (3120). J: 1 wk after orchiectomy, COX-2-ir is
diminished in a few dispersed cells of the distal vas epithelium (arrow); cells on the crest of the rugae (arrowhead)
also are COX-2 negative (3120). K: 2 wk after orchiectomy, the diameter of the distal vas has decreased; its lumen
(p) is smaller and many COX-2-negative epithelial cells are interspersed with those displaying stronger COX-2-ir
(arrows) (3120). L: 4 wk after orchiectomy the diameter of the vas is decreased by ,75% (compare this micrograph
at 3120 with F above at 330). Muscularis (sm) is diminutive, the lamina propria and venous plexus appear
disproportionately large, and the epithelium resembles the proximal vas (G above) both in stature and in the
absence of substantial COX-2-ir. M and N: androgen replacement rat orchiectomized for 4 wk (as L above) received
testosterone injections for the last 2 wk before death. Distal vas at 330 (M) illustrates slightly hypertrophied
muscularis (sm) and intense COX-2-ir in the epithelium surrounding the lumen (p). At 3400 (N), the epithelium
exhibits robust COX-2-ir similar to controls (G above) with unstained blue nuclei and slightly more intense staining
in the infranuclear cytoplasm (inc). The basal cells (arrowheads) are COX-2 negative.
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
chloroform method (1). RNA samples were electrophoresed in
denaturing agarose gels and transferred to nitrocellulose.
Nitrocellulose blots were hybridized as previously described
with a 1.3-kb 32P-labeled cDNA Kpn 1/Xho 1 fragment of the
38 untranslated region of rat COX-2 (3).
Immunoblotting. Weighed samples of vas deferens (proximal and distal segments), testis, epididymis, prostate, and
seminal vesicle were homogenized in 30 mM Tris · HCl, pH
8.0, 100 µM phenylmethylsulfonyl fluoride (1:9 wt/vol). After
a 10-min centrifugation at 10,000 g, the supernatant was
centrifuged for 60 min at 110,000 g to prepare microsomes as
described previously (9). The microsomes were resuspended
in SDS-sample buffer heated to 100°C for 5 min, and the
proteins were separated on 8% SDS gels under reducing
conditions and transferred to Immobilon-P transfer membranes (Millipore, Bedford, MA). The blots were blocked
overnight with 100 mM NaCl-50 mM Tris · Cl, pH 7.4, containing 5% nonfat dry milk, 3% albumin, and 0.5% Tween 20,
followed by incubation for 16 h with rabbit polyclonal antiserum raised against a murine COX-2 peptide (Cayman
Chemical, Ann Arbor, MI) at 2.5 µg/ml dilution. The second
reagent, biotinylated goat anti-rabbit antibody, was detected
using avidin and biotinylated horseradish peroxidase (Pierce,
Rockford, IL) and exposed on film using enhanced chemoluminescence (Amersham).
Immunohistochemistry. Under deep anesthesia with Nembutal (70 mg/kg ip), rats were exsanguinated with ,25
ml/100 g heparinized saline (0.9% NaCl, 2 U/ml heparin,
0.02% sodium nitrite) through a transcardial aortic cannula
and fixed with GPAS (glutaraldehyde-periodate-acid-saline)
as previously described (16). GPAS contains final concentrations of 2.5% glutaraldehyde, 0.01 M sodium meta-periodate,
0.04 M sodium phosphate, 1% acetic acid, and 0.1 M NaCl; it
provides excellent preservation of tissue structure, COX-2
antigenicity, and mRNA. The fixed tissues were dissected and
dehydrated through a graded series of ethanols, embedded in
paraffin, sectioned at 4 µm thickness, and mounted on glass
slides. COX-2 immunoreactivity (COX-2-ir) was immunolocalized with polyclonal rabbit anti-murine COX-2 serum (Cayman) diluted to 2.5 µg/ml. The 1° antibodies were localized
using Vectastain ABC-Elite (Vector, Burlingame, CA) with
COX-2 IN RAT VAS DEFERENS
fixed by perfusion in situ, transected into 6–8 segments
of approximately equal length, and mounted in sequence to provide sections from identified regions.
These sections demonstrated that the vas comprised
two distinct halves with a very abrupt transition. The
proximal half emanating from the cauda epididymis
was smaller in caliber (0.5 mm OD and 0.2 mm ID) than
the distal half (1.3 mm OD and 0.3 mm ID). No
COX-2-ir was observed in any sections from the proximal half, but a thin band of intense staining lined the
lumen of all sections from the distal half.
R229
Histologically, the proximal half of the vas (Fig. 1,
C-E) comprised an adventitial layer (loose connective
tissue) surrounding a 150-µm-thick tunica muscularis
(layer of smooth muscle). The muscularis of the proximal segment was directly apposed to a thin mucosa
that included a 20-µm-thick layer of connective tissue,
the lamina propria, and the 20-µm-tall simple columnar epithelium. The extent of the epithelium was best
illustrated in sections immunostained for lipocortin 1, a
calcium-binding protein (16), that localized to myoepithelial cells at the base of the epithelium (Fig. 1E). The
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
R230
COX-2 IN RAT VAS DEFERENS
apparent in most epithelial cells of the distal vas (Fig.
1J), but reduced staining was apparent at the crests of
the rugae and in a few dispersed epithelial cells. After 2
wk, approximately one-half of the cells in the epithelium were unstained (Fig. 1K). By 4 wk, the caliber of
the distal vas had decreased significantly, primarily
due to loss of muscle (Fig. 1L). The venous plexus and
lamina propria of the mucosa were still substantial, but
the epithelium was seriously diminished. It contained
only a few cells displaying COX-2-ir, and its stature had
decreased to ,25 µm primarily due to the loss of the
infranuclear compartment.
To further test the hypothesis that COX-2 expression
is androgen dependent, a subgroup of the 4-wk orchiectomy rats (above) received androgen replacement
therapy during the last 2 wk of the experiment. After 14
daily injections of testosterone ethanate (10 mg/kg), the
muscularis of the distal vas appeared normal or even
slightly hypertrophied (Fig. 1M); the epithelium was
tall and robust with intense COX-2-ir (Fig. 1N). As in
controls, COX-2-ir in the epithelial cells was more
intense in the infranuclear compartment and staining
was not observed in the nuclei. If the initial 2 wk of
androgen deprivation depleted COX-2-ir in the vas
epithelium, as illustrated in Fig. 1K, then the subsequent 2 wk of androgen replacement restored the
COX-2-ir to normal levels.
Western analysis. Expression of COX-2 protein in
distal vasa deferentia from adult rats was examined
with immunoblots using anti-murine COX-2 serum
(shown with immunohistochemistry to have no detectable crossreactivity with COX-1). The immunoreactive
COX-2 in the microsome fraction from whole homogenates appeared as a distinctive band of molecular mass
,73 kDa (Fig. 2). The homogenates from distal vas
deferens were loaded at 15, 1.5, and 0.15 µg total
protein per lane to provide a three-decade scale of
COX-2-ir. For comparison, homogenates of testis, epididymis, proximal vas, prostate, and seminal vesicle
were loaded at 75 µg per lane; i.e., protein levels 5- to
500-fold greater than the vas lanes. No COX-2-ir was
detected for any of these organs except for the seminal
vesicle. Because the seminal vesicle band is intermediate in intensity between the 1.5 and 0.15 µg bands from
distal vas, we can conclude that COX-2 protein in the
Fig. 2. Total homogenates from male accessory organs probed with
anti-COX-2. In the first 3 lanes, microsomes from distal vas deferens
loaded at 10-fold dilutions (15, 1.5, 0.15 µg of protein per lane)
showed strong bands of COX-2-ir. In the adjacent 5 lanes, microsomes
from testis, epididymis, proximal vas deferens (prox vas), prostate,
and seminal vesicle were loaded at 75 µg/lane (i.e., 5–500 times
greater protein loading than the distal vas lanes). For these 5 organs,
COX-2-ir is observed only in the seminal vesicle.
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
nuclei of the columnar epithelial cells in the proximal
vas were situated in the basal half of the cell, displaced
by the extensive apical cytoplasm that contains Golgi
apparatus and endoplasmic reticulum as demonstrated
with electron microscopy (6).
The distal vas appeared amplified and elaborated
relative to the proximal (compare Fig. 1F with Fig. 1C).
Both the muscularis (400 µm) and mucosa (90–120 µm)
were more than twice as thick as proximal. Additionally, in well-perfused specimens, an extensive network
of vessels (20–40 µm diam) with thin walls and a
continuous endothelial lining separated the muscularis
from the lamina propria (Fig. 1G).
Both the epithelium and lamina propria of the distal
vas mucosa were considerably amplified relative to the
proximal (compare Fig. 1G with Fig. 1D). The lamina
propria was more than twice as thick (50–90 µm) and
extended into longitudinal epithelial rugae (folds) that
gave the lumen its characteristic fluted profile. When
viewed with lipocortin 1 immunolocalization, the simple
columnar epithelial cells of the distal vas are observed
to be much taller (40–50 µm) than proximal (compare
with Fig. 1E). These cells display an extensive lacytextured infranuclear compartment corresponding to
the hypertrophied smooth endoplasmic reticulum detected electron microscopically (6).
Control specimens of the distal vas exhibited COX2-ir throughout the cytoplasm of the columnar epithelial cells, with slightly greater intensity of staining in
the infranuclear cytoplasm than in the supranuclear.
COX-2-ir was uniformly absent from the nuclei of the
epithelial cells. Neither the myoepithelial cells nor any
other cells of the mucosa nor other layers of the vas wall
exhibited notable levels of COX-2-ir.
Vasectomy. To test the hypothesis that COX-2-ir
expression in the vas deferens might be ‘‘induced’’ by
low-grade inflammatory signals generated by sperm
debris being degraded in the lumen, sperm were eliminated by bilateral vasectomy. Each vas deferens was
ligated twice in the scrotum near its junction with the
epididymis and transected between the ligatures. Four
weeks after surgery, the epithelium in the entire vas
distal to the transection appeared healthy and the
lumen of the vas was clear of all traces of sperm. In the
distal vas, the strong COX-2-ir was indistinguishable
from control (Fig. 1I), indicating that COX-2 expression
in the vas deferens was not induced by degenerating
sperm.
Orchiectomy and androgen replacement. Because only
trace amounts of COX-2-ir were observed in the vas
during the first 4 postnatal wk (not illustrated), but
adult levels developed during puberty (weeks 5–7), we
postulated that COX-2 expression could be androgen
dependent. To test this hypothesis, androgens were
eliminated from normal adult male rats by surgical
removal of the testicles after ligature at the corpus
epididymis. The male accessory organs, examined histologically 1, 2, and 4 wk postorchiectomy, exhibited
dramatic atrophy of the prostate and seminal vesicles
and reduction in the caliber and length of the vasa.
After 1 wk of androgen deprivation, COX-2-ir was still
COX-2 IN RAT VAS DEFERENS
DISCUSSION
These studies have demonstrated that the predominant isoform of cyclooxygenase in the male reproductive system of rat is COX-2, and the primary locus of
COX-2 expression is the epithelium of the distal vas
deferens. The intense COX-2-ir begins to appear in the
vas deferens early in puberty and is androgen dependent; i.e., it is depleted after orchiectomy and is restored by testosterone replacement.
Serum and fixative controls. Because conclusions
from these studies depend on valid immunolocalization, our laboratory routinely examines two critical
factors: namely, preservation of tissue antigens and
specificity of antisera. Developing rat kidney and adult
vas deferens, organs shown by Northern blots to have
high levels of the 4.4-kb transcript of COX-2 mRNA,
Fig. 3. Western blots of COX-2-ir in homogenates of distal vas loaded
at 10 µg protein per lane. Control levels (A) are not diminished after
vasectomy for 2 wk (B) or 4 wk (C) but are progressively less after
castration for 2 wk (D) and 4 wk (E); COX-2 expression was restored
to control levels by 2 wk of androgen replacement in rats castrated for
2 wk (F) and 4 wk (G).
Fig. 4. Northern blots of COX-2 mRNA detect a strong signal in
distal (dist) vas deferens from control rats (A) but weak to nil signals
in proximal vas (B) or distal vas after 4 wk androgen deprivation (C).
GADPH mRNA (bottom) confirms equivalent loading in each lane.
were used as experimental models to test the effects of
glutaraldehyde, formaldehyde, and other fixatives on
COX-2 antigen preservation with pH, freezing, and
dehydration variables. Although the fixation sensitivity for COX-2-ir was not as great as described previously for lipocortin 1 (16), tissues fixed with acidified
glutaraldehyde exhibited optimal antigen preservation
and low background. In paraffin sections of these
tissues, the Cayman no. 160106 rabbit anti-murine
COX-2 polyclonal serum stained cells in the distal vas
deferens and renal cortical thick ascending limb, precisely the cells identified to contain COX-2 mRNA by in
situ hybridization (27). Furthermore, the Cayman serum did not produce a signal in cells and tissues that
express COX-1, e.g., epididymis or renal papillary
collecting ducts. These tests provide considerable confidence that the reported immunolocalization in the vas
deferens is specific for COX-2.
Although the results with other sera are beyond the
scope of this venue, it should be mentioned that sera
from Santa Cruz and Transduction Laboratories give
strong staining but are unable to distinguish between
COX-1 and COX-2 in tissues fixed with GPAS or
neutral formaldehyde. It will be prudent to evaluate
with skepticism the growing literature based on studies
using these and other sera. In that regard, preabsorption of anti-COX sera with homogenates of rat vas
deferens provides a tool for distinguishing immunostains of the two COX isoforms, because proximal vas
blocks neither, whereas distal blocks only COX-2 immunoreactivity.
Rat vas deferens. The major dichotomy in prostaglandin synthase activities along the length of the vas
deferens, as documented in the present study, complements several prior observations. In a pioneering study,
Flickinger (6) reported distinctive ultrastructural differences between principal cells of the proximal and distal
vas segments. The distal cells were tall with small
whorls of smooth endoplasmic reticulum (SER) in the
apical cytoplasm and enormous whorls of SER in the
amplified infranuclear cytoplasm. At the time it was
known that tubular arrays of SER membranes likely
were involved in steroid biosynthesis (2), but the import of the whorled configuration was debated. Our
data identifying these subcellular regions as sites of
sustained high level COX-2 expression suggest that
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
distal vas is expressed at levels 50- to 500-fold greater
than in the seminal vesicles. Careful histological examination of the seminal vesicles revealed that COX-2-ir
was absent from the glandular tissues; it appeared in a
sparse population of isolated epithelial cells in the
main ducts that arise embryologically as diverticula of
the vasa deferentia (not illustrated).
To investigate the androgen dependence of COX-2
expression, Western blots were used to detect COX-2
levels in homogenates from the distal vas deferens after
castration and testosterone replacement (Fig. 3). In
lanes loaded with equal protein, the density of the
COX-2 band was strong in control (Fig. 3A) and vasectomized (Fig. 3, B and C) vasa, was decreased after 2 wk
orchiectomy (Fig. 3D), and was virtually absent after 4
wk of androgen deprivation (Fig. 3E). Testosterone
replacement (Fig. 3, F and G) returned the immunoreactive COX-2 to control levels.
Northern analysis. mRNA isolated from the distal
vas deferens of normal adult rats hybridized with a
probe specific for a distinct untranslated region of rat
COX-2 mRNA (9) to produce a strong signal at 4.4 kb
(Fig. 4A). Despite equal loading of total mRNA indicated by equivalent GADPH mRNA (Fig. 4, bottom), no
signal was apparent in samples from either the proximal vas of normal adults (Fig. 4B) or the distal vas 4 wk
after orchiectomy (Fig. 4C).
R231
R232
COX-2 IN RAT VAS DEFERENS
response to chronic salt or volume depletion as well as
after adrenalectomy (McKanna and Harris, unpublished). The current findings of high levels of COX-2
expression in normal vas deferens is a second example
of localized COX-2 expression in the absence of inflammation or dysplasia.
COX-2 expression in the vas appears to be androgen
dependent. In the first 4 postnatal wk, COX-2-ir is
observed in a few isolated cells in the vas epithelium. A
surge of testosterone signals the start of puberty (postnatal day 30); coincidentally, the epithelium of the
distal vas begins to display intense COX-2-ir. After
orchiectomy, the vas regresses along with the prostate
and seminal vesicles, and COX-2-ir diminishes as shown
in Fig. 1K. It will be of interest in future studies to
determine if androgen-responsive elements play direct
roles in regulating sustained transcription of the COX-2
gene.
Perspectives
Although prostaglandins were discovered in human
semen and much has been learned concerning the
prostanoid profiles that characterize various mammals, their roles in male reproductive function have
remained obscure. Documentation of high levels of
COX-2 in the rat vas deferens epithelium identifies a
major site of prostanoid synthesis but also adds another level of complexity to the picture. COX-2 typically
has been implied to be ‘‘inducible’’ and associated with
inflammation; however, in the vas deferens, COX-2
expression is sustained through adult life in an androgen-dependent fashion. The present evidence regarding the localization of high levels of cyclooxygenase may
lead to further experimentation on the roles of prostaglandins in regulating neurogenic contractility in erection and ejaculation.
This work was performed in the George M. O’Brien Center for
Kidney and Urologic Diseases at Vanderbilt University and was
supported by National Institute of Diabetes and Digestive and
Kidney Diseases Grant DK-39261 and by funds from the Department
of Veterans Affairs.
Address for reprint requests: J. A. McKanna, Cell Biology, Vanderbilt Medical School, Nashville, TN 37232.
Received 17 November 1997; accepted in final form 17 March 1998.
REFERENCES
1. Chomczynski, P., and N. Sacchi. Single-step method of RNA
isolation by acid guanidinium thiocyanate-phenol-chloroform
extraction. Anal. Biochem. 162: 156–159, 1987.
2. Christenson, A. K., and D. W. Fawcett. The normal fine
structure of opossum testicular interstitial cells. J. Biophys.
Biochem. Cytol. 9: 653–670, 1961.
3. DuBois, R. N., M. Tsujii, P. Bishop, J. A. Awad, K. Makita,
and A. Lanahan. Cloning and characterization of a growth
factor-inducible cyclooxygenase gene from rat intestinal epithelial cells. Am. J. Physiol. 266 (Gastrointest. Liver Physiol. 29):
G822–G827, 1994.
4. Eliasson, R. Studies on prostaglandin occurrence, formation
and biological actions. Acta Physiol. Scand. Suppl. 158: 1–73,
1959.
5. Euler, U. S. von. On the specific vaso-dilating and plain muscle
stimulating substances from accessory genital glands in man and
certain animals (prostaglandin and vesiglandin). J. Physiol.
(Lond.) 88: 213–234, 1936.
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
whorled SER in the distal vas deferens may represent
an ultrastructural adaptation related to prostanoid
production.
Hamilton and Cooper (8) noted the abrupt transition
between the proximal and distal vas and described
histological distinctions in the distal segment, including taller epithelium and thicker muscularis and submucosal venous plexus. The functional significance of
these features has not been determined, but their strict
correlation with intense COX-2 expression may provide
some clues.
Prostaglandin levels in rat semen are low (4), suggesting that COX-2 products are not secreted into the
lumen but rather are released toward the basal surface
of the epithelium. If such prostaglandins from the
mucosa diffused all the way to the muscularis, they
possibly could modulate the contractility of the vas in
response to adrenergic stimuli, as has been demonstrated in vitro (10, 21, 22). Interestingly, in studies
with rat vas deferens, some investigators found prostaglandin stimulatory effects, whereas others reported
inhibition. The contradictory results possibly reflect
differing proportions of proximal versus distal vas
segments in these experiments.
However, it seems unlikely that prostanoids from the
epithelium would reach the muscularis. En route, the
prostanoid trajectory is interrupted by the submucosal
venous plexus. Although this plexus often draws little
attention because it collapses and appears diminutive
in specimens fixed by immersion, it is extensive after
fixation via vascular perfusion. Its location and structure suggest that it provides a broad avenue for prostaglandins to enter the circulatory system; but what
would be their site of action once in the blood?
The submucosal plexus in the rat distal vas deferens
was noted by Hamilton and Cooper (8), who reported
that the plexus communicates with the corpus cavernosum penis. A venous plexus has not been described in
primates; however, a preponderance of submucosal
arterioles and venules was noted in the ampulla of vas
deferens in adult macaque monkeys (23). Given that
the injection of prostaglandins into the corpora cavernosa currently is an effective remedy for human erectile
impotence (26), future studies may determine the relationship of prostaglandins produced by the epithelium
of the vas deferens to the normal mechanisms of penile
erection.
Regulation of COX-2 transcription. COX-2 was first
identified as an immediate early gene (TIS-10) activated by phorbol esters (24), and subsequent studies
indicated that it can be induced by numerous growth
factors and cytokines (12). Important roles have been
suggested for COX-2 in mediation of inflammation,
regulation of cell growth, prevention of apoptosis (14),
and tumorigenesis (25). COX-2 expression is low in
normal adult tissue but is increased in response to
injury and/or inflammation. An exception is our previous finding that COX-2 is expressed in normal rat
kidney in a subset of cells in the cortical thick ascending limb near the macula densa (9). The numbers of
COX-2 cells in the macula densa region increases in
COX-2 IN RAT VAS DEFERENS
17. Mitchell, J. A., P. Akarasereenont, C. Thiemermann, R. J.
Flower, and J. R. Vane. Selectivity of nonsteroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc. Natl. Acad. Sci. USA 90: 11693–11697, 1993.
18. Needleman, P., J. Turk, B. A. Jakschik, A. R. Morrison, and
J. B. Lefkowith. Arachidonic acid metabolism. Annu. Rev.
Biochem. 55: 69–102, 1986.
19. O’Banion, M. K., H. B. Sadowski, V. Winn, and D. A. Young.
A serum- and glucocorticoid-regulated 4-kilobase mRNA encodes
a cyclooxygenase-related protein. J. Biol. Chem. 266: 23261–
23267, 1991.
20. O’Neill, G. P., and A. W. Ford-Hutchinson. Expression of
mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human
tissues. FEBS Lett. 330: 156–160, 1993.
21. Patra, P. B., R. M. Wadsworth, D. W. Hay, and I. J. Zeitlin.
The effect of inhibitors of prostaglandin formation on contraction
of the rat, rabbit and human vas deferens. J. Auton. Pharmacol.
10: 55–63, 1990.
22. Radomirov, R., and K. Venkova. Responsiveness of rat vas
deferens and stomach smooth muscles after treatment with
indomethacin. Gen. Pharmacol. 17: 425–429, 1986.
23. Ramos, A. S., Jr. Morphologic variations along the length of the
monkey vas deferens. Arch. Androl. 3: 187–196, 1979.
24. Tippetts, M. T., B. C. Varnum, R. W. Lim, and H. R.
Herschman. Tumor promoter-inducible genes are differentially
expressed in the developing mouse. Mol. Cell. Biol. 8: 4570–4572,
1988.
25. Tsujii, M., S. Kawano, and R. N. DuBois. Cyclooxygenase-2
expression in human colon cancer cells increases metastatic
potential. Proc. Natl. Acad. Sci. USA 94: 3336–3340, 1997.
26. Wagner, G., and H. S. Kaplan. The New Injection Treatment for
Impotence: Medical and Psychological Aspects. New York: Brunner/Mazel, 1993.
27. Zhang, M.-Z., J.-L. Wang, H.-F. Cheng, R. C. Harris, and
J. A. McKanna. Cyclooxygenase-2 in rat nephron development.
Am. J. Physiol. 273 (Renal Physiol. 42): F994–F1002, 1997.
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.2 on June 18, 2017
6. Flickinger, C. J. Regional variations in endoplasmic reticulum
in the vas deferens of normal and vasectomized rats. Anat. Rec.
176: 205–223, 1973.
7. Gerozissis, K., and F. Dray. Selective and age-dependent
changes of prostagladin E-2 in the epididymis and vas deferens of
the rat. J. Reprod. Fertil. 50: 113–115, 1977.
8. Hamilton, D. W., and T. G. Cooper. Gross and histological
variations along the length of the rat vas deferens. Anat. Rec.
190: 795–809, 1978.
9. Harris, R. C., J. A. McKanna, Y. Akai, H. R. Jacobson, R. N.
DuBois, and M. D. Breyer. Cyclooxygenase-2 is associated with
the macula densa of rat kidney and increases with salt restriction. J. Clin. Invest. 94: 2504–2510, 1994.
10. Holmquist, F., H. Hedlund, and K. E. Andersson. Pre- and
postjunctional effects of some prostanoids in human isolated vas
deferens. Am. J. Physiol. 260 (Regulatory Integrative Comp.
Physiol. 29): R792–R797, 1991.
11. Kelly, R. W. Immunosuppressive mechanisms in semen: implications for contraception. Hum. Reprod. 10: 1686–1693, 1995.
12. Kujubu, D. A., B. S. Fletcher, B. C. Varnum, R. W. Lim, and
H. R. Herschman. TIS10, a phorbol ester tumor promoterinducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. J. Biol. Chem. 266:
12866–12872, 1991.
13. Litwin, J. A. Does diaminobenzidine demonstrate prostaglandin synthetase? A study on polyunsaturated fatty acid-induced
DAB oxidation in sheep vesicular glands and rabbit kidney
medulla. Histochemistry 53: 301–315, 1977.
14. Lu, X., W. Xie, D. Reed, W. S. Bradshaw, and D. L. Simmons.
Nonsteroidal antiinflammatory drugs cause apoptosis and induce cyclooxygenases in chicken embryo fibroblasts. Proc. Natl.
Acad. Sci. USA 92: 7961–7965, 1995.
15. Masferrer, J. L., K. Seibert, B. Zweifel, and P. Needleman.
Endogenous glucocorticoids regulate an inducible cyclooxygenase enzyme. Proc. Natl. Acad. Sci. USA 89: 3917–3921, 1992.
16. McKanna, J. A., and M. Z. Zhang. Immunohistochemical
localization of lipocortin 1 in rat brain is sensitive to pH, freezing
and dehydration. J. Histochem. Cytochem. 45: 527–538, 1997.
R233