THE ANATOMICAL RECORD 250:488–492 (1998)
Simultaneous Observation of Capillary Nets and Tenascin
in Intestinal Villi
HISASHI HASHIMOTO,1,2* HIROSHI ISHIKAWA,1 AND MORIAKI KUSAKABE2
1Department of Anatomy, Jikei University School of Medicine, Tokyo, Japan
2Division of Experimental Animal Research, Institute of Physical and Chemical Research
(RIKEN), Ibaraki, Japan
ABSTRACT
Background: In order to reveal the biological role of
capillaries in a tissue, it is desirable to study the three-dimensional
distribution of capillary nets in relation to tissue architecture. However,
the simultaneous observation of the three-dimensional distribution of
these nets and other substances has been rarely performed to date. In the
present study, we have developed a novel method for investigating the
three-dimensional distribution of capillary nets simultaneously with that
of extracellular matrix components, such as tenascin, using the confocal
laser scanning microscope.
Methods: Adult male mice were perfused with TRITC-labelled gelatin.
After perfusion, the intestine was irrigated with chilled fixative containing paraformaldehyde and picric acid, dissected, and returned to the same
fixative. The intestine was further sectioned and indirectly immunostained for tenascin using an FITC-labelled antibody.
Results: The three-dimensional distribution of capillary nets and tenascin in villi was simultaneously observed on stereo pairs of pseudo-colored
and superimposed images. Tenascin was distributed at the basement
membrane zone and in the underlying connective tissue but absent in
some regions where capillary nets were running just beneath the epithelium. Substances other than tenascin also can be examined in correlation
with capillary nets.
Conclusions: This method will be useful for investigating the biological
role of capillary nets. Anat. Rec. 250:488–492, 1998. r 1998 Wiley-Liss, Inc.
Key words: confocal laser scanning microscopy; capillary; gelatin; tenascin; three-dimensional reconstruction
Recently, we have reported three-dimensional distribution of tenascin in mouse small intestinal villi in
detail (Hashimoto and Kusakabe, 1997). Briefly, tenascin distributed at the epithelial basement membrane
zone (eBMZ) as well as in the underlying connective
tissue of a villus, but was absent around the tip of a
villus. The absence of tenascin staining in some regions
led to a striped pattern at the eBMZ. In the connective
tissue just beneath the eBMZ, tenascin appeared to be
pushed aside by the capillary. Around the tip of a villus,
tenascin disappeared from the eBMZ and connective
tissue, but it still remained and became prominent
around capillaries. However, the distribution of the
capillary was not directly declared in that study.
Capillary nets in an organ have been investigated by
various methods and, in general, the methods employed
may be divided into two categories. One is to fill
capillaries with a variety of media. This includes observations by dissection or by serial sections and subsequent examination of wholemounts or serial sections
after intravascular injection of materials such as India
ink. In 1955, Batson utilized an acrylic monomer as
r 1998 WILEY-LISS, INC.
injection materials to produce vascular corrosion casts.
Subsequently, plastic resins, such as prepolymerized
methylmethacrylate and Mercox, were introduced as
an injection medium and corrosion casts were observed
macroscopically and also microscopically with either a
dissection microscope or a scanning electron microscope (Murakami, 1971; reviewed by Hodde and Nowell, 1980). The observations of injection replicas by
scanning electron microscopy enable one to analyze
capillary nets at high resolution. Most recent studies on
the microcirculation in a variety of organs have been
performed using this method. However, this procedure
Grant sponsor: Grant-in-aid from the Ministry of Education, Science, Sports and Culture of Japan; Grant numbers: 08670038,
08670825, and 09670032; Grant sponsor: The Special Coordination
Funds of the Science and Technology Agency of the Japanese Government; Grant sponsor: Foundation for Advancement of International
Science.
*Correspondence to: H. Hashimoto, Department of Anatomy, Jikei
University School of Medicine, Nishishinbashi 3-25-8, Minatoku,
Tokyo 105, Japan.
Received 30 September 1997; Accepted 23 December 1997
489
FLUORO-GELATIN ANGIOGRAPHY
requires the injection of methacrylate resin into the
vasculature, polymerization of the resin, and removal
of soft tissues to expose the corrosion cast of the
capillaries of interest. As an alternative to this method,
FITC-dextran has been employed in ophthalmological
research and blood vessels in wholemounts of retinas
have been observed with an epi-illuminated fluorescence microscope (D’Amato et al., 1993).
The other method for investigating capillary nets is
to detect endothelial cells or their basement membranes. Antibodies against vascular endothelium specific antigens were raised and applied to observe endothelial cells (Connolly et al., 1988). Some lectins, such
as Ulex europaeus agglutinin I (UAEI) and Ricinus
communis agglutinin I (RCA), are known to react
specifically with the vascular endothelium and have
been utilized to detect endothelial cells (Connolly et al.,
1988; D’Amato et al., 1993; Rummelt et al., 1994).
UAEI lectin has been reported to be a reliable marker of
the human vascular endothelium (Holthöfer et al.,
1982), but some species and strain specificity of the
UAEI may be present. Immunostaining of laminin or
type IV collagen shows the basement membrane of the
vascular endothelium, thus demonstrating vascular
networks (Connolly et al., 1988). However, fetal or
developing vasculatures may have incomplete basement membranes resulting in the failure of labelling.
Since these molecules should exist in other, more
developed basement membranes, it may be difficult to
identify which is the vascular and which is the epithelial basement membrane in some organs.
In the present study, we developed a novel method for
investigating the three-dimensional distribution of capillary nets using the confocal laser scanning microscope
(CLSM) and attempted to reveal the interrelationship
between the distribution of tenascin and that of capillary nets. For tracing the capillary network, a fluorochrome-labelled gelatin was employed as the injection
material. Gelatin is easily labelled with isothiocyanate
derivatives of fluorochrome because of the abundance of
amide residues. Fluorescence microscopy and CLSM
can easily detect fluorochrome-labelled gelatin without
the removal of all of the soft tissues. Moreover, double
staining with another fluorochrome having a different
emission spectrum is applicable when using fluorochrome-labelled, gelatin-injected specimens. Consequently, an interrelationship between capillary nets
with tenascin can be three-dimensionally revealed. (A
portion of this work was previously reported by Hashimoto et al., 1995.)
MATERIALS AND METHODS
Preparation of Tetramethylrhodamine
Iisothiocyanate-labelled Gelatin
Gelatin from bovine skin (approx. 225 Bloom, No. G
9382, Sigma Co., St. Louis, MO) was swollen in distilled
water and completely dissolved in a hot water bath. The
final concentration of gelatin was adjusted to 20% and
the pH of the solution raised to 11 with a 1N NaOH
solution. Tetramethylrhodamine isothiocyanate (TRITC)
was dissolved in absolute dimethylsulfoxide (DMSO) at
a concentration of 20 mg/ml. This was then gently
poured into the gelatin solution to make the final
weight ratio of TRITC to gelatin 1:400 and allowed to
react in the dark at 37oC overnight under mild agitation. The reacted mixture was then dialyzed in the dark
at 37oC against 0.01 M sodium phosphate-buffered
saline (PBS, pH 7.2) containing 0.01% NaN3. The
dialyzing buffer was changed every day and dialysis
continued for 7-10 days until no free TRITC was found
in the dialyzing buffer. The TRITC-labelled gelatin
solution was subsequently solidified and stored in the
dark at 4oC.
Perfusion With TRITC-labelled Gelatin
The ICR strain of mice were obtained from Japan
Clea (Tokyo, Japan), maintained in our laboratory
(Dept. of Anatomy, Jikei University School of Medicine)
as a closed colony and housed in a temperature (22oC)
and light (14 hr light a day) controlled room with free
access to tap water and diet (CE-2, Japan Clea).
Adult male mice weighing 28-30g were used in this
study. Prior to perfusion, the TRITC-labelled gelatin
solution was dissolved in a hot (60oC) water bath,
diluted with PBS to make a final concentration to 10%,
and was loaded into a 10-ml syringe connected with a
25G needle via a plastic tube. The syringe with plastic
tube and needle was kept at 37oC in a warm water bath
until use. Under sodium pentobarbital anesthesia, a
thoracotomy was performed and the TRITC-labelled
gelatin was gently perfused into the left ventricle and
blood drained from the right atrium. The perfusion
continued until blood no longer flowed from the right
atrium. It took 7 ml of the TRITC-labelled gelatin
solution for a mouse. Thereafter, the intestine was
irrigated with chilled fixative consisting of 0.5% paraformaldehyde and 15% (v/v) of a saturated picric acid
solution in 0.1 M sodium phosphate buffer (pH 7.0),
removed and further fixed in the same fixative overnight at 4oC.
Immunostaining Methods
The intestine was then rinsed several times with
chilled 0.1 M sodium phosphate buffer (pH 7.4) and the
proximal region was obtained as the specimen. The
specimen was then transferred to PBS containing 5%
sucrose, PBS containing 10% sucrose, and finally PBS
containing 20% sucrose and 10% glycerin, 3 hr each, at
4oC. The specimen was sectioned at 150 µm by a
microtome equipped with a cold stage. The sections
were rinsed with chilled PBS and pretreated with 3%
aqueous solution of sodium deoxycholate for 4 hr at 4oC.
After rinsing with chilled distilled water and chilled
PBS, they were incubated with 10% normal goat serum
for 1 hr or more at 4oC, and then incubated with
antihuman tenascin rat monoclonal antibody (Hashimoto and Kusakabe, 1997) at a concentration of 10
µg/ml for 1 day at 4oC. After several rinses with chilled
PBS for 1 hr each, the sections were incubated with
FITC-labelled goat anti-rat IgG antibody (Cappel
#56408, Organon Teknika Co., NC) diluted 1:100 with
PBS containing 1% bovine serum albumin and 0.1%
NaN3 for 1 day at 4oC, subsequently rinsed with chilled
PBS several times for 1 hr each, postfixed with 4%
paraformaldehyde in 0.1 M sodium phosphate buffer
(pH 7.4) for 30 min at 4oC, and rinsed with PBS. Finally,
they were transferred to 0.05M Tris-HCl-buffered saline (TBS, pH 8.0) containing 50% (v/v) nonfluorescent
490
H.HASHIMOTO ET AL.
glycerin and stored at -80oC till observation. For the
tenascin control, the primary antibody was replaced
with a human tenascin specific rat monoclonal antibody
that showed no cross-reactivity to mouse tenascin was
used at a concentration of 10 µg/ml to verify the
tenascin staining.
Observations
The sections were mounted with 0.05M TBS containing 90% (v/v) nonfluorescent glycerin and 10 mg/ml of
1,4-Diazabicyclo-[2.2.2.] octane (DABCO, Wako Pure
Chemical Co., Osaka, Japan), observed with Carl Zeiss
LSM-410 CLSM equipped with 10X, 20X, and 40X
Plan-NEOFLUAR objectives (n.a. 0.3, 0.5, and 0.75,
respectively) and stimulated with both argon laser at
488 nm and He-Ne laser at 543 nm. The emitted
fluorescence was divided with a dichroic mirror at 560
nm. The fluorescence of FITC was obtained through a
band pass filter of 510-525 nm and that of TRITC was
observed through a long pass filter at 590 nm. Serial
images of the optical sections were obtained at the
interval of 1.5-5 µm, depending on the objective lens
used. Three-dimensional reconstruction was performed
with an optional software for LSM-410 prepared by
Carl Zeiss Co.
RESULTS
In the control specimen for tenascin, no fluorescence
was observed except capillary nets filled with TRITClabelled gelatin. In experimental specimens, capillary
nets in villi were filled with the TRITC-labelled gelatin
and the fluorescence was clearly detected throughout
the thickness of the villi with very low background (Fig.
1a). In addition, the three-dimensional distribution of
tenascin was also clearly indicated without any background fluorescence (Fig. 1b); the distribution of tenascin corresponded well with the previous report (Hashimoto and Kusakabe, 1997). The tenascin was distributed
in the eBMZ of villi as well as in the underlying
connective tissue. In some regions of the eBMZ and in
the underlying connective tissue, tenascin was absent
resulting in a striped distribution (Fig. 1b). Around the
tip of villi, tenascin disappeared from the eBMZ, but
remained prominent around tubular structures which
would be capillaries (Fig. 1b). A superimposed image of
capillary nets and tenascin revealed that the capillaries
were distributed in regions where tenascin was absent
in the eBMZ and connective tissue and that prominent
tenascin surrounded the capillaries around the tip of
villi (Fig. 1c).
DISCUSSION
In this study, the three-dimensional distribution of
capillary nets in intestinal villi was revealed following
the injection of TRITC-labelled gelatin. Moreover, the
three-dimensional interrelationship between capillary
nets and tenascin distribution was clearly indicated
using the same specimen. A striped pattern of tenascin
distribution was generated by tenascin in the eBMZ
and underlying connective tissue was pushed aside by
the capillary nets.
In order to reveal the biological role of capillaries in a
tissue, it is desirable to study the three-dimensional
distribution of capillary nets in relation to tissue archi-
tecture. The functional role of capillary nets in a tissue
could be revealed by examining their distribution in
relation to other physiological substances or to specific
cells, such as tumor cells. The simultaneous observation of the three-dimensional distribution of these nets
and other substances has rarely been performed to
date. Wang and Wei (1976), and Wang and Ying (1977)
have attempted partial digestion of a tissue and demonstrated that elastic fibers and collagen bundles interlaced with the capillary network in hamster and rabbit
lungs. Rummelt et al. (1994) have utilized UAEI lectin
and PCNA antibodies to simultaneously demonstrate
tumor blood vessels and proliferating tumor cells in
uveal melanomas.
In our study, we have selected gelatin to make a
vascular replica of intestinal tissue. Gelatin has the
following favorable characteristics: (1) A warm aqueous
gelatin solution has comparatively low viscosity and
never hardens. It is gelled not by a polymerization
reaction, but only by cooling. Therefore, no time limit is
set for injection as long as the tissues are kept at a
suitable temperature. This is particularly valuable for
handling embryonic vasculature. (2) A gelatin molecule
has many residues which covalently bind with fluorochromes and is easily cross-linked with fixatives. Once
the gel is fixed, the solidified gelatin rarely dissolves
even by warming.
This fluorochrome-labelled gelatin method is applicable to most tissues including embryonic and fetal
(Hashimoto et al., 1995) because it is easy to handle and
has no time limit to inject and, moreover, cardiac
contractions are maintained during perfusion of an
embryo and fetus via the umbilical vein. Little is known
about the vascular system during fetal development
because of the difficulties to trace blood vessels. Dollinger and Armstrong (1974) tried to investigate the
chick embryo circulatory system with resin injection
replicas, whereas Kondo et al. (1993) applied the corrosion cast method to observe the vasculature of the
mouse embryo. Recently, a barium sulfate solution was
introduced to visualize the embryonic vasculature using light and scanning electron microscopy (Kondo,
1996). By this method, surface structures were observed with the secondary electron image and blood
vessels by the backscattered electron image. Considering the technical difficulties and the possibilities for
double staining, our novel method will be a more
suitable alternative.
We have combined the fluorochrome-labelled gelatin
method with our novel preparation method for CLSM
(Hashimoto and Kusakabe, 1997). Without the removal
of soft tissues, a fluorochrome-labelled gelatin injected
specimen can be observed with CLSM, in addition to
other desired treatments, such as immunofluorescence,
without encountering any significant problems. By this
method, we noted the three-dimensional distribution of
both capillary nets and tenascin using CLSM and in the
future, this procedure will be useful for investigating
the biological roles of capillary nets in organs, especially during development.
ACKNOWLEDGMENTS
We thank Dr. D. C. Herbert, University of Texas
Health Science Center at San Antonio, for his help
during the preparation of this manuscript. We also
FLUORO-GELATIN ANGIOGRAPHY
Fig. 1. Stereo pair images of capillary nets and tenascin in a villus.
TRITC-labelled gelatin was injected into blood vessels and immunofluorescence staining of tenascin was performed with an FITC-labelled
antibody. Fluorescence of TRITC and FITC was simultaneously obtained with a 40X Plan-NEOFLUAR objective lens (n.a. 0.75). Threedimensional stereo pair images were reconstructed from 50 serial
images taken at an interval of 1.5 µm with an optional software for
LSM-410 prepared by Carl Zeiss Co. Scale bar550 µm.(a) Stereo pair
images of fluorescence of TRITC indicating capillary nets in a villus.
Capillary nets are filled with TRITC-labelled gelatin and their threedimensional distribution is clearly indicated. (b) Stereo pair images of
fluorescence of FITC indicating the distribution of tenascin. Threedimensional distribution of tenascin is well illustrated. In some
regions (white arrows), tenascin is lacking resulting in a striped
distribution. Around the tip of a villus, tenascin disappears from the
eBMZ and the underlying connective tissue (red arrows), whereas it is
491
prominent around tubular structures that are presumably capillaries
(white and green arrowheads). Tenascin intensely surrounds tubular
structures running parallel to the optical axis (green arrowheads). On
a wall of the tubular structures running perpendicular to the optical
axis, however, the fluorescence of tenascin is far more conspicuous at
the vertically oriented portion of the wall than at the horizontal
portion of it (white arrowheads), since the apparent fluorescence
intensity of structures observed with CLSM varies depending on its
orientation with respect to the optical axis (Van Der Voort and
Brakenhoff, 1990). (c) Stereo pair images obtained by superimposing
red-colored TRITC images (capillary nets) and green-colored FITC
images (tenascin). The capillary nets are present where tenascin is
lacking in the eBMZ and connective tissue. Around the tip of a villus,
the tubular structures indicated with arrowheads in (b) are proved to
be capillaries and prominent tenascin surrounds the capillaries.
492
H.HASHIMOTO ET AL.
thank Mr. Fumiyoshi Ishidate (Carl Zeiss Co.) for his
technical advice and Carl Zeiss Co. for providing us
with opportunities to utilize LSM-410 confocal laser
scanning microscope.
Literature Cited
Batson, V. 1955 Corrosion specimens prepared with a new material.
Anat. Rec., 121:425.
Connolly, S.E., T.A. Hores, L.E.H. Smith, and P.A. D’Amore 1988
Characterization of vascular development in the mouse retina.
Microvasc. Res., 36:275–290.
D’Amato, R., E. Wesolowski, and L.E.H. Smith 1993 Microscopic
visualization of the retina by angiography with high-molecularweight fluorescein-labeled dextrans in the mouse. Microvasc.
Res., 46:135–142.
Dollinger, R.K., and P.B. Armstrong 1974 Scanning electron microscopy of injection replicas of the chick embryo circulatory system. J.
Microsc., 102:179–186.
Hashimoto, H., H. Ishikawa, and M. Kusakabe 1995 Three-dimensional observation of capillary nets by confocal laser scanning
microscopy. J. Electron Microsc., 44:233 abstract.
Hashimoto, H., and M. Kusakabe 1997 Three-dimensional distribution of extracellular matrix in the mouse small intestinal villi.
-Laminin and tenascin- Connect. Tiss. Res., 36:63–71.
Hodde, K. C., and J.A. Nowell 1980 SEM of micro-corrosion casts.
SEM/1980/part II, SEM Inc., AMF O’Hare, Chicago, IL 60666,
89–106.
Holthöfer, H., I. Virtanen, A.-L. Kariniemi, M. Hormia, E. Linder, and
A. Miettinen 1982 Ulex europaeus I lectin as a marker for
vascular endothelium in human tissues. Lab. Invest., 47:60–66.
Kondo, S. 1996 Microinfusion method using barium sulfate for visualization of embryonic blood vessels with light and scanning electron microscopy. J. Electron Microsc., 45:163–167.
Kondo, S., R. Suzuki, K. Yamazaki, and K. Aihara 1993 Application of
corrosion cast method for scanning electron microscopic observation of mouse embryo vasculature. J. Electron Microsc., 42:14–23.
Murakami, T. 1971 Application of the SEM to the study of the fine
distribution of the blood vessels. Arch. Histol. Jap., 33:179–198.
Rummelt, V., L.M.G. Gardner, R. Folberg, S. Beck, B. Knosp, T.O.
Moninger, and K.C. Moore 1994 Three-dimensional relationships
between tumor cells and microcirculation with double cyanine
immunolabeling, laser scanning confocal microscopy, and computer-assisted reconstruction: An alternative to cast corrosion
preparations. J. Histochem. Cytochem., 42:681–686.
Van Der Voort, H. T. M., and G.J. Brakenhoff 1990 3-D image
formation in high-aperture fluorescence confocal microscopy: A
numerical analysis. J. Microsc., 158:43–54.
Wang, N. S., F. Wei 1976 Scanning electron microscopy of capillary and
elastic fiber networks in alveoli. Scan. Electron Microsc., 5:
233–238.
Wang, N.S., and W.L. Ying 1977 A scanning electron microscopic study
of alkali-digested human and rabbit alveoli. Am. Rev. Respir. Dis.,
115:449–460.