Schiff Base Adducts of Hemoglobin

4 downloads 150176 Views 1MB Size Report
Schiff base (imine) linkages were formed with amino groups of intracellular hemoglobin. Adducts were isolated by gel electrofocusing and could be dissociated ...
THE JOURNAL OP BIOLOGICAL CHEMISTRY Vol. 252, No 23, Issue of December 10, pp. 8542-8548, Printed

Schiff

W7’7

m U.S.A.

Base Adducts

MODIFICATIONS

THAT

INHIBIT

of Hemoglobin ERYTHROCYTE

SICKLING* (Received

ROBERT H. ZAUGG,

JOSEPH A. WALDER,

From the Department of Biochemistry Northwestern University, Evanston,

AND IRVING

and Molecular Illinois 60201

Sickle cell disease is a hemolytic disorder in which deoxygenated erythrocytes assume a variety of unusual shapes, most typically sickled forms. These abnormally shaped erythrocytes are less deformable than normal red blood cells and, as a result, tend to become trapped in and to occlude the microcirculation. These effects lead to tissue injury and necrosis. The propensity of such cells to sickle is traceable to a markedly decreased solubility of deoxyhemoglobin S relative to the oxygenated form. Molecules of intracellular HbS’ aggregate upon deoxygenation and form helical fibers which distort the cell into characteristic sickle shapes (1, 2). Chemical approaches toward modifying the behavior of hemoglobin S can be based on covalent or noncovalent modifications. Among covalent modifications, blocking of amino groups seems particularly attractive since this can be achieved by relatively mild reagents that might be pharmacologically acceptable. One class of such reagents is exemplified by cyanate (3, 4) and by aspirins (57), the former producing a * This investigation was supported in part by the National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Sec-

tion 1734 solely to indicate this fact. bin

1 The A.

abbreviations

used

are:

HbS,

hemoglobin

S; HbA,

hemoglo-

8542

July

5, 1977)

M. KLOTZ Biology

and the Department

of

Chemistry,

carbamyl substituent and the latter an acyl adduct at amino group sites on the protein. An alternative general procedure for attaching substituents at -NH2 sites takes advantage of Schiff base formation with aldehydes. In Go, the appearance of HbA,,, a glucose adduct of hemoglobin (8, 9), in the blood of normal individuals points strongly to Schiff base formation between glucose and this oxygen carrier. That such sugar adducts are formed in vztro has been demonstrated with hemoglobin (10-12) as well as with a Val-His peptide (13). Furthermore, extensive studies have shown that aldehydic pyridoxal compounds can also be coupled to hemoglobin in vitro (14-18). It has seemed appropriate, therefore, to undertake a broader examination of the reactions

of carbonyl

compounds

effects of these compounds EXPERIMENTAL

with

on sickling

hemoglobin

and

of the

properties.

PROCEDURES

Materi&-All aldehydes and ketones were obtained from commercial sources. p-Dimethylaminobenzaldehyde was purchased from Lapine Scientific, pyridoxal and pyridoxal phosphate from Sigma, and others from Aldrich. These compounds were used as obtained or were purified by fractional distillation or recrystallization when necessary. Potassium cyanate was purchased from Baker. Carrier ampholytes (pH 6 to 8) for isoelectric focusing were obtained from LKB, acrylamide and bisacrylamide from Aldrich, ammonium persulfate and N,N,N’,N’-tetramethylethylenediamine from Bio-Rad. Erythrocyte Preparation-Whole blood was drawn by venipuncture from healthy adults. EDTA was used as anticoagulant. Sickle cell blood was obtained through the Hematology Department, Cook County Hospital, Chicago, Ill., as residual blood from homozygous SS individuals. Erythrocytes were washed twice in isotonic saline solution (0.154 M NaCl) and once in isotonic phosphate buffer (0.123 M sodium phosphate), pH 7.2. Experiments performed on normal whole blood required no prewashing of erythrocytes. Washed sickle cell erythrocytes were resuspended to original hematocrit in fresh, frozen human AB plasma for experiments involving whole sickle blood. Chemical ModqGztions -All compounds were tested at a concentration of 5 rnsr in 20% (v/v) erythrocyte suspensions in isotonic phosphate buffer (pH 7.2). Supplemental tests were conducted in whole blood and in 6% solutions of cell-free hemoglobin. After addition of the test reagent, samples were incubated for i/z h at 37” in a water bath shaker. Incubations with each of several compounds for longer intervals demonstrated that the reaction attained equilibrium within i/z h. Reactions involving erythrocyte suspensions were terminated by centrifugation of the cells, removal of the supernatant, and rapid freezing of the layer of packed erythrocytes by immersion in a dry ice/methanol bath. Reactions with cell-free hemoglobin were ended by freezing the entire sample with dry ice/ methanol. In all experiments, controls lacking reagent were incubated under similar conditions.

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

Normal and sickle erythrocytes were exposed in vitro to millimolar concentrations of 31 different carbonyl compounds. Schiff base (imine) linkages were formed with amino groups of intracellular hemoglobin. Adducts were isolated by gel electrofocusing and could be dissociated by dialysis. Aromatic aldehydes proved more reactive than aliphatic aldehydes, and ketones were unreactive. The influence of various ring substituents on the reactivity of aromatic aldehydes was found to conform closely to traditional concepts regarding electronic and steric effects. Several of the aromatic aldehydes were shown to markedly increase the oxygen affinity of hemoglobins A and S. In particular, 2,4-dihydroxybenzaldehyde and o-vanillin, at concentrations of 5 mu, produced 2- to 3-fold reductions in the P,, (partial pressure of oxygen at half-saturation) of sickle hemoglobin in whole blood. Since low degrees of oxygen saturation promote erythrocyte sickling, compounds of this type significantly inhibit sickling at reduced partial pressures of oxygen.

for publication,

Hemoglobin

Modifications

In this equation, P& and P&, are the partial pressures of oxygen at half-saturation for the control and treated suspensions, respectively. Tests for in Vitro Antisickling Activity -A 4-ml sample of whole sickle cell blood containing 5 rnM reagent and an identical control sample lacking reagent were placed in lo-ml flasks and equilibrated successively with gas mixtures containing 5, 4, 3, 2, and 1% oxygen in 5% CO,. The blood was exposed to each oxygen mixture for 30 min after which an aliquot was removed and fixed in isotonic 2% formalin. A minimum of 300 fixed cells was counted using phase contrast optics at x 450 magnification. The extent of sickling was plotted as a function of oxygen partial pressure. RESULTS

We have investigated the reactions of intracellular HbA with the 31 carbonyl compounds shown in Fig. 1. The extents of reaction, as assessed by isoelectric focusing, were quantitated by integration of densitometric scans of the gels after fixation in 10% trichloroacetic acid. As shown in Table I, aromatic aldehydes (Compounds 1 to 25) generally effected substantial modification of intracellular hemoglobin. Aliphatic aldehydes (Compounds 26 to 29) were less effective, and ketones (Compounds 30 and 31) were unreactive. Carbonyl compounds react with free amino groups of pro’ Electrofocusing of hemolysate from untreated erythrocytes reveals a faint band in this anodal region as well. This is due to HbA,, a glucose adduct of hemoglobin, normally present in trace amounts (Ref. 19).

Inhibit

Sickling

8543 TABLE

Modifications

All reactions cytes in isotonic

of intracellular effects

HbA on oxygen

I with aldehydes affinity

and

ketones;

involved 20% (v/v) suspensions of normal erythrophosphate buffer, pH 7.2, at 37”. Comwund Modified” %APp;“”

% Benzaldehydes 1 Benzaldehyde 2 Salicylaldehyde 3 3-Hydroxybenzaldehyde 4 4-Hydroxybenzaldehyde 5 o-Anisaldehyde 6 m-Anisaldehyde 7 p-Anisaldehyde 8 2-Carboxybenzaldehyde 9 4Carboxybenzaldehyde 10 2-Chlorobenzaldehyde 11 4-Cyanobenzaldehyde 12 4-Dimethylaminobenzaldehyde 13 Helicin 14 2,3-Dihydroxybenzaldehyde 15 2,4-Dihydroxybenzaldehyde 16 o-Vanillin 17 5-Nitrosalicylaldehyde 18 3,4-Dihydroxybenzaldehyde 19 Vanillin 20 Isovanillin 21 Veratraldehyde

60 45 60 30 55 65 25 15 20 so 90 0 10 40 70 55 0 30 50 30 50

5 50 45 20 40 30 10 5 0 35 55 0 15 75 60 75 55 15 30 35 40

Other 22 23 24 25

30 20 0 75

35 60 0 40

25 25 5 0

0 0 0 0

0 0

0 0

aromatic aldehydes trans-Cinnamaldehyde Pyridoxal Pyridoxal phosphate Furfural

Aliphatic aldehydes 26 Valeraldehyde 27 Cyclohexanecarboxaldehyde 28 Glyceraldehyde 29 Glucose Ketones 30 Fructose 31 2-Hvdroxvacetonhenone

n Percentage of total hemoglobin which is modified in the presence of 5 mM compound as assessed by isoelectric focusing. b Percentage of change in P,, in the presence of 3 rnM compound:

teins to form reversible Schiff base linkages (21). That the modified species detected by isoelectric focusing in the present study were Schiff base adducts of hemoglobin is supported by the finding that the modification is reversible. For example, erythrocytes were exposed to 5 mM 4-cyanobenzaldehyde (Compound 11) for i/z h in isotonic phosphate buffer. Hemoglobin extracted from these erythrocytes was greater than 90% modified (Fig. 2, Gel 1). Upon repeated resuspension of treated erythrocytes in reagent-free buffer, the extent of modification was diminished (Fig. 2, Gels 2 and 3) and ultimately reduced to zero (Fig. 2, Gel 4). Similar results were obtained with modified cell-free hemoglobin that was dialyzed against reagent-free buffer. These experiments do not preclude the possibility that the reversible adduct formed is something other than a Schiff base. Model studies suggest that the most likely alternative

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

The extent of chemical modification of hemoglobin was assessed by isoelectric focusing as described by Drysdale et al. (19). Frozen erythrocytes were thawed and completely hemolyzed by addition of 5 volumes of ice cold distilled water containing 0.01 M NaCN. Cellfree hemoglobin was mixed with an equal volume of 0.01 M NaCN. The cyanide converts any methemoglobin (generally less than 10% of the total) to cyanomethemoglobin which co-electrophoreses with Fez+-hemoglobin (20). Gel electrofocusing was conducted in tubes (85 mm x 5 mm, inner diameter) containing 5% (w/v) acrylamide, 0.2% (w/v) bisacrylamide, 2% (v/v) carrier ampholytes, and 5% (v/v) glycerol. Solutions of 0.02 M phosphoric acid (pH 2.2) and 0.01 M NaOH (pH 12) served as anolyte and catholyte, respectively. The temperature of the jacketed electrophoresis chamber was maintained at 4”. Following pre-electrophoresis for 10 min, 200 to 300 pg of Hb in 20 ~1 of 2% ampholytes was applied to each gel tube. A current of 1 mA/gel was maintained until the voltage reached 500 V, after which no further adjustments were made. Overall running time was 3 to 4 h, although the hemoglobin bands reached their equilibrium positions within ii/z h. In every case the modified species appeared in a position anodal to unmodified HbA.* After electrofocusing, the gels were removed from the tubes and fixed in 10% trichloroacetic acid to prevent diffusion of the focused bands. The extent of modification was quantitated by integration of peaks in densitometric scans of the gels. Oxygen Binding Studies -Oxygen affinities of modified erythrocytes were determined (a) in 20% cell suspensions of normal erythrocytes in isotonic phosphate buffer containing 3 rnsr compound, and (b) in whole sickle cell blood with 5 rnsr compound. After addition of the test reagent, adjustments in pH (in the range 7.2 to 7.4) were made as necessary so that control and treated suspensions were identical 10.02 unit. Cell suspensions (in lo-ml flasks) were incubated for i/z h at 37” in a water bath shaker. Gas mixtures containing 5, 4, 3, 2, and 1% oxygen in 5% CO, (Matheson Gas Products) were then administered successively, using a gas proportioner and flow meters supplied by Matheson. Erythrocytes were allowed l/z h to equilibrate at each oxygen level, after which a 0.5. ml aliquot was removed and its oxygen saturation measured on an IL 182 Co-Oximeter. In control experiments, this was shown to be sufficient time to attain equilibrium. The percentage of oxygen saturation was plotted against the partial pressure of oxygen (mm Hg). The P,, value (oxygen tension at 50% saturation) was determined graphically for each of the treated and control suspensions. The following parameter was defined to describe the effect on oxygen binding of treatment with the various test compounds:

That

Hemoglobin

8544 CHO

Modifications

CHO

CHO

&OH

Sickling CHO

0

CHO

CHO

CHO

CHO

OCH,

0

3

CHO

Inhibit

CHO

&OH

2

That

6

CHO

0%

7

CHO O-6 -DGlc

coo-

0

b

IO

8

CH,

I IN\

CH,

12 CHO

CHO

CHO

CHO

CHO

CHO

CHO

@OCH,

@OH OCH,

CHO HO

OCH,

25

OCH,

21

FHO

FHO CHO

‘=a kH,

CHO kH,0~

&H,

28

26 1.

Chemical

F=O HOFH

HOFH

H&OH

bH,

FIG. shown.

HFOH HFOH

HFOH

HFOH

HFOH

CH,OH

29 structures

of carbonyl

compounds

used to modify

is thiohemiacetal formation with -SH groups of available cysteine residues (22). This possibility was ruled out by pretreatment of cell-free hemoglobin with the thiol-specific blocking agent, N-ethylmaleimide, using the procedure of Benesch and Benesch (23). Subsequent exposure of the -SHblocked hemoglobin to aromatic aldehydes 1, 2, 11, and 15 provided modification identical in extent to that obtained in the absence of N-ethylmaleimide. Normal whole blood was treated with aromatic aldehydes 1 to 4, 11, 15, 16, 19, and 20 to assess the influence of serum proteins on the modification of intracellular hemoglobin. In general, the extent of hemoglobin modification in whole blood was 50 to 75% of that obtained from reactions conducted in isotonic buffer. The effect on oxygen affinity of modifying intracellular hemoglobin with 3 mM concentration of each compound is shown in Table I. In general, aromatic aldehydes provided substantial increases in the oxygen affinity of treated erythrocytes, whereas aliphatic aldehydes and ketones elicited no effect. Typical oxygen-binding curves for treated and un-

hemoglobins

COCH,

FH,OH

31

CH,OH

30 A and S. Predominant

ionic

species

at neutral

pH are

treated normal erythrocytes suspended in isotonic phosphate buffer are shown in Fig. 3. Salicylaldehyde (21, o-vanillin (16), and vanillin (19) produced substantial increases in oxygen affinity, characterized by decreases in the P,,. Glyceraldehyde (28) at 3 mM concentration elicited no effect. From Table I, it is apparent that chemical modification of hemoglobin is necessary to increase its oxygen affinity. Thus, compounds which do not modify hemoglobin do not affect oxygen binding. This dependence upon modification was further demonstrated by washing treated erythrocytes repeatedly in reagent-free buffer as described in Fig. 2. A decrease in oxygen affinity resulted which paralleled the decreasing extent of modification. Nevertheless, there is no direct correlation between the extent of modification and the effects on oxygen affinity. In particular, aromatic aldehydes that contain an o-OH substituent (Compounds 2, 14 to 17, and 23) had a disproportionately large effect on oxygen binding, whereas benzaldehyde (11, despite substantial modification, had little effect on oxygen affinity. This general lack of correlation may be due to

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

20

Hemoglobin

Modifications

modifications occurring at different sites on the hemoglobin molecule. Support for this was provided by competition experiments. Benzaldehyde (at 3 and 5 mM concentrations) failed to diminish the increase in oxygen affinity produced by 3 mM 2,4-dihydroxybenzaldehyde (15). Since the effects on oxygen affinity obtained with normal

That

Inhibit

Sickling

8545

erythrocytes were in a favorable direction for the treatment of sickle cell disease, we studied the reactions of a number of aromatic aldehydes with intracellular HbS. Both the extent of modification and the influence on oxygen binding were very similar to those observed with normal erythrocytes. Fig. 4 shows oxygen-binding curves for treated and untreated whole sickle cell blood. With 5 mM salicylaldehyde (2) and ovanillin (161, large increases in oxygen affinity were obtained. Potassium cyanate at this concentration had little effect on oxygen binding. In vitro sickling tests were conducted on whole sickle cell blood containing 5 mM salicylaldehyde (21, o-vanillin (161, and potassium cyanate. The extent of sickling is plotted as a function of the partial pressure of oxygen in Fig. 5. The two aromatic aldehydes tested were approximately lo-fold more effective than cyanate in reducing the extent of sickling at equivalent oxygen tensions. DISCUSSION

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

This report presents the results of a systematic examination of the influence of molecular structure on the reactions of carbonyl compounds with intracellular hemoglobins A and S. We have shown that aromatic aldehydes in particular form relatively stable adducts with hemoglobin which can be resolved by isoelectric focusing. The modified species are focused at positions slightly anodal to unmodified hemoglobin (see Fig. 21, the position suggesting that a positive charge (e.g. ENH2 of lysine) has been partially neutralized. Conjugate acids of Schiff base (imine) linkages between aromatic aldehydes and amino groups typically have pK, values near 7.0 (24). At neutral pH, therefore, the Schiff base adduct will carry on the average a half-positive charge producing a hemoglobin species which is slightly more anionic than unmodified protein. Further support for the existence of Schiff base linkages between aldehyde and hemoglobin was provided by the demonstration that adduct formation is reversible (see Fig. 2). In addition, the possibility that a reversible thiohemiacetal linkage formed was excluded by showing that modification with aldehydes was unaltered by pretreatment of hemoglobin with the thiol-specific agent, N-ethylmaleimide.

loo 50 .IO -.. .- ....i;;m #50 a” _ 8:“# IO-

FIG. 2. Electrofocused gels demonstrating reversibility of chemical modification. Erythrocytes were treated with 5 mM 4-cyanobenzaldehyde (11) for l/z h at 37” in isotonic phosphate buffer, pH 7.2. Cells were then washed free of excess compound and resuspended in reagent-free buffer. This procedure was repeated at hourly intervals for 6 h. Aliquots were removed and frozen at each interval. Isoelectric focusing was performed as described under “Experimental Procedures.” C, untreated Hb; 1 to 4, treated Hb after 0, 1, 2, and 6 resuspensions, respectively.

0

0

lo

I 20 PO, hnl)

I 30

0

0

I IO

FIG. 3 (left). Oxygen-binding curves for normal erythrocytes in isotonic phosphate buffer treated with 3 mM glyceraldehyde (O-O), vanillin (&---a), salicylaldehyde (A---A), o-vanillin (V-V), or untreated (0-O). FIG. 4 (center). Oxygen-binding curves for whole sickle blood treated with 5 mM potassium cyanate (O--O), vanillin (m----m),

I 20 ~0~ (mm)

I 30

I 40

0

0

IO

, 20 pot hnml

I 20

40

2,4-dihydroxybenzaldehyde (A-A), o-vanillin (V-V), or untreated (O-O). FIG. 5 (rig&). Sickling curves for whole sickle blood treated with 5 rnM potassium cyanate (O-O), o-vanillin (+---MI, salicylaldehyde (A-A), or untreated (O-O).

8546

Hemoglobin

In Schiff base type is established: Hb-NH,

formation,

an equilibrium

+ R--CHO

:

Hb-N-CH-R

Modifications of the + H,O

following (1)

Inhibit

Sickling

studied possessing ionizable groups have pK, values equal to or greater than that of 2,4-dihydroxybenzaldehyde. Hence, these compounds should form relatively stable equilibria within the gels, and the observed levels of modification should parallel the equilibrium values that exist within the erythrocyte. In support of this, we find that the relative extents of Schiff base formation with hemoglobin are in general agreement with predictions based on studies of model amines. For simple amines, imine formation decreases in the order: aromatic aldehydes = aliphatic aldehydes > aliphatic ketones > aromatic ketones (25). Within each of these classes, electron-withdrawing substituents enhance Schiff base formation while sterically bulky substituents adjacent to the carbonyl group disfavor adduct formation (26). We found that neither aliphatic nor aromatic ketones (Compounds 30 and 31) formed Schiff base adducts with hemoglobin. On the other hand, aldehydes readily formed Schiff bases with hemoglobin, and aromatic derivatives showed a 2- to 3-fold increased reactivity over aliphatic aldehydes. Weakly electron-donating substituents, such as p-OH and p-OCH, groups (Compounds 4 and 71, caused a lowered reactivity relative to benzaldehyde (see Table I). These same substituents in the meta position either are neutral (m-OH) or slightly electron-withdrawing (mOCH,,) and Compounds 3 and 6 provided modification of HbA that was equivalent to or slightly greater than 1. Strong electron-withdrawing groups, such as 2-chloro and 4-cyano, make 10 and 11 the most reactive compounds tested, whereas strong electron donation by the p-dimethylamino group in 12 reduced the reactivity of that compound to zero. The p-carboxylate group is slightly electron-withdrawing, but 9 provided greatly reduced modification relative to 1 rather than the expected increase. The presence on 9 of a fixed negative charge at neutral pH should reduce the extent of modification detected electrophoretically in a manner analogous to that observed with 5-nitrosalicylaldehyde. However, treatment of erythrocytes with 9 caused no change in the oxygen-binding properties of hemoglobin. Thus, if considerable undetected adduct formed with 9, the modification must have taken place at a site that does not influence oxygen binding.3 Helicin (13), a highly soluble /J-n-glucoside of salicylaldehyde, achieved only slight modification of intracellular HbA. This behavior is probably a reflection of the steric hindrance imposed by the bulky o-substituent. The modification data shown in Table I for the 3,4-disubstituted benzaldehydes (18 to 21) are consistent with the findings just described for monosubstituted compounds. The electron withdrawal provided by the m-OCH, group in 19 and 21 increased the reactivity of these compounds relative to 18 and 20. The latter compounds carry a m-OH group which is neither a withdrawing nor a donating substituent. Both -OH and -OCH, groups are electron-donating in the para position, however, and this effect lowered the reactivity of all four compounds relative to benzaldehyde. Resonance stabilization of the allenic substituent in cinnamaldehyde (22) should disfavor adduct formation and cause decreased modification relative to 1, as is indeed observed. Pyridoxal (23), pyridoxal phosphate (241, and furfural (25) all contain highly reactive carbonyl groups by virtue of strong electron-withdrawing ring systems, but of these only 25 provided increased modification. As a dianion at neutral pH, 24 3 Treatment of modification the erythrocyte

of cell-free hemolysate with 9 indicated was not due to a failure of the compound membrane.

that the lack to permeate

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

An equilibrium between modified and unmodified hemoglobin should also exist within the electrofocusing gel. At a given concentration of free aldehyde, the extent of Schiff base formation should be the same within the gel as it is in solution. However, since little, if any, free aldehyde placed on the gel will migrate to the isoelectric positions of the hemoglobin species, the free concentration of aldehyde at those positions will be derived solely from dissociation of the adduct. Hence, the concentration of free aldehyde, as well as the extent of modification observed in the gel, is a lower limit to the equilibrium value that exists within the erythrocyte. If the free aldehyde is uncharged at the isoelectric positions of the hemoglobin species, then its local concentration would be depleted only by diffusion. Since this is a relatively slow process within the gel, the Schiff base adduct should be comparatively stable. If, on the other hand, the free aldehyde is charged at the isoelectric positions of the hemoglobin species, it will migrate away from the site of focusing in response to the electric field, and the migration will cause the equilibrium in Equation 1 to shift progressively to the left. This dissociating effect of an electric field on an adduct was demonstrated with the complex formed between HbA and 2,4dihydroxybenzaldehyde (Compound 15). Inspection of the focused Hb bands 2 h after onset of electrophoresis showed that about 90% of the hemoglobin was formed into a complex with aldehyde. Inspection of the gels after 3l12 and 5 h of focusing revealed that the extent of adduct formation had decreased to 70 and 50%, respectively. The first pK, for the ionization of the hydroxyl groups of 15 was determined by titration to be 7.45. Thus, at the p1 of the HbA.Schiff base complex (6.901, the free aldehyde will be about 20% ionized. As a result, the free aldehyde will migrate from the position of the hemoglobin species toward the anode, this migration leading to net dissociation of the Schiff base complex. Parallel tests with the adduct of hemoglobin and benzaldehyde (11, which lacks an ionizable group, showed no change in the extent of modification with time of electrofocusing. These results demonstrate both the reversibility of Schiff base formation within the gel and the effect of the electric field on the stability of the equilibrium when the aldehyde carries a net charge. The above analysis of electric field effects explains the paradoxical results obtained with 5nitrosalicylaldehyde (Compound 1’7). Treatment of erythrocytes with this compound caused a significant increase in the oxygen affinity of hemoglobin (%AP,,, = 55) although no modification was detected by electrofocusing (see Table I). Clearly, a certain extent of chemical modification must be occurring if we are to account for the observed changes in the oxygen-binding properties of the protein. The hydroxyl group of 17 is ortho to a formyl group and para to a nitro group and would be expected to have a very low pK, relative to phenol. This expectation was confirmed by spectrophotometric examination of 0.1 mM 5-nitrosalicylaldehyde in a series of 0.05 M acetic acid/sodium acetate buffers, pH 4.0 to 5.5. The pK, of the phenolic -OH was found to be 5.30. At the isoelectric pH of the Hb.Schiff base complex (6.901, the free aldehyde, being more than 97% ionized, would migrate more rapidly than 2,4-dihydroxybenzaldehyde. As a result, the free aldehyde concentration would be depleted before a stable adduct could be observed. Aside from Compounds 8,9, 17, and 24, all other compounds

That

Hemoglobin

Modifications

That

Inhibit

Sickling

8547

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

cannot permeate the erythrocyte membrane and is therefore of intracellular hemoglobin, as well as mcreases m oxygen nonreactive. The low level of observed modification with affinity, equivalent to those observed with normal erythropyridoxal, which is membrane-permeable, is difficult to excytes. Since studies conducted in whole normal blood did not plain, especially in light of its potent effect on oxygen binding. reveal large reductions in the extents of modification relative At neutral pH, 23 is less than 5% ionized (27) so that the low to those obtained in buffer, the remaining studies with sickle level of modification cannot be explained by an electric field erythrocytes were done in AB plasma. effect as in the case of 17. Other factors must be responsible Fig. 4 shows the effects of several aromatic aldehydes on for the large discrepancy between extent of modification and the oxygen binding of sickle cell erythrocytes in whole blood. oxygen binding in this case. As for normal erythrocytes, o-OH-substituted benzaldehydes The available evidence from model studies indicates that produce the greatest increases in oxygen affinity. The oxygenSchiff base formation with small primary amines occurs binding curve for sickle cells treated with an equivalent somewhat more readily with aromatic aldehydes than with concentration of potassium cyanate is also shown. At this aliphatic. Hine et al. (28), in comparing Schiff base equilibria concentration (5 mM), cyanate has little effect on oxygen between methylamine and isobutyraldehyde or 4-pyridinecarbinding. Comparison is made with cyanate since this comboxaldehyde, found the aromatic imine to have a lo-fold pound inhibits sickling by increasing the oxygen affinity of greater equilibrium constant (87 M-‘) than the aliphatic imine HbS (32). Fig. 5 shows that aromatic aldehydes likewise (8.5 Mm’). Pesek and Frost (29) obtained qualitatively similar inhibit erythrocyte sickling. Indeed, salicylaldehyde (2) and results upon equilibrating n-butylamine with isobutyraldeo-vanillin (16) are approximately lo-fold more effective than hyde (K,, = 16 M-‘I, cyclohexanecarboxaldehyde (K,, = 28 cyanate in reducing the extent of sickling at equivalent Mm’), and benzaldehyde (Kc, = 40 M-I). oxygen tensions. Convergence of the curves in Fig. 5 at low Extents of hemoglobin modification obtained in the present ~0, values suggests that these aldehydes have little effect on study (see Table I) can be readily converted to apparent sickling in the absence of oxygen. Thus, as is the case with equilibrium constants. A typical aromatic aldehyde (at a free cyanate, the anti-sickling effects of these compounds appear concentration of 5 mM) eliciting 50% modification would have to be mediated through increases in the oxygen affinity of = 220 Mm’, whereas aliphatic aldehydes 26 and 27, intracellular hemoglobin S. a K, providing 25% modification, have a K,, = 70 M-‘. These Sickle erythrocytes have a diminished oxygen affinity relavalues, which represent lower limits to the true equilibrium tive to normals (compare control curves in Figs. 3 and 4). In values, are 3- to &fold greater than typical Schiff base equilibview of observations reported for 5’-deoxypyridoxal (18) and rium constants determined with small molecule amines. This those described here for a variety of aldehydes, it seems enhancement may reflect a macromolecular environment reasonable to expect that artificial elevation of the oxygen which favors the adduct relative to the starting materials, affinity of sickle blood into the normal range by administraperhaps by excluding water as it forms. tion of certain aromatic aldehydes could alleviate sickling UL Normal erythrocytes showed marked increases in oxygen uiuo. Cyanate inhibits sickling in uivo by this mechanism, affinity when exposed to single, low doses of certain aromatic but therapeutic levels give rise to adverse side effects (33). aldehydes (see Table I). Aliphatic aldehydes and ketones The greater efficacy of the present compounds may permit manifested no effect on oxygen binding despite significant their use at lower dosages, well within an acceptable margin adduct formation in certain instances. Although modification of safety. by aromatic aldehydes was necessary to influence oxygen binding, it was not a sufficient condition. Thus, benzaldehyde, Acknowledgments -We express our appreciation to Drs. a potent modifier, had little or no effect on oxygen affinity. Ashok Pate1 and K. R. P. Rao, Department of Hematology, The most potent effecters of oxygen binding were those Cook County Hospital, Chicago, for providing residual sickle aromatic aldehydes which contain an o-hydroxyl group (2, 14 cell blood used in this study and for helpful discussions. We to 17, and 23). This substituent may direct these compounds are also indebted to Mrs. Janet Goranson for her assistance to specific sites on hemoglobin which strongly influence the with the manuscript. equilibrium. In support of this, we found that OXY e deoxy REFERENCES benzaldehyde (at 3 or 5 mM) did not compete with 2,4-dihy1. Stetson, C. A. (1966) J. Ezp. Med. 123, 341-346 droxybenzaldehyde (3 mM) to diminish the increased oxygen 2. MaEdoff-Fairchild. B.. Swerdlow. P. H.. and Bertles. J. F. affinity elicited by the latter. (1972) Nature 2i9, 217-218 Two sites at which modification should increase the oxygen 3. Cerami, J. A., and Manning, J. M. (1971)Proc. Natl. Acad. SCL. affinity are the NH,-terminal amino groups of the a-chains. U. S. A. 68, 1180-1183 4. Nigen, A. M., Njikam, N., Lee, C. K., and Manning, J. M. Benesch et al. (18) have recently reported increases in oxygen (1974) J. Biol. Chem. 249, 6611-6616 affinity upon attachment of 5’-deoxypyridoxal at these sites. 5. Klotz, I. M., and Tam, J. W. 0. (1973) Proc. Natl. Acad. SCL. U. Likewise, specific carbamylation of these amino groups by S. A. 70. 1313-1315 cyanate leads to an increased oxygen affinity (4). Alterna6. Shamsuddin, M., Mason, R. G., Ritchey, J. M., Honig, G. R., tively, Schiff base formation by aromatic aldehydes in the and Klotz, I. M. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, 4693-4697 region of the P-chain NH, termini could increase the oxygen I. Zaugg, R. H., King, L. C., and Klotz, I. M. (1975) Biochem. affinity by preventing normal binding at this site of 2,3Biophys. Res. Commun. 64, 1192-1198 diphosphoglycerate, a potent negative effector of oxygen bind8. Holmquist, W. R., and Schroeder, W. A. (1966) Biochemistry 5, ing (30). Schiff base formation at Lys 4Oa, which forms a salt 2489-2503 9. Bookchin, R. M., and Gallop, P. M. (1968) Biochem. B~o&s. bridge in deoxyhemoglobin to the COOH-terminal carboxylate Res. Commun. 32. 86-93 group of His 146p (311, may also tip the oxy ti deoxy 10. Bunn, H. F., Haney, D. N., Gabbay, K. H., and Gallop, P. M. equilibrium in favor of the oxy form. (1975) Biochem. Biophys. Res. Commun. 67, 103-109 Exposure of sickle erythrocytes suspended in isotonic phos11. Haney, D. N., and Bunn, H. F. (1976) Proc. Natl. Acad. SCL. U. phate buffer to a variety of aldehydes produced modification S. A. 73, 3534-3538

8548

Hemoglobin

Modifications

12. Abdella, P. M., Ritchey, J. M., Tam, J. W. O., and Klotz, I. M. (1977) B&him. Biophys. Acta 490, 462-470 13. Dixon, H. B. F. (1972) Biochem. J. 129, 203-208 14. Benesch, R. E., Benesch, R., Renthal, R. D., and Maeda, N. (1972) Biochemistry 11, 3576-3582 15. Benesch, R. E., Yung, S., Suzuki, T., Bauer, C., and Benesch, R. (1973) Proc. Natl. Acad. Sci. U. S. A. 70, 2595-2599 16. Benesch, R., Benesch, R. E., and Yung, S. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, 1504-1505 17. Benesch, R., Benesch, R. E., Yung, S., and Edalji, R. (1975) Biochem. Biophys. Res. Commun. 63, 1123-1129 18. Benesch, R., Benesch, R. E., Edalji, R., and Suzuki, T. (1977) Proc. Natl. Acad. Ser. U. S. A. 74, 1721-1723 19. Drysdale, J. W., Righetti, P., and Bunn, H. F. (1971) Biochim. Biophys.

Acta

229,

42-50

20. Bunn, H. F. (1973) Ann. N. Y. Acad. Sci. 209, 345-353 21. Means, G. E., and Feeney, R. E. (1971) Chemical Modification ofproteins, pp. 125-129, Holden-Day, San Francisco 22. Sander, E. G., and Jencks, W. P. (1968) J. Am. Chem. Sot.

90,

That 23.

24. 25.

26. 27.

Inhibit

Sickling

6154-6162 Benesch, R. E., and Benesch, R. (1962) Biochemistry 1, 735-738 Smith, J. W. (1970) in Chemistry of the Carbon-Nitrogen Double Bond (Patai, S., ed) pp. 235-253, Interscience, New York Layer, R. W. (1963) Chem. Rev. 63, 489-510 Jencks, W. P. (1964) Prog. Phys. Org. Chem. 2, 63-128 Williams, V. R., and Neilands, J. B. (1954) Arch. Blochem. Biophys.

28. Hine,

53, 56-70

J., Via,

F. A., Gotkis, J. K., and Craig, J. C., Jr. (1970) Sot. 92, 5186-5193 J. J., and Frost, J. H. (1976) Org. Magnetzc Res. 8, 173-

J. Am.

29. Pesek, 176 30. Benesch,

Chem.

R. E., and Enoki, Y. (1968) Proc. N&Z. 61, 1102-1106 31. Perutz, M. F., Ladner, J. E., Simon, S. R., and Ho, C. (1974) Bzochemistry 13, 2163-2173 32. Diederich, D. (1972) Biochem. Biophys. Res. Commun. 46, 1255 1261 33. Harkness, D. R., and Roth, S. (1975) Progr. Hematol. 9, 157-184 Acad.

R., Benesch,

Sci.

U. S. A.

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

Schiff base adducts of hemoglobin. Modifications that inhibit erythrocyte sickling. R H Zaugg, J A Walder and I M Klotz J. Biol. Chem. 1977, 252:8542-8548.

Access the most updated version of this article at http://www.jbc.org/content/252/23/8542 Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts

Downloaded from http://www.jbc.org/ by guest on August 10, 2017

This article cites 0 references, 0 of which can be accessed free at http://www.jbc.org/content/252/23/8542.full.html#ref-list-1