Quantitation of 3-Deoxyglucosone Levels in Human Plasma

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 342, No. 2, June 15, pp. 254–260, 1997Article No. BB970117

Quantitation of 3-Deoxyglucosone Levels in Human Plasma

Sundeep Lal,*,1 Francis Kappler,* Michael Walker,* Trevor J. Orchard,† Paul J. Beisswenger,‡Benjamin S. Szwergold,* Truman R. Brown**Department of NMR and Medical Spectroscopy, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111;†Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania15261; and ‡Department of Endocrinology, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire 03756

Received December 30, 1996, and in revised form March 26, 1997

from normoglycemic plasma were similar to those re-ported by Niwa et al. (1710 { 750 (SD) nM). These re-3-Deoxyglucosone (3DG), a reactive dicarbonyl, is ansults suggest that 3DG levels measured by ultrafiltra-important intermediate in the formation of advancedtion may represent the free circulating 3DG and thoseglycation end products (AGEs). The AGEs are particu-obtained by ethanol extraction may represent aformlarly important in diabetes since they have been corre-of 3DG bound to a macromolecule (presumbaly pro-lated with the development of diabetic complications.tein). q 1997 Academic PressConsequently, measurements of 3DG are likely to pro-

Key Words: 3-deoxyglucosone; plasma; nonenzymaticvide valuable insights into the role of this metaboliteglycation; diabetes.in the etiology of diabetic complications. While several

methods of 3DG quantitation in human plasma havebeen previously published, a significant discrepancy(over 30-fold) exists in the reported values. Knecht etal. (Arch. Biochem. Biophys. 294, 130–137, 1992) have Although the Diabetes Control and Complication Tri-reported the levels of plasma 3DG in normoglycemics als (1) have firmly established a correlation betweento be 61 nM, using a GC/MS procedure. In contrast to hyperglycemia and the incidence of diabetic complica-this, Niwa et al. (Biochem. Biophys. Res. Commun. 196, tions, the detailed mechanism associating the two re-837–843, 1993) reported 3DG levels to be 1800 nM in mains unknown. Previous studies have suggested thatnormoglycemics, using a totally independent GC/MS formation of advanced glycation end products (AGEs)2

method. To resolve this disagreement and fill the need may be an important factor involved in the etiology offor a robust assay for this dicarbonyl, suitable for abso-diabetic complications (2–5). The initial steps in thelute quantitation, a GC/MS procedure was devised forformation of AGEs are mediated by production of ele-its measurement. Plasma samples were deproteinizedvated concentrations of reactive dicarbonyls such as 3-either by ultrafiltration or by addition of ethanol asdeoxyglucosone (3DG) and methyl glyoxal (6–8). Whiledescribed by Niwa et al. (Biochem. Biophys. Res. Com-the pathways for production and subsequent detoxifi-mun. 196, 837–843, 1993). 3DG in the ultrafiltrate orcation of methyl glyoxal are well known (8, 9), thosethe supernatant was conjugated with 2,3-diamino-for 3DG still remain to be established.naphthalene to produce a stable adduct which was

A critical issue in understanding the pathways in-then converted to a silyl ether and analyzed by GC/volving 3DG is the development of robust and reliableMS. The analyte was monitored by selected ion moni-

toring at an m/z of 295 and 306 and quantitated using methodologies for its measurement. This is a signifi-an internal standard of [U-13C]3DG. Using this ap- cant challenge since 3DG has multiple anomeric formsproach, 3DG levels in plasma deproteinized by ultra- (10), and due to its high reactivity (11–13), it is presentfiltration were found to be significantly elevated from only in low amounts in its free form in vivo. To address58.5 { 14 (SD) nM in normoglycemics to 98.5 { 34 (SD) this challenge, several approaches for detection of 3DGnM in type I diabetics. When deproteinization of the have been reported (14–17). Knecht et al. analyzedplasma was carried out using ethanol, the levels of 3DG

2 Abbreviations used: AGEs, advanced glycation end products;1 To whom correspondence should be addressed at Department of 3DG, 3-deoxyglucosone; DAN, 2,3-diaminonaphthalene; TMS, Tri-

Sil; SIC, selected ion chromatogram; 3-DGA, 2-keto-3-deoxygluconicNMR and Medical Spectroscopy, Fox Chase Cancer Center, 7701Burholme Ave., Philadelphia, PA 19111. acid.

254 0003-9861/97 $25.00Copyright q 1997 by Academic Press

All rights of reproduction in any form reserved.

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca

255MEASUREMENT OF PLASMA 3-DEOXYGLUCOSONE

used for quantitation. This procedure was repeated in triplicate.plasma 3DG levels by reduction to 3-deoxyhexitols withThe concentration of the synthetic 3-[U-13C]DG obtained in thisNaBD4 and subsequent analysis of their acyl deriva-manner was 91.5% of that calculated by 13C NMR. The coefficienttives by GC/MS (14). No 3DG measurements on diabet- of variance of the GC/MS procedure was 1.7%. The final concentra-

ics were, however, reported. Niwa et al. utilized an- tion of the internal standard was based on the value obtained bythis latter GC/MS procedure.other GC/MS method (15, 16), whereby 3DG was mea-

sured as silyl ethers of the oxime derivative of its Collection of plasma and description of the patient population.After an overnight fast, approximately 8–10 ml of blood was drawncarbonyl groups. Using this methodology, a 2-fold ele-in Vacutainers (Becton–Dickinson, NJ) containing EDTA from nor-vation in 3DG levels was noted in diabetics when com-moglycemic volunteers and individuals with type I diabetes. Plasmapared to normals (15, 16). However, in this latter study, was obtained by centrifugation of the blood and was stored at 0707C

3DG concentrations in serum from normals were found until processed.Diabetic population constituted of 16 randomly selected partici-to be 30-fold higher than what was previously reported

pants attending an 8-year follow-up examination of the Pittsburghby Knecht et al.Epidemiology of Diabetes Complication Study, a 10-year prospectiveThe present study was undertaken to resolve thisfollow-up study of a representative of childhood diabetic subjects

significant discrepancy in reported values for concen- described previously (19, 20). Clinical characteristics of these pa-tration of 3DG in human plasma and to develop a reli- tients and the normoglycemic controls are described in Tables I and

II, respectively. Normoglycemic volunteers (n Å 17) were recruitedable and sensitive method for absolute quantitation offrom the staff of Fox Chase Cancer Center.3DG. To this end, we adapted a protocol recently re-

Processing of the plasma. One milliliter of plasma was deprotein-ported by Yamada et al. (17) for measurement of 3DGized using Centrifree filters (Amicon, Inc., Beverly, MA). A 20-mlin the rat plasma, whereby 3DG is conjugated to 2,3-aliquot of 10 mM 3-[U-13C]DG was added as internal standard to 0.35

diaminonaphthalene (DAN) and the resultant non- to 0.5 ml of the plasma followed by 1 ml of 1 mM 2,3-diaminonaphtha-aqueous product analyzed by HPLC using a fluores- lene solution. After an overnight incubation at room temperature,

the mixture was extracted with 4 ml of ethyl acetate and dried com-cence or UV detector. We modified this procedure forpletely by centrifugal evaporation. A 120-ml aliquot of TMS reagentuse with a GC/MS by converting the 3DG–DAN conju-was added under argon to make trimethyl silyl derivatives of thegate to a volatile silyl ether. In addition to this, an 3DG–DAN adduct. The sample vials were capped and vigorously

isotopic internal standard of 3-[U-13C]DG was used to stirred, and after 30 min at room temperature, they were centrifugedcompensate for the reactivity of plasma 3DG. The con- on a bench top centrifuge to settle salts formed during derivatization.

The derivatized sample was then transferred to another vial con-centrations of 3DG in normal and diabetic humantaining a limited volume insert and sealed under argon.plasma were measured. Our values for 3DG in normog-

Comparison of deproteinization methods: Ethanol extraction vs ul-lycemics after ultrafiltration of plasma were in closetrafiltration. Plasma samples from eight normoglycemic individu-agreement with those of Knecht et al. (14). Deproteini- als were deproteinized in two ways: by ultracentrifugation as de-

zation of plasma from normoglycemics with ethanol, as scribed above or by ethanol as described below.described by Niwa et al., resulted in substantially To 0.5 ml of plasma, a 20-ml aliquot of 10 mM 3-[U-13C]DG was

added as internal standard, followed by 2 ml of ethanol. After vigor-higher levels of detectable 3DG, similar to those ob-ous shaking, the mixture was centrifuged on a bench top centrifuge.served by Niwa et al. (15).One milliliter of 1 mM 2,3-diaminonaphthalene was added to theresultant supernatant and the mixture was incubated overnight.Ethanol was completely evaporated, and the dried contents wereMATERIALS AND METHODSreconstituted in deionised water and extracted with ethyl acetate.

Materials. U-12C- and U-13C-labeled 3-deoxyglucosone standards The samples were then derivatized with TMS and prepared for GC/were synthesized as described previously (18). 2-Deoxyglucose, 2- MS analysis as described above.deoxyribose, 2-deoxygalactose, 3-deoxyglucose, and 2,3-diamino-

GC/MS analyses. The derivatized sample was analyzed using anaphthalene were obtained from Sigma (St. Louis, MO), Tri-SilHP 5890 GC equipped with a HP 5971 mass selective detector and(TMS) reagent from Pierce Co. (Rockford, IL), and ethanol fromHP 7673 automatic sampler. A 2-ml sample was injected and the GCQuantum Chemical Co. (Tuscola, IL).separation was performed on a fused silica capillary column (DB-5,

Quantitation of 3-[U-13C]deoxyglucosone. 13C NMR spectra of a 25 m 1 0.25 mm i.d.). The GC/MS temperature program was assolution of 3-[U-13C]DG were acquired along with several deoxy follows: injector port at 2507C and initial column temperature atsugars (2-deoxyglucose, 2-deoxyribose, and 2-deoxygalactose) of 1507C for 1 min, and then ramp to 2907C at 167C/min and held thereknown concentrations. In order to obtain absolute quantitation this for 15 min. The transfer line was maintained at 2507C. Quantitationdata was collected with long pulse delays without nuclear Over- of 3DG was done by selected ion monitoring using the m/z fragmentshausser effect decoupling. Intensities of the methylene resonances of 295 and 306 for plasma 3DG and m/z fragments at 299 and 309in spectra from each deoxy sugar were measured. The concentration for the 3-[U-13C]DG internal standard.of 3-[U-13C]DG was calculated by comparing the intensities of themethylene resonances from the known deoxy sugar standards tothat of 3-[U-13C]DG. To confirm the quantitation of synthetic 3-[U- RESULTS13C]DG standard by 13C NMR, a totally independent GC/MS proce-dure was used in which known amounts of commercially available GC/MS assay for 3DG. To establish optimal condi-3-deoxyglucose and synthetic 3-[U-13C]DG were mixed and then tions for analysis of 3DG in plasma, synthetic U-12C-reduced with NaBH4 according to a procedure described previously and U-13C-labeled 3DG standards, derivatized as de-(14). The silyl ethers of the 3-deoxyhexitols obtained after reduction

scribed under Materials and Methods, were appliedwere monitored by GC/MS. Comparison of the abundance of ionsat m/z 231 (3-deoxyhexitol) and m/z 235 (3-[U-13C]deoxyhexitol) was to the GC/MS. Figures 1a and 1b show the total ion

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca

256 LAL ET AL.

TABLE I

Characteristic of the Type I Diabetic Population

Total M F

Mean (SD) age 35 (6.2) 34 (3.3) 37 (7.7)Mean (SD) age onset 8 (3.9) 5 (3.6) 10 (2.5)Mean (SD) duration 27 (7.1) 28 (5.3) 26 (8.5)Mean (SD) HGBA1C 10.02 (1.21) 9.63 (0.83) 10.34 (1.41)Mean (SD) glucose (mg/dl) 221 (109) 187 (88) 248 (121)Mean body weight (kg) 66.1 (12.0) 73.5 (12.9) 60.2 (7.5)Sex (%) 100 44 56Nephropathy (% overt) 28 25 30CVD (%) — — —Retinopathy (% proliferative) 39 13 60PVD (%) — — —Neuropathy (%) 17 13 20

Note. CVD, cardiovascular disease; PVD, peripheral vascular disease.

chromatogram and the corresponding mass spectra in the plasma samples when compared to saline wasgreater than 99%.obtained from these samples. The proposed fragmen-

tation pattern of the trimethyl silyl derivative of the Concentration of 3DG in normals and diabetics. As3DG–DAN (U-12C and U-13C) adduct is shown in Fig. shown in Figure 6, using the assay described above,2. The total ion chromatogram of a plasma sample the mean concentration of free 3DG in individuals withderivatized according to this protocol is shown in Fig. type I diabetes was 98.5 { 34 nM (n Å 16) and in nor-3a. A peak eluting at the same retention time as syn- moglycemics was 58.5 { 14 nM (n Å 17). The differencethetic 3DG standard had a mass spectrum consisting in these two populations was statistically significantof the dominant fragments unique to the synthetic at P õ 1003 using a two-tailed Student’s t test. If one3DG (Fig. 3b). The identification of this peak as 3DG outlier is removed from the diabetic population, thewas further confirmed by change of elution conditions. mean of the diabetic population reduces to 92 { 17 nMThe fragments at m/z of 295 and 306 were used for and the P value using a two-tailed t test decreases toselected ion monitoring of plasma 3DG. Figure 4 õ1005. This outlier had a kidney transplant, a gradeshows the selected ion chromatogram (SIC) at m/z 295 four retinopathy, and definite neuropathy. The intraas-(a) and m/z 299 (b) from a plasma sample of a diabetic. say and the interassay coefficient of variation of analy-The peak at 15.6 min in the SIC at m/z 295 represents sis values for 3DG were 13 and 22%.the systemic plasma 3DG (Fig. 4a). This peak coelutes Effect of ethanol extraction on 3DG concentrations inwith the m/z 299 peak (Fig. 4b) derived from the 3- plasma. To resolve the discrepancy between results[U-13C]DG internal standard. The concentration of reported by Knecht et al. and Niwa et al., we comparedplasma 3DG was determined by measuring the rela- the levels of 3DG obtained after ultrafiltration oftive amounts of these ions. plasma with the 3DG levels obtained after ethanol

Recovery of 3DG in plasma. Unfiltered plasma and treatment. The levels of ultrafiltred 3DG were foundan equivalent volume of saline solution were spiked to be 58{ 6.1 nM (nÅ 8), similar to the values reportedwith varying amounts of synthetic [12C]3DG. These above. Ethanol treatment of the plasma resulted insamples were processed according to the procedure de- substantial increase in the levels of detectable 3DG toscribed under Materials and Methods for analysis of 1710 { 750 nM (n Å 8). While the range in the valuesplasma 3DG. As shown in Fig. 5, the recovery of 3DG of 3DG from ultrafiltred plasma was from 49.6 to 69.2

nM, the ethanol-treated plasma values ranged from 910to 2910 nM in the same individuals. No correlation was

TABLE II observed between the 3DG levels from the two depro-teinization methods.Characteristic of the Normoglycemic Population

Total M F DISCUSSION

Mean (SD) age 41 (8.4) 43 (8.7) 37 (6.9) 3DG is a reactive dicarbonyl which has been studiedSex (%) 100 59 41 extensively over last several decades by food chemists,Mean body weight (kgs) 72 (14) 81 (13) 60 (9) organic chemists and biochemists (21–32). Its reactiv-

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca


FIG. 1. GC/MS total ion chromatogram and mass spectrum of synthetic 3DG standards. (a) U 12C 3DG standard and (b) U 13C 3DGstandard.

ity by virtue of chemical interaction of its carbonyls histochemical methods (29–31). The potential of 3DGwith amino and thiol groups of proteins (21, 25, 27) and other aldosones as in vivo toxins has also beenhas been extensively demonstrated in in vitro systems. demonstrated in studies in which the proliferation ofThese results have been extended to in vivo systems cell lines is prevented by inhibition of protein and DNAby detection of pyrraline (29–31), a specific reaction synthesis (32–34).product of 3DG with proteins, by HPLC and immuno- Systemic 3DG can be produced in vivo by (i) decompo-

sition of the Amadori product on the glycated protein(26, 41) or (ii) nonenzymatic dephosphorylation of fruc-tose 3-phosphate (12, 13). Although, the relative contri-butions of these two routes in biosynthesis of systemic3DG clearly needs to be evaluated, the rapid turnoverof plasma proteins and the relatively long half-life ofthe Amadori product at physiological pH (56 days) (41),suggests that in vivo, the fructose 3-phosphate path-way may be an important source of 3DG.

As a natural consequence of its high reactivity, thereexist cellular mechanisms for detoxifying this metabo-lite to relatively inert species. Reductases have recentlybeen identified in several tissues in the body whichhave the ability to reduce 3DG on the C-1 to 3-deoxy-fructose (15, 35–37). Another potentially importantpathway for detoxification of 3DG is via oxidation to 2-FIG. 2. Proposed fragmentation pattern of trimethyl silyl deriva-

tive of the 3DG–DAN adduct. keto-3-deoxygluconic acid (3-DGA) by oxaldehyde dehy-

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca

258 LAL ET AL.

FIG. 3. GC/MS total ion chromatogram of plasma from a diabeticindividual. (a) Total ion chromatogram and (b) mass spectrum of theputative 3DG peak.

FIG. 4. Selected ion chromatogram of plasma with [U-13C]3DG in-drogenases (39–40). Both 3DF and 3-DGA have been ternal standard at (a) m/z 295 and (b) m/z 299.measured in human plasma and/or urine samples andfound to be elevated in diabetics (15, 38). A compromisein the detoxification ability of these enzymes may be compared to normoglycemics, in agreement with obser-one of the factors leading to accelerated development vations by Niwa et al. (16, 17).of complications in diabetics (6). The concentration of 3DG in ultrafiltered plasma

Given the potential role of 3DG in pathogenesis of from normoglycemic individuals reported in this studydiabetic complications, it is important that the mostaccurate and quantitative methods for its measure-ment be available. In this study, a previously describedmethod for quantitation of free plasma 3DG after deri-vatization with DAN (18) was adapted to GC/MS sothat a chemically identical internal standard in formof 3-[U-13C]DG could be used for absolute quantitation.Therefore, this procedure presents an advantage overthe previously reported HPLC method which uses 2,3-butadione as an internal standard, which may not ableto adequately compensate for reactivity of 3DG. More-over, using SIC–GC/MS, much higher sensitivity inthe assay of 3DG can be achieved. Extraction with ethylacetate selectively separates 3DG–DAN adduct fromother metabolites and sugars. Consequently, pretreat-ment of plasma samples is considerably minimized. Us-ing this method, a significant increase in the levels of FIG. 5. Recovery of 3DG in plasma from an unfasted volunteer and

saline spiked with synthetic 3DG.plasma 3DG was observed in type I diabetics when

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca


Complications (Rifkin, H., and Porte, D., Eds.), pp. 279–291,Elsevier, New York.

4. Makita, Z., Radoff, S., Rayfield, E. J., Yang, Z., Skolnik, E., Dela-ney, V., Friedman, E. A., and Cerami, A., (1991) N. Engl. J. Med.325, 836–842.

5. Beisswenger, P. J., Makita, Z., Curphey, T. J., Moore, L. L., Jean,S., Brinck-Johnsen, T., Bucala, R., and Vlasssara, H. (1995) Dia-betes 44, 824–829.

6. Brownlee, M. (1994) Diabetes 43, 836–841. [Lilly Lecture, 1993]7. Kato, H., Hayase, F., Shin, D. B., Oimomi, M., and Baba, S.

(1989) Prog. Clin. Biol. Res. 304, 69–84.8. Thornalley, P. J. (1990) Biochem. J. 269, 1–11.9. Phillips, S. A., and Thornalley, P. J. (1993) Eur. J. Biochem. 212,

101–105.10. Kappler, F., Szwergold, B. S., Lal, S., and Brown, T. R., unpub-

lished results.11. Szwergold, B. S., Kappler, F., and Brown, T. R. (1990) Science

FIG. 6. Scattergram illustrating 3DG levels in plasma from nor- 247, 451–454.moglycemic and diabetic individuals.

12. Lal, S., Szwergold, B. S., Taylor, A. H., Randall, W. C., Kappler,F., Knecht, K. W., Baynes, J. W., and Brown, T. R. (1995) Arch.Biochem. Biophys. 318, 191–199.

13. Kato, H., Shin, D. B., and Hayase, F. (1987) Agric. Biol. Chem.agrees very well with the value previously reported51, 2009–2011.

by Knecht et al. (58 nM in comparison to 61 nM) (15)14. Knecht, K. J., Feather, M. S., and Baynes, J. W. (1992) Arch.but less than 1/30th that measured by Niwa et al. (58 Biochem. Biophys. 294, 130–137.

nM vs 1800 nM) (16, 17). This disagreement is resolved 15. Niwa, T., Takeda, N., Yoshizumi, H., Tatematsu, A., Ohara, M.,when 3DG levels after ultrafiltration are compared Tomiyama, S., and Niimura, K. (1993) Biochem. Biophys. Res.

Commun. 196, 837–843.with those obtained after ethanol extraction. Conse-quently, the differences in 3DG levels observed in pre- 16. Niwa, T., Takeda, N., Miyazaki, T., Yoshizumi, H., Tatematsu,

A., Maeda, K., Ohara, M., Tomiyama, S., and Niimura, K. (1995)vious reports by Knecht et al. (14) and Niwa et al. (15)Nephron 69, 438–443.are likely to be due to the different deproteinization

17. Yamada, H., Miyata, S., Igaki, N., Yatabe, H., Miyauchi, Y.,methods used. 3DG values from ultrafiltred plasmaOhara, T., Sakai, M., Shoda, H., Oimomi, M., and Kasuga, M.presumably represent free amounts of this dicarbonyl. (1994) J. Biol. Chem. 269, 20275–20280.

On the other hand, the greatly increased 3DG ob- 18. Madson, A., and Feather, M. S. (1981) Carbohydr. Res. 94, 183–served upon ethanol extraction is likely to be due to 191.ethanol-mediated release of 3DG bound to the plasma 19. Orchard, T. J., Dorman, J. S., Maser, R. E., Becker, D. J., Drash,macromolecules, probably proteins. The consequence A. L., Ellis, D., LaPorte, R. E., Kuller, L. H., Wolfson, S. K., and

Drash, A. L. (1990) Diabetes 39, 1116–1124.and fate of bound 3DG is not clear at this point.20. Orchard, T. J., Dorman, J. S., Maser, R. E., Becker, D. J., Ellis,In this study, the elevated levels of plasma 3DG in

D., LaPorte, R. E., Kuller, L. H., Wolfson, S. K., and Drash, A. L.diabetics compared to normoglycemics support the hy-(1990) Diabetes Care 13, 741–747.pothesis that this dicarbonyl may be causally involved

21. Jellum, E. (1968) Acta Chem. Scand. 22, 1722–1728.in pathogenesis of diabetes complications. This study22. Oka, S. (1969) Agric. Biol. Chem. 33, 554–564.also resolves an important disagreement in the re-23. Rowell, R. M., and Green, J. (1970) Carbohydr. Res. 15, 197–ported values of plasma 3DG. Finally, the procedure

203.reported herein is reliable and sensitive, which allows

24. Shafizadeh, F., and Lai, Y. (1975) Carbohydr. Res. 40, 263–274.for absolute quantitation of 3DG.25. Shin, D. B., Hayase, F., and Kato, H. (1988) Agric. Biol. Chem.

52, 1451–1458.ACKNOWLEDGMENTS 26. Ledl, F., and Schleicher, E. (1990) Angew. Chem. Int. Ed. Engl.

29, 565–706.Support from Grants NEI 08414, NIDDK 44059, NIDDK 50364,27. Wedzicha, B. L., and Edwards, A. S. (1991) Food Chemistry 40,NIDDK 50317, and NIDDK 34818 (to T.J.O.) is gratefully acknowl-

71–86.edged.28. Hirsch, J., Mossine, V. V., and Feather, M. S. (1995) Carbohydr.

Res. 273, 171–177.REFERENCES

29. Miyata, S., and Monnier, V. (1992) J. Clin. Invest. 89, 1102–1. The Diabetes Control and Complications Trial Research Group 1112.

(1993) N. Engl. J. Med. 329, 977–986. 30. Portero-Otin, M., Nagaraj, R. H., and Monnier, V. M. (1995) Bio-chim. Biophys. Acta 1247, 74–80.2. Bucala, R., and Cerami, A. (1992) Adv. Pharmacol. 23, 1–34.

3. Brownlee, M.(1990) in Diabetes Mellitus: Advanced Products of 31. Nagaraj, R. H., Portero-Otin, M., and Monnier, V. M. (1996)Arch. Biochem. Biophys. 325, 152–158.Nonenzymatic Glycosylation and the Pathogenesis of Diabetic

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca

260 LAL ET AL.

32. Shinoda, T., Hayase, F., and Kato, H. (1994) Biosci. Biotech. 37. Liang, Z., Hayase, F., and Kato, H. (1992) Biosci. Biotech. Bio-chem. 56, 1074–1078.Biochem. 58, 2215–2219.

33. Reiffen, K. A., Loffler, M., and Schneider, F. (1980) Z. Natur- 38. Wells-Knecht, K., Lyons, T. J., McCance, D. R., Thorpe, S. R.,Feather, M. S., and Baynes, J. W. (1994) Diabetes 43, 1152–forsch 36, 255–261.1156.34. Reiffen, K. A., and Schneider, F. A. (1984) J. Cancer Clin. Oncol.

107, 206–210. 39. Fujii, E., Iwase, H., Ishii-Karakasa, I., Yajima, Y., and Hotta,K. (1995) Biochem. Biophys. Res. Comm. 210, 852–857.35. Takahashi, M., Fuji, J., Teshima, T., Suzuki, K., Shiba, T., and

Taniguchi, N. (1993) Gene 127, 249–253. 40. Fujii, E., Iwase, H., Ishii-Karakasa, I., Yajima, Y., and Hotta,K. (1994) J. Chromatogr. 660, 265–270.36. Sato, K., Inazu, A., Yamaguchi, S., Nakayama, T., Deyashiki,

Y., Sawada, H., and Hara, A. (1993) Arch. Biochem. Biophys. 41. Zyzak, D. V., Richardson, J. M., Thorpe, S. R., and Baynes, J. W.(1995) Arch. Biochem. Biophys. 316, 547–554.307, 286–294.

AID ABB 0117 / 6b37$$$461 05-09-97 22:30:43 arca

Documents

Quantitation of 3-Deoxyglucosone Levels in Human Plasma