Joint Measurements of Leukocyte Elastase and Myeloperoxidase Promote Identification of the State of Neutrophils in Diabetic Patients

The clinic of diabetes mellitus (DM) offers a number of hypotheses about the leading role of polymorphonuclear neutrophils (PMNs) in both oxidative stress and diabetic complications. However, the results of numerous studies are extremely controversial. Why is it so? We appreciated the clinical significance of simultaneous measurement data of several PMN parameters, which must complement each other. For this purpose, myeloperoxidase (MPO) and elastase (EL) were jointly analyzed in the blood plasma from 160 type 2 diabetes mellitus patients with high levels of HbA1c. A weakly positive correlation (r ∼ 0.56) was observed between MPO and EL analytical data, and any correlation between the concentrations of MPO/EL and HbA1c was absent. Medians of 160 measurements of MPO/EL concentrations were ∼103/190 ng/mL, and 95% of all results were in the range below 320/1016 ng/mL, respectively. The share of DM patients whose concentrations of MPO, EL, or either of two parameters exceeded the corresponding reference values was 65%, 80%, and 82.5%, respectively. These findings—a high intensity of neutrophil degranulation process—indicated that some diabetic conditions promote the transfer of PMNs to an “arousal” or “subactivation” state, which is identical or similar to their activation, providing in vivo an almost inexhaustible source of extremely “aggressive” MPO and EL. Thus, the conjoint MPO/EL measurements confirm the leading role of PMNs in the development of various complications of diabetes. The paradox is that the diagnostic significance of MPO/EL as independent parameters in diabetic patients is unambiguous for a number of reasons.


Introduction
In practical diabetology, besides glycated hemoglobin (HbA 1c ), there are very few biomarkers which could reflect an imbalance in carbohydrate metabolism. 1 For pathologies with the rapid dynamics of inflammatory processes (for example, a postoperative state 2,3 ), polymorphonuclear leukocytes (PMNs or neutrophils) are not only participants but also potential witnesses of the actual pathology state. On the other hand, PMNs as biomarkers are very attractive objects of research, largely due to the numerous and relative simple mon-itoring methods. Consequently, if PMNs are involved in the pathogenesis of any disease, these cells could be tried as biomarkers. Diabetes mellitus (DM) is one such illnesses. 4 The clinic of DM allowed for the suggestion of a number of hypotheses about the leading role of neutrophils in the development of both oxidative stress and diabetic complications. However these hypotheses are not sufficiently supported by reliable analytical results. Generally the data about the integral evaluation of neutrophils' functional state by DM patients are surprisingly contradictory. It is difficult to choose any functional PMN parameter, which would unambiguously depend on either the level of diabetes compensation or disease per se. Moreover even the data about certain PMN protein content in the blood of diabetics are highly ambivalent.
In our studies, we tried to understand the causes of this phenomenon. Using the blood plasma of type 2 diabetes mellitus (T2DM) patients, we performed a parallel analysis of the two most studied PMN enzymes: myeloperoxidase (MPO) and elastase (EL). MPO, for example, can cause a number of pathological processes, such as the formation of reactive oxygen species or hypochlorous acid (HOCl), inducing vascular damage and endothelial dysfunction. [5][6][7][8] EL, for its part, may be responsible for the specific inactivation of C1inhibitor or antithrombin III: the main regulatory proteins of complement and coagulation systems. 9,10 By and large, there are many single studies with one or other PMN-derived proteins which indicate their potential implications by diabetic conditions. Therefore, it was hopeful that the conclusions drawn from joint studies of two independent markers could both confirm and complement each other. As a result, we will be able to (1) evaluate the significance of MPO and EL as biomarkers by diabetes and (2) appreciate the functional state of neutrophils.

Materials and Methods
We have studied two groups of patients with T2DM in compensated (HbA 1c <5.9%) and noncompensated (HbA 1c >7.1-14.7%) states. EDTA-plasma from these patients was received from diagnostic laboratory Synevo (Cluj-Napoca, Romania) after the necessary routine research. The clinical samples recieved were de-identified and no IRB approval was required for further studies. Thus, we have obtained blood plasma with the following well-known characteristics: the concentrations of glucose and HbA 1c . Blood plasma was aliquoted in 100 lL plastic microtubes and stored frozen at À25°C until use.
Quantitative determinations of MPO and EL concentrations in blood plasma were carried out using enzyme immunoassay tests (ELISA) provided by the BioVendor Company (Brno, Czech). All experiments were performed according to the manufacturer's instructions.

Results and Discussion
Numerous data concerning the state of ''diabetic'' PMNs are surprisingly contradictory. We have tried to touch on and understand this problem. It was decided to conduct joint measurements of two different PMN proteins, each of which is capable of characterizing the PMN state by itself. It was hoped that conclusions based on the results of two independent and parallel studies could both confirm and complement each other. Table 1 shows the concentrations of MPO and EL detected in plasma specimens of 160 patients with T2DM. Sixteen samples were obtained from patients with compensated disease (HbA 1c <5.9%, Nr. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16] and 144 samples from patients with poorly compensated illness (HbA 1c >7.1%; Nr. 17-160). The data in Table 1 are arranged in accordance with the growth of HbA 1c concentrations in the samples. The results are also represented graphically in Figures 1 and 2, demonstrating that correlation between the measured concentrations of MPO/EL and the corresponding concentrations of HbA 1c was absent. The concentrations of MPO and PMN-EL in the plasma of healthy donors (n = 4) varied in the selfsame range from 22 to 65 ng/mL. ELISA kits and the data of neutrophil-derived EL levels in plasma of healthy people were provided by BioVendor Company: the median concentration of EL was 35 ng/mL and 95% of all measurements were distributed in the concentration range <65 ng/mL. The values of the similar parameters in the plasma of patients with T2DM were *190 and 1016 ng/mL, respectively. Based on these data, it was estimated that >80% of patients with diabetes had elevated concentrations of EL in blood. Figure 3 represents the frequencies of EL concentration distribution in the plasma samples of healthy probands and patients with T2DM.
Of course, the elevated levels of EL in blood may not necessarily only be due to the diabetic conditions, but do always indicate the presence of persistent inflammatory processes in vivo owing neutrophil activation. To verify this conclusion, MPO measurements were performed in the same clinical samples.
The detected concentrations of MPO in control plasma samples from healthy people were in the same range as in a number of publications. [11][12][13][14] These results confirmed our decision to choose the recommended 95% reference limit of MPO concentrations *90.5 ng/mL. This value was calculated after testing 820 plasma samples of healthy people to ''provide a scientific basis for the further use of MPO in clinical practice.'' 11 As in the case with EL, MPO concentrations in blood, which exceeded this reference value, could be considered an indicator of the presence of in vivo inflammatory processes and infection, or the occurrence of pathologies associated with the activation of neutrophilic leukocytes. The median concentration of MPO in the blood of patients with poorly compensated T2DM was *103 ng/mL, and 95% of all measurements were in the concentration range less than *320 ng/mL. The quota of diabetic patients with MPO increased levels was *65%. To visualize the boundary, which separates healthy and pathological, we compared the frequencies of MPO concentration distribution in the plasma from healthy probands and diabetic patients (Fig. 4).
By and large, MPO and EL are usually considered indicators of nonspecific immunity state. As well known, after neutrophil activation, both MPO and EL enzymes leave the azurophilic granules of white blood cells and enter the extracellular space. Therefore, their concentrations in the blood should be approximately equal and correlated with each other. Indeed, many similar studies in healthy people prove this. From these positions, our results were surprising and characterized by the following two remarkable features: first of all, the 95% confidence range of EL concentration distribution in plasma samples from diabetic patients was more than three times as wide (!), as for MPO. Second, the correlation coefficient between the steadystate concentrations of MPO and EL was only 0.56 (Fig. 5). The value obtained made it possible to assert the existence of only moderate positive correlation between the parameters studied. Based on these observations, it was suggested that diabetic condition could somehow either affect the concentration of the studied enzymes or interfere with the ELISA reaction itself, distorting the analytical results. We tried to identify these processes and evaluate their influence. From this point of view, primarily, the following two unique properties of MPO should be taken into account: First of all, MPO is an autoantigen. This MPO feature was confirmed, for example, by the formation of corresponding antineutrophil cytoplasm antibodies (ANCA) in different vasculitis and a number of other rare (compared with T2DM) disorders. Respectively, chronically elevated MPO concentrations by T2DM can also cause sensitization of the body and, as a consequence, the formation of autoantibodies. [15][16][17][18][19] For example, autoantibodies against MPO were detected in 40% of patients with T2DM. 16 Generally, the simultaneous presence of both antigen and corresponding autoantibodies in samples can have a disruptive effect on the correctness of ELISA results. 20 Another characteristic property of MPO is its ability to adsorb on the walls of blood vessels. 21 This MPO feature was reported in a number of clinical trials: after heparin administration, the blood MPO concentrations were greatly increased (sometimes up to two times), while concentrations of EL remained unchanged. 2,22 In a series of experiments with cultured human endothelial cells, it was also shown that diabetes per se, or high glucose concentrations, contributed to phenotypic changes of the cell surfaces and caused accelerated sorption processes for both MPO and PMNs. 23 These findings were also clearly confirmed by studying microvessels from rats with experimental diabetes. 7,8 The process of PMN-EL occurrence in the blood has also its own remarkable features.
First, like MPO, neutrophilic EL under certain conditions can also play the role as an autoantigen. 24 However, we could not find any reliable information about the presence of EL-specific autoantibodies in DM or about their influence on the results of EL enzyme immunoassay analysis.
From these positions, more attention should be paid to another property of EL. The fact is that once in the extracellular space, EL does not remain intact, and instead builds complexes with some proteins from the socalled ''serpin family'' (serine protease inhibitors). This interaction really affects in vivo both the concentration and enzyme activity. Therefore, taking into account the assumption made, we also evaluated the effect of this phenomenon on EL analyses.
Among the broad family of EL inhibitors, the most active is a1-antitrypsin. Its content in norm (*1 mg/mL) is at least 10,000 times higher than the baseline level of EL in the blood. Consequently, the concentrations of the complexes formed must be correlated with concentrations of EL released from neutrophils. The BioVendore ELISA-Kit for quantitative analysis of PMN-EL is based on this principle and allows analyzing EL in the plasma samples up to 1000 ng/mL. In this regard, it should be emphasized that, although the clinical material under study was obtained mainly from patients with poorly compensated diabetes, 95% of all results were also located in concentration range <1016 ng/mL. Besides, the measured EL values, as well as the range of their concentration distribution in diabetics, were in good agreement with the results of previous studies. 25 Therefore, taking into account that PMN-EL does not adsorb on the surface of epithelium cells, we have concluded that the measured data of EL concentrations are reliable and correct.  (Table 1).
Unfortunately, a similar proposition cannot be used to evaluate the results of MPO measurements-a widespread research object in many disorders, inclusive diabetes. 4 The fact is that due to the unique ability to be adsorbed on the walls of blood vessels, the MPO analytical data most likely are not able to fully reflect the true picture of the enzyme content in vivo. The degree of sorption will depend on many factors that are difficult to take into account. Therefore, even in the case of a massive influx of MPO into the bloodstream, the measured MPO concentrations in T2DM patients can be underestimated in comparison with real ones. Perhaps for this reason, the values and range of MPO concentrations in patients with diabetes did not deviate too far from reference data. The same results were also represented in many other similar studies. 26-28 At the same time, we observed increased concentrations of MPO in 65% of patients with DM.
Based on the foregoing, it should be assumed that more valuable information about MPO through diabetes can be obtained by measuring not the enzyme concentration in blood, but its activity in biopsy samples. In this case, the magnitude of the analytical signal will depend not only on the enzyme dissolved but also adsorbed (immobilized) on the walls of the microvasculature. Indeed, in experiments with tissue pieces from rats with experimental diabetes, it was shown that the ratio of specific activities of MPO diabetes/ control/g of tissue reached >300%. 29 Naturally, the practical use of this approach is unacceptable. However, these data allow us to estimate approximately the degree of MPO sorption and notice its significance in diabetes. Of course, this topic requires special research. However, it can already be argued that MPO adhesion in vivo should proceed more intensively in the capillary network due to the significant increase of the inner blood vessel surface (and, consequently, an increase of sorption area). It is highly likely that this mechanism is dominant in the pathology of diabetic angiopathy, especially considering the significantly extended lifetime of immobilized MPO in vivo (see the following discussion).
Thus, we were able to identify several potential causes, which could significantly affect the ratio of MPO/EL concentrations in vivo. Among them, the main cause is solely the polycationic nature of MPO, which ensures its strong interactions with negatively charged surfaces of endothelial cells. 21 From this point of view, the use of EL as a biomarker seems to be a more reliable and informative parameter. However, it is well known that an increase of PMN-EL in the blood could be associated not only with diabetes but also with a large number of other pathologies. Therefore, in the absence of an adequate reference parameter, which will be specific especially for diabetes, it is unlikely to use only solo EL measurements as a nonspecific biomarker for mass use. 25 In fact, this statement applies not only to EL but also to the neutrophils themselves or any other PMN proteins when trying to use them as diabetic biomarkers. That is why many researchers tried through diabetics to find the relationship between the different characteristics of PMN and HbA 1c levels. However, some discovered relationships between HbA 1c and, for example, MPO activity, 30,31 amounts of neutrophil leukotriene B4, 32 or activity of PMN-alkaline phosphatase 33 strangely did not coincide with the results of similar 26,34 or present studies.
Apparently, the absence of such correlation between the concentrations of MPO/EL and HbA 1c in patients with T2DM can be explained if we take into account the time hierarchy of the processes under consideration and compare the life cycle durations of their participants ( Table 2). As follows from the data presented in Table 2, during the lifetime of one generation of HbA 1c in vivo, at least several tens of PMN generations will be replaced. In other words, substitution of MPO or EL concentrations in blood will proceed much faster than the changes of HbA 1c concentrations. Consequently, the content of HbA 1c in vivo in the framework of supposed correlations must be considered not as a variable, but as a constant value. This means that, even theoretically, there is no and cannot be any relationship between HbA 1c and MPO/EL activity/concentrations. For these reasons, the results of some recent studies, which have also indicated a positive correlation between HbA 1c levels and concentrations of certain proinflammatory proteins (with a clearly shorter life span than hemoglobin), such as angiopoietin-like protein-6 38 or interleukin-6 39 should be carefully reconsidered.

Conclusions
The joint measurements of MPO and EL can independently characterize the same one state of neutrophils from different ''points of view.'' This allowed us to compare analytical meaningfulness of both the tests used and to conclude about the PMN status in patients with T2DM. It was shown that (1) the quota of patients with elevated levels of MPO and EL were 65% and 80%, respectively; (2) between measured concentrations of MPO and EL was observed only a weak correlation, and (3) the 95% range of MPO distribution among diabetic patients was only slightly different from the control and was almost three times less wide as for the EL.
To evaluate the results obtained, it was necessary to take into account not only the physicochemical properties of MPO but also mechanism of its biological activity. Namely the nature of MPO determined its ''dualistic role'' in vivo-on the one hand, the ability to simultaneously protect the organism from various types of bacterial infection and, on the other hand, to attack the capillary network of blood circulation under certain conditions. This phenomenon is based on the same one mechanism of MPO action: at first stage, the polycationic enzyme globules ''adhere'' to the negatively charged surfaces of both different pathogens or endothelial cells. This process can lead to the depletion of MPO in the bloodstream even during powerful enzyme inflow from activated neutrophils. That is why MPO concentrations in diabetic patients may not differ significantly from normal values and the quota of DM patients with elevated MPO levels was depressed compared with EL test.
At the same time, the joint MPO and EL measurements unequivocally indicated the high intensity of PMNs degranulation processes in the blood of diabetic patients. That means that diabetes promotes the transition of neutrophils into a state of ''subactivation,'' which is identical or similar to the state of activated PMNs. Therefore, it can be argued that MPO and EL really take an active part in numerous pathological processes in diabetes. The paradox is that the diagnostic significance of MPO (especially) or EL as independent solo ''diabetic'' biomarkers may be in some cases not only uninformative but even erroneous.

Authors' Contributions
The article was written through contributions of all authors. All authors have given approval to the final version of the article.