Ultrastructural Basis for Pulmonary Capillary Permeability to Autologous Plasma Proteins and to Exogenous Proteinic Tracers

Ultrastructural Basis for Pulmonary Capillary Permeability to Autologous Plasma Proteins and to Exogenous Proteinic Tracers

PuLMONARY CAPILLARY FUNcrION Ultrastructural Basis for Pulmonary Capillary Permeability to Autologous Plasma Proteins and to Exogenous Proteinic Trac...

1MB Sizes 0 Downloads 3 Views


Ultrastructural Basis for Pulmonary Capillary Permeability to Autologous Plasma Proteins and to Exogenous Proteinic Tracers * f. Bigrwn. M.D.; F. Jaubert. M.D.; and M.

e. Jaurand, .'.I.e.

tion of autologous plasma proteins in lung tissue by immunoelectron microscopy and a kinetic study of the transfer of intravenously, intrapleurallv and intratracheallv injected heterologous proteins in anesthetized

adult male rat. (Charles River), weighing 250·300 gm. MATERl.oU" ~l>:TIlODS A"D RESULTS

Localization of Autoloi!ous Plasma Proteins

filtration of plasma proteins from the blood to A noonal the pulmonary lymphatics, as well as to the alveoli,

has been demonstrated by various physiologic and biochemical studies. However, there were some unanswered questions about the morphologic aspects of this filtration such as: the location and concentration of various plasma proteins in the interstitium; the routes, either Intercellular junctions or pinocytic vesicles.' by which macromolecules cross the endothelial and alveolar epithelial membranes; the occurrence of a mono- or bidirecttonal transport. An ultrastructural study of these major points Itas been undertaken, involving a static Iocaliza-

- - - - '-:C--O-=:--

..From the Cllnique de Pneumo-Phtistologfe, Hopital Laennee, Paris, France.

~:~~r:;, 9%Ur~:rJ::ifC~·x,JF~~n, C.Il.I.C. 40 Av. de

Autologous Serum Albumin {69.000 daltons ), leG (160,000 daltons}, Fibrioogen (3-to,OOO daltons I and IgM (950,000 daltons ) were localized in fixed rat lung by means of antibodies labelled with peroxidase. The different methods used for lung fixation and immunoeIectron microscopy have been extensively described.s-" Results of detection of these four plasma proteins in the blood-air barrier tissue were roughly similar for albumin and I~ on the one hand, and on the other, for fibrinogen and I!Z~f. Antirat albumin and anlirat IgG antibodies. Xumerous pinocvnc vesicles. totally or partially filled with albumin and IgC, were found in the capillary endothelium often opening to the cell surface (Fig I}. Some interendothelial spaces were also found positive. hut only with antirat albumin antibodies with occasional unstained portion (Fig 2). The

of both proteins was important at the level of basement membranes. Albumin and I~ were also rf'H'aled


FIGURE 1. Rat lung, immunocytochemical revelation of albumin (airway fixation. glutaraldehyde } The capillary endothelium (En) appears with a positive labelling of the plasmalemma, of an interendothelial space (-,to) and vesicles (v ): The basement membrane (B) is Ii~htly stained (C: capillary lumen). (original magnification X 50,00.». FIGURE 2. Rat lung. immunocvtochemical revelation of fibrinogen (immersion fixation, glutaraldehyde). The capillary endothelium (En) appears with a slight positive labelling of the plasmalemma and of vesicles (v) membrane. The interendothelial spaces and the basement membrane (B) were practically unstained (C: capillary lumen; A: alveolus: EpI: epithelial alveolar cell type ll. (original magnification X 50,000). FIGURE 3. Rat lung, revelation of anti-Hftl' antibodies (immersion fixation, glutaraldehyde). The specific IgG is revealed in the capillary lumen (C) in the basement membranes (B) within an iuterendothelial space (-,to ), and in some vesicles ( v l. The alveolar epithelial cell type 1 t Ep Il surface is heavily stained (RBC: red blood cell: A: alveolus I (original magnification

X 50,000),




at the alveolar surface.a :\0 interepithelial space was stained, hut positive vesicles were observed within the alveolar epithelial type 1 cells. A.ntirat fihrino~(,fl ami antiral IgM antibodies, Xumerous endothelial vesicles were also stained, though less densely and mostly only at the surface of their membrane, their lumen appearing empty (Fig 2). In contrast. the interendotheltal spaces. the basement membranes and the alveolar epithelium were unstained. Localization of .t\ntir>{'roxida~i' Antibodies. Adult rats were immunized with horseradish peroxidase (HRP type 6, RZ 3.0, Sigma Chemical Co, St. Louis, Mo). On day 0, they received 5 mil: of HRP injected into the footpads with Freund's complete adjuvant. On days 20. 21, 22, they received a booster intramuscular injection of 2 mg HRP alone. Anti-HRP antibodies were controlled and identified as I~ by the Ouchterlony method and immunoelectrophoresis. After this procedure. lungs of anesthetized animals were 6xed with glutaraldehyde either by immersion or by tracheal infusion.s Cryostat 50 ~ thick sections of axed lung parenchyma were incubated with 0.5 mg/ml of HRP in cacodylate huffer at 4-C and thereafter rinsed :3 times in the same buffer for 1 hr 30. H HP remaining specifically hound in the tissue was revealed by the diaminobenzidtne IhO~ mixture.a Control.. carried out without prior incubation with HRP showed only the peroxidatic activity of hlood cells. 111is method enabled easv localization of IgG in the lung: tissue with more obvious results than those obtained with antirat IgG antibodies. Moreover, some interendothelial spaces appeared stained, whereas there was no staining when usinK antirat IgG antibodies ( Fig 3). Kinetic.. of Heteroiogou» Proteins

Two kinds of tracers were used in the rat: human albumin (69.lXKldaltons land HRP (-10,000 daltons ):

One 011of human purified albumin in saline solution (4 percent, 40 mg per animal} was injected intravenously to

rats, which were sacrificed at 5 min, 20 min, and 60 min. Blocks were processed for immunoelectron microscopy with specific antihuman albumin antibodies, after absorption against normal rat serum. Human albumin was detected in the capillary lumen and as early as 5 min in endothelial pinocytic vesicles. The interstitial staining was null or very light. Interendothelial spaces were not stained or partially stained towards the luminal portion (Fig 4). However, this method was not reliable enough because of the persistence of some cross reactivity between human and rat albumin. H RP (sigma 6, RZ 3.0) was used in a series of experiments as a tracer injected through three different routes: Intravenous; 2 different doses (5 m~!I00 grn and 17 m~/IOO gm body weight) of HRP dissolved in 1 ml saline solution were injected into anesthetized rats. The lungs were fixed at 2 and 10 min from the Injection of HRP. Intra-alveolar: .5 mg of HRP were injected intratracheally in 0.2 ml saline solution and the lungs fixed by immersion 2 min, 10 min. 30 min and 4 hrs later. Intrapleural: 10 mit of HRP dissolved in 2 ml saline solution were injected into the pleural cavity 2 and 10 min before lung fixation. Because of the possible release of vasoactive substances by f IRP modifying the capillary permeability in the rat.s 0.5 mg/l00 gm body weight of mepyramine were injected intraperitoneally immediately before the injection of HRP. After intravenous injection of a lou' dose 01 HRP (15 mg per animal), basement membranes were only lightly stained after 10 min; the vesicles were opacified by the enzyme while interendothelial spaces were not (Fig 5). After intravenous infection of a Mit!1 dose 01 H RP (50 mg per animal), as early as 2 min after the injection, the

FJ(;'L'RE 4. Rat lung, immunocytochemical revelation of injected human albumin (immersion fixation. glutaraldehyde). The albumin fills only the luminal part of an interenclothelial (_) space. Some vesicles (v ) are positive within the endothelium (En). The basement membrane (B) is unstained. (RBC: red blood cell: In: interstitium), (original magnification X 105,000). FIGlJHE 5. Rat lung, intravenous injection of low dose of HRP (immersion fixation, glutaraldehyde at 2 min). The Interendotheltal space is penetrated hy the tracer only in its luminal part. The basement membrane (B I is unstained. (C; capillary lumen; In: interstitium l. (original magnification X 95,000). FIGL'RE 6. Rat lung, intravenous injection of high dose of HRP (immersion fixation. glutaraldehyde at 2 min). The Interendotheltal space appears filled with the enzyme in all it.s length. 111e narrowed albuminal portion ( ..... ) must correspond to a stretched junction. Tht' basement membrane (B) is highly stained. (C: capillary lumen), (original magnification X 90.000).





interstitium and the basement membranes appeared intensively stained by the enzyme, as did numerous vesicles in endothelial and epithelial type 1 cells. Moreover, interendothelial spaces were also stained in all their length including the IlJUTOwed junctional part (Fig 6). In contrast, the staining in interepithelial spaces wars discontinued, abutting on a possible tight junction towards the alveolar lumen. After intratracheal injection of HRP, the enzyme was rapidly found within the interstitium and in the capillary lumen. as early as at 2 min and obviously at 10 min after the injection. ..4Jter intrapleural injection, as early as 2 min afterwards, it was possible to detect HRP within the basement membranes. Some pictures indicated a transfer by vesicles from the basement membranes to the capillary lumen. However, in these latter 2 types of experiments involving a retrograde transfer of HRP from the interstitium to the capillary lumen, some interendothelial spaces were also filled with the enzyme. DISCUSSION

In this work, the transendothelial transfer of macromolecules the size of albumin or HRP seemed mostly related to pinocytosis. It is noteworthy that even after using the tilting stage in electron microscopy, we never found continuous transendothelial channels formed by fusion of chains of vesicles. Moreover, samples of diaphragm of some animals examined in the same way did not show connecting vesicles, as described by Stmionescu et al' when using hemepeplides of ahout 1900 daltons. However, it was also possible to detect albumin and HRP in some interendothelial spaces. 10 this study carried out in the rat, the passage of HRP through the small pores (interendothelial clefts) was observed only when using high concentration, as previously demonstrated in mice. (I Thus, this challenged route seems a possible one for macromolecules of ahout 50,000 daltons, particularly when the protein concentration is high approximating the concentration of albumin in the plasma. However, how the blood concentration of proteins might influence their pathway across the blood-air membrane is not clear. These points deserve further investigation. IgG molecules (160,000 daltons) must transfer mostly through vesicles, although in our HRP~immunized animals we found anliperoxidase 19G in interendothelial spaces. Regarding larger plasma molecules, such as fibrinogen and IgM, it appeared that if some substantial transfer exists, it must be limited to the endothelium and only due to the large pores system (pinocytic vesicles). However, the negativity of the basement membrane for these proteins remains unsolved and might be related to a lack of detection hy the immunochemical method. Our kinetic studies with HRP provided clues to a bidirectional transfer as previously demonstrated by others: on the one hand from the blood to the interstitium and to the alveolus," and on the other hand, from the alveolus' or the pleura to the interstitium and to the capillaries. The permeability of the alveolar epithelium to macromolecules was limited to albumin, IgG and HRP. It was



only possible through pinocytosis, since the interepithelial clefts were not stained in this study.

ACK;>.IOWLEDGME1\T: The authors wish to thank Mrs. A. M. Laval, M. C. Pinchon and F. Huckendubler for their excellent technical assistance. This work was supported by a grant from the I:\SERM (France), contract no. 75-2~119-06. REFEREXCES

Landis EM, Pappenheimer JR; Chapter 29. Circulation. In Handbook of Physiology: Circulation II. Washington, OC, American Physiology Society. 1968,961-10.34

2 Bignon J, Chahinian P, Feldmann G, et al: Ultrastructural immunoperoxidase demonstration of autologous albumin in the alveolar capillary membrane and in the alveolar lining material in nonnal rats. J Cell BioI 64:503-509. 1975 3 Bignon J. Jaurand Me, Pinchen Me, et al: Immunoelectron microscopic and immunochemical demonstrations of serum proteins in the alveolar lining material of the rat lung. Am Rev Respir Dis 113:109.120, 1976 4 Schneeberger EE, Kamovsky MJ: The ultrastructural basis of alveolar capillary membrane permeability to peroxidase

used as a tracer. J Cell BioI37:781-793, 1968 5 Simionescu X, Simionescu M, Palade CE: Permeability of muscle capillaries to small heme-pepttdes. Evidence for existence of patent transendothelial channels. J Cell Biol

64:586-607, 1975 6 Clementi F: Effect of horseradish peroxidase on mice lung capillaries permeability. J Histochem Cytochem 18:887-

892.1970 7 Bensch KG, Dominguez E:\.~f: Studies in the pulmonary air-tissue barrier. Part IV: Cytochemical tracing: of macromolecules during absorption. Yale J Btol Med 43:236-241,


Comparison of Single Transit and Equilibration Studies of 22Na+ Distribution in the Lung* Richard M. £91'08,.M.D.; Robert S. Y. Chang, M.D.; and Philip Silverman

..From the Division of Respiratory Physiology and Medicine, Harbor General Hospital-UCl.A School of Medicine, Torranee.

Supported by XIH grants HL 00132 and HL 18606. Reprint requests: Dr. EDros, Harbor General Hospital, 100 \Vest Carson, Torrance. California 90509