Spatial orientation of microtubules in contractile fibroblasts in vivo.

Spatial Orientation of Microtubules in Contractile
Fibroblasts In Vivo
ROSS RUDOLPH AND MARILYN WOODWARD
Division of Plastic Surgery and the Department of Surgery, University of California, San
Diego, and the Veterans Administration Hospital, La Jolla, California 921 61
ABSTRACT
Contracting fibrous tissues from skin wounds in pigs, and scars
around silicone implants in humans, contained fibroblasts t h a t had multiple
bundles of 60-80 A microfilaments with electron-dense bodies, features typical
of contractile fibroblasts both in vivo (myofibroblasts) and in vitro. These in
vivo fibroblasts contained many 220 A diameter microtubules, which paralleled
plasma membranes in the long axis of the cells. The spatial orientation of the
microtubules in these myofibroblasts strongly suggests a bracing or scaffolding
function.
Extensive study of cultured fibroblasts t h a t
a r e both mobile and contractile has demonstrated microtubules and microfilaments t h a t
appear intimately related to cell movement
and mitosis (Porter, ’76; Goldman e t al., ’73).
Morphologically similar fibroblasts occur in
contracting wounds and scars in living animals and humans. Gabbiani and associates
(’72, ’76) described contractile fibroblasts,
which they named “myofibroblasts,” t h a t
share electron microscopic characteristics of
smooth muscle cells and fibroblasts. Gabbiani
e t al. (’72, ’76) described these cells as containing bundles of 60-80 A microfilaments with
electron dense bodies, convoluted nuclei, desmosomes, and basal lamina. Immunologically
(Gabbiani e t al., ’73) and pharmacologically
(Gabbiani e t al., ’72; Majno et al., ’711, myofibroblasts s h a r e many characteristics of
smooth muscle cells and are thought to be responsible for wound and scar contraction in
vivo. Gabbiani e t al.’s descriptions of in vivo
contractile fibroblasts did not stress t h e presence of microtubules, but recent studies using
colchicine t o control wound contraction in
vivo (Ehrlich e t al., ’77) indicate t h e need to
determine if microtubules in myofibroblasts
have a n orientation t h a t might explain this
drug effect.
MATERIALS AND METHODS
Tissue harvesting
On the backs of ten Pitman-Moore 30-40 kg
pigs, two 5 x 5 cm full thickness skin wounds
were created. These wounds were protected
ANAT. REC. (1978) 191: 169-182.
with a leather coat for two weeks, and healed
by slow contraction and epithelialization.
From each wound, multiple small biopsies
measuring 1 x 1 x 3 mm were taken weekly
for eight weeks, and then every four weeks for
a total of 20 weeks of observation. A total of 35
biopsies was taken from t h e wounds, with care
being taken to avoid biopsying the same area.
Weekly area measurements of t h e granulating wounds were done via tracings of clear
plastic followed by planimetry.
Human specimens were obtained from contracted tissue around twenty silicone breast
implants, at the time of surgical revision.
Tissue preparation
Tissue for light microscopy was embedded in
paraffin blocks and stained with hematoxylin
and eosin after formaldehyde fixation. The
specimens for electron microscopy were placed
immediately in room temperature (21-22OC)
solution of 4% paraformaldehyde and 5%
glutaraldehyde buffered with 0.1 M sodium
cacodylate, pH 7.2-7.4.After four hours of fixation, tissues were rinsed three times with
cacodylate buffer a n d postfixed w i t h 2%
cacodylate buffered osmium tetroxide at 4°C.
Tissue blocks were progressively dehydrated
with ethanol, followed by propylene oxide and
then propylene oxide-Epon mixtures, proceeding to Epon. Epon blocks were placed in
Received Oct. 4, ‘77. Accepted Jan. 13, ‘78.
‘ This investigation was supported in part by the Medical Research Service of the Veterans Administration, and by NIH Grant
G M 24648-01.
169
170
ROSS RUDOLPH AND MARILYN WOODWARD
embedding molds with fresh Epon, and cured
overnight a t 60°C. Thick sections were stained
with methylene blue. Thin sections were
stained with uranyl acetate and lead citrate,
and were examined at 16,000 to 30,000 x
using a Zeiss 10 electron microscope. Myofibroblast population was estimated by seeking
microfilament bundles in cells in thin sections; from each tissue block ( 5 5 in all), thick
sections were cut and used for orientation of
thin sections. Thick section face blocks usually contained 60-80 cells, which were examined
for myofibroblast characteristics.
RESULTS
Pig wounds. Wounds within the first week
of surgery became thickly crusted with a
non-purulent exudate, and began to contract.
Rapid contraction of the wounds occurred,
with decrease to 42.1 % Standard Error of the
Mean 5.2% surface area at four weeks. By
eight weeks, wound area was stable at 25.0 2
S.E.M. 7.0% of surface area. Weekly biopsies
under light microscopy showed a typical appearance of granulation tissue containing new
fibroblasts, new collagen bundles, and multiple small capillaries. During the eight weeks
t h a t i t took wounds to heal, the granulation
tissue matured with formation of larger collagen bundles and growth of epithelium across
the wound.
Human tissue. Gross scar around 20 silicone breast implants showed contraction leading to hardness and rounding up of the silicone
implant, typical of the contraction process
t h a t can occur around such implants (Domanskis and Owsley, ’76; Wilflingseder e t
al., ’74). Light microscopy showed dense connective tissue, with collagen bundles parallel
to the surface of the scar around the implant.
Electron microscopy
In the pig wounds, within the first three
weeks many fibroblasts were seen containing
abundant Golgi apparatus and rough endoplasmic reticulum with small surface area,
indicating active protein synthesis (fig. 1).
These fibroblasts were surrounded by collagen
bundles. A large number of the fibroblasts,
estimated to be as high as 80% by two weeks,
contained features characteristic of both fibroblasts and smooth muscle cells. Fibroblast
cytoplasm contained large bundles of parallel
60-80 A filaments with electron-dense bodies
(figs. 1-5).Paralleling the occurrence of these
microfilament bundles in fibroblasts, micro-
tubules were seen with specific orientation. In
fibroblasts in 2-week-old healing pig wounds
(fig. l), microtubules of considerable length
were seen t h a t paralleled the large bundles of
microfilaments. Figure 2 shows another fibroblast in a two week old contracting wound,
with microtubules t h a t appear to run directly
into bundles of fine parallel filaments a t the
edge of the cell membrane. Well defined desmosomes were also seen (fig. 3) in these fibroblasts.
Figure 4 shows a fibroblast from a wound at
five weeks, containing long parallel microtubules. In addition, microtubules appear to
crisscross t h e cell. Microtubules remained
plentiful during t h e two-to-eight week period
when the large bundles of microfilaments,
typical of in vivo “myofibroblasts” were also
plentiful. After 8 and 12 weeks of observation,
wound myofibroblasts became less plentiful
and at 16 to 20 weeks, no cells were seen containing the large bundles of parallel microfilaments. Microtubules were rarely seen in fibroblasts in these wounds, and instead 100-120 A
intermediate filaments were especially prominent. When microtubules were present, they
were seen only a s short lengths and of random
orientation, without the specific spatial orientation found in the fibroblasts seen earlier.
In human contracted scar (fig. 5 ) , fibroblasts were also seen containing large bundles
of microfilaments with electron-dense bodies.
Spatially oriented microtubules paralleled the
microfilament bundles, and joined smaller
bundles at the edge of t h e cells.
DISCUSSION
Wound contraction is a basic process of animal wound healing, and this process has recently been identified as a function of specialized fibroblasts (Gabbiani e t al., ’72, ’76;
Majno et al., ’71). These contractile fibroblasts, in addition to having rough endoplasmic reticulum and being surrounded by collagen, have large bundles of fine parallel microfilaments with electron-dense bodies. The
microfilament bundles are similar to those
seen in smooth muscle cells, and are theorized
(Gabbiani e t al., ’72, ’76) as being the active
component of the in vivo contractile fibroblast. These fibroblasts also contain convoluted nuclei, desmosomes, and basal lamina,
characteristics again of smooth muscle cells.
Immunologic staining with fluorescein a n tibody to human smooth muscle cells heavily
stains t h e myofibroblasts (Gabbiani e t al.,
MICROTUBULES IN CONTRACTILE FIBROBLASTS
’73). Pharmacologic treatment of human and
animal granulation tissue strips with agents
t h a t would normally s t i m u l a t e o r relax
smooth muscle (Gabbiani et al., ’72; Majno et
al., ’72; Madden e t al., ’74; Ryan e t al., ’74)
similarly stimulates or relaxes t h e granulation tissue, again suggesting t h a t i t contains
cells t h a t react like smooth muscle cells. Fibroblasts with smooth muscle-like features
have also been found in normal organs such as
adrenal gland (Bressler, ’731, rat testicle
(Gorgas and Bock, ’741, duodenal villi (Giildner e t al., ’721, and lung (Kapanci e t al., ’74).
Pathologically contracted tissues such as
burn scars, cirrhosis and Dupuytren’s contracture also contain myofibroblasts (Ryan e t al.,
‘74; Gabbiani e t al., ’72).
Many studies of contractile fibroblasts in
tissue culture have shown structural characteristics much like those seen in vivo. Mobile
fibroblasts in tissue culture contain bundles
of microfilaments (Wessells e t al., ’71; Vasiliev and Gelfand, ’761, along with larger (100120 A) intermediate filaments, and microtubules. Probably t h e myofibroblasts of Gabbiani e t al., and the contractile fibroblasts
seen in tissue culture, are the same cells. In
teleological terms, probably fibroblasts in tissue culture have contractile properties because in t h e living animal they contract and
close wounds.
In tissue culture studies, substances such as
colchicine and cytochalasin B have specific effects on contractile fibroblasts. Cytochalasin
B causes disruption of the bundles of microfilaments (Wessells e t al., ’711, while colchicine prevents assembly of microtubules
(Vasiliev and Gelfand, ’76). Recent studies by
Ehrlich e t al. (’77) utilized these agents not in
tissue culture but in live animals to see if drug
effects observed in vitro had parallels in living
tissue. Contracting wounds were treated topically with these agents, and i t was found t h a t
colchicine but not cytochalasin B inhibited
the contraction of wounds. Thus, Ehrlich e t al.
(’77) stated t h a t microtubules might play a
significant role in t h e contraction of myofibroblasts in vivo.
Microtubules have not been previously included as a prominent feature of myofibroblasts. In studies from Gabbiani’s group (Ryan
et al., ’74), description was made of “scattered
microtubules” of 220 A diameter but no specific spatial orientation was noted. Our study
shows t h a t in contractile fibroblasts in vivo,
microtubules are quite prominent but only
171
during a relatively limited time span when
t h e classic myofibroblasts are plentiful (Rudolph e t al., ’77). The microtubules occur parallel to the long bundles of 60-80 A microfilaments. In addition, both in humans and in
pigs, they occur parallel to and joining smaller
bundles of microfilaments at the cell borders.
The especially close relationship of the microtubules and microfilaments is particularly
evident in figures 4 and 5 , and suggests t h a t
microtubules may be important in facilitating
effective microfilament contraction.
The spatial orientation of the microtubules
- parallel to t h e long axis of the fibroblasts
and occasionally crossing the cell a t long
angles - suggests a bracing or scaffolding
function. Microtubules have been considered
as important structural elements of cellular
cytoskeleton (Porter, ’761, probably without
significant inherent contractile ability. Thus,
t h e inhibitory effect of colchicine on wound
contraction may be via reduction of structural
bracing within myofibroblasts, preventing effective cellular Contraction.
LITERATURE CITED
Bressler, R. S. 1973 Myoid cells in the capsule of the
adrenal gland and in monolayers derived from cultured
adrenal capsules. Anat. Rec., 177: 525-531.
Domanskis, E. J., and J. Q. Owsley 1976 Histological investigation of the etiology of capsule contracture following
augmentation mammaplasty. Plast. Reconstruc. Surg.,
58: 689-693.
Ehrlich, H. P., G. Grislis and T. K. Hunt 1977 Evidence for
the involvement of microtubules in wound contraction.
Am. J. Surgery, 133: 706-709.
Gabbiani, G., B. J. Hirschel, G. B. Ryan, P. R. Statkov and
G. Majno 1972 Granulation tissue as a contractile organ: A study of structure and function. J. Exp. Med., 135:
719-734.
Gabbiani, G., M. LeLouis, A. J. Bailey, S. Bazin and
A. Delaunay 1976 Collagen and myofibroblasts of
granulation tissue. Virchow’s Arch. B. Cell. Path., 21:
133-145.
Gabbiani, G., G. B. Ryan, J. P. Lamelin, P. Vassalli, G.
Majno, C. A. Bouvier, A. Cruchaud and E. F. Luscher
1973 Human smooth muscle auto-antibody: Its identification as antiactin antibody and a study of its binding
to “non-muscular” cells. Am. J. Path., 72: 473-484.
Goldman, R. D., G. Berg, A. Bushnell, C. Chang, L. Dickerman, N. Hopkins, M. Miller, R. Pollack and E. Wang 1973
Fibrillar systems in cell motility. In: Locomotion of Tissue Cells. Elsevier, Amsterdam, pp. 83-107.
Gorgas, K., and P. Bdck 1974 Myofibroblasts in the rat
testicular capsule. Cell Tissue Res., 154: 533-541.
Guldner, F. H., J. R. Wolff and D. Grafkeyserlingk 1972 Fibroblasts as a part of the contractile system in duodenal
villi of rat. Z. Zellforsch. Mikrosk. Anat., 135: 349-360.
Kapanci, Y., A. Assimacopoulos, C. Irle, A. Zwahlen and
G. Gabbiani 1974 “Contractile interstitial cells” in pulmonary alveolar septa; A possible regulator of ventilationiperfusion ratio? J. Cell. Biol., 60: 375-392.
Madden, J . W., D. Morton, J r . and E. E. Peacock,
172
ROSS RUDOLPH AND MARILYN WOODWARD
Jr. 1974 Contraction of experimental wounds. I. I n hibiting wound contraction by using a topical smooth
muscle antagonist. Surgery, 76: 8-15.
Majno, G., G. Gabbiani. B. J. Hirschel. G . B. Ryan a n d P. R.
Statkov 1971 Contraction of granulation tissue in uitro:
Similarity to smooth muscle. Science, 173: 548-550.
Porter, K. R. 1976 Motility in cells. I n : Cell Motility.
Cold Spring Harbor Conference on Cell Proliferation.
Vol. 111. Cold S p r i n g Harbor Laboratory, New York.
pp. 1-28.
Rudolph, R.. S. Guber, M. Suzuki a n d M. Woodward 1977
T h e life cycle of t h e myofibroblast. Surg.. Gyn.. & Obstet.,
145: 389-394.
Ryan, G. B.. W. J. Cliff, G. Gabbiani, C. Irle, D. Montandon.
P. Statkov a n d G. Majno 1974 Myofibroblasts in h u m a n
granulation tissue. H u m a n P a t h . , 5: 55-67.
Vasiliev, J. M., a n d I. M. Gelfand 1976 Effects of colcemid
on morphogenic processes a n d locomotion of fibroblasts.
In: Cell Motility, Cold Spring Harbor Conference on Cell
Proliferation. Vol. 111. Cold Spring Harbor Laboratory,
New York, pp. 279-304.
Wessells, N. K., B. S. Spooner, J. F. Ash, M. 0.Bradley, M. A.
Luduena, E. L. Taylor, J. T. Wrenn a n d K. M. Yamada
1971 Microfilaments i n cellular a n d developmental processes. Science, 171: 135.143.
Wilflingseder, P., A. Propst a n d G. Mikuz 1974 Constructive fibrosis following silicone implants in mammary a u g mentation. Chir. Plastica 2: 215-229.
PLATE 1
EXPLANATION OF FIGURE
1 Fibroblast from a 2-week-old contracting wound in a pig. Rough endoplasmic reticulum is active with many ribosomes. Long microtubules (arrows1 of 220 A diameter
a r e prominent, r u n n i n g parallel to t h e cell membrane a n d t o each other. A bundle of
60 8, parallel microfilaments with electron-dense bodies (asterisk) is typical of in
vivo contractile fibroblast (“myofibroblast”). X 40,000.
MICROTUBULES I N CONTRACTILE FIBROBLASTS
Ross Rudolph and Marilyn Woodward
PLATE 1
173
PLATE 2
EXPLANAlION OF F I G U R E
2 Terminal process of a fibroblast from a 2-week-old wound in a pig. Microtubules a r e
frequent and appear to r u n parallel and into short bundles of 60 A microfilaments
a t cell membrane, some of which extend into t h e extracellular space (arrow).
Pinocytic vesicles and rough endoplasmic reticulum a r e prominent. X 45,000.
174
MICKOTUBULES IiX CONTHACTILE FIBROBLASTS
Ross Rudolph and Marilyn Woodward
PLATE 2
175
PLATE 3
EXPLANATION OF FIGURE
3 Myofibroblasts in a 2-week-old wound in a pig. Desmosome and gap junction
(asterisk a t upper left) are present. Microtubules (arrows) parallel the cell membranes. X 40,000.
176
MICROTUBULES IN CONTRACTILE FIBROB1,ASTS
Ross Rudolph a n d Marilyn Woodward
177
PLATE 4
EXPLANATION OF FIGURE
4 Fibroblast in a 5-week-old contracting wound in a pig. Prominent long microtubules
. microfilaments with electron-dense bodies
parallel thick bundles of 60-80 h
(asterisk).Microtubules also traverse the cell a t an oblique angle, suggesting a bracing function (arrows). X 40,000.
178
MICROTUBU1,ES IN ('ONTRACTILE FIBROBLASTS
Ross Rudolph a n d M a r i l y n Woodward
PI.A'TE 4
179
PLATE 5
EXPLANATION OF FIGURE
5 Fibroblast in a human contracting fibrous scar around a silicone rubber implant,
four weeks after onset of contraction. Prominent microtubules (arrows) appear to
join large bundles of microfilarnents. x 40,000.
180
MICROTUBULES I N CON'IKA('TILE FIBKOBLASTS
Ross Kudolph a n d Marilyn Wondward
PI.ATE 5
181