Intratracheal elastase challenge of laboratory animals has long been established as a model for observing the physiological and morphological changes that result from alveolar destruction, the hallmark of emphysema. However, instillation of elastase suspended in buffer results in widespread inflammation and variable emphysematous lesions, which has made the identification of specific cellular and molecular events associated with the onset of emphysema difficult to define. Here we establish a bead-based elastase delivery system that induces localized tissue destruction, a key event in the initiation of emphysema. Elastase was coupled to bisacrylamide beads, which were shown to retain enzymatic activity prior to intratracheal administration in mice. C57BL/6 mice were given a single dose of 40,000 beads, which became distributed throughout the small airways and parenchyma of the lung. Elastase-coupled beads resulted in a quantifiable loss of alveolar tissue immediately surrounding the beads, an effect that was not observed with beads that lacked protein altogether or with beads containing elastase inactivated by an irreversible inhibitor. Furthermore, beads bound with active elastase elicited local recruitment of mononuclear cells, including macrophages, and polymorphonuclear neutrophils to the site of bead deposition, a feature consistent with the cellular infiltration observed following conventional solubilized elastase challenges. This work identifies a novel bead-based enzyme delivery system that also extends the elastase model of emphysema to permit the characterization of mechanisms that drive alveolar surface area loss following elastin degradation in focal emphysematous lesions.
chronic obstructive pulmonary disease (COPD) encompasses chronic bronchitis and emphysema, the latter of which is marked by the irreversible destruction of lung tissue. Despite considerable study in a variety of models, specific mechanisms underlying emphysematous loss of alveolar walls and respiratory surface area remain largely elusive. One possible explanation for the progressive nature of emphysema is that put forth by the protease-antiprotease hypothesis (13). This theory is supported in part by the early development of emphysema in patients displaying α1-antitrypsin deficiency, but also in animal models by the induction of airspace enlargement resulting from intratracheal aspiration of various proteases capable of degrading lung tissue, namely neutrophil, macrophage, and pancreatic elastases (14, 17, 24).
Elastase-induced emphysema results in chronic tissue destruction despite the fact that the exogenous enzyme is inactivated within the first 24 h following administration (25, 27). This persistent destruction of lung parenchyma is similar to that observed in human emphysema patients, including those who have stopped smoking, but for whom the pulmonary pathology is nonresolving (22, 30). Frequently accompanying the histopathology observed in elastase-induced emphysema is an increased number of neutrophils and/or macrophages in the airways (3, 8–11, 19, 20, 28, 29, 31). These findings have given support to the idea that pulmonary insults that initiate emphysema result in the continual recruitment of inflammatory neutrophils and macrophages, cells that carry and secrete the elastolytic enzymes necessary to degrade elastin and subsequently disrupt lung architecture over time (2, 6, 8, 23). However, it is still unclear what drives these cells to initially infiltrate to the site of acute damage and what stimulates them to sustain the self-perpetuating and progressive pathology observed in chronic emphysema.
One feature of the elastase model that has hindered the study of the specific cells and molecular events associated with the onset of emphysema is the widespread inflammation and unpredictable localization and distribution of tissue destruction that results from solubilized elastase challenge (12). Given the heterogeneity in responses to elastase delivered indiscriminately throughout the lungs, defining the onset of emphysema as first a disease of the small airways or that of alveolar parenchyma has proven difficult (18). As such, it would be useful from a therapeutic intervention standpoint to know if responses to acute emphysematous injury arise from inflammation that radiates outwards from the small airways into attached acini, or if focal damage originates in parenchymal lesions and spreads to involve adjacent alveoli and respiratory bronchioles. In addition, the administration of solubilized elastase precludes the characterization of the cells directly involved with the development of pathology vs. bystanders. Developing a system to identify those cells which respond initially to discrete foci of elastase-induced damage should provide insight into critical cellular and molecular activation pathways involved in emphysema and would serve as a useful extension of the existing elastase model.
Here we describe a new model that permits the targeted delivery of elastase covalently coupled to bisacrylamide beads for the study of localized tissue destruction in the lung. This model also allows for the identification of cells directly involved in chemotaxis to the site of elastase deposition. In addition, this method has broader application as a means to deliver a whole range of proteins into the lung for the purposes of studying inflammatory and spatial thresholds in the pulmonary microenvironment.
MATERIALS AND METHODS
Protein-bead coupling protocol.
Three types of beads were generated for the purposes of this study: beads bound with active elastase (AE beads), beads bound with elastase inactivated by an irreversible inhibitor (IE beads), and control beads containing no protein (NP beads). Covalent linkage of elastase protein to 50- to 80-μm beads was achieved through coupling reactions containing 10 mg of bisacrylamide polymer beads functionalized with azlactone (Ultralink Biosupport, Thermo Scientific) and 2 mg of porcine pancreatic elastase (high-purity, Elastin Products). Beads and elastase were combined in 1 ml of a 0.6 M sodium citrate and 0.2 M tricine coupling solution at pH 8.5. The bead mixtures were gently rotated for 2 h at room temperature, after which the beads were pelleted by centrifugation at 1,200 g for 5 min and the coupling solution containing any unbound protein was removed. Unoccupied azlactone sites on the beads were then quenched with 1 ml of 1 M Tris-HCl at pH 7.5 with gentle rotation for 2.5 h at room temperature. To generate IE beads, the irreversible inhibitor N-(methoxysuccinyl)-Ala-Ala-Pro-Val-chloromethyl ketone (Sigma) was added to the quenching solution in a 50-fold molar excess over elastase for the final 30 min. NP control beads were generated by first quenching the azlactone sites with Tris-HCl prior to the addition of elastase in coupling solution. All bead types were subsequently washed for 15 min each with Dulbecco's phosphate-buffered saline (1× PBS), once with 1.0 M NaCl, and twice more with 1× PBS. Washed beads were suspended in 1× PBS and enumerated on a glass slide under a Nikon SMZ1500 stereo microscope prior to administration. Test and control beads were stored on ice and administered within 1 h following completion of the coupling procedure.
Assessment of coupling efficiency and elastase enzymatic activity.
SDS-PAGE analysis of enzyme eluted from 10,000 NP, AE, or IE beads was utilized as a semiquantitative method for determining the success of protein-bead coupling. A standard curve was generated by loading known quantities of elastase protein (Elastin Products). To verify the retention of enzymatic activity following elastase-bead coupling, and to confirm the loss of activity following inhibitor binding, 10,000 beads from each of the three bead types were mixed with 10 μg of bovine neck ligament elastin (Invitrogen) and incubated overnight at room temperature before SDS-PAGE analysis. All samples were heated at 95°C for 20 min in Laemmli sample buffer (Bio-Rad) containing 5% 2-mercaptoethanol prior to gel loading. Insoluble material was pelleted and the soluble fraction was loaded onto a 12% polyacrylamide gel (Mini-Protean TGX, Bio-Rad) and resolved at 200 V in the presence of 1× Tris/Glycine/SDS buffer (Bio-Rad). Gels were stained with 0.025% Coomassie Brilliant Blue R-250 (Thermo Scientific) with gentle agitation for 4 h. Gels were destained in a solution of 40% methanol and 7% acetic acid for 30 min, followed by immersion overnight in a solution containing 5% methanol and 7% acetic acid, and subsequently digitalized on a FluorChemQ imager (Protein Simple).
All experiments were performed using wild-type, 8- to 10-wk-old male C57BL/6J mice (Jackson Laboratories) under the approval of the Johns Hopkins University Animal Care and Use Committee. Mice were housed in a specific pathogen-free facility, continually provided filtered air and water, maintained on a 12:12-h light/dark cycle, and fed autoclaved food ad libitum. Mice were anesthetized by open-drop exposure to isoflurane (20% vol/vol in propylene glycol) and a single dose of 40,000 beads was well suspended in 50 μl of sterile 1× PBS and administered by placing the mice supine on a 60° incline board, extending the tongue to prevent swallowing, pipetting the volume onto the back of the oropharynx, and obstructing the nares to promote intratracheal aspiration of the bead solution (5). A similar protocol was followed for the administration of soluble challenges consisting of either 6 U of elastase dissolved in 50 μl 1× PBS, 6 U of elastase dissolved in 5% dimethyl sulfoxide (DMSO) containing a 50-fold molar excess of N-(methoxysuccinyl)-Ala-Ala-Pro-Val-chloromethyl ketone elastase inhibitor, or PBS and 5% DMSO vehicle controls.
At days 2 and 21 following soluble elastase challenges, or at days 2, 21, and 42 following bead administrations, mice were anesthetized with a lethal dose of avertin prior to tracheostomy. Lungs were inflated at a constant pressure of 30 cmH2O for 5 min with zinc-buffered formalin (Z-Fix, Anatech). The trachea was tied and the intact lungs were excised from the chest cavity and submerged in formalin for at least 48 h. After fixation, lungs were randomly cut into 2- to 3-mm-thick blocks, which were then embedded in paraffin. Five-micrometer sections were cut and stained with hematoxylin and eosin (H and E) or Masson's trichrome and imaged on a Nikon E800 upright microscope using the Spot Advanced camera and software (Diagnostic Instruments).
Quantification of tissue destruction and cellular infiltration.
To assess the magnitude of local tissue destruction around the beads, we designed a protocol to quantify the number of alveolar walls in the immediate vicinity of each bead. Beads for analysis were accepted if they met the following criteria. First, beads needed to be surrounded by respiratory parenchyma, and thus any beads residing completely or partially in airways were not included. Second, beads were not included if there were any other beads closer than three radii from a bead being analyzed. All beads meeting these criteria in 20–40 randomly designated fields (acquired at 20× magnification) from each mouse were analyzed by counting the number of alveolar walls that intercepted the edge of a circle surrounding the bead with twice the bead diameter (illustrated in Fig. 3A). Utilizing the same inclusion/exclusion criteria, a similar protocol was followed to quantify the number of mononuclear and polymorphonuclear cells within two bead diameters of 20 randomly sampled beads visualized at 40× magnification from each mouse. Emphysema scores and cell counts were graphed and analyzed by two-way ANOVA with Bonferroni posttests to compare all groups using Prism software (GraphPad).
Unstained paraffin-embedded lung sections were deparaffinized and rehydrated using 3 min washes in three changes of xylenes, once in 100% ethanol, once in 95% ethanol, and once in diH2O. Antigen retrieval was performed by submerging slides in 10 mM citric acid at pH 6.0 and heating to 90°C in a microwave for 10 min. Tissue sections were outlined with a hydrophobic barrier pen (ImmEdge, Vector Labs) and blocked for 20 min in 10% normal goat serum (Vector Labs). Sections were stained overnight with a polyclonal primary antibody raised in rabbit to porcine pancreatic elastase (1:500, Thermo Scientific). Sections were subsequently washed twice with 1× PBS, incubated for 1.5 h with an Alexa-Fluor 488-conjugated goat anti-rabbit IgG secondary antibody (1:500, Invitrogen), washed twice again with 1× PBS, and mounted with a medium containing 4′,6-diamidino-2-phenylindole (DAPI HardSet, Vector Labs) to visualize nuclei. Images were captured using epifluorescence on a Nikon E800 microscope using the Spot Advanced camera and software.
Development and delivery of elastase beads.
The conventional solubilized elastase challenge protocol was carried out to generate data for comparative analysis to the novel bead method and yielded results consistent with previous reports utilizing varying doses of elastase (9, 16, 21, 26). Intratracheal aspiration of 6 U of elastase solubilized in saline resulted in widespread lung inflammation marked by hemorrhage, cellular infiltration, and edema within the parenchyma of C57BL/6 mice 2 days following administration (Fig. 1). This inflammation gives rise to extensive destruction of alveolar tissue and the development of emphysema as evidenced by a drastic enlargement in airspace at 21 days post-elastase challenge. Inflammation, tissue destruction, and the resulting emphysema could all be prevented by pretreating the elastase with an irreversible inhibitor, N-(methoxysuccinyl)-Ala-Ala-Pro-Val-chloromethyl ketone, which covalently binds to the active site of elastase and blocks interaction with its elastin substrate (4). While this protocol has its utility for modeling the later stages of emphysema, the widespread damage and inflammation resulting from the administration of solubilized elastase makes it difficult to define the cellular and molecular mechanisms that are required to induce and regulate the progressive destruction of respiratory epithelium. Thus we devised this novel variation of the elastase model that would facilitate defining the mechanisms involved in the early induction of emphysema.
To generate focal emphysematous lesions, active elastase was covalently linked to bisacrylamide copolymer beads (active elastase or AE beads) that are functionalized with an azlactone group that enables formation of an amide bond when mixed with amine-rich proteins. Elastase-bead linkage was confirmed by SDS-PAGE, which suggested that coupling resulted in ∼1 μg/0.125 U of elastase per 10,000 beads or, assuming uniform coupling, roughly 1 pg of elastase per bead (Fig. 2A). Beads for which the azlactone sites were first quenched with a primary amine buffer, Tris-HCl, prior to the addition of elastase were devoid of protein (no protein or NP beads), indicating specific elastase coupling as opposed to passive enzyme uptake by the porous beads. The NP beads were also used to account for any nonspecific reactivity to the bisacrylamide beads alone. To control for responses to the porcine-derived elastase and to assess the requirement for active enzyme on the beads, elastase beads subsequently mixed with the N-(methoxysuccinyl)-Ala-Ala-Pro-Val-chloromethyl ketone inhibitor (inactive elastase or IE beads) were also generated. As a qualitative measure of enzymatic activity, AE, IE, and NP beads were incubated with elastin in vitro. Only the AE beads retained enzymatic activity, as demonstrated by the ability to degrade elastin, a function that was absent in both IE and NP beads (Fig. 2B).
Intratracheal aspiration of a single dose of 40,000 AE, IE, or NP beads (carrying ∼4 μg/0.5 U of total elastase) resulted in successful delivery of the beads into the small airways and parenchyma of the lung, a distribution that was readily observed in histological sections where the beads were labeled with an anti-elastase antibody and a fluorescently conjugated secondary antibody (Fig. 2C). Immunostaining also confirmed the attachment of elastase protein to the AE and IE beads, but not to the NP beads, at days 2 and 21 postadministration in vivo (Fig. 2C). The same pattern and intensity of fluorescence were also present in samples collected 42 days after bead challenge (data not shown). It is interesting to note that elastase also appears to be coupled to the internal matrix of the beads as indicated by the pattern of fluorescence observed with the anti-elastase antibody for both the AE and IE beads.
Quantification of elastase bead-induced alveolar tissue destruction.
To quantify the level of alveolar tissue destruction induced by the AE beads relative to IE or NP control beads, we developed a system for evaluating the loss of alveolar walls immediately surrounding the beads. In brief, a circle equal to two times the diameter of an individual bead was drawn around each bead that met specific inclusion criteria (see materials and methods), and the number of alveolar walls that intersected each circle was counted (Fig. 3A). Compared with the control IE and NP beads, AE beads were surrounded by significantly fewer alveolar wall intercepts at days 2, 21, and 42 following administration (Fig. 3B). These data indicate that AE beads induce significantly more tissue destruction in the local lung environment immediately surrounding the beads, but also that within this 42-day time frame the damage was not progressive in nature (Fig. 3B).
Pulmonary reactivity to elastase beads.
H and E staining of lung sections visualized at 40× magnification and containing IE or NP beads at day 2 revealed modest cellular reactivity at the site of bead deposition, whereas AE beads were surrounded by a robust and localized cellular infiltrate (Fig. 4A). By day 21, the cellular infiltration around the AE beads had waned, leaving behind focal emphysematous lesions reminiscent of those that arise from solubilized elastase challenges, but on a more localized scale (Fig. 4A). Little or no change was observed for IE or NP beads between days 2 and 21 postadministration, indicating the need for active elastase on the beads to achieve the observed cellular and focal emphysema-like phenotype (Fig. 4A). The quality and magnitude of the response at day 21 was also similar to that observed at day 42 for all bead groups (data not shown).
Analysis of the cellular profile surrounding AE beads in the parenchyma indicated the presence of a comparable level of mononuclear cells, including macrophages, within two bead diameters relative to the NP and IE beads at 2 and 21 days after bead instillation (Fig. 4B). In contrast, there was a significant increase of polymorphonuclear cells, which appeared to be exclusively neutrophils, within two bead diameters of AE beads at day 2, a response that had waned by day 21 and was absent altogether in close proximity to NP or IE beads (Fig. 4C). That AE beads evoked a localized neutrophil migration on day 2, while NP and IE beads were surrounded almost exclusively by mononuclear cells at both time points, indicates the need for active elastase to elicit neutrophils in the early stages following bead administration.
Development of elastase beads.
In this study we have documented the ability to deliver elastase protein to focal regions of the lung. While this approach could be used to deliver a wide variety of different proteins, we have validated the method using elastase, a protein commonly used in the generation of emphysema in animal models. In the conventional approach, the elastase is given either intratracheally or intranasally, and such delivery results in a heterogeneous pattern of tissue destruction. The unpredictable destruction likely results from the uncontrolled distribution of the liquid vehicle in which the elastase is dissolved. With the approach described in this study, although there is still an uncontrolled distribution of beads, the delivery of the elastase is now easily identified in discrete regions in close proximity to the beads. This change allows one to examine the immediate localized effects of elastase on inflammatory cell traffic and tissue destruction.
The first problem we needed to resolve with regard to this method was to find a bead with properties that would allow for the stable binding of enzymatically active elastase. After considering different materials, including Sepharose 4B beads (7, 15), we selected the bisacrylamide beads used in this study for their relatively uniform ∼50- to 80-μm size. We then modified the coupling procedure to optimize for the chemical linkage of elastase to the beads. This involved a number of trials varying the type, pH, and salt concentration of the coupling buffer with different concentrations of elastase. We also varied the time, temperature, and volume of the coupling reaction before settling on the conditions reported here.
In testing the efficiency and effectiveness of the approach, there were several concerns that we needed to address. These included the spatial distribution of the beads, the amount of elastase actually bound to the beads, the effect of control beads with no bound protein, whether an active form of elastase is needed for tissue destruction, and how long the bound elastase remained active. With regard to the spatial distribution of beads in the mouse lung, although we knew from the literature that chemicals instilled into the lungs by placing them on the back of the tongue and obstructing the nares would be widely distributed, we had no knowledge of how discrete-sized particles would disperse into the terminal airspaces (5). However, the results from this study showed that the 50- to 80-μm beads seemed to be distributed as broadly as chemicals in a liquid vehicle. Although we did not do sufficient sampling for a statistical analysis of the complete spatial distribution, we did take random 5-μm histological sections from all lobes of the lung. In these lobes examined, we consistently observed beads in all sections. From a simple mathematical consideration, if 40,000 beads were randomly dispersed in a 20-mm-high mouse lung that is cut into 4,000 5-μm sections, we should find 10 beads per complete transverse slice of the lung. In the real-world situation, where our sections were only from whole lobes and the distribution was not completely uniform, we would often find more or less than 10 beads per section. While not perfect, the instillation procedure clearly resulted in a widely dispersed distribution of beads.
With regard to the amount of elastase bound to the bead, there is only limited conjecture that we can make, since there are many unknowns. Although the procedure we used was designed to maximize the binding of elastase to the beads, we have no easy way to quantify how much active elastase is bound to the surface of the beads or the activity status of the elastase that appears to be present in the internal aspect of the beads. In addition, we still have no way to assess the contribution that the deeper-bound elastase makes to the cellular infiltrate and subsequent pathology. Nevertheless, even though we do not know how many fractional units of elastase are bound to each bead, we do know that the bound enzyme retains sufficient activity to degrade elastin (Fig. 2B) and to cause local tissue destruction (Fig. 3B), which was the whole objective of this novel experimental model.
The extent and composition of the cellular reactions induced by the beads at day 2 postinstillation revealed that in contrast to the NP and IE control beads that attracted only mononuclear cells, many with the morphological characteristics of macrophages, the AE beads induced a robust, mixed polymorphonuclear neutrophil and mononuclear cellular infiltrate (Fig. 4). Inasmuch as the magnitude and nature of the cellular accumulation around the AE and control beads recapitulated the profile of inflammatory cells commonly observed in the solubilized elastase system (3, 8–11, 19, 20, 28, 29, 31), we conclude that the bead-based system will be useful in defining the early cellular and molecular events that initiate emphysema. The observation that macrophages were attracted to the beads, whether or not the beads were associated with protein, probably reflects the general role that macrophages play in the walling off of foreign bodies (1). In addition, the presence of neutrophils at day 2 and not at day 21 suggests that the bead-associated elastase was likely inactivated at the latter time point, possibly due to the action of α-2-macroglobulin (26, 27).
The level of tissue destruction seen at 2 days was maintained for up to 42 days after the bead delivery (Fig. 3B), suggesting not only a lack of tissue repair, but also the absence of progressive damage. We originally anticipated that there might be continual destruction of the local tissue adjacent to the bead. That this was not observed suggests that a required threshold of activation was not achieved, potentially because the elastase became quickly inactivated, as has been shown for the more global whole lung elastase delivery (26, 27). This is an area worth further investigation.
From the histological analysis it was clear that not all the beads were found in immediate proximity to alveoli, as often individual or clusters of beads were located in larger airways. Since the beads had a nominal size of 50–80 μm, they are too large to wedge into single mouse alveoli. However, they were commonly found lodged in alveolar ducts or the short respiratory bronchioles in mice, and it was these beads that were analyzed for adjacent tissue destruction. The question of whether the beads seen in airways were there simply because they could not lodge in a stable airspace structure, or because the structure they initially wedged into became sufficiently damaged such that the bead moved, cannot be easily assessed. It is also possible that instilling fixative at 30-cmH2O pressure dislodged some beads during fixation. Nevertheless, despite this uncertainty, a majority of the beads were in fact observed in terminal airspaces, and these were the ones analyzed.
In summary, the work presented here clearly shows that it is possible to bind tissue-destructive proteins such as elastase to inert beads and instill them into the lungs. The potential to study localized tissue destruction and inflammatory cell recruitment induced by AE beads was demonstrated over a 42-day time period after administration. However, since it is likely that these beads will remain permanently in the lung, continuing analysis could have been done over an indefinite period for the life of the mouse. Although we only tested elastase-bound beads, theoretically the same procedure could be used for binding a variety of proteins to study local antibody or other pro/anti-inflammatory responses in the lung.
This work was funded by National Heart, Lung, and Blood Institute Grant HL-10342 (W. Mitzner).
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: J.M.C., A.L.S., and W.M. conception and design of research; J.M.C. performed experiments; J.M.C. and W.M. analyzed data; J.M.C., A.L.S., and W.M. interpreted results of experiments; J.M.C. prepared figures; J.M.C., A.L.S., and W.M. drafted manuscript; J.M.C., A.L.S., and W.M. edited and revised manuscript; J.M.C., A.L.S., and W.M. approved final version of manuscript.
We thank Xin Guo for assistance with the histological tissue processing.
- Copyright © 2013 the American Physiological Society