Cryopreserved aortic homografts contain viable smooth muscle cells capable of expressing transplantation antigens

Cryopreserved aortic homografts contain viable smooth muscle cells capable of expressing transplantation antigens Frozen aortic tissue is increasingly used as homografts in reconstructive cardiovascular surgical procedures. The viability of cells within these cryopreserved tissues, their identity, and their potential immunogenicity have been the subject of controversy. We cultured cells from cryopreserved human aortic homografts that reacted with a monoclonal antibody that recognizes muscle actin isoforms, identifying them as smooth muscle cells. Under basal conditions, these smooth muscle cells contained messenger ribonucleic acid for class I human leukocyte antigens detected by northern blotting and expressed class I human leukocyte antigen on their surfaces as measured by enzyme-linked immunoassay and immunohistochemistry. Unstimulated smooth muscle cells contained no class II human leukocyte antigen messenger ribonucleic acid as determined by northern blotting and displayed almost no class II surface antigen as determined by enzyme-linked immunoassay. Interferon gamma (1000 U /m1, 72 hours), a product of activated T lymphocytes, not only increased the expression of class I human leukocyte antigens by smooth muscle cells, but induced class II human leukocyte antigen messenger ribonucleic acid and elevated surface expression from 22 ± 7 to 819 ± 35 enzyme-linked immunoassay units (n = 4). Immunohistochemistry revealed few class II-positive smooth muscle cells under basal culture conditions, but all cells showed high levels of DR antigen after exposure to interferon gamma for 3 days. Similar results were obtained in two independent isolates. We conclude that cryopreserved aortic homografts can contain viable smooth muscle cells capable of expressing major histocompatibility antigens that might render them immunogenic and susceptible to rejection by the recipient's immune system. (J THORAC CARDIOVASC SURG 1993;106:1173-80)

Robert N. Salomon, MD,a, b Gary B. Friedman, MD,a Allan D. Callow, MD, Phfr," Douglas D. Payne, MD,c and Peter Libby, MD,d Boston, Mass.

From the U.S. Department of Agriculture, Human Nutrition Research Center on Aging at Tufts University"; the Departments of Pathologyband Surgery,", New England Medical Center; Vascular Medicine and Atherosclerosis Unit, Brigham and Women's Hospital," Boston, Mass. Supported in part by the U.S. Department of Agriculture, Agricultural Research Service contract 53-3K06-5-1 O. The contents of this publication do not necessarily reflect views or policies of the U.S. Department of Agriculture nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. Dr. Libby is an Established Investigator of the American Heart Association. Dr. Salomon was a Fellow of the American Heart Association, Massachusetts Affiliate, during the course of this work. Received for publication Oct. 5, 1992. Accepted for publication Feb. 17, 1993. Address for reprints: Robert N. Salomon, MD, Bioavailability Laboratory, Human Nutrition Research Center on Aging, 711 Washington St., Boston, MA 0211!.

12/1/46700

Cryopreserved autologous aortic homografts from unrelated donors have been used for years as a source of material in cardiovascular reconstructive procedures. These tissues are structurally suitable for this purpose and have yielded good surgical results in a number of series. 1-3 The present study was undertaken to determine if cryopreserved homografts contained viable smooth muscle cells (SMCs), and if these SM Cs retained the capacity to express targets for rejection by the recipient's immune system. The surface structures involved in allograft rejection are highly polymorphic proteins known by various names, such as histocompatibility antigens, transplantation antigens, or, in the case of human beings, human leukocyte antigens (HLA). These molecules are encoded in a region of the genome known as the major histocompatibility complex (MHC) and are inherited in a simple mendelian fashion. Because there are many alleles for I I 73

The Journal of Thoracic and Cardiovascular Surgery December 1993

I I 7 4 Salomon et al.

Fig. 1. Cells cultured from cryopreserved aortic homografts stain positively for muscle-specific 0: actins. Immunostaining was performed with HHF 35, a monoclonalantibody specific for muscle 0: actins. Cells show typical SMC staining of 0: actin cables within the cytoplasm. (Hematoxylin counterstain; original magnification 200X.) each locus in the population, tissues obtained at randomfrom unrelated donors likely differ significantly from the host's own HLA antigens and can thus elicit rejection. These surface antigens are divided into two major types. Class I antigens (HLA A, B, and C loci) serve as recognition structures for T lymphocytes that bear the CD8 surface marker often associated with cytolytic function. CD8+ killer T cells constitute the effector limb of the cellular immune response. These cytolytic T cells, interacting with foreign class I HLA, mediate the damage to the allograft during rejection. Most nucleated cells express class I MHC antigens constitutively. Class II antigens provide the crucial stimulus for initiation of the cellular immune response by selective recognition by helper T lymphocytes that bear the CD4 surface marker. CD4+ helper cells require interaction with class II molecules to initiate the afferent limb of cellular immunity. In contrast to class I molecules, class II antigens (HLA-DR, -DP, and -DQ loci) are expressed by few cells other than leukocytes under usual circumstances in vivo. Recently, it has become clear that non-bone marrowderived cells, including vascular wall cells that might populate cryopreserved aortic homografts (e.g., endothelium and SMCs), can also express class II antigens in an inducible manner. If viable cells survive the preparation and storage that aortic homografts undergo and retain the ability to express HLA, they may not evade immune rejection with attendant undesirable clinical sequellae. In the present study, we cultured SMCs from the aortic portions of cryopreserved aortic homografts prepared for clinical use. We report that these homograft-derived

SMCs constitutively bear class I HLA antigens and exhibit regulated class II HLA-DR expression. Methods Cytokines. Recombinant human interferon gamma (IFN-,) was providedby Genentech, Inc., South San Francisco,Calif., recombinant human leukocyteIFN -0: wassuppliedby Dr. Peter Sorter of Hoffmann-LaRoche, Inc., Nutley, New Jersey, and natural human IFN-iJ was purchased from Lee Biomolecular Research, Inc., San Diego, California. Antibodies. Human class I histocompatibility antigens were detected with monoclonalantibody W6j32 (immunoglobulin G 2a).4 A hybridoma cell line expressing this antibody was supplied by Dr. Mark 1. Green of the Universityof Pennsylvania. The immunoglobulinG was isolated from hybridoma culture supernatants by protein A affinitychromatography. Class II histocompatibilityantigens were detected with either monoclonal antibody 1-2 (immunoglobulin G 2a)5 or monoclonal antibody 949, both supplied as mouse ascites fluid by Drs. Lee Nadler and Stuart F. Schlossman of the Dana-Farber Cancer Institute, Boston, Massachusetts. A mouse immunoglobulin G 2a K myeloma protein (UPC 10;Sigma Chemical Co., S1. Louis, Mo.) was used as a class-matched irrelevant antibody for control purposes. Muscle-specific actins were demonstrated with mouse monoclonalantibody HHF 35 (Enzo Biochem Inc., New York, N.Y.). Endothelial cells were identified with the lectin Ulex europaeus I (Vector Laboratories, Burlingame, Calif.) and a polyclonal rabbit antibody against human von Willebrand factor (Dako Corporation, Santa Barbara, Calif.). Nucleic acid probes. Class I HLA messenger ribonucleic acid (mRNA) was detected with a 1.8 kb probe derived by means of Bgi II cleavage of a genomic clone of HLA-B7 complementary deoxyribonucleic acid (cDNA).6 mRNA for class II antigens was detected with a 550 bp Pst I fragment of DBIO probe for HLA-DRo:.7 Cell cultures. Aortic homograft-derived SMCs were pre-

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6

Salomon et al.

117 5

Fig. 2. IFN-I' increases class I MHC gene expression in SMCs cultured from human aortic homografts. SMCs from human aortic homografts (HSMC) and endothelial cells from human saphenous vein (EC) were separately incubated under usual culture conditions or in the presence of recombinant human IFN-I' (500 U /ml) for 72 hours. RNA was hybridized with a probe for class I HLA derived from a HLA-B7 eDNA clone. Hybridization bands are present in all lanes. Exposure to IFN-I' causes a large increase in expression of class I mRN A. The position of 18S ribosomal RNA is indicated as a size marker.

pared by two methods from unused aortic portions of cryopreservedaortic homografts treated with antibiotics (CryoLife Inc., Marietta, Ga.) obtained at the time of reconstructive cardiac operation. SMCs were either grown from explants of homograft aorta or isolated enzymatically from the inner third of the tunica media. In the latter case, the adventitia and abluminal two thirds of the tunica media were removed before dissociation of the tissue with collagenase. These cells were cultured in Dulbecco's modified Eagle's medium containing fetal bovine serum (10%) with HEPES buffer (25 mmol/L). All components of tissue culture media were tested for endotoxin contamination with the chromogenic Limulus amoebocyte lysate assay. Only media that contained a final endotoxin concentration ofless than 40 pg/rnl were used for these studies to avoid stimulation of endogenous cytokine production by the endotoxin-sensitive SMCs. Four of eight attempts to culture SMCs from aortic homografts were successful (two of three from explantation and two of five from enzymatic dissociation). Human vascular endothelial cells were prepared from explants of unused portions of saphenous veins obtained at the time of coronary artery bypass operation.' This use of normally discarded human tissue was approved by the institutional Human Investigation Review Committee. RNA extraction and nucleic acid hybridization. RNA was isolated from cultured SMCs by phenol extraction after lysis in guanidinium isothiocyanate.? After electrophoresis on agarose gels (1.2%) that contained formaldehyde (2.2 mol/L), the RNA was transferred to nylon membranes (Hybond-N; Amersham Co., Arlington Heights, Ill.). Membranes were examined with ultraviolet transillumination to determine RNA integrity and equal loading of all lanes. DNA probes were labeled by random hexanucleotide priming with 32P-labeled nucleotide triphos-

phates. Prehybridization, hybridization, and autoradiography were performed with standard techniques.l? Measurement of cell surface antigen level by enzymelinked immunosorbent assay. SMCs derived from aortic homografts were cultured in 96 well plates for study by cell-based enzyme-linked immunosorbent assay (ELISA).II After treatment with the test stimulus, the cell layers were washed with phosphate-buffered saline solution containing bovine serum albumin (10 mg/rnl) to block nonspecific binding sites. The wells were incubated with various dilutions of the test antibody or buffer alone for 2 hours at 4° C. The cell layers were then washed three times with phosphate-buffered saline solution with bovine serum albumin then incubated for I hour with horse antimouse antibody conjugated with biotin (Vector Laboratories, Burlingame, Calif.) diluted I: 1000. After an additional wash in phosphate-buffered saline solution with bovine serum albumin, the cell layers were incubated for 30 minutes with streptavidin-alkaline phosphatase diluted I: I000 (Zymed Laboratories, South San Francisco, Calif.). After this incubation, the cell layers were washed three more times with phosphatebuffered saline solution with bovine serum albumin and finally with phosphate-buffered saline solution alone. The cell-associated enzyme activity was then measured by incubation for IS to 30 minutes with paranitrophenyl phosphate (I mg/rnl) in diethanolamine (0.1 mol/L, pH 10.3) that contained levamisole (I mmol/L, Sigma) to block endogenous phosphatase activity. II The reaction was stopped by the addition of two volumes of sodium hydroxide (2 mol/L), and the absorbance at 410 nm was read in an automated microplate photometer. Control incubations performed with equivalent dilutions of the irrelevant class-matched antibody UPC 10, which is used with most experiments, yielded negligible absorbance.

The Journal of Thoracic and' Cardiovascular Surgery December 1993

I I 7 6 Salomon et al.

control

alpha

beta

gamma

INTERFERON

Fig. 3. IFN-y increases the expressionof class II HLA on homograft-derived SMC. Cultures of aortic homograftderived SMC were incubated for 72 hours with or without IFN-a, (3, or y (1000 U jml). The expression of class I or II MHC antigen on the surface of the cultured cells was measured by ELISA. Under control conditions cells expressed class I but little or no class II antigen. IFN--y induced a marked increase in class II expressionand a more modest but statistically significant increase in class I expression (p < 0.01). Neither IFN-a nor -(3 induced a significant increase in either class I or II expression.

Immunohistochemical evaluation of cultured SMCs. SMCs were cultured in Labtek four-well chambers (Miles Laboratories, Naperville, Ill.) with or without the presence of IFN--y, 1000 U jml, for 3 days. After treatment with the test stimulus, the cell layers were fixed with acetone at 4° C for 60 secondsand evaluated by immunohistochemicalstudies with the antibodies described previously by standard avidin-biotin immunoperoxidase techniques. Cell layers were washed for 10 minutes with phosphate-buffered saline solution that contained horse serum ( 1%). Then primary antibody was diluted I:I00 in phosphate-buffered saline solution and incubated with the cell layers for I hour at ambient temperature. After being washed with phosphate-buffered saline solution, biotinylated secondary antibody was applied for 40 minutes. The slides were again washed and incubated for 30 minutes with Vectastain ABC reagent, a mixture of avidin and biotinylated horseradish peroxidase (Vector). The slides were washed again and developed in peroxidase substrate solution (50 ml acetate buffer [0.1 mol/ L, pH 5.0] that contained 25 ,ILl of a 30% hydrogen peroxide solution and aminoethylcarbazole, 10 mg dissolved in 2.5 ml N,N-dimethylformamide). Controls in all immunohistochemical experiments included omissionof primary antibody, the use of class-matched irrelevant primary antibodies, and parallel reactions with panels of cells of known positiveor negative reactivity. Results Immunohistochemical identification of cultured cells. Cells cultured from homografts showed characteristic staining of actin cables when stained with the muscle-specific antibody, HHF 35 (Fig. I). This staining pat-

tern was retained even after several passages in culture. In addition, these cells exhibited typical morphologic characteristics of SMCs in vitro, including a pattern of growth in hills and valleys. Cells in these isolates did not stain with reagents that reacted with human endothelial cells (a polyclonal antibody against human von Willebrand factor and the lectin Ulex europaeus I) under similar conditions. Class I HLA expression in cultured cells. Under basal conditions, cells cultured from homografts contained mRNA that hybridized with a cDNA probe for HLA-B7 that recognizes a conserved region of class I HLA genes (Fig. 2). ELISA demonstrated that the surfaces of these cells bound the HLA class l-specific monoclonal antibody W6/32 (Fig. 3). Immunohistochemical examination of these cells also revealed basal expression of class I MHC antigens that stained the surfaces of the cells in a diffuse and uniform manner (Fig. 4, A). Previous studies have shown that IFN-y, a product of activated T lymphocytes and natural killer cells, can induce expression of MHC molecules in many cell types including vascular endothelial cells and SMCS. IO, 12, 13 Therefore, we tested the influence of this lymphokine on MHC expression in the homograft-derived cells. After exposure for 72 hours to recombinant human IFN-y, the mRNA level for class I MHC increased several-fold (Fig. 2). Recombinant human IFN-y moderately increased the

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6

Salomon et al.

1 17 7

Fig. 4. Immunohistochemical identification ofclass I and class II MHC antigens onSMCs cultured fromhuman aortic homografts. SMCs cultured from homografts were cultured with or without IFN-')' for 72 hours and stained with class I or class II antibody by standard ABC techniques. Both unstimulated (A) and stimulated (B) cells were positive forclass I antigens, with stimulated cells showing a moderate increase instaining intensity. Stimulated cells showed intense staining for class II antigens (C). Unstimulated cells were overwhelmingly class II negative (D). (Hematoxyin counterstain; original magnification 200X.)

surface expression of class I antigen, determined quantitatively by ELISA, compared with that of cells not exposed to recombinant IFN-'Y (p < 0.01, Student's t test). IFN-a and IFN-,6 did not change class I expression (Fig. 3). Inspection ofSMC layers exposed to IFN-'Yfor 3 days and stained immunohistochemically for class I antigens with monoclonal antibody W6/32 showed a moderate increase in staining intensity compared with that of nonstimulated controls (Fig. 4, B), which is consistent with the change in surface expression of these antigens documented by quantitative ELISA (Fig. 3). Class II HLA expression in cultured cells. Although all cultures of homograft-derived cells showed class I MH C genes under basal conditions, we found little or no class II MHC mRNA in unstimulated cells examined with an HLA-DR probe that encodes well conserved a-chain sequences(Fig. 5). The surfaces of untreated cells contained only low levels of immunoreactive class II MHC antigen measured by ELISA with monoclonal antibody 1-2 (Fig. 3). Incubation with IFN-'Y for 72 hours caused marked accumulation of class II MHC mRNA by homograftderived cells (Fig. 5). This increase in class II mRNA

corresponded to an 18-fold increase in expression of HLA-DR as measured by ELISA (Fig. 3). Significant increases in the expression of class II MHC antigen did not occur until the second day of exposure to IFN-'Y and increased approximately linearly over the next 3 days of exposure (Fig. 6, A). IFN-'Y induced expression of class II MHC antigens by homograft-derived SMCs in a concentration-dependent manner (Fig. 6, B). Immunohistochemical study demonstrated induction of class II MHC antigen involving the majority of cells exposed to IFN-'Y for 3 days (Fig. 4, C). The pattern of expression was heterogeneous, varying from intense staining of discrete intracytoplasmic perinuclear clusters to diffuse surface staining. The overwhelming majority of cells not exposed to INF-'Y were completely negative for class II antigen (Fig. 4, D). Discussion Preserved human aortic homografts have been used for many years in cardiovascular reconstructions in children and adults. Long-term follow-upstudies have established the long-term viability of many of these implants. Barratt- Boyes and associates I reported a 10-year actuarial

I I78

The Journal of Thoracic and Cardiovascular Surgery December 1993

Salomon et al.

MMA IFN Fig. 5. IFN-'Y induces class II MHC gene expression in SM Cs cultured from human aortic homografts. SMCs from human aortic homografts (HSMC) and endothelial cells from human saphenous vein (EC) were separately incubated under usual culture conditions, or in the presence of recombinant human IFN-'Y (500 U / ml) for 72 hours. RNA was hybridized with an HLA-DR-a cDNA probe. Intense bands are present in lanes containing RNA from' homograft SMCs and human saphenous vein ECs incubated with IFN-')'. Essentially no signal is seen from unstimulated cells. The position of I8S ribosomal RNA is indicated as a size marker.

graft survival of almost 60%. Angell and associates/ reported similar graft survival at II years in their series. Bodnar and associates' reported greater than 50% survival of implanted homografts after 20 years. These favorable long-term results have been obtained with homografts subjected to a variety of treatments, some involving prolonged incubation with antibiotics and others involving cryopreservation. The increased recognition of the limitations and durability of bioprosthetic valves and the commercial availability of cryopreserved aortic homografts have renewed interest in this therapeutic approach. Yet, a number of questions persist regarding the biologic implications of implantation of cryopreserved aortic homografts. Whether these homografts contain viable cells has been controversial. Barratt-Boyes and associates! considered cells in homograft valves sterilized with antibiotic solution to be nonvital. Angell and associates? described viable donor cells in histologic sections of valves explanted 6 months after the operation and interpreted these findings as a demonstration of the replicative capacity of homograft-derived fibroblasts. O'Brien and associates':' reported the culture of cells from cryopreserved pulmonary valve leaflets treated with antibiotics before implantation and the presence of viable donor-derived fibroblasts on an explanted valve leaflet 9 years after implantation.!" Others believed that retention of donor

fibroblasts or ingrowth of host fibroblasts yield repopulation of aortic homogratts." Although a number of authors believe that cells seen within homografts are fibroblasts, the identity of these cells has not been established with certainty. The possibility that these cells may represent vascular SMCs has been neglected. Another issue is the potential antigenicity of aortic homografts. Bodnar and associates' have furnished evidence that adsorption of soluble antigens from calf serum used during processing of human aortic homografts can evoke an immune response. This finding might explain a tendency toward a higher rate of primary failure observed in this group in grafts preserved in a solution containing calf serum. Yankah and associates'" reported endothelial cells capable of expressing HLA class I and II antigens on explanted cryopreserved allografts removed because of incompatibility. Recent advances in vascular biology permit new approaches to these important basic questions that underlie the clinical use of human aortic homografts. This study applied tools of modern cellular and molecular biology to approach the issues raised previously. About half of the cryopreserved aortic homografts examined in this study yielded cells by tissue culture. The lack of cell growth from the remaining specimens could relate to differences in donors, preparation and storage, or cell culture conditions. Immunostaining of these cells indicated their

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6

Salomon et al.

1.2

1.2

c: CD

c: o

C)

;:

c:

<

~ .c ;: as

.Ci)

0.9

...CD

'e c(

0.3

Class II ----.---. Class I

o

.! ::I: A

lil

W

::J

< ...J

CJ)

~

Co

>< .g

::J 0.6

Co ::::i

E o (,)

0.9

UI

...... .l!l

w

I I 79

::I:

o.o.q:=;;--r---r---r----, 1 2 3 4

0.6

-c

(/)

::::i

!!:!. 0.3

0.0 " f - - - ' T - - - r - - - - , 1 10 100 1000

o

Days

B

Gamma

Interferon

(U/ml)

Fig. 6. IFN-')' produces time-dependent and dose-dependent increases in the expression of class II MHC antigen onthesurface ofhomograft-derived SMCs. A, Cultures ofhomograft-derived SMCswere exposed to IFN-')' (1000 UImO forthe indicated timeperiods. Class I andclass II MHC expression were measured by ELISA. Data shown are mean ± standard deviation, n = 6. B, SMCsderived from human aortic homografts were treatedwith IFN-')' at the indicated concentrations for 72 hours. Measurement of class II expression was by ELISA. Data shown are mean ± standard deviation, n = 6.

smoothmuscleorigin.Although the ideal tissuesourcefor cells analyzed in this study would have been portions of leaflets from cryopreserved homografts, we believe that the viability and immunologic properties of SMCs derived from adjacent, identically processed aortic portionsof the homografts are likelyto be identical to SMCs derived from leaflets. Both in vivoand in vitro, immunologic properties of vascular SMCs seem to be conserved in cells derived from such diverse sources as saphenous veins.!" arteries, 13 and evenatheroscleroticplaques.17 We did not succeed in propagating endothelial cells from thesecryopreserved specimens, although some endothelium-likecells were seen on frozen sections of the tissues. Lack of endothelial cell growth is not surprising because the monolayerof endothelial cells coveringthe intima of human vessels is vulnerable to mechanical injury and desiccation during the processing of the specimens. In addition, our tissueculture methods werenot designedfor optimal harvest or propagation of endothelial cells. An important aspect of this study addresses the ability of the viable cells cultured from homografts to express HLA, surface structures that permit recognition of cells as foreign. We found that every isolate of SMCs from homografts expressed class I antigens at both the RNA and protein levels. Thus, viableSMCs preservedthe ability to express class I HLA, structures recognized by

CD8-bearing T cellswhich often mediate cytolytic function in the context of allograft rejection. SMCs cultured from homografts expressed low levels of HLA-DR, a representative class II antigen. The bulk of the SMCs in vitro expressed high levels of HLA-DR mRNA and protein after stimulation with the lymphokine IFN--y. Foreign class II HLA can trigger the immune system of allograft recipients by providing the recognition structures required for interaction with CD4bearing helper T cells. In summary, SMCs derived from the aortic portions of cryopreserved aortic homografts retain the ability to expressboth class I HLA required for recognition by the effector limb of the cellular immune response and class II HLA generally involved in the initiation of cellular immunity. In these respects, homograft-derived SMCs resemble their counterparts cultured from normal human blood vessels. 10, 13, 18 Functionalstudiesin vitro haveestablished that human vascular SMCs induced to express maximal levels of HLA by exposureto IFN-')'in culture stimulate T cells from unrelated individuals much less effectively than do similarlytreated endothelialcells.This difference may be due to the expression of accessory functions by endothelial cells not shared by SMCs in vitro.'? This relatively weak ability to stimulate an allogeneic response may underlie the clinical success of cryopreserved

I I 80

Salomon et al.

homografts despite the ability of viable SMCs within these tissues to express histocompatibility genes. Alternatively, appropriate stimuli for induction of class II antigens required to initiate the cellular immune response may be lacking under normal circumstances in implanted homografts. In conclusion, our results demonstrate that aortic homografts commercially prepared by cryopreservation contain viable SMCs that can be propagated in tissue culture. These cells retain certain functional capacities after cryopreservation and passage in vitro, including the ability to express histocompatibility genes. The functional consequences of this potential immunogenicity require further study. Our data indicate how application of contemporary cell biology techniques can be used to address basic issues in cardiovascular transplantation.

I.

2.

3.

4.

5.

6.

7.

REFERENCES Barratt-Boyes BG, Roche AH, Subramanyan R, Pemberton JR, Whitlock RM. Long-term follow-up of patients with the antibiotic-sterilized aortic homograft valve inserted freehand in the aortic position.Circulation 1987;75:76877. Angell WW, Angell JD, Oury JH, Lamberti JJ, Grehl TM. Long-term follow-up of viable frozen aortic homografts: a viable homograft valve bank. J THORAC CARDIOVASC SURG 1987;93:815-22. Bodnar E, Matsuki 0, Parker R, Ross DN. Viable and nonviable aortic homografts in the subcoronary position:a comparative study. Ann Thorac Surg 1989;47:799-805. Barnstable CJ, Bodmer WF, Brown G, et al. Production of monoclonal antibodies to group A erythrocytes, HLA and other human cell surface antigens: new tools for genetic analysis. Cell 1978;14:9-20. Nadler LM, Stashenko P, Hardy R, Pesando JM, Yunis EJ, Schlossman SF. Monoclonal antibodies defining serologically distinct HLA-D/DR related la-like antigens in man. Hum Immunol. 1981;2:77-90. Barbosa JA, Kamarck ME, Biro PA, Weissman SM, Ruddle FH. Identification of human genomic clonescoding the major histocompatibility antigens HLA-A2 and HLAB7 by DNA-mediated gene transfer. Proc Natl Acad Sci USA 1982;79:6327-31. Wake CT, Long EO, Strubin M, et al. Isolation of eDNA clones encoding HLA-DR alpha chains. Proc Natl Acad Sci USA 1982;79:6979-83.

The Journal of Thoracic and Cardiovascular Surgery December 1993

8. Foxall T, Auger KR, Callow AD, Libby P. Adult human endothelial cell coverageof small caliber Dacron and PTFE vascular prostheses in vitro. J Surg Res 1986;41: 158-72. 9. Chomczynski P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156-9. 10. Warner SJC, Friedman GB, Libby P. Regulation of major histocompatibility gene expression in cultured human vascular smooth muscle cells. Arteriosclerosis 1989;9:27988. 1I. Posner MR, Antoniou D, Griffin J, Schlossman SF, Lazarus H. An enzyme-linked immunosorbent assay (ELISA) for the detection of monoclonal antibodies to cell surface antigens on viable cells. J Immunol Methods 1982;48:2331. 12. Pober JS, Collins T, Gimbrone MA Jr, et al. Lymphocytes recognize human vascular endothelial and dermal fibroblast Ia antigens induced by recombinant immune interferon. Nature 1983;305:726-30. 13. Hansson GK, Jonasson L, Holm J, Clowes MK, ClowesA. Gamma interferon regulates vascular smooth muscle proliferation and Ia expression in vivo and in vitro. Circ Res 1988;4:712-9. 14. O'Brien MF, Johnston N, Stafford G, et al. A study of the cells in the explanted viable cryopreserved allograft valve. J Card Surg 1988;3(3 Suppl):279-87. 15. Gonzalez LL, Spotnitz AJ, Mackenzie JW, et al. Homograft valve durability: host or donor influence? Heart Vessels 1990;5:102-6. 16. Yankah AC, Wottge HU, Muller-Hermelink HK, et al. Transplantation of aortic and pulmonary allografts, enhanced viability of endothelial cells by cryopreservation, importance of histocompatibility. J Card Surg 1987;2(1 Suppl):209-20. 17. Hansson GK, Jonasson L, Holm J, Claesson-Welsh L. Class II MHC antigen expression in the atherosclerotic plaque: smooth muscle cells express HLA-DR, HLA-DQ and the invariant gamma chain. Clin Exp Immunol 1986; 64:261-8. 18. Pober JS, Collins T, Gimbrone MA Jr, Libby P, Reiss CS. Inducible expression of class II major histocompatibility complex antigens and the immunogenicity of vascular endothelium. Transplantation 1986;41 :141-6. 19. Guinan EC, Smith BR, Doukas JT, Miller RA, Pober JS. Vascular endothelial cells enhance T cell responses by markedly augmenting IL-2 concentrations. Cell Immunol 1989;118(1):166-77.