Istituto di Medicina Interna, Malattie Endocrine e Metaboliche, Università di Catania, Ospedale Garibaldi (L.F., R.B., B.C., S.G., D.S., R.V., V.T.), 95123 Catania; Consiglionazionale della Ricerca, Institute of Clinical Physiology and Department of Internal Medicine, University of Pisa (S.C., E.F.), 56100 Pisa; Divisione ed Unita di Ricerca di Endocrinologia, Istituto Scientifico Ospedale Casa Sollievo della Sofferenza, San Giovanni Rotondo (V.T.), 71013 Foggia; and Divisione di Endocrinologia 1, Università La Sapienza (M.D.), 00100 Rome, Italy; and the Diabetes Research Laboratory, Mount Zion Hospital, University of CaliforniaSan Francisco (M.D., B.A.M.), San Francisco, California 94143-1616
Address all correspondence and requests for reprints to: Lucia Frittitta, M.D., Endocrinologia, Ospedale Garibaldi, P.zza S. M. di Gesù, 95123 Catania, Italy. E-mail: segmeint{at}mbox.unict.it
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A soluble PC-1, with mol wt and enzymatic activity similar to those of tissue PC-1, was measurable in human plasma by a specific enzyme-linked immunosorbent assay and was inversely correlated to skeletal muscle PC-1 content (r = -0.5; P < 0.01). The plasma PC-1 concentration was decreased (P < 0.05) in insulin-resistant (22.7 ± 3.0 ng/mL; n = 25) compared to insulin-sensitive (36.7 ± 4.5; n = 25) nondiabetic subjects and was correlated negatively with the waist/hip ratio (r = -0.48; P < 0.001) and mean blood pressure (r = -0.3; P < 0.05) and positively with high density lipoprotein/total cholesterol (r = 0.38; P < 0.01) and both the M value and the plasma free fatty acid level decrement at clamp studies (r = 0.28; n = 50; P = 0.05 and r = 0.43; n = 22; P < 0.05, respectively). A plasma PC-1 concentration of 19 ng/mL or less identified a cluster of insulin resistance-related alterations with 75% accuracy.
In conclusion, PC-1 circulates in human plasma, and its concentration is related to insulin sensitivity. This may help to plan studies aimed at understanding the role of PC-1 in insulin resistance.
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The cellular mechanisms causing insulin resistance are not yet fully understood (3, 4, 5, 6, 7, 8, 9, 10). Recently, overexpression of several potential inhibitors of the insulin receptor tyrosine-kinase activity, a key step in insulin signaling, has been described in insulin-resistant subjects (11, 12, 13, 14). Although questioned by some researchers (15), these inhibitors include PC-1 (11), a class II (cytoplasmic amino-terminus) membrane glycoprotein with multiple enzymatic activities (pyrophosphatase and phosphodiesterase) that is expressed in many tissues and inhibits insulin signaling either at the level of the insulin receptor (11) or downstream at the level of the S-6 kinase activity (16). In both cell models and human tissues an elevated PC-1 content is associated with defective insulin receptor tyrosine kinase activity. In MCF-7 cells transfected with and overexpressing PC-1 there is a marked inhibition of insulin receptor tyrosine kinase activity and insulin action (11). In addition, PC-1 content is elevated in skeletal muscle, sc adipose tissue, and cultured skin fibroblasts of insulin-resistant nondiabetic subjects (17, 18, 19). In the same tissues, insulin receptor tyrosine kinase activity is impaired (17, 18, 19). These observations suggest that an elevated PC-1 content in insulin target tissues may play an important role in the development of insulin resistance. To test this hypothesis, large scale prospective studies are required. These studies, however, are difficult to perform because of the need for repeated tissue biopsies in a large number of subjects during a long observation period.
A soluble form of PC-1 has been detected in rat cell culture medium, rat serum (20), and human synovial fluid (21). Its biological significance is presently unknown. Whether a soluble form of PC-1 is present in human plasma and whether its concentration reflects insulin sensitivity are also unknown. If this is the case, however, soluble PC-1 measurement would considerably enhance the feasibility of large scale studies aimed at understanding the role of PC-1 in human insulin resistance and related abnormalities.
To address this issue, we set up a sensitive, specific enzyme-linked immunosorbent assay (ELISA) method to measure the PC-1 concentration in human plasma. We then studied 50 healthy subjects with a wide range of body mass indexes (BMIs) and insulin sensitivities.
![]() |
Experimental Subjects |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
![]() |
Materials and Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Insulin-stimulated glucose disposal measurement by the
euglycemic hyperinsulinemic clamp (22) was performed with a constant
insulin infusion (40 mU/m2·min) and a variable
glucose infusion to maintain plasma glucose within 10% of the baseline
value. The M value (milligrams per kg BW/min) ranged from 2.211.1
(mean ± SE, 5.37 ± 0.21; median, 5.1). The 50
healthy subjects were subdivided into insulin-sensitive (M values
>5.1, i.e. the median value) and insulin-resistant subjects
(M value <5.1; Table 1).
Plasma PC-1 measurement
Wells in Maxisorb plates (Nunc, Roskilde, Denmark) were
precoated (overnight incubation at 4 C) with an affinity-purified
polyclonal antibody to PC-1 (provided by Dr. Yamashima). After washing
with TBST buffer (20 mmol/L Tris, 150 mmol/L NaCl, and 0.05% Tween-20)
to remove unbound antibody, wells were blocked with 150 mL TBST
containing 1% BSA (30 min at 56 C) and washed again with TBST. Then,
human plasma (1030 µL diluted to a total volume of 100 µL with 50
mmol/L HEPES buffer, pH 7.6, containing 0.05% Tween-20, 1 mmol/L
phenylmethylsulfonylfluoride, 2 mmol/L orthovanadate, 1% BSA, and 1
mg/mL bacitracin) was added to each well, and PC-1 was allowed to bind
overnight at 4 C. After extensive washing with TBST, a biotinylated
anti-PC-1 monoclonal antibody was added. After 2 h at 22 C,
peroxidase-streptavidin was added, and after 30 min wells were washed
again with TBST, and 100 µL biotinyl-tyramide solution were added.
After 15-min incubation at 22 C, wells were washed with TBST, and
streptavidin-horseradish peroxidase solution was added (30 min at 22
C). After further extensive washing, the peroxidase activity was
determined colorimetrically by adding 3,3',5,5'-tetramethylbenzidine at
a concentration of 0.4 g/L in an organic base and measuring the
absorbance at 451 nm. In this assay, human plasma produced a dilution
slope that paralleled the PC-1 standard purified as previously
described (11) (Fig. 1). Intra- and
interassay coefficients of variation were less than 10%.
|
The phosphodiesterase activity of PC-1 was measured by the hydrolysis of thymidine 5'-monophosphate p-nitrophenyl ester (PNTP) (15). The reaction was carried out at saturating concentrations of the substrate by incubating 540 µL human serum with 1 µmol PNTP in 250 µL buffer (0.1 mol/L 2-amino-2-methyl-propanol and 7.5 mmol/L Mg(OAc)2, pH 9.4) for 60 min at 37 C. The reaction was stopped by the addition of 2 mL 0.1 N NaOH. Liberated p-nitrophenol was quantified spectrophotometrically by reading at 401 nm. Data were calculated as nanomoles of PNTP hydrolyzed in 1 h by 1 mL plasma.
Muscle PC-1 measurement
Muscle tissue specimens (external oblique) were obtained at
elective abdominal surgery (cholecystectomy) from 27 additional healthy
subjects. After adipose tissue was dissected, and blood was removed,
specimens were immediately frozen in liquid nitrogen. Soluble extracts
were subsequently prepared as previously described. Briefly, muscle
tissue (150 mg) was pulverized under liquid nitrogen and then
homogenized in 2 mL buffer (50 mmol/L HEPES, 150 mmol/L NaCl, and 2
mmol/L PMSF, pH 7.6) at 4 C using a Polytron homogenizer
(Brinkmann Instruments, Inc., Westbury, NY) for 10 s
at medium speed. Triton X-100 was added to a final concentration of
1%, the homogenates were solubilized for 60 min at 4 C and centrifuged
at 100,000 x g for 60 min at 4 C, and then
supernatants were used for measurement of the PC-1 content (as
described above), which was normalized for protein content (23).
Western blot studies
For Western blot analysis, human serum (2025 mL) was first
applied to a wheat-germ agglutinin column, and the glycoproteins were
eluted with acetylglucosamine 0.3 mol/L. Ten nanograms of PC-1
(measured by ELISA) were then immunoprecipitated from eluates with an
affinity-purified anti-PC-1 polyclonal antibody coated with protein
A-Sepharose for 16 h at 4 C. After centrifugation at 10,000
x g for 5 min, the pellet was washed three times with
buffer [0.5 mol/L NaCl, 10 mmol/L sodium phosphate (pH 7.4), 0.5%
Nonidet P-40, 2 mmol/L ethylenediamine tetraacetate, 0.1% SDS, and
0.04% BSA] and boiled in sample buffer [0.5 mol/L Tris-HCl (pH 6.8),
glycerol, 10% SDS, 0.05% bromophenol blue, and ß-mercaptoethanol].
The sample was than centrifuged at 10,000 x g for 5
min, and the supernatant was subjected to SDS-PAGE in 7.5%
polyacrylamide gel under reducing conditions. Proteins were transferred
(2 h at 4 C) to nitrocellulose membranes that were first blocked with
10% BSA, washed with TBST buffer (10 mmol/L Tris, 150 mmol/L NaCl, and
0.5% Tween-20), and then incubated with 1 µg/mL anti-PC-1 monoclonal
antibody. After 16 h at 4 C, membranes were incubated for 1 h
with rabbit antimouse serum conjugated with horseradish peroxidase, and
the reaction was developed with an enhanced chemiluminescence (ECL)
detection system (Amersham International, Aylesbury, UK).
The specific signaling was revealed by autoradiography.
Statistical analysis
Students t test and one-way ANOVA were used to compare mean values. Relationships between variables were evaluated by simple and multiple stepwise regression analysis. Data are presented as the mean ± SEM.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
When repeatedly measured every 30 min for 4 h in three subjects, the PC-1 plasma concentration did not change significantly, with observed variations being smaller than the intraassay coefficient of variation (data not shown). In the 50 subjects studied, the plasma PC-1 concentration ranged from 2.582.5 ng/mL (mean ± SEM, 29.7 ± 2.9 ng/mL; median, 26.0). No difference was observed between obese (BMI, >28 kg/m2) and nonobese (26.9 ± 4.4 and 31.0 ± 3.7 ng/mL, respectively) subjects.
Plasma PC-1 retained the pyrophosphatase activity typical of tissue
PC-1 (17, 20). Pyrophosphatase activity was significantly
(P < 0.01) higher in seven plasma specimens with high
PC-1 (56 ± 6.1 ng/mL; range, 33.572) than in seven plasma
specimens with low PC-1 (15 ± 3.1 ng/mL; range, 3.926.3):
28 ± 3 vs. 6 ± 1 nmol PNTP hydrolyzed/min·mL,
respectively. When we measured plasma and muscle PC-1 in a subset of 27
healthy subjects, a negative correlation (r = -0.5;
P < 0.01) between muscle (range, 4.582.5 ng/mg
protein) and plasma (range, 2.520 ng/mL) PC-1 was observed (Fig. 2).
|
When studied by PAGE and immunoblot, plasma PC-1 showed a similar
mol wt as skeletal muscle PC-1 and PC-1 purified from Chinese hamster
ovary (CHO) cells transfected with full-length human PC-1 complementary
DNA (i.e. the standard PC-1 used in our ELISA; Fig. 3).
|
In the 25 insulin-sensitive healthy subjects, the mean PC-1 plasma
level was significantly (P < 0.05) higher (36.7
± 4.5 ng/mL) than that in the 25 insulin-resistant healthy subjects
(22.7 ± 3.0 ng/mL; Fig. 4).
Moreover, when all 50 subjects were considered together, the plasma
PC-1 concentration was negatively correlated with both the waist/hip
ratio (r = -0.48; P < 0.001; Fig. 5A
) and mean blood pressure (r =
-0.3; P < 0.05; Fig. 5B
). In contrast, a positive
correlation was observed with the high density lipoprotein (HDL)/total
cholesterol ratio (r = 0.38; P < 0.01; Fig. 5C
)
and both the M value (r = 0.28; n = 50; P =
0.05) and the insulin-induced decrement in plasma FFA levels during the
clamp (r = 0.43; n = 22; P = 0.05; Fig. 6
, A and B). No significant correlation
was observed between plasma PC-1 and BMI (r = 0.1;
P = NS). When data were adjusted for BMI, sex, and age,
plasma PC-1 was still correlated with both the waist/hip ratio and the
decrement in plasma FFA (P < 0.05), but not with the
other parameters.
|
|
|
Plasma PC1 levels and the clustering of the insulin resistance syndrome
Looking for the cluster of insulin resistance syndrome, tertiles were obtained for the following parameters (plasma glucose at 0 and/or 120 min of OGTT, plasma insulin at 0 and/or 120 min of OGTT, serum triglycerides, HDL/total cholesterol ratio, and mean blood pressure). Clustering was defined as four or more of the mentioned parameters being in the worst tertile (the highest for glycemia, insulinemia, triglyceridemia, and mean blood pressure and the lowest for HDL/total cholesterol ratio). When the 50 healthy subjects were subdivided into tertiles according to their plasma PC-1 concentration (range, 2.517, 18.534, and 3582.5 ng/mL in tertiles 1, 2, and 3, respectively), the cluster of insulin resistance was more frequent in subjects from tertile 1 (7 of 17, 41%) than in tertiles 2 (1 of 16, 6.3%) and 3 (1 of 17, 5.9%; P < 0.05, by ANOVA). A plasma PC-1 level of 19 ng/mL or less identified this cluster of alterations typical of insulin resistance with 75% accuracy (75% sensitivity and 75% specificity).
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Previous studies have shown that in insulin-resistant subjects, the PC-1 content is increased in insulin target tissues (including muscle and adipose tissues) (17, 18, 19, 24) and that the PC-1 may act through inhibition of the insulin receptor tyrosine kinase activity. The reason why insulin-resistant individuals have an increased PC-1 content in tissues but a reduced PC-1 concentration in plasma is not known. One possibility is that the PC-1 content is increased in tissues because its release into extracellular fluids is impaired, as suggested by the negative correlation we observed between plasma and skeletal muscle PC-1 content. Another possibility is that in insulin-resistant subjects, circulating PC-1 is cleared at a higher rate from plasma (either bound or degraded).
The observation that the PC-1 plasma concentration is not affected by the acute hyperinsulinemia obtained in the glucose clamp study is in concert with the recent observation that the PC-1 tissue content is not affected by chronic hyperinsulinemia in patients with insulinoma (25) and suggests that hyperinsulinemia does not regulate PC-1 tissue content and/or release.
PC-1 released from rat cultured cells has a smaller mol wt than cellular PC-1, and it is likely to be cleaved intracellularly, with no evidence of plasma membrane-associated cleavage (20). In contrast, human plasma PC-1 has a mol wt similar to that of tissue PC-1. This finding is compatible with either the active secretion of intact PC-1 from human tissues or the binding into plasma of cleaved PC-1 to an unknown circulating factor(s).
Cell membrane PC-1 as well as other ecto-enzymes may hydrolyze nucleotides to nucleosides, which subsequently are taken up by cells through nucleoside transporters (26). This would act as a salvage pathway for those cells that have a low capacity to synthesize purine nucleotides by collecting the presynthesized nucleosides from the extracellular fluid. As human plasma PC-1 retains its phosphodiesterase activity, it may contribute to the metabolism of purine nucleotides.
Whatever its biological significance, plasma PC-1 is significantly related to several features of the insulin resistance syndrome. If a cause-effect relationship between PC-1 and the insulin resistance syndrome is established, measuring plasma PC-1 may easily identify subjects with PC-1-related insulin resistance. Breaking up such a heterogeneous condition into distinct phenotypes may be the key for future treatment of insulin resistance and its clinical outcomes (i.e. type 2 diabetes and/or CHD).
![]() |
Footnotes |
---|
Received May 6, 1999.
Revised June 22, 1999.
Accepted June 29, 1999.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|