CLINICAL APPLICATION OF PLASMA PROTEIN C DETERMINATION
George Zhu*1, AW Broekmans2, RM Bertina2
1The Institute of oncology, Tehran University of Medical sciences, Tehran, Iran.
2University Hospital Leiden, The Netherlands.
Objective: Protein C, a vitamin K-dependent coagulation factor, is involved in blood coagulation. Activated protein C inactivates Va and VIIIa and stimulates fibrinolysis. In this process, protein S serve as an important factor for activated protein C. Furthermore,excess protein S drives cancer cell proliferation and cell survival through oncogenic receptor Axl(Anexelekto). We determined ranges of protein C both in healthy individuals and distinct hospitalized patients.
Methods: A total of 100 patients with different diagnostic diseases and 50 healthy individuals were included in their plasma protein C determination. A rabbit antibody against human protein C was used for the quantitative estimation of plasma protein C antigen by using rocket immunoassay.
Results: In healthy individuals protein C antigen(PC:Ag) ranges o.6439- 1.4752 µ/ml. The mean coefficient of variation(CV) of length of rocket was calculated to be 12.45%. PC:Ag within laboratory variation was 11.47%. Plasma protein C antigen was destroyed at 56℃ for 30 minutes, whereas no significant decrease of protein C was found at 4℃ refrigerator for one week.
Conclusion: The results showed that plasma protein C antigen was considerably high in 22 diabetes mellitus. On the other hand,the PC:Ag was significantly decreased in 19 liver cirrhosis(p< 0.001) and was positively correlated with serum albumin levels(p< 0.05). In 20 acute leukemias,on the average,there was slightly lower values in PC:Ag,and accompanied with significant decrease of PC:Ag in 5 FAB M5 subtype and in 9 hyper- leukocytes acute leukemias. However,the 3 acute promyelocytic leukemia (APL) with overt laboratory picture of DIC(disseminated intravascular coagulation) had protein C concentration no lower than the remaining 2 patients with infectious DIC, which suggested the coagulopathy in APL might be due to mechanisms different from other forms of DIC.
Keywords: Antigen, clinical determination, plasma protein C, rocket immunoassay.
INTRODUCTION
Protein C is a vitamin K-dependent protein which was discovered by Stenflo in 19761. It was composed of a light and heavy chain linked by interchain sulfide bridges (Figure 1)2. In its activated form protein C inactivates the activated factor V (Va) and VIIIa and accelerates fibrinolysis3. Activated protein C(APC) requires the presence of protein S4. Protein S is a plasma protein that serve as an important cofactor for activated protein C in the blood anticoagulant system4,5. Protein S also act as a mitogen on distinct cell types and is a ligand for Tyro3, a member of the Axl family of oncogenic receptor tyrosine kinases6,7. The TAM family of proto-oncogenic receptor kinases (Tyro3, Axl and Mer) is implicated in many human cancers8-11.The full length Tyro-3, Axl, and Mer protein contain 890,894,and 999 amino acids, respectively8. Protein S is expressed in lung carcinoma tissue, predominantly in tumors of squamous cell origin, and in head and neck and oral squamous cell carcinoma12. Most of the protein S is synthesized in the liver by hepatocytes together with the closely related protein Gas6, as an activating ligand for TAM family. Gas6 and protein S share 43% amino acid sequence identity and have the same domain structure with the exception of thrombin cleavage sites which present in protein S but not Gas68. The role of the TAM cognate ligands Gas6 and protein S was demonstrated in homeostatic regulation of the immune, vascular and nervous systems. Excess protein S drives oral squamous cell carcinoma proliferation and cell survival through high expression and the activation of oncogenic receptor Axl7,11,13-16.
Griffin et al.,17 reported the first family with a congenital protein C deficiency. Bertina and Broekmans et al.,4,18,19 reported the first Dutch family with a congenital protein C deficiency. Using an immunologic assay for protein C20, we carried out the measurement of protein C in patients with clinical various diseases especially bleeding diathesis.
MATERIALS AND METHODS
Blood collection:
Venous blood was collected in 1/10 volume of 0.11 M sodium citrate. Plasma was prepared by centrifugation of blood at 3000 rpm for 10 min. Platelet free plasma of 100 patients and 50 healthy volunteers were stored at -20℃ for the assessment of protein C levels. Three tubes with rabbit anti-protein C serum was kindly supplied by Prof. RM Bertina, Leiden University Hospital, The Netherlands. The period of experimental procedure was completed in early 1986-89. Procedures:
A 12.5 ml of 1% indubiose A37 or commeciable agarose solution is obtained by boiling in gel buffer (PH8.8 Tris-BB or Tris-BB-EDTA) during continuous stirring. The solution is allowed to cool to 55℃, and a 60 µl amount of rabbit protein C antiserum and a 0.05ml of 0.11 mol/l sodium citrate solution was added during careful mixing. A 100 x 60 x 2 mm gel is prepared on small glass plate by pouring antibody- containing agarose (50~55℃) into a coated glass plate placed on a horizontal table. After 10 minutes the solidified gel is placed in a refrigerator (4℃) in a moist atmosphere. The gel should harden before wells can be punched.
Wells are punched out with a gel puncher, using a template.
A row of wells (about 4 mm in diameter) with their centers at least 5 mm apart is punched in the gel. After appropriate dilution the samples (13 µl) are deposited in the wells. The ionic strength of the samples should preferably be approximately the same as that of the electrophoresis buffer. Subsequently, samples are applied after the glass plate has been placed on the electrophoresis apparatus and a low voltage has been applied (0.5 V/cm). Standards in suitable dilutions are deposited in the middle holes of the row.
Electrophoresis is performed at 120 voltes (4-6 mA/plate) at room temperature 12~30℃ for 20 hours overnight.
The length of rocket immunoprecipitate was measured by using a ruler, which is linearly correlated to the amount of antigen. Results (including normal controls and patients) are plotted on log-log paper (length of rocket in mm versus percentage antigen), based on the construction of the standard curve.
RESULTS
Protein C antigen in healthy individuals
Protein C antigen was measured in the plasma of 50 healthy individuals who contributed to our pooled normal plasma. A mean protein C antigen of 1.0578 µ/ml was calculated with a SD of 0.1886 µ/ml; individual protein C values ranged from 0.6439 µ/ml to 1.4752 µ/ml. The distribution of age on the protein C antigen concentration was evaluated in 50 healthy individuals. The results are summarized in Table 1, which indicated that no statistical differences in protein C antigen were found for the different age groups (p>0.5). Also, no difference was observed between men (mean 1.0613 µ/ml) and women (mean 1.0550 µ/ml). Data from those in Table 1 are the representation of Mean±Standard deviation without significant difference of mean for the different age groups (p> 0.5).
The qualitative evaluation of protein C antigen determination
CVT= √(CVs)2 +(CVp)2
CVT in PC antigen levels within laboratory was calculated to be 11.47%.
Protein C antigen in distinct diseases
Diabetes Mellitus(DM)
In a group of 22 patients with diabetes mellitus (DM) an average value of protein C antigen was 1.4238 µl/ml, which is considerably higher than the average value in healthy individuals (1.0578 µ/ml). Among 8 DM, the protein C antigen level was over 1.4752 µl/ml. The increased plasma protein C antigen was in turn diabetic ketoacidosis (n=4, 1.5361 µl/ml, range 1.3516~1.8434 µ/ml),diabetic angiopathy (n=5, 1.4551 µ/ml, range 1.1888~2.0894 µl/ml), DM without vascular complication (n=9, 1.4403 µl/ml, range 1.0834~1.8926 µl/ml) and diabetic ketosis (n=4, 1.2352 µl/ml, range 1.1955~1.2663 µ/ml). Plasma PC:Ag level was positively correlated with urine glucose level (r=0.9581, p<0.001).
Pregnancy-Induced Hypertension Syndrome (PIHS)
In 15 pregnancy-induced hypertension syndrome (PIHS), 10 severe preeclampsia were observed, and a significant increase of protein C antigen (PC:Ag) value (1.4673 µ/ml, range 0.7781 ~2.2525 µl/ml), while in 2 moderate PIHS the PC:Ag has normal limit range (0.7787,0.9481 µl/ml respectively), another 3 mild patients with PIHS had decreased plasma PC:Ag level(0.5603 µl/ml). Intriguing, repeat PC:Ag level was recovered to normal value (1.1455 µ/ml) in one mild patient with PIHS after partum. Plasma protein C antigen was positively correlated with urine protein (r=0.9518, p<0.001).
Uremia
In a group of 10 uremia there were no significant changes of protein C antigen concentration when compared to normal controls (1.1387 vs 1.0578 µ/ml, p>0.2)
Thrombotic diathesis
To screening congenital protein C deficiency, we detected 6 patients with cerebral haemorrhagic infarction and 3 deep venous thrombosis, no PC deficiency was found in this study.
Liver cirrhosis
19 patients with liver cirrhosis included viral liver cirrhosis 14 cases and other 1 primary biliary cirrhosis, 1 scistosomal cirrhosis, 1 alcoholic liver cirrhosis and 1 primary hepatocellular carcinoma, respectively. The average value of protein C antigen was 0.5501 µ/ml (range:0.1436~0.9291µ/ml), which was significantly lower than the average normal controls(1.0578 µ/ml). Moreover, 13(68.4%) liver cirrhosis the protein C antigen was below the lower limit of normal control (range: 0.1436~0.6188 µ/ml). 14 patients with viral liver cirrhosis the protein C antigen was 0.5068 µ/ml. Three severe liver cirrhosis had its the lowest value of protein C antigen 0.1436, 0.1846 and 0.1919 µ/ml respectively. The PC:Ag level was 1.0883 µ/ml in one liver cirrhosis complicated with diabetes mellitus. Plasma PC:Ag level was positively correlated with serum albumin value(r=0.9680, p<0.05).
Leukemias
In 20 acute leukemias, there was slightly decreased in protein C antigen level (0.9340 vs 1.0578 µ/ml), but no statistical difference was found between acute leukemia and normal controls. When further analysis of acute leukemia with different cytological subtypes morphology according to FAB classification, it was shown that there was remarkably decrease of PC:Ag(0.8879 µ/ml, p<0.05) in 5 FAB M5 subtype. The decreased PC:Ag level(0.7833 µ/ml) was also found in 9 hyper-leukocytes acute leukemias (WBC 23.6~280 x109/l). Intriguing, there were no significant difference in terms of protein C antigen between 3 acute promyelocytic leukemia (APL) with DIC, whereas 2 patients with infectious DIC(1 nodular panniculitis and 1 septic shock) the average value of protein C antigen was 0.2328 and 0.1646 µ/ml respectively.
Protein C deficiency plasma
1 vial protein C deficient plasma which provided by RM. Bertina was used to measure the protein C antigen < 0.01%.
DISCUSSION
This underlines the need for each laboratory to construct its own reference values for PC antigen in the laboratory determination of different methods5,21. We set up successfully the electroimmunoassay (EIA) of the protein C antigen with the laurell method. Based on our laboratory condition, the mean PC:Ag level in 50 healthy individuals was 1.0578±0.1886 µ/ml, which was slightly higher PC antigen levels than that of Bertina's results. To evaluate the performance and specificity of PC:Ag assays5,21,the ICTH subcommittee on protein C organized an international collaborative study in which 13 lymphilized plasma samples were distributed among 17 different laboratories. Three different types of methods has been used: ELISA's using polyclonal anti-protein C IgG (n=9), electro- immunoassays (n=10) and other methods (three laboratories used monoclonal antibodies against protein C, one laboratory performed a RIA with affinity purified polyclonal antibodies against protein C and one laboratory used an immunoradio- metric assay of protein C antigen). No statistically significant results were obtained with these three different methods. ELISA, RIA and IRMA methods were found to be more sensitive than the electro-immunoassay (EIA). The mean coefficient of variation (CV) for protein C antigen in various plasma was calculated to be 22%, with CV within laboratory variation 11.7% and between laboratory variation 17.8%. In this current study, the mean CV in the construction of standard curve for protein C antigen was 12.45%, with CV within laboratory 11.47%. The results indicated that the protein C antigen measurement was specific, practice and available in clinical useful. The measurement of plasma PC:Ag has become relevant in clinical medicine since the important demonstration that hereditary protein C deficiency is linked to a history of recurrent venous thrombosis22. PC is found to be low in a number of acquired condition such as during the postoperative and neonatal periods23, DIC21,23,24, liver disease21,23-25 and oral anticoagulant treatment (eg. warfarin)4,17,26. Elevated PC levels, on the other hand, have been detected in a number of unrelated condition, such as ischemic heart disease, diabetes, late pregnancy27 and in women taking oral contraceptive with a high estrogen content28.
Type 2 Diabetes mellitus is characterized by vascular complication with accelerated micro and macro vascular thrombotic disease. A hypercoagulable state is appreciated in this disease. In this study, PC:Ag levels of patients especially in ketoaccidosis and diabetic angiopathy were significantly higher than that of the control group(P<0.001). The results were consistent with the reports by Vigno27, Takahashi30, Saito31, and Garcia32. Moreover, evaluation of protein C and protein S levels in patients with DM receiving well- managed T2DM (type 2 DM) had significantly higher antigen levels and activity of PC(5.78 vs 4.64 µl/ml),PS and ATIII compared to those that were poorly- managed33. The decreased PC antigen was observed in type I diabetes mellitus34,35. This may reflect the possibility that diabetic ketoacidosis (or angiopathy) and hypercoagulate may cause endothelial injury, which lead to elevated levels of thrombomodulin(TM) and compensatory increase the PC synthesis by hepatocytes. Patients with the nephrotic syndrome and uremia are at increased risk of developing renal venous and arterial thromboembolism. In the measurement of PC:Ag levels in 10 uremia, there was no significant difference compared to normal controls. Cosio36 measured the plasma concentration of protein C in 17 patients with severe proteinuria, the results was not significantly difference between the healthy subjects and 14 patients with chronic renal insufficiency (CRI).
But protein C was elevated in 71% of patients with proteinura, and was inversely correlated with serum albumin concentrations. Demaschell37 found that PC functional activity and the antigen level were normal in 30 uremic patients before dialysis. Data of current study are consistent with the previous observation. In this current study, PC:Ag levels in 19 different types of liver cirrhosis were significantly lower than those in healthy individuals (0.5501 vs 1.0578 µl/ml, p< 0.001). 68.4% of liver cirrhosis had PC:Ag level below the lower limit of normal controls. Moreover, plasma PC:Ag level was positively correlated with serum albumin values. In earlier study, Griffin24 reported that eleven of 15 patients with clinically severe liver disease (9 with positive FDP, 6 with negative FDP) has decreased plasma protein C levels. Mannucci21,23 also found that in 58 chronic liver diseases protein C levels were lower than 60 healthy subjects, the decreased degrees roughly proportional to the severity of the disease. These findings indicated that protein C deficiencies occur the dysregulation of hepatic PC synthesis, and might play a role as a predictor index in liver diseases38.
Acute leukemias varied in PC antigen level. In this current study, mean PC antigens in 20 acute leukemias were slightly lower than that of the normal individuals. Among FAB subtypes, the decreased PC:Ag concentration were found in those with M5 subtype and hyperleukocytic acute leukemias. Moreover, the PC:Ag had no lower in a limited 3 patients with APL complicated by DIC, which suggested the coagulopathy in APL might be due to mechanisms different from other forms of DIC such as infectious disease (eg. septic shock)39,40. The results was similar to previous studies by Griffin et al.,24 and Rodeghiero et al.,25. In clinics, two types of hereditary protein C deficiency can be recognized4, 22. In patients with type I protein C deficiency, the plasma concentration of protein C activity and/or protein C antigen will be below the lower limit of the normal range,while the ratio between protein C activity and protein C antigen is still within the normal range. A type I protein C deficiency inherits as an autosomal dominant trait. In type II protein C deficiency, the patients protein C antigen will be within the normal range, while the protein C activity and the ratio between protein C activity and protein C antigen are below the lower limit of the normal range. In both types of protein C deficiency the plasma concentration of the other vitamin K dependent coagulation factors should be within normal range. The presence of 1 vial PC deficiency plasma which was provided by RM Bertina was the lower than 0.01 µ/ml, and which was consisted with the criteria of type I severe protein C deficincy41-45. In Dutch family, more than 80 patients from 30 unrelated families have been identified with a type I protein C deficiency. Individual protein C values may vary between 0.30 and 0.76 µl/ml4,22. Up to now, only five patients with a type II protein C deficiency have been reported22. It seems likely the main prevalence of clinical type I protein C deficiency. Like protein C deficiency, familial protein S deficiency is also reported46-49. At present, some case reports and case series suggest that protein C concentrates may improve the outcome in patients with congenital or acquired protein C deficiency45,50-52. The use of protein C concentrates in adult septic shock53-55 is also testable.
AUTHOR’S CONTRIBUTION
All authors have worked equally in this work.
ACKNOWLEGEMENTS
This work was completed during the previous period of Master degree of medicine. The author greatly thanks for the kind technical assistance of my supervisor Prof. RM Bertina, University Hospital Leiden, The Netherlands.
CONFLICT OF INTEREST
No conflict of interest associated with this work.
REFERENCES
PMID: 1245477