Atherosclerosis, 97 ( 1992) 97- 106 0 1992 Elsevier Scientific Publishers Printed and Published in Ireland
ATHERO
91 Ireland,
Ltd. All rights reserved.
0021-9150/92/$05.00
04922
Homocysteine catabolism: levels of 3 enzymes in cultured human vascular endothelium and their relevance to vascular disease Jian Wang, Nicholas
P.B. Dudman,
David E.L. Wilcken
and Judith
Department of Medicine, The Prince Henry Hospilal, University of New South Wales, Little Bay (Sydney), (Australia)
F. Lynch New South Wales 2036
(Received 7 April, 1992) (Revised. received 29 July, 1992) (Accepted 3 August, 1992)
Summary
Elevated plasma homocysteine enhances the risk of thrombosis and premature arteriosclerosis. We have assessed the activity of the 3 prime enzymes of homocysteine metabolism in cultured human venous endothelial cells, in a study of their possible protective roles. In cells from 4 individuals, cultured in Dulbecco’s modified Eagle medium, the mean activity f S.D. of cystathionine /3-synthase (nmol of product/h per mg of cell protein, at 37°C) was 3.58 f 3.11 at pH 8.6. The assay used was our newly developed amino acid analyser-based procedure. The activity of 5-methyltetrahydrofolate:homocysteine methyltransferase at pH 7.4 was 4.12 + 1.25 and betaine:homocysteine methyltransferase (BHMT) was undetectable (< 1.4 nmol/h per mg protein). Cells were also cultured in a medium aimed at stimulating methionine biosynthesis, containing methionine-deficient Dulbecco’s modified Eagle medium to which r_-homocystine (100 pmol/l) and methylcobalamin (1 pmol/l) had been added. In these cells 5methyltetrahydrofolate:homocysteine methyltransferase activity increased to 7.95 f 1.45, P < 0.001, there was a non-significant decrease in cystathionine P-synthase activity to 2.16 f 1.52 and BHMT activity was still undetectable. These cells were more resistant to in vitro homocysteine-induced detachment than were cells from the same line cultured in Dulbecco’s modified Eagle medium alone. Our findings establish that human endothelial cells express 2 of the 3 primary enzymes of homocysteine catabolism. They suggest that persons who are deficient in cystathionine /3-synthase or 5-methyltetrahydrofolate:homocysteine methyltransferase activity may not only develop homocysteinemia, but also have vascular endothelium which is more susceptible to damage by homocysteine than persons with normal enzyme levels. Key words: Homocysteine catabolism; Endothelial cells; Cystathionine /3-synthase; 5_Methyltetrahydrofolate:homocysteine methyltransferase; Betaine:homocysteine methyltransferase; Premature vascular disease; Homocystinuria Correspondence to: Dr Nicholas P.B. Dudman, Department of Medicine, The Prince Henry Hospital, Little Bay (Sydney), New South Wales 2036, Australia. Tel.: Australia 02 694 5694; Fax: Australia 02 311 3483.
98
Introduction
Thrombosis and premature arteriosclerosis are regular features of homocystinuria due to recessively inherited inborn errors of homocysteine catabolism [ 11.These errors most generally include deficiency of cystathionine /3-synthase activity and more rarely various flaws in the metabolism of folates and vitamin B12 which lead to deficient remethylation of homocysteine to methionine. Despite the autosomal recessive inheritance, heterozygosity appears to be linked with an increased risk of premature vascular disease [2]. The possibility that vascular lesions in homocystinuria may be induced by dysfunction of the endothelium is supported by several strands of evidence, including: (i) The widespread distribution of both thrombosis and premature arteriosclerosis in homocystinuria; (ii) homocysteine-induced changes in cultured vascular endothelial cells including greater fragility [3] and thrombosispromoting responses in prostacyclin output [4], activation of factor V [5] and activation of protein C [6]; (iii) endothelial desquamation and presumably dysfunction of non-desquamated endothelium in baboons infused with homocysteine (as the thiolactone) [7]; (iv) the finding that vascular endothelial cells cultured from a heterozygote for cystathionine fl-synthase deliciency were more sensitive than were control cells in vitro to damage by homocystine and methionine, both of which would be expected to be metabolised intracellularly to homocysteine [8]. These findings suggested to us that normal endothelium could have the ability to metabolise homocysteine, enabling the endothelium in normal subjects to metabolise and remove potentially harmful levels of homocysteine. In patients with homocystinuria and in heterozygotes, the already elevated circulating levels of homocysteine might remain unmetabolised in the vascular endothelium, posing a chronic vascular risk: In this paper we have explored that possibility by assessing, in cultured human vascular endothelial cells, three key enzymes of homocysteine catabolism [ 11. These are cystathionine &synthase (CS; EC 4.2.1.22), which condenses homocysteine with serine to form cystathionine, and the two enzymes which remethylate homocysteine to form
methionine, 5-methyltetrahydrofolate:homocysteine methyltransferase (FHMT; EC 2.1.1.13) and betaine:homocysteine methyltransferase (BHMT; EC 2.1.1.5). A fourth enzyme, $lO-methylenetetratrahydrofolate reductase, which is also involved in the folate dependent homocysteine remethylation, is the subject of separate study. We have explored the question of whether changes in the cellular activity levels of these enzymes could be induced by culture in an appropriate medium. In addition to culture in the previously employed Dulbecco’s modified Eagle medium [9], we cultured vascular endothelial cells for several cell divisions in medium supplemented with homocystine and vitamin B12, but with lower concentrations of methionine, prior to determining their enzyme levels. The latter medium parallels the lower circulating levels of methionine and elevated homocysteine found in patients with homocystinuria due to a deficiency of homocysteine remethylation [ 11. Having identified changes in cellular enzyme levels in response to this altered medium, we then studied the effects of these changes on the known [3] detachment of endothelial cells from their substrate induced by homocysteine in vitro. Materials and Methods
Dulbecco’s modified Eagle medium (DMEM) and Dulbecco’s phosphate-buffered saline (Ca2+from Flow and Mg 2+-free) were obtained Laboratories Co. (McLean, VA). DMEM without methionine and human serum fibronectin were obtained from Cytosystems Co. (Sydney, Australia). Amino acid analysis confirmed that the amino acid contents of DMEM and DMEM without methionine were essentially identical with the exception of their methionine content. L-Homocystine, sodium heparin from bovine lung, sodium pyruvate, serine, betaine, S-adenosylmethionine and methylcobalamin were supplied by Sigma Chemical Co. (MO). D,L-Homocysteine was obtained from Fluka Co. (Switzerland). Endothelial cell growth factor (ECGF) was purchased from Boehringer-Mannheim (FRG). Fetal calf serum was obtained from Commonwealth Serum Laboratories Co. (Melbourne, Australia). 5[Methyl-‘4C]methyltetrahydrofolic acid, barium salt and P.C.S. scintillant (code No. 196097) were
99
obtained from Amersham Co. (Buckinghamshire, England). &Mercaptoethanol was bought from BDH Chemicals (Australia). Bio-Rad AG- 1X8 anion exchange resin was obtained from Bio-Rad Laboratories (Richmond, CA). Tissue culture flasks and plates were bought from Nunc Co. (Roskilde, Denmark) and from Flow Laboratories Co.. Fresh chicken liver was supplied by Inghams Enterprises Pty. Ltd. (Sydney, Australia). Human umbilical cords were obtained from natural or caesarian births at Sutherland Hospital, Sydney. Lithium S buffer was supplied by Beckman Instruments Inc. (Palo Alto, CA). Culture of endothelial cells
Venous endothelial cells were isolated from 6 human umbilical cords derived from mature normal newborns by collagenase treatment of vessels [lo], and cultured in tissue culture flasks precoated with human serum fibronectin [ll]. The cell lines were designated A23, A33, ECJL, ECl, EC2 and EC4. Cells were cultured in DMEM containing 20 mg/l ECGF, 100 mgil heparin, 15 mmol/l HEPES, penicillin (50 I.U./ml), streptomycin (50 &ml), NaHC03 (44 mmolil) and 16% fetal calf serum, in a 5% COz/air atmosphere. Where stated, cells were cultured in DMEM-H in which the above mixture was prepared using methionine-free DMEM and was supplemented with 100 pmol/l L-homocystine and 1 pmol/l methylcobalamin.
used as a control in assays determining the activity of the enzymes. Protein concentrations were measured by the Lowry procedure [12]. Enzyme activities are expressed here as nmol of product formed/h per mg protein and all values are mean f SD.. When the performance of a particular line of cells cultured in DMEM was compared with the performance of cells of the same line cultured in DMEM-H, cells of the same passage number were used in both media. CS assay The incubation mixture for the CS assay in 0.2 ml contained 2.5 mmol/l L-serine, 0.1 mmol/l Tris-HCl buffer (pH 8.6), 30 mmol/l D,Lhomocysteine and 400 pg protein added last. After 2 h incubation at 37°C 50 ~1 of 200/o sulphosalicylic acid was added, containing 0.2 mmol/l y-aminobutyric acid as an internal standard. Cystathionine and y-aminobutyric acid concentrations in the samples were then quantitated by amino acid analysis (see below). BHMT assay
BHMT activity was assayed by production of methionine from homocysteine and betaine, followed by performic acid oxidation of the products and quantitation by amino acid analysis of the methionine sulphone which was produced, as reported elsewhere [ 131. Amino acid analysis
Preparation of enzymatic extracts
After culture in DMEM or DMEM-H for 14 days, the human umbilical vein endothelial cells at passage lo-13 were trypsinised, collected in medium containing 16% serum and washed twice with Dulbecco’s phosphate-buffered saline and resuspended in 0.05 mmol/l sodium phosphate buffer, pH 7.4. The cells were sonicated using a Sonifer 250 sonicator with a tapered micro tip and centrifuged at 23 000 x g for 10 min at 4°C. The supernatant was used as the enzyme source. Fresh chicken liver (5 g) was homogenised in sodium phosphate buffer (pH 7.4, 50 mmol/l, 20 ml, 4°C) and centrifuged (23 000 x g, 10 min, 4°C). The supernatant was stored at -70°C. Chicken liver supernatant possessed substantial activity of each of the 3 enzymes studied and was
Sample solutions were centrifuged (21 000 x g for 10 min at 4°C) and the supernatants were filtered (0.22 pm filter). Aliquots of 100 ~1 were chromatographed in a Beckman 6300 automated amino acid analyzer fitted with a 25-cm lithium column (type 338050) and eluted with Beckman lithium citrate buffers A, D and E. Chromatographic data were collected and analysed using a Shimadzu C-R4A Chromatopac integrator. The quantities of amino acids in each assay supernatant were calculated based on the recovery of yaminobutyric acid in that supernatant. In the CS assays, some samples were chromatographed several times under different conditions, to ensure that the cystathionine peak did not coincide with any other peaks as had been reported to occur previously [ 141.
100
FHMT
assay
FHMT activity was determined by the formation of [ “C]methionine from 5-[methyl- 14C]methyltetrahydrofolate and homocysteine in the dark as described [ 151.
assay chromatographs after a period of incubation of the assay mixture, which required the simultaneous presence of homocysteine, serine and an enzyme source. Our BHMT assay has already been similarly validated [ 131.
Endothelial cell detachment
Enzymatic activities
Endothelial cells (ECl) which had been cultured for 18 days in DMEM or DMEM-H were trypsinised and plated into tibronectin-treated tissue culture multi-well plates at 2.6 x lo4 per well in 0.5 ml medium supplemented with serum, antibiotics and 10 mmol/l D,L-homocysteine. After incubation for 24 h, photomicrographs of the cells were taken using a Leitz microscope with phasecontrast optics. The strength of cellular attachment to tissue culture dishes of cells treated in this way was tested by a detachment assay [16] in which the cells were reproducibly sluiced with saline from a specially designed nozzle. Endothelial cells from 2 other cell lines (A23 and A33) were cultured for 14 days in DMEM or DMEM-H, after which A33 cells were exposed to a range of concentrations of D,L-homocysteine for 24 h, and A23 cells were exposed to 5 mmoY1 D,Lhomocysteine for different periods of time up to 48 h. Following these homocysteine treatments, the strength of attachment of each cell line to the substrate was assessed using the sluicing assay. Cells which remained attached to the multiwell surface after sluicing were trypsinised and counted. There were 6 wells of cells exposed to each different concentration of homocysteine.
Using chicken liver extract as the enzyme source, our CS assay after 120 min was linearly dependent on the amount of enzyme added, from 0 to 40.6 nmol/h. After culture in DMEM for 14 days, all endothelial cell lines expressed CS activity with a mean value f S.D. of 3.58 f 3.11. The CS activity of cells cultured in DMEM-H for 14 days decreased in 3 cell lines and increased marginally in the fourth (Fig. 1). The mean CS activity f S.D. of cells cultured in DMEM-H was 2.16 f 1.52, P = NS compared with cells grown in DMEM. FHMT activity was also expressed in all cell lines and cells cultured in DMEM-H developed
A
DMEM
A DMEM-H
A
f A
Statistics
Activity of each enzyme in each cell line was measured in duplicate. Enzyme levels in endotheha1 cell lines grown in DMEM and in DMEM-H were compared using the t-test as applied to small numbers of samples [17].
AA 0'
Results
A EC1
Endothelial
Fig. I.
In the present study CS assays were quantitated by amino acid analysis, using our previously unreported procedure. The product cystathionine was identified by its elution time and peak shape in comparison with authentic L-cystathionine and because a peak of cystathionine only occurred in
ECJL
B
EC2
cell
EC4
lines
CS activity in extracts of cultured human venous endothelial cells prepared from 4 different umbilical cords. Cells were extracted and assayed after culture in Dulbecco’s modified Eagle medium (A) or in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12 (A). Each cell line, whether cultured in either medium, was assayed in duplicate. In some instances duplicate results overlapped substantially.
101
almost double the activity of those cultured in DMEM. The mean f S.D. FHMT activities of cells cultured in DMEM and DMEM-H, respectively, were 4.12 f 1.25 and 7.95 f 1.45, P < 0.01 (Fig. 2). No cell line expressed detectable BHMT activity whether cultured in DMEM or DMEM-H. Endothelial cell detachment We obtained evidence by micrography, dose response curves and kinetic measurements, all of which showed that when our cultured endothelial cells were exposed to homocysteine in medium, the cells which had previously been cultured in DMEM-H were more resistant to homocyteineinduced damage or detachment than were cells cultured in DMEM. Cells from line EC1 were passaged at the same density in DMEM and DMEM-H. After culture for 18 days, the same cell numbers were obtained from DMEM and DMEM-H, indicating that DMEM-H did not affect the rate of cell replication. However, cells in 3 wells which were cultured in DMEM were seriously damaged by a 24-h exposure to 10 mmol/l homocysteine (Fig. 3a). Almost 100% of these cells
A A
have rounded up, become detached and probably are in the process of apoptosis, since numbers of small vesicles are present. By contrast, cells cultured in DMEM-H are still almost fully attached to the culture dish (Fig. 3b) and show little rounding up activity. The strength of attachment of these cells to their multiwell surface was tested also by sluicing. When EC1 cells previously cultured in DMEM were exposed to 10 mmol/l D,L-homocysteine for 24 h prior to sluicing, detachment on slucing was 50.3% f 9.6% (n = 5) greater than in cells not exposed to homocysteine. Cells which had previously been cultured in DMEM-H showed no homocysteine-enhanced detachment even at 10 mmol/l homocysteine, while homocysteine at 3 mmol/l and below was ineffectual with cells in either DMEM or DMEM-H. A repeat experiment with another endothelial cell line (A33) resulted in a dose response curve for D,L-homocysteine (Fig. 4) in which homocysteine concentrations required to produce the same percentage of detachment were approximately 3.9 times higher for endothelial cells cultured in DMEM-H than for DMEM-cultured cells. With this cell line also, visual monitoring of cells immediately before sluicing showed that cells in medium containing 10 mmol/l D,L-homocysteine were largely detached if they had been cultured in DMEM, but remained attached and appeared essentially normal if cultured in DMEM-H. A 48-h study of the detachment kinetics of A23 cells in DMEM or DMEM-H containing 5 mmol/l D,L-homocysteine (Fig. 5) showed (i) that exposure of cells in DMEM to homocysteine followed by sluicing caused a progressive cell loss for the first 36 h and then became steady; (ii) that culture of cells in DMEM-H prevented homocysteineinduced cell detachment. Discussion
Endothelial
cell
lines
Fig. 2. Activity of FHMT in extracts of cultured human venous endothelial cells from 4 umbilical cords. Conditions are as for Fig. I.
Our results have shown that all endothelial cell lines studied expressed both CS and FHMT activities. While there was a wide range of CS activities among isolates from the various cell lines, wide inter-person variation of CS levels has been noted before in other human cell types as well, ineluding human cultured skin tibroblasts [18]. No
Fig. 3. Photomicrographs of human venous endothelial cells from cell line EC1 which have been exposed to medium containing 10 mmolil D,L-homocysteine for 24 h. Before this exposure the cells had been cultured for 18 days in either Dulbecco’s modified Eagle medium (panel A), or in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12 (panel B). Cells were washed twice in phosphate buffered saline prior to homocysteine exposure, and after it they were examined by phase contrast. Overall magnification x 1100.
cells expressed significant BHMT activity. Previously liver has been shown to express CS, FHMT and BHMT activities [14,19,20], whilst CS and FHMT are also produced in phytohaemagglutinin-stimulated peripheral blood lymphocytes [21,22] and in cultured skin libroblasts [23]. We have found that neither of these last two cell types express significant BHMT activity [ 131. Cells cultured in DMEM-H clearly responded to the changed medium by expressing increased levels of FHMT (Fig. 2). This change presumably allowed the cells in DMEM-H to continue growing at a rate identical to their rate of growth in DMEM.
Increased FHMT activity could have occurred in response to a decrease in the available methionine, or an increase in the concentration of homocystine or vitamin B12, or a combination of these factors. Because the increased levels of FHMT and the decreased levels of CS in 3 cell lines together would have the effect of conserving the available methionine in DMEM-H, the first explanation is supported. This would be in line with several studies which have shown that restriction of dietary methionine intake leads to a higher proportion of the bodily homocysteine being remethylated, rather than being processed to cysteine
103
Log concentration
of D,L-homocysteine
(mol/L)
Fig. 4. Dose response for endotheliat cells which have been exposed to various concentrations of D.L-homocysteine for 24 h and then sluiced. A, Cells cultured in Dulbecco’s modified Eagle medium; 0, cells cultured in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12. Each triangle or circle represents the mean value of 6 wells and the error bars represent + (or -) SD. Where the behaviour of cells at a particular D,L-homocysteine concentration differs sigkicantly, depending upon the culture medium used, this is denoted by *** (P < 0.001).
120
r
Time (hr) Fig. 5. Changing detachment with time of cells exposed to 5 mmoVl D.L-homocyteine and then sluiced. A, Cells cultured in Dulbecco’s modified Eagle medium; 0, cells cultured in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12. Control values were normahsed to 100%. Each point represents the mean value of 3 wells and the error bars represent + (or -) S.D. Where the behaviour of cells at 36 h or 48 h differs significantly, depending upon the culture medium used, this is denoted by ??(P < 0.05) or *** (P < 0.001).
104
and SOd2- and excreted [24]. This alteration of the metabolic balance is associated with increasing hepatic FHMT and decreasing CS activity [25]. The alternative explanation that the cells in DMEM-H have readjusted their enzyme levels to decrease the concentrations of potentially damaging homocysteine seems to be less likely. The decrease of CS levels in medium DMEM-H in 3 of the 4 cell lines does not directly conform to this hypothesis. Interestingly, the size of this decrease in cell lines ECl, EC2, and EC4 is less than the size of the increase in FHMT activity of the same cells, so that the overall effect of the changes in enzyme levels in all four cell lines is both to conserve methionine and to lower homocysteine levels simultaneously. In support of this the experiments in which endothelial cells, cultured in DMEM-H, were exposed to homocysteine for up to 48 h shows that cells cultured in DMEM-H were more resistant to homocysteine-induced damage and detachment than were cells cultured in DMEM. This induced resistance to homocysteine suggests that endothelial detachment induced by brief exposure to millimolar homocysteine, as reported earlier [3], could have been substantially diminished if the cells had previously been cultured in medium containing homocyst(e)ine. In relation to the possible response of cultured endothelial cells to the increased concentration of vitamin B12 in DMEM-H, Kamely et al. [26] have reported that total FHMT increased in cultured baby hamster kidney cells as the cobalamin concentration of the medium was raised from 0.5 to 1.5 pmol/l, with similar findings in cultured human tibroblasts. The increased concentration of vitamin B12 in DMEM-H may thus have contributed to the higher levels of FHMT that developed in endothelial cells in that medium. Studies conducted in both humans and laboratory animals demonstrated that deficiency of B12 is associated with increased plasma concentrations of homocysteine [27]. It has also been proposed that B12 deficiency may be related to an increased frequency of vascular disease [28]. Therefore the use of B12 may reduce cardiovascular risk in homocysteinemia patients by induction of FHMT activity. Our results add to the previous finding that
endothelial cells which are heterozygous for deliciency of cystathionine /3-synthase are more susceptible to damage following exposure to homocystine or methionine, than cells which are not genetically compromised in this way [8]. Since the present study establishes that human endothelial cells express CS activity, we can infer that homocysteine produced by metabolism of either homocystine or methionine, in the absence of adequate catabolism via cystathionine synthesis, can damage the cells. This has an important clinical consequence. In people who have a heterozygous deficiency of cystathionine fl-synthase, metabolism of high dietary methionine will lead to plasma homocysteine levels rising above normal [29]. Low levels of cystathionine /3-synthase in the endothelium of heterozygotes would be expected to result in their endothelial cells being less able to catabolise this potentially damaging compound than the endothelial cells of normal subjects with a full quota of CS activity. This could lead to endothelial changes and development of arteriosclerosis. In this context it may be significant that premature vascular disease has been commonly associated with deficiencies of either CS activity or folate-dependent remethylation [l], but not with deficiency of betaine-dependent remethylation. We suggest that this reflects an ability of vascular endothelium to protect itself from damaging concentrations of homocysteine with both CS and FHMT, whose levels may be appropriately reset if homocysteine levels rise due to BHMT deficiency. Genetically based BHMT deficiency has not been described as far as we are aware, but it would be expected to have less impact on the endothelium than does deficiency of CS or FHMT because BHMT is apparently not directly involved in this protective role. The previously puzzling homocysteinemia and enhanced vascular disease found in persons with a deficiency of folate-dependent remethylation may be explained by the substantial down-regulation of CS activity in diets low in methionine [25]. This is reflected in cells from ECl, EC2 and EC4 when exposed to decreasing concentrations of methionine (Fig. 1). The absence of an inborn deficiency of CS activity in patients with deficient folate-dependent remethylation would support the expectation of a
105
normal flow of homocysteine down the cystathionine pathway, resulting in normal circulating homocysteine levels. However, plasma levels are in fact elevated [ 11. Because of their remethylation defect, these patients have lower than normal levels of methionine in blood [l] and presumably in tissues as well. We propose that this may lead to down-regulation of CS and that the homocyteinemia in these patients may thus result from a functional deficiency of CS activity. In conclusion, we have found that venous endothelial cells cultured independently from four human umbilical cords express CS and FHMT activity, but no significant BHMT activity. Cells from all cell lines, when cultured in DMEM-H rather than DMEM, express increased FHMT activity and in three of the four cell lines decreased CS activity as well. These results suggest that persons who are heterozygous for CS or FHMT deliciencies and therefore have lower than normal activity of these enzymes may not only develop moderate homocysteinemia during periods of high dietary methionine, but also have vascular endothelium which is more susceptible to damage by homocysteine than the endothelium in persons with normal CS and FHMT activity. Patients homozygous for these enzyme deficiencies have little residual enzyme activity and in these the endothelium would be particularly vulnerable. The mechanisms described in this study could account for their precocious vascular disease. Acknowledgements
We thank the staff of Sutherland Hospital, Sydney, for their kind help in obtaining umbilical cords and MS C. Hicks for generous assistance with endothelial cell experiments. This work was supported by grants from the National Health and Medical Research Council of Australia and the Ramaciotti Foundations. References Mudd, S.H., Levy, H.L. and Skovby, F., Disorders of transsulfuration. In: &river, C.R., Beaudet, A.L., Sly, W.S., and Valle, D. (Eds.), The Metabolic Basis of Inherited Disease, 6th Edn., McGraw-Hill, New York, 1989, pp. 693-734. Clarke, R., Daly, L., Robinson, K., Naughten, E.,
3
4
5
6
7
8
9
10
II
12
13
14
15
I6
I7
18
Cahalane, S., Fowler, B. and Graham, I., Hyperhomocysteinemia: an independent risk factor for vascular disease, N. Engl. J. Med., 324 (1991) 1149. Wall, R.T., Harlan, J.M., Harker, L.A. and Striker, G.E., Homocysteine-induced endothelial cell injury in vitro: a model for the study of vascular injury, Thromb. Res., 18 (1980) 113. Panganamala, R.V., Karpen, C.W. and Merola, A.J., Peroxide mediated effects of homocysteine on arterial prostacyclin synthesis, Prost. Leuk. Med., 22 (1986) 349. Rodgers, G.M. and Kane, W.H., Activation of endogenous factor V by a homocysteine induced vascular cell activator, J. Clin. Invest., 77 (1986) 1909. Rodgers, G.M. and Conn, M.T.. Homocysteine, an atherogenic stimulus, reduces protein C activation by arterial and venous endothelial cells, Blood, 75 (1990) 895. Harker, L.A., Ross, R., Slichter, S.J. and Scott, C.R., Homocysteine-induced arteriosclerosis: the role of endothelial cell injury and platelet response in its genesis, J. Clin. Invest., 58 (1976) 731. de Groot, P., Willems, C., Boers, G.H.J., Gonsalves, M.D., van Aken, W.G. and van Mourik, J.A., Endothelial dysfunction in homocystinuria, Eur. J. Clin. Invest., 13 (1983) 405. Gospodarowicz, D. and Ill, C., Extracellular matrix and control of proliferation of vascular endothelial cells, J. Clin. Invest., 65 (1980) 1351. Jaffe, EA., Culture and identification of large vessel endothelial cells. In: Jaffe, E.A. (Ed.), Biology of Endothelial Cells, Martinus Nijhoff, Boston, 1984, pp. 1-13. Maciag, T., Hoover, G.A., Stemerman, M.B. and Weinstein, R.. Factors which stimulate the growth of human umbilical vein endothelial cells in vitro. In: Jaffe. E.A. (Ed.), Biology of Endothelial Cells, Martinus Nijhoff, Boston, 1984, pp. 87-96. Lowry, H.O., Rosebrough, W.J., Farr, A.L. et al., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265. Wang, J., Dudman, N.P.B., Lynch, J.F. and Wilcken, D.E.L., Betaine:homocysteine methyltransferase - a new assay for the liver enzyme and its absence from human skin libroblasts and peripheral blood lymphocytes. Clin. Chim. Acta, 204 (1991) 239. Gaul], G.E., Rassin, D.K. and Sturman, J.A., Enzymatic and metabolic studies of homocystinuria: effects of pyridoxine, Neuropadiatrie, I (1969) 199. Peytreman, R., Thorndyke, J. and Beck, W.S.. Studies on NS-methyltetrahydrofolate:homocysteine methyltransferase in normal and leukemic lymphocytes, J. Clin. Invest., 56 (1975) 1293. Dudman, N.P.B., Hicks, C. and Wilcken, D.E.L., A sluicing device for reproducibly measuring the strength of attachment of cultured cells to their substrate, Anal. Biochem., 164 (1988) 35. Bland, M. (Ed.), Analysis of the Means of Small Samples Using the t Distribution, Oxford University Press. Oxford, 1987, pp. 165-187. McGill, J.J., Mettler, G., Rosenblatt, D.S. and Striver, C.R., Detection of heterozygotes for recessive alleles,
106 Homocyst(e)inemia: paradigm of pitfalls in phenotypes, Am. J. Med. Genet., 36 (1990) 45. 19 Gaull, G.E., Berg, W.V., Raiha, N.C.R. and Sturman, J.A., Development of methyltransferase activities of human fetal tissues, Pediatr. Res., 7 (1973) 527. 20 Skiba, W.E., Wells, MS., Mangum, J.H. and Awad, W.M., Jr., Betaine-homocysteine Smethyltransferase (human), Methods Enzymol., 143 (1987) 384. 21 Hallam, L.J., Sawyer, M., Clark, A.C.L. and Van Der neonatal Weyden, M.B., Vitamin Bl2-responsive megaloblastic anemia and homocystinuria wish associated reduced methionine synthase activity, Blood, 69 (1987) 1128. 22 Goldstein, J.L., Campbell, B.K. and Gartler, SM., Cystathionine synthase activity in human lymphocytes: induction by phytohemagglutinin, J. Clin. Invest., 51 (1972) 1034. 23 Uhlendorf, B.W., Conerly, E.B. and Mudd, S.H., Homocystinuria: studies in tissue culture, Pediatr. Res., 7 (1973) 645. J.D. and Martin, J.J., Methionine 24 Finkelstein, metabolism in mammals - distribution of homocysteine
between competing pathways, J. Biol. Chem., 259 (1984) 9508. 25 Finkelstein, J.D., Methionine metabolism in mammals, J. Nutr. Biochem., 1 (1990) 228. 26 Kamely, D., Littletield, J.W., Erbe, R.W., Regulation of 5-methyltetrahydrofolate:homocysteine methyltransferase activity by methionine, vitamin B12 and folate in cultured baby hamster kidney cells, Proc. Nat. Acad. Sci. USA, 70 (1973) ,2585. 27 Selhub, J. and Miller, J.W., The pathogenesis of homocysteinemia: interruption of the coordinate regulation by 5remethylation adenosylmethionine of the and transsulfuration of homocysteine, Am. J. Clin. Nutr., 55 (1992) 131. 28 Brattstrom, L., Israelsson, B., Lindegarde, F. and H&berg, B., Higher total plasma homocysteine in vitamin B12 deficiency than in heterozygosity for homocystinuria due to cystathionine /T-synthase deficiency, Metabolism, 37 (1988) 175. 29 Wilcken, D.E.L., Reddy, G.S.R. and Gupta, V.J., Homocysteinemia, ischemic heart disease and the carrier state for homocystinuria, Metabolism, 32 (1983) 363.
ATHERO
91 Ireland,
Ltd. All rights reserved.
0021-9150/92/$05.00
04922
Homocysteine catabolism: levels of 3 enzymes in cultured human vascular endothelium and their relevance to vascular disease Jian Wang, Nicholas
P.B. Dudman,
David E.L. Wilcken
and Judith
Department of Medicine, The Prince Henry Hospilal, University of New South Wales, Little Bay (Sydney), (Australia)
F. Lynch New South Wales 2036
(Received 7 April, 1992) (Revised. received 29 July, 1992) (Accepted 3 August, 1992)
Summary
Elevated plasma homocysteine enhances the risk of thrombosis and premature arteriosclerosis. We have assessed the activity of the 3 prime enzymes of homocysteine metabolism in cultured human venous endothelial cells, in a study of their possible protective roles. In cells from 4 individuals, cultured in Dulbecco’s modified Eagle medium, the mean activity f S.D. of cystathionine /3-synthase (nmol of product/h per mg of cell protein, at 37°C) was 3.58 f 3.11 at pH 8.6. The assay used was our newly developed amino acid analyser-based procedure. The activity of 5-methyltetrahydrofolate:homocysteine methyltransferase at pH 7.4 was 4.12 + 1.25 and betaine:homocysteine methyltransferase (BHMT) was undetectable (< 1.4 nmol/h per mg protein). Cells were also cultured in a medium aimed at stimulating methionine biosynthesis, containing methionine-deficient Dulbecco’s modified Eagle medium to which r_-homocystine (100 pmol/l) and methylcobalamin (1 pmol/l) had been added. In these cells 5methyltetrahydrofolate:homocysteine methyltransferase activity increased to 7.95 f 1.45, P < 0.001, there was a non-significant decrease in cystathionine P-synthase activity to 2.16 f 1.52 and BHMT activity was still undetectable. These cells were more resistant to in vitro homocysteine-induced detachment than were cells from the same line cultured in Dulbecco’s modified Eagle medium alone. Our findings establish that human endothelial cells express 2 of the 3 primary enzymes of homocysteine catabolism. They suggest that persons who are deficient in cystathionine /3-synthase or 5-methyltetrahydrofolate:homocysteine methyltransferase activity may not only develop homocysteinemia, but also have vascular endothelium which is more susceptible to damage by homocysteine than persons with normal enzyme levels. Key words: Homocysteine catabolism; Endothelial cells; Cystathionine /3-synthase; 5_Methyltetrahydrofolate:homocysteine methyltransferase; Betaine:homocysteine methyltransferase; Premature vascular disease; Homocystinuria Correspondence to: Dr Nicholas P.B. Dudman, Department of Medicine, The Prince Henry Hospital, Little Bay (Sydney), New South Wales 2036, Australia. Tel.: Australia 02 694 5694; Fax: Australia 02 311 3483.
98
Introduction
Thrombosis and premature arteriosclerosis are regular features of homocystinuria due to recessively inherited inborn errors of homocysteine catabolism [ 11.These errors most generally include deficiency of cystathionine /3-synthase activity and more rarely various flaws in the metabolism of folates and vitamin B12 which lead to deficient remethylation of homocysteine to methionine. Despite the autosomal recessive inheritance, heterozygosity appears to be linked with an increased risk of premature vascular disease [2]. The possibility that vascular lesions in homocystinuria may be induced by dysfunction of the endothelium is supported by several strands of evidence, including: (i) The widespread distribution of both thrombosis and premature arteriosclerosis in homocystinuria; (ii) homocysteine-induced changes in cultured vascular endothelial cells including greater fragility [3] and thrombosispromoting responses in prostacyclin output [4], activation of factor V [5] and activation of protein C [6]; (iii) endothelial desquamation and presumably dysfunction of non-desquamated endothelium in baboons infused with homocysteine (as the thiolactone) [7]; (iv) the finding that vascular endothelial cells cultured from a heterozygote for cystathionine fl-synthase deliciency were more sensitive than were control cells in vitro to damage by homocystine and methionine, both of which would be expected to be metabolised intracellularly to homocysteine [8]. These findings suggested to us that normal endothelium could have the ability to metabolise homocysteine, enabling the endothelium in normal subjects to metabolise and remove potentially harmful levels of homocysteine. In patients with homocystinuria and in heterozygotes, the already elevated circulating levels of homocysteine might remain unmetabolised in the vascular endothelium, posing a chronic vascular risk: In this paper we have explored that possibility by assessing, in cultured human vascular endothelial cells, three key enzymes of homocysteine catabolism [ 11. These are cystathionine &synthase (CS; EC 4.2.1.22), which condenses homocysteine with serine to form cystathionine, and the two enzymes which remethylate homocysteine to form
methionine, 5-methyltetrahydrofolate:homocysteine methyltransferase (FHMT; EC 2.1.1.13) and betaine:homocysteine methyltransferase (BHMT; EC 2.1.1.5). A fourth enzyme, $lO-methylenetetratrahydrofolate reductase, which is also involved in the folate dependent homocysteine remethylation, is the subject of separate study. We have explored the question of whether changes in the cellular activity levels of these enzymes could be induced by culture in an appropriate medium. In addition to culture in the previously employed Dulbecco’s modified Eagle medium [9], we cultured vascular endothelial cells for several cell divisions in medium supplemented with homocystine and vitamin B12, but with lower concentrations of methionine, prior to determining their enzyme levels. The latter medium parallels the lower circulating levels of methionine and elevated homocysteine found in patients with homocystinuria due to a deficiency of homocysteine remethylation [ 11. Having identified changes in cellular enzyme levels in response to this altered medium, we then studied the effects of these changes on the known [3] detachment of endothelial cells from their substrate induced by homocysteine in vitro. Materials and Methods
Dulbecco’s modified Eagle medium (DMEM) and Dulbecco’s phosphate-buffered saline (Ca2+from Flow and Mg 2+-free) were obtained Laboratories Co. (McLean, VA). DMEM without methionine and human serum fibronectin were obtained from Cytosystems Co. (Sydney, Australia). Amino acid analysis confirmed that the amino acid contents of DMEM and DMEM without methionine were essentially identical with the exception of their methionine content. L-Homocystine, sodium heparin from bovine lung, sodium pyruvate, serine, betaine, S-adenosylmethionine and methylcobalamin were supplied by Sigma Chemical Co. (MO). D,L-Homocysteine was obtained from Fluka Co. (Switzerland). Endothelial cell growth factor (ECGF) was purchased from Boehringer-Mannheim (FRG). Fetal calf serum was obtained from Commonwealth Serum Laboratories Co. (Melbourne, Australia). 5[Methyl-‘4C]methyltetrahydrofolic acid, barium salt and P.C.S. scintillant (code No. 196097) were
99
obtained from Amersham Co. (Buckinghamshire, England). &Mercaptoethanol was bought from BDH Chemicals (Australia). Bio-Rad AG- 1X8 anion exchange resin was obtained from Bio-Rad Laboratories (Richmond, CA). Tissue culture flasks and plates were bought from Nunc Co. (Roskilde, Denmark) and from Flow Laboratories Co.. Fresh chicken liver was supplied by Inghams Enterprises Pty. Ltd. (Sydney, Australia). Human umbilical cords were obtained from natural or caesarian births at Sutherland Hospital, Sydney. Lithium S buffer was supplied by Beckman Instruments Inc. (Palo Alto, CA). Culture of endothelial cells
Venous endothelial cells were isolated from 6 human umbilical cords derived from mature normal newborns by collagenase treatment of vessels [lo], and cultured in tissue culture flasks precoated with human serum fibronectin [ll]. The cell lines were designated A23, A33, ECJL, ECl, EC2 and EC4. Cells were cultured in DMEM containing 20 mg/l ECGF, 100 mgil heparin, 15 mmol/l HEPES, penicillin (50 I.U./ml), streptomycin (50 &ml), NaHC03 (44 mmolil) and 16% fetal calf serum, in a 5% COz/air atmosphere. Where stated, cells were cultured in DMEM-H in which the above mixture was prepared using methionine-free DMEM and was supplemented with 100 pmol/l L-homocystine and 1 pmol/l methylcobalamin.
used as a control in assays determining the activity of the enzymes. Protein concentrations were measured by the Lowry procedure [12]. Enzyme activities are expressed here as nmol of product formed/h per mg protein and all values are mean f SD.. When the performance of a particular line of cells cultured in DMEM was compared with the performance of cells of the same line cultured in DMEM-H, cells of the same passage number were used in both media. CS assay The incubation mixture for the CS assay in 0.2 ml contained 2.5 mmol/l L-serine, 0.1 mmol/l Tris-HCl buffer (pH 8.6), 30 mmol/l D,Lhomocysteine and 400 pg protein added last. After 2 h incubation at 37°C 50 ~1 of 200/o sulphosalicylic acid was added, containing 0.2 mmol/l y-aminobutyric acid as an internal standard. Cystathionine and y-aminobutyric acid concentrations in the samples were then quantitated by amino acid analysis (see below). BHMT assay
BHMT activity was assayed by production of methionine from homocysteine and betaine, followed by performic acid oxidation of the products and quantitation by amino acid analysis of the methionine sulphone which was produced, as reported elsewhere [ 131. Amino acid analysis
Preparation of enzymatic extracts
After culture in DMEM or DMEM-H for 14 days, the human umbilical vein endothelial cells at passage lo-13 were trypsinised, collected in medium containing 16% serum and washed twice with Dulbecco’s phosphate-buffered saline and resuspended in 0.05 mmol/l sodium phosphate buffer, pH 7.4. The cells were sonicated using a Sonifer 250 sonicator with a tapered micro tip and centrifuged at 23 000 x g for 10 min at 4°C. The supernatant was used as the enzyme source. Fresh chicken liver (5 g) was homogenised in sodium phosphate buffer (pH 7.4, 50 mmol/l, 20 ml, 4°C) and centrifuged (23 000 x g, 10 min, 4°C). The supernatant was stored at -70°C. Chicken liver supernatant possessed substantial activity of each of the 3 enzymes studied and was
Sample solutions were centrifuged (21 000 x g for 10 min at 4°C) and the supernatants were filtered (0.22 pm filter). Aliquots of 100 ~1 were chromatographed in a Beckman 6300 automated amino acid analyzer fitted with a 25-cm lithium column (type 338050) and eluted with Beckman lithium citrate buffers A, D and E. Chromatographic data were collected and analysed using a Shimadzu C-R4A Chromatopac integrator. The quantities of amino acids in each assay supernatant were calculated based on the recovery of yaminobutyric acid in that supernatant. In the CS assays, some samples were chromatographed several times under different conditions, to ensure that the cystathionine peak did not coincide with any other peaks as had been reported to occur previously [ 141.
100
FHMT
assay
FHMT activity was determined by the formation of [ “C]methionine from 5-[methyl- 14C]methyltetrahydrofolate and homocysteine in the dark as described [ 151.
assay chromatographs after a period of incubation of the assay mixture, which required the simultaneous presence of homocysteine, serine and an enzyme source. Our BHMT assay has already been similarly validated [ 131.
Endothelial cell detachment
Enzymatic activities
Endothelial cells (ECl) which had been cultured for 18 days in DMEM or DMEM-H were trypsinised and plated into tibronectin-treated tissue culture multi-well plates at 2.6 x lo4 per well in 0.5 ml medium supplemented with serum, antibiotics and 10 mmol/l D,L-homocysteine. After incubation for 24 h, photomicrographs of the cells were taken using a Leitz microscope with phasecontrast optics. The strength of cellular attachment to tissue culture dishes of cells treated in this way was tested by a detachment assay [16] in which the cells were reproducibly sluiced with saline from a specially designed nozzle. Endothelial cells from 2 other cell lines (A23 and A33) were cultured for 14 days in DMEM or DMEM-H, after which A33 cells were exposed to a range of concentrations of D,L-homocysteine for 24 h, and A23 cells were exposed to 5 mmoY1 D,Lhomocysteine for different periods of time up to 48 h. Following these homocysteine treatments, the strength of attachment of each cell line to the substrate was assessed using the sluicing assay. Cells which remained attached to the multiwell surface after sluicing were trypsinised and counted. There were 6 wells of cells exposed to each different concentration of homocysteine.
Using chicken liver extract as the enzyme source, our CS assay after 120 min was linearly dependent on the amount of enzyme added, from 0 to 40.6 nmol/h. After culture in DMEM for 14 days, all endothelial cell lines expressed CS activity with a mean value f S.D. of 3.58 f 3.11. The CS activity of cells cultured in DMEM-H for 14 days decreased in 3 cell lines and increased marginally in the fourth (Fig. 1). The mean CS activity f S.D. of cells cultured in DMEM-H was 2.16 f 1.52, P = NS compared with cells grown in DMEM. FHMT activity was also expressed in all cell lines and cells cultured in DMEM-H developed
A
DMEM
A DMEM-H
A
f A
Statistics
Activity of each enzyme in each cell line was measured in duplicate. Enzyme levels in endotheha1 cell lines grown in DMEM and in DMEM-H were compared using the t-test as applied to small numbers of samples [17].
AA 0'
Results
A EC1
Endothelial
Fig. I.
In the present study CS assays were quantitated by amino acid analysis, using our previously unreported procedure. The product cystathionine was identified by its elution time and peak shape in comparison with authentic L-cystathionine and because a peak of cystathionine only occurred in
ECJL
B
EC2
cell
EC4
lines
CS activity in extracts of cultured human venous endothelial cells prepared from 4 different umbilical cords. Cells were extracted and assayed after culture in Dulbecco’s modified Eagle medium (A) or in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12 (A). Each cell line, whether cultured in either medium, was assayed in duplicate. In some instances duplicate results overlapped substantially.
101
almost double the activity of those cultured in DMEM. The mean f S.D. FHMT activities of cells cultured in DMEM and DMEM-H, respectively, were 4.12 f 1.25 and 7.95 f 1.45, P < 0.01 (Fig. 2). No cell line expressed detectable BHMT activity whether cultured in DMEM or DMEM-H. Endothelial cell detachment We obtained evidence by micrography, dose response curves and kinetic measurements, all of which showed that when our cultured endothelial cells were exposed to homocysteine in medium, the cells which had previously been cultured in DMEM-H were more resistant to homocyteineinduced damage or detachment than were cells cultured in DMEM. Cells from line EC1 were passaged at the same density in DMEM and DMEM-H. After culture for 18 days, the same cell numbers were obtained from DMEM and DMEM-H, indicating that DMEM-H did not affect the rate of cell replication. However, cells in 3 wells which were cultured in DMEM were seriously damaged by a 24-h exposure to 10 mmol/l homocysteine (Fig. 3a). Almost 100% of these cells
A A
have rounded up, become detached and probably are in the process of apoptosis, since numbers of small vesicles are present. By contrast, cells cultured in DMEM-H are still almost fully attached to the culture dish (Fig. 3b) and show little rounding up activity. The strength of attachment of these cells to their multiwell surface was tested also by sluicing. When EC1 cells previously cultured in DMEM were exposed to 10 mmol/l D,L-homocysteine for 24 h prior to sluicing, detachment on slucing was 50.3% f 9.6% (n = 5) greater than in cells not exposed to homocysteine. Cells which had previously been cultured in DMEM-H showed no homocysteine-enhanced detachment even at 10 mmol/l homocysteine, while homocysteine at 3 mmol/l and below was ineffectual with cells in either DMEM or DMEM-H. A repeat experiment with another endothelial cell line (A33) resulted in a dose response curve for D,L-homocysteine (Fig. 4) in which homocysteine concentrations required to produce the same percentage of detachment were approximately 3.9 times higher for endothelial cells cultured in DMEM-H than for DMEM-cultured cells. With this cell line also, visual monitoring of cells immediately before sluicing showed that cells in medium containing 10 mmol/l D,L-homocysteine were largely detached if they had been cultured in DMEM, but remained attached and appeared essentially normal if cultured in DMEM-H. A 48-h study of the detachment kinetics of A23 cells in DMEM or DMEM-H containing 5 mmol/l D,L-homocysteine (Fig. 5) showed (i) that exposure of cells in DMEM to homocysteine followed by sluicing caused a progressive cell loss for the first 36 h and then became steady; (ii) that culture of cells in DMEM-H prevented homocysteineinduced cell detachment. Discussion
Endothelial
cell
lines
Fig. 2. Activity of FHMT in extracts of cultured human venous endothelial cells from 4 umbilical cords. Conditions are as for Fig. I.
Our results have shown that all endothelial cell lines studied expressed both CS and FHMT activities. While there was a wide range of CS activities among isolates from the various cell lines, wide inter-person variation of CS levels has been noted before in other human cell types as well, ineluding human cultured skin tibroblasts [18]. No
Fig. 3. Photomicrographs of human venous endothelial cells from cell line EC1 which have been exposed to medium containing 10 mmolil D,L-homocysteine for 24 h. Before this exposure the cells had been cultured for 18 days in either Dulbecco’s modified Eagle medium (panel A), or in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12 (panel B). Cells were washed twice in phosphate buffered saline prior to homocysteine exposure, and after it they were examined by phase contrast. Overall magnification x 1100.
cells expressed significant BHMT activity. Previously liver has been shown to express CS, FHMT and BHMT activities [14,19,20], whilst CS and FHMT are also produced in phytohaemagglutinin-stimulated peripheral blood lymphocytes [21,22] and in cultured skin libroblasts [23]. We have found that neither of these last two cell types express significant BHMT activity [ 131. Cells cultured in DMEM-H clearly responded to the changed medium by expressing increased levels of FHMT (Fig. 2). This change presumably allowed the cells in DMEM-H to continue growing at a rate identical to their rate of growth in DMEM.
Increased FHMT activity could have occurred in response to a decrease in the available methionine, or an increase in the concentration of homocystine or vitamin B12, or a combination of these factors. Because the increased levels of FHMT and the decreased levels of CS in 3 cell lines together would have the effect of conserving the available methionine in DMEM-H, the first explanation is supported. This would be in line with several studies which have shown that restriction of dietary methionine intake leads to a higher proportion of the bodily homocysteine being remethylated, rather than being processed to cysteine
103
Log concentration
of D,L-homocysteine
(mol/L)
Fig. 4. Dose response for endotheliat cells which have been exposed to various concentrations of D.L-homocysteine for 24 h and then sluiced. A, Cells cultured in Dulbecco’s modified Eagle medium; 0, cells cultured in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12. Each triangle or circle represents the mean value of 6 wells and the error bars represent + (or -) SD. Where the behaviour of cells at a particular D,L-homocysteine concentration differs sigkicantly, depending upon the culture medium used, this is denoted by *** (P < 0.001).
120
r
Time (hr) Fig. 5. Changing detachment with time of cells exposed to 5 mmoVl D.L-homocyteine and then sluiced. A, Cells cultured in Dulbecco’s modified Eagle medium; 0, cells cultured in this medium made deficient in methionine but supplemented with L-homocystine and vitamin B12. Control values were normahsed to 100%. Each point represents the mean value of 3 wells and the error bars represent + (or -) S.D. Where the behaviour of cells at 36 h or 48 h differs significantly, depending upon the culture medium used, this is denoted by ??(P < 0.05) or *** (P < 0.001).
104
and SOd2- and excreted [24]. This alteration of the metabolic balance is associated with increasing hepatic FHMT and decreasing CS activity [25]. The alternative explanation that the cells in DMEM-H have readjusted their enzyme levels to decrease the concentrations of potentially damaging homocysteine seems to be less likely. The decrease of CS levels in medium DMEM-H in 3 of the 4 cell lines does not directly conform to this hypothesis. Interestingly, the size of this decrease in cell lines ECl, EC2, and EC4 is less than the size of the increase in FHMT activity of the same cells, so that the overall effect of the changes in enzyme levels in all four cell lines is both to conserve methionine and to lower homocysteine levels simultaneously. In support of this the experiments in which endothelial cells, cultured in DMEM-H, were exposed to homocysteine for up to 48 h shows that cells cultured in DMEM-H were more resistant to homocysteine-induced damage and detachment than were cells cultured in DMEM. This induced resistance to homocysteine suggests that endothelial detachment induced by brief exposure to millimolar homocysteine, as reported earlier [3], could have been substantially diminished if the cells had previously been cultured in medium containing homocyst(e)ine. In relation to the possible response of cultured endothelial cells to the increased concentration of vitamin B12 in DMEM-H, Kamely et al. [26] have reported that total FHMT increased in cultured baby hamster kidney cells as the cobalamin concentration of the medium was raised from 0.5 to 1.5 pmol/l, with similar findings in cultured human tibroblasts. The increased concentration of vitamin B12 in DMEM-H may thus have contributed to the higher levels of FHMT that developed in endothelial cells in that medium. Studies conducted in both humans and laboratory animals demonstrated that deficiency of B12 is associated with increased plasma concentrations of homocysteine [27]. It has also been proposed that B12 deficiency may be related to an increased frequency of vascular disease [28]. Therefore the use of B12 may reduce cardiovascular risk in homocysteinemia patients by induction of FHMT activity. Our results add to the previous finding that
endothelial cells which are heterozygous for deliciency of cystathionine /3-synthase are more susceptible to damage following exposure to homocystine or methionine, than cells which are not genetically compromised in this way [8]. Since the present study establishes that human endothelial cells express CS activity, we can infer that homocysteine produced by metabolism of either homocystine or methionine, in the absence of adequate catabolism via cystathionine synthesis, can damage the cells. This has an important clinical consequence. In people who have a heterozygous deficiency of cystathionine fl-synthase, metabolism of high dietary methionine will lead to plasma homocysteine levels rising above normal [29]. Low levels of cystathionine /3-synthase in the endothelium of heterozygotes would be expected to result in their endothelial cells being less able to catabolise this potentially damaging compound than the endothelial cells of normal subjects with a full quota of CS activity. This could lead to endothelial changes and development of arteriosclerosis. In this context it may be significant that premature vascular disease has been commonly associated with deficiencies of either CS activity or folate-dependent remethylation [l], but not with deficiency of betaine-dependent remethylation. We suggest that this reflects an ability of vascular endothelium to protect itself from damaging concentrations of homocysteine with both CS and FHMT, whose levels may be appropriately reset if homocysteine levels rise due to BHMT deficiency. Genetically based BHMT deficiency has not been described as far as we are aware, but it would be expected to have less impact on the endothelium than does deficiency of CS or FHMT because BHMT is apparently not directly involved in this protective role. The previously puzzling homocysteinemia and enhanced vascular disease found in persons with a deficiency of folate-dependent remethylation may be explained by the substantial down-regulation of CS activity in diets low in methionine [25]. This is reflected in cells from ECl, EC2 and EC4 when exposed to decreasing concentrations of methionine (Fig. 1). The absence of an inborn deficiency of CS activity in patients with deficient folate-dependent remethylation would support the expectation of a
105
normal flow of homocysteine down the cystathionine pathway, resulting in normal circulating homocysteine levels. However, plasma levels are in fact elevated [ 11. Because of their remethylation defect, these patients have lower than normal levels of methionine in blood [l] and presumably in tissues as well. We propose that this may lead to down-regulation of CS and that the homocyteinemia in these patients may thus result from a functional deficiency of CS activity. In conclusion, we have found that venous endothelial cells cultured independently from four human umbilical cords express CS and FHMT activity, but no significant BHMT activity. Cells from all cell lines, when cultured in DMEM-H rather than DMEM, express increased FHMT activity and in three of the four cell lines decreased CS activity as well. These results suggest that persons who are heterozygous for CS or FHMT deliciencies and therefore have lower than normal activity of these enzymes may not only develop moderate homocysteinemia during periods of high dietary methionine, but also have vascular endothelium which is more susceptible to damage by homocysteine than the endothelium in persons with normal CS and FHMT activity. Patients homozygous for these enzyme deficiencies have little residual enzyme activity and in these the endothelium would be particularly vulnerable. The mechanisms described in this study could account for their precocious vascular disease. Acknowledgements
We thank the staff of Sutherland Hospital, Sydney, for their kind help in obtaining umbilical cords and MS C. Hicks for generous assistance with endothelial cell experiments. This work was supported by grants from the National Health and Medical Research Council of Australia and the Ramaciotti Foundations. References Mudd, S.H., Levy, H.L. and Skovby, F., Disorders of transsulfuration. In: &river, C.R., Beaudet, A.L., Sly, W.S., and Valle, D. (Eds.), The Metabolic Basis of Inherited Disease, 6th Edn., McGraw-Hill, New York, 1989, pp. 693-734. Clarke, R., Daly, L., Robinson, K., Naughten, E.,
3
4
5
6
7
8
9
10
II
12
13
14
15
I6
I7
18
Cahalane, S., Fowler, B. and Graham, I., Hyperhomocysteinemia: an independent risk factor for vascular disease, N. Engl. J. Med., 324 (1991) 1149. Wall, R.T., Harlan, J.M., Harker, L.A. and Striker, G.E., Homocysteine-induced endothelial cell injury in vitro: a model for the study of vascular injury, Thromb. Res., 18 (1980) 113. Panganamala, R.V., Karpen, C.W. and Merola, A.J., Peroxide mediated effects of homocysteine on arterial prostacyclin synthesis, Prost. Leuk. Med., 22 (1986) 349. Rodgers, G.M. and Kane, W.H., Activation of endogenous factor V by a homocysteine induced vascular cell activator, J. Clin. Invest., 77 (1986) 1909. Rodgers, G.M. and Conn, M.T.. Homocysteine, an atherogenic stimulus, reduces protein C activation by arterial and venous endothelial cells, Blood, 75 (1990) 895. Harker, L.A., Ross, R., Slichter, S.J. and Scott, C.R., Homocysteine-induced arteriosclerosis: the role of endothelial cell injury and platelet response in its genesis, J. Clin. Invest., 58 (1976) 731. de Groot, P., Willems, C., Boers, G.H.J., Gonsalves, M.D., van Aken, W.G. and van Mourik, J.A., Endothelial dysfunction in homocystinuria, Eur. J. Clin. Invest., 13 (1983) 405. Gospodarowicz, D. and Ill, C., Extracellular matrix and control of proliferation of vascular endothelial cells, J. Clin. Invest., 65 (1980) 1351. Jaffe, EA., Culture and identification of large vessel endothelial cells. In: Jaffe, E.A. (Ed.), Biology of Endothelial Cells, Martinus Nijhoff, Boston, 1984, pp. 1-13. Maciag, T., Hoover, G.A., Stemerman, M.B. and Weinstein, R.. Factors which stimulate the growth of human umbilical vein endothelial cells in vitro. In: Jaffe. E.A. (Ed.), Biology of Endothelial Cells, Martinus Nijhoff, Boston, 1984, pp. 87-96. Lowry, H.O., Rosebrough, W.J., Farr, A.L. et al., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265. Wang, J., Dudman, N.P.B., Lynch, J.F. and Wilcken, D.E.L., Betaine:homocysteine methyltransferase - a new assay for the liver enzyme and its absence from human skin libroblasts and peripheral blood lymphocytes. Clin. Chim. Acta, 204 (1991) 239. Gaul], G.E., Rassin, D.K. and Sturman, J.A., Enzymatic and metabolic studies of homocystinuria: effects of pyridoxine, Neuropadiatrie, I (1969) 199. Peytreman, R., Thorndyke, J. and Beck, W.S.. Studies on NS-methyltetrahydrofolate:homocysteine methyltransferase in normal and leukemic lymphocytes, J. Clin. Invest., 56 (1975) 1293. Dudman, N.P.B., Hicks, C. and Wilcken, D.E.L., A sluicing device for reproducibly measuring the strength of attachment of cultured cells to their substrate, Anal. Biochem., 164 (1988) 35. Bland, M. (Ed.), Analysis of the Means of Small Samples Using the t Distribution, Oxford University Press. Oxford, 1987, pp. 165-187. McGill, J.J., Mettler, G., Rosenblatt, D.S. and Striver, C.R., Detection of heterozygotes for recessive alleles,
106 Homocyst(e)inemia: paradigm of pitfalls in phenotypes, Am. J. Med. Genet., 36 (1990) 45. 19 Gaull, G.E., Berg, W.V., Raiha, N.C.R. and Sturman, J.A., Development of methyltransferase activities of human fetal tissues, Pediatr. Res., 7 (1973) 527. 20 Skiba, W.E., Wells, MS., Mangum, J.H. and Awad, W.M., Jr., Betaine-homocysteine Smethyltransferase (human), Methods Enzymol., 143 (1987) 384. 21 Hallam, L.J., Sawyer, M., Clark, A.C.L. and Van Der neonatal Weyden, M.B., Vitamin Bl2-responsive megaloblastic anemia and homocystinuria wish associated reduced methionine synthase activity, Blood, 69 (1987) 1128. 22 Goldstein, J.L., Campbell, B.K. and Gartler, SM., Cystathionine synthase activity in human lymphocytes: induction by phytohemagglutinin, J. Clin. Invest., 51 (1972) 1034. 23 Uhlendorf, B.W., Conerly, E.B. and Mudd, S.H., Homocystinuria: studies in tissue culture, Pediatr. Res., 7 (1973) 645. J.D. and Martin, J.J., Methionine 24 Finkelstein, metabolism in mammals - distribution of homocysteine
between competing pathways, J. Biol. Chem., 259 (1984) 9508. 25 Finkelstein, J.D., Methionine metabolism in mammals, J. Nutr. Biochem., 1 (1990) 228. 26 Kamely, D., Littletield, J.W., Erbe, R.W., Regulation of 5-methyltetrahydrofolate:homocysteine methyltransferase activity by methionine, vitamin B12 and folate in cultured baby hamster kidney cells, Proc. Nat. Acad. Sci. USA, 70 (1973) ,2585. 27 Selhub, J. and Miller, J.W., The pathogenesis of homocysteinemia: interruption of the coordinate regulation by 5remethylation adenosylmethionine of the and transsulfuration of homocysteine, Am. J. Clin. Nutr., 55 (1992) 131. 28 Brattstrom, L., Israelsson, B., Lindegarde, F. and H&berg, B., Higher total plasma homocysteine in vitamin B12 deficiency than in heterozygosity for homocystinuria due to cystathionine /T-synthase deficiency, Metabolism, 37 (1988) 175. 29 Wilcken, D.E.L., Reddy, G.S.R. and Gupta, V.J., Homocysteinemia, ischemic heart disease and the carrier state for homocystinuria, Metabolism, 32 (1983) 363.
