In vivo assessment of the pulmonary microcirculation in elastase-induced emphysema using probe-based confocal fluorescence microscopy

Abstract

Introduction: The alveolar capillary bed, which appears essential for the maintenance of alveolar septa, is altered in pulmonary emphysema. Until recently, techniques that allow its analysis in vivo in spontaneously breathing conditions were lacking. Fibered confocal fluorescence microscopy (FCFM) is a new technique that enables distal lung microstructures imaging in vivo. FCFM can be coupled with I.V fluorescein injection to image the pulmonary capillary network. The aim of this study was to assess the lung microcirculation in vivo using FCFM and I.V fluorescein in rats with experimental emphysema.

Results: In vivo pulmonary microcirculation imaging was possible in 7/7 elastase animals and in 6/7 controls. Using FCFM, intercapillary distances and alveolar facets diameters were found significantly higher in the elastase group compared with controls (49.5 vs. 41.8 µm p < 0.001, and 118.5 vs. 95.1 µm p < 0.001, respectively). Ex vivo mean interwall distance (MIWD) was correlated with the alveolar facets diameters measured in vivo (r_s = 0.65 ; p = 0.016).

Methods: 14 Sprague-Dawley rats were assigned to intratracheal instillation of porcine pancreatic elastase (n = 7) or saline (n = 7). The subpleural microcirculation was assessed using FCFM in spontaneously breathing rats, through a 2mm thoracic window using a continuous aspiration system, after I.V. injection of fluorescein-dextran. FCFM sequences were recorded and the image analysis was performed separately by two observers, blindly to the animal group. Fluorescence intensity (FI), maximal intercapillary distances, and alveolar facets diameters measured with FCFM were compared between groups, and to ex vivo lung morphometric measurements (MIWD).

Conclusion: FCFM allows the quantitative assessment of the microcirculation alterations due to emphysema in vivo.

Keywords: :

• animal study
• capillary
• chronic obstructive pulmonary disease
• endoscopy
• fibred confocal fluorescence microscopy
• microcirculation
• morphometry
• pulmonary alveoli

Introduction

Pulmonary emphysema is anatomically defined as the abnormal and permanent destruction of the distal airways up to the terminal bronchiole, with the destruction of the walls and no evident fibrosis.¹ The pathogenesis of pulmonary emphysema is complex, involving abnormal inflammatory response to inhalation of noxious particles and/or gases,² and protease/antiprotease imbalance.³ Recently, the disruption of a lung cellular and molecular maintenance program involving alterations of the pulmonary capillary bed was proposed as one of the mechanisms leading to emphysema.^4-7 Early ex vivo studies have described the loss of the pulmonary capillary bed in emphysematous lungs,⁸ and COPD patients have a reduced pulmonary capillary length and density.⁹ The quantitative assessment of emphysema has been largely studied with the development of software techniques for the analysis of CT or MRI images.^10-15 However, there is currently no mean to assess the modifications of the human pulmonary microcirculation in vivo with the progression of the disease, even if new technologies such as optical microangiography (OMAG) emerge for the study of pulmonary microcirculation in animals,¹⁶ or other organs, skin and retina, in humans.^17,18

Fibered confocal fluorescence microscopy (FCFM), also called probe-based confocal laser endomicroscopy (pCLE), is a new minimally invasive endoscopic technique that enables to image a living tissue at the microscopic level. The procedure makes it possible to explore the proximal bronchi as well as alveolar regions in real time during bronchoscopy in patients under topical anesthesia.^19-22 We previously demonstrated that the main endogenous fluorophore, observed with pCLE at 488 nm excitation wavelength, is the elastin, which is a major component of the distal lung interstitial network, and is present in the axial backbone of the alveolar ducts and alveolar entrances.^23,24 In this human study, the elastic fibers and the microvessels could be measured during the procedure, and the first morphometric analysis of the normal acinar and alveolar structures in vivo have been reported.²⁰ However, pulmonary capillaries cannot be observed without intravascular contrast.²⁰

The aim of this experimental study is to investigate whether FCFM is able to assess and quantify emphysema in vivo, on the basis of the loss of the pulmonary capillary network in elastase-induced emphysema.

Results

Observation of the subpleural microcirculation in vivo using FCFM and I.V. fluorescein

To achieve the subpleural microcirculation imaging, the miniprobe was introduced into the thorax through a 2 mm surgical opening in the right 5th intercostal space, and a customized aspiration system was used to hold the miniprobe in order to prevent the lung to collapse (Fig. 1). Special care was taken to avoid any mechanical damage onto the lung surface. After the smooth tip of the imaging probe was properly placed against the pleura, 0.8 ml/kg of 1% w/v fluorescein-dextran 150 kD was administered intravenously. The microscopic fluorescent imaging was then observed in real-time and recorded for further quantitative analysis. The imaging procedure was interrupted immediately if the lung collapsed

Figure 1. Fiberd confocal fluorescence microscopy (FCFM) associated to a continuous aspiration system to assess the subpleural microcirculation in spontaneously breathing animals. The FCFM probe is sealed into a cylindrical chamber that is connected to an aspiration device (vacuum pump) providing a continuous depression of 25 cm H₂O into the chamber. The extremity of the chamber, where the probe-tip emerges, is sealed to the lateral thoracic wall. The chest had been surgically prepared by removing the thoracic muscles layers up the internal intercostal muscle. The probe is then applied through the last muscle onto the pleural surface.

Before fluorescein injection, subpleural 488 nm FCFM microimaging of normal and emphysematous lung did not show any fluorescent signal. In contrast, a strong and clear intravascular signal could be observed immediately after the fluorescein-dextran IV injection, with a maximal intensity obtained after 2 min, which remained stable during 15 min after the injection (Supplemental Materials). This allowed subpleural microvascular imaging in 7/7 animals in the elastase group and 6/7 in the control group, where the procedure was unsuccessful in one case because of lung collapse.

Two main patterns could be observed, corresponding to subpleural capillary network and extra-alveolar vessels (Fig. 2), and, later during the imaging acquisition, the edges of the subpleural alveolar facets (Fig. 3), framing the peripheral alveoli.

Figure 2. The subpleural capillary network observed with fiberd confocal fluorescence microscopy (FCFM) after IV injection of fluorescein. (A) The subpleural capillary network is easily recognized, and extra-alveolar vessels can be observed (*). (B) The intercapillary distance was measured on selected frames displaying typical capillary network. The ten maximal intercapillary distances were selected for the quantitative analysis, and are displayed as median ± interquartile range for each animal. Inter-individual variability, assessed by the inter-quartile range of the medians was [30–55] µm in the control group, and [42–66] µm in the elastase group. (C) The intercapillary distance was higher in the elastase-treated animals (median = 49.5µm [IQR (40.8 - 62.4)]) than in controls (median = 41.8µm [IQR (32.7 – 51.9)]) (p < 0.001, Mann-Whitney test).

Figure 3. The subpleural alveolar facets could be observed with fiberd confocal fluorescence microscopy (FCFM) after a few minutes of examination. (A) The ten maximal diameters were selected for the quantitative analysis. (B) The 10 measures per animal are displayed as median ± interquartile range. Inter-individual variability, assessed by the inter-quartile range of the medians was [84–106] µm in the control group, and [104–152] µm in the elastase group. (C) The alveolar facets diameters were significantly higher in the elastase-treated animals (median = 118.5µm [IQR (97.2 - 145.9)]) than in controls (median = 95.1µm [IQR (81.6 - 106.8)]) (p < 0.0001, Mann-Whitney test).

The diameters of the alveolar facets were significantly higher in elastase-instilled animals than controls (median = 118.5 µm [IQR (97.2 - 145.9)] in the elastase group vs. 95.1 µm [IQR (81.6 - 106.8)] in the control group ; p < 0.0001, Mann-Whitney test) (Fig. 3). The measurements of the maximal intercapillary distance are illustrated in Figure 2. It was significantly higher in the elastase-treated animals than in controls (median = 49.5µm [IQR (40.8 - 62.4)] in the elastase group vs. 41.8µm [IQR (32.7 – 51.9)] in the control group ; p < 0.001, Mann-Whitney test).

The intravascular fluorescence intensity was normally distributed in both elastase and control groups, and was significantly lower in treated animals as compared with controls (mean ± SD were 764.5 ± 341 vs. 926 ± 258 arbitrary units; p = 0.004, unpaired t test, Fig. 4).

Figure 4. The fluorescence intensity (FI) was measured on ten frames representative of the capillary network, excluding frames displaying large extra-alveolar microvessels. (A) The 10 measures per animal are displayed as median ± interquartile range. Inter-individual variability, assessed by the inter-quartile range of the medians was [789–1120] arbitrary units (A.U.) in the control group, and [424–951] A.U. in the elastase group. (B) The mean FI was significantly lower in the elastase-treated animals (mean ± SD = 764 ± 341 A.U.) than in controls (mean ± SD = 926 ± 258 A.U.) (p = 0,004, unpaired t-test)

Ex vivo measurements of distal airspace

Examples of airspace enlargement from ex vivo microscopic slides are shown in Figure 5. The MIWD was higher in the elastase group compared with controls (44.8 vs. 38.3 µm) (p = 0.001, Mann Withney test) (Fig. 5). There was a good correlation between the MIWD and the mean diameter of the alveolar facets (r_s = 0.65, p = 0,016) (Fig. 6). No correlation could be found between the MIWD and the fluorescence intensity (r_s = -0.36 ; p = 0.22), and the mean maximal intercapillary distance (r_s = -0.24 ; p = 0.43).

Figure 5. The mean interwall distance (MIWD) was used to assess pulmonary emphysema ex vivo. It was significantly higher in the elastase group (B) (median = 44.8 [IQR (43.7 – 45.3)]) than in controls (A) (median = 38.3µm [IQR (36.9 – 40.4)]) (p = 0.001, Mann-Whitney test).

Figure 6. The size of the subpleural alveolar facets diameters was correlated with the mean interwall distance (r_s = 0.65 ; p = 0.016, Spearman correlation test).

Discussion

Historical ex vivo studies have demonstrated the alteration of the pulmonary microcirculation during emphysema. Liebow first described in 1959 the loss of the capillary bed in the human emphysematous lung.⁸ His findings were further confirmed by Reid and Heard,²⁵ who observed a decrease in the arteriolar branching in emphysematous tissue, in relation to the severity of the parenchymal destruction. Later, unbiased morphometric methods ex vivo showed the decreased length and length density of the capillary network in emphysematous patients.⁹

In the present study, the use of fibered confocal fluorescence microscopy (FCFM) with an exogenous fluorophore allowed to assess the lung microcirculation changes in vivo, in spontaneously breathing animals. The technique was able to accurately assess the alteration of the subpleural microvascular network in elastase-induced emphysema. These data were consistent with ex vivo morphometric analysis based on the mean interwall distance measurements. The FCFM measurements from the control group showed a median maximal diameter of the alveolar facets of 95 µm [IQR (82 – 107)]. These results are similar to the findings of a previous study using intravital laser confocal microscopy in vivo.²⁶ Using FCFM the alveolar facet diameters appear higher than the alveolar diameter usually found in rats ex vivo (70 – 80 µm).^27,28 An explanation could be that the published alveolar diameter is derived from averaged measurements per field of view, whereas our results are based on subpleural maximal alveolar diameters. It is admitted that histostereology is the gold-standard for the assessment of pulmonary anatomy at the microscopic level,²⁹ but histostereology relies on ex vivo and time-consuming tissue preparations. Our results showed that the alveolar facet diameter was correlated with the ex vivo morphometric analysis, therefore validating the technique to assess emphysema in vivo.

Compared with controls, in vivo exploration of emphysema by FCFM showed a 25% increase of alveolar facet maximal diameter, as well as a smaller (10%) but significant increase of the maximal intercapillary distance. This is may be due to the fact that the subpleural capillary bed is scarcer as compared with more central parts of the lungs. Along the same line, no correlation could be found between the mean interwall distance measured on histological preparations and the intercapillary distance in vivo. This probably also reflects the heterogeneity of the emphysema-related pulmonary lesions, which could not be assessed through a small thoracic window. Heterogeneity of emphysema may be particularly pronounced in the elastase intratracheal instillation model, with predominant distribution in the central parts of the lungs as a result of differences in elastase concentrations between central and peripheral airways, as compared with other experimental models.³⁰

The fluorescence intensity was significantly lower in the emphysematous rats compared with controls, corresponding to a looser microcirculation network. This result was consistent with the increased intercapillary distance in emphysematous rats. Even if the fluorescence intensity assessment was successfully used with the FCFM associated software in other animal and human studies,^31-33 another parameter, more accurate, to assess the lung microcirculation would be blood flow modifications. Recent progress in intravital microscopy recently allowed to image changes in pulmonary blood flow. Hanna et al. used a closed-chest pulmonary window that allowed real-time images to be captured by fluorescence and multispectral absorption microscopy.³⁴ However blood-flow measurements in this study were not performed, as, if the impact of pressure artifacts on vessel morphology were limited by the technique, its consequences on blood flow could not be appreciated.

To our knowledge, this study is the first to focus on pulmonary microcirculation in emphysema in vivo. Several ex vivo animal studies on emphysema have analyzed the peripheral capillary alterations.^30,35 Whereas they have consistently found a decrease in the subpleural capillary density, they were limited by potential artifacts due ex vivo inflation of the lung,³⁰ which was not the case in our in vivo study.

Until now, studies on the pulmonary microcirculation had used intravital microscopy in healthy isolated perfused lung³⁶ and fluorescent microspheres²⁷ or intravital confocal microscopy and fluorescein contrast in acute respiratory distress syndrome in rats,^26,37 or in models of hypoxia or drug-induced vasoconstrictive responses in mice.³⁸ These studies relied on mechanical ventilation, which also could disturb the architecture of the subpleural microcirculation.³⁹ Recent development of intravital microscopy using two-photon microscopy (TPM) was successfully applied to the pulmonary microcirculation analysis.^40,41 The main advantage of this technique is the possibility to increase the depth of focus into the tissue. However, the technique has some limitations related to motion artifacts due to a slow acquisition rate. Using customized TPM system, respiratory gating, and post-acquisition image processing, Presson and al. showed that the pulmonary microcirculation could be imaged in living animals, without respiratory and cardiac motion artifacts, with a comparable precision as in pump-perfused animals.⁴¹

These techniques offer very high-resolution and stable images of the pulmonary microcirculation with limited motion-artifacts. However, they require high-technicity equipment and a long preparation time.⁴¹

FCFM has been previously used in two animal experimental studies on isolated perfused lung⁴² and in vivo in a model of acute lung injury.⁴³ Compared with these studies, our in vivo FCFM approach makes it possible to study spontaneously breathing animals, a condition that may be crucial in emphysema, even if the aspiration system may affect the in vivo morphometric measurements. In the future, this approach could also be applied to longitudinal studies, therefore allowing therapeutics assessment in emphysema. For instance, studies have shown that the inhibition of VEGF receptors causes emphysema in adult rats,⁵ a mechanism that could be targeted in vivo using specific fluorescent probes and confocal laser endomicroscopy.⁴⁴ FCFM could be used, for example, to assess the effect of therapies that would restore the VEGF-VEGFR-2 pathway for the treatment of emphysema in animal models.⁴⁵

Because of the fragility of the upper respiratory tract, the transbronchial acinar imaging was not possible using the confocal miniprobe dedicated to small animals. The transthoracic approach, however, can be applied to smaller animals, i.e mice, that are usually used in emphysema studies and other lung conditions. On the other hand, this limitation can also be overcome using a transbronchial exploration in bigger animal models. Along the same line, FCFM can be easily employed for human explorations, during bronchoscopy in spontaneously breathing patients under topical anesthesia.

Human studies using FCFM for distal lung explorations are still limited.²⁰ The technique has been recently used in a preliminary study in 13 patients, finding significant alveolar entrance enlargement and distortion or thickening of the remaining elastic fibers during emphysema.⁴⁶ Preliminary data using FCFM and IV fluorescein in humans also showed that imaging of pulmonary capillaries in vivo is possible,⁴⁷ in conjunction with elastin assessment. Along with these early results, our data indicate that FCFM could be used in the future for in vivo assessment of the capillary bed alterations in emphysema in humans.

In conclusion, this study demonstrates the possibility to assess the pulmonary microcirculation in vivo in spontaneously breathing small animals, and validates the technique for in vivo measurements of the distal lung structures. Even if it did not provide new fundamental data for the comprehension of the pathophysiological changes occurring in emphysema, or new therapeutic data, fundamental research on emphysema pathogenesis may benefit, in the future, from the possibilities of in vivo microvascular imaging using endomicroscopy. Future studies using FCFM may aim at describing and understanding other pulmonary conditions, such as interstitial lung diseases,^33,48,49 and new studies are also needed to evaluate the adaptation of the technique to longitudinal studies in emphysema or other experimental models focusing on vascular-related disorders.

Methods

Animals

Fourteen 14 week-old Sprague-Dawley male rats were used in this study. All animals were housed in a temperature-controlled room (21 ± 1°C.) and were maintained on a 12:12-h light:dark cycle with free access to water and food. Animals were assigned to intra-tracheal instillation⁵⁰ either of porcine pancreatic elastase (750 units/ kg; Elastin Products Company) diluted in 200 µL of physiological saline solution (n = 7), or of the same volume of physiological saline solution alone (n = 7). Elastase concentration was 130 IU / mg where 1 IU is the amount of enzyme able to catalyze 1 mg of elastin at pH 8.8 and 37°C. Elastase was prepared immediately before intratracheal instillation. All procedures were performed under general anesthesia with intraperitoneal injection of ketamine and chlorpromazine (50 mg/kg and 5 mg/kg, respectively).

Fibered confocal fluorescence imaging was performed 8 weeks after intratracheal instillation. All the animals were sacrificed immediately after the imaging procedure by lethal intravenous injection of 200 mg/kg thiopenthal, and the lungs were collected for histology.

All the animal procedures were completed in accordance to the institutional animal care and use committee (N-03-07-08/11/06-12).

Fibered confocal fluorescence microscopy procedure in vivo

The catheter-based confocal fluorescence imaging of the lungs was performed with a CellVizio^® Lab (Mauna Kea Technologies) in spontaneously breathing animals anesthetized by intra-peritoneal injection of thiopenthal (100 mg/kg). The confocal fluorescence imaging system uses an excitation laser source at 488 nm and enables the real-time observation of fluorescent structures with an image flow rate at 9 to 12 images per second. The lateral resolution is 3.5 µm within a 600 × 600 µm field of view (FOV). Before each imaging procedure the system was calibrated for the 1.4 mm diameter Proflex S-1400^® miniprobe (Mauna Kea Technologies), with the dedicated calibration kit according to the manufacturer recommendations.

Quantitative analysis of the subpleural vascular network using fluorescence endomicroscopy

Quantitative measurements were performed concurrently by two observers (MS, LT), blindly to the animal group (elastase or saline), and from the histological results. In order to assess the density of the microvascular network, the entire FCFM sequences were analyzed. At least one representative frame showing the characteristic subpleural capillary network (capillaries were recognized based on an averaged diameter of 10 µm), as well as one frame displaying alveolar facets were selected for the analysis by each observer for each animal. The diameter of the ten largest subpleural alveolar facets and 10 measurements of the longest maximal intercapillary distance were analyzed, on each frame, using the Signal Quantification Toolbox of the Medviewer^® software (Medviewer 1.1.1^®, Mauna Kea Technologies) (Fig. 7).

Figure 7. Examples of measurements of (top) intercapillary distances, and (bottom) alveolar facets diameters.

The intensity of the intravascular fluorescent signal was assessed using the software’s Signal Quantification Toolbox, with the lower and upper level thresholds of the look-up table set to 0 and 8,000, respectively. Fluorescence intensity (FI) was defined as the median of pixel intensities over the entire selected image. For FI analysis, ten representing frames, with the highest fluorescence intensity, displaying the subpleural microcirculation were retrieved from the video file of each animal, blindly to the animal group. Frames displaying vessels with a diameter of 25 µm or more were not considered for the FI analysis. Median FI values, facet diameters and intercapillary distances were compared between elastase-treated animals and controls. The high rate of acquisition of FCFM (9 frames/second) as well as a slight pressure of the probe tip onto the surface of the pleura during respiration allowed limiting motion artifacts. In order to minimize related pressure artifacts, the slightest pressure onto the lung surface was applied, resulting in a discontinuous contact between the lung and the probe-tip during the respiration cycles. For quantitative measurements, this allowed to select images with the minimal compression effect from the last frames before the contact is lost.

Ex vivo lung tissue processing and histology

Animals were sacrificed immediately after the imaging procedure and the respiratory tract including the trachea was collected. A catheter was carefully introduced into the trachea and tightly ligatured to avoid any air or liquid leak. Lungs were immediately fixed in an ethanol and acetic acid mix solution (Hydrosafe^®, Labonord) at a constant pressure of 25 cmH₂O for 7 d. Thereafter the lungs were kept in the same fixative until they were processed for histomorphometric analysis.

Lung lobes were separated and cut in 6 sagittal fragments, except the accessory lobe, which was saved for further experiments. The samples were then embedded in paraffin, cut into 5µm slices, and stained with hematoxylin-eosin-safran (HES).

Ex vivo histomorphometric measurements

One HES-stained slide from each block (i.e., 6 slides per lung lobe) was used for the morphometric analysis. The slides were scanned (MiraxScan^®, Zeiss) and observed with the dedicated software (MiraxViewer^®, Zeiss) at x58 magnification. The magnified slices were integrally photographed, and the ImageJ software⁵¹ was used to extract 1.89 mm² FOVs. Sixty FOVs per animal were considered for the morphometric analysis, and were distributed as follows: 1 FOV per slide from the right upper lobe (i.e., 6 FOVs), 1 from the right middle lobe (i.e., 6 FOVs), 3 from the right lower lobe (i.e., 18 FOVs) and 5 from the left lung (i.e., 30 FOVs). Fields of view displaying extra-alveolar microvessels and/or bronchial structures were not considered for the analysis.

The automated quantification of the mean diameter of the alveolar structure was performed with a custom-written software (Matlab^®, Mathworks). Briefly, the software drew 50 vertical and horizontal lines that intersected with the alveolar walls and ducts, on each of the 60 FOVs per animal. The mean length of the segments between the intersections, the mean interwall distance (MIWD),⁵² was then computed and compared between elastase-treated animals and controls.

Statistical analysis

The statistical analysis was performed with the GraphPad Prism^® software (GraphPad Software, Inc.). The unpaired t test and Mann-Whitney test were used when appropriate to compare the fluorescence intensity median values, intercapillary distances, alveolar facet diameters, and the MIWD values between animal groups. The Spearman test was used to correlate FCFM measurements and MIWD. Statistical significance was set at a p value < 0.05.

Acknowledgments

This study was supported by ADIR Association.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Supplemental Materials

Supplemental materials may be found here: www.landesbioscience.com/journals/intravital/article/23471