Transport of macromolecules across microvascular walls: the two-pore theory

In this review we summarized the evidence favoring the concept that the major plasma proteins are passively transported across vascular walls through water-filled pathways by means of convection and diffusion. With regard to solute transport, a majority of microvascular walls seems to show a bimodal size selectivity. This implies the presence of a high frequency of functional small pores, restricting proteins, and an extremely low number of non-size-selective pathways, permitting the passage of macromolecules from blood to tissue, here denoted large pores. We discussed the general behavior of such a heteroporous system. A major consequence of two-pore heteroporosity is that large-solute transport must mainly occur due to convection through large pores at low filtration rates, that is, at normal or even zero lymph flows. Indeed, convection must be the predominating transport mode for most solutes across large pores when the net filtration rate is zero. Under these (transient) conditions, the convective leak of macromolecules across large pores will be counterbalanced by absorption of essentially protein-free fluid through protein-restrictive pores. In a heteroporous membrane, proteins can thus be transported by solvent drag across vascular walls in the absence of a net convection. Normally the steady-state transcapillary fluid flow (lymph flow) is about equally partitioned among small and large pores, which makes lymph essentially a "half and half" mixture of protein-free ultrafiltrate and plasma. With increasing fluid flows, however, the plasma filtrate will be progressively diluted, eventually reaching a protein concentration largely in proportion to the fractional hydraulic conductance accounted for by the large pores (alpha L). Under these high lymph flow conditions, not only the large-pore transport but also the small-pore transport (of smaller macromolecules) will become convective. At low lymph flows, however, the small-pore transport of smaller macromolecules is usually mostly diffusive. An important implication of capillary heteroporosity is that single-pore formalism is inadequate for correctly evaluating the capillary sieving characteristics. With the use of homoporous transport formalism, the "lumped" macromolecular PS and sigma will therefore vary as a function of transcapillary fluid flow (Jv). However, it is approximately correct to use single-pore formalism for conditions when Jv is very high during steady state. Thus, if minimal sieving coefficients can be measured for macromolecules, then these values will accurately reflect (1 - sigma).(ABSTRACT TRUNCATED AT 400 WORDS)