INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Optimal lung preservation and storage is an important first step in successful lung transplantation (1, 2). Because surfactant alterations have been suggested to contribute significantly to the pathophysiology of transplantation-associated lung injuries, a procedure that stabilizes the pulmonary surfactant system is required for optimal lung preservation (as reviewed in reference 1). Analysis of bronchoalveolar lavage (BAL) revealed impairment of biophysical surfactant properties after clinical (3) and experimental lung transplantation (4, 5).

Using electron microscopy, we previously showed that the amount of tubular myelin, the surface-active intraaveolar surfactant form, was decreased relative to the surface-inactive subtypes in lungs after ischemia and reperfusion (I/R) and preservation with modified EuroCollins (EC) solution (6). Although EC is still used by most transplant centers throughout the world (7), several other solutions have been reported to improve lung preservation (8). Using an extracorporeal heart-lung model of the rat, postischemic respiratory function has been shown to be improved when the potassium concentration of the EC solution was reduced (9). This complies with the generally accepted principle of avoiding high potassium concentration because of its injurious effects on surfactant-producing type II pneumocytes and endothelial cells (8). Further improvement of functional parameters have been reported with extracellular-type flush solutions (low potassium, high sodium) (8) of which Celsior (CE) has been shown to reduce microvascular permeability and to improve oxygen tension (10). We recently found that the amount of intraalveolar edema is reduced, oxygenation of the perfusate improved, and the volume of the parenchymal air space increased after preservation with CE compared with potassium-reduced (40 mmol) EC40 solution (11, 12). We proposed that these differences may be the result of the different potential of the respective flush solutions to stabilize pulmonary surfactant.

On the basis of these findings, the present study sought to test the hypothesis that differences in I/R injury after lung preservation with distinct flush solutions, i.e., potassium-reduced EC (EC40) and CE, result from differences in the stability of intraalveolar surfactant. We therefore investigated the cumulative effects induced by the whole sequence of transplantation-related events on surfactant ultrastructure and pulmonary function, which includes flush perfusion, cold ischemic storage, and subsequent reperfusion of the lung, rather than looking at the single event. To assess I/R injury, hemodynamic and respiratory data were recorded in an established isolated organ preparation (9), and edema formation was sterologically quantified by light microscopy (11, 13). Cell-type specific stainings were used to look for alveolar epithelial damage by Bauhinia purpurea lectin histochemistry (13, 14) and for vascular endothelial damage by immunohistochemical staining for von Willebrand factor (15). In order to investigate the integrity of distinct surfactant subtypes, electron microscopy was used, which proved to be an appropriate technique to quantify intracellular and intraalveolar surfactant in the organ (6, 16).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Fifteen male inbred Sprague-Dawley rats (Crl:CD; Charles River, Sulzfeld, Germany) with a mean body weight of 460 ± SD 33.2 g were anaesthetized intraperitoneally with pentobarbital (Nembutal 1 mg/kg body weight), intubated by tracheostomy, and heparinized via the vena cava inferior (100 IU). All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals published by the National Institute of Health (NIH Publication 85-23, revised 1985).

Lung Preparation and Perfusion

Operation and excision of the heart-lung block was performed as described recently (9). During surgery the animals were ventilated with room air at a tidal volume of 5 ml and a respiration rate of 40 breaths/ min with positive end-expiratory pressure (PEEP) of 3 cm H₂O. For optimal preservation of the ultrastructure, control lungs (n = 5) were fixed in situ immediately before excision (see FIXATION). Experimental lungs were assigned randomly to two groups (n = 5 per group) either preserved with potassium-reduced EC40 or with CE (IMTIX, Pasteur Merieux, France) (for compositions see Table 1). EC40 was supplemented with prostacyclin (6 µg/100 ml) (Epoprostenol; Flolan, Wellcome, Beckenham, UK), which has been reported to have a protective effect on pulmonary microcirculation and tissue (19). Lungs were flushed via the pulmonary artery with 20 ml of cold (4° C) preservation solution at hydrostatic pressure of 20 cm H₂O and stored for 2 h at 10° C. This was followed by a 50-min reperfusion with Krebs-Henseleit-buffer (8.0 ml/min at 37° C) containing bovine red blood cells (hematocrit of 38 to 40%) using a quattro head roller pump (Mod- Reglo-Digital; Ismatec, Zürich, Switzerland).

Functional Parameters

Hemodynamic and respiratory data were recorded during reperfusion as described previously (9, 11). Blood gas determinations and peak inspiratory pressure (PIP) were recorded every 10 min after the onset of reperfusion. The perfusate oxygenation (PO₂), defined as the difference between oxygen tension of the perfusate collected from the left atrium after oxygenation (PO_2ox) and of the deoxygenated perfusate of the preload pool (PO_2deox), was used to assess the capability of the lungs for gas exchange. After reperfusion the right lung was fixed for light and electron microscopy (EM).

Fixation

The right lungs were fixed by vascular perfusion for 15 to 20 min via the pulmonary artery at a hydrostatic pressure adjusted to 15 cm H₂O and at constant airway pressure of 12 cm H₂O. The fixative used was a mixture of 1.5% glutaraldehyde and 1.5% paraformaldehyde in 0.15 M HEPES buffer (300 mOsm/kg buffer osmolality at pH 7.35) (Merck, Darmstadt, Germany). At the end of perfusion fixation, the main right bronchus and pulmonary artery were tightly clamped and the lungs were stored in cold fixative for 5 to 7 d until further processing.

Sampling

Systematic random sampling of lung tissue performed at each sampling step is a prerequisite to yield samples that are representative of the whole organ and thus can be analyzed using stereological methods (18). Therefore, systematic uniform random samples were taken according to standard methods (20). By means of a tissue slicer the organ was cut into 10 to 12 horizonal slices of 3 mm thickness. Starting with a random number, one slice was collected for EM, the next for plastic, and the third for paraffin embedding. A transparent point grid with 11 × 11 points (distance between points, 10 mm) was then superimposed over the collection of slices to be used for EM. Whenever a point hit the surface of a lung slice, a 3-mm³ tissue block was excised defining the point as the lower left corner of the tissue block. By this method five to seven blocks were obtained from each single lung.

Immunohistochemistry and Lectin Histochemistry

Lectin histochemistry was performed on paraffin sections of the whole slices to evaluate injuries of the alveolar epithelial wall using the biotinylated form of the type I pneumocyte-specific marker Bauhinia purpurea lectin (BPL) (Vector Laboratories, Burlingame, CA) as previously described (14). For immunohistochemistry, dewaxed sections were treated with 0.05% pronase (for 15 min at 37° C) prior to incubation (1 h at 37° C) with anti-von Willebrand factor antibody (Dako, Glostrup, Denmark) at a dilution of 1:500 to look for vascular endothelial damage (15).

Tissue Processing

Processing of the tissue blocks for stereological analysis was performed as described earlier (18). In brief, tissue blocks were osmicated in 1% OsO₄ in 0.1 M cacodylate buffer for 2 h and stained en bloc overnight with semisaturated aqueous uranyl acetate to stabilize phospholipids. For histologic analysis, samples were dehydrated in an ascending series of alcohols, and whole tissue slices were embedded in a methacrylate resin (Technovit 7100; Kulzer, Heraeus, Germany) as described recently (11). Sections 1 µm thick were cut from each slice (three to four per lung) using a Reichert Supercut 2050 and stained with eosin-hematoxylin. For ultrastructural analysis the tissue was dehydrated through a graded series of acetone and transferred to Araldite. Five tissue blocks were chosen at random and ultrathin sections were cut from each block and counterstained with lead citrate in an Ultrostainer (Leica, Bensheim, Germany). Quantitative and stereological analysis was performed using a Leitz Laborlux 11 (Leitz, Wetzlar, Germany) and a Zeiss Axioskop light microscope equipped with a projection tube and using an EM 900 (LEO, Oberkochen, Germany) for transmission electron microscopy.

Stereological Analysis

The sections were analyzed by established stereological methods using a systematic quadrats subsampling scheme to generate test fields over the whole section distributed in a systematic random fashion (20). All parameters were determined by means of point and intersection counting, the latter being introduced as a novel approach to evaluate the volume of the phospholipid layers composing loosely arranged surfactant forms such as tubular myelin and vesicular forms. As point counting reflects the biochemically determined amount of phospholipids only in the condensed surfactant forms, i.e., lamellar and multilamellar bodies, the phospholipid amount of surfactant forms with loosely arranged phospholipid membranes may be overestimated. In contrast, by intersection counting the surface of the phospholipid layers was determined and multiplied with the thickness of the layer to obtain an estimate of the absolute volume of the lipid layers composing these forms.

For light microscopy an eyepiece containing an integration plate (25 points) was used at a final magnification ×100 to determine the volumes of parenchyma and nonparenchyma. To evaluate the extent of lung edema a grid (81/12/4 points) was projected into the image level of the microscope at a total magnification ×1,000. For electron microscopy a coherent multipurpose test system with 624 test points and 312 test lines was applied on micrographs at a final magnification ×17,500.

The following sterological parameters were estimated: volume densities (by point counting) (1) of tubular myelin (tm) V_vtm and (2) unilamellar vesicles (ul) V_vul referred to total intraaveolar surfactant forms; total volumes (by intersection counting) of (3) tubular myelin Vtm, which was distinguished into luminal Vtm_lum and epithelial Vtm_epi according to the location of tm in contact with or without contact to the alveolar epithelium, (4) unilamellar vesicles Vul, which were classified into small (s-ul: 0.05 to 0.3 µm) and large (1-ul: > 0.3 µm) forms according to size-associated differences in surface activity (21); the total volumes (by point counting) of (5) lamellar bodies (lb) Vlb, (6) type II pneumocytes (pII) VpII, and (7) intraalveolar edema (ed) Ved. Additionally, the side dimension of the tm lattice (mesh width) was estimated by determination of the barrier thickness according to Weibel (22).

A mean of 486 ± SD 228 points and of 525 ± SD 236 intersections with intraalveolar surfactant subtypes were counted per individual lung.

Statistics

Data referring to individual lungs are given as discrete values obtained according to standard stereological formulas. Mean values are given ± SD unless otherwise indicated. Differences between the experimental groups were tested for significance by the nonparametric Mann-Whitney rank sum test or by Kruskal-Wallis one-way ANOVA on ranks. Provided that normality and equal variance were given, parametric one-way ANOVA was used followed by post hoc multiple comparisons (Tukey test). All statistical analyses and graphic presentations were performed using the SigmaStat 2.0 and Sigmaplot 3.0 software programs (Jandel Scientific, Erkrath, Germany); p values < 0.05 were considered to be significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Functional Parameters

After 10 min of reperfusion, perfusate oxygenation PO₂ was significantly higher (p = 0.008) in lungs preserved with CE (142.1 ± 16.4 mm Hg) than in those treated with EC40 (49.2 ± 22.3 mm Hg) (Figure 1). After 50 min oxygen tension had decreased in both groups (CE: 126.0 ± 32.4 mm Hg; EC40: 38.5 ± 18.4 mm Hg). Peak inspiratory pressures (PIP), were higher in lungs preserved with EC40 (14.0 ± 3.1 cm H₂O) than in those treated with CE (11.4 ± 2.1 cm H₂O), although at a level of significance of p = 0.056 only. In most lungs PIP values stayed constant throughout the reperfusion period (Figure 1). Because tidal volume and PEEP were kept constant in our experimental settings, an increase in PIP indicates a decrease in lung compliance.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Oxygen tensions and peak inspiratory pressures measured after 10 and 50 min of reperfusion in I/R lungs preserved with either EuroCollins (EC) or Celsior (CE). Boxes comprise the 25th and 75th percentiles, mean values (dotted lines) and median values (solid lines) ± SD. *p < 0.05 determined by Mann-Whitney rank sum test.

Qualitative Structural Findings

Histologically, intraalveolar edema and edematous swellings of peribronchovascular cuffs were found in I/R lungs, whereas the controls were almost free from edema formation (Figures 2a-2c). In both groups vascular endothelial cell-specific immunostaining for von Willebrand factor showed that vascular endothelial cells smoothly adhered to their basal lamina (Figures 2d-2f). Ultrastructural changes in the blood-gas barrier were rarely seen and type I epithelial cell-specific lectin histochemistry did not reveal any gaps in the alveolar epithelium in any of the groups studied (Figures 2b and 2c). However, mitochondrial swellings in type II cells, edematously swollen type I pneumocytes, and areas of fragmented cells or denuded basal lamina sometimes occurred after preservation with EC40, whereas none of the changes were found in lungs preserved with CE. In the EC40 group atelectasis was found to be more prominent than in lungs treated with CE and in the controls (Figure 2b). In control lungs small and narrow lattices of tm were found close to the epithelium (Figure 3A), whereas after I/R larger and wider lattices of tm without any contact to the alveolar wall appeared in the alveolar lumen (Figure 3B). Vesicular surfactant forms were either small (0.05 to 0.3 µm) and uniform or large (> 0.3 µm) and heterogenous. Both subgroups were evaluated separately according to results from density gradient centrifugation of bronchoalveolar lavages showing size-associated differences in surface activities of these forms (21).

View larger version (129K): [in this window] [in a new window]   Figure 2.   Lectin histochemistry and immunohistochemistry to detect epithelial and endothelial injury in paraffin sections of control lungs (a, d ), and I/R lungs either preserved with EuroCollins (b, e) or with Celsior (c, f  ). (a-c) Type I cell-specific staining with Bauhinia purpurea lectin (arrowheads). (d-f  ) Vascular endothelial cell-specific staining for von Willebrand factor (arrowheads). Asterisks indicate edematous peribronchovascular cuffs in b and c compared with control in a and atelectatic alveoli in c. Magnification in a, b, and c: ×204; in d and e: ×409; in f  : ×1,022.

View larger version (102K): [in this window] [in a new window]   Figure 3.   Electron micrographs showing intraalveolar surfactant subtypes in control lung (A) and in I/R lung (B) preserved with EuroCollins. The side dimensions (mesh width) of tubular myelin (tm) were determined by evaluation of the barrier thickness (inset in A). Alveolar lumen (a), capillary (c), blood-gas barrier (arrowhead ), endothelial cell (ec), multilamellar body (mb), and unilamellar vesicles (uv).

Quantitative Structural Findings

In control animals tubular myelin formed 10.6 ± 4.5% (CE: 13.5 ± 7.6%; EC40: 12.0 ± 6.7%) of total intraalveolar surfactant, unilamellar vesicles (small and large) formed 43.3 ± 11.9% (CE: 40.7 ± 13.0%; EC40: 28.9 ± 3.0%) and multilamellar forms of variable phospholipid composition and size formed 46.1 ± 13.2% (CE: 45.9 ± 13.4%; EC40: 59.2 ± 7.9%). After I/R these relative amounts remained unchanged except for lungs preserved with EC40 where the volume density of unilamellar vesicles (small and large) significantly decreased (p = 0.029) in favor of multilamellar subtypes.

In the controls almost all tm was found in contact with the alveolar epithelium (Vtm_epi) continuous with the alveolar lining layer (Table 2 and Figure 3A). After preservation with CE and EC40 a significant amount (p = 0.008) of tm was displaced into the alveolar lumen (tm_lum) (Table 2 and Figure 3B). Compared with the controls, the luminally displaced tm was characterized by a significant increase in mesh widths (p = 0.023), indicating a loosening up of the lattice structure (Table 2 and Figure 4). In EC40-preserved lungs Vtm_epi was significantly reduced (p = 0.018) compared with the other groups, whereas equivalent amounts of tm_epi, as in the controls, were found in lungs preserved with CE, which corresponded to approximately 50% of their total tm pool. The tm lattice of this epithelial subfraction has similar mesh widths in all groups and was therefore considered to be the functional surface-active tm (Figure 3A, inset).

View larger version (15K): [in this window] [in a new window]   Figure 4.   Side dimensions (mesh width) of tubular myelin lattices after I/R versus control. In the controls almost all tubular myelin was found in close contact with the alveolar epithelium (Coepi), whereas after I/R the luminally dislocated fractions of tubular myelin (CElum, EClum) displayed wider lattices. Boxes comprise the 25th and 75th percentiles, mean values (dotted lines) and median values (solid lines) ± SD.

The total phospholipid amount of both, tm and unilamellar vesicles, was increased after I/R and the ratios of the total of small (ul) to large (tm) surfactant aggregates remained unchanged (Table 2). Subfractionation of the ul revealed a higher amount of phospholipids in the large ul (l-ul) than in the small ul (s-ul) which, according to Gross and coworkers (21) were considered to be the "pure" surface-inactive form. Compared with the controls the ratios of Vs-ul and Vtm_epi revealed a significant increase after treatment with EC40 (p = 0.034), which was rather due to the decrease in tm_epi than to a rise in s-ul (Table 2).

Although the volumes of type II pneumocytes (VpII) were in the same range, slight reduction in the volume was observed in I/R lungs. This may be the result of surfactant release during reperfusion, since it was accompanied by decreased volumes of lamellar bodies (Vlb) compared with that in the controls (Table 2).

The volumes of intraaveolar edema varied considerably and were significantly increased (p = 0.042) in the I/R groups compared with those in the controls (180 ± 114 mm³) (Figure 5). The highest volume of intraalveolar edema was measured in lungs preserved with EC40 (EC40: 916 ± 673 mm³; CE: 482 ± 207 mm³) corresponding to 8 to 37% of total parenchyma in EC40, 8 to 17% in CE-preserved lungs, and approximately 3% in the controls. The volumes of the parenchymal air space (Figure 5) were significantly (p = 0.009) reduced after I/R, with lowest values found in the EC40 group (EC40: 2,634 ± 788 mm³; CE: 3,344 ± 941 mm³; Co: 5,186 ± 714 mm³), which is in accordance with the appearance of atelectatic regions predominantly occurring after preservation with EC40 (Figure 2b).

View larger version (20K): [in this window] [in a new window]   Figure 5.   Volumes of parenchymal air space and intraalveolar edema (mean ± SD) in control lungs (Co) and lungs after I/R and preservation with either Celsior (CE) or EuroCollins (EC) solutions. Evaluation was performed at the light microscopy level on plastic-embedded whole lung transects. *p < 0.05 versus control determined by Kruskal-Wallis ANOVA followed by post hoc multiple comparison (Turkey test)

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The surfactant system is in continuous flux. Removal of surfactant material from the surface-covering monolayer is part of conversion and clearing processes occurring in the alveoli (24). In several animal models of lung injury the ratio of surface-inactive small (SA) to surface-active large (LA) surfactant aggregates, corresponding to unilamellar vesicles (ul) and tubular myelin (tm), respectively, was increased (1). Increased SA/LA ratios were reported from patients with ARDS and after clinical lung transplantation, suggesting an increase of intraalveolar surfactant conversion (1). These findings correspond to ultrastructural alterations in intraalveolar surfactant reported in a previous study of our group (6). Comparing different preservation solutions with respect to oxygenation and alveolar architecture, we proposed different protective potentials of flush solutions for the pulmonary surfactant after ischemia and reperfusion (I/R) (11). The present study was conducted to test this working hypothesis.

In order to avoid extensive I/R damage of the blood-gas barrier, which could indirectly increase disturbances at the surfactant level by serum protein-associated inactivation (25), EC solution was applied in its improved reduced-potassium modification and with beneficial supplementation of prostacyclin (19). I/R injury was documented to occur in the rat lung preserved with conventional EC solution already after short periods of ischemia (9). The improved protection potential of CE solution was described earlier (10). Compared with previous studies (9, 13) reduced amounts of intraalveolar edema that were similar in both groups were associated with an improved preservation of the alveolar epithelium as shown by lectin histochemistry with a type I cell specific marker (14). Staining for the von Willebrand factor, a vascular endothelial-cell-specific marker (15), did not reveal any gross damage to the vascular endothelium, indicating that the swellings of the peribronchovascular cuffs after I/R were related to reversible hydrostatic fluid accumulations. Preservation associated differences at the functional level, i.e., oxygen tension and peak inspiratory pressure, both of which were improved in lungs preserved with CE, paralleled the improvement of surfactant stability observed in the CE group. This indicated primary effects of the preservation solution at the surfactant level rather than secondary effects induced by endothelial and/or epithelial damage and the subsequent presence of edema fluid in the alveoli. This proposal is further supported by the fact that alveolar collapse was much more prominent with EC40 than with CE, which contributed to the significantly lower volume of the parenchymal air space in the EC40 group.

It is generally agreed that the surface activity of the pulmonary surfactant is structurally related to its close contact with the aqueous hypophase, separating surfactant monolayer and alveolar epithelium (23). Luminal displacement of tm, however, as found in lungs after I/R, was not seen in control lungs and therefore has to be considered as disintegration of surface-active surfactant phospholipids from the monolayer, which was further indicated by the increase in mesh width of luminally displaced tm. The relative intraalveolar surfactant composition was similar to the amounts determined by others (16). However, the high fraction of displaced tm (> 50%) in lungs after I/R suggests that clearing processes cannot cope with this load of material leading to the accumulation of tm in the alveolar lumen.

The question arises as to whether this rise in tm observed after I/R results from an increase in the rate of surfactant secretion by type II pneumocytes, which have been assumed to be responsible for persistent impairment of biophysical surfactant properties after lung transplantation (2, 3). Flush perfusion with EC might not be as effective for rat lungs as for other species (26) which may be reflected by the greater abundance of mitochondrial swellings in type II cells after I/R and preservation with EC40 compared with CE. This may explain the previously observed decrease in the relative amount of tm after preservation with conventional EC (6), which contains high amounts of potassium (115 mmol K⁺) shown to be disadvantageous in rats (9). The slight reduction in type II cell volume measured after I/R corresponds to reduction of lamellar body volume and can be considered to be a consequence of phospholipid secretion rather than to be a sign of type II cell shrinkage. Prostacyclin was reported to enhance stretch-induced surfactant secretion in type II cells (27), which may be the cause for the increased amounts of tm and multilamellar bodies measured in the alveoli after preservation with the prostacyclin-containing EC40 solution. CE solution contains calcium, which has been implicated as a cellular second messenger in surfactant secretion (28) and may therefore be considered to induce the rise in intraalveolar surfactant in CE-preserved lungs. However, the effect of calcium on surfactant secretion was shown to be mediated from intracellular rather than from extracellular stores (29). It appears more likely that calcium may interfere with the formation of intraalveolar surfactant and its organizing components such as surfactant protein A (SP-A), which was shown to be calcium-dependent (30).

It has been suggested that SP-A stabilizes large surfactant aggregates and that changes in SP-A/phospholipid interactions might be involved in rapid adsorption of tm to the air-liquid interface (31). Increasing the SP-A-to-lipid ratio tended to increase the dimension of the mesh width (32) which, in the present study, were found to be in the range of freeze-fractured material (29 to 54 nm) (33). Because wider lattices were also seen in the epithelial fraction of tm after I/R, it is suggested that the loosening up of these structures precedes detachment and can therefore be considered as an early indicator of disturbances at the SP-A/phospholipid level.

SP-A was shown to inhibit large aggregate conversion (31), whereas increased SA/LA ratios and decreased levels of SP-A were reported after clinical (3) and experimental (5) lung transplantation. Taking into account that mainly s-ul constitute the surface-inactive catabolic fraction of intraalveolar surfactant (21) and the tm_epi represents the functional component of the tm pool in the alveoli, we found increased ratios of Vs-ul/Vtm_epi, which is in accordance with findings by other groups (1). In EC40 lungs the higher ratio is due to a decrease in tm_epi rather than to a rise in s-ul, and the opposite is true for lungs treated with CE. This indicated that both, surfactant adsorption to the air-liquid interface as well as metabolic surfactant conversion, can be influenced by lung preservation. Several factors are proposed to cause alterations in the intraalveolar surfactant microstructure. Oxidation was shown to impair the biophysical properties of surfactants in vitro affecting the unsaturated lipid components and surfactant protein A (SP-A) (34). In addition to the above proposed beneficial supplementation with calcium, the composition of CE (histidine, mannitol, lactobionate, and reduced glutathione) is targeted at inactivating free radicals, which has been speculated to imitate oxidative injury (10). This further supports our hypothesis that an initial imbalance between phospholipids and surfactant proteins is induced by I/R and subsequently leads to irreversible alterations of intraalveolar surfactant and lung function.

This implication is qualified by the observation that SP-A knockout mice, which form a negligible amount of tubular myelin, have apparently normal lung function (35). Qualitative and quantitative evaluation of surfactant proteins by immunoelectron microscopy, monitoring of oxygen radical formation, and assessment of respiratory parameters in lungs after I/R are necessary to further elucidate the initiation of these processes.

In summary, we have shown that improved lung function during the reperfusion period is associated with beneficial effects of lung preservation on the stability of tm lattices. These observations suggest that the well established effects of I/R on phospholipids and SP-A might be a consequence of early surfactant disturbances at the phospholipid/surfactant protein level that affect tm ultrastructure and attachment to the hypophase. Thus, increasing the potential of lung preservation procedures to stabilize the integrity of endogenous pulmonary surfactant appears to be an important aspect for further improvements in lung transplantation.