| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Instituto de Investigaciones Científicas, Universidad de Guanajuato, 36000 Guanajuato, Mexico.
2
Instituto de Investigaciones Médicas,
Universidad de Guanajuato, 37320 Leon, Mexico.
a Author for correspondence. Fax 473–26252; e-mail katarzyn{at}quijote.ugto.mx
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
= 280 nm) and spectrofluorometric for AGEs (ex =
247 nm, em = 440 nm). Sample pretreatment was carried
out in microcentrifuge tubes: Serum (20 µL) was deproteinized with
trichloroacetic acid (480 µL, 0.15 mol/L) and lipids were extracted
with chloroform (100 µL). Twenty microliters of the filtered aqueous
layer was injected to the flow system and the relation between
fluorescence and absorption signals was measured. A peptide-derived AGE
calibrator was used for calibration. Within-day and between-day CVs
were 6.7% and 9.1%, respectively, at an AGE concentration
corresponding approximately to that in healthy individuals. Mean
results (±SD) in 10 healthy individuals were 10.1% ± 1.0%, in 21
patients with diabetes without complications 18.0% ± 6.2%, in 25
patients with complications 24.1% ± 15.4%, and in 12 diabetic
patients in end-stage renal disease 92% ± 30%. Comparison with an
ELISA procedure (x, in arbitrary units/L) yields a
regression equation y = 0.713x + 1.24
(Sy||x = 6777, r
= 0.8477, n = 41). |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
AGE–proteins can be determined by immunochemical methods (6)(12)(13)(14). For such assays, AGE-specific antibody is required, as first described by Makita et al. (6) and used in later studies (13)(14). However, no commercial kit is available at this time. Serum and tissue AGEs have also been measured by radioreceptor assay (9)(15), for which cell cultures are needed; the procedure is labor and time consuming.
AGEs exhibit characteristic fluorescence (6). With excitation in the range 350–390 nm, fluorescence emission has been measured at 440–470 nm for AGE detection (8)(9)(16)(17)(18). In one study (9), the protein concentration in all samples was adjusted to 1 g/L and fluorescence measurements were expressed as the percentage of relative fluorescence as compared with that of an AGE–albumin calibrator. The results obtained did not agree well with the results of radioreceptor assay (9).
In the present work a simple analytical procedure is proposed for AGE measurement. The experimental conditions are discussed, results of 68 serum samples from healthy persons and diabetic patients are presented, and statistical correlation between those results and the results obtained by ELISA is evaluated.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
A Spectronic 3000 Diode Array Milton Roy spectrophotometer was used for absorbance measurements of bovine serum albumin (BSA) and AGE-BSA calibrators.
reagents
BSA, glucose, proteinase K, trichloroacetic acid (TCA),
chloroform, phenylmethylsulfonyl fluoride (PMSF), and Bradford reagent
were obtained from Sigma. Suprapur perchloric acid was from Merck.
Phosphate buffer solution (0.2 mol/L, pH 7.4) was prepared from respective sodium salts (Sigma).
AGE antibody was a kind gift from R. Bucala (The Rockefeller University, New York, NY). Stock enzyme solution (8 g/L) was prepared in Tris buffer (0.05 mol/L, pH 8) with 0.2 g/L sodium azide (Sigma) and stored at -20 °C for up to 2 weeks. Working solutions were obtained by diluting this stock solution (1:20) in 0.02 mol/L phosphate buffer.
Stock solution of 0.1 mol/L PMSF in ethanol was prepared daily. For proteinase K inhibition this solution was diluted 1:200 in phosphate buffer.
AGE–albumin was prepared by incubating albumin (50 g/L) with 0.5 mol/L glucose in 0.2 mol/L phosphate buffer (pH 7.4) for 60 days. The control samples of albumin without glucose were also incubated in these same conditions (37 °C, sterile conditions, darkness). After incubation, dialysis against PBS was carried out to remove unbound material.
AGE–peptide calibrator was obtained by hydrolysis of calibrator AGE-BSA (50 g/L) with proteinase K: 10 µL of proteinase K solution (8 g/L) was mixed with 90 µL of AGE-BSA and incubated for 24 h at 37 °C. In parallel, BSA was incubated with proteinase K to obtain a peptide calibrator. The reproducibility of AGE–peptide preparation was satisfactory: For three hydrolyzed AGE-BSA calibrators, prepared with the same proteinase K, the difference in AGE content was not greater than 4% as measured by our method.
clinical samples
Serum samples from healthy individuals and 58 diabetic patients
were analyzed. One control group of 10 healthy persons (1) and
three experimental groups were defined: (2) 21 diabetic patients
without chronic diabetes-relevant complications, (3) 25 diabetic
patients with chronic complications (neuropathy, nephropathy,
retinopathy, etc.), and (4) 12 diabetic patients with end-stage renal
disease.
on-line double- (spectrophotometric and
spectrofluorometric) detection flow system
Serum samples (20 µL) were mixed with 480 µL of TCA (0.15
mol/L) in microcentrifuge tubes, and 100 µL of chloroform was added.
The tubes were shaken vigorously to complete the precipitation of
proteins and to extract lipids to organic phase and then centrifuged
(10 min, 13 000g), and 20 µL of the aqueous layer was
injected into the flow system. Water flow rate was 0.5 mL/min
(deionized water, HPLC grade, Labconco); spectrophotometric detector
was set at 280 nm for detection of peptides (reference wavelength 550
nm). By using spectrofluorometric detector, AGE fluorescent signal was
obtained at 440 nm with excitation at 247 nm (bandwidth 17 nm, cutoff
filter 370 nm). The samples were run in triplicate and peak height mode
was used for signal measurements.
immunochemical assay
Serum samples (200 µL) were digested with proteinase K. To do
so, samples (100 µL) were placed in the microcentrifuge tubes and 200
µL of the working proteinase K solution was added. After incubation
(37 °C, 12 h) samples were centrifuged (10 min,
13 000g) and 200 µL of clear solution was mixed with 50
µL of PMSF solution (diluted 1:200). Volumes of 50 µL of such
prepared samples were taken to competitive ELISA (6).
Proteinase K could contain some AGEs, so the enzyme control sample
(adding 100 µL of phosphate buffer instead of serum) was prepared and
run in parallel.
The method of Bradford (19) was used for quantification of proteins (BSA and AGE-BSA after dialysis).
statistical analysis
The detection limit for AGEs in the proposed method was evaluated
as follows (20):
 |
 |
Within-day precision was evaluated as CV for 20 measurements of one sample, while measuring 10 other samples in between.
Between-day precision was evaluated as CV for three measurements of one sample during 20 days.
Differences between AGE results obtained in the control group and each of three groups of diabetic patients were studied with the t-test for unequal variances (Welch test) (21). Significance was accepted with P <0.05.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
em = 440 nm) of albumin (BSA) and AGE-BSA was seen
with excitation at 247 nm (Fig. 1
a). Emission fluorescence spectra (ex = 247 nm,
Fig. 1b
) showed maximal spectral shift at 440 nm.
|
Previous work (9) reported that AGEs in the low-molecular-mass serum fraction showed more marked differences between healthy subjects and diabetic patients than did AGEs in total serum.
In Fig. 2
, fluorescence excitation and emission spectra of hydrolyzed BSA
and peptide-derived AGE calibrator (obtained by degradation of AGE-BSA
with proteinase K) are presented. Comparison of Fig. 1a
and Fig. 2a
reveals that fluorescence peaks for BSA and AGE-BSA with excitation at
228 nm (Fig. 1a
) disappeared in excitation spectra of hydrolyzed
calibrators (Fig. 2a
), confirming that this peak corresponded to
peptide bonds. As can be observed in Figs. 1b
and 2b
, emission spectra
of BSA and AGE-BSA were also changed after degradation of proteins: The
difference between signals for hydrolyzed calibrators (Fig. 2b
) was
larger than for BSA and AGE-BSA (Fig. 1b
). The obtained results
indicate that, after protein degradation, some interferences affecting
AGE fluorescence signal while using spectrofluorometric detection
conditions ex = 247 nm and em = 440
nm were reduced (compare Fig. 1b
and Fig. 2b
). We experimentally
confirmed that wavelengths corresponding to excitation and emission
spectral maxima of AGEs in calibrators were not affected in the pH
range 1 to 8.
|
The idea of this work was to use these conditions of
spectrofluorometric detection for determination of AGEs in serum
(ex = 247 nm, em = 440 nm). A
low-molecular-mass serum fraction was obtained by deproteinization of
serum with TCA. The advantage of selected fluorescence conditions was
the improved sensitivity for AGEs, which in turn enabled sample
dilution and better reproducibility of protein precipitation
(difference in fluorescence signal for 10 repeated precipitation
procedures did not exceed 2%).
Four pooled serum samples were prepared by mixing seven serum aliquots
(20 µL) from: (1) healthy persons, (2) diabetic patients without
chronic complications, (3) diabetic patients with chronic
complications (neuropathy, nephropathy, etc.), and (4) diabetic
patients with end-stage renal disease. These pools were treated with
0.15 mol/L TCA and, in Fig. 3
a and b, excitation and emission fluorescence spectra obtained
for centrifuged and diluted (pools 1–3 25 times, pool 4 75 times)
samples are presented. As can be observed in Fig. 3a
, the fluorescence
signals obtained in deproteinized serum pools (ex =
247 nm, em = 440 nm) corresponded to fluorescence signal
obtained for calibrators of albumin- and peptide-derived AGEs (Fig. 1a
, 2a
). Moreover, significant differences can be observed (Fig. 3a
)
between relative fluorescence signals (ex = 247 nm,
em = 440 nm) in four serum pools, in agreement with
expected content of AGEs: low in healthy persons and increasing in
diabetic patients depending on the possible complications. It seems
possible that fluorescence observed in the excitation wavelength range
260–300 nm was emitted by degradation products of serum proteins.
Indeed, the highest relative fluorescence in this range was observed
for pooled serum from end-stage renal disease patients, whose serum
contained peptides accumulated because of impaired renal function (Fig. 3a
). Fluorescence emission spectra (ex = 247 nm) of the
four pooled serum samples (1–4) are presented in Fig. 3b
. The
progressive shift of these spectra (2–4) towards the position of
maximum on the peptide-derived AGE spectra (max = 420
nm, Fig. 2b
) can be observed with the expected increase of AGE content
in serum of diabetic patients with respect to that of healthy persons.
These results confirmed that after serum deproteinization, and using
247 nm for excitation (em = 440 nm), the AGE
fluorescence signal can be measured in serum. Furthermore, in such
experimental conditions, AGE content in four pools prepared from serum
of different groups of patients was clearly distinguished.
|
In the previous reports, AGE serum content was measured as a fraction
of total serum proteins (6)(11) or a fraction
of serum peptides (9). Calibration was carried out with a
single calibrator of AGE-BSA (1 mmol/L AGE-BSA = 12
A350nm) (6)(9). In the
present work, two detectors were connected on-line in a flow system to
measure simultaneously signals corresponding to AGEs and to peptides in
the sample. Spectrophotometric detection was used for total peptides
and spectrofluorometric detection for AGEs. In Fig. 4
the example of measurement record for eight samples
injected in triplicate is presented. The value of the ratio between
relative fluorescence (FLDex247nm,em440nm) and
absorbance (A280nm) was taken as the analytical
signal for AGEs (S):
 |
|
The analytical signals for hydrolized calibrators of AGE-BSA
(SAGE-BSA) and BSA (SBSA)
were measured at physiological pH (pH 7.4). The ratio between these
signals was significantly lower than the ratio between specific AGE
signals measured as absorbance at 350 nm
(6)(9)
(SAGE-BSA/SBSA = 4.8 and
AAGE-BSA/ABSA = 12.3).
This indicates that in such conditions the proposed procedure did not
assure measurement of signal specific for AGEs. In further development
it was observed that, in acid medium, relative fluorescence
(FLDex247nm,em440nm) was suppressed, but the ratio
SAGE-BSA/SBSA increased.
This suggests that fluorescence not specific for AGEs was quenched in
acid medium. Then, effect of TCA and perchloric acid, commonly used for
deproteinization of serum, on the relation between AGE signals in the
two hydrolized calibrators was studied.To do so, 20 µL of each
calibrator was treated with different concentrations of TCA or
perchloric acid (480 µL) and we determined that, after addition of
acid, peptides did not precipitate (absorbance of peptide-derived AGE
calibrator at 280 nm not altered after acid treatment). In the presence
of TCA, relative fluorescence (FLDex247nm,em440nm) was
always lower than in the presence of perchloric acid. For TCA
concentrations up to 0.15 mol/L and for perchloric acid concentrations
up to 0.4 mol/L the ratio
SAGE-BSA/SBSA increased
with increasing acid concentration (Fig. 5
) up to
SAGE-BSA/SBSA = 12.3,
which corresponded exactly to the relation obtained between absorbances
of AGE-BSA and BSA measured at 350 nm. These results clearly indicate
that analytical signal (S) obtained with calibrators in the
presence of 0.15 mol/L TCA or perchloric acid >0.4 mol/L was specific
for AGEs, as compared with earlier reports
(6)(9). A very similar effect of the two acids
was observed while deproteinizing four serum pools (1–4). The most
marked differences between AGE signals in diabetic (2–4) and control
(1) serum pools were obtained with 0.15 mol/L TCA and 0.50 mol/L
perchloric acid, and these concentrations were selected for further
studies. It should be mentioned that a significant fluorescence blank
was observed for perchloric acid.
|
As no AGE standard is available, external calibration was carried out
with an AGE–peptide calibrator derived from an AGE-BSA calibrator (50
g/L) by hydrolysis with proteinase K. This calibrator was diluted (five
times) to obtain the same peptide content as in deproteinized serum (10
g/L) and prepared for analysis in the same way as serum (see
Materials and Methods: treatment with acid, extraction,
etc.). The obtained sample corresponded to a calibrator 1 mmol/L
AGE-BSA = 12 A350nm, used in the previous
reports (6)(9). For this sample, analytical
signal S was measured as described above, and taken as 100%
of AGE-derived peptides (saturation of BSA with AGEs before hydrolysis
with proteinase K). This solution was used as a "stock"
AGE–peptide calibrator, and multipoint calibration was carried out in
the proposed measurement system. To do so, a series of calibration
solutions was prepared by appropriate dilution of stock AGE–peptide
calibrator. These calibration samples contained different AGE
concentrations and constant peptide concentration, achieved by addition
of hydrolyzed BSA (in each sample the final peptide content was 0.4
g/L). In Fig. 6
the measurement records obtained for calibration samples (10%
to 100% AGEs) with the two detectors are presented. It should be
mentioned that by using the proposed procedure it was impossible to
prepare a blank solution (0% AGEs) because BSA contains some natural
AGEs. AGEs in the hydrolyzed BSA calibrator were determined as
described previously (1 mmol/L BSA = 12
A350nm) (6)(9), and the
evaluated content corresponded to 8.1% of AGEs in the scale proposed
here. In Table 1
, analytical characteristics obtained while treating samples
with TCA and perchloric acid are presented. As expected, better
sensitivity was obtained with perchloric acid (compare slope values of
linear regression) because of quenching of AGE fluorescence in the
presence of TCA. However, relative standard deviations for slope and
for intercept were much higher for perchloric acid. Also the detection
limit, within-day, and between-day precision were poorer for this acid,
which should be ascribed to a high fluorescence blank. Thus, for
analytical application 0.15 mol/L TCA was chosen.
|
|
The proposed procedure was applied to determination of AGEs in serum
from (1) healthy persons, (2) diabetic patients without
diabetes-relevant complications, (3) diabetic patients with chronic
complications (neuropathy, nephropathy, etc.), and (4) diabetic
patients with end-stage renal disease. The results obtained are given
in Table 2
. Variances of the results obtained in the four groups (1–4)
were calculated, and the Welch test (21) for unequal
variances was performed between results for the control group (1) and
each of three groups of diabetic patients (2–4). The obtained results
indicated that samples from groups (1) and (2–4) are statistically
different (P <0.05).
|
Finally, ELISA was carried out for AGE quantification in 41 serum
samples selected from the samples previously analyzed by the proposed
procedure to cover the observed range of AGE content. The procedure of
immunochemical assay described previously (6) was slightly
modified. Following the kind suggestions of R. Bucala, degradation of
serum proteins was carried out (proteinase K, protocol given in
Materials and Methods) to facilitate the access of specific
antibody to AGEs. In Fig. 7
, the relation between AGE content in serum samples obtained by
using spectrofluorometric–spectrophotometric procedures and ELISA is
plotted. Statistically significant correlation between the results
obtained with the two methods can be observed in this Figure
(regression equation y = 0.713x + 1.24,
Sy|x = 6777,
r = 0.8477, n = 41).
|
We conclude that the simplicity of the proposed method should enable its automation and use in routine clinical laboratories. Moreover, the small sample volume requirement (~1 µL of serum) may allow use of a filter-paper spot test in the future.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  S. S Lund, L. Tarnow, C. D A Stehouwer, C. G Schalkwijk, T. Teerlink, J. Gram, K. Winther, M. Frandsen, U. M Smidt, O. Pedersen, et al. Impact of metformin versus repaglinide on non-glycaemic cardiovascular risk markers related to inflammation and endothelial dysfunction in non-obese patients with type 2 diabetes Eur. J. Endocrinol., May 1, 2008; 158(5): 631 - 641. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Aragno, R. Mastrocola, G. Alloatti, I. Vercellinatto, P. Bardini, S. Geuna, M. G. Catalano, O. Danni, and G. Boccuzzi Oxidative Stress Triggers Cardiac Fibrosis in the Heart of Diabetic Rats Endocrinology, January 1, 2008; 149(1): 380 - 388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Persson, P. Rossing, P. Hovind, C. D.A. Stehouwer, C. Schalkwijk, L. Tarnow, and H.-H. Parving Irbesartan Treatment Reduces Biomarkers of Inflammatory Activity in Patients With Type 2 Diabetes and Microalbuminuria: An IRMA 2 Substudy Diabetes, December 1, 2006; 55(12): 3550 - 3555. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Roberts, M. C. Thomas, D. Fernando, N. Macmillan, D. A. Power, and F. L. Ierino Low molecular weight advanced glycation end products predict mortality in asymptomatic patients receiving chronic haemodialysis Nephrol. Dial. Transplant., June 1, 2006; 21(6): 1611 - 1617. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Stam, C. G. Schalkwijk, C. van Guldener, P. M. ter Wee, and C. D. A. Stehouwer Advanced glycation end-product peptides are associated with impaired renal function, but not with biochemical markers of endothelial dysfunction and inflammation, in non-diabetic individuals Nephrol. Dial. Transplant., March 1, 2006; 21(3): 677 - 682. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Forbes, S. R. Thorpe, V. Thallas-Bonke, J. Pete, M. C. Thomas, E. R. Deemer, S. Bassal, A. El-Osta, D. M. Long, S. Panagiotopoulos, et al. Modulation of Soluble Receptor for Advanced Glycation End Products by Angiotensin-Converting Enzyme-1 Inhibition in Diabetic Nephropathy J. Am. Soc. Nephrol., August 1, 2005; 16(8): 2363 - 2372. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lassila, K. K. Seah, T. J. Allen, V. Thallas, M. C. Thomas, R. Candido, W. C. Burns, J. M. Forbes, A. C. Calkin, M. E. Cooper, et al. Accelerated Nephropathy in Diabetic Apolipoprotein E-Knockout Mouse: Role of Advanced Glycation End Products J. Am. Soc. Nephrol., August 1, 2004; 15(8): 2125 - 2138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Forbes, L. T. L. Yee, V. Thallas, M. Lassila, R. Candido, K. A. Jandeleit-Dahm, M. C. Thomas, W. C. Burns, E. K. Deemer, S. R. Thorpe, et al. Advanced Glycation End Product Interventions Reduce Diabetes-Accelerated Atherosclerosis Diabetes, July 1, 2004; 53(7): 1813 - 1823. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Candido, J. M. Forbes, M. C. Thomas, V. Thallas, R. G. Dean, W. C. Burns, C. Tikellis, R. H. Ritchie, S. M. Twigg, M. E. Cooper, et al. A Breaker of Advanced Glycation End Products Attenuates Diabetes-Induced Myocardial Structural Changes Circ. Res., April 18, 2003; 92(7): 785 - 792. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Forbes, M. E. Cooper, V. Thallas, W. C. Burns, M. C. Thomas, G. C. Brammar, F. Lee, S. L. Grant, L. A. Burrell, G. Jerums, et al. Reduction of the Accumulation of Advanced Glycation End Products by ACE Inhibition in Experimental Diabetic Nephropathy Diabetes, November 1, 2002; 51(11): 3274 - 3282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chen, S. Brodsky, H. Li, D. J. Hampel, T. Miyata, T. Weinstein, U. Gafter, J. T. Norman, L. G. Fine, and M. S. Goligorsky Delayed branching of endothelial capillary-like cords in glycated collagen I is mediated by early induction of PAI-1 Am J Physiol Renal Physiol, July 1, 2001; 281(1): F71 - F80. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. S. Ooi and A. A. House Cardiac troponin T in hemodialyzed patients Clin. Chem., July 1, 1998; 44(7): 1410 - 1416. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, TCA; 
, perchloric acid.
