Cellular and Intercellular Transport Pathways in Exchange Vessels 1 - 3 EUGENE M. RENKIN

Endothelial Transport Pathways The following is a review of current knowledge of endothelial transport pathways and transport mechanisms, with special reference to pulmonary alveolar and tracheobronchial exchange vessels. The "exchange vessels" within a tissue or an organ are the capillaries and pericytic (nonmuscular) venules. The walls of these vesselsconsist of a single layer of thin, flat endothelial cells supported on a collagenous basal lamina (figure 1).In most exchange vessel endothelia, permeation of water and solutes takes place both through the cells and through the junctional spaces between them (1, 2). Water and small lipidsoluble substances (0 2 , N2 , CO 2 ) penetrate directly through the cellsurfaces, which make up > 99.98070 of total endothelial surface. It has been proposed that large lipid molecules insoluble in water avoid the aqueous cytoplasmic phase by lateral diffusion within the lipid bilayer bounding the junctional regions (3). Cations (Na", K+), anions (Cl, HC0 3- ) , and small water-soluble solutes (glucose, amino acids) are believed to pass through open intercellular junctions, which make up less than 0.02% of the total surface area. The so-called "tight" junctions are wide enough to allow permeation of these substances, yet narrow enough to hold back plasma proteins. These openings are often referred to as junctional "pores" or interendothelial cell "slits." The restrictive element may not be the junction itself but fluid-filled spaces in a fibrous or gel matrix ("glycocalyx") filling the junction (4). It has been estimated that the width of most of these openings is 8 to 10 nm. According to Starling's hypothesis, relative impermeability to the plasma proteins is an important factor in control of transcapillary fluid exchange. However, small amounts of plasma proteins and other large molecules do penetrate exchange vessel walls, possibly by exchange of vesiclesbetween luminal and abluminal surfaces, through channels formed by transient fusion of vesicles, and/or through a few "large pores" in the intercellular junctions (1, 2, 5). Exchange vessel endothelia may be classified as continuous, fenestrated, discontinuous, and "tight-junction" (high resistance) types on the basis of structure and permeability characteristics. Only the first two types are found in the lungs and airways, and only these are shown in figure 1.

Continuous Endothelium The continuous endothelium is the most abundant and widespread type. It lines the chambers of the heart, the walls of all arteries and veins, arterioles, and muscular venules as well as capillaries and postcapillary venules in skeletal, smooth, and cardiac mus528

SUMMARY The endothelium of lung alveolar capillaries is of the continuous type, that of airway exchange vessels (capillaries and pericytic venules) includes both continuous and fenestrated types. Water and small lipophilic solutes penetrate via the endothelial cells (cell membrane pathway) as well as through intercellular junctions. Hydrophilic solutes are limited to junctional pathways and cytoplasmic vesicles. Permeation of hydrophilic solutes is progressively restricted with increasing molecular size, as by a sieve, with many openings 8 nm and a few 40 to 60 nm wide. In response to local tissue Injury or to certain chemical mediators, larger junctional pathways may be opened, greatly increasing permeability to large molecules. Both alveolar capillaries and airway exchange vessels exhibit this response, but the effective stimuli may differ (e.g., alveolar capl1Jarles are insensitive to histamine and bradykinin). Hydrophilic solutes are transported by diffusion, convection, and vesicular exchange (transcytosis). For small Ions and molecules (radii < 2 nm), diffusion is the dominant transport mode; contributions of convection and transcytosls are negligibly small. Because diffusion decreases with increasing molecular size, all three mechanisms may contribute substantially to transport of large molecules (radii> 2 nm). Fenestrated endothelia have higher hydraulic conductivities and are more permeable to small ions and molecules than are continuous endothelia. However,their permeablllties to plasma proteins are about the same. Lung alveolar capillary endothelium has lower hydraulic conductivity and lower solute permeablllties than do other continuous endothelia (heart, skeletal muscle). Airway exchange vessel endothelium has about the same permeability to serum albumin as alveolar capillary endothelium. AM REV RE5PIR DI5 1992; 146:528-531

cle, skin, and connective tissues. Pulmonary capillaries lining the walls of the alveoli and most of those underlying the bronchial epithelium are of this type (6, 7). Maintenance of normal permeability of most continuous endothelia requires the presence of plasma proteins, notably, albumin and orosomucoid (n-l acid glycoprotein). These substances are believedto interact with the junctional fibrous matrix (8). It is not clear whether permeabilities of other types of endothelium depend on plasma proteins.

Fenestrated Endothelium The fenestrated endothelium is found in exchange vessels of secretory and excretory organs: exocrine and endocrine glands, gastrointestinal mucosa, kidney (glomerular and peritubular capillaries), and the choroid plexuses of the brain. In dog lungs it is present in airway tissue only around serous glands (6), but in rats it is also present in tracheobronchial mucosal capillaries (7). Much of the cell surface (50 to 95%) is like that of continuous endothelium, but the remainder is less than 0.05 urn thick and bears numerous fenestrae. These are circular structures 50 to 60 nm in diameter that may appear as open perforations in electron micrographs or may be closed by thin membranous diaphragms. In all cases the basal lamina is complete. The functional significance of fenestrae is problematical; closed, they appear to provide a cell-membrane pathway that bypasses the cytoplasm; open, they provide direct access of plasma to the basal lamina. Fenestrated capillaries generally show no greater permeability to plasma proteins than continuous capillaries, but they show considerably higher

permeability to water, ions, and small solute molecules (1, 2).

Discontinuous and Tight Junction Endothelia Discontinuous and "tight junction" endothelia are of limited distribution: the former is found in sinusoids of liver,bone marrow, and spleen, the latter is found in the brain, retina, and spinal cord. They will not be described further in this review. Injury and Inflammation In response to local tissue injury or inflammation additional transport pathways for large molecules may be opened and existing pathways made less restrictive. The response is complex, and varies among different animals, organs, and tissues (7). One phase of the injury response is retraction of a fewendothelial cells in pericytic venules, opening micron-widegaps resembling those of discontinuous endothelium. The basal lamina is not perforated. Extravasation of plasma proteins and other large molecules is greatly increased (> 50-fold). Much of the leakage is localized to the venular gaps while they remain open

1 From the Department of Human Physiology, University of California, Davis, School of Medicine, Davis, California. 2 Supported by Grant HL-I8010 from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Eugene M. Renkin, Department of Human Physiology,University of California, Davis, School of Medicine, Davis, California 95616.

529

ENDOTHELIAL TRANSPORT PATHWAYS

Fig. 1. Diagrams at electron-microscopic level of A. Continuous and B. Fenestratedendothelia, to illustrate presumed transport pathways.(N = cell nucleus; M = mitochondrion; BM = basement membrane). Pathways: (1)through cells (cell membranes and cytoplasm), (2) byvesiculartransit, (3-5) through intercellular junctions (via "small pores;' "large pores:' and membrane surface), (6)through channels formed by transient fusion of luminal and ablurninal vesicles, (7) through open fenestrae, and (8) through diaphragms of closed fenestrae. From reference 1, with permission.

A

Microvascular Exchange Mechanisms \,

Ultrafiltration Ultrafiltration is convectivemovement of fluid and permeating solutes, volume flow from plasma to tissue, or vice versa. It is controlled by the balance of hydrostatic and osmotic forces across exchange vessel walls and depends on the relative impermeability of exchange vesselendothelium to plasma proteins. Rates of fluid movement in either direction (lv, volume/time) may be described by the following equation, which incorporates several refinements to Starling's original hypothesis (13):

=

Fenestrated endothelium Renal glomerulus Intestinal mucosa Continuous endothelium Mesentery Heart Skeletal muscle Lung Tight-junction endothelium Brain Transport epithelium Renal tubule Proximal Distal Intestinal mucosa

B

(6, 9), and because they are so conspicuous, it has often been assumed that they are the only source of increased protein leakage. However, macromolecular extravasation is increased in both capillaries and venules in the absence of or apart from the gaps, though to a lesser extent (two- to five-fold) (10), and permeability remains high long after the visible gaps close (11, 12). Both alveolar and airway exchange vessels are capable of exhibiting this response when adequately stimulated. However, alveolar capillaries are unresponsive to histamine, bradykinin, and serotonin, which are highly effective stimuli for airway vessels (6, 7).

Jv

TABLE 1 HYDRAULIC CONDUCTIVITIES (Lp)*

LpA [(Pc - Pi) - a(llp -lli)] (1)

Positive lv means fluid movement out of the capillary, negative lv means fluid movement into the capillary. L p ("hydraulic conductivity") represents endothelial permeability to ultrafiltrate, A is capillary surface area, P, and Pi are fluid pressures in capillary and interstitium, respectively, and IIp and lli are the respective colloid osmotic pressures. The product LpA is called the "capillary filtration coefficient"; it is often represented as KF or CFC; the format LpA emphasizes its two components. Because microvascular endothelium is not completely impermeable to plasma proteins, lli is not zero; it may vary from 0.2 and 0.8 IIp in different organs. The symbol a (sigma) represents the "reflection coefficient" of the endothelium, a measure of its efficiency in retaining protein: 1.00 is perfect; a textbook figure for total plasma proteins in most capillaries is 0.95.

Hydraulic conductivities of fenestrated and continuous capillary endothelia in several mammalian organs are listed in table 1; values for the "tight junction" capillaries of the brain and for representative transport epithelia are included to extend the comparison. Hydraulic conductivities of fenestrated endothelia are higher than those of continuous endothelia. Valuesfor continuous endothelia cover a wide range (nearly two orders of magnitude), with lung (alveolar) capillaries lowest, but still 25 times greater than brain capillaries. Hydraulic conductivities of continuous endothelia are similar in magnitude to those of transport epithelia. No data are available for tracheobronchial exchange vessel endothelium, though isolated, perfused preparations suitable for its measurement have been described (14).

Diffusion Diffusion is the principal mechanism for endothelial exchange of respiratory gases, inorganic ions, and small solutes in general. Kinetic movement of individual ions and molecules results in net solute transfer from regions of high concentration to those of low concentration, without volume transfer. Fick's law states that the rate of diffusion (ls) of a solute through a thin membrane is proportional to membrane permeability (P) to that solute, to membrane surface area (A), and to the difference in concentration of the solute on either side (i\C): ls = PA i\C

(2)

Solute permeabilities in exchange vessel endothelia depend on both membrane and solute characteristics. In general, permeabilities to lipophilic solutes are much higher than to hydrophilic solutes of similar molecular size because they can penetrate through the cellmembrane pathway. For hydrophilic solutes, permeability decreases with increasing molecular size, and for macromolecules, permeability decreases with increasing negative charge (15, 16). Hydrophilic solute permeabilities for continuous endothelia of cardiac and skeletal muscle and lung are compared with fenestrated capillaries of intestine in figure 2. The ordinate is permeability on a logarithmic scale covering six orders of magnitude; the abscissa

1.5 x 10-' 1.3 x 10-'

5.0

X

8.6

X

2.5 8.4

X

10-10 10-11 10- 11 10- 12

3.0

X

10-13

2 3 9

X

X

10-10 10- 11 10- 12

X

X

represents the Einstein-Stokes radii of the solute molecules up to 10 nm. The dotted curve represents the decline of free diffusibility in water at 37° C. Permeabilities for solutes smaller than 2 nm were measured by indicator diffusion at high blood flow rates and approximate true diffusive permeabilities (16). Values for larger molecules were calculated from lymph/plasma concentration ratios at basal lymph flows (2, 5, 17), and they include convective and vesicular components (see below). Permeability to small solutes in all organs falls more steeply than free diffusibility, suggesting that diffusion is restricted by junctional structures ("pores" or fiber network) with openings approximating 4 nm in radius or half-width. Extension of finite, though greatly reduced, permeability to larger molecules is the basis for postulating "large pore" or vesicular transport (I, 18). These would contribute negligibly to transport of ions and small molecules. Permeability of fenestrated intestinal exchange vessel endothelium to ions and small molecules is an order of magnitude or more greater than that of muscle and lung. However, permeability to large molecules is no greater than that in muscle. Permeabilities for skeletal and cardiac muscle endothelia are practically identical over their entire range (and are represented on the graph by a single curve). The greater PA products for cardiac muscle are due to the eightfold greater surface area of exchange vessels in the heart. However, permeabilities for lung alveolar capillaries are consistently lower than those for skeletaland cardiac muscle(approximately one-fourth as large). No comparable small-molecule data are available for airway exchange vessels, although indicator transit methods used to study solute metabolism (14)could be adapted to measure small-solute PA products. Barrow and coworkers (19) measured flow and albumin concentration in sheep tracheal lymph (postnodal) in control conditions and after irritation of the mucosa by cotton smoke. Fromtheir control data and from my own rough estimate of exchange vesselsurface area (i,500 cmvg) (table 2), I calculated P(albu-

530

EUGENE M. RENKIN

10. 3

"'-t:s..,...... ....

10- 4 u

~

~ € >-

h, 10. 5

~

10- 6

I~

..

iii

" .............

10· 7

,, ,, ,, ,, ,, ,, ,

s. c muscle

intestine

"

W

::;

\

a: w

e,

----n

10. 8

'D

--

lung (alv)

10· 9 10

0

Fig. 2. Permeabilities of exchange vessels to hydrophilic solutes in relation to molecular size (Einstein-Stokes radius). Continuous endothelium: (s) skeletal and (c) cardiac muscle (same curve), lung (predominantly alveolar capillaries). Fenestrated endothelium: small intestine. The dolled line at the top shows the decline of free diffusibility with increasing size. Molecules for s,c muscle, left to right: Na'CI-, urea, glucose, sucrose, inulin, albumin, IgG; same for lung and intestine except: lung, mannitol at 0.4 nm, intestine 51CrEDTA at 0.5 nm, IgM at 9.1 nm. Data from references 1, 5,16, and 17.

MOLECULAR RADIUS, nm

min) = 9 x 10-9 cm/s, This is about the same as for lung alveolar capillaries (1 x 10-8 cm/s) (figure 2). Measurement of large-solute permeabilities of intrapulmonary airways through lymph are not feasible because of the drainage of alveolar lymph through parabronchiallymphatics (20). However, measurement of tissue accumulation of tracer macromolecules appears to be possible. Dr. Kramer, working in my laboratory, was able to measure basal albumin permeability-surface area products (PA, equivalent to albumin clearances) and small-vessel plasma volumes (Vp) in this way for various parts of the tracheobronchial tree and for lung parenchyma of anesthetized rats (21). His data are listed in the first two rows of table 2. The fourth row of figures displays albumin permeabilities calculated by dividing the PA products by estimates of exchange vessel surface calculated from the measured plasma volumes (A = 4Vp/exchange vessel radius, 3 x 10-4 cm; row 3 in the table). The permeabilities are low for all sampled airway segments, and they are approximately the same as for lung parenchymal capillaries. [Our estimate of pulmonary capillary surface is almost the same as that obtained by morphometric procedures (3,000 cmt/g), but our figures for albumin permeability of alveolar capillaries and tracheal exchange vessel endothelium are only onefourth of the values obtained from lymph studies. This may reflect species differences between rat and dog/sheep, but the disparity

is at least partly due to oppositely directed technical errors inherent in the two methods. Lymph (output) clearances tend to be too high, whereas blood-to-tissue (input) clearances tend to be low. The correct value probably lies somewhere between them.]

Transport of Plasma Proteins and Other Large Molecules (Einstein-Stokes radii> 2 nm) Three major mechanisms are believed to contribute to macromolecular transport: convection, diffusion, and vesicular exchange (transeytosis)(I5, 17). The latter two processes have similar dynamics, and they are sometimes lumped together as "dissipative" mechanisms. For smaller solutes, convection and vesicular exchange are relatively unimportant, and they are usually neglected. However, diffusion alone cannot account for transport of macromolecules; either convection (through large pores) or vesicular exchange or both must be considered (18). The relative importance of these three mechanisms is presently a matter of contention. Convection refers to transport of solute molecules entrained in a stream of ultrafiltrate. The concentrations of small solutes in the filtrate are close to those in plasma, but large solutes are restricted or "sieved" by the membrane. At high filtration rates their concentrations in the filtrate approach (I o)Cp , where C p is plasma solute concentration. The expression (I - 0) is called the "sol-

Js = Jv(l - cr)C p + PA ~C Fx + a Qv

TABLE 2

CONVECTION

PERMEABILITY OF LUNG AND AIRWAY EXCHANGE VESSELS TO ALBUMIN

PA, ~I g-' Vp, ~I g-1' A, cm' g-,t P, cm s-4

h-1 '

vent drag" coefficient of a solute; 0 is the reflection coefficient of equation 1. Solute transport by convection is thus equal to Jv(l - O)Cp , Jv being the rate of fluid flow (4, 15). Diffusion may be independent, as described above, or it may be superimposed on convection. In the absence of convective flow through the same pathway, diffusion through exchange vessel endothelium follows Fick's law for a thin, limiting membrane (equation 2). This relation is valid for a linear gradient of solute concentration across the membrane. When there is an appreciable fluid velocity in the diffusion channels, solute concentration gradients within channel entrances are made less steep by convective entry of fluid, and thus diffusion falls below values predicted by this equation. Evaluation of diffusion transport under these conditions requires multiplication of equation 1 by a nonlinear correction function (F x) based on the ratio of convective solute entry to solute diffusion entry [Fx = x(e- X - 1), whereas x = Jv(I o)/PA] (5, 15). Vesicularexchange has usually been envisioned as a diffusion process in which the vesicles and their contents are the statistically distributing units (IS, 22, 23). Luminal vesicles filled with plasma detach and move randomly within the cytoplasm (Brownian motion) until they reach the abluminal surface or return to the side they started from. Those reaching the abluminal surface become attached, reopen, and exchange their contents with interstitial fluid. Thus, net transport of the solutes they contain is from high to low concentration, as in molecular diffusion. However, exchange is much slower than molecular diffusion because ofthe large size of the vesicles and the high viscosity of endothelial cytoplasm, Vesicular solute transport (Jsv) depends on the volume of vesicular contents carried from plasma to interstitial fluid per unit time (Qv), and the partition of solute molecules between plasma and the vesicular contents (a). Jsv = a Qv ~C; the form is the same as Fick's law (equation 2), with aQv replacing PA. Because there is no convective flow through this pathway, no correction factor is required. Total solute transport is the sum of these three components:

Trachea

Mainstem Bronchi

Intrapulmonary Airways

Lung

15.6 112 1,500 2.9 x 10-'

12.8 132 1,800 2.0 x 10-'

16.6 181 2,400 1.9 x 10-'

24.6 199 2,700 2.5 x 10-'

Definition of abbreviations: PA = permeability-surface-area product: Vp = small-vessel plasma volumes; A = membrane surface area: P = membrane permeability. • From Kramer et a/. (21). t A = 2Vb!rc, where Vb is blood volume (= 2Vp, small-vessel plasma volume) and rc is exchange vessel radius (" 3 I'm). ~ P = PAIA.

DIFFUSION

~C

(3)

VESICULAR

Reflection coefficients. As Jv is increased by progressive elevation of capillary pressure (equation 1), the convective term in equation 3 comes to predominate over the others (the vesicular term remains constant, and the diffusive term decreases because F x falls toward zero), and it is possible to estimate solute drag coefficients (I - 0) and thus reflection coefficients (0) for macromolecular solutes: 1- 0 = Js/JvC p (4, 15). This method tends to underestimate 0 if Jv cannot be raised high enough, especially if diffusive and vesicular transport coefficients are large.

831

ENDOTHELIAL TRANSPORT PATHWAYS

1.0

intestine

Fig. 3. Microvascular reflection coefficients in relation to molecular size. Values for molecular radii below 2 nm were measured from theirtransient osmoticeffects (16, 17, 27).Values formolecular radii above 2 nmwere calculated from limiting Iymph-to-plasma concentration ratios at elevated capillary pressures. They are minimal estimates (2, 5, 17, 20). Molecules fors,c muscle, left to right: Na'CI-, urea, sucrose, inulin, albumin, IgG; same forlungand intestine at equivalent molecular radii.

-

•. ··0

-

0""'-

08

lung (alv)

0.6

P

0.4

..

0

-,

5, C muscle

0.2

0.0

+----r------,------r-----,----

,a

a

MOLECULAR RADIUS, nm

Graphs of (J versus molecular radius for exchange vesselsofcardiac and skeletal muscle (represented by a single curve), lung, and intestine are presented in figure 3. Largemolecule reflection coefficients measured in this way (from minimal lymph/plasma ratios) are approximately the same for exchange vessels in muscle and intestine: albumin, 0.90; IgO, 0.95; IgM, 0.98 (5). For lung (alveolar) capillaries they are lower: 0.80 for albumin, 0.94 for IgO (20, 24). However all these values may be too low and the corresponding solute drag coefficients too high. Recent direct measurements of solute-filtrate coupling in tissues and organs of intact animals and in single capillaries perfused with whole plasma have yielded reflectionscoefficients for albumin of> 0.98 and for IgO of> 0.99 in organs with continuous capillaries (25). If these values are correct, then the solvent drag coefficient for albumin is < 0.02, and that for IgO is < 0.01 (instead of 0.10and 0.05, respectively) and convective transport through large pores cannot account for observed basal transport of either. In organs with fenestrated capillaries, the tissue uptake method yields reflection coefficients close to those obtained from measurements on lymph (26). Reflection coefficients for small molecules Were measured osmotically (16). In muscle capillaries they are low: for NaCl, urea, and sucrose they are less than 0.1. However, reflection coefficients for these solutes in lung capillaries are much higher, 0.3 to 0.4 (27). The difference is indicative of an exceptionally large contribution of the cell membrane pathway (pathway 1 in figure 1) to hydraulic conductivity in lung (alveolar) capillaries (13). Assuming that the reflection coefficient of endothelial cells to NaCI is 1.0, then about 300/0 of total LpA can be assigned to the cells and 70% to the intercellular pathways. Comparable values of 5 and 95%, respectively, can be assigned to muscle on the same basis, and they are characteristic of continuous capillaries in general. This observation, together with the low total hydraulic conductivity (table 1)and low small-solute permeabilities (figure 2), suggests that the open junctional pathways are

less extensive in pulmonary capillary endothelium than in other continuous endothelia. No small-molecule data are available for fenestrated endothelium nor for airway exchange vessels. Acknowledgment The writer wishes to thank Dr. George Kramer for his review of the manuscript and for his helpful discussion of the airway permeability data presented, and Pamela Lowart for secretarialassistance.

References 1. Renkin EM. Multiple pathways of capillary permeability. Circ Res 1977; 41:735-43. 2. Parker JC, Perry MA, Taylor AE. Permeability of the microvascular barrier. In: Staub NC, Taylor AE, eds. Edema. New York: Raven Press, 1984; 143-87. 3. Scow RO, Blanchette-Mackie EJ, Smith Ie. Role of capillary endothelium in the clearance of chylomicrons: a model for lipid transport from blood by lateral diffusion in cell membranes. Circ Res 1976; 39:149-62. 4. Curry FE, Michel CC. A fiber matrix model of capillary permeability. Microvasc Res 1980; 20:96-9. 5. Taylor AE, Granger DN. Exchange of macromolecules across the microcirculation. In: Renkin EM, MichelCC, eds. Handbook of physiology, Section 2: The cardiovascular system. Vol IV: Microcirculation. Bethesda: American Physiological Society 1984; 467-520. 6. Pietra GP, Magno M. Pharmacological factors influencing permeability of the bronchial microcirculation. Fed Proc 1978; 37:2466-70. 7. McDonald DM. The ultrastructure and permeability of tracheobronchial blood vessels in health and disease. Eur Respir J 1990; 3(Suppl 12: 572s-85s). 8. Haraldsson B, Rippe B. Serum factors other than albumin are needed for the maintenance of normal capillary permeability in rat hindlimb muscle. Acta Physiol Scand 1985; 123:427-36. 9. Svensjo E, Arfors K-E, Grega GJ. Morphological and physiological correlation of bradykinininduced macromolecular efflux. Am J Physiol1979; 236:H600-6. 10. Carter RD, Joyner WL, Renkin EM. Effect of histamine and some other substances on molecular selectivity of the capillary wall to plasma proteins and dextran. Microvasc Res 1974; 7:31-48. 11. Svensjo E, Joyner W1. The effects of inter-

mittent and continuous stimulation of microvessels in the cheek pouch of hamsters with histamine and bradykinin on the development of venular leaky sites. Microcirc Endothelium Lymphatics 1984; 1:381-96. 12. Curry FE, Joyner WL, Rutledge JC. Graded modulation of frog microvessel permeability to albumin using ionophore A23187. Am J Physioll990; 258:H587-98. 13. Michel CC. Fluid movements through capillary walls. In: Renkin EM, Michel CC, eds. Handbook of physiology. Section 2: The cardiovascular system. VolIV: Microcirculation. Bethesda: American Physiological Society 1984; 375-409. 14. Grantham CJ, Jackowski JT, Wanner A, Ryan US. Metabolic and pharmacokinetic activity of the isolated sheep bronchial circulation. J Appl Physiol 1989; 67:1041-7. 15. Curry FE. Mechanics and thermodynamics of transcapillary exchange. In: Renkin EM, Michel CC, eds. Handbook of physiology. Section 2: The cardiovascular system. Vol IV: Microcirculation. Bethesda: American Physiological Society 1984; 309-74. 16. Crone C, Levitt DG. Capillary permeability to small solutes. In: Renkin EM, Michel CC, eds. Handbook of physiology. Section 2. The cardiovascular system. VolIV: Microcirculation. Bethesda: American Physiological Society 1984; 411-66. 17. Renkin EM, Curry FE. Transport of water and solutes across capillary endothelium. In: Giebisch G, Tosteson DC, Ussing HH, eds. Membrane transport in biology. Vol IV. New York: Springer-Verlag, 1978; 1-45. 18. Renkin EM. Transport pathways and processes. In: Simionescu N, Simionescu M, eds. Epithelial cell biology. New York: Plenum, 1988; 51-68. 19. Barrow RE, Morris SE, Linares HA, Herndon DN. Tracheal venous blood and lymph collection: a model to study airway injury in sheep. J Appl Physiol 1991; 70:1645-9. 20. Taylor AE, Parker rc. PulmonaryinterstitiaI spaces and lymphatics. In: Fishman AP, Fisher AB, eds. Handbook of physiology. Section 3: The respiratory system. Vol I: Circulation and nonrespiratory functions. Bethesda: American Physiological Society 1984; 167-230. 21. Kramer GC, Ashley KD, Wong MS, Renkin EM, Sheikh AA. Blood to tissue transport rates of albumin in airways and whole lung of rat (abstract). Proc IUPS 1989; XVII:278. 22. LoudonMF, Michel CC, White IF. The labelling of vesicles in frog endothelial cells with ferritin. J Physiol (Lond) 1979; 296:97-112. 23. Clough G, Michel Ce. The role of vesicles in the transport of ferritin through frog endothelium. J Physiol (Lond) 1981; 315:127-42. 24. Parker RE, Roselli RJ, Brigham K1. Effects of prolonged elevated microvascular pressure on lung fluid balance in sheep. J Appl Physiol 1985; 58:869-75. 25. Renkin EM, Gustafson-Sgro M, Sibley 1. Coupling of albumin flux to volume flow in skin and musclesof anesthetized rats. Am J Physiol1988; 255:H458-66. 26. Renkin EM, Rew K, Wong M, O'Loughlin D, Sibley 1. Influence of saline infusion on bloodtissue albumin transport. Am J Physiol 1989; 257:H525-33. 27. Perl W, Chowdhury P, Chinard FP. Reflection coefficients of dog lung endothelium to small hydrophilic solutes. Am J Physiol 1975; 228: 797-809.

Cellular and intercellular transport pathways in exchange vessels.

The endothelium of lung alveolar capillaries is of the continuous type, that of airway exchange vessels (capillaries and pericytic venules) includes b...
568KB Sizes 0 Downloads 0 Views