In Vitro Cell. Dev. Biol. 27A:914-920, December 1991 © 1991 Tissue Culture Association 0883-8364/91 $01.50+0.00

A METHOD FOR THE HARVEST, CULTURE, AND CHARACTERIZATION OF HUMAN ADULT ATRIAL MYOCARDIAL CELLS: CORRELATION WITH AGE O F D O N O R

DAVID A. SMITH, JOHN L. GLOVER, LAURACE E. TOWNSEND, AND DIANE E. MAUPIN

William Beaumont Hospital, 3601 West Thirteen Mile Road, Royal Oak, Michigan 48073

(Received 26 December 1990; accepted 17 July 1991)

SUMMARY

Myocardial cell culture methdds are now well established for animal and fetal human tissue. We present here a method for harvesting and culturing adult human atrial myocardiocytes. Cells are obtained from fresh atrial tissue normally discarded after being removed to cannulate the right atrium during open heart surgery. The atrial tissue is minced and then digested using collagenase. The single cell suspension is initially cultured in serum-containing growth medium, then transferred to defined medium, selective for myocardial cell growth. The cells are characterized by immunoperoxidase stains and transmission electron microscopy. The cultured cells stain positive for myoglobin, whereas control cultured fibroblasts and endothelial cells do not. Electron microscopy shows the presence of numerous myofibrils, Z-bodies, pleomorphic mitochondria, and secretory granules. The chronological age of the donor was an important factor in culturing the adult tissue, the younger tissue correlated with a higher success rate. This method provides a means for in vitro study of human adult myocardial cells and provides guidelines for appropriate atrial tissue to use.

Key words: human; atrial; myocardiocyte; culture; harvest; characterization; aging.

INTRODUCTION

Techniques for culturing myocardial cells from rats (9), chicks (6), guinea pigs (11), and cats (5) have been established and rou- tinely utilized for cardiac muscle research. Halbert et al. in 1973 (8) reported the growth of dissociated fetal myocardiocytes in cul- ture after previous reports of fetal cardiac explant cultures (1,7). Claycomb et al. in 1989 (2), using human fetal atrial and ventricu- lar cardiac muscle cells, established cultures of dissociated myocar- diocytes with spontaneous contraction and uhrastructural charac- teristics of myocardial cells, including organized sarcomeres, inter- calated discs, and transverse tubules with couplings. They also demonstrated atrial granules in atrial cells and electron dense gran- ules associated with the golgi cisternae in ventricular cells. How- ever, to date no one has cultured neonatal or adult human myocar- diocytes. Yet these cells are extremely important to have in culture for growth factor studies pertinent to regeneration of heart cells after myocardial infarction or other ischemic insult.

Two lines of evidence support the potential to culture adult hu- man myocardial cells. First, culture techniques have been described for adult mammalian myocardial cells. Nag et al. in 1983 (10) described the successful long-term culture of cardiac myocytes from adult rats and observed morphologic changes reminiscent of embryonic or neonatal cardiac muscle cells. Second, Rumyantsev in 1974 (12) reported proliferative activity of a hmited number of atrial myocardiocytes after experimental infarction of the left ventri- cle in rats, whereas ventricular cells around the infarcted area showed no such response. He noted that adult mammalian atrial myocardiocytes are often found to proliferate in response to massive ventricular infarction, crush injury to the atrial wall, burning of the

atrial wall, as well as hypoxia. The purpose of this study was to determine if we could harvest and culture adult human myocardio- cytes by applying established techniques in animals to samples of atrial wall obtained at cardiac surgery.

MATERIALS AND METHODS

Cell harvest. Atrial tissue was obtained from right atrial appendages harvested from cardiovascular surgery patients undergoing procedures re- quiting "heart-lung bypass." The patients were randomly selected and in- cluded both sexes, with ages ranging from 38 to 81 yr. These appendages were removed routinely to insert the venous bypass cannulae into the right atria. They were obtained for study immediately after removal and placed in ice-saline slush. After rinsing in saline, the tough epicardial covering was removed using a scalpel to reduce the amount of connective tissue included in the cell harvest. The remaining "pure" atrial muscle was minced into small (0.5- to 1.0-ram a pieces and placed in cold Hanks' balanced salt solution (HBSS) without calcium or magnesium (Whittaker, WalkerviUe, MA). The atrial pieces were then transported from the operating suite to the tissue culture laboratory.

Digestion and plating. The minced atrial tissue was digested in O. 14% collagenase solution (Worthington, Freehold, N J) at a concentration of 1.43 mg/ml. The pieces were placed in 35 ml of this solution and digested in a shaker at 37 ° C at 125 rpm for 1 h. The supernatant was removed from the atrial tissue and centrifuged at 3500 rpm for 10 min at 37 ° C. Another 35 ml of collagenase solution was placed with the minced tissue while the supernatant was spinning, and the digestion continued for another hour. The cell suspension formed a pellet during centrifugation. The supernatant collagenase solution was removed and set aside for use in the third diges- tion. The cell pellet was resuspended in 2 ml of Eagle's minimal essential medium (EMEM) with Earle's salts (Whittaker) containing 30% newborn bovine serum (Whittake0 and 0.1% antibiotic solution--10 000 U/ml penicillin G, 10 000 #g/rot streptomycin, and 25 #g/ml Amphotericin B (GIBCO, Grand Island, NY). This process was continued for a total of eight digestions. The first pellet was discarded and digestions two to eight pooled.

9 1 4

HUMAN ADULT MYOCARDIOCYTE CULTURE 915

Fic. 1. Harvested atrial cells settled at the bottom of the culture plate at 24 hours. A, myocardiocyte; B, nonmyocardiocyte. ×250.

The cell concentration was checked using a hemacytometer and adjusted to 1 X l0 s cells/ml with EMEM. The cells were plated on 35-mm gelatin- coated dishes (Coming, Coming, NY) and incubated at 37 '~ C in 5% C% atmosphere. Medium was changed every 3 days for the first 2 wk of growth, then every 5 to 7 days thereafter. When the cultures spread out and ap- proached confluence, they were treated with trypsin and transferred to 60-mm gelatin-coated dishes (Coming) in EMEM. As the cultures again spread out and approached confluence they. were treated with trypsin and transferred to T-75 flasks (Coming) in MCDB 107 (Sigma, Saint Louis, MO).

Immunoperoxidase staining. A portion of tile cells grown in MCDB 107 were plated on four chamber gelatin-coated slide culture plates (Lab Tek, Naperville, IL). Control cells were human umbilical endothehal cell and human skin fibroblast cultures (Beaumont Research Institute, Royal Oak, MI) that were grown in M199 with 20% fetal bovine serum, 1% L-gluta- mine, 0.1% of 5 mg/ml insulin-5 mg/ml transferrin-5 /ag/ml selenions

it3

0 3.5

N 5.0

2.5 r,,.)

2.0

0 1.5

¢) ,.Q 1.o

~ 0 ,5

Z 0.0

0

4.0 , ~ , 1 ) I t I I

~r'73

x x

x x X X

x ) < × X

NX

X><] x x ~ K ~ X X ~ K X

XX NX NX

J NN 1 2 5 4 5

Digestions

~ X ;x x r x x [ x x 'k'x'

[ x x LX.X X)< r ' xx × x x x ~ × x 'XX x ~ XX ~ X X ~ X)< b<'X x '~ ×>< X X x ~ X × DQX x ) , x x b<X x ~ x x >(X x ~ x × r x x x x XX 'XX x ~ XX b,(X ×>< ~<x X X X)~ × x b~X x)< ~ x [XX x ~ ×)< ~,(X x x X')<

6 7 8 9

(Hours) Fro. 2. Cell yield vs. time of digestion. Hemacytometric cell counts at

the time of harvest for each digestion.

0 45

4O

I~ 35

m,-I 30

r.~ 25

0 2o

15 0

P ~ 10

5

Z 0

I I I I ' " I I I I I

, , I 1 _ 1 0 1 2 3 4 5 6 7 8 9 10

Digestions (Hours) Fro. 3. Cell viability vs. time of digestion. Planimetric cell counts at 1

wk for each digestion.

acid (Collaborative Research Inc., Waltham, MA), 0.6 ml heparin (0.015% in M199), 0.1% antibiotic-antimycotic solution (GIBCO Laboratories: 10 000 U/rot sodium penicillin G, 100 000 #g/ml streptomycin sulfate, and 25 #g/ml amphotericin B), and 300 #g/ml of endothehal cell growth supplement (ECGS) from Biotechnology Research Institute, Rockville, MD. When control cultures and harvested cells spread out and approached con- fluence they were rinsed with HBSS and fixed with 10% formalin for 10 min. The chambers were removed and the cells remaining on the plates were stained with immunoperoxidase stains for smooth muscle alpha actin (Lipshaw, Detroit, MI), striated muscle specific myosin (Sigma), myoglobin (Dako, Carpinteria, CA), Factor VIII (Lipshaw), and atrial natrinretic factor peptide (Research and Diagnostic Antibodies, Berkeley, CA). The plates were then examined using light microscopy.

Electron microscopy. A portion of cells growing in MCDB 107 were plated on 96-well gelatin-coated plates (Coming). When they spread out

F[c. 4. Cells cultured in EMEM nearing confluence, immediately be- fore passage. X50.

916 SMITH ET AL.

FIG. 5. Cells cultured in MCDB 107 with characteristic "bat winged" morphology. X 125.

and approached confluence they were rinsed with HBSS and fixed with 2.5% glutaraldehyde, 0.2 M cacedylate buffer, pH 7.4 (Polysciences, Inc., Warrington, PA), post-fixed with 1% osmium tetroxide (Polyscienees, Inc.), embedded in Epon LX-112 resin (Ladd's Research, Burlington, VA), stained with 0.03% lead citrate (Eastman Kodak, Rochester, NY) and satu- rated uranyl acetate (Pelco Co., Tustin, CA) in 50% ethyl alcohol, and then examined under transmission electron microscopy.

<

B5

BO

75

70

65

60

55

50

45

40

35

30

$

i

|

I I

GROWTH NO GROWTH Fig. 6. Myocardial tissue samples described in Table 1 separated ac-

cording to growth characteristics and plotted by age of tissue donor. Aver- age age of cells that grew 43.8 -+ 7.2; average age of cells that did not grow 62.7 + 11. P = 0.0095.

RESULTS

Cell attachment and morphology. Suspended cells settled very quickly, Usually within 1 h after plating, the majority of harvested

TABLE 1

GROWTH CHARACTERISTICS OF MYOCARDIAL TISSUE SAMPLES

Patient History Age Comments

1 AVR, no CAD 60 very few cells attached minimal growth

2 CAD, CABG×3 69 no growth Previous MI

3 AVR, no CAD 55 little adherence (10 cells) dead by 21 days

4 CABG, CAD 40 3 passages EMEM 5 CABG, CAD 44 7 passages EMEM

3 passages MCDB-107 6 CABG, CAD 56 3 passages EMEM 7 CABG×3, HTN 81 no growth

CAD 8 CABG, CAD 61 no growth 9 CABG, CAD 50 very few cells adhered

no growth 10 CABG, CAD 38 excellent growth

1 passage EMEM 5 passages MCDB-107

11 CABG, CAD 41 excellent growth 1 passage EMEM 5 passages MCDB-107

Key: AVR = Aortic valve replacement; CAD = coronary artery disease; CABG = Coronary artery bypass graft; CABG×3 = Triple bypass; HTN = hypertension.

cells were seen on the bottom of the culture dish; they were not adherent, however, as gentle agitation resuspended them. They re- quired 48 to 72 h before attachment and cytoplasmic spreading occurred. This is consistent with the experience of Nag et aL (10) culturing adult rat myocardiocytes. The settled cells were inspected using phase contrast microscopy and categorized as myocardial or nonmyocardial on the basis of their appearance. Myocardial cells were rectangular or spindle shaped, and had cross striations, a single nucleus with prominent nucleolus, and occasional branching. Nonmyocardial cells were round or oval and lacked clearly identifi- able cross striations (Fig. 1). The cells were 42.8% myocardial and 57.2% nonmyocardial.

Cell yields. Cell yields, including both myocardial and nonmyo- cardial cells, were relatively constant for digestions two to eight (Fig. 2). Cell counts ranged from 1.5 × 10S/ml at 2 h to 3.4 × 10S/ml at 6 h. Cell viability, however, varied with the length of digestion (Fig. 3). Planimetric cell counts measured at 1 wk for cultures plated separately for each digestion showed a steady in- crease in cell viability in digestions one through seven, peaking at 6 to 7 h then dropping abruptly after that, so that viability was nil at 9 h or more of digestion.

Growth characteristics. After attaching to the gelatin matrix, cells spread out with numerous cytoplasmic extensions seen extend- ing from the cells. Within 2 wk of plating, the cells had sufficiently spread to reach confluence (Fig. 4). "Splitting" the cells with tryp- sin and transferring them to 60-mm dishes and then to T-75 flasks resulted in similar spreading to reach confluence. Utilizing EMEM, the maximum number of passages before deterioration of cells was 7. Using MCDB 107, a medium shown by Suzuki et al. (14) to be selective for the growth of myocardial cells, the maximum number of passages dropped to three. All cultures went through at least 1

HUMAN ADULT MYOCARDIOCYTE CULTURE 917

Fro. 7. lmmunoperoxidase staining characteristics of isolated myocardiocytes, control fibroblasts and control endothelial cells. A, endothelial cells with cytoplasmic granules staining for Factor VIII; B, myoeardiocytes with no Factor VIII staining; C, fibroblasts with no staining for myoglobin; D, myocardiocytes with diffuse cytoplasmic staining for myoglobin; E, fibroblasts with no significant actin filament staining; F, myocardiocytes with aefin filament staining; G, myocardial cells with no primary ANP antibody; H, myocardial cells, with perinuclear staining of ANP. X500.

918 SMITH ET AL.

Fro. 8. Electron micrograph of presumptive myocardiocytes cultured in MCDB 107. A, free ribosomes; B, pleomorphic mitochon- dria; C, dense Z-body with attached myofibrils; D, myofibrils; E, autophagosome. )< 16 300 (inset × 1575).

passage in EMEM; cells placed initially in MCDB 107, regardless of their age, failed to attach and grow. The phase contrast microscopic appearance of cells cultured in MCDB 107 was described as "bat winged" (Fig. 5) and was similar to cells described in human fetal atrial myocardial cell cultures (2). Growth characteristics are corre- lated with age of patient and pathology associated with myocardial tissue in Table 1. Fig. 6 illustrates the correlation of growth with age. The younger patient tissue obviously has increased growth potential relative to older patient myocardial tissue, with some overlap,

Characterization. Monoclonal antibody immunoperoxidase stains were performed on presumptive myocardiocytes using fibro- blasts and endothelial cells as controls. The stains showed charac- teristic staining of harvested cells for the proteins actin, myosin, myoglobin, and atrial natriuretic factor whereas control cells did not. Factor VIII stains of endothelial cells highlighted their charac- teristic cytoplasmic granules whereas the same stains revealed no such granules in the presumptive myocardiocytes. Actin stains in the presumptive myocardiocytes demonstrated numerous cytoplas- mic filaments whereas myosin and myoglobin stains showed more diffuse cytoplasmic staining (Fig. 7}. Atrial natriuretic peptide (ANP) showed definitive perinuclear staining. Electron microscopy of the presumptive myocardiocytes revealed that they were secre- tory cells containing numerous free ribosomes, rough endoplasmic reticulum, and secretory granules. The cells also contained numer- ous microfilaments. No well-organized sarcomeres or T-tubules were seen; however, dense Z-bodies containing attached microfila- ments, and pleomorphic mitochondria are prominent (Fig. 8). The

electron micrographic appearance is similar to that seen in adult rat myocardiocyte cultures at i 4 days undergoing myofibrillar reorgani- zation (10). Figure 9 illustrates adult human tissue from which myocardiocytes were isolated (age: 44 yr).

DISCUSSION

This study is of significance because it provides a method for the harvest and culture of adult human myocardial cells. These cells are harvested from human heart tissue with ischemic pathology, exactly the tissue that is of interest for identifying human growth factors that can prevent or partially reverse damage to the heart. Also important are the guidelines established for age of tissue. This raises the ques- tion of whether human atrial tissue is irreversibly affected by age or is able to respond to growth factors not present in our culture system.

We have found no published reports of adult human myocardial cell culture systems. Evidence cited here shows that it is possible to culture adult myocardial cells from animal tissue and that atrial cells are unique by nature of their neuroendocrine function. This encour- aged us to pursue a human adult atrial myocardial cell culture system. However, initial culture methods which followed those use- ful for rat tissue produced discouraging results; optimizing methods of digestion and harvest finally produced cells that persisted through several passages.

Identification of the myocardial cells was done using antibodies specific for muscle myosin, myoglobin, and atrial natriuretic pep- tide. Immunoperoxidase stains were consistent with cells of cardiac

HUMAN ADULT MYOCARDIOCYTE CULTURE 919

Fro. 9. Electron micrograph of myocardial tissue used for harvest of myocardiocytes. X16 300.

muscle origin, whereas control fibroblasts did not reflect the same findings. Repeat stains on the same harvest and stains performed on additional harvests verified these findings. Transmission electron microscopy (TEM) was performed to confirm the myocardial origin of these cells. We found uhrastructural organization consistent with adult rat myocardial cells undergoing myofibrillar reorganization as observed by Nag and associates (10). The above evidence strongly suggests that we have successfully cultured adult human atrial myo- cardial cells even though organized sarcomeres and t-tubules are absent at this stage of myofibrillar reorganization. Our method for harvesting cells from the atrial appendage has also allowed us to culture endothehal cells from adult atrial tissue by using selective media and cloning techniques.

Atrial myocardiocytes, themselves, represent a heterogeneous population of cells, The common atrial myocyte is specialized not only as a contractile muscle cell but also as an endocrine cell secret- ing ANP. This hormone has potent diuretic and hypotensive effects and inhibits renin and aldosterone secretion (3,4). Atrial myocytes therefore have the machinery responsible for synthesis, processing, and releasing ANP. The specific organelles involved are rough en- doplasmic reticulum, free ribosomes, Golgi apparatus and atrial specific granules. Other uhrastructural details unique to atrial myo- cytes are a) absent or few transverse tubules and b) tremendous variation in the organization of sarcomeres, from dense and well-or- ganized, as seen in ventricular cells, to scanty and haphazard (13). The ultrastructural characteristics of atrial myocytes, their neurose- cretory potential, and their response to injury in vivo differ so greatly from their ventricular and skeletal muscle counterparts that

they must be considered a different class of striated muscle cells. These cells have the potential to survive in culture for up to 7 passages. The immunostaining that we illustrated for ANP confirms the presence of atrial myocardiocytes in our cultures.

Further TEM characterization and immunostaining of human adult presumptive myocardiocytes during the course of their growth cycle in culture is planned. These cells will be extremely valuable in studies of adult heart cell response to growth factors with respect to prohferation and differentiation.

ACKNOWLEDGEMENTS

Henry Mauer, M.D., Nicholas Tepe, M.D., and Joseph Bassett, M.D., provided fresh atrial tissue. John Long, M.D. and the Immunopathology Lab performed immunoperoxidase staining. Marilyn Seymour and Darlene Reitz-Vick provided excellent technical and photographic assistance. The Beaumont Research Institute funded this project.

REFERENCES

1. Chang, T. D.; Cumming, G. R. Chronotropic responses of human heart tissue cultures. Circ. Res. 30:628-633; 1972.

2. Claycomh, W. C.; Delcarpio, J. B.; Guice, S. E., et al. Culture and characterization of fetal human atrial and ventricular cardiac muscle cells. In Vitro Cell. Dev. Biol. 25:1114-1120; 1989.

3. DeBold, A. J. Atrial natriuretic factor: a hormone produced by the heart. Science 230:767; 1985.

4. Flynn, T. G.; Davies, P. L. The biochemistry and molecular biology of atrial natriuretic factor. Biochem. J. 232:313; 1985.

5. Follmer, C. H.; Ten Eick, R. E.; Yeh, J. Z. Sodium kinetics in cat atrial myocytes. J. Physiol. 384:169-197: 1987.

6. Friedman, B. J.; Beihn, R.; Friedman, J. P. The effect of hypoxia on

920 SMITH ET AL.

thallium kinetics in cultured chick myocardial cells. J. Nucl. Med. 28(9):1453-1460; 1987.

7. Garry, W. E.; Townsend, S. E. Neural responses and reactions of the heart of a human embryo. Am. J. Physiol. 152:219-224; 194.8.

8. Halbert, S. P.; Bruderer, R.; Thompson, A. Growth of dissociated beating human heart cells in tissue culture. Life Sei. 13:969-975; 1973.

9. Harary, I.; Farley, B. In vitro studies of single rat heart cells. I. Growth and organization. Exp. Cell. Res. 29:451-456; 1963.

10. Nag, C. G.; Lee, M. I.; Kosiur, J. R. Adult cardiac muscle cells in long-term serum-free culture: myofibrillar organization and expres- sion of myosin heavy chain isoforms. In Vitro Cell. Dev. Biol. 26:464-470; 1990.

11. Nakajima, T.; Yoshihisa, K.; lto, H., et al. Anticholinergic effects of quinidine, disopyramide and procainamide in isolated atrial myo- eytes. Circ. Res. 64(2):297-303; 1989.

12. Rumyantsev, P. P. Ultrastructural reorganization, DNA synthesis and mitotic division of myocytes in atria of rats with left ventricle infarc- tion. Virehows Arch. B. 15:357-378; 1974.

13. Severs, N. J. Isolated adult cardiomyocytcs, ln: Piper, H. M.; Isen- berg, G., eds. Constituent cells of the heart and isolated cell models in cardiovascular research. Structure and metabolism, vol. 1. Boca Raton, FL: CRC Press; 1989:16-18.

14. Suzuki, T.; Ohta, M.; Hoshi, H. Serum-free, chemically defined me- dium to evaluate the direct effects of growth factors and inhibitors on proliferation and function of neonatal rat cardiac muscle cells in culture. In Vitro Cell. Dev. Biol. 25:601-606; 1989.