Differential Inhibition of Maillard Protein Fluorescenceby Nitric Oxide Donors

Gerardo Suarez1 and Xiao-Hong Wang

Department of Biochemistry and Molecular Biology, New York Medical College, Valhalla, New York 10595

Received October 30, 1998

The nitric oxide donors nitroprusside (NP) andS-nitroso-N-acetylpenicillamine (SNAP) were addedrepeatedly over a prolonged period into a proteinfructation system of 0.05 M fructose and BSA. Theseadditions inhibited Maillard reaction advanced-stage fluorescence generation in a dose-dependentmanner without affecting initiation of glycation.NP caused 66% inhibition whereas SNAP causedonly 30% inhibition at maximum dose. The lowerinhibition by SNAP possibly reflects an interfer-ence caused by N-acetylpenicillamine and mediatedby a metal-dependent enhanced free-radical gener-ation. We propose that the inhibition of fluores-cence results from mutual annihilation between ni-tric oxide and free radicals, such as OH z , producedduring fructation. In vivo generated nitric oxidemay play a protective role in cells against the dele-terious effect of free radicals that are associatedwith the augmented fructose autoxidation and fruc-tation that occurs in diabetes. © 1998 Academic Press

Key Words: nitric oxide; fructation; nitroprusside;S-nitrosothiols; Maillard; fluorescence; free radi-cals.

Evidence has been presented in support of thegeneration of reactive oxygen species (ROS)2 alongthe Maillard reaction pathway of protein glycation

and concurrent sugar autoxidation (1). ROS is anexpression that denotes free radicals or oxidizingagents that are not necessarily free radicals but thatmay contribute to their generation. ROSs that havebeen found linked to the Maillard reaction includethe superoxide (O2

•2) (2) and hydroxyl ( zOH) (3) rad-icals, H2O2 (4), and alcoxyl radicals derived fromfree sugars and Amadori-like adducts (1, 2, 5). Theseintermediate molecular species are thought to play arole in the formation of advanced Maillard reactionfluorophores (3, 5) which have excitation maxima inthe near UV region.

Nitric oxide (NO z), shown to be a major mediatorof the effect of the endothelial-derived relaxing fac-tor, is an endogenously generated gas with multiplefunctions which include, besides the modulation ofvascular tone, participation in neural transmissionand cytotoxicity toward tumor cells and invasiveorganisms (6). As a result of having an unpairedelectron, NO z is also a free radical of high reactivitywhich explains its half-life of seconds in an aqueousmedium (7). As a free radical NO z participates inheteromolecular annihilation reactions with otherfree radicals. Of special interest is its reaction withthe hydroxyl radical due to its generation duringglycation (3). It has been recently demonstrated thatNO z condenses with OH z radicals that are generatedby either UV photolysis or by the Fenton reactionwith production of nitrite (8). Hence, NO z could be,1 To whom correspondence should be addressed at Department

of Cell Biology and Anatomy, New York Medical College, BSB242A, Valhalla, NY 10595. Fax: (914) 594-4653.

2 Abbreviations used: BSA, bovine serum albumin; NO z , nitricoxide; NAP, N-acetylpenicillamine; NP, sodium nitroprusside;

ROS, reactive oxygen species; PBS, phosphate-buffered saline;SNAP, S-nitroso-N-acetylpenicillamine.

NITRIC OXIDE: Biology and ChemistryVol. 2, No. 6, pp. 475–480 (1998)Article No. NO980200

4751089-8603/98 $25.00Copyright © 1998 by Academic PressAll rights of reproduction in any form reserved.

in principle, an important modulator of the potentialtoxic effects of free radicals generated in conjunctionwith nonenzymatic glycation.

We now describe evidence that NO z inhibits ad-vanced Maillard fluorescence and propose that thiseffect is mediated by its free radical reactivity to-ward intermediate ROS occurring in the glycationpathway. To provide a stringent test for this hypoth-esis we have chosen fructose as a glycating sugarbecause it induces about 10 times the fluorescencegenerated during glycation by glucose (9). In addi-tion, fructose is also of physiological relevance be-cause of its involvement in the complications of di-abetes (10).

MATERIALS AND METHODS

Bovine serum albumin (BSA), crystalline, fattyacid free, and fructose (b-D(2)-fructose) were fromSigma Chemical Co. St. Louis, Missouri. S-Nitroso-N-a-acetylpenicillamine (SNAP), a nitric oxide do-nor, was synthesized from N-a-acetylpenicillamine(NAP) as described (11). Sodium nitroprusside (NP)was purchased from Fisher Scientific.

In vitro glycation. BSA was incubated at a con-centration of 6 mg/ml in the presence of fructose atan initial concentration of 0.05 M in 0.05 M phos-phate, 0.1 M NaCl, pH 7.45 (phosphate-buffered sa-line, PBS), under sterile conditions in a humidifiedatmosphere, at 37°C. Since the concentrations ofNO z donors drop steeply due to their instability,these compounds were added periodically at inter-vals of 4–5 days. Following an initial 4-day incuba-tion period at various nitric oxide donor concentra-tions that ranged from 0.47 to 3.2 mM, subsequentadditions of corresponding equal amounts (0.7–4.9mmol, 0.3 ml) were made four times for each incu-bation condition. All additions were made fromfreshly prepared stock solutions. Incubation mix-tures were protected from exposure to light duringtheir preparation and incubation. The total amountof nitric oxide donor added ranged from 28 to 196mmol/mol fructose. Incubation and additions wereconducted under sterile conditions. Sterilization wasaccomplished by passage of the stock solutionsthrough ultrafiltration membranes with 0.45-mmpores (Acrodisc, Gelman). Following incubation for a

total period of 21 days the mixtures were dialyzed at4°C against a 20-fold volume excess of PBS witheight changes of the buffer. The dialyzed sampleswere stored at 4°C until analyzed.

Analysis of Maillard reaction products in proteins.Quantification of indicators of the Maillard reactionwas accomplished, in general, as described bySuarez et al (9). Fluorescence was measured at anexcitation wavelength of 340 nm and an emissionwavelength of 420 nm with a Perkin–Elmer lumi-nescence spectrometer and the emission readingswere normalized to both a constant protein concen-tration of 10 mg/ml and a constant reading of thestandard 4 (tetrabutylenediamine), provided by themanufacturers. Fluorescence was corrected for fluo-rescence of nonfructated BSA. Mean values of dupli-cate assays were used for the quantifications.

RESULTS

Because of the fast decay of nitric oxide donors insolution their effect on the Maillard reaction wasevaluated in terms of the total mass of donor addedrelative to the initial fructose concentration ratherthan as a function of their initial concentration. Thisexpression of results also obviates dilution effectsresulting from the periodic additions which totaled a1.4-fold at the end of the incubation. As shown inFig. 1 additions of low amounts of NP induced amodest but reproducible enhancement of fluores-cence during fructation; however, higher concentra-tions of NP inhibited fructation-induced fluores-cence generation up to a maximum of 66%. As isshown by a representative experiment in Fig. 2,bottom curve, the alternative NO donor SNAP alsocaused inhibition, ranging from 4 to 30 %, that is,less than half the inhibition caused by NP at maxi-mal molar ratio. Since NAP, a by-product of theNO z-releasing reaction, is more stable than SNAPand, thus, can participate in further reactions forlonger time periods, we compared its effect with thatof SNAP. Figure 2 (top curve) shows that in thepresence of NAP fluorescence during fructationreached values that were more than double those inits absence. A large fraction of the maximum effectwas attained at the lowest level of NO z donor tested.

SNAP did not modify the extent of Amadori groupgeneration during fructation (Fig. 3).

476 SUAREZ AND WANG

Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.

DISCUSSION

As supported by the effect of both NO donors,nitric oxide appears to inhibit the fluorogenic stagesof the Maillard reaction initiated by fructation.However, the much lower SNAP inhibition couldresult from interference caused by SNAP reactionproducts. As recently shown, NO z release from anS-nitrosothiol compound results from a complex se-ries of reactions with a free sulfhydryl group (aspresent in BSA), generating the oxidized thiol dimeras a by-product (12). Under the conditions of our

experiments the dithio dimer could exchange thiolgroups with BSA thus regenerating in part the re-duced monomeric thiol, NAP. The generation of flu-orescence in the presence of NAP alone (Fig. 2, topcurve) suggests that the net effect of SNAP is acomposite of the separate effects of NAP and nitricoxide. As has been proposed for a prooxidant effect ofpenicillamine (13), the NAP sulfhydryl group couldchelate and reduce trace amounts of metals. Thereduced metal would subsequently participate intwo one-electron transfers to molecular oxygenforming hydrogen peroxide. Thus, through severalredox cycles, both reduced metal and H2O2 wouldcoexist and would generate OH z via the Fenton re-action. The observed fluorescence enhancement by

FIG. 2. Effect of NAP or SNAP on fructose-dependent Maillardfluorescence generation. NAP (E) or SNAP (F) was added in thesame amounts and at the intervals described in the legend to Fig.1. Fluorescence was measured by the same method.

FIG. 1. Effect of nitroprusside on fructose-dependent Maillardfluorescence generation. To a fructation system (0.05 M fructose,6 mg/ml BSA) several partial doses of NP were added at 4- to5-day intervals to complete the total mass of added donor relativeto sugar as indicated (0–196 mmol/mol), during a total incubationperiod of 21 days. Fluorescence (excitation 340 nm; emission 420nm) was measured on dialyzed samples as described in text.

477INHIBITION OF THE MAILLARD REACTION BY NITRIC OXIDE

Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.

NAP could thus be explained by the known promo-tion of Maillard fluorescence by OH z (3). The nitro-prusside effect might not be antagonized, at least tothe same extent, by free-radical-generating activi-ties of products arising after NO z release. On thebasis of these considerations NAP controls should beincluded in all studies that utilize SNAP as aNO z donor.

NO z probably does not affect de novo glycation,i.e., does not inhibit the initial stages of the Maillardreaction. This is supported by the observation thatSNAP did not modify the extent of Amadori groupgeneration during fructation (Fig. 3). This observa-

tion restricts the targets of nitric oxide to the ad-vanced stages of the Maillard reaction.

In conclusion, our results argue in favor of a spe-cific effect of NO z on advanced stages of the Maillardreaction that mediate fluorescence generation. Themechanism of NO z inhibition of fluorescence is un-known at present. As a plausible explanation forthis inhibition we suggest that NO z interferes withreactions between free sugars or Amadori groupsand hydroxyl radicals by trapping the latter.

X2Fe112NO 1 H2O23 X2Fe112NOz O z H

} { }H O [1]

X2Fe112NO X2Fe1112NO X2Fe111

O z H 3 O2 z H 3 1OH2 1 HNO2}{} P P

H O H z O [2]

Following very closely the model proposed for theantioxidant effect of NO z referred to above (8), in aglycation reaction a free sugar, an Amadori group,phosphate ions, or protein residues (X) would formcomplexes with transition metals at lower oxidationstate, such as Fe11, which in turn could coordinatewith hydrogen peroxide, an intermediate product ofthe Maillard reaction (4), initiating in this way theFenton reaction (8) and OH z generation. NO z coulddisplace hydrogen peroxide from the metal butwould be able to also bind hydrogen peroxide (reac-tion [1]). One-electron transfer from the metal to thebound hydrogen peroxide and subsequent hetero-lytic fission would generate HO2 and HO z at thesite in closeproximity to both the oxidized metal andthe NO z . HO z , due to its proximity to NO z , wouldreadily generate nitrite, by free-radical annihilation(reaction [2]). Summing up, the antioxidant effect ofNO z is a consequence of its site-specific OH z radical-scavenging capacity (14), which is further supportedby a recent demonstration of a correlation betweenthe NO z-releasing potential of NP and its OH z-scav-enging activity (15). Maillard fluorescence inhibitionwould result because NO z prevents the availabilityof z OH which is required in reactions that lead tothe generation of dicarbonyls (1, 3) which, in turn,are precursors of Maillard reaction fluorophores.

FIG. 3. Effect of NAP (E) and SNAP (F) on fructose-dependentAmadori (Heyns) group generation. Additions of NAP or SNAPwere performed as described in legend to Fig. 1. Amadori groupconcentrations were measured on the basis of periodate-inducedformaldehyde release (see text).

478 SUAREZ AND WANG

Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.

During in vitro protein fructation peroxynitritecould also be produced by the annihilation of Mail-lard generated superoxide anion and added nitricoxide. Peroxynitrite has been shown to be a strongoxidant, and this could explain the enhanced fluo-rescence at low NO z /fructose ratio (Fig. 1). However,it has been shown that an excess nitric oxide/superoxide anion molar ratio abolishes the prooxi-dant effect of peroxynitrite (16). Thus at higher lev-els of added nitric oxide only its inhibitory effectwould become evident, in accordance with our re-sults.

Because of the short half-life of NO z (7) it would beunreasonable to expect significant inhibition byNO z in prolonged glycation reactions if a very fastrelease from the donor is assumed. However, a half-life of 4.6 h has been documented for SNAP underexperimental conditions similar to those prevailingin the present study (17). This much greater stabil-ity, in conjunction with the fact that NO z is capableof reacting with sulfhydryl groups of protein cys-teine residues with the formation of even more sta-ble S-nitroso derivatives (18, 19), strengthens thefeasibility of slow delivery of NO z to make it effectivefor long periods.

If a Maillard reaction inhibitory capacity can bedemonstrated for NO z in vivo, then it is conceivablethat NO z might play a role in cellular self-detoxification from glycation or sugar autoxidationproducts, especially in those cells where NO z syn-thesis and free-radical generation due to fructationare concurrent processes. This notion is supportedby a protection by NO z of oxidative modification oflow-density lipoprotein caused by mouse macro-phages (20). In addition, nitric oxide donors abro-gate the cytotoxicity caused by reactive oxygen spe-cies in rat fibroblasts and mesencephalic dopa-minergic cells (21).

REFERENCES

1. Wolff, S. P., Jiang, Z. Y., and Hunt, J. V. (1991). Proteinglycation and oxidative stress in diabetes mellitus and age-ing. Free Rad. Biol. Med. 10, 339–352.

2. Sakurai, T., and Tsuchiya, S. (1988). Superoxide productionfrom nonenzymatically glycated protein. FEBS Lett. 236,406–410.

3. Hunt, J. V., Dean, R. T., and Wolff, S. P. (1988). Hydroxylradical production and autoxidative glycosylation. Biochem.J. 256, 205–212.

4. Jiang, Z.-Y., Woollard, A. C. S., and Wolff, S. P. (1990). Hy-drogen peroxide production during experimental protein gly-cation. FEBS Lett. 268, 69–71.

5. Wolff, S. P., and Dean, R. T. (1987). Glucose autoxidation andprotein modification. Biochem. J. 245, 243–250.

6. Moncada, S., Palmer, R. M. J., and Higgs, E. A. (1991). Nitricoxide: Physiology, pathology, and pharmacology. Pharmacol.Rev. 43, 109–142.

7. Kikuchi, K., Nagano, T., Hayakawa, H., Hirata, Y., and Hi-robe, M. (1993). Real time measurement of nitric oxide pro-duced ex vivo by luminol-H2O2 chemiluminescence method.J. Biol. Chem. 268, 23106–23110.

8. Kanner, J., Harel, S. and Granit, R. (1991). Nitric oxide as anantioxidant. Arch. Biochem. Biophys. 289, 130–136.

9. Suarez, G., Rajaram, R., Oronsky, A. L., and Gawinowicz,M. A. (1989). Nonenzymatic glycation of bovine serum albu-min by fructose (Fructation). J. Biol. Chem. 264, 3674–3679.

10. Suarez, G. (1989). Nonenzymatic browning of proteins andthe sorbitol pathway. In The Maillard Reaction in Aging,Diabetes and Nutrition (Baynes, J. W., and Monnier, V. M.,Eds.), pp. l4l–162, A. R. Liss, New York.

11. Field, L., Dilts, R., Ravichandran, R., Lenhert, P. G., andCarnahan, G. E. (1978). An unusually stable thionitrite fromN-acetyl-D,L-penicillamine: X-ray crystal and molecularstructure of 2-(acetylamino)-2-carboxy-1,1-dimethylethylthionitrite. J. Chem. Soc. Chem. Commun. 1978, 249–250.

12. Singh, S. P., Wishnok, J. S., Keshive, M., Deen, W. M., andTannenbaum, S. R. (1996). The chemistry of theS-nitrosoglutathione/glutathione system. Proc. Natl. Acad.Sci. USA 93, 14428–14433.

13. Miki, M. (1994). Penicillamine as antioxidant. Methods En-zymol. 234, 542–547.

14. Chevion, M. (1988). A site-specific mechanism for free radicalinduced biological damage: The essential role of redox-activetransition metals. Free Rad. Biol. Med. 5, 27–37.

15. Dalloz, F., Maupoil, V., Lecour, S., Briot, F., and Rochette, L.(1997). In vitro studies of interactions of NO z donor drugswith superoxide and hydroxyl radicals. Mol. Cell. Biochem.177, 193–200.

16. Miles, A. M., Bohle, D. S., Glassbrenner, P. A., Hansert, B.,Wink, D. A., and Grisham, M. B. (1996). Modulation ofsuperoxide-dependent oxidation and hydroxylation reactionsby nitric oxide. J. Biol. Chem. 271, 40–47.

17. Ignarro, L. J., Lippton, H., Edwards, J. C., Baricos, W. H.,Hyman, Albert. L., Kadowitz, P. J., and Gruetter, C. A.(1981). Mechanism of vascular smooth muscle relaxation byorganic nitrates, nitrites, nitroprusside and nitric oxide: Ev-idence for the involvement of S-nitrosothiols as active inter-mediates. J. Pharmacol. Exp. Ther. 218, 739–749.

479INHIBITION OF THE MAILLARD REACTION BY NITRIC OXIDE

Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.

18. Stamler, J. S., Simon, D. I., Osborne, J. A., Mullins, M. E.,Jaraki, O., Michel, T., Singel, D. J., and Loscalzo, J. (1992).S-Nitrosylation of proteins with nitric oxide: Synthesis andcharacterization of biologically active compounds. Proc. Natl.Acad. Sci. USA 89, 444–448.

19. Stamler, J. S., Jaraki, O., Osborne, J., Simon, D. I., Keaney, J.,Vita, J., Singel, D., Valeri, C. R., and Loscalzo, J. (1992). Nitricoxide circulates in mammalian plasma as an S-nitroso adduct ofserum albumin. Proc. Natl. Acad. Sci. USA 89, 7674–7677.

20. Hogg, N., Struck, A., Goss, S. P. A., Santanam, N., Joseph,J., Parthasarathy, S., and Kalyanaram, B. (1995). Inhibi-tion of macrophage-dependent low density lipoproteinoxidation by nitric oxide donors. J. Lipid Res 36, 1756 –1762.

21. Wink, D. A., Hanbauer, I., Krishna, M. C., DeGraff, W.,Gamson, J., and Mitchell, J. B. (1993). Nitric oxide protectsagainst cellular damage and cytotoxicity from reactive ox-ygen species Proc. Natl. Acad. Sci USA 90, 9813–9817.

480 SUAREZ AND WANG

Copyright © 1998 by Academic Press. All rights of reproduction in any form reserved.