Murine pulmonary myocardium: Developmental analysis of cardiac gene expression

DEVELOPMENTAL DYNAMICS 2001 17-128 (1994)

Murine Pulmonary Myocardium: Developmental Analysis of Cardiac Gene Expression W. KEITH JONES, ALEJANDRO SANCHEZ, AND JEFFREY ROBBINS Department of Pediatrics, Division of Cardiovascular Molecular Biology, Children’s Hospital Research Foundation, Cincinnati, Ohio 45229-3039

ABSTRACT Long-standing observations that cardiac muscle exists in the walls of the pulmonary and caval veins have recently been confirmed at the molecular level (Lyons et al. [1990] J. Cell Biol. 111:2427-2436; Springall et al. [19881 Thorax 43:44-52; Subramaniam et al. [19911 J. Biol. Chem. 266:24613-24620). Using ventricle- and atrial-specific riboprobes, we determined that the pulmonary myocardium exhibits an atrial pattern of cardiac-specific gene expression. Additionally, the developmental pattern of expression was studied using a riboprobe specific to the a-cardiac myosin heavy chain (a-MHC) gene transcript. We find that a-MHC gene expression is first detectable in the lung between 13.9-14.3 days post-coitum. Extension of the a-MHC specific hybridization signal into the pulmonary venous bed progresses through the neonatal period. The data are consistent with the hypothesis that the extension of a-MHC gene expression into the lung occurs via the migration of atrial myoblasts into the vein during atrial septation and remodeling of the sinus venosus and pulmonary venous trunk. 0 1994 Wiley-Liss, Inc.

Key words: Myosin, Pulmonary development, Mouse, Pulmonary myocardium, Car- diac gene expression

INTRODUCTION Striated muscle was first described in human pulmo-

nary and caval veins over 150 years ago (Rauschel, 1836). This tissue, named “pulmonary myocardium” (pm) by Favaro (Favaro, 1910), has been observed in numerous species including the shrew, mouse, rat, hamster, guinea pig, dog, deer, human, and bird (Kramer and Marks, 1965; Klika, 1976; Nathan and Gloobe, 1970; Endo et al., 1992a,b). Ultrastructural analyses showed that this muscle consists of striated mono- and bi-nucleate cells having narrow I-bands and intercalated discs and the sarcomere and single cell morphology are indistinguishable from cardiomyocytes (Karrer, 1959a,b; Ludatscher, 1968).

The pm is supplied with both adrenergic and cholin- ergic nerves (deAlmeida et al., 1975), and responds to electrical and pharmacological stimuli in a manner similar to atrial muscle (MacLeod and Hunter, 1966; deAlmeida et al., 1975). Cheung demonstrated that the

0 1994 WILEY-LISS, INC.

pm in the guinea pig is electrically active and normally is dominated by the pace making activity of the SA node (Cheung, 1980). This is consistent with earlier work which showed that the pm of dogs constitutes a functional syncytium with the atria (Spach et al., 1972; deAlmeida et al., 1975). Toshimori et al. and Asai et al. (1987) determined that the active form of atrial natriuretic peptide (ANP) in the lung is localized to atrial- like storage granules within the striated muscle of the pulmonary veins in the rat, pig, and ox.

Characterization of the murine pm has begun at the molecular level. Springall et al. showed that the atrial natriuretic peptide gene mRNA and immunoreactive peptide are localized to the striated musculature of extrapulmonary and intrapulmonary veins of the rat (Springall et al., 1988). Lyons and colleagues demonstrated that the a-myosin heavy chain (a-MHC) and myosin light chain 1A (MLClA) genes are expressed in the intrapulmonary veins of 15.5 day post-coitum (p.c.1 murine embryos (Lyons et al., 1990). Subramaniam et al. showed that the a-MHC gene is expressed in the caval veins and small pulmonary venules in the adult, and that the a-MHC promoter is able to direct expression of a chloramphenicol acetyl transferase transgene in the pm (Subramaniam et al., 1991).

Developmental analyses of the pm using electron microscopy, revealed the occurrence of mature striated muscle in the larger pulmonary veins of the mouse at day 19 p.c. (Klika, 1976). Histological analyses of earlier embryonic stages revealed discontinuous groups of myoblast-like cells (non-striated) near the presumptive hylus of 14 day p.c. embryos. These data are consistent with either migration of atrial myoblasts from the atrium or with the differentiation of cardiac myoblasts within the developing veins from progenitors pre-exist- ing in the splanchnopleuric mesoderm.

In order to determine if the pm behaves a t the molecular level like the ventricle or atrium, we undertook a transcriptional analysis of the pm during development. Both atrial- and ventricle-specific riboprobes were used to characterize the transcriptional pattern of expression. Additionally, we monitored the early development of the pm using in situ analyses. The data are

Received December 10, 1993; accepted March 2, 1994. Address reprint requests to Dr. Jeffrey Robbins, Department of Pe-

diatrics, Division of Cardiovascular Molecular Biology, Children’s Hospital Research Foundation, Cincinnati, OH 45229-3039.

118 JONES ET AL

consistent with atrial myoblast migration into the developing pulmonary vasculature during fetal development.

RESULTS Developmental Analyses of the Murine Pulmonary Myocardium

Developmental embryologists have described the process by which the pulmonary venous network is formed. The main branch of the vein grows from the atrium beginning around day 11 P.c., and eventually anastomoses with the pulmonary venous plexus within the developing lung (Rugh et al., 1990; Larsen et al., 1993). Atrial septation occurs from day 10 p.c. to shortly after birth, concomitant with the remodeling of the sinus venosus. The timing of the remodeling of the atria, the sinus venosus, and the growth of the pulmonary vein has complicated studies of the origin of the vein.

Because the a-MHC transcript is expressed in the heart tube even as i t is formed (Lyons et al., 1990; Klika, 1976), we chose a-MHC gene expression as a marker for a developmental study of the pm. Initial experiments sought to determine the developmental timing of a-MHC gene expression in the pm. Staged embryos were collected from FVBiN mice and subjected to in situ hybridization with the a-MHC riboprobe (Su- bramaniam et al., 1991).

Parasagittal sections of 13.3 day p.c. embryos demonstrate strong a-MHC specific hybridization in the atrium (Fig. 1A). No hybridization signal is detectable in the lung though both pulmonary arteries and veins are apparent. Weak hybridization in the atrial-proximal portion of the pulmonary vein, exter- nal to the lung, was noted (Fig. 1A). Bright field photomicrographs (Fig. 1B,D) revealed that apparent signal in the lumen of the atrium and pulmonary arteries is due to sequestration of probe by red blood cells; this is distinguishable from signal over the atrial wall such as in the top left portion of the atrium in Figure 1A,B. We observe that the pulmonary vein originates from the region of the left atrium proper and not from the sinus venosus. The hybridization signal in the atrium is weakest in the dorsal-most aspect and is undetectable in the region of the sinus venosus. In our laboratory, under the experimental conditions used, the corresponding sense strand a-MHC riboprobe did not produce a signal in any of the tissues studied (e.g., Subramaniam et al., 1991; Sanchez et al., 1991).

Within a 24-hr period, there is rapid extension of a-MHC hybridization signal past the hylus of the lung bud and into the intrapulmonary veins (Fig. 2). In 13.9 day p.c. embryos, the a-MHC hybridization signal within the pulmonary vein is strongest proximal to the atrium, and falls to background levels near the entry point of the vein into the lung. By 14.3 days P.c., the veins are more strongly positive for a-MHC hybridiza-

tion signal and the signal extends farther along their length than previously (Fig. 2Ci.

Also between 13.9 and 14.3 days px., the sinus venosus becomes positive for a-MHC specific-hybridization (Figs. 2A, 3A,B). At the earlier timepoint, the sinus venosus, inferior vena cava (at this stage called the posterior cardinal vein), bulbus arteriosus, and cardiac outflow tract are negative for a-MHC hybridization (Fig. 2A). By 14.3 days P.c., the lung bud has expanded, and is more obviously multi-lobed (Fig. 2C,D). Pulmo- nary arteries visible in sections adjacent to that shown in Figure 2C are negative for a-MHC specific hybridization. The spatial relationship between the sinus, the atrium, and the venous system of the embryo and the adult are shown in panel C of Figure 3.

Analyses of later developmental stages reveal that the a-MHC hybridization signal progressively extends to the more distal veins within the lung (Fig. 4A,C). By 2 days post-birth, the a-MHC hybridization signal extends throughout a large distal portion of the lung, and is as intense in the small veins as in the heart itself (Fig. 4C). The proximal vena cava shows strong a-MHC specific hybridization a t this stage as well.

The results of this developmental analysis show that the a-MHC transcript is detectable in the pm throughout embryonic and fetal development; a pattern consistent with expression in the murine atrium. Although the a-MHC gene is constitutively expressed in the atrium during all stages of development, it is transcrip- tionally activated in the ventricle a t birth, replacing the 6-MHC transcript as the major transcript in this compartment during embryonic and fetal development (Lyons et al., 1990; Lompre et al., 1981, 1984; Ng et al., 1991). Therefore, it was of interest to investigate the embryonic pattern of p-MHC gene expression in the Pm.

The a-MHC and P-MHC riboprobes were hybridized to cross-sections of 15.5 day p.c. embryos (Fig. 4A,B). Although there is considerable p-MHC transcript in the ventricle a t this stage (Ng et al., 1991), this gene is not expressed a t detectable levels in the pm (Fig. 4B), while the a-MHC is (Fig. 4A). Thus MHC gene expression in the developing pm displays an atrial-like pattern. We noted significant a-MHC specific hybridization in the ventricle at 15.5 days p.c. This has been previously documented a t similar timepoints (Lyons et al., 19901, and there is quantitative evidence for transcriptional up-regulation of the a-MHC gene late in fetal life, prior to its maximal activation post-birth (Ng et al., 1991).

Cardiac Compartment-Specific Gene Expression in the Adult Pulmonary Myocardium

We sought to further define the phenotype of gene expression characteristic of the pulmonary cardiomyocytes relative to that of atrial and ventricular cardiomyocytes. The transcriptional switch which occurs a t birth in murine ventricle is regulated primarily by thy-

PULMONARY MYOCARDIAL DEVELOPMENT 119

Fig. 1. The wMHC gene transcript is not detected in 13.3 day p.c. embryonic lung. Parasagittal sections through a 13.3 day p.c. FVB/N embryo. The sections shown in (A) and (C) were hybridized with the a-MHC riboprobe. The atrial wall is positive for a-MHC hybridization signal (A). The white arrow in panel A points out a portion of the pulmonary vein proximal to the atrium. The white arrow in panel C points out the

pulmonary vein in a section adjacent to that shown in panel A. (6) and (D) are phase-contrast photographs of the same sections shown in panels A and C. The black arrows point out the same features as the white arrows in panels A and C. a, atrium; Ib, lung bud; sv, sinus venosus. Magnifica- tion is the same for panels A-D; the black bar in panel D is 300 km in length.

roxin, which increases expression of the a-MHC gene, and decreases that of the P-MHC gene (Mahdavi et al., 1986; Hoh et al., 1978; Ng et al., 1991; Lompre et al., 1984). In the small rodent, induced hypothyroidism results in a transcriptional switch in the adult ventricle, causing a-MHC transcription to decrease and P-MHC transcription to increase (Mahdavi et al., 1986; Hoh et al., 1978; Ng et al., 1991; Lompre et al., 1984; Subramaniam et al., 1991). The expression of the a-MHC gene in the atrium is relatively unaffected by hypothyroidism (Subramaniam et al., 1991; Ng et al., 1991 j.

Experiments were carried out to determine the effect of induced hypothyroidism upon a-MHC and p-MHC transcription in the pm. Adult mice were made hypothyroid by treatment with 5-propyl-2-thiouracil (PTU), and portions of the heart and lungs processed for in situ hybridization (Fig. 5). As expected, hypothyroidism results in significant down-regulation of a-MHC transcription in the ventricle; only an ex- tremely weak hybridization signal is apparent (Fig.

5B). The expected ventricular a-MHC -tp-MHC switch occurs, and while the p-MHC hybridization signal is intense in the hypothyroid ventricle, it is undetectable in the atrium (Fig. 5A). Notably, no P-MHC hybridization signal was apparent in the lung (Fig. 5C). In the hypothyroid animal, the a-MHC hybridization signal remains intense in the pm (Fig. 5B); thus, we conclude that the pm displays an atrial-like response to the hypothyroid state.

A set of cardiac and cardiac compartment-specific riboprobes were made in order to extend the analysis of cardiac gene transcription in the pm (see Experimental Procedures), The a-MHC and P-MHC riboprobes have previously been described (Sanchez et al., 1991). Post- birth in the mouse, the a-MHC gene is expressed in both the atria and ventricle, and the P-MHC gene is not expressed at detectable levels in either cardiac compartment (Lyons et al., 1990; Ng et al., 1991). An atrial natriuretic factor (ANF) riboprobe was obtained (see Experimental Procedures); ANF gene expression occurs predominantly in the atrial cardiac compartment

120 JONES ET AL.

Fig. 2. a-MHC-specific hybridization extends into the embryonic lung early on day 14 p.c. (A), (B) are dark-field and bright-field (hematoxylin stained) photographs of adjacent mid-sagittal sections through a 13.9 day p c embryo. (C), (D) are dark-field and bright-field (hematoxylin stained) photographs of a sagittal section through a 14.3 day p.c. embryo. The sections were hybridized with the a-MHC riboprobe. The apparent signal in the dorsal aspect of the inferior vena cava in panel A is due to seques-

tration of probe by blood cells in the region. The white arrow in panels A and C points out the pulmonary vein near the presumptive hylus of the lung. a, atrium; ba, bulbus arteriosus; ivc, inferior vena cava (at this stage also called the posterior cardinal vein); lb, lung bud; Ii, liver; sv, sinus venosus; v, ventricle. Magnification is the same for panels A-D; the black bar in panel D is 500 pm in length.

post-birth (Zivin et al., 1984). A riboprobe specific for the ventricle-specific myosin light chain gene (MLC2; Lee et al., 1992) was also prepared (see Experimental Procedures). ANF, a-MHC, p-MHC, and MLCB probes were hybridized to cross-sections of 2-day neonate mice and lung isolated from adult (8-week-old) animals. In the neonate (Fig. 6), a-MHC riboprobe hybridizes to the atrium, ventricle and pm equally well. No hybridization is seen within the walls of the pulmonary arteries. Bright field microscopy revealed the high background within the lumen of the pulmonary arteries results from sequestration of the probe by red blood cells which failed to wash out of the vessels. The MLCB isoform detected by the riboprobe is expressed strongly in the ventricle but is not present in the atrium or in the pm

(Fig. 6B). ANF is expressed strongly in the atrium and in the pm, and to a lesser extent, in the ventricles (Fig. 6C). Finally, p-MHC gene expression is not detected in the neonate heart, or in the pm (Fig. 6D). The analysis of adult lung yielded similar results, a-MHC and ANF specific hybridization signals were detected within the tunica media of the pulmonary veins (Fig. 7A,C), while MLC2 specific hybridization signal was not above the background signal within the atria (Fig. 7B). Staining of parallel sections with phosphotungstic acid-hematoxylin revealed that the a-MHC positive tissue is striated muscle (Fig. 7D). These data are consistent with the developmental analyses (Figs. 2,3); pm exhibits an atrial pattern of transcription throughout development.


C pulmonary vein

of sinus venosus

region of atrial septum

stem of presumptive vena

transverse sinus

left horn

left ventricle

Fig. 3. On day 14.3 P.c., a-MHC transcript is detected in some portions of the sinus venosus. (A) and (B) show parasagittal sections from the same 14.3 day p.c. embryo shown in Figure 2C,D. The sections were hybridized with the a-MHC riboprobe. The section in panel A is the most exterior in the series, with that of panel B being approximately 150 p,m closer to the embryo's midline. The section shown in the previous figure (Fig. 2C,D) is located between the two. White arrows in panels A and B indicate the positions of the a-MHC positive regions of the sinus venosus in these sections. These areas correspond to the ventral and inferior portions of the sinus venosus. (C) shows the spatial relationships be-

DISCUSSION All the data presented indicate that the pm is com-

prised of atrial cardiac muscle. Examination of phosphotungstic acid-hematoxylin stained sections (Fig. 7D) reveal that the a-MHC positive muscle layer is located within the tunica media of the pulmonary vein (separated from the venous lumen by a layer of epithe- lial cells), and is striated muscle. These results agree with and extend previous structural (Klika, 1976; Kar- rer, 1959a,b; Ludatscher, 1968) and molecular analyses (Lyons et al., 1990; Springall et al., 1988; Subrama- niam et al., 1991).

cava

tween the embryonic veins (1, 2, and 3), the sinus venosus, the atria, and the adult veins. The heart is depicted in dorsal view. The diagram is not meant to represent any one stage of embryonic development. The structures indicated by dashed lines are the early embryonic veins; 1 , the cardinal veins; 2, the umbilical veins; 3, the vitelline veins. The pulmonary vein is indicated (the darkly shaded region is the portion which is ab- sorbed into the left atrium). The position of the atrial septum is indicated by a dashed line. a, atrium; Ib, lung bud; li, liver; sv, sinus venosus; v, ventricle. Magnification in panels A and B are the same; the white bar in panel B is 500 pn in length.

The data suggest that the atrium, and not the sinus is the embryologic origin of the pulmonary vein. The consensus of previous studies is that the vein derives from the sinu-atrial region prior to the time it is distinguishable from the sinus venosus (Auer, 1948; Dor et al., 1987). Our work demonstrates that the pulmonary vein extends from the atrium in a region of the sinu-atrial transition which is positive for a-MHC hybridization (Figs. 1, 2A,B). At this stage we find that the sinus venosus is negative for a-MHC gene expression, while the atrium is strongly positive (Fig. 1C).

The main branch of the pulmonary vein originates as

122 JONES ET AL.

Fig 4 Progressive expansion of a-MHC-specific hybridization signal within the pulmonary venous bed during development (A) and (8) show dark-field photographs of transverse sections through a 15 5 day embryo, hybridized with the a-MHC and 6-MHC specific riboprobes respectively The white arrow in panel A points out a section through a small pulmonary vein distal from the pulmonary hylus The white arrow in panel B indicates an artery within the lung The apparent hybridization is trapping

an outgrowth of the atrium around day 11 P.c., and undergoes anastomosis with the pulmonary venous plexus sometime before day 13.3 P.c.; a t this time we observe pulmonary veins extending deep within the lung buds. We detect no a-MHC specific gene expression within embryonic lung until 14.3 days P.c., after the actual extension of the pulmonary vein into the lung. Since the expression of a-MHC transcript is an early characteristic of developing cardiomyocytes (Ly- ons et al., 1990; Sanchez et al., 1991), we conclude that this cell type does not exist in the intrapulmonary venous tract until this time. No hybridization was detected with cardiac specific probes within the pulmonary arterial wall or cardiac outflow tract a t any of the developmental stages examined (Figs. 2, 4, 6 ) .

It appears that the pm infiltrates the lung late on embryonic day 13 or early on day 14 px. There is a measurable extension of the a-MHC hybridization signal within the main pulmonary vein past the hylus between 13.9 and 14.3 days p.c. (Fig. 2A,C). By day 5 post-birth, a-MHC specific hybridization extends to small venules 50 pm in diameter (Fig. 4 and Subrama- niam et al., 1991). Figure 8 summarizes our observations of pulmonary myocardial expansion. At later stages, the a-MHC positive tissue is observed in more distal positions within the pulmonary venous bed. This suggests a progressive migration of atrial cardiomyocytes into the branching network of pulmonary veins.

Cardiac-like muscle has been reported within the hepatic portal vein based upon ultrastructural observations (Yokota and Yamauchi, 1985). These cells were

of probe by red blood cells within the lumen of the vessel. No p-MHC specific hybridization is detected within the lung. (C) shows a transverse section through the cardiac region of 2-day post-birth neonate, hybridized to the a-MHC riboprobe and subjected to dark-field photomicroscopy. a, atrium; ao, aorta; la, left atrium; Iu, lung; Iv, left ventricle; pv, pulmonary vein; ra, right atrium; rv, right ventricle; vc, vena cava. The white bars in panels A, B, and C measure 225 pn, 300 pm, and 640 pm, respectively.

observed only in the hepatic end of the vein, as irreg- ular patches of cells within the adventitia andlor tunica media of the venous wall. However, only 5 of 13 mice and 2 of 4 rats exhibited this striated muscle. This differs considerably from the situation in the large pulmonary and caval veins, where the cardiac muscle is a continuous layer within the tunica media of the venous wall in all animals examined (Karrer, 1959a,b; Lu- datscher, 1968; this study).

Regions of the sinus venosus which are inferior and directly adjacent to the atria become positive for a-MHC hybridization by day 14.3 p.c. (Fig. 3A,B). Be- tween 12.5-16 days px., the sinus venosus (gray shaded area of Fig. 3C) undergoes a process of intus- susception, shifts to the right, and is incorporated into the right atrial wall. Clearly this process involves the spread of a-MHC gene expression into the portion of the atrial wall which once was a-MHC negative sinus venosus. Thus we conclude that a-MHC positive myocardium invades the sinus venosus prior to its absorp- tion by the right atrium and shortly after extension into the pulmonary vein has begun.

The caval veins are positive for a-MHC specific hybridization (Subramaniam et al., 1991; Lyons et al., 1990). Beginning on days 12.5-13.5 P.c., the caval veins develop from the remnants of the cardinal veins (embryonic vein 1 in Fig. 8C) at the junction with the right horn of the sinus venosus. The presumptive caval veins empty into the right sinus horn and, eventually, the incorporation of the sinus into the right atrium results in the two caval veins opening directly into the


Fig. 5. Transcription of the a-MHC gene in the pm is unaffected by experimentally induced hypothyroidism. (A) shows a section through the isolated heart of an adult mouse made hypothyroid by experimental ma- nipulation (Subramaniam et al., 1991). The section was hybridized with the p-MHC riboprobe and viewed by dark-field microscopy. As expected, a strong hybridization signal occurs in the ventricle. (C) shows a transverse section through the lung of the same animal. The tissue was probed with the p-MHC riboprobe, and viewed by dark-field microscopy. No hybridization signal is seen within the lung. The white arrow points to the same pulmonary vein indicated in panel D. (B) shows a section Io-

right atrium (Fig. 8C). Similarly, the atrial-proximal portion of the pulmonary veins undergoes limited incorporation into the wall of the left atrium (shaded portion of pulmonary vein in Fig. 8C). Thus the caval and distal pulmonary veins bear a similar relationship to the veins which contribute to the atrial wall.

The data are consistent with the idea that structures closely associated with the developing atrium are col- onized by atrial cardiomyocytes. This “myocardial expansion” may be associated with the process of sinu-

cated between those depicted in panels A and C, which was hybridized with the a-MHC riboprobe. Though the level of a-MHC specific hybridization in the ventricle is low relative to the level of p-MHC specific hybridization, the a-MHC specific hybridization signal is strong within the pm (panel B, white arrow). (D) is a phase-contrast photomicrograph of the a-MHC-positive structure indicated by the arrow in panel B. The morphology is typical of a large pulmonary vein, having a continuous layer of cardiac muscle (black arrow). a, atrium: v, ventricle. Magnification is the same in panels A-C; the white bar in panel B is 500 pm in length, the black bar in panel D is 170 wm in length.

atrial remodeling and primary atrial septation (Fig. 3C; Larsen et al., 1993; Arey et al., 1954). Such an expansion would presumably be required for the development of myocardium in the portions of the atrial wall derived from the pulmonary vein and sinus venosus. The possibility remains, however, that the pm results from differentiation of myocardium from progenitor cells present in the splanchnopleuric mesoderm which is the common precursor for the cardiac tube and pulmonary vasculature (Larsen, 1993). Careful cell lin-

124 JONES ET AL.

Fig. 6. Cardiac-specific gene expression in 2-day neonate heart and lung. (A-E) show serial sections through the cardiac level of a 2-day FVB/N neonate, hybridized with the following riboprobes and subjected to dark-field photomicroscopy: (A), a-MHC; (B), MLCZ; (C), ANF; (D), p-MHC. The section in panel E is an adjacent section which was stained

with hematoxylin-eosin, and photographed under bright-field conditions. The position of the pulmonary vein is marked with a white arrow in panel A. a, atrium; pa, pulmonary artery; pv, pulmonary vein; v, ventricle. Mag- nification is the same in panels A-€; the white bar in panel € is 940 prn in length.

eage studies will be necessary to determine unequivo- cally the embryonic origin of the pulmonary myocardium. Since the portal vein is derived from a portion of the right vitelline vein, which also drains into the sinus venosus (embryonic vein 3 in Fig. 3C) , it is possible that cardiac muscle observed there (Yokota and Ya- mauchi, 1985) is due to occasional invasion by sinu- atrial cardiomyocytes.

Whether the pm is a developmental by-product or a

functional organ cannot be judged until the question of physiological function is addressed. In this regard, it is interesting to note that in mature fish, the sinus venosus functions as a reservoir for the collection of blood during atrial contraction (Arey et al., 1954). The mam- malian sinus probably plays a similar role during early development (Arey et al., 1954). However, as a result of atrial remodeling, necessary for the development of a four-chambered heart, the sinus is lost. The develop-


Fig. 7. Cardiac-specific gene expression in adult pulmonary myocardium. (A-C) show adjacent transverse sections through intrapulrnonary veins from an 8-week-old FVB/N mouse. The sections shown in panels A-C were hybridized with the following riboprobes and subjected to dark- field photomicroscopy: (A), a-MHC; (B), MLCZ; (C), ANF. The region boxed in white in panel C is shown magnified in (D). The section in panel

ment of muscular veins proximal to the cardiac inflow tracts (i.e., the pulmonary and caval veins), may sub- stitute for the lost sinus function. This is consistent with the hypothesis that the pm functions to pre-load the atrium after contraction (in the absence of a reservoir). However, evidence that pulmonary myocardial excitation propagates from the atrium toward the lung (MacLeod and Hunter, 1966; deAlmeida et al., 1975; Cheung, 1980) is easier to reconcile with the hypothesis that the muscle serves as a pulmonary venous valve. Assigning a function to the pm will have to await experimental evidence for a physiological role of the organ.

The existence of atrial cardiac muscle within the pulmonary and caval venous systems has not generally been recognized as a possible factor in pulmonary hypertension and vascular constriction. Studies of pulmonary vasoconstriction usually focus upon the pressure response to treatment, or monitor smooth muscle contraction and arterial constriction. Particularly the

D was stained with phosphotungstic acid-hernatoxylin, and photographed under bright-field conditions, and shows striated muscle within the wall of the vein lpv, lumen of the pulmonary vein Magnification is the same in panels A-C The white bar in panel C is 340 pm in length, the white bar in panel D is 90 +rn in length

studies which utilize mouse and rat need to consider the presence of intrapulmonary cardiac muscle. Inter- estingly, it has been shown that the pm undergoes expansion in mice exposed to chronic altitude hypoxia. These animals subsequently developed right ventricular hypertrophy (Jarkouska and Ostadal, 1983). Addi- tionally, right ventricular hypertrophy is often associated with chronic pulmonary hypertension in humans. The possibility that growth andlor pathology of the pm is involved in human disease such as pulmonary hypertension and altitude sickness should be given some consideration.

EXPERIMENTAL PROCEDURES Collection of Staged Embryos and Treatment of Tissues

Developmentally staged embryos were collected from FVBIN mice (Charles River Laboratories, Inc., Wilm- ington, MA). Female mice were subjected to increased population density in order to suppress estrus 24-72 hr

126 JONES ET AL

C .

Fig. 8. Diagram depicting the stages of cardiac and pulmonary vascular development presented in,Figures 1-4. (A) corresponds to panel A of Figure 2, a 13.9 day p.c. embryonic section (near mid-sagittal). The shaded areas represent the a-MHC hybridization signal, which at this stage is strongest in the atrium. The wMHC hybridization signal is visible in the portion of the pulmonary venous tract proximal to the atrium. The pulmonary veins already extend into the lung, while the a-MHC hybridization signal is restricted to the extrapulmonary vein at this stage. (6) corresponds to panel C of Figure 2, a 14.3 day p.c. embryo (near mid- sagittal section). At this stage, the a-MHC hybridization signal is detected

prior to mating (Scharmann and Wolff, 1980). Sub- sequently, females were split into groups of 4, and exposed to soiled bedding from the cage of a male breeder for 48 hr to induce estrus. Female mice were then added to a male’s cage in late afternoon, and checked for copulatory plugs every 3-6 hr, depending upon the desired breeding window (this being day 0 post-coitum). Successfully bred females were sacri- ficed, the embryos dissected, rinsed with chilled (4°C) sterile RNase-free PBS (80 g NaC1, 2.0 g KC1, 14.4 g Na,HPO, and 2.4 g KH,PO, per liter a t pH = 7.41, and fixed in 4% paraformaldehyde (Electron Micros- copy Science, Fort Washington, PA), 2.5% gluteralde- hyde (Sigma, St. Louis, MO) in PBS overnight a t 4°C.

B

within the intrapulmonary veins. (C) corresponds to panel A of Figure 4, a 15.5 day p.c. embryo, hybridized with the a-MHC specific riboprobe. While this panel is a transverse section, the pulmonary veins are drawn in diagrammatic form to illustrate the progressive appearance, relative to preceding and following stages, of the a-MHC hybridization signal into the pulmonary venous bed. (D) corresponds to panel C of Figure 4, a transverse section through a 2-day neonate. Again, the pulmonary veins are drawn in diagrammatic form to illustrate the extent of the a-MHC specific hybridization signal. At this stage, a-MHC gene expression is observed through a large portion of the pulmonary venous bed.

Neonatal and adult tissues were dissected and fixed as previously described (Subramaniam et al., 1991). Fixed tissues were rinsed with chilled PBS, and cryoprotected in 30% sucrose in PBS overnight a t 4°C. The next day, tissues were rinsed with chilled PBS briefly, and embedded in Tissue-Tek OCT (Miles, Elkhart, IN) a t -80°C. Embryo sections (cut a t -20°C) 5-7 pm thick were attached to 1,3-aminopropyltri- ethoxysilane (Sigma) coated microscope slides and air dried for 2-3 hr a t room temperature.

In Situ Hybridization Post-fixation was carried out as described (Subrama-

niam et al., 1991). Hybridizations were carried out ac-


cording to Cox et al. with modifications (Cox et al., 1984). Briefly, sections were hybridized overnight at 55°C in 4 x SSC (1 x SSC is 150 mM sodium chloride, 15 mM sodium citrate), 10% dextran sulfate, 5 x Den- hardts, 100 +g/ml denatured salmon sperm DNA, 100 pgiml yeast tRNA, 10 mM dithiothreitol, and 1 x lo5 cpm/pl of riboprobe (1 x lo6 cpm per section). For all the probes stringent washes included two post-hybridization washes in 1 x SSC, 50% formamide, 10 mM dithiothreitol a t 65°C for 30 min, with an intervening RNase A treatment. After washing, the slides were de- hydrated, and dipped into Ilford K5 emulsion (Electron Microscopy Sciences), air-dried, and the emulsion exposed in the dark at 4°C for 7-20 days. Developed sections were photographed using the dark-field and phase-contrast optics of an Olympus BHTU microscope.

Slides containing tissue sections were stained with hematoxylin-eosin, or phosphotungstic acid-hematoxylin, which stains the contractile elements of striated muscle (Gude et al., 1982). Slides which had been subjected to in situ hybridization were photographed under dark-field conditions, then rehydrated and stained. In other cases, sections adjacent to those hybridized were available for staining.

PREPARATION OF RIBOPROBES The constructs for the synthesis of the a-MHC and

b-MHC specific riboprobes have been previously described (Sanchez et al., 1991). The a-MHC and b-MHC riboprobes are homologous to the 3’ untranslated portions of the respective transcripts. These sequences are known to be absolutely isoform specific (Robbins et al., 1990). Our previous work with the a-MHC riboprobe showed it to be specific for cardiac myocytes; no hybridization was detected in smooth or skeletal muscle (e.g., Subramaniam et al., 1991; Sanchez et al., 1991). The rat ANF riboprobe was the generous gift of Dr. K. Chien (Department of Medicine, U.C. San Diego, La Jolla, CAI. The antisense riboprobe is 120 base pairs in length and is highly homologous to the murine ANF mRNA (Zivin et al., 1984). The sequence of the antisense ANF riboprobe used is 5‘-GGATCCGTCG ACCTGCAGCT CCAGGAGGGT ATTCACCACC TCT-

CGCAA GGCTTGGGAT CTTTTGCGAT CTGCTC- GAGC AGATTTGGCT GTTATCTTCG GTA-3’. The MLCZ riboprobe construct was a subcloned PCR product made using 10-day embryoid body RNA as tem- plate, and specific primers synthesized using known sequence data for the ventricle-specific isoform (Lee et al., 1992). The sequence of the MLC2 riboprobe (sense strand) is 5‘-GGGCTGATCC TGAAGAGACC AT TC TCA ACG CAT TCA AGG T G TT TGATCCC GAGGGCAAAG GGTCACTGAA GGCTGACTAT GTCCGGGAGA TGCTGACCAC ACAAGCAGAG AGGTTCTCCA AAGAGGAGAT CGACCAGATG TTCGCAGCCT TTCCCCCTGA CGTTACCGGC A AT CTTGATT ATA AGA ATTT GGTCCACATC

CAGTGGC AATGCGACCA AGCTGTGTGA CACAC-

ATTACCCACG GAGAAGAGAA GGACTGAGCC CTGAACCACA GCCTCAGGTG ACCCACAGCC CACTCTCCAT CCCS’. Each riboprobe construct was confirmed by DNA sequence analysis, and the specific- ity of all riboprobes was demonstrated by the ability of [32Pl UTP labeled cRNAs to hybridize to single frag- ments on a Southern blot containing murine genomic DNA (data not shown).

Preparation of riboprobes for in situ hybridization was carried out with [35S] UTP or [35S] GTP by stan- dard protocols from linearized templates (Melton et al., 1984). The cRNA products were purified and concen- trated as previously described (Subramaniam et al., 1991).

ACKNOWLEDGMENTS This work is supported by American Heart Associa-

tion Grant SW-92-13-1, and by National Institutes of Health Grant HL 41496. The authors thank Joseph Palermo and Marquita Neyland Allen for technical as- sistance with the cloning and sequencing of the MLC 2 sequences, and Dr. Jeffrey Whitsett for a critical read- ing of the manuscript.

REFERENCES Arey, L.B. (1954) “Developmental Anatomy.” Philadelphia: W.B.

Saunders. Asai, J., Nakazato, M., Toshimori, H., Matsukura, S., Kangawa, K.,

and Matsuo, H. (1987) Presence of atrial natriuretic polypeptide in the pulmonary vein and vena cava. Biochem. Biophys. Res. Coni- mun. 146:1465-1470.

Auer, J. (1948) Development of the human pulmonary vein and its major variations. Anat. Rec. 101581-594.

Barton, P.J.R., Robert, B., Cohen, A., Garner, I., Sasoon, D., Weydert, A., and Buckingham, M.E. (1988) Structure and sequence of the myosin alkali light chain gene expressed in adult cardiac atria and fetal striated muscle. J. Biol. Chem. 263:12669-12676.

Cheung, D.W. (1980) Electrical activity of the pulmonary vein and its interaction with the right atrium in the guinea-pig. J. Physiol. 314: 445-456.

Cox, K.H., DeLoen, D.V., Angerer, L.M., and Angerer, R.C. (19841 Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetric RNA probes. Dev. Biol. 101:485-502.

deAlmeida, O.P., Miklos, B., de Carvalho, M., and de Carvalho, A. (1975) The cardiac muscle in the pulmonary vein of the rat: A mor- phological and electrophysiological study. J. Morphol. 145:409- 434.

Dor, X., Corone, P., and Johnson, E. (1987) Origin of the common pulmonary vein, sinus venosus septation, atrial situs and theory of the “sinus man.” Arch. Mal. Coeur 80:483-498.

Endo, H., Kurohmaru, M., Nishida, T., Hattori, S., and Hayashi, Y. (1992a) Cardiac musculature of the intrapulmonary vein in the musk shrew. J. Vet. Med. Sci. 54:119-123.

Endo, H., Kurohmaru, M., Nishida, T., and Hayashi, Y. (1992h) Car- diac musculature of the cranial and caudal vena cava and the pulmonary vein in the fowl. J. Vet. Med. Sci. 54:479-484.

Favaro, G. (1910) Contributi all’istologica umana e compararata du vasi pulmonari. Int. Monatschr. Anat. Physiol. 27:375.

Gude, W.D., Cosgrove, G.E., and Hirsch, G.P. (1982) “Histological Atlas of the Laboratory Mouse.’’ New York: Plenum Press.

Hoh, J.F.Y., McGrath, P.A., and Hale, P.T. (1978) Electrophoretic analysis of multiple forms of ra t cardiac myosin: Effects of hypophy- sectomy and thyroxine replacement. J. Mol. Cell. Cardiol. 10:1053- 1076.

Jarkouska, D., and Ostadal, B. (1980) The effect of intermediate high

128 JONES ET AL

altitude hypoxia on the pulmonary myocardium in the mouse. Folia Morphol. 1:116-119.

Jarkouska, D., and Ostadal, B. (1983) Intermittent high altitude hypoxia-induced structural changes in the pulmonary myocardium in young mice. Virchows Arch. 43:327-336.

Karrer, H. (1959a) The striated musculature of blood vessels I. Gen- eral cell morphology. J. Biophys. Biochem. Cytol. 6:383-403.

Karrer, H. (195913) The striated musculature of blood vessels 11. Cell interconnections and cell surface. J. Biophys. Biochem. Cytol. 8: 135 -150.

Klika, E. (1976) “The Myocardium of the Intrapulmonary Veins in Mammals.” (monograph).

Kramer, A.W., and Marks, L.S. (1965) The occurrence of cardiac muscle in the pulmonary veins of rodents. J . Morphol. 117:135-150.

Larsen, W.J. (1993) “Human Embryology.” New York: Churchill Liv- ingstone.

Lee, K.L., Ross, R.S., Rockman, H.A., Harris, A.N., O’Brien, T.X., van Bilsen, M., Shubeita, H.E., Kandolf, R., Brem, G., Price, J., Evans, S., Zhu, H., Michael-Franza, W., and Chein, K.R. (1992) Myosin light chain-2 luciferase transgenic mice reveal distinct regulatory programs for cardiac and skeletal muscle-specific expression of a single contractile protein gene. J . Biol. Chem. 267:15875-15885.

Lompre, A.M., Mercadier, J.J., Wisnewsky, C., Bieveret, O., Pantal- oni, C., d’Albis, A,, and Schwartz, K. (1981) Species and age-depen- dent changes in the relative amounts of cardiac myosin isoenzymes in mammals. Dev. Biol. 84286-290.

Lompre, A., Nadal-Ginard, B., and Mahdavi, V. (1984) Expression of cardiac ventricular a- and b-myosin heavy chain genes is developmentally controlled and hormonally regulated. J. Biol. Chem. 259: 6437-6446.

Ludatscher, R. (1968) The structure of the muscular walls of the rat pulmonary vein. Anatomy 103:345-357.

Lyons, G., Schiaffno, S., Sassoon, D., Barton, P., and Buckingham, M. (1990) Developmental regulation of myosin gene expression in mouse cardiac muscle. J . Cell Biol. 111:2427-2436.

MacLeod, D., and Hunter, E. (1966) The pharmacology of the cardiac muscle of the great veins of the rat. Can. J. Physiol. 45:463-473.

Mahdavi, V., Strehler, E., Periasamy, M., Wieczorek, D., Izumo, S., Grund, S., Strehler, N., and Nadal-Ginard, B. (1986) Sarcomeric myosin heavy chain gene family organization and pattern of expression. UCLA Symp. Mol. Cell. Biol. 29:345-361.

Melton, D.A., Krieg, P.A., Redbagliati, M.R., Maniatis, T., Zink, K., and Green, M.R. (1984) Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucleic Acids Res. 12:7035- 7056.

Murray, J.F. (1986) “The Normal Lung.” Philadelphia: W.B. Saun- ders.

Nathan, H., and Gloohe, H. (1970) Myocardial atrio-venous junctions and extensions (sleeves) over the pulmonary and caval veins. Tho- rax 2.5317-324.

Ng, W., Grupp, I., Subramaniam, A,, and Rohbins, J.R. (1991) Cardiac myosin heavy chain mRNA expression and myocardial function in the mouse heart. Circ. Res. 69:1742-1750.

Rauschel, F. (18361 “De Arteriarum e t Venaram Structura.” (monograph) Breslau.

Robhins, J . , Gulick, J., Sanchez, A., Howles, P., and Doetschman, T. (1990) Mouse embryonic stem cells express the cardiac myosin heavy chain genes during development in vitro. J. Biol. Chem. 265: 11905-11909.

Rugh, R. (1990) “The Mouse.” New York: Oxford University Press. Sanchez, A., Jones, W.K., Gulick, J . , Doetschman, T.C., and Robbins,

J. (1991) Myosin heavy chain gene expression in mouse emhryoid bodies. J . Biol. Chem. 266:22419-22426.

Scharmann, W., and Wolff, D. (1980) Production of timed pregnant mice by utilization of the Whitten effect and a simple cage system. Lab. Anim. Sci. 30:206-208.

Spach, M.S., Facc, R.C., Barr, R.C., and Jewett, P.H. (1972) Spread of excitation from the atrium into thoracic veins in human beings and dogs. Am. J . Cardiol. 302344-854.

Springall, D.R., Bhatnagar, M., Wharton, J., Hamid, Q., Gulbenkian, S., Hedges, M., Meleagros, L., Bloom, S.R., and Polak, J.M. (1988) Expression of the atrial natriuretic peptide gene in the cardiac muscle of rat extrapulmonary and intrapulmonary veins. Thorax 43: 44-52.

Subramaniam, A., Jones, W.K., Gulick, J., Wert, S., Neuman, J., and Robbins, J. (1991) Tissue-specific regulation of the a-myosin heavy chain gene promoter in transgenic mice. J. Biol. Chem. 266:24613- 24620.

Toshimori, H., Toshimori, K., Ohura, C,, and Matsuo, H. (1987) Im- munohistochemical study of atrial natriuretic factor polypeptides in the embryonic, fetal and neonatal rat heart. Cell Tissue Res. 248: 627-633.

Wilkinson, D.G., Bailes, J.A., Champion, J.E., and McMahon, A.P. (1987) A molecular analysis of mouse development from 8 to 10 days post coitum detects changes only in embryonic globin expression. Development 99:493-500.

Yokota, R., and Yamauchi, A. (1985) Occurrence of cardiac muscle in the hepatic portal vein wall of the mouse and rat. Anat. Rec. 212: 23-32.

Zivin, R.A., Condra, J.H., Dixon, R.A.F., Seidah, N.G., Chretien, M., Mener, M., Chamberland, M., and Drouin, J. (1984) Molecular cloning and characterization of DNA sequences encoding rat and human atrial natriuretic factors. Proc. Natl. Acad. Sci. U.S.A. 81: 6325-6329.

Murine pulmonary myocardium: Developmental analysis of cardiac gene expression

Documents