Role of Leukotriene A₄ Hydrolase Aminopeptidase in the Pathogenesis of Emphysema

Leukotriene A₄ hydrolase (LTA₄H) has been known as a bifunctional enzyme. Although two enzymatic activities share an overlapping substrate site, their biological functions are distinctive (1, 2). The LTA₄H epoxy hydrolase (EH) converts leukotriene A₄ (LTA₄) to leukotriene B₄ (LTB₄), which is a potent inducer of neutrophil, macrophage, and T lymphocyte chemotaxis in human diseases (3–10). The LTA₄H aminopeptidase degrades the N terminus of peptides. Several studies demonstrated that a chemotactic triamino acid peptide, proline-glycine-proline (PGP), is produced due to breakdown of collagen by prolyl-endopeptidase, and PGP has been shown to induce chemotaxis of neutrophils by binding to CXCR2 (11–19). Recently, LTA₄H aminopeptidase has been reported to break down and clear PGP, thus mitigating the influx of neutrophils into murine lungs postinfluenza infection (20).

Pulmonary emphysema is a major manifestation of chronic obstructive pulmonary disease. It is characterized by alveolar destruction in patients due to infiltration of neutrophils, lymphocytes, and macrophages into cigarette smoke (CS)–exposed lungs (3, 8, 14, 21–23). Although a number of mechanisms were proposed to explain the pathogenesis of emphysema, its molecular pathogenesis is not yet clearly understood. Neutrophil-rich inflammation in emphysematous lungs of smokers led us to hypothesize that the LTA₄H aminopeptidase activity and bioproduction/clearance of PGP may play an important role during the development of CS-induced neutrophilic inflammation and emphysema.

Our laboratory has previously reported that the LTA₄H enzymatic activities make important contribution to the development of emphysematous tissue alterations (3, 24, 25). LTA₄H activity was found to influence severity of emphysematous alveolar remodeling in murine lungs exposed to transgenically overexpressed IL-13 (3, 24). LTA₄H EH activity was found to contribute to emphysematous alveolar remodeling and neutrophilic infiltration into lungs postexposure of intranasal elastase (3, 24). Whereas a number of studies have characterized the importance of the LTA₄H EH, no studies have investigated the biological contributions made by the LTA₄H aminopeptidase during the development of emphysema. Therefore, we first investigated the CS-induced alterations in the localization and enzymatic activity of the LTA₄H protein. These studies demonstrated that CS significantly increased the amount of LTA₄H protein in murine lungs and led to specific patterns of LTA₄H protein localization in lung tissues. Chronic exposure to CS also caused acidification of the bronchoalveolar lavage fluid (BALF) in mice, which suppressed the enzymatic activity of the LTA₄H aminopeptidase in the BALF. All of these events promoted exaggerated bioproduction of PGP and LTB₄. When the activity of the LTA₄H aminopeptidase was restored by selectively augmenting it without changing the EH activity, murine lungs were protected from CS-induced emphysematous damage due to reduction in the levels of PGP and neutrophilic infiltration into the lungs without changes in the levels of LTB₄. These studies demonstrate that the LTA₄H aminopeptidase pathway is an important contributor for the CS-induced neutrophilic inflammation, PGP clearance, and emphysematous alveolar remodeling independent of the LTA₄H EH pathway.

The solubility of 4MDM was enhanced by formulation with 2-hydroxypropyl-β-cyclodextrin (CDX) and dextrose. The vehicle was prepared as an aqueous solution of CDX and dextrose without 4MDM. Maximum tolerated dose was first determined with WT mice (n = 10), and then the CDX-4MDM dose equal to 25% of the maximum tolerated dose was used in all subsequent in vivo studies. The solutions were administered as drinking water for mice in cages. The levels of 4MDM in BALF were quantified first by enriching 4MDM using C₁₈ Sep-Pak (Waters) and then measuring the levels by using HPLC equipped with a UV detector (HPLC-UV). The HPLC setup consisted of the following: Shimadzu CBM controller, Shimadzu LC20AD pump, Shimadzu SPD20A UV detector, Shimadzu CTO20A column oven, Waters C₁₈ Symmetry HPLC column, and Shimadzu SIL20A autoinjector. The Shimadzu EZStart software was used to operate the instrument and for data analysis. The HPLC method parameters were as follows: binary conditions with a solvent rate of 1 ml/min for 15 min and simultaneous UV detection at λ = 237 nm. Each sample was injected twice. First, 100 μl unspiked samples was injected. Second, 90 μl samples spiked with 10 μl known amounts of 4MDM was injected. Peaks corresponding to the 4MDM of each sample were identified by superimposing the 4MDM-spiked UV tracing onto the unspiked UV tracing of the samples. The levels of 4MDM were assessed by calculating the area under the curve of the BALF samples to known standard curves of 4MDM.

The 129sv wild-type (WT) mice were purchased from National Cancer Institute. The LTA₄H knockout (KO) mice were provided by B. Koller (University of North Carolina) (26). Animal use was approved by the University of Virginia Institutional Animal Care and Use Committee. WT and LTA₄H KO mice, 8–12 wk of age, were exposed to CS using a Teague TE-2 smoking apparatus. The 3R4F research cigarettes were purchased from the University of Kentucky. Cigarettes were combusted at the rate of three cigarettes every 9 min, 5 h per day, 5 d per week, over 5 mo. Circulating air was trapped with an inline 22-μm filter attached to the air circulation system, and total particulate matters were monitored in the air. Mice were studied 0, 1, 4, and 20 wk postexposure to CS. Mice were exposed to either ambient air or CS while being treated with either vehicle or a pharmaceutical agent (4MDM), which selectively augments the LTA₄H aminopeptidase activity without affecting the LTA₄H EH activity (25).

Total lung volume and lung compliance were assessed by the Flexivent (SCIREQ), as previously described (3, 25, 27, 28). In brief, animals were deeply anesthetized with ketamine and xylazine mixture (60/5 mg/kg weight); the trachea was cannulated using p10 tubing; the sternum was opened; the diaphragm was cleared by opening the abdominal cavity; and animals were ventilated at a respiratory rate of 120 breaths/min with positive end expiratory pressure 3 cm H₂O per a prewritten macro program. This technique ensured that the live animals’ voluntary effort could not influence the physiologic values detected by the Flexivent. Once the animals were acclimated to the Flexivent ventilator, the prescribed Flexivent algorithm was performed, as previously described (3, 25, 27, 28), and the total lung volume and compliance were calculated using the software supplied with the ventilator. Animals were sacrificed following physiologic measurements, and tissues were harvested for postmortem physiologic and morphometric assessment.

As previously described (3, 24, 25, 29), animals were anesthetized, the trachea was cannulated, and the lungs were removed en block and inflated at 25 cm H₂O pressure of 1% melted low–melting-point agarose gel (Promega) in PBS. The trachea was tied to keep the lungs inflated and then fixed in 10 ml paraformaldehyde for 18 h. The fixed lungs were stained with H&E. Alveolar size was determined by mean cordlength (Lm), as previously described (3, 24, 25, 29, 30). Sequential digital pictures (at least 10 per animal) of the entire lungs were captured by an Axiostar microscope (Carl Zeiss Microimaging) and then processed by NIH Image 1.63 on a Macintosh computer with a macro downloaded from the National Institutes of Health server user-macro directory (http://rsb.info.nih.gov/nih-image/download/contrib/ChordLength.SurfaceArea).

Quantitative analysis of PGP in the BALF by HPLC-mass spectroscopy proved to be highly unreliable presumably due to the matrix effects of the aromatic carbons generated from exposure to cigarette combustion. These aromatic carbons saturated the chromatography columns and made the columns unusable even after the first run of chromatography. Therefore, a two-step preparation of the BALF was performed, as follows. First, the BALF was purified by normal flow HPLC. The retention time of a standard solution of PGP was determined to guide fractionation of PGP in the BALF samples. Second, the levels of PGP were quantified using a Thermo Electron TSQ Quantum Access MAX mass spectrometer system with a Protana nanospray ion source interfaced to a custom-packed 8 cm × 75-μm (internal diameter) Phenomenex Jupiter 10 μm C₁₈ reverse-phase capillary column. A 0.5-μL aliquot of each fractionated extract was injected, and the samples were eluted from the column using a reverse-phase gradient from 0 to 100% acetonitrile at a flow rate of 0.5 μL/min over 0.5 h. The nanospray ion source was operated at +3 kV. The multiple reaction monitoring used was 270.2 − 172.9, 116.1.

The software Virtual Molecular Dynamics version 1.8.7 (VMD) was used to visualize and render the final figure (31). The manual positioning of the ligands in the LTA₄H substrate-binding site was accomplished using the Molefracture Plugin version 1.3, which is available in the VMD package. The software GROMACS was used for energy minimization of the enzyme and ligands after solvation of the protein and ligands in a TIP3P water model (32). The ligands were parameterized using the SwissParam server (33). The minimization was carried out using the CHARMM27 forcefield, which is available in the GROMACS package (34). The E296Q mutant of the LTA₄H cocrystallized with the tripeptide Arg-Ala-Arg (Protein Data Bank: 3B7T) was used as a template to build the Pro-Gly-Pro substrate in the peptidase-binding pocket of the enzyme (35). A crystal structure of a WT LTA₄H (Protein Data Bank: 3FTV) was used for all subsequent modeling of the protein (36). On the basis of mutation data, the C terminus carboxylate was positioned to form a hydrogen bond with the side chain of Arg⁵⁶³ (37). Hydrolysis of the peptide occurs at the N terminus. Therefore, the N-terminal amine and the carboxyl oxygen were positioned to chelate with the Zn²⁺ atom. The LTA₄, a precursor to LTB₄, was docked following mutation data. The substrate was arranged to set two hydrogen bonds between the Arg⁵⁶³ residue and the carboxylic acid moiety of LTA₄ (37).

The concentration of the LTA₄H was determined by a custom-developed ELISA. Two LTA₄H Abs were purchased from commercial vendors. The levels of LTA₄H in the BALF and lung homogenate soup were determined by using double-paired coating Ab (Novus EPR5713) and biotinylated detection Ab (R&D Systems). Recombinant LTA₄H was purchased from Creative BioMart recombinant murine LTA₄H with His-tag as a standard. A total of 50 μL samples was used in duplicates to quantify the amount of the LTA₄H. Samples were analyzed against the standard curve generated from known quantities of recombinant mouse LTA₄H enzyme. The ODs were measured with a Dynex Technology TRIAD ELISA reader controlled by Concert-Triad Serves software (version 2.0.0.12) at a wavelength 450 nm for the measurement of LTA₄H.

Whole-lung single-cell suspension was analyzed by FACS analysis. Briefly, mouse lungs were harvested and then digested in lung digestion media with 1.0 mg/ml collagenase A (Roche Diagnostics) in RPMI 1640 (Quality Biological), and erythrocytes were lysed with ammonium–chloride–potassium buffer (Quality Biological). Cells isolated from the lungs were stained with PerCP-labeled CD45 (BD Biosciences), allophycocyanin-Cy7–labeled CD11b (BD Biosciences), PE-labeled Ly6G (BD Biosciences), and FITC-Labeled LTA₄H (Bioss). Infiltrating leukocytes were gated from nonleukocytes by the expression of CD45. Next, all CD45^high cells were gated into Ly6G^high and CD11b^high cells (neutrophils), as previously described (3). For the purpose of detecting cells containing the LTA₄H in their intracellular space, cells were permeabilized with flow cytometry permeabilization buffer (R&D Systems) after being stained with CD45 Ab. Cells expressing LTA₄H were separated into two groups, leukocytes (CD45^high) and nonleukocytes (CD45^low). Multicolor detection of the stained cells was performed on a FACScan flow cytometer (BD Biosciences). FlowJo (version 8.8.6.) was used to analyze the data.

TUNEL stain was performed with TACS 2Td-DAB in situ apoptosis detection kit per manufacturer’s protocol (Trevigen). Paraffin-embedded lung blocks were at 8 μm thickness and then stained with reagents provided in the TACS 2Td-DAB kit. At least five animals in each group were studied. Ten adjacent pictures were taken from lungs of each animal at original magnification power ×40; positively stained epithelial cells (brown) were counted and then averaged to conduct statistical analyses.

Immunohistochemistry was performed according to the commercial vendor’s protocol (Vectostain Elite ABC kit) with Biogenex Ag retrieval citra plus, Biogenex power block, murine LTA₄H Ab and isotype control Ab (Novus), and diaminobenzidine Vector peroxidase substrate kit.

Animals were anesthetized, as described above at specified time points, and BALF was collected and processed after pH was measured per previously published method (3, 20, 24, 25). Concentration of the LTA₄H protein was determined by an ELISA developed in our laboratory. Concentrations of keratinocyte-derived chemokine and MIP2 in BALF were determined by commercially available ELISA kits (R&D Systems). Quantitative analysis of PGP in the BALF by HPLC-mass spectroscopy was achieved in two steps. First, BALF was purified by HPLC and then PGP was quantified by a Thermo Electron TSQ Quantum Access MAX mass spectrometer system. Levels of LTB₄ were assessed by a commercially available LTB₄ enzyme immunoassay kit (R&D Systems) (3, 24, 25).

Recombinant human LTA₄H enzyme was produced, and the enzymatic properties of the LTA₄H aminopeptidase were assessed with a method previously published by our laboratory (25). A total of 25 μL BALF from each mouse was used to determine in vivo LTA₄H aminopeptidase activity.

Total RNA from lung tissue was isolated by TRIzol (3, 24, 25, 29, 30). SYBR Green real-time RT-PCR was performed and then standardized with β-actin to semiquantify the genes transcribing for metalloproteinase (MMP)8 and MMP9. Primers used for PCR are listed in Table I.

All data were expressed as mean ± SEM and assessed for significance by t test or one-way ANOVA with Bonferroni subgroup comparison, as appropriate. All statistics were performed using Prism software (GraphPad). In all analyses, a p value <0.05 was considered significant.

The 129sv WT mice developed progressively severe emphysematous tissue alteration with CS exposure, and this was demonstrated by all three primary end points, premortem lung physiology (total lung capacity and static compliance) and postmortem histology (Lm) (Fig. 1).

Levels of the LTA₄H protein in the BALF and whole-lung soup increased and peaked post-4 wk CS exposure. Then levels of the LTA₄H protein in the BALF of the CS-exposed mice decreased to the levels comparable to the age-matched, ambient air (AA)–exposed mice by 20 wk (Fig. 2A). However, levels of the LTA₄H protein in the whole-lung protein soup of the CS-exposed mice were still significantly elevated as compared with the AA-exposed mice by 20 wk (Fig. 2B). Significantly more LTA₄H was found in intracellular staining of CD45⁺ and CD45⁻ cells from the CS-exposed lungs as compared with AA-exposed lungs (Fig. 2C). The LTA₄H protein was located in the cytosol of the airway epithelial cells in AA-exposed mice, but the LTA₄H protein was localized into the nuclei of the airway epithelial cells in CS-exposed mice (Fig. 2D). Levels of LTB₄ and PGP were significantly increased in the BALF of CS-exposed mice as compared with AA-exposed mice (Fig. 3A, 3B). A significantly increased number of CD45^highCD11b^highLy6G^high cells infiltrated the CS-exposed lungs as compared with AA-exposed lungs (Fig. 3C). BALF was found to be significantly more acidic in mice exposed to CS for 20 wk (pH 6.7) as compared with mice exposed to AA for 20 wk (pH 7.2) (Fig. 3D). Enzymatic activity of the LTA₄H aminopeptidase was found to be significantly suppressed in vitro at pH 6.7 as compared with pH 7.2 (Fig. 3E).

Previously, we published the design and biological effects of 4MDM with which the LTA₄H aminopeptidase can be selectively augmented without affecting the EH activity (25). However, its application was limited due to low water solubility. Therefore, we developed a water-soluble formulation of 4MDM by encapsulating it with CDX. The hydrophobic cavity of the CDX structure encapsulates the water-insoluble 4MDM, whereas the hydrophilic exterior of the CDX affords a complex with improved water solubility (Fig. 4A). A 4:1 molar ratio of CDX to 4MDM (CDX-4MDM) remained homogenous aqueous solution over 72 h (Table II). After CDX-4MDM was administered to mice as drinking water mixed with dextrose (1 g dextrose per mL water), 4MDM was found in the BALF within 24 h (Fig. 4B). No mortality or gross abnormalities were observed after oral administration of CDX-4MDM for 5 mo (n = 33).

Molecular mechanism of selective augmentation of the LTA₄H aminopeptidase by 4MDM can be rationalized with a molecular model. In silico model of PGP and 4MDM in the substrate pocket of LTA₄H is presented with LTA₄ in purple and 4MDM in orange (Fig. 5A). The binding model for the 4MDM suggests that 4MDM augments the LTA₄H aminopeptidase by interacting with Phe³¹⁴ and subsequently perturbing the chelation of Glu³¹⁸ to the zinc atom and enhancing turnover of the peptide hydrolysis reaction (38). The long aliphatic tail of LTA₄ is predicted to extend into this pocket and preclude simultaneous binding of 4MDM. Therefore, 4MDM is not expected to alter LTA₄H EH activity, that is, processing of LTA₄ to LTB₄. In vivo effects of the CDX-4MDM were assessed by quantifying the levels of PGP and LTB₄. In CS-exposed mice treated with vehicle or CDX-4MDM, levels of LTB₄ were not significantly different (Fig. 5B). However, levels of PGP were significantly reduced in the mice treated with CDX-4MDM as compared with mice treated with vehicle post-20 wk CS exposure (Fig. 5C). A significantly fewer number of CD45^highCD11b^highLy6G^high cells was found in the CS-exposed mice treated with CDX-4MDM as compared with CS-exposed mice treated with vehicle (Fig. 5D). The selectivity of 4MDM at the LTA₄H enzyme target was confirmed by demonstrating that treatment with CDX-4MDM had no effects in mice with a null mutation at the LTA₄H loci (Fig. 5E). After 20 wk CS exposure, assessment by premortem total lung capacity and static compliance and postmortem Lm demonstrated that selective augmentation of the LTA₄H aminopeptidase activity protected lungs from CS-induced emphysematous alveolar tissue alteration (Fig. 6).

Effects of 4MDM on the enzymatic characteristics of the LTA₄H aminopeptidase at pH 6.7 were determined by using in vitro aminopeptidase activity assay with a previously published method (25). The aminopeptidase activity was augmented with positive dose response to 4MDM (Fig. 7A). The k_cat (enzymatic turnover number) of the LTA₄H aminopeptidase increased from 17.2 s⁻¹ in vehicle to 89.8 s⁻¹ in the presence of 8.0 μM 4MDM at pH 6.7 (Table III). Treatment with 8.0 μM 4MDM at pH 6.7 increased the aminopeptidase activity by >3-fold as compared with treatment with vehicle at pH 7.2 (27.5/s to 89.8/s) (Table III). Then the in vivo activity of the LTA₄H aminopeptidase was measured in the BALF. The specificity of this assay on measuring the LTA₄H aminopeptidase in the BALF was determined by demonstrating minimally detectable LTA₄H aminopeptidase activity in mice with null mutation at the LTA₄H loci as compared with mice with WT LTA₄H loci (Fig. 7B). The LTA₄H aminopeptidase activity was then found to be substantially augmented in the BALF from CS-exposed WT mice treated with oral CDX-4MDM as compared with CS-exposed WT mice treated with vehicle for 5 mo (Fig. 7C). These studies demonstrated that the loss of LTA₄H aminopeptidase activity caused by reduction in pH was rescued by treatment with CDX-4MDM, which augmented the LTA₄H aminopeptidase activity with minimal off-targeting effects. Recovery of enzymatic activity indicates that, at low pH, the LTA₄H protein is functionally suppressed but structurally intact.

In contrast to PGP, hydrolysis of LTA₄ is at the C-12 position and is stereoselective to give exclusively the R epimer. Asp³⁷⁵ is an essential residue for this process and presumably directs the addition of a water molecule to the C-12 carbon of the LTA₄ molecule (37). Therefore, the LTA₄ in the LTA₄H substrate-binding pocket was assembled by positioning the Re face of the C-12 atom opposite the Asp³⁷⁵ residue. The 4MDM ligand was docked with the unsubstituted phenyl ring of 4MDM positioned to form π-π stacking interactions with the phenyl ring of Phe³¹⁴, and the 4-methoxyphenyl substituent was positioned to interact with the hydrophobic side chain of Val³⁶⁷ through van der Waal interactions. The proposed binding mode is in accordance with the docking studies of Lai and coworkers (39) on related analogs.

Mice after 20 wk CS exposure had a significantly higher number of alveolar epithelial cells staining positive in TUNEL assay as compared with AA exposure (Fig. 8). However, selective augmentation of the LTA₄H aminopeptidase significantly reduced apoptosis in CS-exposed mice as compared with CS-exposed, vehicle-treated mice.

PGP is known to bind to CXCR2 and induces chemotaxis of neutrophils (13, 19, 20, 40). Levels of KC in the BALF post-CS exposure were not significantly different between the mice treated with CDX-4MDM and vehicle (Fig. 9A). However, levels of MIP2 in the BALF post-CS exposure were significantly reduced in mice treated with CDX-4MDM as compared with mice treated with vehicle (Fig. 9B). MMP8 and MMP9 have been found to be associated with PGP bioproduction and subsequent infiltration of neutrophils into the diseased tissues (12, 15, 18–20). Real-time RT-PCR of the whole-lung RNA demonstrated that the levels of genes transcribing for MMP8 and MMP9 were significantly increased post-20 wk CS exposure as compared with 20 wk AA exposure (Fig. 9C, 9D). Levels of gene transcribing for MMP8 were not significantly altered post-CS exposure when compared between mice treated with vehicle or CDX-4MDM. However, selective augmentation of the LTA₄H aminopeptidase significantly reduced the levels of gene transcribing for MMP9 post-20 wk CS exposure as compared with vehicle treatment.

An exaggerated activity of the LTA₄H enzyme frequently coexists with and is believed to contribute to the pathogenesis of a variety of diseases associated with neutrophils, including sepsis, cystic fibrosis, nonsteroid-dependent asthma, and chronic obstructive pulmonary disease (5, 41–45). Several studies have highlighted the importance of the exaggerated LTA₄H EH activity and its bioproduct, LTB₄, which contribute to neutrophilic inflammation and tissue remodeling (46–50). We have previously demonstrated that LTB₄ biosynthesis plays an important role in the emphysematous form of COPD induced by transgenic IL-13 and intranasally administered elastase (24, 25). However, new evidence has emerged to suggest that the second enzymatic activity of the LTA₄H, namely the aminopeptidase, may play an important role involving its ability to break down and clear PGP, a product of collagen breakdown with chemotactic properties.

Our current studies provide further evidence to suggest that the LTA₄H aminopeptidase makes an important contribution to CS-induced neutrophilic inflammation and emphysema in lungs. First, we demonstrated that CS exposure alters the levels and distribution of the LTA₄H protein, which promote bioproduction of LTB₄ and hinder clearance of PGP. After 5 mo of CS exposure, LTA₄H protein accumulated in the nuclei of the airway epithelial cells with minimal accumulation in the airspaces. Examination of the pH in the BALF after 5 mo CS exposure demonstrated that CS exposure acidified the BALF significantly (pH 6.7), and, at pH 6.7, the LTA₄H aminopeptidase activity was significantly more suppressed than pH of the AA-exposed BALF (pH 7.2). We hypothesized that breakdown and clearance of PGP occur primarily in the airspaces and that the bioproduction of LTB₄ occurs primarily in nuclei in which all necessary molecular components are present. Localization of the LTA₄H protein into nuclei and associated induction of LTB₄ bioproduction have been described by others in the past (51). Combination of these changes after CS exposure was thought to create a cellular microenvironment in which persistently elevated levels of LTB₄ and PGP would occur.

Traditional transgenic strategies using knockins or knockouts are limited due to their inabilities to selectively modify the individual functions of the LTA₄H bifunctional activities. All available pharmaceutical agents are limited because these agents are nonselective inhibitors of the LTA₄H EH and aminopeptidase. We overcame this limitation by developing 4MDM, which selectively augments the LTA₄H aminopeptidase without affecting the LTA₄H EH activity. This tool provided us with a new ability to probe the bifunctionality of the LTA₄H enzyme. We developed a water-soluble formulation for chronic administration suitable for a murine model of emphysema induced by chronic CS exposure. We then demonstrated that selective augmentation of the LTA₄H aminopeptidase protects lungs from CS-induced emphysematous alveolar remodeling by enhancing the clearance of PGP and reducing neutrophilic inflammation independent of LTB₄ bioproduction.

Our results demonstrated several notable observations related to the bifunctional activities of the LTA₄H and pathogenesis of emphysema. First, CS exposure in pulmonary system induces complex biomolecular alterations within the bifunctional enzymatic activities of the LTA₄H protein. We demonstrated that CS exposure causes distinctive localization of the LTA₄H protein in lung tissues and acidification of the airspaces that alter the LTA₄H aminopeptidase activity. These are, in our opinion, important biomolecular events that have been previously underappreciated. Even though a number of animal studies showed promising pharmaceutical benefits by using LTA₄H inhibitors that indiscriminately block both enzymatic activities, no agents have been successfully translated to the bedside as Food and Drug Administration–approved therapeutics (52–55). Our studies may provide additional biological insights as to how these nonselective modifiers of the LTA₄H enzymatic activities may inadvertently cause unwanted and potentially harmful biological effects. This could occur by unknowingly canceling the “good” LTA₄H aminopeptidase while blocking the “bad” LTA₄H EH with nonselective LTA₄H modifiers. Localization of the LTA₄H protein into the nuclei also raises a possibility that the LTA₄H may potentially behave as a regulator of transcriptional factors such as MAPK or NF-κβ. This is an intriguing hypothesis that merits further investigation in the future.

Second, our molecular modeling and the properties of 4MDM suggest that the aminopeptidase activity of the LTA₄H is a druggable target. The challenge with targeting a multifunctional enzyme is modifying one activity of the enzyme without affecting the other activity. Selective augmentation of the aminopeptidase activity is possible because of the large substrate-binding pocket of the LTA₄H enzyme (Fig. 5A). A number of available mutation data suggests that 4MDM would occupy the space between Val³⁶⁷ and Phe³¹⁴ and form π-π stacking interaction with phenyl ring of Phe³¹⁴ and van der Waal interactions with the hydrophobic side chain of Val³⁶⁷. In this model, 4MDM and PGP can simultaneously bind to the LTA₄H substrate site. Mutation data suggest that the aliphatic tail of LTA₄ must extend deep into the pocket, which would prevent simultaneous binding with 4MDM (37). It appears that this model offers a plausible explanation as to why 4MDM does not cause measurable changes in the LTA₄ EH activity in vivo. Further studies involving cocrystallization of 4MDM with the LTA₄H is necessary to make a conclusive determination for its mechanism selective for aminopeptidase.

Third, the animals treated with CDX-4MDM showed reduction in the levels of MIP2, whereas the levels of KC were unchanged. This was an unanticipated epiphenomenon. Reactive oxygen species have been found to activate macrophages, and this interaction was found to induce neutrophil chemotaxis via production of MIP2 (56). We speculate that the selective augmentation of the LTA₄H aminopeptidase may alter interaction between alveolar macrophages and reactive oxygen species and reduce the levels of MIP2. This is a speculation that merits more studies in the future.

Fourth, CS exposure upregulated the genes transcribing for MMP8 and MMP9, which are described to contribute to the bioproduction of PGP (12, 15, 17, 57). Treatment with 4MDM selectively downregulated the genes transcribing for MMP9, whereas it did not for MMP8. Previously, others have reported that the MMP9 is potentially involved in a feed-forward mechanism, which perpetuates the in vitro bioproduction of PGP due to CS exposure (12, 16). These findings suggest that the selective augmentation of the LTA₄H aminopeptidase could potentially halt this feed-forward cycle as a part of its therapeutic effects.

Although CDX-4MDM provided an opportunity to probe the biology of the LTA₄H bifunctional activities, there are limitations in our study worth mentioning. First, our current study appears to be potentially divergent from what we have observed in the LTA₄H KO animals (3). In our previous study, we reported that the null mutation at the LTA₄H loci led to significant protection against emphysema induced by intranasal elastase administration. The elastase-induced model of emphysema is a one-hit model induced by a single-dose administration of porcine elastase. The CS-induced model of emphysema is repeated-insult model (daily smoking) induced over 5 mo, and chronicity and repeating nature of the insults by CS most likely induced pulmonary responses divergent from the elastase-induced model of emphysema. Second, all of our analyses were observational related to the development of emphysema and neutrophils. Secondary in vivo confirmation of our results would require depletion of neutrophils. Inherent limitations in available methods hindered this attempt. Mice with null mutation at the CXCR2 loci would alter nonneutrophil-dependent biological mechanisms such as angiogenesis, which make important contributions to the pathogenesis of emphysema (58–60). Neutrophil depletion with anti-Ly6G Ab was attempted in our laboratory for as long as 1-mo duration (3). However, administration of this Ab longer than 1 mo caused serum sickness in animals and was determined to be unsuitable for a chronic model like that of 20 wk CS exposure.

In summary, our studies highlighted cellular mechanisms associated with the perturbations in the bifunctional enzymatic activities of LTA₄H, which is responsible for emphysematous alveolar remodeling induced by CS exposure. Our studies provided mechanistic insights that the development of the emphysematous destruction in lungs may be ameliorated by interventions that selectively regulate the LTA₄H aminopeptidase. This report features the unexplored bifunctional activities of the LTA₄H pathway as noteworthy sites for future investigations designed to evaluate disease susceptibility, disease progression, and therapeutic utility in emphysema.

Two authors (Y.M.S. and M.P.) are part owners of a biotech company that is attempting to commercialize chemical agents used in this manuscript for disease indications described in this manuscript (emphysema).