- Email: [email protected]
Developmental changes in actin and myosin heavy chain isoform expression in smooth muscle
ARCHIVES
Vol.
OF BIOCHEMISTRY
AND
284, No. 2, February
BIOPHYSICS
1, pp. 232-237,199l
Developmental Changes in Actin and Myosin Heavy Chain lsoform Expression in Smooth Muscle’ Thomas
J. Eddinger2
Department
Received
of
Physiology,
June 25,1990,
and Richard University
and in revised
form
of
A. Murphy3 Virginia
School
September
lo,1990
of
Medicine,
Smooth muscle cells express isoforms of actin and myosin heavy chains (MHC). In early postnatal animals the nonmuscle (NM) actin and MHC isoforms in vascular (aorta) smooth muscle were present in relatively high percentages. More than 30% of the MHC and 40% of the actin isoforms were NM. The relative percentage of the NM isoforms decreased significantly as the animals reached maturity, with NM MHC less than 10% and NM actin less than 30% of the totals. Concurrent with this decrease in NM isoforms was an increase in the smooth muscle (SM) isoforms. The relative changes and time frame in which these changes occurred were very similar for the actin and MHC isoforms. In arterial tissue there were species differences for changes with development in the two SM MHC isoforms (SMl and SM2). The ratio of SMl:SMB in young rat aorta was approximately 0.5, while this same ratio was approximately 3 in young swine carotid. Both adult rats and swine had a SMl:SMB MHC ratio of approximately 1.2. Rat bladder smooth muscle showed no significant change in NM vs SM ratio between young and old rats, while the SMl:SMS ratio decreased from 2.7 to 1.7 between these age groups. The shifts in (Yand B actin were similar to those in the vascular tissue, B 1991 Academic PMS, IN. but of much smaller magnitude.
Six different isoactins are known to exist in vertebrates (1, 2). In smooth muscles there are two smooth muscle (SM)4 specific actin isoforms (a-SM and -r-SM) and two nonmuscle (NM) isoforms (P-NM and -r-NM) (2-4). i This research was supported by NIH Grant 5-POl-HL19242 and a NIH postdoctoral Fellowship (l-F32-HL07528) to T.J.E. ’ Present address: Biology Department, Marquette University, Milwaukee, WI 53233. 3 To whom correspondence should be addressed at Box 449, Health Sciences Center, University of Virginia, Charlottesville, VA 22908. 4 Abbreviations used: SM, smooth muscle; NM, nonmuscle; MHC, myosin heavy chains; PSS, physiological saline solution; EDTA, ethylenediaminetetraacetic acid; IEF, isoelectric focusing; SDS, sodium dodecyl sulfate.
Charlottesville,
Virginia
22908
There is evidence for three unique myosin heavy chains (MHC) in smooth muscle (SMl, SM2, NM) (5-10) and two unique 20-kDa regulatory light chains (one SM and one NM) (11,6). Two 17-kDa alkali light chain isoforms are also present (12-15). Molecular biological approaches continue to verify and complement biochemical studies in identifying various protein isoforms in smooth muscles (20-kDa myosin light chains (16,17); myosin heavy chains (18-20)).
It is not known whether the various heavy and light chains combine readily to form many native myosin variants, nor whether such variants functionally differ. Developmental changes in the actin isoforms have been observed. Kocher et al. (21) and Owens and Thompson (22) reported that the fractional content of ol-SM-actin increases significantly with development in the rat aorta (with P-NM-actin decreasing). These changes may be due both to an increase in ol-SM-actin synthesis and to a decrease in a-SM-actin degradation (22). McConnell et al. (23) suggested that MHC expression in rabbit bladder smooth muscle is developmentally and neurally regulated. Swine tracheal smooth muscle was reported to show a decrease in the SMl:SMB ratio with increasing age, with no NM MHC present at any age (24). Borrione et al. (25) and Kuro-o et al. (26) reported that developing rabbit aorta expresses NM MHC only during early stagesof development. It is replaced by both SMl and SM2 in the adult animal. SMl and SM2 are the only two isoforms detectable in adult rabbit aorta (25, 26). This paper describes the developmental expression of both actin and MHC isoforms in rat and MHC isoforms in swine. Changes in the expression of the smooth vs nonmuscle isoforms for both actin and MHC occurred in a similar time frame in the rat aorta. Change in NM vs SM isoforms was much lessin the rat bladder. The relative expression of the SM-specific MHC isoforms appears to vary with respect to both tissue type and species. MATERIALS AND Tissue preparation. days),
adult
METHODS
Sprague-Dawley SpragueeDawley rats (retired
232 All
rat pups (3, 6, 10, 14, and 21 breeder females), and piglets
Copyright 0 1991 rights of reproduction
0003.9861/91 $3.00 by Academic Press, Inc. in any form reserved.
ACTIN
AND
MYOSIN
HEAVY
(2-3 kg, 5-7 days) were euthanized. Tissue samples were removed and placed in cold physiological saline solution (PSS): 140 mM NaCl, 4.7 mM KCl, 1.2 mM Na2HPOI, 1.2 mM MgSO,, 1.6 mM CaCl,, 5.6 mM glucose, 0.02 mM ethylenediaminetetraacetic acid (EDTA), and 2 mM 3-(N-morpholino)-propanesulfonic acid, pH 7.4. Tissues from adult swine (So-100 kg) were obtained from a slaughterhouse and transported in cold PSS. With the exception of certain control tissues, all blood vessels were cleaned of the adventitial layer by stripping (27). This was done under a dissecting microscope with the tissue submerged in PSS. To verify that the removal of the adventitia from the media was complete, samples from both young (3-day) and old (adult) rat aortas were processed for transmission electron microscopy (28). Nonvascular tissues were cleaned of adhering connective tissue. The mucosal layer was stripped away from stomach samples. Bladders and uteri were opened and scraped on both surfaces prior to grinding in homogenization buffer. It was necessary to pool tissue from rat pups. Approximately 15 3day-old rat pups were required for a single sample, with fewer pups required as they got larger. Two pups were used for a single sample at 21 days. Tissues were analyzed individually for the adult rats and all swine samples. Gel electrophoresis. Tissues were ground using glass homogenizers in isoelectric focusing (IEF) buffer (1% sodium dodecyl sulfate (SDS), (4). They were then sonicated and 10% glycerol, 20 mM dithiothreitol) stored at -2O’C until run on gels. Myosin heavy chains were resolved on SDS-polyacrylamide gels using the methods of Giulian et al. (29). Actin isoforms were resolved using a modification of O’Farrell’s (30) two-dimensional IEF/SDS electrophoresis technique described by Fatigati and Murphy (4) using a 5.0-6.0 pH gradient (Pharmalyte, Sigma Chemical Co., St. Louis, MO). Western blots. Western blots were performed as described previously (6). Antisera specific for the NM MHC (rabbit anti-human platelet myosin) or NM + SM MHC (rabbit anti-bovine uterus myosin) were obtained from Biomedical Technologies, Inc. (Stoughton, MA). SMspecific antisera were generated in chickens using purified rat smooth muscle myosin as the antigen. Densitometty. Proteins were stained using standard procedures with Coomassie brilliant blue R. Gels were scanned using a Quick Scan Jr. densitometer with custom-modified high-resolution optics (Helena Laboratories, Beaumont, TX). Peak areas were determined by integration of these scans. Loading curves were performed to verify linearity of loading volume with integrated peak area. Statistics. The relative amounts of protein isoforms were determined for each sample. Multiple lanes of a given sample were averaged prior to determining the mean and standard errors for different samples of each isoform. Variability between repeat lanes and gels was usually l3% and never more than 5%. A Student t test was used to compare values between young and adult swine. A Newman-Keuls multiple range test was used to compare the data of various rat age groups. Significant differences are reported at P < 0.05.
CHAIN
ISOFORM
233
EXPRESSION
dothelial cells were noted on the luminal surface of the samples followed by several layers of alternating smooth muscle cells and elastic lamena and the adventitial layer. No adventitial layer was present in the stripped tissues. The effect of stripping on fractional isoform composition was more significant for actin than for MHC. For all age groups, the relative cu-SM-actin content was 5% greater and the /‘3-NM-actin content was 5% lower when the tissue samples were stripped of their adventitia prior to processing than when they were not. Separation of the three MHC isoforms by gel electrophoresis was verified by Western blotting (Fig. 1). As we reported previously (5, 6) the upper two bands were recognized by smooth muscle antisera, while the lower band was recognized by nonmuscle antisera. These gels completely resolve the three MHC isoforms so that quantitation may be done on Commassie blue-stained gels. Typical gel results for actin and MHC isoforms from rat aorta for each age group are shown in Figs. 2 and 3. The two-dimensional gel system does not resolve r-NMactin from y-SM-actin. The changes with age in the fractional content of the actin and myosin isoforms are shown in Fig. 4. The r-SM + NM-actin isoforms decreased from 19 to 11% from 3 days to adult. The cr-SM-actin increased from 40 to 64% as P-SM-actin decreased from 40 to 25% over this age range (Fig. 4). The differences between the 3- or 6-day-old animals and all the older animals were significant. Changes were essentially complete by 21 days of age in the rat aorta. The NM MHC was present at the greatest levels in the young animals and decreased with development (Fig. 4, lower panel). This approximately 25% decrease in NM MHC (greater than 30% at 3 and 6 days to less than 10% in the adult) was accounted for by the SMl MHC, which showed a 25% increase between 3 days and adult. All changes between the age groups were significant (with
SDS Gel
Western SM
NM
Blots SM + NM
RESULTS
The contribution of actin and/or MHC isoforms from adventitial fibroblasts or endothelial cells to estimates of aortic smooth muscle cells was minimized by stripping the media from the adventitia. This procedure also caused abrasion of the intimal surface, removing most of the endothelial cells. These “stripped” medial preparations and the remaining adventitia were run on the appropriate gels along with unstripped samples. Young (3-day) and adult unstripped rat aortas that had been cleaned of loose fatty connective tissue and those which had been stripped of their adventitia were fixed, sectioned, stained, and examined by transmission electron microscopy. Some en-
-+*- x FIG. 1. High-molecular-weight region of Coomassie blue-stained 8% SDS-polyacrylamide gel and Western blots of the same region of identical gels. Two loadings of adult rat aorta are shown for each condition. The CBB-stained gel shows the three MHC isoforms. The Western blots verify that the upper two bands are “smooth muscle” MHC, while the lower band is “nonmuscle” MHC. The SM antiserum was chicken anti-rat stomach myosin. The NM antiserum was rabbit anti-human platelet myosin. The NM + SM antiserum was rabbit anti-bovine uterus myosin.
234
EDDINGER
AND
MURPHY
ACTIN
077
2
MW
MYOSIN HEAVY CHAIN
I 42K
5!35
250
PHFIG. 2. Actin-containing portions of Coomassie blue-stained, twodimensional IEF (X-axis)/SDS (Y-axis) gels showing fractional content of actin isoforms from rat aortic medial preparations at various ages. 01, ol-SM-actin; /3, non-NM-actin; y, y-SM+ y-NM-actin.
0.0-w 0
5
15
10
AGE
the exception of the 3- and lo-day NM MHC values). While SM2 fell in fractional content at I4 and 21 days, there was no significant difference between the values observed for the younger ages (3,6, and 10 days) and the adult animals. Figure 5 shows the relationship of the a-SM- and pNM-actin isoforms to the SM and NM MHC isoforms over this age range. The combined SM MHC fractional composition (SMl and SM2) of 69% in the R-day-old rats increased to 93% in the adult animals. The cr-SM-actin increased from 40 to 64%. These increases in SM protein were countered by the decrease in the NM proteins. NM MHC decreased from 31% at 3 days to 7% in the adult rat, while the P-NM-actin decreased from 40 to 25%. Plotting SM- or NM-actin and MHC on the same graph shows the parallel changes with age for increases in the SM isoforms and the decreases in the NM isoforms (Fig. 5). There were also changes in the MHC isoforms with development in swine. The fractional composition of the
6
ADULT
(DAYS)
FIG. 4. Developmental changes in fractional content of the actin and MHC isoforms in rat aortic media. Values shown are means f SE (when the SE exceeds the symbol size) from densitometric evaluation of Coomassie blue-stained gels as shown in Figs. 2 and 3. Sample numbers for each age group for actin and MHC isoforms, respectively, are: 3 days, N = 10,9; 6 days, N = 6.5; 10 days, N = 5,5; 14 days, N = 7,7; 21 days, N = 10,lO; adult, N = 11,13. Samples were pooled for each N to obtain enough tissue at younger ages.
three MHC isoforms present in three types of smooth muscle at two different ages are shown in Table I. Each tissue exhibited somewhat different changes. However, the general trends were for a fall in NM MHC, in large
l.Or
SM MYOSIN
(Days)
Age 3
20
10
14
21
A NM MYO:
LL
0.0I’ 0
5
10
15
20
ADULT
AGE (LMYS) FIG. 3. High-molecular-weight region of Coomassie SDS-polyacrylamide gels showing fractional content of from rat aortic medial preparations at various ages. A, Resolved are the 196-kDa (NM), 200-kDa (SM2), and MHC variants. Separation variability is due to different
blue-stained 8% myosin isoforms adult female rat. 204-kDa (SMl) gels.
FIG. 5. Changes in fractional content of a-SMand p-NM-actin and SM and NM MHC isoforms present at various ages in rat aortic media. Values are means f SE (where the SE exceeds the symbol size) from densitometric evaluation of Coomassie blue-stained gels. Sample number and pooling are described in Fig. 4.
ACTIN
AND
MYOSIN
HEAVY
CHAIN
TABLE
Fractional
Content
of MHC
Isoforms
Present
ISOFORM
I
in Young (2-3 kg, 5-7 Days) and Adult
Piglet n Carotid Stomach Uterus
SMl
7 8 5
NM
14.4 * 1.4 17.5 * 2.1 11.9 f 1.3
39.4 f 1.7 7.8 k 1.0 35.0 i- 1.1
part reflecting an increase in SM2. Unlike the data from rat aorta, the swine carotid pattern differed for the SM MHC isoforms (Table II). In the young rat aorta the fractional content of SMl was approximately half that of SM2. In the young swine carotid, SMl was approximately three times SM2. In both animals the adult ratio of SMl: SM2 was approximately 1.2. The pattern in the rat bladder differed from that in the rat aorta (Fig. 6; compare with Fig. 4). a-SM- and P-NMactin fractional content showed a significant increase and decrease, respectively, but of much smaller magnitude (7 and 5% respectively vs 25 and 15% in the rat aorta). ySM + NM-actin fractional content was high and showed no changes with age. The NM MHC showed no significant differences between the 3-day and the adult animals. This contrasts with the 25% decrease seen in NM MHC in the rat aorta over these same ages. SMl and SM2 MHC fractional contents were more similar to changes in the swine tissues than to changes in the rat aorta (i.e., SMl started out high and decreased rather than starting low and increasing). SMl showed a decrease (0.66 to 0.58) and SM2 an increase (0.25 to 0.35) in fractional content between 3 days and adult (Fig. 6). The changes in SMl and SM2 relative content at 3 days of age and adult were significant,
kg) Swine
swine
n
SMl
SM2
5 5 5
51.1 2 1.1 51.6 + 5.5 45.4 2 5.3
40.1 + 1.3 44.1 f 5.1 37.1 3~ 4.2
Note. Values shown are means * SE from densitometric evaluation of Coomassie blue-stained age groups are significantly different (P < 0.05) with the exception of SMl in the uterus.
TABLE
(So-100 Adult
SM2
46.2 + 1.6 74.7 + 2.6 53.1 f 1.4
235
EXPRESSION
gels. All isoform
NM
comparisons
but always exceeded the 1.7 SMl:SMS the adult animal.
8.8 f 0.8 4.3 f 0.5 17.5 f 2.3 between
these two
ratio observed in
DISCUSSION
The possibility that protein contributions from cells in the adventitial layer affected the measurements was ruled out by the use of electron microscopy showing complete removal of the adventitia from the stripped medial preparations. Samples from each age group (unstripped and stripped of adventitia) were also run on SDS denaturing gels in addition to the stripped samples. In unstripped aortas the fractional content, on average, was 5% lower
II
Comparison of Fractional Content for the Three MHC Isoforms Present in the Medial Tissue of Swine Carotid and Rat Aorta at Young (6 Days) and Adult Ages (Market Weight Swine and Breeder Female Rats)
2 LL
O5 04 SM2
03
/.fe
t. 0.2
II.
SMl
SM2
NM
0.1 i
Rat aorta B-Day rat pup Breeder female Swine carotid B-Day piglet 80 to loo-kg swine
$-----,i---,
0.0-M
5 13
20.2 -t 1.4 50.3 3L 0.9
44.5 f 1.0 42.4 f 0.9
35.3 + 1.1 7.3 f 0.4
7 5
46.2 k 1.6 51.1 f 1.1
14.4 +- 1.4 40.1 f 1.3
39.4 * 1.7 8.8 2 0.8
Note. Values shown are means * SE. All isoform comparisons these two age groups are significantly different (P < 0.05).
between
3
Age FIG. 6.
Adult
21
(Days)
Changes in fractional content of the actin and MHC isoforms in rat bladder. Values shown are means f SE (when SE exceeds the symbol size) from densitometric evaluation of Coomassie blue-stained gels. Sample numbers for each age group for actin and MHC isoforms, respectively, are: 3 days, N = 5,4; 21 days, N = 5,5; adult, N = 7,7.
236
EDDINGER
for cr-SM-actin and 5% higher for P-NM-actin than in the stripped media by itself. There was no consistent effect on the r-SM + NM-actin. Stripping caused no significant change in the fractional content of the SM or NM MHC. The large actin content and small myosin content of nonmuscle cells present in the adventitia could explain why actin isoform fractional content was affected to a greater extent than MHC. Fibroblasts have a relatively high percentage of P-NM-actin. These results show that the adventitial layer can contribute enough P-NM-actin to significantly increase the relative fraction of this isoform over that present in the medial smooth muscle cells alone. The changes in the actin isoform fractional content with development in rat aorta (Figs. 2 and 4) confirm other reports (21, 22). Small differences in the absolute values may be explained by differences between rat strains and/or incomplete removal of the adventitia. Removal of the adventitia is technically difficult in the young animals. Adventitial and possibly endothelial cell contamination can alter the relative percentages of the isoactins. Occasional endothelial cells were observed in some of the tissues prepared for electron microscopy. However, the total contribution of any given isoform from the remaining endothelial cells to the medial samples was very small. The endothelial cells make up only a small fraction of the aorta, and only a small percentage of these cells remained on the medial samples. In addition, the quantities of actin and myosin present in endothelial cells are much lower than those in smooth muscle cells. The major change in isoform composition from young to old rats occurred between 1 and 3 weeks of age (Fig. 4). Changes in the fractional content of actin and MHC isoforms occurred during a similar time frame and to a similar extent (Fig. 5). This correlation may reflect coordinated regulation of the NM or SM MHC and actin isoform genes. A functional significance for the NM and SM isoforms remains to be established. Young piglets and adult swine were used to compare the MHC isoform fractional content in vascular, gastric, and uterine smooth muscle (Table I). All three tissue types show qualitative changes similar to those observed in the rat aorta (SM vs NM isoforms). The magnitude of these changes varied considerably among the tissues studied. The ratio of SMl to SM2 MHC decreased between the young and adult pigs from 4 to 1.3 in the carotid, from 4.2 to 1.2 in the stomach, and from 4.5 to 1.2 in the uterus. These changes contrast with those in the rat aorta, where the SMl:SM2 MHC ratio increased from 0.5 to 1.2 between young and adult animals. In the rat bladder, the SMl:SMB MHC ratio decreased from 2.7 to 1.7. This was similar to, but of a smaller magnitude than, the change in rabbit bladder. McConnell et al. (23) reported that the ratio of SMl:SMB in rabbit bladder smooth muscle decreased from 3.8 to 0.7 between newborn and adult animals. Mohammad and Sparrow (24) observed a SMl:SM2 ratio decrease from 2.1 to 1.0 in swine trachea smooth
AND
MURPHY
muscle between young (30-60 kg) and adult (90-140 kg) animals. They reported no NM MHC in this tissue. In human bronchus smooth muscle, Mohammad and Sparrow (10) found no change in the SMl:SMB ratio with age (0.69 in both infants and adults), but abundant NM MHC. The trend was for a relatively high neonatal SMl:SMB MHC ratio that decreased with development. The exceptions were the rat aorta, where the ratio increased, and in the human bronchus (24) or chicken gizzard (31), where no changes were found in the SMl:SM2 ratio with development. Both Kuro-o et al. (26) and Borrione et al. (25) reported developmental regulation of MHC in rabbit aorta. This was another unique developmental pattern in that there was apparently no SM2 MHC in the embryonic and early neonatal animals. SM2 appears later in development with a concomitant decrease in NM MHC. In the adult there was no NM MHC and the SMl:SMS ratio was approximately 1.8 (25, 26). These differences in SMl and SM2 expression with age were species-specific, as the rat aorta and bladder did not show the same changes. All of the swine tissues studied showed similar directional changes in the SMl:SM2 ratio with development. With the exception of the rat bladder, all of the tissues we examined showed a decrease in NM MHC with development. This NM:SM MHC change parallels the NM:SM actin isoform shift in time and magnitude. The coordinated regulation of these changes would be most readily controlled at the gene level. Unlike the reports for rabbit aorta (25, 26), the NM MHC was also always present over the age range examined. The physiological significance of this NM to SM isoform shift is unknown. There may be several domains in the cell where the NM and SM isoforms function. This has been suggested for the thin filaments (32, 33). More recently, Drew (34) used immunoelectron microscopy of native thin filaments from swine stomach to show that the actin isoforms were not segregated in different thin filaments or along the thin filaments. These results argue strongly against separate populations of NM-actinand SM-actin-containing thin filaments (although heterogeneity might occur due to the differential presence of other thin filament constituents). Drew (34) suggested that the different isoforms may reflect gene regulation without functional implications. By contrast, DeNofrio et al. (35) reported that NM actin was present in membrane ruffles, pseudopods, and stress fibers of pericytes, while SM-actin was localized in stress fibers. These results may differ because of the cell types examined, but need further study. Separate mRNAs for SMl and SM2 have been identified and sequenced (19, 20). The native myosin composition and the role of NM MHC are still unresolved. Sartore et al. (8) reported a differential distribution of SM MHC isoforms among cells and differential immunolocalization of SMl and SM2 within the same cell. Segregation may allow for distinct regulatory and/or
ACTIN
AND
MYOSIN
HEAVY
functional roles. The NM variants may be involved in proliferative, synthetic, and secretory functions, while the SM isoforms underlie contraction. There is also evidence now for two different NM MHC genes and mRNAs (36, 37). Both of the mRNAs were detected in a variety of tissues, including smooth muscle cells. The quantity of mRNA present varied between tissues,but did not appear to change with development. The two NM MHC proteins have not been resolved to date on SDS gels, and it remains unclear if both of these NM MHC are expressed in aortic smooth muscle. This paper confirms earlier reports of changes in contractile protein isoform composition with development. In addition, it provides an explanation for the diversity of results reported in the literature. Species, tissue, and age differences in experimental animals can account for the enormous differences in contractile protein isoform expression reported in the literature. This study also suggests that there is a coordinated regulation of the contractile protein actin and myosin isoforms. ACKNOWLEDGMENTS
CHAIN
ISOFORM
10. Mohammad,
M. A., and Sparrow,
Purdy-Ramos, E. and photography, Swine tissue was Research support HL07528.
Van Neil, and M. Forbes for and Kathy Dobbins for her donated by Smithfield Packwas provided by NIH Grants
11. Mougios,
V., and Barany,
12. Sobieszek, 13. Helper,
J. I., and Gibson,
2. Vandekerckhove,
W. (1976)
J., and Weber,
Cell 9, 7933805.
K. (1978)
J. Mol.
Biol.
126, 783-
3. Vandekerckhove, J., and Weber, K. (1979) Dif-ferentiation 14,123133. 4. Fatigati, V., and Murphy, R. A. (1984) J. Biol. Chem. 259, 14,38314,388. 5. Eddinger, T. J., Gaylinn, B. D., Monical, P. L., and Murphy, R. A. Biophys.
J. 53,
579a.
F., Janmot,
7. Sartore, Saggin, 179,
S., De Marzo, N., Borrione, L., Fabbri, L., and Schiafhno,
Cell
P. A., Monical,
P. L.,
J. Physiol.
C997-
257,
A. C., Zanellato, A. M. C., S. (1989) Eur. J. Biochem.
79-85.
8. Sartore, napieco,
S. Borrione, A. C., Zanellato, A. M. C., Tonello, M., ScanG., Pauletto, P., and Dal Palu, C. (198913) J. Muscle Res. Motil. 10, 167-168.
9. Martin,
A. F. (1988)
Biophys.
J. 53,
577a.
Biophys.
Acta
872,
7, 417-425. J. Biol. Chem.
Electrophoresis
D. R. (1988)
C., Ropert,
S., and d’Albis,
A. (1986)
Eur.
J.
160,507-513.
Biochem.
15. Hasegawa,
Y., Ueno,
H., Horie,
K., and Morita,
F. (1988)
J. Biochem.
103,15-18. 16. Taubman,
M. B., Grant,
J. W., and Nadal-Ginard,
B. (1987)
J. Cell
104,1505-1513.
Biol.
17. Kumar, C. C., Mohan, S. R., Zavodny, P. J., Narula, S. K., and Leibowitz, P. J. (1989) Biochemistry 28,4027-4035. 18. Nagai, R., Larson, D. M., and Periasamy, M. (1988) Proc. Natl. Acad. Sci. USA 85, 104771051. 19. Nagai, R., Kuro-o, M., Babij, P., and Periasamy, M. (1989) J. Biol. Chem.
264,9734-9737.
20. Babij, P., and Periasamy, M. (1989) J. Mol. Biol. 210, 673-679. 21. Kocher, O., Skalli, O., Cerutti, D., Gabbiani, F., and Gabbiani, G. (1985) Circ. Res. 56, 829-838. 22. Owens, G. K., and Thompson, M. M. (1986) J. Biol. Chem. 261, 13,373-13,380. J. D., Ewalt,
D. H., and Boone,
T. B. (1988)
J. Cell
Biol.
107,36a. 24. Mohammad, M. A., and Sparrow, M. P. (1988) FEBS Lett. 228, 1099112. 25. Borrione, A. C., Zanellato, A. M. C., Scannapieco, G., Pauletto, P., and Sartore, S. (1989) Eur. J. Biochem. 183, 413-417. M., Nagai, A., Takaku,
R., Tsuchimochi, F. (1989) J. Biol.
H., and Daly,
28. Purdy-Ramos,
M. M.
(1970)
H.,
Katoh,
Chem.
Proc.
264,
H., Yazaki, Y., 18,272-18,275.
Sot.
Exp.
Biol.
Med.
S., and
Forbes,
M.
S. (1984)
J. Electron
Microsc.
1,271&277.
29. Giulian,
G. G., Moss,
R. L., and Greaser,
M. L. (1983)
Anal.
Biochem.
129,277-287. 30. O’Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021. 31. Takano-Ohmuro, H., Obinata, T., Mikawa, T., and Masaki, J. Biochem.
32. Fiirst,
T. J., Martino, R. A. (1989) Amer.
Biochim.
263, 15,748-15,753. 14. Cavaille,
Tech.
(1988)
J. 260,
135,364-368.
802.
6. Gaylinn, B. D., Eddinger, Hunt, D. F., and Murphy, c1004.
M. (1986)
A., and Jertschin, P. (1986) D. J., Lash, J. A., and Hathaway,
27. Wolinsky, 1. Garrels,
Biochem.
305c308.
26. Kuro-o, Ohkubo,
REFERENCES
M. P. (1989)
421-426.
23. McConnell,
The authors thank S. the electron microscopy word processing prowess. ing Co. (Smithfield, VA). 2POl HL19242 and lF32
237
EXPRESSION
93,
T. (1983)
903-908.
D. 0.. Cross, R. A., De Mey,
J., and Small,
J. V. (1986)
EMBO
J. 5, 251-257.
33. Small, J. V., Fiirst, D. O., and De Mey, J. (1986) J. Cell Biol. 102, 210-220. 34. Drew, J. S. (1989) Ph.D. dissertation, University of Virginia, Charlottesville, University Microfilms, Ann Arbor, MI. 35. DeNofrio, D., Hoock, R. C., and Herman, I. M. (1989) J. Cell Biol.
109,191-202. 36. Kawamoto, 37. Katsuragawa, Eur.
S., and Adelstein, Y., Yanagisawa, J. Biochem. 184,611-616.
R. S. (1989) J. Cell Biol. M., Inoue, A., and Masaki,
109, 275a. T. (1989)
Vol.
OF BIOCHEMISTRY
AND
284, No. 2, February
BIOPHYSICS
1, pp. 232-237,199l
Developmental Changes in Actin and Myosin Heavy Chain lsoform Expression in Smooth Muscle’ Thomas
J. Eddinger2
Department
Received
of
Physiology,
June 25,1990,
and Richard University
and in revised
form
of
A. Murphy3 Virginia
School
September
lo,1990
of
Medicine,
Smooth muscle cells express isoforms of actin and myosin heavy chains (MHC). In early postnatal animals the nonmuscle (NM) actin and MHC isoforms in vascular (aorta) smooth muscle were present in relatively high percentages. More than 30% of the MHC and 40% of the actin isoforms were NM. The relative percentage of the NM isoforms decreased significantly as the animals reached maturity, with NM MHC less than 10% and NM actin less than 30% of the totals. Concurrent with this decrease in NM isoforms was an increase in the smooth muscle (SM) isoforms. The relative changes and time frame in which these changes occurred were very similar for the actin and MHC isoforms. In arterial tissue there were species differences for changes with development in the two SM MHC isoforms (SMl and SM2). The ratio of SMl:SMB in young rat aorta was approximately 0.5, while this same ratio was approximately 3 in young swine carotid. Both adult rats and swine had a SMl:SMB MHC ratio of approximately 1.2. Rat bladder smooth muscle showed no significant change in NM vs SM ratio between young and old rats, while the SMl:SMS ratio decreased from 2.7 to 1.7 between these age groups. The shifts in (Yand B actin were similar to those in the vascular tissue, B 1991 Academic PMS, IN. but of much smaller magnitude.
Six different isoactins are known to exist in vertebrates (1, 2). In smooth muscles there are two smooth muscle (SM)4 specific actin isoforms (a-SM and -r-SM) and two nonmuscle (NM) isoforms (P-NM and -r-NM) (2-4). i This research was supported by NIH Grant 5-POl-HL19242 and a NIH postdoctoral Fellowship (l-F32-HL07528) to T.J.E. ’ Present address: Biology Department, Marquette University, Milwaukee, WI 53233. 3 To whom correspondence should be addressed at Box 449, Health Sciences Center, University of Virginia, Charlottesville, VA 22908. 4 Abbreviations used: SM, smooth muscle; NM, nonmuscle; MHC, myosin heavy chains; PSS, physiological saline solution; EDTA, ethylenediaminetetraacetic acid; IEF, isoelectric focusing; SDS, sodium dodecyl sulfate.
Charlottesville,
Virginia
22908
There is evidence for three unique myosin heavy chains (MHC) in smooth muscle (SMl, SM2, NM) (5-10) and two unique 20-kDa regulatory light chains (one SM and one NM) (11,6). Two 17-kDa alkali light chain isoforms are also present (12-15). Molecular biological approaches continue to verify and complement biochemical studies in identifying various protein isoforms in smooth muscles (20-kDa myosin light chains (16,17); myosin heavy chains (18-20)).
It is not known whether the various heavy and light chains combine readily to form many native myosin variants, nor whether such variants functionally differ. Developmental changes in the actin isoforms have been observed. Kocher et al. (21) and Owens and Thompson (22) reported that the fractional content of ol-SM-actin increases significantly with development in the rat aorta (with P-NM-actin decreasing). These changes may be due both to an increase in ol-SM-actin synthesis and to a decrease in a-SM-actin degradation (22). McConnell et al. (23) suggested that MHC expression in rabbit bladder smooth muscle is developmentally and neurally regulated. Swine tracheal smooth muscle was reported to show a decrease in the SMl:SMB ratio with increasing age, with no NM MHC present at any age (24). Borrione et al. (25) and Kuro-o et al. (26) reported that developing rabbit aorta expresses NM MHC only during early stagesof development. It is replaced by both SMl and SM2 in the adult animal. SMl and SM2 are the only two isoforms detectable in adult rabbit aorta (25, 26). This paper describes the developmental expression of both actin and MHC isoforms in rat and MHC isoforms in swine. Changes in the expression of the smooth vs nonmuscle isoforms for both actin and MHC occurred in a similar time frame in the rat aorta. Change in NM vs SM isoforms was much lessin the rat bladder. The relative expression of the SM-specific MHC isoforms appears to vary with respect to both tissue type and species. MATERIALS AND Tissue preparation. days),
adult
METHODS
Sprague-Dawley SpragueeDawley rats (retired
232 All
rat pups (3, 6, 10, 14, and 21 breeder females), and piglets
Copyright 0 1991 rights of reproduction
0003.9861/91 $3.00 by Academic Press, Inc. in any form reserved.
ACTIN
AND
MYOSIN
HEAVY
(2-3 kg, 5-7 days) were euthanized. Tissue samples were removed and placed in cold physiological saline solution (PSS): 140 mM NaCl, 4.7 mM KCl, 1.2 mM Na2HPOI, 1.2 mM MgSO,, 1.6 mM CaCl,, 5.6 mM glucose, 0.02 mM ethylenediaminetetraacetic acid (EDTA), and 2 mM 3-(N-morpholino)-propanesulfonic acid, pH 7.4. Tissues from adult swine (So-100 kg) were obtained from a slaughterhouse and transported in cold PSS. With the exception of certain control tissues, all blood vessels were cleaned of the adventitial layer by stripping (27). This was done under a dissecting microscope with the tissue submerged in PSS. To verify that the removal of the adventitia from the media was complete, samples from both young (3-day) and old (adult) rat aortas were processed for transmission electron microscopy (28). Nonvascular tissues were cleaned of adhering connective tissue. The mucosal layer was stripped away from stomach samples. Bladders and uteri were opened and scraped on both surfaces prior to grinding in homogenization buffer. It was necessary to pool tissue from rat pups. Approximately 15 3day-old rat pups were required for a single sample, with fewer pups required as they got larger. Two pups were used for a single sample at 21 days. Tissues were analyzed individually for the adult rats and all swine samples. Gel electrophoresis. Tissues were ground using glass homogenizers in isoelectric focusing (IEF) buffer (1% sodium dodecyl sulfate (SDS), (4). They were then sonicated and 10% glycerol, 20 mM dithiothreitol) stored at -2O’C until run on gels. Myosin heavy chains were resolved on SDS-polyacrylamide gels using the methods of Giulian et al. (29). Actin isoforms were resolved using a modification of O’Farrell’s (30) two-dimensional IEF/SDS electrophoresis technique described by Fatigati and Murphy (4) using a 5.0-6.0 pH gradient (Pharmalyte, Sigma Chemical Co., St. Louis, MO). Western blots. Western blots were performed as described previously (6). Antisera specific for the NM MHC (rabbit anti-human platelet myosin) or NM + SM MHC (rabbit anti-bovine uterus myosin) were obtained from Biomedical Technologies, Inc. (Stoughton, MA). SMspecific antisera were generated in chickens using purified rat smooth muscle myosin as the antigen. Densitometty. Proteins were stained using standard procedures with Coomassie brilliant blue R. Gels were scanned using a Quick Scan Jr. densitometer with custom-modified high-resolution optics (Helena Laboratories, Beaumont, TX). Peak areas were determined by integration of these scans. Loading curves were performed to verify linearity of loading volume with integrated peak area. Statistics. The relative amounts of protein isoforms were determined for each sample. Multiple lanes of a given sample were averaged prior to determining the mean and standard errors for different samples of each isoform. Variability between repeat lanes and gels was usually l3% and never more than 5%. A Student t test was used to compare values between young and adult swine. A Newman-Keuls multiple range test was used to compare the data of various rat age groups. Significant differences are reported at P < 0.05.
CHAIN
ISOFORM
233
EXPRESSION
dothelial cells were noted on the luminal surface of the samples followed by several layers of alternating smooth muscle cells and elastic lamena and the adventitial layer. No adventitial layer was present in the stripped tissues. The effect of stripping on fractional isoform composition was more significant for actin than for MHC. For all age groups, the relative cu-SM-actin content was 5% greater and the /‘3-NM-actin content was 5% lower when the tissue samples were stripped of their adventitia prior to processing than when they were not. Separation of the three MHC isoforms by gel electrophoresis was verified by Western blotting (Fig. 1). As we reported previously (5, 6) the upper two bands were recognized by smooth muscle antisera, while the lower band was recognized by nonmuscle antisera. These gels completely resolve the three MHC isoforms so that quantitation may be done on Commassie blue-stained gels. Typical gel results for actin and MHC isoforms from rat aorta for each age group are shown in Figs. 2 and 3. The two-dimensional gel system does not resolve r-NMactin from y-SM-actin. The changes with age in the fractional content of the actin and myosin isoforms are shown in Fig. 4. The r-SM + NM-actin isoforms decreased from 19 to 11% from 3 days to adult. The cr-SM-actin increased from 40 to 64% as P-SM-actin decreased from 40 to 25% over this age range (Fig. 4). The differences between the 3- or 6-day-old animals and all the older animals were significant. Changes were essentially complete by 21 days of age in the rat aorta. The NM MHC was present at the greatest levels in the young animals and decreased with development (Fig. 4, lower panel). This approximately 25% decrease in NM MHC (greater than 30% at 3 and 6 days to less than 10% in the adult) was accounted for by the SMl MHC, which showed a 25% increase between 3 days and adult. All changes between the age groups were significant (with
SDS Gel
Western SM
NM
Blots SM + NM
RESULTS
The contribution of actin and/or MHC isoforms from adventitial fibroblasts or endothelial cells to estimates of aortic smooth muscle cells was minimized by stripping the media from the adventitia. This procedure also caused abrasion of the intimal surface, removing most of the endothelial cells. These “stripped” medial preparations and the remaining adventitia were run on the appropriate gels along with unstripped samples. Young (3-day) and adult unstripped rat aortas that had been cleaned of loose fatty connective tissue and those which had been stripped of their adventitia were fixed, sectioned, stained, and examined by transmission electron microscopy. Some en-
-+*- x FIG. 1. High-molecular-weight region of Coomassie blue-stained 8% SDS-polyacrylamide gel and Western blots of the same region of identical gels. Two loadings of adult rat aorta are shown for each condition. The CBB-stained gel shows the three MHC isoforms. The Western blots verify that the upper two bands are “smooth muscle” MHC, while the lower band is “nonmuscle” MHC. The SM antiserum was chicken anti-rat stomach myosin. The NM antiserum was rabbit anti-human platelet myosin. The NM + SM antiserum was rabbit anti-bovine uterus myosin.
234
EDDINGER
AND
MURPHY
ACTIN
077
2
MW
MYOSIN HEAVY CHAIN
I 42K
5!35
250
PHFIG. 2. Actin-containing portions of Coomassie blue-stained, twodimensional IEF (X-axis)/SDS (Y-axis) gels showing fractional content of actin isoforms from rat aortic medial preparations at various ages. 01, ol-SM-actin; /3, non-NM-actin; y, y-SM+ y-NM-actin.
0.0-w 0
5
15
10
AGE
the exception of the 3- and lo-day NM MHC values). While SM2 fell in fractional content at I4 and 21 days, there was no significant difference between the values observed for the younger ages (3,6, and 10 days) and the adult animals. Figure 5 shows the relationship of the a-SM- and pNM-actin isoforms to the SM and NM MHC isoforms over this age range. The combined SM MHC fractional composition (SMl and SM2) of 69% in the R-day-old rats increased to 93% in the adult animals. The cr-SM-actin increased from 40 to 64%. These increases in SM protein were countered by the decrease in the NM proteins. NM MHC decreased from 31% at 3 days to 7% in the adult rat, while the P-NM-actin decreased from 40 to 25%. Plotting SM- or NM-actin and MHC on the same graph shows the parallel changes with age for increases in the SM isoforms and the decreases in the NM isoforms (Fig. 5). There were also changes in the MHC isoforms with development in swine. The fractional composition of the
6
ADULT
(DAYS)
FIG. 4. Developmental changes in fractional content of the actin and MHC isoforms in rat aortic media. Values shown are means f SE (when the SE exceeds the symbol size) from densitometric evaluation of Coomassie blue-stained gels as shown in Figs. 2 and 3. Sample numbers for each age group for actin and MHC isoforms, respectively, are: 3 days, N = 10,9; 6 days, N = 6.5; 10 days, N = 5,5; 14 days, N = 7,7; 21 days, N = 10,lO; adult, N = 11,13. Samples were pooled for each N to obtain enough tissue at younger ages.
three MHC isoforms present in three types of smooth muscle at two different ages are shown in Table I. Each tissue exhibited somewhat different changes. However, the general trends were for a fall in NM MHC, in large
l.Or
SM MYOSIN
(Days)
Age 3
20
10
14
21
A NM MYO:
LL
0.0I’ 0
5
10
15
20
ADULT
AGE (LMYS) FIG. 3. High-molecular-weight region of Coomassie SDS-polyacrylamide gels showing fractional content of from rat aortic medial preparations at various ages. A, Resolved are the 196-kDa (NM), 200-kDa (SM2), and MHC variants. Separation variability is due to different
blue-stained 8% myosin isoforms adult female rat. 204-kDa (SMl) gels.
FIG. 5. Changes in fractional content of a-SMand p-NM-actin and SM and NM MHC isoforms present at various ages in rat aortic media. Values are means f SE (where the SE exceeds the symbol size) from densitometric evaluation of Coomassie blue-stained gels. Sample number and pooling are described in Fig. 4.
ACTIN
AND
MYOSIN
HEAVY
CHAIN
TABLE
Fractional
Content
of MHC
Isoforms
Present
ISOFORM
I
in Young (2-3 kg, 5-7 Days) and Adult
Piglet n Carotid Stomach Uterus
SMl
7 8 5
NM
14.4 * 1.4 17.5 * 2.1 11.9 f 1.3
39.4 f 1.7 7.8 k 1.0 35.0 i- 1.1
part reflecting an increase in SM2. Unlike the data from rat aorta, the swine carotid pattern differed for the SM MHC isoforms (Table II). In the young rat aorta the fractional content of SMl was approximately half that of SM2. In the young swine carotid, SMl was approximately three times SM2. In both animals the adult ratio of SMl: SM2 was approximately 1.2. The pattern in the rat bladder differed from that in the rat aorta (Fig. 6; compare with Fig. 4). a-SM- and P-NMactin fractional content showed a significant increase and decrease, respectively, but of much smaller magnitude (7 and 5% respectively vs 25 and 15% in the rat aorta). ySM + NM-actin fractional content was high and showed no changes with age. The NM MHC showed no significant differences between the 3-day and the adult animals. This contrasts with the 25% decrease seen in NM MHC in the rat aorta over these same ages. SMl and SM2 MHC fractional contents were more similar to changes in the swine tissues than to changes in the rat aorta (i.e., SMl started out high and decreased rather than starting low and increasing). SMl showed a decrease (0.66 to 0.58) and SM2 an increase (0.25 to 0.35) in fractional content between 3 days and adult (Fig. 6). The changes in SMl and SM2 relative content at 3 days of age and adult were significant,
kg) Swine
swine
n
SMl
SM2
5 5 5
51.1 2 1.1 51.6 + 5.5 45.4 2 5.3
40.1 + 1.3 44.1 f 5.1 37.1 3~ 4.2
Note. Values shown are means * SE from densitometric evaluation of Coomassie blue-stained age groups are significantly different (P < 0.05) with the exception of SMl in the uterus.
TABLE
(So-100 Adult
SM2
46.2 + 1.6 74.7 + 2.6 53.1 f 1.4
235
EXPRESSION
gels. All isoform
NM
comparisons
but always exceeded the 1.7 SMl:SMS the adult animal.
8.8 f 0.8 4.3 f 0.5 17.5 f 2.3 between
these two
ratio observed in
DISCUSSION
The possibility that protein contributions from cells in the adventitial layer affected the measurements was ruled out by the use of electron microscopy showing complete removal of the adventitia from the stripped medial preparations. Samples from each age group (unstripped and stripped of adventitia) were also run on SDS denaturing gels in addition to the stripped samples. In unstripped aortas the fractional content, on average, was 5% lower
II
Comparison of Fractional Content for the Three MHC Isoforms Present in the Medial Tissue of Swine Carotid and Rat Aorta at Young (6 Days) and Adult Ages (Market Weight Swine and Breeder Female Rats)
2 LL
O5 04 SM2
03
/.fe
t. 0.2
II.
SMl
SM2
NM
0.1 i
Rat aorta B-Day rat pup Breeder female Swine carotid B-Day piglet 80 to loo-kg swine
$-----,i---,
0.0-M
5 13
20.2 -t 1.4 50.3 3L 0.9
44.5 f 1.0 42.4 f 0.9
35.3 + 1.1 7.3 f 0.4
7 5
46.2 k 1.6 51.1 f 1.1
14.4 +- 1.4 40.1 f 1.3
39.4 * 1.7 8.8 2 0.8
Note. Values shown are means * SE. All isoform comparisons these two age groups are significantly different (P < 0.05).
between
3
Age FIG. 6.
Adult
21
(Days)
Changes in fractional content of the actin and MHC isoforms in rat bladder. Values shown are means f SE (when SE exceeds the symbol size) from densitometric evaluation of Coomassie blue-stained gels. Sample numbers for each age group for actin and MHC isoforms, respectively, are: 3 days, N = 5,4; 21 days, N = 5,5; adult, N = 7,7.
236
EDDINGER
for cr-SM-actin and 5% higher for P-NM-actin than in the stripped media by itself. There was no consistent effect on the r-SM + NM-actin. Stripping caused no significant change in the fractional content of the SM or NM MHC. The large actin content and small myosin content of nonmuscle cells present in the adventitia could explain why actin isoform fractional content was affected to a greater extent than MHC. Fibroblasts have a relatively high percentage of P-NM-actin. These results show that the adventitial layer can contribute enough P-NM-actin to significantly increase the relative fraction of this isoform over that present in the medial smooth muscle cells alone. The changes in the actin isoform fractional content with development in rat aorta (Figs. 2 and 4) confirm other reports (21, 22). Small differences in the absolute values may be explained by differences between rat strains and/or incomplete removal of the adventitia. Removal of the adventitia is technically difficult in the young animals. Adventitial and possibly endothelial cell contamination can alter the relative percentages of the isoactins. Occasional endothelial cells were observed in some of the tissues prepared for electron microscopy. However, the total contribution of any given isoform from the remaining endothelial cells to the medial samples was very small. The endothelial cells make up only a small fraction of the aorta, and only a small percentage of these cells remained on the medial samples. In addition, the quantities of actin and myosin present in endothelial cells are much lower than those in smooth muscle cells. The major change in isoform composition from young to old rats occurred between 1 and 3 weeks of age (Fig. 4). Changes in the fractional content of actin and MHC isoforms occurred during a similar time frame and to a similar extent (Fig. 5). This correlation may reflect coordinated regulation of the NM or SM MHC and actin isoform genes. A functional significance for the NM and SM isoforms remains to be established. Young piglets and adult swine were used to compare the MHC isoform fractional content in vascular, gastric, and uterine smooth muscle (Table I). All three tissue types show qualitative changes similar to those observed in the rat aorta (SM vs NM isoforms). The magnitude of these changes varied considerably among the tissues studied. The ratio of SMl to SM2 MHC decreased between the young and adult pigs from 4 to 1.3 in the carotid, from 4.2 to 1.2 in the stomach, and from 4.5 to 1.2 in the uterus. These changes contrast with those in the rat aorta, where the SMl:SM2 MHC ratio increased from 0.5 to 1.2 between young and adult animals. In the rat bladder, the SMl:SMB MHC ratio decreased from 2.7 to 1.7. This was similar to, but of a smaller magnitude than, the change in rabbit bladder. McConnell et al. (23) reported that the ratio of SMl:SMB in rabbit bladder smooth muscle decreased from 3.8 to 0.7 between newborn and adult animals. Mohammad and Sparrow (24) observed a SMl:SM2 ratio decrease from 2.1 to 1.0 in swine trachea smooth
AND
MURPHY
muscle between young (30-60 kg) and adult (90-140 kg) animals. They reported no NM MHC in this tissue. In human bronchus smooth muscle, Mohammad and Sparrow (10) found no change in the SMl:SMB ratio with age (0.69 in both infants and adults), but abundant NM MHC. The trend was for a relatively high neonatal SMl:SMB MHC ratio that decreased with development. The exceptions were the rat aorta, where the ratio increased, and in the human bronchus (24) or chicken gizzard (31), where no changes were found in the SMl:SM2 ratio with development. Both Kuro-o et al. (26) and Borrione et al. (25) reported developmental regulation of MHC in rabbit aorta. This was another unique developmental pattern in that there was apparently no SM2 MHC in the embryonic and early neonatal animals. SM2 appears later in development with a concomitant decrease in NM MHC. In the adult there was no NM MHC and the SMl:SMS ratio was approximately 1.8 (25, 26). These differences in SMl and SM2 expression with age were species-specific, as the rat aorta and bladder did not show the same changes. All of the swine tissues studied showed similar directional changes in the SMl:SM2 ratio with development. With the exception of the rat bladder, all of the tissues we examined showed a decrease in NM MHC with development. This NM:SM MHC change parallels the NM:SM actin isoform shift in time and magnitude. The coordinated regulation of these changes would be most readily controlled at the gene level. Unlike the reports for rabbit aorta (25, 26), the NM MHC was also always present over the age range examined. The physiological significance of this NM to SM isoform shift is unknown. There may be several domains in the cell where the NM and SM isoforms function. This has been suggested for the thin filaments (32, 33). More recently, Drew (34) used immunoelectron microscopy of native thin filaments from swine stomach to show that the actin isoforms were not segregated in different thin filaments or along the thin filaments. These results argue strongly against separate populations of NM-actinand SM-actin-containing thin filaments (although heterogeneity might occur due to the differential presence of other thin filament constituents). Drew (34) suggested that the different isoforms may reflect gene regulation without functional implications. By contrast, DeNofrio et al. (35) reported that NM actin was present in membrane ruffles, pseudopods, and stress fibers of pericytes, while SM-actin was localized in stress fibers. These results may differ because of the cell types examined, but need further study. Separate mRNAs for SMl and SM2 have been identified and sequenced (19, 20). The native myosin composition and the role of NM MHC are still unresolved. Sartore et al. (8) reported a differential distribution of SM MHC isoforms among cells and differential immunolocalization of SMl and SM2 within the same cell. Segregation may allow for distinct regulatory and/or
ACTIN
AND
MYOSIN
HEAVY
functional roles. The NM variants may be involved in proliferative, synthetic, and secretory functions, while the SM isoforms underlie contraction. There is also evidence now for two different NM MHC genes and mRNAs (36, 37). Both of the mRNAs were detected in a variety of tissues, including smooth muscle cells. The quantity of mRNA present varied between tissues,but did not appear to change with development. The two NM MHC proteins have not been resolved to date on SDS gels, and it remains unclear if both of these NM MHC are expressed in aortic smooth muscle. This paper confirms earlier reports of changes in contractile protein isoform composition with development. In addition, it provides an explanation for the diversity of results reported in the literature. Species, tissue, and age differences in experimental animals can account for the enormous differences in contractile protein isoform expression reported in the literature. This study also suggests that there is a coordinated regulation of the contractile protein actin and myosin isoforms. ACKNOWLEDGMENTS
CHAIN
ISOFORM
10. Mohammad,
M. A., and Sparrow,
Purdy-Ramos, E. and photography, Swine tissue was Research support HL07528.
Van Neil, and M. Forbes for and Kathy Dobbins for her donated by Smithfield Packwas provided by NIH Grants
11. Mougios,
V., and Barany,
12. Sobieszek, 13. Helper,
J. I., and Gibson,
2. Vandekerckhove,
W. (1976)
J., and Weber,
Cell 9, 7933805.
K. (1978)
J. Mol.
Biol.
126, 783-
3. Vandekerckhove, J., and Weber, K. (1979) Dif-ferentiation 14,123133. 4. Fatigati, V., and Murphy, R. A. (1984) J. Biol. Chem. 259, 14,38314,388. 5. Eddinger, T. J., Gaylinn, B. D., Monical, P. L., and Murphy, R. A. Biophys.
J. 53,
579a.
F., Janmot,
7. Sartore, Saggin, 179,
S., De Marzo, N., Borrione, L., Fabbri, L., and Schiafhno,
Cell
P. A., Monical,
P. L.,
J. Physiol.
C997-
257,
A. C., Zanellato, A. M. C., S. (1989) Eur. J. Biochem.
79-85.
8. Sartore, napieco,
S. Borrione, A. C., Zanellato, A. M. C., Tonello, M., ScanG., Pauletto, P., and Dal Palu, C. (198913) J. Muscle Res. Motil. 10, 167-168.
9. Martin,
A. F. (1988)
Biophys.
J. 53,
577a.
Biophys.
Acta
872,
7, 417-425. J. Biol. Chem.
Electrophoresis
D. R. (1988)
C., Ropert,
S., and d’Albis,
A. (1986)
Eur.
J.
160,507-513.
Biochem.
15. Hasegawa,
Y., Ueno,
H., Horie,
K., and Morita,
F. (1988)
J. Biochem.
103,15-18. 16. Taubman,
M. B., Grant,
J. W., and Nadal-Ginard,
B. (1987)
J. Cell
104,1505-1513.
Biol.
17. Kumar, C. C., Mohan, S. R., Zavodny, P. J., Narula, S. K., and Leibowitz, P. J. (1989) Biochemistry 28,4027-4035. 18. Nagai, R., Larson, D. M., and Periasamy, M. (1988) Proc. Natl. Acad. Sci. USA 85, 104771051. 19. Nagai, R., Kuro-o, M., Babij, P., and Periasamy, M. (1989) J. Biol. Chem.
264,9734-9737.
20. Babij, P., and Periasamy, M. (1989) J. Mol. Biol. 210, 673-679. 21. Kocher, O., Skalli, O., Cerutti, D., Gabbiani, F., and Gabbiani, G. (1985) Circ. Res. 56, 829-838. 22. Owens, G. K., and Thompson, M. M. (1986) J. Biol. Chem. 261, 13,373-13,380. J. D., Ewalt,
D. H., and Boone,
T. B. (1988)
J. Cell
Biol.
107,36a. 24. Mohammad, M. A., and Sparrow, M. P. (1988) FEBS Lett. 228, 1099112. 25. Borrione, A. C., Zanellato, A. M. C., Scannapieco, G., Pauletto, P., and Sartore, S. (1989) Eur. J. Biochem. 183, 413-417. M., Nagai, A., Takaku,
R., Tsuchimochi, F. (1989) J. Biol.
H., and Daly,
28. Purdy-Ramos,
M. M.
(1970)
H.,
Katoh,
Chem.
Proc.
264,
H., Yazaki, Y., 18,272-18,275.
Sot.
Exp.
Biol.
Med.
S., and
Forbes,
M.
S. (1984)
J. Electron
Microsc.
1,271&277.
29. Giulian,
G. G., Moss,
R. L., and Greaser,
M. L. (1983)
Anal.
Biochem.
129,277-287. 30. O’Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021. 31. Takano-Ohmuro, H., Obinata, T., Mikawa, T., and Masaki, J. Biochem.
32. Fiirst,
T. J., Martino, R. A. (1989) Amer.
Biochim.
263, 15,748-15,753. 14. Cavaille,
Tech.
(1988)
J. 260,
135,364-368.
802.
6. Gaylinn, B. D., Eddinger, Hunt, D. F., and Murphy, c1004.
M. (1986)
A., and Jertschin, P. (1986) D. J., Lash, J. A., and Hathaway,
27. Wolinsky, 1. Garrels,
Biochem.
305c308.
26. Kuro-o, Ohkubo,
REFERENCES
M. P. (1989)
421-426.
23. McConnell,
The authors thank S. the electron microscopy word processing prowess. ing Co. (Smithfield, VA). 2POl HL19242 and lF32
237
EXPRESSION
93,
T. (1983)
903-908.
D. 0.. Cross, R. A., De Mey,
J., and Small,
J. V. (1986)
EMBO
J. 5, 251-257.
33. Small, J. V., Fiirst, D. O., and De Mey, J. (1986) J. Cell Biol. 102, 210-220. 34. Drew, J. S. (1989) Ph.D. dissertation, University of Virginia, Charlottesville, University Microfilms, Ann Arbor, MI. 35. DeNofrio, D., Hoock, R. C., and Herman, I. M. (1989) J. Cell Biol.
109,191-202. 36. Kawamoto, 37. Katsuragawa, Eur.
S., and Adelstein, Y., Yanagisawa, J. Biochem. 184,611-616.
R. S. (1989) J. Cell Biol. M., Inoue, A., and Masaki,
109, 275a. T. (1989)