Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Human leukocyte elastase (HLE) is an important inflammatory mediator involved in tissue destruction and functional disturbance in acute and chronic lung diseases, such as acute respiratory distress syndrome, some forms of emphysema, and cystic fibrosis (CF). The proteolytic activity of HLE in the lungs is regulated by several endogenous serine protease inhibitors, mainly ₁-proteinase inhibitor (₁-PI) and secretory leukoprotease inhibitor (SLPI). Measurements in vitro showed that both ₁-PI and SLPI bind HLE rapidly with second-order rate constants on the levels of 10⁷ and 10⁶ M¹ s¹, respectively (1). The actual association rates depend on a number of components in the microenvironment where the inhibitory reactions take place. Endogenous and exogenous reactive oxygen species oxidize methionine residues at the reactive sites of ₁-PI and SLPI, leading to substantial reduction of the association rates (5). Proteases originating from inflammatory cells and invading microorganisms cleave reactive site loops of the inhibitors, thereby abolishing their inhibitory activity (11). Polyanions are another class of substances that modulate the inhibitory reactions but do not directly modify primary structures of the inhibitors. Polyanions that may exist in the lung secretions and affect the inhibitory reactions include glycosaminoglycans (1, 10, 15), nucleic acids (16, 17), and alginate, the slime exopolysaccharide secreted by Pseudomonas aeruginosa (18). These polyanions were documented to reduce the association rate of HLE with ₁-PI (1, 2, 17, 18), whereas they increase the association rate of HLE with SLPI (3, 4, 16). Mechanisms by which polyanions modulate the reaction rates remain to be elucidated.

DNA is a major polyanion in the secretions of inflamed lungs. Previously we reported that DNA isolated from sputum markedly promotes the inhibition of HLE by native SLPI (16). Due to massive action of recruited neutrophils and other inflammatory cells, high levels of oxidative stress frequently exist in the secretions of inflamed lungs. Under such conditions, a large fraction of SLPI should be converted into its oxidized form. The present study demonstrates that sputum DNA promotes the inhibition of HLE by oxidized SLPI as well. We also show that sputum DNA significantly reduces the association rate of HLE and ₁-PI. The efficacy of sputum DNA in upregulating the activity of oxidized SLPI and in downregulating the activity of ₁-PI was compared with those of heparin and alginate. Further, DNA binding sites of HLE which may be involved in the up- and downregulation were probed by an elastase-specific oligonucleotide aptamer. The data obtained allow us to highlight the contribution of DNA in bronchial secretions to the balance between HLE and its endogenous inhibitors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

DNA Preparations

Sputum samples were collected from patients with CF and mechanically ventilated tracheostomy patients treated in the Division of Pulmonary and Critical Care, Hospital of State University of New York at Stony Brook, NY. DNA from individual sputum samples was purified by the thiocyanate-phenol method as described previously (16). Double-stranded DNA (dsDNA) and the [+] strand of single-stranded DNA (ssDNA) from phage M13mp18 were purchased from Bayou Biolabs (Harahan, LA). dsDNA from phage  was purchased from Worthington (Lakewood, NJ). Calf thymus DNA was from Sigma (St. Louis, MO). Some of the DNA preparations were thermally denatured to enrich single-stranded domains; i.e., DNA was heated in boiling water for 10 min and then rapidly cooled in ice. DNA concentrations were measured at 260 nm with the assumption that one unit of absorbance at this wavelength was equivalent to 37 µg/ml of ssDNA or 50 µg/ml of dsDNA. Sputum DNA concentrations were measured at 260 nm as for dsDNA because they contained less than 5% of single-stranded domains (16). The oligonucleotide aptamer for an exosite of HLE, 5'-TAGCGATACTGCGTGGGTTGGGGCGGGTAGGGCCAGCAGTCTCGT-3' (Oligo-I) (19), was synthesized by BioServe Biotechnologies (Laurel, MD). To facilitate the formation of secondary structural domains, Oligo-I in 36 mM, 1,3-bis[tris(hydroxymethyl)-methylamino]propane (bis- tris propane) buffer, pH 7.4, containing 94 mM NaCl and 6 mM KCl, was incubated at 85°C for 10 min and then slowly cooled to room temperature. Oligo-I concentration was measured at 260 nm with the extinction coefficients of 13,800, 6,500, 10,500, and 7,900 M¹ cm¹ for the bases A, C, G, and T, respectively.

Other Polyanions

Heparin sodium salt (grade I-A, 195 USP U/mg) extracted from porcine intestinal mucosa was purchased from Sigma. Alginate sodium salt was purified from culture of P. aeruginosa ATCC 39324 (18).

Enzyme and Inhibitors

HLE purified from neutrophil granules and ₁-PI from human plasma were purchased from Athens Research and Technology (Athens, GA). Recombinant SLPI was purchased from R&D Systems (Minneapolis, MN). The active site concentration of HLE was titrated with N-benzyloxycarbonyl-Ala-Ala-Pro-azaAla p-nitrophenyl ester (Enzyme System Products, Livermore, CA) (20). The active site concentrations of ₁-PI and SLPI were measured with titrated HLE. The oxidized form of SLPI that contained four methionine sulfoxide residues (OSLPI) was prepared as described (10). Briefly, SLPI (10 to 15 µM) was oxidized with newly synthesized N-chlorotaurine (2 mM) in Dulbecco's modified phosphate-buffered saline at room temperature for 3 h. The reaction was terminated by addition of L-methionine (20 mM) to quench residual N-chlorotaurine. Aliquots of OSLPI were stored at 20°C and used within 2 wk.

Kinetic Assays

Kinetic constants were determined in 36 mM bis-tris propane/ HCL, pH 7.4, containing 0.1 M NaCl, 0.1 mg/ml Triton X-100, and 0.1 ml/ml dimethyl sulfoxide, ionic strength 0.15, at 25°C. The substrates of HLE used were MeO-Suc-Ala-Ala-Pro-Val-p nitroanilide (pNA) (Sigma) for association rate k_a < 10⁷ M¹ s¹, and MeO-Suc-Lys(pic)-Ala-Pro-Val-pNA (Bachem Bioscience, Philadelphia, PA) for k_a > 10⁷ M¹ s¹. p-Nitroaniline released from the substrates by HLE in the presence of ₁-PI or SLPI, with or without DNA, was monitored at 405 nm for 10 to 60 min. The data collected were digitized and fitted to integrated rate equations by the nonlinear regression data analysis program, Enzfitter (Elsevier, Cambridge, UK). For the reversible association of HLE with SLPI, the integrated rate equation used was Equation 1 (4):

(1)

where [P] and [P₀] are the concentrations of product at any time t and time zero, respectively; ₀ and _s are the initial and steady-state velocities of substrate hydrolysis, respectively;  is an apparent second-order rate constant for the translation from ₀ to _s; and  is a variable for fitting the data, as defined by Equation 2:

(2)

where [E₀] and [I₀] are the initial concentrations of enzyme and inhibitor, respectively. The fitting yielded a set of curve parameters, ₀, _s, , and [P₀], from which k_a and the dissociation rate constant (k_d) were calculated according to Equations 3 and 4, respectively (4):

(3)

(4)

The ratio of k_d to k_a is defined as the inhibition constant (K_i):

(5)

By substituting Equations 3 and 4 into Equation 5 and rearranging, Equation 6 is obtained:

(6)

This is exactly the same equation as that derived by Henderson for tight binding competitive inhibitors (21). For some measurements with OSLPI in which precise computation of  from progress curves was difficult, Equation 6 was used to determine K_i from the slope of a linear plot of (1 + [S]/K_m)/(_s/₀) versus [I₀]/(1  _s/₀).

For the irreversible association of ₁-PI and HLE, the data collected were fitted to Equation 7 (18):

(7)

The fit yielded a parameter A, from which k_a was computed according to Equation 8:

(8)

Equations 1 and 7 apply to inhibitory reactions in which [I₀] is at least somewhat greater than [E₀], but do not require that [I₀] be much greater than [E₀]. For example, they may be used to analyze the inhibition of 1 nM HLE by 2 nM SLPI (Equation 1) or by 2 nM ₁-PI (Equation 7).

All results are reported as means ± standard deviation, which were calculated on the basis of data obtained from three or more experiments.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Downregulation of ₁-PI Activity by Sputum DNA

Figure 1 shows that sputum DNA delays the inhibition of HLE by ₁-PI. The DNA used was extracted from a mechanically ventilated patient's sputum sample (sample 2 in Table 1). As an inhibitor, sputum DNA alone inhibits 18% of the amidolytic activity (Figure 1, dashed line). However, in a reaction solution containing HLE, its substrate MeO-Suc-Ala-Ala-Pro-Val-pNA, and ₁-PI, more p-nitroaniline was released in the presence of sputum DNA (Figure 1, upper progress curve) than in the absence of DNA (Figure 1, bottom progress curve). Without DNA, HLE was completely inactivated by ₁-PI within 4 min, whereas with DNA complete inactivation was not achieved until 14 min. The corresponding association rates computed from the two progress curves are 8.4 and 2.3 × 10⁶ M¹ s¹, respectively. Thus, reduction of the association rate by a factor of 3.7 resulted in a 10-min delay before complete inactivation of HLE under these particular conditions.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Demonstration that sputum DNA reduces the association rate of 1-PI and HLE. The experiments were carried out in pH 7.4 buffer, ionic strength 0.15, at 25°C, with 983 µM MeO-Suc-Ala-Ala-Pro-Val-pNA. The concentrations of HLE, 1-PI, and sputum DNA were 6 nM, 20 nM, and 100 µg/ml, respectively. Solid straight line, HLE alone; dashed straight line, HLE plus DNA; bottom progress curve, HLE plus 1-PI; upper progress curve, HLE plus 1-PI plus DNA. The arrows indicate when HLE was completely inactivated.

The action of sputum DNA is dose-dependent, as shown by the relevant curve in Figure 2. When the DNA concentration was increased from 0 to 100 µg/ml, the association rate was decreased gradually from 9.7 ± 0.6 to 2.0 ± 0.3 × 10⁶ M¹ s¹. DNA concentrations in sputum specimens from patients with various pulmonary disorders were documented from 100 µg/ml to 9 mg/ml (22, 23). The highest DNA concentration used for the experiments in Figure 2, 100 µg/ml, is at the lowest limit of the reported range. However, even at this concentration, the DNA solution had already become visibly more viscous than water. An accurate measurement of association rate should consider the effects of viscosity, but due to technical difficulties no correction for viscosity was made in the present study. Therefore, the association rates reported here should be regarded as approximate values. With the concentration of DNA fixed at 100 µg/ml, we measured the effects of six sputum DNA samples on the association of ₁-PI and HLE. The results are summarized in Table 1. The reduced association rates fall in a narrow range from 2.0 to 5.7 × 10⁶ M¹ s¹, all of which are lower than the value in the absence of DNA. Because DNA concentrations in most sputum samples are higher or much higher than 100 µg/ml, on the basis of the data in Table 1 it can be concluded that DNA in sputum substantially delays the inhibition of HLE by ₁-PI.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Polyanion concentration dependence of the association rate of 1-PI and HLE. Open circles, heparin; filled circles, sputum DNA; open squares, alginate.

Figure 2 also gives the dose-response curves with alginate and heparin. Alginate effectively reduced the association rate of HLE with ₁-PI. However, the effect of this polyanion was obviously weaker than those of sputum DNA and heparin at all of the concentrations tested. The effect of sputum DNA on association of HLE with ₁-PI was greater than that of heparin at concentrations greater than 50 µg/ml, but the effect of heparin was greater at concentrations lower than 50 µg/ml. This concentration dependence is a reflection of the multiphasic dose-response curve of heparin, with a minimum value at a heparin concentration of 1 µg/ml. The molecular mechanism underlying the multiphasic dose-response curve for heparin is uncertain at present, but might be due to different states of aggregation of heparin polysaccharide chains at different concentrations. The distinctive shape of the dose-response curve of heparin allowed us to estimate the minimum association rate for ₁-PI and HLE in the presence of heparin, 7.4 ± 0.1 × 10⁵ M¹ s¹. When compared with this minimum value, it would appear that some species of DNA may be more effective than heparin in downregulating the activity of ₁-PI. For example, in the presence of M13mp18 ssDNA, denatured calf thymus DNA, and denatured phage  DNA at 100 µg/ml, the measured association rates of ₁-PI and HLE were all lower than 7.4 ± 0.1 × 10⁵ M¹ s¹ (Table 1).

Upregulation of Oxidized SLPI Activity by Sputum DNA

Among several types of oxidants generated by activated neutrophils, those produced by the myeloperoxidase- H₂O₂-Cl system are the species to which SLPI is most sensitive (8, 24). We chose N-chlorotaurine to prepare oxidized SLPI because this chloramine is the most abundant, long-lived oxidant produced by the myeloperoxidase- H₂O₂-Cl system (25, 26). By using the method outlined in MATERIALS AND METHODS, all four of the methionine residues in SLPI are oxidized to sulfoxides (10). Oxidation of the four residues raised the inhibition constant for SLPI against HLE from 58 ± 1 pM to 31 ± 4 nM. Before oxidation, the values of association and dissociation rate constants were 3.9 ± 0.1 × 10⁶ M¹ s¹ and 2.3 ± 0.03 × 10⁴ M¹ s¹, respectively, whereas after oxidation, the values were 2.3 ± 0.3 × 10⁵ M¹ s¹ and 7.0 ± 0.2 × 10³ M¹ s¹, respectively.

Figure 3 shows that sputum DNA promotes the inhibition of HLE by OSLPI. The DNA used was prepared from a CF sputum sample (sample 8, Table 2). As shown in the figure, the amidolytic traces with HLE alone (Figure 3, solid line) and with HLE plus DNA (Figure 3, dashed line) were linear. The trace with HLE plus OSLPI (Figure 3, dotted line) was slightly bent, reflecting the slow-binding mode of inhibition by OSLPI. At steady state, DNA alone and OSLPI alone, respectively, inhibited 9.6 and 26% of the amidolytic activity. However, when both DNA and OSLPI were present in the solution, the residual amidolytic activity was reduced to as low as 2.1%, as compared with 4.4% residual activity of HLE in the presence of native SLPI alone (Figure 3, solid progress curve). These data show that sputum DNA greatly enhances the inhibitory activity of OSLPI.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Demonstration that sputum DNA promotes the inhibition of HLE by OSLPI. The experiments were performed under the same conditions as those of Figure 1, except that the concentration of substrate was 1,021 µM. The concentrations of HLE, OSLPI, SLPI, and DNA were 6 nM, 20 nM, 20 nM, and 10 µg/ml, respectively. Solid straight line, HLE alone; dashed straight line, HLE plus DNA; dotted progress curve, HLE plus OSLPI; solid progress curve, HLE plus SLPI; dot-dashed progress curve, HLE plus DNA plus OSLPI.

It can be seen in Figure 3 that the progress curve with native SLPI is markedly nonlinear, whereas the progress curve with OSLPI plus DNA is nearly linear. Nearly linear progress curves were also observed with other sputum DNA samples under most conditions. Only under specific conditions, in which the concentrations of HLE, substrate, DNA, and OSLPI were carefully selected, were more nonlinear progress curves obtained with the oxidized inhibitor. Analysis of these nonlinear curves shows that DNA considerably increased the association rate of OSLPI and HLE, but only slightly decreased the dissociation rate of OSLPI/HLE complex. The association rate was estimated to be increased from the order of 10⁵ M¹ s¹ to the order of 10⁷ M¹ s¹, whereas the dissociation rate was still on the order of 10³ s¹ (data not shown). Thus, in the presence of DNA, OSLPI behaves as a relatively rapidly associating and dissociating inhibitor of HLE, with progress curves approximating the linear pattern characteristic of rapidly equilibrating inhibitors. Fitting the digitized data from these progress curves to Equation 1 no longer produced reliable  values, and it is therefore not valid to use the computed  values to further calculate association and dissociation rates. As an alternative, we used the Henderson equation (Equation 6) to determine the K_i. Table 2 lists K_i values in the presence of eight sputum DNA samples at 5 µg/ml. The values of K_i were found to be distributed over a range from 23 to 920 pM. Even when the sputum DNA sample with the least effect on OSLPI inhibition was used (Tables 1 and 2, sample 6), the affinity of OSLPI for HLE was increased by a factor of 34. The data support the conclusion that sputum DNA markedly enhances the inhibitory activity of OSLPI, and in the presence of DNA the inhibitory activity of OSLPI may be comparable to or even higher than native SLPI alone.

Figure 4 compares the effects of sputum DNA and heparin on the K_i for inhibition of HLE by OSLPI. Data with alginate are not compiled in this figure. Alginate also enhanced the inhibition of HLE by OSLPI but with lower activity. At a concentration of 100 µg/ml, alginate decreased the K_i from 31 ± 4 to 1 ± 0.1 nM. As shown by the data in Table 2, the effect of the sputum DNA sample with the lowest enhancing activity at 5 µg/ml was still greater than the effect of alginate at 100 µg/ml. The data in Figure 4 show that the activity of heparin is stronger than that of sputum DNA. In the presence of 0.7 µg/ml heparin, the K_i for OSLPI was determined to be 30 ± 2 pM, whereas with 5 µg/ml of sputum DNA, the K_i was 34 ± 1 pM.

View larger version (13K): [in this window] [in a new window]   Figure 4.   Polyanion concentration dependence of the Ki of OSLPI against HLE. Open circles, heparin; filled circles, sputum DNA.

Probing the DNA-Binding Sites on HLE

In addition to DNA isolated from sputum, we also examined the effects of DNA from other sources on inhibition of HLE by ₁-PI and OSLPI. The results are compiled in Tables 1 and 2, along with the results obtained with sputum DNA. The data demonstrate that regardless of the source, all of the DNA samples tested reduced the association rate of HLE with ₁-PI yet enhanced the association with OSLPI. Among the DNA species studied, M13mp18 ssDNA possessed the greatest activity. In the presence of 100 µg/ml M13mp18 ssDNA, the association rate of ₁-PI and HLE was determined to be 4.9 ± 0.3 × 10⁵ M¹ s¹, which is lower than the minimum rate obtained with heparin, 7.4 ± 0.1 × 10⁵ M¹ s¹. With 0.7 µg/ml M13mp18 ssDNA, the inhibition constant for OSLPI was 30 ± 1.4 pM (result not shown in Table 2), as compared with the K_i of 30 ± 2 pM determined in the presence of 0.7 µg/ml heparin (Figure 4). Phage  dsDNA illustrates the effect of strandedness on the modulating action of DNA. Native  dsDNA had the least effect on the inhibition by ₁-PI as well as OSLPI. When the dsDNA was denatured, however, the effects on the two inhibitors became comparable to those measured with M13mp18 ssDNA (Tables 1 and 2). Data in the two tables suggest that (1) with respect to modulation of the inhibitory activities of ₁-PI and OSLPI, ssDNA is more potent than dsDNA, and (2) thermal denaturation significantly increases the activity of dsDNA. In an earlier paper we provided several lines of evidence that ssDNA and single-stranded domains in sputum DNA have greater efficacy than dsDNA in accelerating the association of HLE and native SLPI (16). Data in Tables 1 and 2 also support the same conclusion for the modulation of ₁-PI and OSLPI activities.

To modulate the inhibitory reactions, DNA might be expected to interact with specific sites on HLE and SLPI (₁-PI does not bind DNA [17]), but no DNA-binding site has yet been characterized on HLE or SLPI. Recently, by using the Systematic Evolution of Ligands by Exponential Enrichment (SELEX) technique, several sets of single-stranded oligonucleotides with high affinity to HLE were isolated from large random-sequence libraries of RNA and DNA (19, 27, 28). Among the aptamers found, a G-rich oligodeoxyribonucleotide of 45-nt-long, which we call Oligo-I, has been studied in detail (19). We used Oligo-I to probe DNA-binding sites on HLE by examining its effects on the inhibition of HLE by ₁-PI and OSLPI.

Figure 5 shows the effects of Oligo-I on the association of ₁-PI and HLE. In the presence of 5 µg/ml (334 nM) Oligo-I, the association rate of ₁-PI and HLE was reduced to 4.5 ± 0.2 × 10⁶ M¹ s¹ (Figure 5, third bar from the top). The dissociation constant of Oligo-I and HLE was reported to be 17 nM (19). At an aptamer concentration of 334 nM, the specific site for Oligo-I on HLE was almost saturated (the calculated degree of saturation is 95%), and therefore the effect of specific binding of the aptamer on ₁-PI association may be assumed to be nearly maximal at this concentration. The value of the association rate in the presence of Oligo-I is roughly comparable to or slightly larger than those determined for the six sputum DNA samples in Table 1. Though the association rates in Table 1 were measured with DNA concentration of 100 µg/ml, the site(s) on HLE for sputum DNA are not necessarily saturated at this concentration. When the reaction solution contained 334 nM Oligo-I plus 100 µg/ml sputum DNA, the association rate was further decreased to 3.5 ± 0.1 × 10⁶ M¹ s¹ (Figure 5, bottom bar). Because the site for Oligo-I has already been saturated, sputum DNA must bind to additional sites on HLE. From the data in Figure 5, the following points can be deduced: (1) Binding of Oligo-I on its specific site does reduce the association rate of ₁-PI and HLE; (2) because the diminution of the association rate by Oligo-I alone is modest, there must be additional sites for binding of DNA which contribute significantly to deceleration of ₁-PI binding; and (3) the decelerating effects on ₁-PI binding can result from multisite interactions between HLE and DNA.

View larger version (12K): [in this window] [in a new window]   Figure 5.   Effects of Oligo-I on the association rate of 1-PI and HLE. The assays were carried out in pH 7.4 buffer, ionic strength 0.15, at 25°C with 1,418 µM MeO-Suc-Ala-Ala-Pro-Val-pNA. The concentrations of HLE, -PI, sputum DNA, and Oligo-I were 6 nM, 15 nM, 100 µg/ml, and 5 µg/ml (334 nM), respectively.

Next we studied the effects of Oligo-I on the association of HLE and OSLPI. The data obtained are expressed as percentages of residual amidolytic activity at steady state (Figure 6). At the concentrations employed here, sputum DNA, Oligo-I, and OSLPI individually inhibited HLE activity to a small extent; the respective residual activities were 94, 81, and 74%, respectively, under the experimental conditions used. In the presence of OSLPI plus Oligo-I the residual activity was 54%, which is slightly smaller than 60% (0.81 × 0.74 × 100% = 59.94), the residual amidolytic activity calculated on the assumption that OSLPI and Oligo-I inhibit HLE independently and additively. These results demonstrate that Oligo-I itself has a negligible effect on the inhibition of HLE by OSLPI. In contrast, sputum DNA shows a pronounced effect; in the presence of OSLPI plus sputum DNA, the residual steady-state activity was as low as 1.7% (the shortest bar in Figure 6). Further addition of Oligo-I to the reaction solution containing OSLPI and sputum DNA, however, increased the residual activity to 21%. In this set of experiments, the concentration of Oligo-I used was 145 nM, a concentration at which we estimated that 89% of the specific binding site for the aptamer would be saturated. This observation suggests that addition of Oligo-I must displace some sputum DNA molecules from the site, thereby diminishing the enhancing effects. Data in Figure 6 support the following conclusions: (1) Binding of Oligo-I to its specific site on HLE alone does not significantly enhance the affinity of HLE for OSLPI, but (2) binding of sputum DNA to sites that include the specific site for Oligo-I contributes to the promoting effects observed.

View larger version (17K): [in this window] [in a new window]   Figure 6.   Effects of Oligo-I on the inhibition of HLE by OSLPI. The assays were carried out in pH 7.4 buffer, ionic strength 0.15, at 25°C with 964 µM MeO-Suc-Ala-Ala-Pro-Val-pNA. The concentrations of HLE, OSLPI, Oligo-I, and sputum DNA were 6 nM, 30 nM, 145 nM, and 5 µg/ml, respectively. The reaction was allowed to proceed for 3 to 15 min, velocity of substrate hydrolysis during the last minute was measured and used as the velocity at steady state. The results shown were normalized with the velocity of substrate hydrolysis by HLE alone, (5.81 ± 0.17) µM/min, as 100%.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Oxidized SLPI

Studies in vitro revealed that oxidants released by activated neutrophils diminish the inhibitory activity of SLPI against HLE (8, 24). Similar compromise to SLPI inhibition may also happen in vivo. In many acute and chronic pulmonary disorders, massive activation of recruited neutrophils not only results in an imbalance between proteases and antiproteases, but also an imbalance between oxidants and antioxidants. For instance, high concentrations of chloramines have been detected in CF sputum specimens (26). In such an environment, it is expected that a large fraction of SLPI, if not all the inhibitor, would be oxidized. This condition prompted our evaluation of the inhibitory activity of oxidized SLPI. Oxidized SLPI itself is a fairly strong inhibitor of HLE. The K_i for oxidized SLPI in which a single active site-methionine residue was oxidized was reported to be 10 nM (9), whereas the K_i for the inhibitor in which all four methionine residues were oxidized was determined to be 21 (10) or 31 nM (this work). The present study shows that in the presence of sputum DNA, the activity of oxidized SLPI may be appreciably enhanced. We examined eight sputum DNA samples in all for this study, and at a concentration of 5 µg/ml, which is much lower than those in vivo, three of the samples decreased the K_i for OSLPI to values < 34 pM (Table 2), as compared with a K_i of 58 pM for native SLPI alone. The other five samples also enhanced the activity of OSLPI significantly, if not to a degree greater than that of the native inhibitor alone. Data from this study and in the literature support the following conclusions: (1) Although oxidation diminishes SLPI activity significantly, the residual activity of oxidized SLPI is still not negligible, and (2) polyanions which exist in the airways, such as DNA (this study) and glycosaminoglycans (10), enhance the activity of oxidized SLPI to a level comparable to that of native SLPI. Because oxidative stress is very common in many lung diseases, we hypothesize that oxidized SLPI is a functionally active form of SLPI and may still contribute substantially to the total lung antielastase defense. Many laboratories have reported the antielastase capability of whole sputum. A large portion of the elastase-inhibitory activity in the secretions previously ascribed to SLPI should probably be more accurately attributed to oxidized SLPI.

An Unusual Biologic Effect of Sputum DNA

Besides the effect on rheologic properties, a number of other biologic effects have been ascribed to DNA in sputum (for a brief list, see 16). Data from this and our previous studies (16) reveal a new biologic effect of sputum DNA: modulation of elastase inhibition by ₁-PI and SLPI. Table 1 shows that sputum DNA moderately reduces the association rate of ₁-PI and HLE. In the presence of the sputum DNA samples of the highest activity (DNA samples 1 and 2, Table 1), the association rate for ₁-PI and HLE was reduced by a factor of less than one order of magnitude at the DNA concentration used. As shown in Figure 1, this level of diminution is sufficient to delay the complete inhibition of HLE by ₁-PI for minutes under certain conditions; during this lag phase, the protease has the chance to access other natural targets. In contrast, sputum DNA promotes the inhibition of HLE by OSLPI with much higher efficacy. Even with the sputum DNA sample of the lowest activity (sample 6, Table 2) at a low concentration (5 µg/ml), the affinity of OSLPI to HLE was increased by a factor of more than one order of magnitude. A similarly marked effect was also observed for the DNA-accelerated association of HLE and native SLPI (16). We also compared the efficacies of sputum DNA, heparin, and alginate in modulating HLE inhibition (Figures 2 and 4). The efficacy of sputum DNA is considerably greater than that of alginate, and slightly lower than that of heparin at most of the concentrations tested. However, heparin, which was used as a model compound for glycosaminoglycans, is rarely present in the lung secretions. Glycosaminoglycans reportedly found in the lung secretions with higher concentrations include chondroitin sulfate (29) and hyaluronic acid (30). Hyaluronic acid has no effect on the inhibition of HLE by ₁-PI and SLPI and the activity of chondroitin sulfate is much lower than that of heparin (1, 10). Therefore, DNA appears to be the major polyanion in lung secretions that modulates the inhibition of HLE by its endogenous inhibitors. In the fluids lining the respiratory epithelium, ₁-PI, SLPI, and oxidized SLPI often exist simultaneously. In the presence of DNA, the increased association rates of HLE with SLPI and oxidized SLPI may be higher than the diminished association rate of HLE with ₁-PI. More HLE would be bound by SLPI or oxidized SLPI initially. However, because ₁-PI and HLE form an irreversible complex whereas SLPI and HLE form a reversible complex, SLPI-bound HLE would be gradually transferred to ₁-PI until the latter is saturated. The temporary sequestration of active HLE as an HLE/SLPI complex should contribute to the total balance between proteases and antiproteases in the lungs.

The DNA-Binding Sites on HLE

Our data demonstrate that ssDNA or single-stranded domains in sputum DNA are much more potent than dsDNA in modulating the inhibition of HLE by its inhibitors. It may be inferred from the various published applications of the SELEX technique to develop oligonucleotide aptamers for different biologic molecules that single-stranded oligonucleotides are able to form sequence-dependent, folded structures with a variety of shapes that fit a broad spectrum of binding sites. ssDNA molecules and single-stranded domains in sputum DNA may also be expected to form local folded structures by which they interact with binding sites on HLE or SLPI. These conformers are less likely to form within dsDNA molecules. We used Oligo-I, an elastase-specific aptamer that was identified by the SELEX technique, to probe the binding sites on HLE. On the basis of analysis of nuclear magnetic resonance spectra, two structural motifs were characterized within Oligo-I: its G-rich central portion condenses into an intramolecular G-quartet, and its 3'- and 5'-tails complement each other to allow formation of a duplex (19). Occupation by Oligo-I of its binding site on HLE does not block access of a synthetic substrate such as MeO-Suc-Ala-Ala-Pro-Val-pNA to the extended substrate-binding domain (19). However, the K_m and k_cat for this oligopeptide substrate are changed from 145 to 159 µM and from 24 to 16 s¹, respectively, upon the addition of 334 nM Oligo-I. Thus, although the Oligo-I binding site may be considered to be an exosite which does not directly overlap the catalytic center in HLE, occupation of this exosite by Oligo-I seems to induce a conformational change in the extended substrate binding domain, thereby affecting the kinetic parameters. Data in Figure 5 show that binding of Oligo-I also retards the inhibition of HLE by ₁-PI. If single-stranded domains in sputum DNA form folded structures with shapes similar to Oligo-I, the DNA might also be expected to reduce the association rate. The activity of Oligo-I in reducing the association rate is weaker than most of the sputum DNA samples listed in Table 1, with the exception of sample 4. This fact suggests the presence of additional DNA binding sites on HLE, at least one of which may induce more extensive conformational changes in the protease.

Binding of Oligo-I to its exosite on HLE affects inhibition by OSLPI. Oligo-I alone does not greatly enhance the affinity of HLE for OSLPI (Figure 6). Oligo-I slightly increases the association rate of native SLPI with HLE, but has no effect on the dissociation rate of SLPI/HLE complex (data not shown). These results suggest that binding of Oligo-I on HLE does not prevent the inhibitor from approaching HLE, consistent with the inference that the site for Oligo-I is an exosite. The observation that Oligo-I reduces the effect of sputum DNA on the association of OSLPI and HLE has implications not only for the apparent competition between DNA and Oligo-I for a binding domain on HLE, but also for the mechanism by which DNA enhances inhibition of HLE by OSLPI. Two mechanisms have been proposed to explain the accelerated association of SLPI and HLE by polyanions: (1) Activation of SLPI: Binding of polyanions induces a conformational change in the inhibitor, which promotes association of SLPI with HLE (3). Although we cannot rule out a direct effect of polyanions on SLPI, we can conclude that this effect does not extend to Oligo-I. (2) Formation of a ternary complex of SLPI and HLE on a single polyanion chain: Once both molecules are bound to the polyanion, one- dimensional diffusion of the reactants along the chain may be expected to result in more rapid association (10). If it is assumed that the dimensions of Oligo-I are not large enough to permit formation of a ternary complex with both HLE and SLPI, then its ability to partially displace sputum DNA from the exosite (or other neighboring sites) would effectively diminish the effects of sputum DNA on association of the protease and its inhibitor. The data presented here suggest that DNA-binding sites on HLE as well as SLPI contribute to the accelerated association, supporting a model of a ternary complex in which the enzyme and inhibitor can associate more rapidly than when both molecules are free in solution.