Membrane protein damage and methylation reactions in chronic renal failure

Kidney International, Vol. 50 (1996), pp. 358—366

EDITORIAL REVIEW

Membrane protein damage and methylation reactions in chronic renal failure Natural proteins are prone, once they are synthesized in the isomerization including human epidermal growth factor, human endoplasmic reticulum, to spontaneous, non-enzymatic, molecu- growth hormone releasing hormone, calmodulin, lysozyme [1, 2]. Relevant to the mechanisms so far discussed is the spontaneous lar damage. This is represented by examples such as oxidation, glycosylation, racemization, isomerization and deamidation reac- formation of D-aspartyl residues in proteins. It is acknowledged tions, which affect a variety of different amino acid residues. Rates that a subset of these racemized residues can in fact be generated, of these non-enzymatic reactions depend on the intrinsic struc- particularly in long-lived proteins, through the slow racemization tural stability of protein molecules, as well as on the cellular and of the succinimide intermediate. D-aspartyl residues have been found to occur in the /3-amyloid core protein of Alzheimer's extracellular micro-environment. It has been proposed that these modifications in their entirety disease. In this condition it has been shown that the presence of should be referred to as "protein fatigue," and the proteins such residues can significantly affect the solubility of this protein subject to such posttranslational alterations, as "fatigue damaged and its deposition in cerebral tissues [7—I proteins" [1]. The terms are taken from the terminology used in Enzymatic repair of damaged proteins the field of aerospace engineering. Structural modifications thus determined can significantly alter Normal cells possess the enzymatic machinery to counteract the physical-chemical properties of a protein and therefore impair some of the structural and functional effects of spontaneous its function. These changes, even if minor, involving for example protein damage. It is obvious that the energy expenditure of just one residue over hundreds, are nevertheless able to trigger a synthesizing a protein ex novo is much higher than the one significant modification of function. Damage accumulation can be required to simply modify single residues of a protein. One of such dramatic in the case of long lived proteins, particularly in cells and examples is represented by the protein L-isoaspartate-(D-aspartissues where protein synthesis is permanently inactivated (such as tate) 0 methyltransferase (PCMT type II; type II MTase; EC the erythrocyte) so that damaged proteins cannot be replaced by 2.1.1.77), which selectively catalyzes the S-adenosylmethionine newly synthesized and functionally intact macromolecules. (AdoMet)-dependent methyl esterification of protein and peptide

Among the different protein fatigue damages, attention has been recently focused on the spontaneous alterations involving asparaginyl and aspartyl residues. Spontaneous deamidation of asparaginyl residues, and, less frequently, isomerization of aspartyl residues, can generate atypical L-isoaspartyl residues. The mechanism of these reactions has been clarified, and it involves the formation of a common succinimidyl intermediate, followed by the opening of the ring, yielding, in at least 70% of cases, an abnormal L-isoaspartyl residue, which is linked to the next residue

through a /3-isopeptide bond, instead of the normal a-peptide bond [1, 2]. Deamidation is capable of inducing major alterations of the three-dimensional organization of a protein molecule [1, 2].

For example, alterations associated with the formation of an L-isoaspartyl residue, such as the presence of an additional methylcne group in the protein backbone or the negative charge on the free cx-carboxyl group, can be expected to affect the local

free carboxyl groups at the level of altered L-isoaspartyl (and D-aspartyl) residues [1, 2]. Since 1886 when Willhelm His, the scientist who located the heart's bundle named after him, described the first known methyl transfer reaction, leading from pyridine to methylpyridine, more

than 100 methyltransferases have been also discovered, with hundreds of genes that should encode for the relevant enzymatic proteins. AdoMet-dependent methyltransferases are by far the most numerous group of these enzymes, so that this sulfonium compound is considered the universal methyl donor, which is probably second only to ATP as regards to the great number of reactions in which it serves as a substrate. The enzymes known to date deliver a number of different services in the various living organisms where they have been found, ranging from the biosynthesis of natural compounds and drug metabolism, to regulatory functions and housekeeping purposes (Table 1).

stability of an a helix where present. As a matter of fact the Several classes of AdoMet-dependent protein methyltransalterations induced by deamidation on protein structure and ferases have been so far identified, including, among others, four function have been extensively investigated using calmodulin as a different classes of enzymes, just to mention only those which model protein [3—6]. In addition, several other proteins have been transfer a methyl group to specific free carboxyl groups of protein reported to undergo asparaginyl deamidation and/or aspartyl and peptide substrates [1, 2]. In particular, relevant to the issue of

protein repair is the function of PCMT type II, a ubiquitous enzyme that has been recently identified to be involved in the repair of protein and peptide damages induced by the above Received for publication November 15, 1995 and in revised form March 1, 1996 Accepted for publication March 25, 1996

© 1996 by the International Society of Nephrology

mentioned spontaneous deamidation of asparaginyl or isomerization of normal aspartyl residues [1, 2, 10]. Brain and blood are, in mammals, the tissues endowed with the highest enzyme activity [2]. The enzyme is essentially cytosolic in human red cells, while a 358

Perna et at: Protein methylation in uremia

Tabk 1. Classification of AdoMet-dependent methyltransferases

• Small molecule methyltransferases

including: —amino acid MTases —phospholipid MTases ---—neurotransmitter MTases

—etc. DNA MTases

Nucleic acids__.__. RNA MTases

• Macromolecule methyltransferases

Proteins—

N-methyltransferases

0methyltransferasesa Two major groups of enzymes can be distinguished, based on the nature

of the methyl accepting substrate. Different major classes of protein methyltransferases have been characterized based on the atom to which the methyl group is transferred. Protein 0-methyl transferases () include four different subclasses of enzymes, which differ for both the nature of the

methylesterified amino acid residue and the biological function of the reaction. PCMT type II recognizes and modifies altered aspartyl residues and this reaction initiates the repair of an abnormal isoaspartyl residue.

359

methionine, which is not incorporated into proteins, since protein

synthesis is virtually absent in mature erythrocytes [22]. This feature has been extensively utilized for selectively identi'ing in vivo methyl esterified proteins through SDS/PAGE. These mainly include band 3 integral protein, as well as band 2.1 (ankyrin) and

4.1 cytoskeletal components, and calmodulin [4, 23—25]. As a matter of fact, RBCs are an excellent model system to study the reaction catalyzed by the type II PCMT, which virtually represents the only AdoMet-dependent reaction active in these cells. S-adenosylhomoeysteine (AdoHcy), the demethylated product of AdoMet, is the natural, competitive inhibitor of this reaction, as well as of all AdoMet-dependent enzymatic methylations [26—311.

All AdoMet-dependent methyltransferases are regulated in vivo by the [AdoMet/[AdoHcy] ratio. This ratio is therefore consid-

ered to be a very sensitive indicator of the presence and the degree of a potential inhibition not only of this reaction, hut also of all AdoMet-depent transmethylations. AdoHcy concentration is normally kept low, intracellularly, by its reversible enzymatic hydrolysis, catalyzed by AdoHcy hydrolase, to adenosine and homocysteine (Hey). Thermodynamics actually favor the biosynthetic reaction over hydrolysis, but what happens in vivo is the effective splitting of AdoHcy to adenosine and Hey, because these

products are rapidly metabolized (Fig. 2). Adenosine (Ado) membrane bound form has also been identified in rat tissues enters the purine nucleotide pool [26—311. including brain and kidney cortex [111. The human erythrocyte enzyme has been purified to homogeneity from human red blood

The metabolic fate of ilcy Hey can be metabolized intracellularly through several ways (Fig. 3): (i) to cystathionine by a reaction catalyzed by cystathithe homologous enzyme from other mammalian species [2, 121. onine f3-synthase (C/3S), a vitamin B6-dependent enzyme, which Two major isoenzymes with different isoelectric points have been utilizes serine as the other substrate. CPS is positively modulated isolated in different cells types including human erythrocytes. In by AdoMet. This is the transsulfuration pathway that leads to the humans their sequences show very little variation, particularly in formation of cysteine as the final product through the action of a the C-terminal region, which account for the observed physical- second enzyme, cystathionase [32, 33]; (ii) alternatively, Hey can chemical differences [13, 141. A minor isoform with an interme- be remethylated to methionine through the action of methionine diate isoelectric point between the two major isoenzymes has also synthase, a methyleobalamin-dependent enzyme. This step also been identified in human erythrocytes [10]. The red cell isoen- requires methyltetrahydrofolate (methylTHF) formed by the enzymes of PCMT type 1! display an identical substrate specificity zyme methyleneTHF reductase (MTHFR), which is negatively when assayed in vitro using purified peptide substrates [2, 151. In modulated by AdoMet [32, 33]. An impairment of either metathe intact erythrocytes, major substrates for this repair enzyme are bolic pathways, due to a deficiency of CI3S, or of MTHFR (or methionine synthase), respectively, but also to a relative or membrane proteins [1, 2]. Evidence for a role in the repair of isopeptide bonds present at absolute deficit of their relevant cofactors, can he responsible for L-isoaspartyl sites, has been obtained by independent research a defective metabolism of Hey. In this respect it has been recently proposed that hyperhomocysteinemia may also be the result of a groups [16—20]. The repair mechanism is partly enzymatic, including the methyl esterification step, and partly spontaneous, includ- disturbance of the AdoMet-dependent coordinate regulation of ing the non-enzymatic hydrolysis of the methyl ester via suecin- remethylation and transsulfuration of homocysteine, leading to imide, leading to the conversion of the L-isoaspartyl into a normal the interesting hypothesis that any disruption of either pathway L-aspartyl residue (Fig. 1). As a demonstration of the effective- will result in a general impairment of Hey metabolism [33]. (iii) cells (RBCs) [2, 121. It acts as a monomer with a molecular weight of 25 kDa, whose sequence is highly conserved when compared to

ness of this repair on protein function, is the elegant example Hey deamination was found to be negligible in humans [331. provided by Aswad's group with the calcium-binding protein Conversely, it has been recently shown that Hey can be converted calmodulin [3, 21]. If deamidated in vitro the protein loses most of to Hey thiolactone in mammalian cells, as the result of an error its biological activity, which is in turn quantitatively restored after

this protein is enzymatically methyl esterifled. Evidence of the effectiveness of this repair function have been provided by independent research groups [16, 17, 20], who have also clarified the mechanism, as it has been depicted above. Substrates and products of the PCMT type II catalyzed reaction As well as other cell types, RBCs are able to synthesize the methyl donor AdoMet from ATP and exogenously administered

editing activity of methionyl tRNA synthetase, in order to prevent Hey misincorporation within proteins [34]. Hey thiolactone, is a highly reactive, toxic metabolite, which can also be reconverted into Hey by the action of esterases of cells and plasma. Its complex

metabolism is beyond the scope of this article, however, for an exhaustive review of this topic we refer the reader to McCully [35—37],

Intracellularly produced Hey can also enter the plasma at this point, or it can be metabolized elsewhere as in the kidney, or excreted in the urine, As far as the erythrocyte is concerned, we

360

Perna et al: Protein methylation in uremia

0

H H2C C H3 0 H

N —R'

R C—N

Hg

L-succlnimidyl-peptide

CO OH

RyN-9-yN-R" L-a spa rty I -pep tide Fig. 1. Mechanism of PCMT type 11-dependent repair of isopeptide bonds in proteins and peptides. The first reaction of this mechanism is the enzymatic

methyl esterification of the free a-carboxyl group of isoaspartyl residue, yielding the isoaspartyl methyl ester of the peptide (step 1; grey-dotted background). The following steps occur spontaneously: the hydrolysis of the methyl ester through the formation of the succinimide derivative of the peptide (step 2). The hydrolysis methyl ester, on either side of the nitrogen atom, yields the "repaired" peptide, containing a normal aspartyl residue

(step 4; normal peptide within a black frame), or again the isoaspartyl residue (step 3). The yield of the repair process at each methylation/demethylation cycle is about 30%, depending on the nature of the peptide substrate. However, a quantitative conversion of the isoaspartyl peptide into the normal peptide can be obtained through repeated cycles of the whole pathway.

III.

AdoHcy Hydrolase I f

E

If

Horn Sbie

OSYfYSt

Fig. 2. AdoHcy enzymatic hydrolysis. Schematic representation of the reversible hydrolysis of AdoHcy catalyzed by AdoHcy hydrolase, yielding adenosine and homocysteine. The biosynthetic reaction is thermodynamically favored, while in vivo AdoHcy hydrolysis is

S adenosylmethionine dependent Transmethylations

Out

In

insured due to the rapid removal of its products. Intracellular accumulation of AdoHcy can be expected under conditions of defective removal of its products due, for example, to hyperhomocysteinemia.

may expect this to be by far the most important mechanism for Hey metabolism, in the presence of significant and chronic eliminating intracellularly produced Hey. in fact, Hey has been elevations of homocysteinemia. shown to leak out from RBCs in blood samples stored in vitro [38], The ways by which Hey is removed from the body are raising and since the RBC membrane is permeable to an inward flux of much interest because of a recent observation that hyperhomoHey as well [281, it has been established that the intracellular cysteinemia is an independent risk factor of vascular disease, that concentration of its derivative, AdoHcy, is also dependent on the is, acute coronary ischemia, stroke, as well as peripheral artery, concentration of Hey in the extracellular medium [28]. Therefore, but also venous thrombosis [39—431. The relationship between the level of enzymatic methylation of membrane proteins may be high homocysteine and vascular disease is as strong as the one also considered as a fine indicator of a general derangement of between smoking or high blood pressure [40]. Recent studies have

Perna et al: Protein methylation in uremia

361

Fig. 3. General pathways of homocysteine metabolism. The main ways for the intracellular formation and metabolism of homocysteine have been depicted. The intracellular space is indicated by a light-grey colored rectangle. The erythrocyte has limited metabolic capabilities (highlighted by a white area).

shown that Hey induces a proliferation of vascular smooth muscle cells, the most prominent hallmark of atherosclerosis [44], besides

promoting several other alterations of factors related to the pathogenesis of atherosclerosis, such as lipoprotein(a), oxygen radicals, etc. [45—49]. It has been postulated that several thousands of cardiovascular deaths can be prevented by lowering plasma homocysteine, either by identifying hyperhomocysteine-

Bostom et al, who demonstrated that homocysteine concentration is lower in the renal vein than in the renal artery, in the face of a very low urinary excretion [58, 59]. It is also possible that (4) remethylation to methionine may be

impaired, either in the kidney or elsewhere, due to a relative deficiency of folic acid or its active derivatives, or to (5) decreased transsulfuration to cystathionine, due to a deficiency of serine. It

mic subjects and treating them with folic acid and its active is well known that low plasma levels of serine are present in compounds, or by supplementing breads, cereals, and other uremia [60], but of course this could be instead the consequence of an increased transsulfuration pathway activity. dietary components with folic acid in low dosages [501. High levels of Hey have been found in the plasma of chronic renal failure (CRF) subjects: while normal levels range between 6 Protein methylation in CRF and 12 .LM, end-stage renal disease subjects have values between CRF is characterized by several alterations of the erythrocyte 25 and 40 rM [51—54], a very high range indeed, considering that high risk cardiovascular subjects are those with more than 15 to 16 membrane, expressed by hemolysis, reduced deformability and /kM [40—42]. In CRF, cardiovascular disease is among the most increased osmotic fragility, and of the activity of various mem-

common cause of death, accounting for more than one third in international registries [55], the role of hyperhomocysteinemia as an independent risk factor for vascular disease has also been recently assessed in hemodialysis patients [56]. The high levels of Hey found in uremia can be related to (1) reduced urinary excretion. However, recent results that Guttormsen et al presented at the International Conference on Homocys-

brane proteins, such as the erythrocyte Cl/HC03 anion exchanger, by a reduced erythrocyte Na/K pump activity, and by a defect in the activity of the Na/K/2Cl cotransport [61-68]. The end result is certainly a state of deranged cell membrane function.

Recent results obtained by us show that in CRF patients

compared to control healthy individuals, a significant reduction of teine Metabolism [57], show that urinary excretion can be actually membrane protein methyl esterification is detectable that is due to increased in CRF patients, and that the urinary excretion of a significant increase of the natural inhibitor AdoHcy intracellular homocysteine as such is trivial, accounting for only 0.2% of daily concentration [69]. The patient population was divided into two homocysteine production from the body. (2) Also, there can be a groups, one treated conservatively on a low protein dietary decreased tubular metabolism of homocysteine, but this too is regimen, and with a creatinine clearance ranging between 10 and considered to be a minor determinant, since the GFR correlates 40 mI/mm, and one on standard hemodialysis therapy. In particular, a significant reduction in total membrane protein with plasma Hey but not with urinary a2 microglobulin (a marker of tubular damage) in CRF, and Hey clearance is reduced much methyl esterification was observed either by normalizing the data

less than creatinine clearance in CRF. Much more important for mg of proteins (Fig. 4A), or by cell number (Fig. 4B). seems to be (3) a decreased renal metabolism, as shown by Moreover, the reduction was present in both groups, but was more

362

Perna et al: Protein methylation in uremia

A

5

100000

4

80000

0 C',

60000

I0

0)

40000 20000

1

0

*

B

40

0 Fig. 5. Intracellular AdoMetiAdoffcy concentration ratio, in eythrocytes from CRF patients. AdoMet and AdoHey intracellular concentrations have been measured in erythrocytes from CRF patients (both under conservative treatment or under hemodialysis). The two compounds have been simultaneously determined using HPLC. The results are expressed as [AdoMet]/[AdoHcy] in both patient groups compared to normal controls. Since Adoflcy is a competitive inhibitor of AdoMet-dependent enzymes and because of the close values of the Km for AdoMet and Ki for AdoHcy,

U) 30 a) C.)

20

lower values of this ratio indicate the existence of an intracellular environment less suitable for correct methylation. Symbols are: (U) controls; () non-dialysis; dialysis; *P < 0.001. Further details are in

()

110

[69]. Reproduced with permission from J Clin Invest 91:2497—2503, 1993 (The American Society for Clinical Investigation).

0 Fig. 4. Enzymatic methylation of membrane proteins in intact e,ythrocytes

from CRF patients. Erythrocytes have been freshly prepared from CRF patients (both under conservative treatment or under hemodialysis) and from matched normal controls. Methyl esterification has been carried out by incubating these cells for one hour at 37°C in the presence of methyl labeled methionine, as the AdoMet in viva precursor. At the end of the incubation cells have been washed and membranes prepared after hemolysis in hypotonic medium. Methylation of membrane proteins has been evaluated by liquid scintillation counting. Results are expressed as cpm/ million cells (A) or cpm/mg protein (B). Symbols are: (U) control, N = 24; < 0.01. Further details () non-dialysis, N = 12; dialysis, N = 12; are in [69].

()

*

inhibited by a powerful compound, AdoHey. Cytosolic GOT activity, a marker of cell age [70], did not correlate with either methylation levels or with AdoHcy or AdoMet concentrations, demonstrating that these parameters were not influenced by cell age. The data, taken as a whole, are consistent with the notion that in CRF structural damages accumulate in erythrocyte membrane proteins and are not adequately repaired. Role of Hey in the increase of intracellular AdoHcy In the effort to investigate the origin of AdoHcy increase, we measured plasma and RBC Hey levels, AdoHcy, AdoMet, and Ado intracellular levels, and AdoHcy hydrolase specific activity. A

pronounced in the patients with end-stage kidney failure. Char- hemodialysis patient group was compared to normal healthy acterization of proteins through SDS/PAGE showed that the controls, Plasma and red cell Hey were significantly higher in the

decrease was strikingly evident for cytoskeletal component patient group compared to control subjects, and were both ankyriri, band 2.1 [69], which is known to be involved in the positively and significantly correlated to AdoHey intracellular maintenance of membrane stability and integrity [25]. A reduction was present also for the other methyl accepting substrates, that is

levels. Ado and AdoMet levels, and AdoHey hydrolase specific activity were not significantly different between the two groups

band 3 (AE-1 protein), band 4.1 and 4.2, but did not reach

[71]. In addition, we identified positive correlation between

statistical significance. In addition, HPLC analysis of the same RBC samples subject to methylation evidenced a several-fold rise in AdoHcy intracellular levels, while AdoMet concentration was not different from con-

homocysteinemia and the intracellular concentration of AdoHcy in CRF patients (Fig. 6) [71]. To further substantiate the role that increased Hey plays in the elevation of AdoHcy, we designed a series of in vitro experiments. Erythrocytes from control and dialysis patients were incubated with or without 50 LM Hey (a concentration comparable to plasma levels actually found in vivo in these patients), and tracing Ado. The enzymatic formation of labeled AdoHcy from Hey at that

trols, resulting in a lower [AdoMet]/[AdoHcy] ratio in both the patient groups (Fig. 5). The specific activity of the enzyme PCMT type II was not different from controls [69], indicating that this enzyme is malfunctioning not because of some intrinsic damage to

its structure or to altered protein content, but because it is particular concentration and Ado appeared to be significantly

363

Perna et al. Protein methylation in uremia

12

100 -

10

90 -

C

0 C

0 C 0 0 >, 0

I0 •0

.

8-

I800

U

6-

70 -

4-

U

0 60 Cu

C-)

.

2

a

50 D

0

10

5

C

—

I

I

15 20 25 30 35 40 45

Plasma Hcy concentration, l.tM Fig. 6. Correlation between hornocysteinemia and intracellular AdoHcy

0

concentration in CRF patients. Total plasma Hcy concentration was

0

50

100

150 200

250

300

Incubation time, minutes

determined by HPLC. The intracellular concentration of AdoHcy was also determined, by HPLC, in erythrocyte samples from the same patients. Full symbols: CRF patients; open symbols: controls. A positive simple correlation between plasma [Hey] and intracellular [AdoHey] is documented

Fig. 8. Reduction of intracellular AdoHcy concentration as a result of Hey removal. Erythrocytes isolated from CRF patients where incubated in the

P < 0.001). (From Perna AF et al [71], reprinted with

evaluated by HPLC. AdoHcy concentration in the cells incubated without Hey is expressed as percentage of the values obtained in parallel samples

(r =

0.71,

permission).

presence or in the absence of 50 jM Hey. At the indicated time points aliquots were withdrawn and intracellular AdoHcy concentration was

incubated with Hey. (From Perna AF ct al [71], reprinted with permission).

1500 E

Therefore, we were able to conclude that removing Hey from the

incubation medium leads to the gradual disappearance of the elevated levels of AdoHcy, which are in fact of a reversible nature (Fig. 8). The results allowed us to conclude that plasma and red cell Hey levels actually found in uremia can be effectively responsible for the intracellular accumulation of the toxic compound AdoHcy.

1000

500 -

Conclusions and future perspectives The overall importance of our work taken as a whole lies, in our

opinion, in the fact that (1) a pathophysiological link between

0-

AdoHcy, a potent inhibitor of methyltransferases, and Hey can be Control

Uremic

pinpointed in uremia; (2) AdoHey can thus be considered a prospective highly toxic metabolite of Hey, since it is able to

Fig. 7. AdoHcy formation in isolated etythrocytes from CRF patients. Erythrocytes isolated from both CRF patients and normal subjects where

interfere with many AdoMet-dependent pathways involved in the

trace amounts of radioactive Ado. After 5' cytosol was rapidly prepared and AdoHey analysis performed by HPLC. Radioactivity associated with AdoHey peak was measured. (From Perna AF et al [71], reprinted with permission).

variations can be subject to pharmacological treatment, with

incubated in the presence (R) or in the absence () of 50 M Hey and regulation of crucial cell functions; and (3) furthermore, these

increased, in vitro, in erythrocytes from both control and uremic patients (Fig. 7) [71]. In another experiment, two different subsets of erythrocytes from uremic patients were incubated in vitro in the presence and absence of 50 .tM Hey. A significant reduction of intracellular AdoHcy was observed in the subset incubated without Hey over time, compared to the identical samples incubated in presence of

50 .LM Hey, with a half life of approximately 270 minutes.

compounds such as folic acid and its active derivatives. Regarding the first point, the demonstration that the levels of Hey actually found in uremia not only can influence the uremic patient risk of cardiovascular disease, but also can determine an

increase in the intracellular levels of a known inhibitory compound, is in our opinion of major importance. Most experimental evidence on the effects of Hey on various cellular parameters are designed to assess its role in diseases such as genetic homocystinuria, because it represented the first model where the relationship between Hey and premature atherosclerosis was observed. But in this disease Hey levels are in the high micromolar range [47—49]. On the contrary, we showed that even a modest increase, such as it is present in uremic patients, can determine tangible effects that

364

Perna et al: Protein methylation in uremia

Table 2. Cellular functions which may be affected by AdoHcy in uremia Process/reaction

Reference

Cell growth and carcinogenesis Lymphocyte chemokinetic response Insulin release from the pancreas Interferon synthesis Norepinephrine uptake and release in brain Catecholamine degradation Conversion of phosphatidylethanolamine to phosphatidylcholine Brain histamine content Serotonin and dopamine turnover

(73—77)

(78, 79) (80, 81) (82, 83) (84—87)

(29, 31) (29, 31) (88, 89) (90, 91)

4. OTA TM, CLARKE S: Multiple sites of methyl esterification of calmodulin in intact erythrocytes. Arch Biochem Biophys 279:320—327, 1990 5. OT. IM, CLARKE 5: Calcium affects the spontaneous degradation of aspartyl/asparaginyl residues in calmodulin. Biochemistiy 28:4020— 4027, 1989 6. OTA TM, CLARKE S: Enzymatic methylation of L-isoaspartyl residues derived from aspartyl residues in affinity-purified calmodulin: The role

of conformational flexibility in spontaneous isoaspartyl formation. J Biol Chem 264:54—60, 1989 7. ROHER AE, LOWENSON JD, CLARKE 5, WOLKOW C, WANG R, COTrER RJ, REAROON TM, ZURCHER-NEELY HA, HEINRIKSON RL, BALL MJ,

GREENBERG BD: Structural alterations in the peptide backbone of f3-amyloid core protein may account for its deposition and stability in Alzheimer's disease. J Biol Uhem 268:3072—3083, 1993 8. ROI-IER AE, LOWENSON JD, CLARKE 5, WooDs AS, COTI-ER RI, GOWING E, BALL MJ: l3-amyloid-(1-42) is a major component of

cerebrovascular amyloid deposits: Implications for the pathology of Alzheimer disease. Proc NatlAcad Sci USA 90:10836—10840, 1993

can be quantified in vitro. In addition, uremia represents the first pathological condition in which a role of AdoHcy as methylation inhibitor is evidenced. In fact, even if there is a disruption in the protein methylation pattern in another pathology (such as spherocytosis), it is not due to AdoHcy [23, 72]. The second point stated above is the most intriguing, because it calls for several speculations. AdoMet-dependent reactions exist

in the transfer of a methyl group, and are catalyzed by several different enzymes to a variety of acceptors, such as proteins, nucleic acids, phospholipids, neurotransmitters like norepinephrifle, and other small molecules, with different biological functions. AdoHcy affects these enzymes in a fashion that depends on

the Ki values for AdoHcy and Km values for AdoMet [29]. Therefore, if it is proven that AdoHcy also increases in cell types other than erythrocytes in response to the Hcy elevation in blood, it will be likely that a similar inhibitory action would be exerted on

other cellular functions modulated by AdoMet-dependent enzymes (Table 2). Regarding the third point, several studies have shown that Hcy plasma levels can be lowered in uremia by treatment with folic acid, a compound that favors remethylation of Hey to methionine [92—94]. Hemodialysis can only partially lower plasma Hcy [52], while folates abate its levels in an even fashion. Folate levels are

not low in uremia either in plasma or in RBCs, but a relative deficit has been hypothesized [92]. Finally, our in vitro experi. ments indicate that if Hey levels are lowered, AdoHcy concentration can be lowered as well. ALESSANDRA F. PERNA, Dmoo tNGROSSO, PATRIZIA GALLETTI, VINCENZ0 ZAPPIA, and NATALE G. DE SANTO

Second University of Naples, Naples, Italy

Reprint requests to Alessandra F. Pema, M.D., Ph.D., Dipartimento di Pediatria, Cattedra di Nefrologia, Seconda Università di Napoli Federico II, Via S. Pansini, 5 - Padiglione 17, 80131 Napoli, Italy.

9. FABIAN H, SZENDREI GI, MANTSCH HH, GREENBERG BD, OTVOS L

JR: Synthetic post-translationally modified human A13 peptide exhibits a markedly increased tendency to form /3-pleated sheets in vitro. Eur I Biochem 221:959—964, 1994 10. INGR0SSO D, CLARKE 5: Human erythrocyte D-aspartyl/L-isoaspartyl methyltransferase: Enzymes that recognize age-damaged proteins, in Red Blood CellAging, edited by MAGNANI M, DE FLORA A, New York,

Plenum Press, 1991, p 263 11. Boivi D, GINGRAS D, BELIVEAU R: Purification and characterization

of a membrane-bound protein carboxyl methyltransferase from rat kidney cortex. J Biol Chem 268:2610—2615, 1993 12. INGROSSO D, FOWLER AV, BLEIBAUM J, CLARKE S: Sequence of the

D-aspartyl/L-isoaspartyl methyltransferase from human erythrocytes: Common sequence motifs for protein, DNA, RNA and small mole-

cule S-adenosylmethionine-dependent methyltransferases. J Biol Chem 264:20131—20139, 1989 13. INGR0sSO D, KAGAN RM, CLARKE 5: Distinct C-terminal sequences of

isozymes I and II of the human erythrocyte L-isoaspartyl/D-aspartyl protein methyltransferase. Biochem Biophys Res Commun 175:351— 358, 1991 14. MACLAREN DC, KAGAN RM, CLARKE 5: Alternative splicing of the

human isoaspartyl protein carboxyl methyltransferase RNA leads to

the generation of a C-terminal-RDEL sequence in isoenzyme II. Biochem Biophys Res Common 185:277—283, 1992 15. OrA TM, GILBERT JM, CLARKE 5: Two major isozymes of the protein

D-aspartyl/L-isoaspartyl methyltransferase from human erythrocytes.

Biochem Biophys Res Common 151:1136—1139, 1988 16. GALLErTI P, CIARDIELLO A, INoRosso D, Di DONATO A, D'ALESSIO

G: Repair of isopeptide bonds by protein carboxyl 0-methyltrans-

ferase: Seminal ribonuclease as a model system. Biochemistry 27:1752— 1757, 1988 17. MCFADDEN PN, CLARKE 5: Conversion of isoaspartyl peptides to

normal peptides: Implications for the cellular repair of damaged proteins. Proc NatlAcad Sci USA 84:2595—2599, 1987 18. JOHNSON BA, MURRAY ED, ClARKE S, GLASS DB, ASWAD DW:

Protein carboxyl methyltransferase facilitates conversion of atypical L-isoaspartyl peptides to normal L-aspartyl peptides. I Biol Giem 262:5622—5629, 1987

19. LOWENSON JD, CLARKE 5: Spontaneous degradation and enzymatic repair of aspartyl and asparaginyl residues in aging red cell proteins analyzed by computer simulation. Gerontology 37:128—151, 1991 20. BRENNAN TV, ANDERSON JW, JIA Z, WAY000D EB, CLARKE S:

Repair of spontaneously deamidated phosphocarrier protein cata-

References 1. GALIErrI P, INGROSSO D, MANNA C, CLEMENTE G, ZAPPIA V: Protein damage and methylation mediated repair in the erythrocyte. Biochem J 306:313—325, 1995

2. OTA TM, CLARKE 5: The function and enzymology of protein Daspartyl/L-isoaspartyl methyltrasferases in eukaryotic and prokaryotic cells, in Protein Mcthylation, edited by PAIK WK, KIM S, Boca Raton, CRC Press, 1990, pp 179—194 3. JOHNSON BA, LANOMACK EL, AswAo DW: Partial repair of deami-

dation-damaged calmodulin by protein carhoxylmethyltransferase.

J Biol Chem 262:12283—12287, 1987

lyzed by the L-isoaspartate-(D-Aspartate) O-methyltransfcrase. J Biol Chem 269:24586—24595, 1994 21. ASWAD DW, JOHNSON BA, LANGMACK EL, SISIROKAWA JM: Modifi-

cation of isoaspartyl peptides and proteins by protein carboxyl methyltransferase from bovine brain, in Advances in Post-translational Modifications of Proteins and Aging, edited by ZAPPIA V, GALLET1I P, PORTA R, WOLD F, New York, Plenum Press, 1988, pp 247—259 22. ODEN KL, CLARKE 5: S-adenosyl-L-methionine synthetase from hu-

man erythrocytes: Role in the regulation of cellular S-adenosylmethionine levels. Biochemistry 22:2978—2986, 1983 23. INGROSSO D, D'ANGELO 5, PERNA AF, IOLASCON A, MIRAGLIA DEL GIUDICE E, PERRO1TA 5, ZAPPIA V, GALLETFI P: Increased membrane

Perna et at: Protein methylation in uremia protein methylation in hereditary spherocytosis: A marker of cytoskeletal disarray. Eur J Biochem 228:894—898, 1995

D, NAN'! A, GRAGNANIELLO V, IOLASCON A, PINTO L: Increased methyl esterification of membrane proteins in aged red-blood cells. Preferential esterification of ankyrin and band-

24. GALLETrI P, INGROSSO

4.1 cytoskeletal proteins. EurJ Biochem 135:25—31, 1983 25. FREITAG C, CLARKE S: Reversible methylation of cytoskeletal and

membrane proteins in intact human erythrocytes. J Biol Chem 256: 6102—6108, 1981 26. ZAPPIA V, ZYDEK-CWICK CR, SCHLENK F: The specificity of S-

adenosylmethionine derivatives in methyl transfer reactions. J Biol Chem 244:4499—4505, 1969 27. DELA HABA G, CANTONI GL: The enzymatic synthesis of S-adenosyl-

L-homocysteine from adenosine and homocysteine. J Biol Chem 234:603—608, 1959 28. BARBER JR, CLARKE S: Inhibition of protein carboxyl methylation by

S-adenosyl-L-homocysteine in intact erythrocytcs. Physiological consequences. J Biol Chem 259:7115—7122, 1984 29. CANTON! GL, RICHARDS HH, CHIANG PK: Inhibitors of S-adenosylhomocysteine hydrolase and their role in the regulation of biological methylation, in Transmethylation, edited by USDIN E, BORCHARD RT, CREVEI.ING CR, Amsterdam, Elsevier/North Holland, 1979, pp 155— 164 30. OLWA A, GAI.LETrI P, ZAPPIA V, PAIK WK, KIM S: Studies on

substrate specificity of 5-adenosylmethionine: Protein-carboxyl methyltransferase from calf brain. Eur J Biochem 104:595—602, 1980 31. CANTONI GL, CL-HANG PK: The role of S-adenosylhomocysteine and S-adenosylhornocysteine hydrolase in the control of biologic methylations, in Natural Sulfur Compounds: Novel Biochemical and Structural Aspects, edited by CAVALLIN! D, GAULL GE, ZAPPIA V, New York, Plenum Press, 1980, p 67 32. STIPANIJK M: Metabolism of sulfur-containing amino acids. Ann Rev Nutr 6:179—209, 1986 33. SELHUB J, MIL!.ER JW: The pathogenesis of homocysteinemia: Inter-

ruption of the coordinate regulation by S-adenosylmethionine of the remethylation and transsulfuration of homocysteine. Am J Clin Nutr 55:131—138, 1992 34. JAKUBOSKI H, GOLDMAN E:

Synthesis of homocysteine thiolactone by methionyl-tRNA synthetase in cultured mammalian cells. FEBS Let!

317:237—240, 1993 35. MCCULLY KS: Chemical pathology of homocysteine. I. Atherogenesis.

Ann Clin Lab Sci 23:477—493, 1993 36. MCCULLY KS: Chemical pathology of homocysteine. II. Carcinogen-

esis and homocysteine. Thiolactone metabolism. Ann Clin Lab Sci 24:27—59, 1994

37. MCCUI.LY KS: Chemical pathology of homocysteine. III. Cellular function and aging. Ann Clin Lab Sci 24:134-152, 1994 38. ANDERSSON A, ISAKSSON A, HULTBERG B: Homocysteine export from

365

homocysteine: A link to atherosclerosis. Proc NatI Acad Sd USA 91:6369—6373, 1994 45. HARPEL PC, CHANG VT, BORT W: Homocysteine and other sulfhydryl

compounds enhance the binding of lipoprotein(a) to fibrin: A potential biochemical link between thrombosis, atherogenesis, and sulfhydryl compound metabolism. Proc NatlAcad Sci USA 89:10193—10197, 1992 46. PREIBISCF! G, KUFFNER, ELSTNER EF: Biochemical model reactions on the prooxidative activity of homocysteine. J Biosci 48e:58—62, 1992 47. HAJJAR KA: Homocysteine-induced modulation of tissue plasminogen activator binding to its endothelial cell membrane receptor. J Clin Invest 91:2873—2879, 1993 48. RODGERS GM, KANE WH: Activation of endogenous Factor V by a homocysteine-induced vascular endothelial cell activator. J Clin invest 77:1909—1916, 1986 49. HAYASFII T, HONDA G, StJZUKI K: An atherogenic stimulus homocys-

teine inhibits cofactor activity of thrombomodulin and enhances

thrombomodulin expression in human umbilical vein endothelial cells. Blood 79:2930—2936, 1992 50. OAKLEY GP: Folic acid fortification—Prevention opportunity of the decade? Irish JMed Sd 164:(S)14, 1995 51. WILCKEN DEL, GUPTA Vi, REDDY SO: Accumulation of sulphurcontaining amino acids including cysteine-homocysteine in patients on maintenance haemodialysis. Clin Sci 58:427—430, 1980 52. LAIDLAW SA, SM0uN LA, DAVIDSON WD, KOPPLE JD: Sulfur amino

acids in maintenance hemodialysis patients. Kidney in! 32(Suppl 22):S191—S196, 1987 53. CHAUVEAU P, CHADEFAUX B, C0UDE M, AUPETIT J, HANNEDOUCHE T, KAMOUN P, JUNGERS P: Increased plasma honiocysteine concen-

tration in patients with chronic renal failure. Miner Electrol Metab

18:196—198, 1992 54. CHAUVEAU P, CHADEFAUX B, COUDE M, AUPETIT J, HANNEDOUCUE T, KAMOUN P, JUNGERS P: Hyperhomocysteinemia, a risk factor for

atherosclerosis in chronic uremic patients. KidnEy mt 43(Suppl 41): S72—S77, 1993

55. PARFREY PS: Cardiac and cerebrovascular disease in chronic uremia. Am J Kid Dis 21:77—80, 1993 56. BACHMANN J, TEPEL M, RA!DT H, RIEZLER R, GRAEFE U, LANGER K, ZIDEK W: Hyperhomocysteinemia and the risk for vascular disease in hemodialysis patients. JAm Soc Nephrol 6:121—125, 1995 57. GUTrORMSEN AB, SVARSTAD E, UELAND PM, REFSUM H: Elimination

of homocysteine from plasma in subjects with end stage renal failure. Irish J Med Sci 164:(S)8—9, 1995 58. BOSTOM AG, BROSNAN JT, HALL B, NADEAU MR, SELHUB J: Homo-

eysteine metabolism by the rat kidney in vivo. Irish J Med Sci 164:(S)2—3, 1995

59. Bosrorsi AG, BROSNAN JT, HALL B, NADEAU MR, SELHUB J: Net

uptake of plasma homocyteine by the rat kidney in vivo. Atherosclerosis 116:59—62, 1995

erythrocytes and its implication for plasma sampling. Clin ('hem

60. JONTOFSOHN R, TRIVISAS G, KATZ N, KLUTHE R: Amino acid content

38:1311—1315, 1992

of erythrocytes in uremia. Am J ('lin Nutr 31:1956—1960, 1978 61. SHAW AB: Haemolysis in chronic renal failure. Br Med J 2:213—216,

39. CLARKE R, DALY L, ROBINSON K, NAUGHTEN E, CAHALANE 5, FOWLER B, GRAHAM I: Hyperhomocysteinemia: An independent risk factor for vascular disease. N Engl J Med 324:1149—1155, 1991 40. STAMPEER MJ, MALINOW MR, WILLE-tT WC, NEWCOMER LM, UPSON B, ULLMANN D, TISHLER PV, HENNEKENS CH: A prospective study of plasma homocyst(e)ine and risk of myocardial infarction in US

1967 62. AKMAL M, TELFER N, ANSARI AN, MASSRY SG: Erythrocyte survival

in chronic renal failure. Role of secondary hyperparathyroidism. J Clin Invest 76:1695—1698, 1985 63. INAUEN W, STAUBLI M, DESCOEUDRES C, GALEAZZI RL, STRAUB PW:

physicians. JAMA 268:877—88 1, 1992 41. MALINOW MR, NIETO FJ, SzKIO M, CHAMIILESS LE, BOND G: Carotid

Erythrocyte deformability in dialysed and non-dialysed uraemic pa-

artery intimal-medial wall thickening and plasma homocyst(e)ine in asymptomatic adults. The atherosclerosis risk in communities study.

64. MCGRATH i:r, D0UGtAs AF, MCCLEAN E, BROWN JH, DOHERTY CC, JOHNSTON GD, P00LER G, ARCILBOLD R: Oxidative stress and eiyth-

Circulation 87:1107—1113, 1993 42. NYGARD 0, VOLLSET SE, REFSUM H, STENSVOI.D I, TVERDAI. A, NORDREHAUG JE, UELAND PM, KVALE G: Total plasma homocysteine

rocyte membrane fluidity in patients undergoing regular dialysis. Clin ('him Acta 235:179—188, 1995 65. MADUELL F, FERNANDEZ I, DIEZ J: Alterations of the Cl 7HC03

and cardiovascular risk profile. The Hordaland homocysteine study.

anion exchanger in erythrocytes of uraemic patients. Nephrol Dial

JAIVIA 274:1526—1533, 1995 43. RoBINsoN K, MAYER EL, MILLER DP, GREEN R, VAN LENTE F, GUPTA A, KOTI-KE-MARCI-LANT K, SAvors SR, SELHUB J, Nissrs SE,

Transplant 5:1018—1022, 1990 66. KAJI D, KAHN T: Nat-K pump in chronic renal failure. Am J Physiol

M, TOP0L EJ, JACOBSEN DW: Hyperhomocysteinemia and low pyridoxal phosphate: Common and independent reversible risk KUTNER

factors for coronary artery disease. Circulation 92:2825—2830, 1995 44. Tsi C, PERRELLA MA, YOsHIzuMI M, HSIEN C-M, HABE E, SCiiLAGE!. R, LEE M-E: Promotion of vascular smooth muscle cell growth by

tients. Eur J Clin invest 12:173—176, 1982

252:F785—F793, 1987 67. KELLY RA, CANESSA ML, STEINMAN TI, MITCH WE: Hemodialysis

and red cell cation transport in uremia: Role of membrane free fatty

acids. Kidney bit 35:595—603, 1989 68. CORRY DB, LEE DBM, TUEK ML: A kinetic study of cation transport in erythrocytes from uremic patients. Kidney mt 32:256—260, 1987

366

Perna et al: P,vtein methylation in uremia

69. PERNA AF, lNc,Rosso D, GAuAcrrl P, ZAPPIA V, DR SANTO NG:

Enzymatic methyl esterification of erythrocyte membrane proteins is

edited by FRIEDMAN EA, BEYER M, DL SANTO NG, GIORDANO C,

New York, Field and Wood, 1989, pp 129—133

impaired in chronic renal failure: Evidence for high levels of the

84. HEILBREDER F, SCHAFFERHANS K, GOTZ R, BAUSEWEIN K, HEIDLAND

natural inhibitor S-adenosylhomocysteine. J Gun Invest 91:2497—2503,

A: Plasma catecholamines in renal failure, in Hormonal and Metabolic

1993

70. HARAM 5, CARRIERO D, SEAMAN C, PIoMELLI S: The mechanism of

decline of age-dependent enzymes in the red blood cell. Enzyme 45:47—53, 1991 71. PERNA AF, INGROSSO D, Du SANTO NG, GALLETFI F, ZAPPIA V:

Mechanism of erythrocyte accumulation of methylation inhibitor

S-adenosylhomocysteine in uremia. Kidney mt 47:247—253, 1995 72. INoRosso D, D'ANOELO 5, PERROYIA 5, D'URZO G, IOLASCON A, PERNA AF, GALLE1TI F, ZAPPIA V, MIRAGLIA DEL GIUDIcE E:

ytoske1etal behaviour in spectrin and in hand 3 deficient spherocytic red cells: Evidence for a differentiated splenic conditioning role. BrJ

Haematol 93:38—41, 1996

73. HOFFMAN RM: Altered methionine metabolism, DNA methylation and oncogene expression in carcinogenesis. Biochim Biophys Acta 738:49—87, 1984 74. NEWIIERNE MP, ROGERS AE: Labile methyl groups and the promotion

of cancer. Ann Rev Nutr 6:407—432, 1986 75. DUERRE JA, Di MARIA P, KIM 5, PAIR WK: Current status of protein methylation in earcinogenesis. Grit Rev Oncol 2:97—108, 1991

76. PORT FK: Fatal ncoplasms in dialysis patients: A population-based study. Am J Kid Dis 24:119—1 25, 1989 77. IsEKI K, OSAWA A, FIJKIYAMA K: Evidence for increased cancer deaths in chronic dialysis patients. Am J Kid Dis 22:308—313, 1993 78. MCFADDEN RG, FRAIIER LG: Inhibitors of membrane transmethyla-

tion reactions prevent the lymphocyte chcmokinetic response. Immuno! Lett 26:211—215, 1990

79. PIKE MC, Di MEESTER CA: Inhibition of phosphoinositide metabolism in human polymorphonuclear leucocytcs by S-adenosylhomocysteine. J Biol Chem 263:3592—3599, 1988 80. FADDA GZ, HAJJAR SM, PERNA AF, ZHOU X-J, LIPSON LG, MASSRY

SO: On the mechanism of impaired insulin secretion in chronic renal failure.] C/in Invest 87:255—261, 1991 81. BEST L, LEBRUN F, SACEDA M, GARCIA-MORALES P, HUBINoT C,

Derangements in Renal Failure, edited by 1-ILIDLAND A, QUELIHORST E, HEILBREDER E, RITZ E, MASSRY SG, Basel, Karger, 1986, pp 14—27 85. SM000RZEWSRI M, CAMPESE VM, MASSRY SO: Abnormal norepi-

ncphrine uptake and release in brain synaptosomcs in chronic renal failure. Kidney mt 36:458—465, 1989

86. FULLER RW: Inhibitors of norepinephrine N-methyltransferasc, in Transmethylation, edited by USDIN E, BORCI-IARD RT, CREVELING CR,

Amsterdam, Elscvicr/North Holland, 1979, pp 25 1—259 87. SAMET MK, RUTLEDGE CO: Inhibition of norepinephrine uptake by inhibitors of methylation, in Biocheniistmy of S-Adenosylmethionine and Related Compounds, edited by USDIN E, BORCHARDT RT, CREVELINO CR, London, MacMillan Press, 1982, pp 195—198 88. DUCH DS, BOWERS 5, EDELSTEIN M, NIcI 01 CA: Histamine: Elevation of brain levels by inhibition of histamine N-methyl transferase, in Transmethylation, edited by USDIN E, BORCI lARD RT, CREVELING CR,

Amsterdam, Elsevier/North Holland, 1979, pp 287—295 89. SCIIMID G, PRZUNTER H, FRICRE L, HEIDLAND A, HEMPEL K:

Increased histidine and histamine content in the braIn of chronic uremic rats. Am J C/in Nutr 31:1665—1668, 1978

90. EwOIT GR, BARCIIAS JD: The transmcthylation hypothesis of schizophrenia, in Transmethylation, edited by USDIN E, BORCIIARD RT, CREVELING CR, Amsterdam, Elsevier/Norlh Holland, 1979, pp 307—3 18

91. BIASI0LI 5, FERIANI M, CHIARAMONTE 5, PORENA F, PI-:TROSINO L, ZAMBELLO A, CESARO A, PRADELLA M, CAVALLINI L: The scrotonin-

ergic system in uremia: Relationship to hormonal status, in Prevention of Progressive Uremia, edited by FRIEDMAN EA, BIYER M, DR SANTO NO, GIORDANO C, New York, Field and Wood, 1989, pp 79—82 92. WILCKEN DEL, DUDMAN NPB, TYRRELI, PA, ROBERTSON MR: Folic

acid lowers elevated plasma homocysteine in chronic renal insufficiency: Possible implications for prevention of vascular disease. Metabolism 37:697—70), 1988

JUVENI M, HERCHUELZ A, MAI,AISSE-LAGAE F, VAIVERDE I, MAL-

93. ARNADOYrIR M, BRAITSTROM LA, SIMONSI:N 0, THYSELL H, HtJLT-

AISSE WJ: Impairment of insulin release by methylation inhibitors.

BERG B, ANDERSSON A, NILSSON-EHLE F: The effect of high-dose

Biochem Pharmacol 33:2033—2039, 1984

pyridoxine and folic acid supplementation on serum lipid and plasma homocysteine concentrations in dialysis patients. C/in Nephrol 40:236— 240, 1993

82. Aui R: Effect of inhibitors of methylation on early and late interferon synthesis in bovine kidney cell coltures. J Intemferon Res 1:203—218, 1981

83. LAMPERI 5, CAROZZI S: y-Interferon as an in vitro enhancing factor of peritoneal macrophage defective bactericidal activity during continous

ambulatory peritoneal dialysis, in Prevention of Progressive Uremia,

94. BOSTOM AG, SHEMIN D, LAPANE KL, HUME AL, YOBURN D, NADEAU

MR, BENDICH A, SELHUB J, ROSENBERG IH: High dose B-vitamin

treatment of hyperhomocysteinemia in dialysis patients. Kidney mt 49:147—152, 1996