Skip Navigation



NDT Advance Access published online on April 3, 2008
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfn049
This Article
Right arrow Extract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/7/2147    most recent
gfn049v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Rippe, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rippe, B.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2008]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org


PRO/CON DEBATE: free water transport, small pore transport and the osmotic pressure gradient three-pore model of peritoneal transport

Bengt Rippe

Department of Nephrology, University Hospital of Lund, Sweden

Correspondence and offprint requests to: Bengt Rippe, Department of Nephrology, Lund University, University Hospital of Lund, S-211 85 Lund, Sweden. Tel: +46-46-171247; Fax: +46-46-2114356; E-mail: bengt.rippe{at}med.lu.se

Keywords: aquaporins; capillary permeability; interstitium; lymphatic absorption; ultrafiltration



   Introduction
Top
 Introduction
Key physiologic features of...
 How does an endothelial...
 Conclusions
 References
 
In this issue of NDT, Flessner in a commentary [1] argues that the three-pore model (TPM) of peritoneal transport, although mathematically a powerful predictor of solute transport and ultrafiltration (UF) in peritoneal dialysis (PD), may be too simple as a tool for understanding the physiology of transperitoneal exchange. Flessner then disregards the fact that the TPM can be modified in a very simple fashion by taking both the capillary and the interstitial barriers into account in the modelling. This has in fact already been done by adding a second heteroporous barrier [2] or an interstitial gel–matrix barrier in series with the capillary membrane in the TPM; the latter model denoted the ‘three-pore membrane/fibre matrix model’ [3]. Flessner also brings up now the 30-year-old controversy whether the endothelial ‘fuzzy’ surface layer, the glycocalyx, has size-selective sieving properties or not.

In response to Flessner's criticism of the TPM, I would like to cite Leonardo DaVinci: ‘Simplification is the ultimate form of sophistication’. The TPM is based on decades of basic capillary physiologic research, and the many simplifications made in the model have been done with due consideration of all the complexities of the peritoneal barrier. The usefulness of the model is illustrated by the fact that it can predict solute and fluid transport not only for glucose, but also for alternative osmotic agents, such as icodextrin and amino acid solutions [4,5], and also for solutions with altered electrolyte composition [6,7]. This is in contrast to the distributed model, which is in an early developmental phase. For example, it has great problems in predicting UF (glucose osmosis) from the peritoneal tissue to the peritoneal cavity. In contrast, the three-pore membrane/fibre matrix model can explain how osmotic fluid flow can occur from the peritoneal tissue to the peritoneal cavity against a glucose concentration gradient, and against a hydrostatic pressure gradient, at the same time as fluid and macromolecules move from the peritoneal cavity to the peritoneal tissue, by hydrostatic pressure driven flow.

Concerning the ‘glycocalyx (fibre matrix) hypothesis’, it may be regarded as a ‘soft’ hypothesis since it is hard to test experimentally. Actually, the pore theory and the fibre matrix theory are mathematically similar, and there is no way of actually distinguishing between the two models. In fact, the bimodal (‘small pore’/‘large pore’) solute selectivity of the TPM can be accurately described by the presence of a dense, but heterogeneous, fibre matrix filling out all spaces in between the endothelial cells of the capillary barrier.

In this commentary I shall first review the key features of the TPM and how these have contributed to a deeper understanding of the physiology of peritoneal exchange. Especially I will discuss the puzzling discrepancy between the clearance of tracer macromolecules from the peritoneal cavity into the tissues (KE) and the net reabsorption of fluid from the peritoneal cavity to plasma. Second, arguments will be raised that the endothelial glycocalyx does not satisfy the criteria for being a highly size-selective barrier in the TPM.



   Key physiologic features of the TPM
Top
 Introduction
 Key physiologic features of...
How does an endothelial...
 Conclusions
 References
 
The capillary wall is the dominating peritoneal barrier, being heteroporous. Despite reflection coefficients ({sigma}) near zero, sieving coefficients ({theta}) for small solutes are 0.5–0.6 and not near unity [(1–{sigma})]
The anatomic barrier separating the blood from the peritoneal fluid in PD contains a number of structures placed in series, mainly the peritoneal capillary walls, the interstitium and the mesothelium. According to the classical TPM, the capillary wall is by far the dominating transport barrier and it therefore, as a good approximation, neglects the other serial barriers under normal conditions. The second important principle of the TPM (not included in older models) is that the capillary barrier is heteroporous, with a bimodal size-selectivity for solute transport (‘small pores’/‘large pores’) combined with a transcellular pathway exclusive for water (aquaporin-1; AQP-1). In fact, long before the cloning of AQP-1 [8], the TPM predicted that the phenomenon of ‘sodium sieving’ must be the result of passage of water devoid of solute through water-only transcellular pores. This paradigm was recently amply confirmed by PD experiments in mice lacking AQP-1 [9]. The consequence of the presence of water-only pores is that small solute sieving coefficients (filtrate-to-plasma concentration ratios; {theta}) are usually 0.5–0.6 under conditions of crystalloid (e.g. glucose) osmosis, instead of near unity, as had been the case in the absence of aquaporins. At the same time, small solute reflection coefficients ({sigma}), characterizing the osmotic efficiency of the solute, are very low (below 0.1), as in most continuous capillary walls [10,11]. This is in fact a key feature of the TPM and is the basis for the understanding of the difference in osmotic efficiency between colloid osmotic-type agents, such as icodextrin ({sigma} > 0.4) and crystalloid osmotic-type agents, such as glucose ({sigma} < 0.1). In older models [12] or newer [13] where {sigma} for small solutes are set >0.4 [12], it is impossible to correctly model UF for colloid osmotic-type agents [14].

Transendothelial macromolecule transport occurs by convection through large pores and not by ‘transcytosis’
Yet another unique feature of the TPM is that it handles the transport of macromolecules separately from that of small solutes. The size-selectivity of the TPM for macromolecules larger than albumin is compatible with the presence of very infrequent large pores (radius ~250 Å; 0.01% of the total number of pores) across which plasma proteins can reach the interstitium and the peritoneal cavity by pressure-driven flow, unaffected by the oncotic or colloid osmotic pressures acting across small and transcellular pores. Thus, large pore transport is dictated only by changes in the transcapillary hydrostatic pressure gradient and/or the number of large pores available for transport, both of which can increase dramatically in states of inflammation. While the TPM has consistently predicted that the transvascular transport of macromolecules must occur through large pores, many morphologists and cell biologists have considered this process to be due to transendothelial shuttling of proteins by plasmalemmal vesicles (caveolae), so-called transcytosis. However, it was recently shown that caveolin-1 deficient mice completely lacking endothelial caveolae (vesicles), and therefore being incapable of transcytosis, actually show an increased, not a reduced, transport of macromolecules across their microvessels [15]. It thus seems clear that transcytosis does not significantly contribute to the transport of proteins between the plasma and the peritoneal cavity.

The reabsorption of isotonic fluid from the peritoneal cavity to plasma occurs via the small capillary pores due to the Starling mechanism, because of the high plasma to peritoneal colloid osmotic pressure gradient ({Delta}{pi}) during PD
Another key feature of the TPM is that it predicts the reabsorption of isotonic fluid devoid of colloid (e.g. isotonic NaCl) as occurring through small pores, with only a minor contribution of lymphatic absorption (0.2–0.3 mL/min). This transport behaviour of the peritoneum has recently been shown for 1.36% glucose solutions using a modified ‘LaMilia technique’ (see below) [16], but it is more difficult to show for hypertonic glucose in short (4 h) dwell times, when net reabsorption is low or absent, as in the study by Parikova et al. in this issue of NDT [17]. Using the ‘LaMilia technique’ [18], the UF occurring through small pores (SPUF) is based on the assessment of the transperitoneal clearance of Na+ because Na+ almost exclusively traverses the small pores by convection during the first 0–1 h of the dwell. Subtracting SPUF from total (net) UF, the ultrasmall pore-mediated UF (USPUF) can be calculated. However, the technique becomes increasingly inaccurate (even after corrections for Na+ diffusion) for dwell times longer than 1–2 h [19]. This problem, combined with the absence of net reabsorption (fluid loss) in the study by Parikova et al. [17], may be responsible for the problems of showing back-filtration through the small pores in their study. Another source of ambiguity seems to be related to the estimation of intraperitoneal (i.p.) volume as a function of dwell time using a macromolecular volume marker, the clearance of which out of the peritoneal cavity, usually denoted KE, will be discussed next.

The clearance of a macromolecular marker from the peritoneum to peritoneal tissues (KE) is a complex parameter determined by several different processes: convection into the tissue, ‘volume recirculation’ between tissue and cavity, lymphatic reabsorption, and capillary small pore fluid, but not macromolecule, reabsorption
Because the TPM is based on direct volumetric measurements of drained peritoneal volume versus time it avoids the complexity associated with the use of a macromolecular volume marker, continuously disappearing at a high rate from the peritoneal cavity. The clearance of a macromolecular marker into the peritoneal tissues, KE, can increase due to increases in peritoneal hydrostatic pressure, also causing some oedema in the tissue, as discussed at some length by Flessner [1]. In a steady state, however, KE is mainly dependent on four processes, namely (1) the convective fluid transport into the tissue, drained by (2) lymphatic absorption (~0.2 mL/min) and (3) back-filtration of fluid, but not of marker, through the small pores in the capillary walls (~0.9 mL/min), and also, as suggested here, by (4) ‘volume recirculation’ between the peritoneal cavity and the peritoneal tissue (~0.7–1.0 mL/min). The latter process is not associated with net fluid movement, but with net macromolecule permeation, into the tissue. Binding of a marker to the peritoneal tissue and marker diffusion also contribute to KE, which is thus larger than the sum of lymphatic reabsorption and capillary small pore back-filtration.

‘Volume recirculation’ is a concept taken from capillary physiology and is a phenomenon that occurs in heteroporous (e.g. large pore/small pore) membranes. It has been described in detail in a number of previous publications [20–22]. In isolated perfused muscle capillary beds, volume recirculation was demonstrated under conditions of zero net fluid movement from plasma to the interstitium. Such conditions can be created by lowering the capillary hydrostatic pressure slightly (by 0.5–1.5 mmHg) from the normal level, so that net filtration (lymph flow) ceases. Under such conditions the normal filtration of albumin and other proteins occurring through the large pores, unopposed by any oncotic gradients, is exactly counterbalanced by reabsorption of tissue fluid (devoid of proteins) via the small pores due to the altered Starling balance (following upon the reduction in capillary hydrostatic pressure) (Figure 1a). Small pore back-filtration is due to the fact that the effective colloid osmotic pressure across the small pores is now larger than the effective transcapillary hydrostatic pressure gradient. Under such ‘isovolumetric’ conditions, filtration through the large pores, containing fluid and plasma proteins, is in exact equilibrium with the reabsorption of tissue fluid, devoid of proteins, through the small pores. Thus, there is net movement of proteins (albumin) out of the capillaries in the absence of net fluid transport.


Figure 1
View larger version (29K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 (a) ‘Volume recirculation’ in a two-pore membrane under conditions when there are no net volume shifts across the membrane. Colloid osmotic pressure exerted by the proteins on the left side (due to a higher protein concentration here) across small (protein-restrictive) pores is counterbalanced by hydrostatic pressure. The latter causes fluid and proteins to move across the large pores to the right side, whereas only protein-poor fluid returns to the left side. Thus, there is a net convective flux (solvent drag) of proteins from left to right in the absence of a net fluid shift (from [22]). (b) In analogy with the situation across heteroporous capillary walls, there should be a substantial component of volume recirculation also in the interphase between the peritoneal cavity and the interstitium. The large pore equivalent is represented by large fluid channels penetrating the tissue gel (randomly). Solid arrows represent transport of fluid (and small solutes), while grey arrows represent transport of macromolecular tracer. Tracer macromolecules are depicted as grey circles. Through the tissue channels fluid and macromolecules can move by hydrostatic pressure-driven flow. Some of the fluid is returned to the peritoneal cavity through the tissue gel, which retains macromolecules, yet being nearly freely permeable to water and small solutes. There is, however, at least some resistance offered by the gel matrix, which is enough to direct the fluid flow from the peritoneal capillaries towards the cavity. Even under conditions of net reabsorption of fluid from the peritoneal cavity to plasma a ‘volume recirculation term’ (acting like a ‘diffusive’ term) will enhance the elimination of macromolecular tracer from the peritoneal cavity, and make its clearance (KE) larger than the net uptake of fluid into capillaries and lymphatics.

 
All the elements needed to set up volume recirculation are present at the interphase between the peritoneal cavity and the peritoneal tissue. As discussed by Flessner [1] the interstitial tissue is extremely permeable, yet offering at least some resistance to solute and water transport, except in areas with free fluid channels. Volume recirculation between the peritoneal cavity and tissue can occur due to the higher hydrostatic pressure in the peritoneal cavity than in the tissue, and also due to the usually much higher crystalloid osmotic pressure in the peritoneal cavity than in the tissue. Even in the absence of these conditions, a ‘volume recirculation term’ has to be added to the equations that describe transport through a heteroporous barrier [21]. It is thus likely that the peritoneal tissue is heterogeneous, with a matrix of fibres, which is penetrated by very large parallel channels of fluid (regions of extremely low size-selectivity). Along these channels, fluid and macromolecules can move into the tissue by hydrostatic pressure-driven flow. At the same time (at least during the early phase of the dwell) fluid and small solutes are ‘dragged’ in the opposite direction out from the peritoneal tissue to the cavity by crystalloid (glucose) osmosis through the interstitial fibre matrix (Figure 1b). A heterogeneous interstitial fibre matrix can thus explain how vectorial osmotic water flow can occur through the normal peritoneum against the glucose concentration gradient (and along the glucose osmotic gradient), and also against the hydrostatic pressure gradient, at the same time as fluid and macromolecules move into the tissue through regions devoid of fibres driven by hydrostatic pressure and unopposed by crystalloid osmotic pressure.

In summary, the classical TPM can predict the measured drained volume as a function of time and does not need to predict the virtual volume that appears from the volume of distribution of an i.p. macromolecular marker in the peritoneal cavity and tissue. KE is a very complex parameter, which is only partly dependent on net fluid reabsorption from the peritoneal cavity to lymphatics and blood capillaries. According to the TPM, it is not correct to add KE to the net UF during the dwell to establish a transcapillary UF (TCUF) curve since only a part of KE is related to (net) fluid transport across the peritoneum. According to the TPM, TCUF is just the net UF + the lymphatic reabsorption term (0.2–0.3 mL/min). The net fluid loss from the peritoneal cavity (~1.2 mL/min) is the sum of the small pore back-filtration (~0.9 mL/min) and the lymphatic reabsorption term. Since TCUF has a different meaning in the study of Parikova et al. [17], leading to higher values of both TCUF and small pore UF than predicted by the TPM, this may have contributed to the deviation of the experimentally assessed small pore UF from the TPM predicted small pore UF at 2–4 h. Had net reabsorption prevailed, as for 1.36% glucose, it is likely that small pore ‘back-filtration’ would have been demonstrated [16].



   How does an endothelial glycocalyx affect the TPM?
Top
 Introduction
 Key physiologic features of...
 How does an endothelial...
Conclusions
 References
 
The answer to this question is ‘little’, or even ‘not at all’! The glycocalyx (fibre matrix) model is in fact much older than the TPM. Already in 1940, Chambers and Zweifach [23] suggested that an ‘intercellular cement’ would represent a molecular size-selective sieving material in the slits (clefts) that separate endothelial cells from each other in the capillary wall. With the advent of the pore theory of Pappenheimer et al. [24], it was suggested that the equivalent small pores might lie within the ‘intercellular cement’ or be determined by other structures within the interendothelial clefts. These structures were later found to be ‘junctional strands’ formed by integral membrane adhesion proteins (occludin and claudin-5), which form incomplete lines of cell-to-cell adhesion of endothelial cells at the luminal aspect of the interendothelial cleft [25].

In 1980 Curry and Michel proposed ‘the fibre matrix theory’ [26], implying that the selectivity of the interendothelial clefts would be dictated by the presence of a matrix of fibrous molecules filling out all pores and having a high density (~5–6% of total water volume) and a radius of 0.5–0.6 nm. The rationale for suggesting a fibre matrix filling out the capillary pores is the striking similarity of the high size-selectivity (high {sigma} to albumin) among different capillaries, or among various capillary beds, despite their great variability with respect to both hydraulic conductance (‘water permeability’, ‘LpS’) and small solute permeability–surface area products (PS; also denoted mass transfer area coefficients; MTAC). PS and LpS also seem to co-vary to a great extent. An alternative explanation to the rather uniform selectivity of capillaries, despite highly variable exchange (pore) areas is that the dimensions of the size-selective structures (slits/small pores) may be invariant from capillary to capillary, while the number of pores is highly variable.

The structural counterpart to the fibre matrix has been assumed to be the endothelial surface coat, or its innermost layer, the ‘glycocalyx’. The glycocalyx was discovered by Luft [27] using a cationic dye, ruthenium-red, that was found to stain the endothelial surface both at the luminal and the abluminal side of the endothelium, and also within the interendothelial clefts. Until recently, the glycocalyx was considered to be a 0.1-µm-thick endocapillary layer consisting of endothelial cell-derived proteoglycans, glycoproteins and glycosaminoglycans, and also containing plasma proteins (albumin and orosomucoid). The glycocalyx confers a negative charge to the endothelial surface and seems to be a major contributor to the negative charge of the glomerular filtration barrier [28], while being of little functional importance as a charge barrier in the peritoneum [29,30]. It is also crucial for the proper interactions of leucocytes with the endothelium, especially in inflammation, for the mechanosensing of shear rate, and for the regulation of the coagulation cascade [31]. However, at present, there is no clear-cut evidence that the endothelial glycocalyx would really contribute to the size-selectivity of the capillary wall.

The glycocalyx hypothesis has been debated among capillary physiologists for over three decades, and remains very controversial, since it cannot be unambiguously tested. This is due to the fact that the fibre matrix model can be made to more or less exactly fit the behaviour of the two-pore model by adjusting its critical determinants, (1) the fibre density (or the matrix ‘void volume’), (2) the fibre radius and (3) the surface area (over unit diffusion path length, A0/{Delta}X) of the matrix available for exchange. For example, a matrix having fibres of radius 0.5 nm and 5.5% matrix density (fractional void volume = 94.5%), offers a size-selectivity corresponding to that of the small pores of the TPM (radius ~4.5 nm). A very ‘loose’ matrix (~99.6% void volume) occupying a minimal ‘surface area’ would correspond to the very few large pores of the TPM. Mathematically, there is thus no fundamental difference between the two different ways of the modelling transcapillary solute transport. Actually, the TPM is based on an equivalent pore concept, which describes the physiological behaviour of the porous pathways, and not the anatomic. Therefore, transperitoneal transport can be modelled using either equivalent pore models, slit models or the fibre matrix model. If the fibre matrix theory is to be applied, then the fibre matrix should be made partly heterogeneous in order to describe transport as occurring through small pores (slits) or large pores (large slits), while the third pathway is offered by the AQP-1.

Why is pore theory preferred to glycocalyx theory in the TPM?

  1. The glycocalyx is a poorly defined ‘fuzzy’ layer of endothelial cell-anchored proteoglycans and glycosaminoglycans. Depending on staining and fixation techniques its thickness varies from 0.1 µm to several µm. The glycocalyx is usually regarded as very ‘loose’ and fragile. However, to offer a size-selectivity similar to that of small pores, its density should be >5% (fractional void volume <95%) for a fibre radius of 0.5–0.6 nm. The morphological appearance of the glycocalyx, however, indicates that it is too loose and fragile to offer a substantial sieving barrier for uncharged solutes.
  2. The permeability of the endothelial surface layer demonstrated so far is drastically different from that of normal capillary walls! Vink and Duling [32] thus demonstrated that the endothelial surface layer was largely impermeable to dextran 70 (and red cells), but highly permeable to dextran 40, albumin and fibrinogen. This does not comply with the numerous transcapillary (lymph) sieving studies performed using dextran of different sizes. They, in contrast, show that the permeability of monodisperse dextran 40 is usually only 20% higher than that of dextran 70 [33,34]. Furthermore, the transcapillary clearance of fibrinogen is low [21,35]. Indeed, Krediet et al. have demonstrated that the transperitoneal fractional clearance of dextran 70 and 40 only differed moderately [29], as recently confirmed by our laboratory [30]. Furthermore, the peritoneal clearance of albumin (negatively charged) and dextran 40 (neutral, and only slightly larger than albumin) were nearly identical [29,30]. Across the glomerular filtration barrier, however, dextran 40 shows at least a hundredfold higher permeability than albumin [36]. Thus, the permeability characteristics of the endothelial capillary surface layer, such as described by Vink and Duling, is markedly different from the sieving properties of most capillary walls.
  3. In inflammation focal endothelial cell retractions cause an increased number of dynamic intercellular gaps (large pores) by active contraction of the endothelial cell actin cytoskeleton [37–40]. This is a highly dynamic, rapid and reversible process, occurring within seconds and reversible within minutes. Conceivably, the glycocalyx may be rapidly disrupted in inflammation, but could the glycocalyx really be resynthesized and reorganized within seconds (or minutes), similar to the dynamic behaviour of the (contractile) endothelial cell actin cytoskeleton?
  4. There is no physiological technique available at present that can accurately assess the glycocalyx (volume). The techniques used by Vink and Duling and associates, based on comparing red cell (or dextran 70) intravascular distribution volume(s) with the distribution volume of dextran 40 are largely defective. This is due to the very inhomogeneous distribution of red cells among various vascular beds. A fundamental principle applied in tracer dilution measurements of volume is that the tracer must be evenly distributed in the volume to be under study. However, the well-known Fåhreus–Lindquist effect (for review see e.g. [41]) makes red cells flow much faster through small arteries and arterioles, in e.g. heart and skeletal muscle, causing the haematocrit in these tissues to be very low. Also, ‘plasma-skimming’ effects in certain tissues, such as in the renal medulla, cause regionally very low haematocrit values. Conversely, in parenchymatous organs, such as in the liver and the lung, haematocrit is high. Red cells are thus unevenly distributed in the various vascular beds of the circulation and are therefore disqualified as intravascular volume markers! Another problem in the studies of Vink and associates is the extremely high clearance of dextran 40 from the circulation, thought to be due to its permeation through the endothelial glycocalyx [42], but in reality being dependent on its marked excretion to the urine by glomerular filtration [36]. The Vink and Duling techniques are also based on labelling polydisperse dextran fractions. The permeabilities of these fractions are not only dependent on the average size of the fraction used, but also critically dependent on the (non-characterized) grade of polydispersity.
  5. We have so far found no evidence that the (negatively charged) glycocalyx is specifically altered in, for example, ischaemia-reperfusion (I/R) injury, or in other pathophysiological situations. After 60 min of total renal artery occlusion and I/R injury, the functionally most important change of the glomerular filtration barrier was a reduced size-selectivity with only a very minor component of reduced charge selectivity [43], indicating that the negatively charged glycocalyx was only partly involved in these changes. In early microalbuminuric experimental diabetes mellitus in rats, we recently observed, using Ficoll as sieving marker, that the glomerular filter displayed a reduced size-selectivity with only secondary changes in charge selectivity, confirming a previous Ficoll sieving study by Andersen et al. [44]. Thus, the glomerular glycocalyx is not likely to be involved in these changes [45].
  6. If the recently proposed ‘glycocalyx-junction-break model’ of capillary permeability described by Weinbaum and Curry (reviewed in [46] and [47]), discussed by Flessner [1], were correct, then the size-selectivity of the small pore pathway would be accounted for by two separate barriers in series: first the glycocalyx at the luminal entrance of the interendothelial slit, and second, more downhill, the interendothelial junctional strands, which are its leaky ‘breaks’. This means that there would be a ‘protected region’ underneath the glycocalyx and above the plane of the junctional strands. This in turn implies that the Starling equilibrium would be nearly unaffected by the interstitial colloid osmotic pressure acting from the tissue side of the junctional strands because of limited permeation of albumin into the ‘protected’ region. If that were really true, then the action of i.p. colloidal osmotic agents, such as icodextrin or albumin, would be negligible. However, both albumin and icodextrin can produce considerable i.p. osmotic pressure and enhance fluid movement from plasma to the peritoneal cavity, which can be simulated by the TPM [4,10,48]. These facts seem to invalidate the glycocalyx-junction-break model of capillary permeability.



   Conclusions
Top
 Introduction
 Key physiologic features of...
 How does an endothelial...
 Conclusions
References
 
A moderately extended TPM, including the capillary wall and a serial barrier consisting of heterogeneously and loosely arranged fibres in the interstitium, can correct some of the apparent shortcomings of the two-dimensional (planar) TPM. Such a model has the potential of explaining how (tracer) macromolecules move from the peritoneal cavity into the tissues driven by hydrostatic pressure in the partial absence of net tissue (volume) changes, according to the so-called volume recirculation principle. Also, the pathophysiological changes occurring in long-term PD associated with submesothelial fibrosis and neoangiogenesis can be amply described by an extended TPM [3].

The fibre matrix theory, introduced nearly 30 years ago by Curry and Michel, remains an alternative to the pore theory for explaining the selectivity of small and large pores in the capillary wall of the TPM, while the ultrasmall pores of the TPM undoubtedly must be represented by AQP-1. If applying the fibre matrix model, the bimodal selectivity of the capillary wall requires the presence of a non-uniform fibre matrix, essentially showing a functional ‘two-pore’ behaviour. The endothelial glycocalyx seems, however, to be too delicate to offer the size-selectivity of a dense fibre matrix. In inflammation, there is ample evidence of large pore (gap) formation by focal endothelial cell retractions, as a consequence of rapid reorganization of the endothelial actin cytoskeleton. Over the last 30 years, these processes have been amply documented [46], whereas the potential permeability changes occurring due to alterations of the fragile endothelial glycocalyx have not been unambiguously demonstrated. The sieving properties of the glycocalyx demonstrated so far are confusing and not at all compatible with the well-documented sieving behaviour of microvascular walls to macromolecules [21,47,49]! However, in the renal microcirculation, the glycocalyx is likely to form a negatively charged gel-like structure with implications for the charge selectivity of the glomerular filtration barrier [28]. Functionally, a fibre matrix is not likely to be as dynamic as the endothelial cells in producing permeability changes in response to e.g. inflammation or I/R. Furthermore, there is good evidence that changes in colloid osmotic pressure in the peritoneal cavity by adding albumin or icodextrin [4,10,48] will markedly affect the capillary Starling equilibrium to produce an enhanced fluid movement from plasma to the peritoneum. Therefore, the hypothesis that the endothelial glycocalyx would be a size-selective primary barrier placed in front of the semi-leaky junctional strands in the clefts of the capillary endothelium must be put into serious doubt.

The glycocalyx is a poorly defined ‘fuzzy’ layer of cell-anchored proteoglycans and glycoproteins surrounding most cells of the body, including the peritoneal endothelial and mesothelial [50] cells. It is difficult to assess by conventional staining and fixation techniques, and its presence defies physiological testing. With its delicate ‘architecture’ is it not likely to represent a highly size-selective sieving barrier to uncharged macromolecules. Furthermore, why should the endothelial glycocalyx be of such importance and not the mesothelial glycocalyx? The incorporation of an endothelial glycocalyx as a separate capillary barrier in the TPM will neither enhance the ability of the TPM to predict peritoneal transport, nor will it deepen our understanding of the physiology and pathophysiology of peritoneal exchange.



   Acknowledgments
 
This study was supported by grants from the Swedish Medical Research Council (grant no. 08285) and the Swedish National Heart-and-Lung Institute (grant no. 20065026). Kerstin Wihlborg is acknowledged for skilful secretarial assistance.

Conflict of interest statement. None declared.



   References
Top
 Introduction
 Key physiologic features of...
 How does an endothelial...
 Conclusions
 References
 

  1. Flessner M. Distributed model of peritoneal transport: implications of the endothelial glycocalyx. Nephrol Dial Transplant (2008) doi: 10.1093/ndt/gfn055.
  2. Venturoli D, Rippe B. Transport asymmetry in peritoneal dialysis: application of a serial heteroporous peritoneal membrane model. Am J Physiol Renal Physiol (2001) 280:F599–F606.[Abstract/Free Full Text]
  3. Rippe B, Venturoli D. Simulations of osmotic ultrafiltration failure in CAPD using a serial three-pore membrane/fiber matrix model. Am J Physiol Renal Physiol (2007) 292:F1035–F1043.[Abstract/Free Full Text]
  4. Rippe B, Levin L. Computer simulations of ultrafiltration profiles for an icodextrin-based peritoneal fluid in CAPD. Kidney Int (2000) 57:2546–2556.[CrossRef][Web of Science][Medline]
  5. Rippe B, Venturoli D. Peritoneal transport kinetics with amino acid-based and glucose-based peritoneal dialysis solutions. Perit Dial Int (2007) 27:518–522.[Free Full Text]
  6. Simonsen O, Wieslander A, Venturoli D, et al. Mass transfer of calcium across the peritoneum at three different peritoneal dialysis fluid Ca and glucose concentrations. Kidney Int (2003) 64:208–215.[CrossRef][Web of Science][Medline]
  7. Rippe B, Venturoli D, Simonsen O, et al. Fluid and electrolyte transport across the peritoneal membrane during CAPD according to the three-pore model. Perit Dial Int (2004) 24:10–27.[Abstract/Free Full Text]
  8. Preston GM, Carroll TP, Guggino WB, et al. Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein. Science (1992) 256:385–387.[Abstract/Free Full Text]
  9. Ni J, Verbavatz JM, Rippe A, et al. Aquaporin-1 plays an essential role in water permeability and ultrafiltration during peritoneal dialysis. Kidney Int (2006) 69:1518–1525.[CrossRef][Web of Science][Medline]
  10. Rippe B, Perry MA, Granger DN. Permselectivity of the peritoneal membrane. Microvasc Res (1985) 29:89–102.[CrossRef][Web of Science][Medline]
  11. Rippe B, Haraldsson B. Capillary permeability in rat hindquarters as determined by estimations of capillary reflection coefficients. Acta Physiol Scand (1986) 127:289–303.[Web of Science][Medline]
  12. Pyle WK, Moncrief JW, Popovich RP. Peritoneal transport evaluation in CAPD. In: CAPD Update, Continuous Ambulatory Peritoneal Dialysis (1981) New York: Mason Publishing. 35–42.
  13. Waniewski J, Dutka V, Stachowska-Pietka J, et al. Distributed modeling of glucose-induced osmotic flow. Adv Perit Dial (2007) 23:1–5.[Medline]
  14. Vonesh EF, Rippe B. Net fluid absorption under membrane transport models of peritoneal dialysis. Blood Purif (1992) 10:209–226.[Web of Science][Medline]
  15. Rosengren B-I, Rippe A, Rippe C, et al. Transvascular protein transport in mice lacking endothelial caveolae. Am J Physiol Heart Circ Physiol (2006) 291:H1371–H1377.[Abstract/Free Full Text]
  16. Asghar RB, Davies SJ. Pathways of fluid transport and reabsorption across the peritoneal membrane. Kidney Int (2008) 73:985–986.[Medline]
  17. Parikova A, Smit W, Zweers MM, et al. Free water transport, small pore transport and the osmotic pressure gradient. Nephrol Dial Transplant (2008) doi: 10.1093/ndt/gfm768.
  18. La Milia V, Di Filippo S, Crepaldi M, et al. Mini-peritoneal equilibration test: a simple and fast method to assess free water and small solute transport across the peritoneal membrane. Kidney Int (2005) 68:840–846.[CrossRef][Web of Science][Medline]
  19. Venturoli D, Rippe B. Validation by computer simulation of two indirect methods for quantification of free water transport in peritoneal dialysis. Perit Dial Int (2005) 25:77–84.[Abstract/Free Full Text]
  20. Rippe B, Kamiya A, Folkow B. Transcapillary passage of albumin, effects of tissue cooling and of increases in filtration and plasma colloid osmotic pressure. Acta Physiol Scand (1979) 105:171–187.[Web of Science][Medline]
  21. Rippe B, Haraldsson B. Transport of macromolecules across microvascular walls. The two-pore theory. Physiol Reviews (1994) 74:163–219.[Abstract/Free Full Text]
  22. Rippe B, Haraldsson B. How are macromolecules transported across the capillary wall. News In Physiol Sci (1987) 2:135–138.[Abstract/Free Full Text]
  23. Chambers R, Zweifach BW. Capillary endothelial cement in relation to permeability. J Cell Comp Physiol (1940) 15:255–272.[CrossRef][Web of Science]
  24. Pappenheimer JR, Renkin EM, Borrero LM. Filtration, diffusion and molecular sieving through peripheral capillary membranes. A contribution to the pore theory of capillary permeability. Am J Physiol (1951) 167:13–46.[Free Full Text]
  25. Wissig SL. Identification of the small pore in muscle capillaries. Acta Physiol Scand Suppl (1979) 463:33–44.[Medline]
  26. Curry FE, Michel CC. A fiber matrix model of capillary permeability. Microvasc Res (1980) 20:96–99.[CrossRef][Web of Science][Medline]
  27. Luft JH. Fine structures of capillary and endocapillary layer as revealed by ruthenium red. Fed Proc (1966) 25:1773–1783.[Web of Science][Medline]
  28. Haraldsson B, Sorensson J. Why do we not all have proteinuria? An update of our current understanding of the glomerular barrier. News Physiol Sci (2004) 19:7–10.[Abstract/Free Full Text]
  29. Krediet RT, Koomen GC, Koopman MG, et al. The peritoneal transport of serum proteins and neutral dextran in CAPD patients. Kidney Int (1989) 35:1064–1072.[Web of Science][Medline]
  30. Rippe B, Axelsson J, Asgeirsson D, et al. Polysacharides are hyperpermeable across the glomerular filtration barrier, but not across the continuous peritoneal capillaries. Abstract SU-PO513. J Am Soc Nephrol (2007) 18:695A.
  31. Van Den Berg BM, Nieuwdorp M, Stroes ES, et al. Glycocalyx and endothelial (dys) function: from mice to men. Pharmacol Rep (2006) 58(Suppl):75–80.[Web of Science][Medline]
  32. Vink H, Duling BR. Capillary endothelial surface layer selectively reduces plasma solute distribution volume. Am J Physiol Heart Circ Physiol (2000) 278:H285–H289.[Abstract/Free Full Text]
  33. Grotte G. Passage of dextran molecules across the blood-lymph barrier. Acta Chir Scand Suppl (1956) 211:1–84.[Medline]
  34. Renkin EM, Garlick DG. Blood-lymph transport of macromolecules. Microvasc Res (1970) 2:392–398.[CrossRef][Medline]
  35. Renkin EM, Joyner WL, Sloop CH, et al. Influence of venous pressure on plasma-lymph transport in the dog's paw: convective and dissipative mechanisms. Microvasc Res (1977) 14:191–204.[CrossRef][Web of Science][Medline]
  36. Asgeirsson D, Venturoli D, Fries E, et al. Glomerular sieving of three neutral polysaccharides, polyethylene oxide and bikunin in rat. Effects of molecular size and conformation. Acta Physiol (Oxf) (2007) 191:237–246.[CrossRef][Medline]
  37. McDonald DM. Endothelial gaps and permeability of venules in rat tracheas exposed to inflammatory stimuli. Am J Physiol (1994) 266:L61–L83.[Web of Science][Medline]
  38. van Nieuw Amerongen GP, Natarajan K, Yin G, et al. GIT1 mediates thrombin signaling in endothelial cells: role in turnover of RhoA-type focal adhesions. Circ Res (2004) 94:1041–1049.[Abstract/Free Full Text]
  39. Rippe B, Grega GJ. Effects of isoprenaline and cooling on histamine induced changes of capillary permeability in the rat hindquarter vascular bed. Acta Physiol Scand (1978) 103:252–262.[Web of Science][Medline]
  40. Haraldsson B, Zackrisson U, Rippe B. Calcium dependence of histamine-induced increases in capillary permeability, studied in isolated rat hindlimbs. Acta Physiol Scand (1986) 128:247–258.[Web of Science][Medline]
  41. Folkow B, Neil E. Circulation (1971) New York, London, Toronto: Oxford University Press. 1–593.
  42. Zuurbier CJ, Demirci C, Koeman A, et al. Short-term hyperglycemia increases endothelial glycocalyx permeability and acutely decreases lineal density of capillaries with flowing red blood cells. J Appl Physiol (2005) 99:1471–1476.[Abstract/Free Full Text]
  43. Rippe C, Rippe A, Larsson A, et al. Nature of glomerular capillary permeability changes following acute renal ischemia-reperfusion injury in rats. Am J Physiol Renal Physiol (2006) 291:F1362–F1368.[Abstract/Free Full Text]
  44. Andersen S, Blouch K, Bialek J, et al. Glomerular permselectivity in early stages of overt diabetic nephropathy. Kidney Int (2000) 58:2129–2137.[CrossRef][Web of Science][Medline]
  45. Rippe C, Rippe A, Torffvit O, et al. Size and charge selectivity of the glomerular filter in early experimental diabetes in rats. Am J Physiol Renal Physiol (2007) 293:F1533–F1538.[Abstract/Free Full Text]
  46. Michel CC, Curry FE. Microvascular permeability. Physiol Reviews (1999) 79:703–761.[Abstract/Free Full Text]
  47. Curry FR. Microvascular solute and water transport. Microcirculation (2005) 12:17–31.[Web of Science][Medline]
  48. Zakaria ER, Rippe B. Osmotic barrier properties of the rat peritoneal membrane. Acta Physiol Scand (1993) 149:355–364.[Web of Science][Medline]
  49. Taylor AE, Granger DN. Exchange of macromolecules across the microcirculation. In: Handbook of Physiology. The Cardiovascular System. Microcirculation (1984) 4. Bethesda, MD: American Physiological Society. 467. section 2, chapter 11.
  50. Heldin P, Pertoft H. Synthesis and assembly of the hyaluronan-containing coats around normal human mesothelial cells. Exp Cell Res (1993) 208:422–429.[CrossRef][Web of Science][Medline]
Received for publication: 15. 1.08
Accepted in revised form: 23. 1.08


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nephrol Dial TransplantHome page
A. M. Gomes, M. P. Fontan, A. Rodriguez-Carmona, A. Sastre, H. D. Cambre, A. L. Muniz, and T. G. Falcon
Categorization of sodium sieving by 2.27% and 3.86% peritoneal equilibration tests--a comparative analysis in the clinical setting
Nephrol. Dial. Transplant., November 1, 2009; 24(11): 3513 - 3520.
[Abstract] [Full Text] [PDF]

Home page
 NDT PlusHome page
A. M. Coester, W. Smit, D. G. Struijk, and R. T. Krediet
Peritoneal function in clinical practice: the importance of follow-up and its measurement in patients. Recommendations for patient information and measurement of peritoneal function
NDT Plus, April 1, 2009; 2(2): 104 - 110.
[Abstract] [Full Text] [PDF]

Home page
 pdiHome page
B. Rippe
HOW TO ASSESS TRANSPORT IN ANIMALS?
Perit. Dial. Int., February 1, 2009; 29(Supplement_2): S32 - S35.
[Abstract] [Full Text] [PDF]

Home page
 Nephrol Dial TransplantHome page
O. Devuyst and E. Goffin
Water and solute transport in peritoneal dialysis: models and clinical applications
Nephrol. Dial. Transplant., July 1, 2008; 23(7): 2120 - 2123.
[Full Text] [PDF]

This Article
Right arrow Extract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/7/2147    most recent
gfn049v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Rippe, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rippe, B.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?