Proton nuclear magnetic resonance spectral profiles of urine in type II diabetic patients

^a Address correspondence to this author at: Istituto di Chimica e Chimica Clinica, Università Cattolica del Sacro Cuore, Largo F.Vito 1–00168 Rome, Italy. Fax 039-6-30501918.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Serial urine samples of 33 type II diabetic patients and 20 control subjects were examined by ¹H nuclear magnetic resonance (NMR). Metabolites including lactate, citrate, glycine, alanine, hippurate, trimethylamine-N-oxide, and dimethylamine were identified in all subjects although in higher concentrations in diabetic patients. Other analytes, such as creatine, acetate, betaine, and ketone bodies, were found more frequently and in greater concentrations in diabetics than in controls. In addition, although lactate, citrate, alanine, and hippurate concentrations increased with increasing glycosuria and glycohemoglobin, trimethylamine-N-oxide and dimethylamine were present at high concentrations even in diabetics with good metabolic control. ¹H NMR spectroscopy permitted us to explore the relationships among the metabolites present in the urine samples and to obtain information about the disease status in type II diabetic patients.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
¹ H nuclear magnetic resonance (NMR)¹ is a powerful technique that allows study of biological fluids, providing an overview of important H-containing substances in samples. The main advantage of this technique is that it can furnish a rapid metabolic picture with little or no pretreatment of samples. NMR spectroscopy of urine samples might represent an eligible analytical method and a powerful diagnostic tool. Urine samples are more readily analyzed than blood because of their low protein content, and the registered signals may give important information about renal function in addition to the metabolic status of the subject.

In a previous report (1), we studied urine ¹ H NMR spectra of healthy subjects, using a standardized procedure, and we quantified the major metabolites that are consistently present. The aim of this study was to compare ¹ H NMR urine spectra of type II diabetic patients and nondiabetic subjects: (a) to define metabolic profiles characteristic of type II diabetes, (b) to obtain a reliable indication of the degree of metabolic control, and (c) to identify possible fingerprints of the biochemical changes that may accompany type II diabetes.

Furthermore, because the metabolite excretion pattern is also related to effective renal function, ¹ H NMR analysis of urine samples may be indicative of pathological processes that may occur in renal parenchyma as a consequence of the diabetic disease.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
chemicals and other materials
Deuterium oxide containing 7.5 mg/g of sodium 3-(trimethylsilyl)-[2,2,3,3,-H₄]-1-propionate (TSP) and all standard materials used to confirm peak assignments were from Sigma Chemical Co.

Wilmad 5-mm (Wilmad Glass Co.) NMR tubes (507-PP, ultra Imperial) were used to record the spectra.

¹ h nmr experiments
¹ H NMR spectra were registered at 25 °C on a Varian Gemini 300 apparatus (Varian) operating at 300 MHz. Urine (0.9 mL) was added to 0.10 mL of D₂O containing 7.5 g/kg TSP. The signal of TSP was used as the chemical shift reference (0.0 ppm).

Sample pH was adjusted to 5.80 ± 0.05 by adding the appropriate amount of concentrated HCl or NaOH. A standard single-pulse technique was used to collect the free induction decays. The following experimental conditions were applied: 40° pulse, acquisition time of 2 s, and pulse recycle time of 3 s to permit T₁ relaxation. The large H₂O signal was partially suppressed by the application of a gated secondary irradiation field (off during acquisition) at the H₂O resonance frequency. For each sample, 120 free induction decays were typically collected using a spectral width of 4000 Hz. No line-broadening was applied, and data were zero-filled once before Fourier transformation. Peaks were assigned considering the chemical shift relative to TSP and the spin-spin coupling patterns. Assignments were confirmed by adding standard materials to the samples.

Metabolite quantification was performed using peak heights corrected for the proton number contributing to the resonance and normalized with respect to the signal of creatinine. Thus, the obtained value for each metabolite was expressed as mmol/mol creatinine present in the sample.

patients
Fifty type II diabetic patients (28 men and 22 women; mean age, 65.3 ± 13 years) were recruited. In the morning, blood samples were collected: glucose (glucose oxidase colorimetric method), urea nitrogen (blood urea nitrogen, Berthelot modified method) and creatinine (alkaline picrate in kinetic) were determined on a Hitachi 717 device, glycohemoglobin (HbA_1c) was determined by HPLC (GlycolLab Analyzer Instrumentation Laboratory).

Within the same day, patient urine samples were collected in three time periods: 0800–1200, 1200–2000, 2000–0800.

The three daily samples were analyzed for qualitative evaluation of pH, hemoglobin (Hb), ketone bodies, and nitrite (Clinitek Atlas Analyzer-Ames), and for the quantitative determination of glucose and albumin (Behring nephelometric analyzer).

Samples were frozen at -80 °C for ¹ H NMR analysis, which was performed within 1 month.

All patients showed reference plasmatic values (mean ± SE) of blood urea nitrogen (8.4 ± 0.67 mmol/L) and creatinine (99.8 ± 18.5 µmol/L). Nonetheless, 17 of 50 patients who showed microalbuminuria values >20 mg/L were excluded from the study to eliminate subjects with a suspected renal involvement.

Twenty healthy subjects (mean age, 56 ± 7 years) were enrolled on the basis of the following criteria: no history or clinical evidence of renal disease, hypertension, diabetes mellitus, or other metabolic disorders. Their urine samples were collected and analyzed as reported for diabetics.

Statistics
Data are given as mean ± SE; differences between groups were calculated using ANOVA if the data were gaussian distributed or by the nonparametric Mann–Whitney test for unpaired values. P <0.05 was taken as significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
The degree of metabolic control of our patients was evaluated on the basis of HbA_1c and fasting glycemia values, which showed mean values of 7.5% ± 0.17%, and 10.2 ± 0.31 mmol/L, respectively. Moreover, we found glycosuria in 24 of 33 patients and ketonuria in 5 of 33 patients, and we observed poor metabolic control more frequently in women than in men.

Table 1 lists the chemical shift values of the chemical function used to identify the metabolites observed in ¹ H NMR spectra, using standardized conditions (1). We quantified the signal of interest, using creatinine peak heights as the internal reference, and calculated the relative metabolite concentration. Obtained results allowed us to evaluate (a) the metabolite concentrations in healthy and in diabetic subjects, (b) the intraday variability, and (c) the possible use of some of these metabolites as fingerprints of diabetic status.

Spectral profiles of healthy and diabetic subjects were partially similar (Fig. 1 ). However, although some metabolites were consistently present in both groups, their concentrations were significantly different. Mean and respective SE values are reported in Table 2 . These data were calculated using the mean daily excretion of each metabolite for all the studied subjects (n = 33). Hence, we have utilized data obtained from 99 urine samples. From Table 2 , it is clearly evident that all metabolites showed significantly increased concentrations in diabetics with respect to controls.

View larger version (18K): [in this window] [in a new window]   Figure 1. 1H NMR spectral profiles of a healthy and a diabetic subject.

To explain the increased metabolite excretion observed in diabetics, urinary concentrations were analyzed as a function of glycated Hb and glycosuria values (Tables 3 and 4). The behavior of each metabolite is described below.

We found that the concentrations of lactate, alanine, citrate, and hippurate in diabetics without glycosuria were similar to those of healthy subjects and that they increased dramatically with the increase of glucose waste. On the other hand, these metabolites were significantly increased in patients with reference values of HbA_1c (<6.4% of total Hb). This finding may be explained by the finding that 40% of these patients were also glycosuric.

Dimethylamine and trimethylamine-N-oxide concentrations were also increased in diabetics with good metabolic control, evaluated either by glycosuria or by HbA_1c concentrations.

Hippurate significantly increased in diabetic patients only in the presence of pathological glycated Hb concentrations. Finally, the concentration of glycine was significantly decreased in diabetic patients, and it increased significantly as glycosuria concentrations increased. We did not consider data of samples with glucose urine concentrations >5.5 mmol/L because of the possible overlapping of the glycine peak with part of the glucose signals.

Analysis of the three serial samples of each subject showed that, in almost every case, metabolite concentrations were quite constant during the day. However, hippurate showed a great intraday variability in diabetics, as already observed in controls (1), and lactate concentrations significantly increased (P = 0.01) in samples collected after feeding (morning samples, 85.6 ± 11.1 mmol/mol creatinine; afternoon samples, 120.1 ± 18.4 mmol/mol creatinine; night samples, 90.0 ± 12.8 mmol/mol creatinine).

In addition to all the biochemical components considered above, ¹ H NMR spectra of urine from diabetics indicated the presence of other metabolites, not usually determined by conventional urinalysis and not found in healthy subjects. In particular, at least one of the three ketone bodies was observed in greater percentage than in results using dipsticks (20 of 33 patients vs 5 of 33 patients). Moreover, 11 of these patients showed all three ketone bodies simultaneously. The respective mean ± SE values were as follows: acetone, n = 18, 14.3 ± 1.5 mmol/mol creatinine; acetoacetate, n = 12, 30.1 ± 2.7 mmol/mol creatinine; and 3-OH-butyrate, n = 13, 70.3 ± 8.9 mmol/mol creatinine. The relative proportions of ketone bodies found in urine (3-OH-butyrate, 61%; acetoacetate, 27%; and acetone, 2%) reflect those present in plasma (2).

Acetate and betaine, always present in small amounts in urine samples of healthy subjects (mean values, 8.3 ± 1.3 and 14.1 ± 0.9 mmol/mol creatinine, respectively), were significantly higher in diabetics (acetate, 161 ± 32.6 mmol/mol creatinine; betaine, 53.4 ± 7.1 mmol/mol creatinine in patients without glycosuria and 68.4 ± 6.1 mmol/mol creatinine in patients with glycosuria <5.5 mmol/L). Statistical analysis of betaine excluded samples with glucose concentration >5.5 mmol/L because of the possible overlapping of the resonance of interest with glucose peaks.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
This study of urine ¹ H NMR profiles of type II diabetic patients has demonstrated increased excretion of the metabolites consistently observed in healthy subjects and has pointed out the appearance of compounds usually not quantifiable in healthy subjects.

The variations of metabolite concentrations, evaluated as a function of glycosuria and glycated Hb, permitted us to obtain some information about the mechanisms involved in metabolite excretion. In fact, systemic cellular abnormalities occurring in diabetes can lead to pathological concentrations of several metabolites in plasma and, consequently, to their urinary overflow. Moreover, urine composition in diabetic disease may also be the result of target effects of glucose on renal cells, effects evidenced by impaired solute re-absorption from the tubular lumen and reduced efficiency of the epithelium (3)(4)(5).

In particular, the higher citrate excretion might depend not on the increased plasmatic concentrations but on increased citrate production in tubular cells and/or from reduced citrate re-absorption from the tubular fluid because of glucose overflow. In this respect, preliminary experiments on the plasma of diabetics did not show any difference with respect to controls (data not presented), suggesting an independence of citrate excretion from plasma concentration, according to data in the literature (6).

Moreover, although the urinary citrate concentration was found dramatically higher in diabetics with glycosuria, we did not observe a positive correlation between them, probably because glycosuria is affected by such variables as glomerular filtration rate, tubular function, and the presence of cells and bacteria.

Glucose overflow, which stimulates some glycolytic enzymatic activities in tubular cells (7)(8), may also account for peak resonances relative to succinate and pyruvate observed in some patients.

Lactate was always found at higher concentrations in diabetics compared with healthy subjects, as a consequence of high lactate hematic concentrations. In this respect, it should be pointed out that lactate is the only metabolite increased in samples collected after feeding, probably because of the impaired metabolic response to food intake in diabetic patients. Moreover, to explain the high lactate concentrations in the urine of diabetics with intense glycosuria, we cannot exclude a role of glucose on lactate tubular transport mechanisms. Although our patients were negative for nitrite, it must be regarded that lactate may also be linked to the presence of bacteria (9).

The increase of alanine in diabetic patients may be associated with a reduced insulin-mediated suppression of plasma amino acid concentrations. Because we have observed that alanine concentrations are linked to metabolic status, we suggest that this component may provide information about the relative unresponsiveness of amino acid metabolism to insulin (10). Moreover, increased alanine concentrations may also be indicative of an initial tubular damage (11). The simultaneous determination of the three ketone bodies is essential to evaluate their specific relationships. However, it is without clinical significance to consider acetoacetate and acetone separately, because the former breaks down spontaneously to the latter (12). Conventional urinalysis inevitably underestimates ketone bodies because dipsticks do not determine 3-OH-butyrate and underestimate acetoacetate for the reasons reported above (2).

More information about real metabolic status was obtained by analyzing the concentration of acetate, which appeared similar to that of healthy subjects in patients with good metabolic control, and it was significantly increased when metabolic conditions were altered. Although in the present study we did not consider subjects with pathological plasmatic and urinary values of renal function, we cannot exclude that acetate can also be the expression of an early tubular involvement, because it was found increased in subjects with overt renal insufficiency (13).

Trimethylamine-N-oxide and dimethylamine, synthesized in the medullar cells of the kidney, are osmolyte molecules that counteract protein perturbations produced by a high urea concentration (14)(15). In diabetes, the over-excretion of these metabolites may be linked to the hyperosmotic effect of glucose, or it may be indicative of a papillary dysfunction (11)(13).

Another osmolyte with intriguing importance observed in diabetic spectral profiles is betaine. The concentrations that we measured by ¹ H NMR in diabetics were similar to those measured by HPLC and reported by other authors (16), and they appeared significantly increased (P <0.001) compared with controls.

Lever et al. (16) hypothesized that increases of betaine in diabetes may be ascribed to three possible mechanisms: (a) an excessive production of sorbitol, another renal osmolyte (17); (b) the glycation of the betaine transport system; and (c) a secondary response to a tubular dysfunction.

Our data cannot exclude any of these hypotheses for the following reasons: The first one cannot be excluded because sorbitol resonances partially overlap with those of glucose; the last two hypotheses cannot be eliminated because we have not found a relation between betaine excretion, HbA_1c concentrations, and routine markers of tubular dysfunction. Thus, according to Lever et al. (16), we may conclude that high concentrations of betaine in the urine of diabetic patients may indicate some form of abnormal renal function not yet evinced by microalbuminuria and/or by tubular function markers. Furthermore, betaine excretion may be the response of the inner medulla tissue to diuretic state and to urinary osmolality (17)(18)(19). In addition, it must be recalled that betaine may be diagnostically important not only as a sign of tubular distorsion but also for its protective effect on bacteria growth in hyperosmotic urine (17).

In conclusion, the main advantages of NMR application to the study of disease status in diabetic patients may be summarized as follows: (a) a nonselective picture of many urinary metabolites is furnished; (b) ketone bodies and their relationships may be studied; (c) the abnormal presence of amino acids, useful in diabetics as test for insulin responsiveness, can be shown; and (d) metabolites such as citrate, acetate, and osmolytes may provide information about tubular and medullar function. The last point may be very important in the clinical evaluation of diabetic nephropathy because tubular damage appears to precede glomerular proteinuria, suggesting a tubular role in the initiation of diabetic renal disease (20)(21)(22).

	Footnotes

 
Istituto di Chimica e Chimica Clinica e Centro di Studio per la Chimica dei Recettori e delle Molecole Biologicamente Attive, CNR, Università Cattolica del Sacro Cuore, 00168 Rome, Italy.

¹ Nonstandard abbreviations: NMR, nuclear magnetic resonance; TSP, sodium 3-(trimethylsilyl)-[2,2,3,3,-²H₄]-1-propionate; and Hb, hemoglobin.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Zuppi C, Messana I, Forni F, Rossi C, Pennacchietti L, Ferrari F, et al. ¹H NMR spectra of normal urines: reference ranges of the major metabolites. Clin Chim Acta 1997;265:85-97. [Web of Science][Medline] [Order article via Infotrieve]
2. Sacks DB. Carbohydrates. Burtis CA Ashwood ER eds. Tietz textbook of clinical chemistry 2nd ed. 1994:928-1001 WB Saunders Philadelphia. .
3. Körner A, Eklöf AC, Celsi G, Aperia A. Increased renal metabolism in diabetes. Mechanism and functional implications. Diabetes 1994;43:629-633. [Abstract]
4. Christiansen SJ, Gammelgaard J, Frandsen M, Parving HH. Increased kidney size, glomerular filtration rate and renal plasma flow in short-term insulin-dependent diabetes. Diabetologia 1981;20:451-456. [Web of Science][Medline] [Order article via Infotrieve]
5. Pugliese G, Tilton RG, Williamson JR. Glucose-induced metabolic imbalances in the pathogenesis of diabetic vascular disease. Diabetes 1991;7:35-59.
6. Nicholson JK, Timbrell JA, Sadler PJ. Proton NMR spectra as indicators of renal damage. Mercury induced nephrotoxicity in rats. Mol Pharmacol 1985;27:644-651. [Abstract]
7. Belfiore F, Iannello S, Campione R, Volpicelli G. Le basi metaboliche delle complicanze diabetiche. G Ital Diabetol 1992;12:31-48.
8. Belfiore F, Iannello S, Campione R, Volpicelli G, Caraffa F. Metabolic effects of hyperglycemia in various tissues. Front Diabetes 1990;10:124-133.
9. Foxall PJD, Mellotte GJ, Bending MR, Lindon JC, Nicholson JK. NMR spectroscopy as a novel approach to the monitoring of renal transplant function. Kidney Int 1993;43:234-245. [Web of Science][Medline] [Order article via Infotrieve]
10. Tessari P. Effects of insulin on whole-body and regional amino acid metabolism. Diabetes Metab Rev 1994;10:253-285. [Web of Science][Medline] [Order article via Infotrieve]
11. Lundina TA, Knubovets TL, Sedov KR, Markova SA, Sibeldin LA. Variability of kidney tubular interstitial distorsions in glomerulonephritis as measured by ¹H NMR urinalysis. Clin Chim Acta 1993;214:165-173. [Web of Science][Medline] [Order article via Infotrieve]
12. Nicholson JK, O'Flynn MP, Sadler PJ, Macleod AF, Juul SM, Sonksen PH. Proton-nuclear-magnetic-resonance of serum, plasma and urine from fasting normal and diabetic subjects. Biochem J 1984;217:365-375. [Web of Science][Medline] [Order article via Infotrieve]
13. Gartland KPR, Bonner FW, Nicholson JK. Investigation into the biochemical effects of region specific nephrotoxins. Mol Pharmacol 1989;35:242-250. [Abstract]
14. Balaban RS, Burg MN. Osmotically active organic solutes in the renal inner medulla. Kidney Int 1987;31:562-564. [Web of Science][Medline] [Order article via Infotrieve]
15. Holmes H, Bonner FW, Sweatman BC, Lindon JC, Beddel CR, Rahr E, et al. Nuclear magnetic resonance spectroscopy and pattern recognition analysis of the biochemical processes associated with the progression of and recovery from nephrotoxic lesions in the rat induced by mercury(II) chloride and 2-bromoethanolamine. Mol Pharmacol 1992;42:922-930. [Abstract]
16. Lever M, Sizeland PCB, Bason LM, Hayman CM, Robson RA, Chambers ST. Abnormal glycine betaine content of the blood and urine of diabetic and renal patients. Clin Chim Acta 1994;230:69-79. [Web of Science][Medline] [Order article via Infotrieve]
17. Garcia-Perez A, Burg MB. Renal medullary organic osmolytes. Physiol Rev 1991;71:1081-1115. [Abstract/Free Full Text]
18. Nakanishi T, Uyama O, Nakahama H, Takamitsu Y, Sgita M. Determinants of relative amounts of medullary organic osmolytes: effects of NaCl and urea differ. Am J Physiol 1993;264:F472-F479. [Abstract/Free Full Text]
19. Moeckel G, Dasser HG, Chen TJ, Schmolke M, Guder WG. Bicarbonate-dependent betaine synthesis in rat kidney. Contrib Nephrol 1994;110:46-53. [Medline] [Order article via Infotrieve]
20. Yaqoob M, Patrick AW, McClelland P, Stevenson A, Mason H, Whites MC, et al. Relationship between markers of endothelial dysfunction, oxidant injury and tubular damage in patients with insulin-dependent diabetes mellitus. Clin Sci 1993;85:557-562. [Medline] [Order article via Infotrieve]
21. Yaqoob M, McClelland P, Patrick AW, Stevenson A, Mason H, Bell GM. Tubular damage in microalbuminuric patients with primary glomerulonephritis and diabetic nephropathy. Am J Nephrol 1990;10:475-481. [Web of Science][Medline] [Order article via Infotrieve]
22. Mogensen CE, Christensen CK, Vittinghus E. The stages in diabetic renal disease with emphasis on the stage of incipient diabetic nephropathy. Diabetes 1983;32(Suppl 2):64-78.