INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

EDEMA can occur within a few minutes after the onset of an acute inflammatory reaction in skin and develops when the capillary filtration rate exceeds the lymphatic drainage (3). The fluid flux across the capillary wall is determined by the transcapillary pressures; i.e., the interstitial fluid pressure (P_IF), the hydrostatic capillary pressure (P_C), the colloid osmotic pressure in plasma (COP_P) and the interstitium (COP_IF). Under normal conditions, P_IF acts to maintain a constant tissue fluid volume and counteracts edema formation (3) because an increased transcapillary filtration will raise interstitial fluid volume (IFV) and thereby P_IF, which in turn will act across the capillary to limit further filtration as well as enhancing lymph drainage from the tissues (3).

Contrary to this commonly accepted role for P_IF in maintaining constant IFV, we have observed a dramatic lowering of P_IF in the initial phase of several acute inflammatory reactions (6, 24, 25, 29, 37). The lowering of P_IF will increase the transcapillary filtration pressure (39). This observation has led to the development of the concept that the loose connective tissues, via P_IF, can be "active" in transcapillary fluid exchange as opposed to its commonly accepted role of being a "passive controller" in maintaining constant transcapillary fluid flux and thereby constant IFV (35). The cellular events evolved in the lowering of P_IF appears to be related to the properties of the extracellular matrix (ECM) macromolecules and the cell surface receptors toward ECM components, the integrins. The rationale for linking ECM molecules and interstitial cells to the lowering of P_IF and edema formation is primarily based on experiments in which cellular adhesion receptors toward ECM components (₁-integrins) are perturbed. Blockade of the ₁-integrins in the rat skin by the subdermal injection of anti-₁-integrin IgG causes a decrease of P_IF concomitant with edema formation (38, 41, 42), suggesting that the connective tissue cells can actively exert control of P_IF via their collagen-binding ₁-integrins (39). The control of P_IF is most likely achieved by a balance between the swelling properties of the hyaluronan-glycosaminoglycan gel (30) and the ability of connective tissue cells to physically exert tension on the collagen and microfibril network that restrain the swelling gel (30). According to this concept, the force required to exert tension on the ECM fibers is generated by the cytoskeleton and is transmitted across the cell surface to the ECM components. Integrins link ECM components with intracellular cytoskeletal components (9, 10, 22) and elicit intracellular signaling when clustered or ligand occupied (11, 34, 53). Integrins transduce mechanical information between the extracellular environment and the cell interior (44, 47) coupling ECM fibers and cytoskeletal elements mechanically via integrins or integrin-directed membrane complexes. It appears likely that alterations in cytoskeletal function should change the cellular tension on the ECM and thereby the P_IF according to the model presented above.

The aim of this study was twofold: 1) to investigate the effect on P_IF by selectively disrupting different parts of the cytoskeletal structure in connective tissue cells of dermal tissue and 2) to evaluate whether the cytoskeleton is turned over continuously in vivo, because experimental data have demonstrated that the cytoskeleton in vitro is continuously regenerated (48). However, this has to our knowledge not been previously investigated in vivo. The experiments were performed in two separate experimental groups by the investigation of the effect of the subdermal injection of the microfilament and microtubule-disrupting drugs on P_IF or edema formation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals

Measurements

Interstitial fluid pressure. The measurement of P_IF has been described previously (5, 52). Briefly, P_IF was measured by a micropuncture technique with the use of micropipettes with tip diameters of 3-7 µm, which were connected to a servo-controlled counterpressure system (50, 52). Punctures were performed on the dorsal side of the hind paw through intact skin with the use of a micromanipulator (Leitz; Heerbrugg, Switzerland) and under the guidance of a stereomicroscope (Wild M5; Heerbrugg, Switzerland). Pressure measurements were recorded in the following sequence: 1) with an intact circulation, and 2) for 90 min after circulatory arrest. In all experiments subdermal injections of test substances were performed directly after circulatory arrest. An average was obtained for each of the following time periods: 0-10, 11-20, 21-30, 31-45, 46-60, and 61-90 min. The test drugs were deposited subdermally (2 µl) using a 5-µl chromatography syringe (Hamilton 800 Series; Sutton, UK) with a 34-gauge needle (38). Measurements of P_IF were performed 0.2- to 0.5-mm below the skin surface and at the edge of the injected volume (5).

Total Tissue Water and Albumin Extravasation

Tissue samples. Tissue samples were obtained by removing the skin on the dorsal side of the hind paws. The samples were placed immediately in preweighed vials that were bottled and weighed as soon as possible. The wet weight of the tissue samples ranged from 0.1 to 0.2 g. Radioactivity was measured (see Albumin extravasation rate), and the samples were dried at 65°C for 2-3 wk until they reached a constant weight (usually 2-3 wk) to obtain the water content and dry tissue weight.

Total tissue water. Total tissue water (TTW) in the tissue samples was estimated as the water content per gram of dry tissue weight [(wet weight  dry weight)/(dry weight)].

Albumin extravasation rate. Albumin extravasation rate (E_ALB) was measured as the 25-min extravascular space of ¹²⁵I-labeled human serum albumin (¹²⁵I-HSA) (Institute for Energy Technology; Kjeller, Norway) in the same tissue samples that were used to obtain TTW. ¹²⁵I-HSA (0.05 MBq) in 0.3 ml of saline was administered intravenously immediately after the subdermal injection (2 µl) of the test substances. After a period of 25 min, ¹³¹I-HSA (0.05 MBq) in 0.3 ml of saline was injected intravenously, and 5 min thereafter blood samples were collected by cardiac puncture. The rat was killed by an intravenous injection of saturated potassium chloride. Tissue samples were obtained as described in Tissue samples. Radioactivity in tissue and blood samples was determined in a gamma-counting system (LKB Wallac 1285; Turku, Finland) with automatic background subtraction and spillover correction. Distribution volumes were calculated as plasma equivalent spaces (i.e., counts per minute per gram dry tissue weight divided by counts per milliliter plasma). E_ALB was calculated as the difference between the plasma equivalent distribution volume of ¹²⁵I-HSA and that of ¹³¹I-HSA. All calculations were made per gram dry tissue weight.

Experimental Groups

Group 1: P_IF. P_IF was measured as described above. All test substances were injected subdermally in a volume of 2 µl. Experiments were carried out using the following: 1) saline controls (n = 6): PBS; 2) DMSO controls (n = 6): 10% DMSO in PBS; 3) antimicrotubuli agents: colchicine at 3 mM (n = 6) or nocodazole at 0.33 mM (n = 6); and 4) antimicrofilamental agents: cytochalasin D in concentrations of 1.0 (n = 6), 0.4 (n = 6), or 0.12 mM (n = 6).

Group 2: TTW and E_ALB. Test substance (2 µl) was injected into the right paw, and the left paw was given 2 µl of either sterile PBS or 10% DMSO in PBS as a control. Experiments were carried out using the following: 1) saline controls (n = 6): 2 µl of 10% DMSO in PBS was injected subdermally into the right paw and 2 µl of PBS was injected into the left paw as a control. 2) Antimicrotubuli agents: 2 µl of colchicine (3 mM) were injected subdermally (n = 8), with PBS used as the control in the left paw. In a separate series (n = 5) 2 µl nocodazole (0.33 mM) were injected using 2 µl of 10% DMSO as a control in the left paw. 3) Antimicrofilamental agents: 2 µl of cytochalasin D (1 mM) were injected (n = 6) with the use of 2 µl of 10% DMSO as a control in the left paw.

Test Substances

Statistical Methods

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Group 1: P_IF

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Effect of subdermal injection of cytochalasin D on interstitial fluid pressure (PIF) in rat paw at decreasing concentrations. Also shown is DMSO control. Values are means ± SD. *P < 0.05 compared with DMSO control at same time points.

Group 2: TTW and E_ALB

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The cytochalasins constitute a group of fungal metabolites that readily permeate cell membranes in vivo and exert profound effects on cell shape, disrupt actin organization, and inhibit cell movements (16, 49). In vitro, these compounds inhibit actin polymerization by binding to the rapidly growing (barbed) end of F-actin filaments and block all association and dissociation reactions at these filament ends (7, 8, 12, 16). In this study we show that subdermal injections of cytochalasin D, but not nocodazole or colchicine, lead to a decrease of P_IF and rapid edema formation in rat paw skin.

Subdermal injections of cytochalasin D lowered P_IF. This is an effect that most likely relates to a specific action of actin. However, in our model, targets other than actin cannot be totally excluded. The following observations argue for specificity. First, the measured effect on P_IF is dose dependent. Second, the effect is rapid and is seen within 10-15 min after administration of the drug. Finally, the time required to obtain a measurable response on P_IF after cytochalasin D administration is in agreement with the in vitro observations where alterations in cellular processes can be observed within seconds or minutes after the drug challenge.

Colchicine and nocodazole are important drugs in the study of microtubule function. In vitro, the microtubuli display dynamic instability and undergo continual cycles of assembly and disassembly as a result of tubulin polymerization (31). Most microtubuli have half-lives in the order of minutes in most cell types studied. Colchicine inhibits tubulin polymerization into microtubules (1a), whereas nocodazole depolymerizes microtubuli (13). These two drugs inhibit the dynamic functions of the microtubule system. Microtubule-disrupting agents have been reported to increase the isometric tension exerted by fibroblasts cultured in a three-dimensional collagen lattice (26, 27). Although both nocodazole and colchicine inhibit cell elongation, these drugs do not inhibit formation and the extension of pseudopodia by fibroblasts cultured in collagen gels (46). As discussed below, connective tissue cells likely participate in the control of P_IF by exerting a tension on a collagen-microfibril network. According to this model for the control of P_IF, as well as taking into account the effects of microtubule-disrupting drugs on the tension exerted by cultured fibroblasts, one would expect that P_IF should increase after injection of colchicine and nocodazole in rat dermis. The finding that neither of these drugs had an effect on P_IF suggests that the connective tissue cell tension in rat dermis was not dependent primarily on the microtubule function. Furthermore, our data demonstrate that albumin extravasation and total tissue water content are not affected by colchicine and nocaodazole, suggesting that the endothelial barrier is not dependent on microtubule integrity.

P_IF was measured after circulatory arrest had been induced to eliminate the potential vascular effects caused by the different drugs that thereby limited transcapillary fluid filtration and prevented edema formation. This procedure did not affect P_IF for up to 90 min compared with control situations with intact circulation (52). Different drugs were injected subdermally, and P_IF was measured by micropipettes at the edge of the injected volume; i.e., about 2 mm from the center of the injection site (5). This implies that the different test substances become diluted as they diffuse into the surrounding tissue, which leaves some uncertainty about the effective concentration of the drugs in the area of measurement. This would explain the relatively high concentration of cytochalasin D (100-1,000 µM), which was required to give a measurable effect on the P_IF in our model system; i.e., around 100-fold higher than necessary to alter cellular processes in tissue culture systems (4, 19, 32, 46). A sphere with a volume of 2 µl has a radius of about 0.8 mm and doubling of the radius will raise the volume in the sphere eight times while the concentration in the sphere falls correspondingly, i.e., eight times. Because the measurements of P_IF were made approximately 2 mm from the center of injection, this will cause a dilution to <10% of that in the injectate. The same phenomenon has been observed with several other test substances, where the lowest effective concentration in vivo has been between 10 and 100 times larger than in cell culture experiments (5, 40, 42).

Under normal conditions P_IF acts to maintain normal interstitial volume and to counteract edema formation (3). An increased fluid flux large enough to increase interstitial fluid volume will, depending on the tissue compliance, eventually raise P_IF and thereby restrict further fluid filtration to the tissue. This mechanism is known to be one of the "safety factors" against edema formation. Contrary to this commonly accepted role for P_IF in the normal control of interstitial volume, our observations of lowering of P_IF concomitant with edema formation during acute inflammations demonstrate that the tissue can "actively" enhance capillary filtration and eventually cause edema formation. The lowering of P_IF will provide a driving force only for the initial and rapid part of the edema formation, because once the inflammatory edema is sufficiently large, P_IF will become positive, as a function of the tissue compliance, and counteract further fluid filtration. Maintenance of the edema formation will thereafter be due to other factors like increased capillary permeability together with increased hydrostatic pressure.

Several series of experiments suggest that the final events resulting in the lowering of the P_IF involve connective tissue cells and structural components of the connective tissue (39). The evidence for involvement of structural components as well as participation of the connective tissue cells is based on in vivo studies of P_IF and edema formation in dermal tissue and fibroblasts cultured in vitro. Experiments by Meyer (30) showed that pieces of isolated loose connective tissue placed into an isotonic buffer had a tendency to swell. This swelling phenomenon has been attributed to the tissue content of hyaluronan (30). Our previous findings (38, 42) that subdermal injection of anti-₁ integrin IgG induces edema formation concomitant with lowering of P_IF, together with observations of the ability of fibroblasts to compact a three-dimensional collagen lattice in vitro, form the basis for proposing a mechanical model for the cellular control of the P_IF by connective tissue cells (43). Platelet-derived growth factor (42) and phosphatidylinositol-3' kinase (1) have been shown to be involved in the control of P_IF as well as influence the collagen gel contraction, which provides further evidence for the cellular control of P_IF and thereby tissue fluid homeostasis. According to this model, the lowering of P_IF occurs after connective tissue cells loosen their attachment on either a collagen or a microfibril network, which constrains a hyaluronan-proteoglycan gel with an intrinsic tendency to swell (30). When the restraining force is released in vivo, the tissue will swell, but the initial lack of excess fluid will restrict swelling, and P_IF will decrease until the tissue expansion is balanced by the negative P_IF and the stress in the fiber networks. Our observation of a decrease of P_IF following subdermal injections of cytochalasin D further substantiates this hypothesis in that this substance disrupts the actin filaments of the cytoskeleton and thereby limits the ability of the cells to generate enough force to exert a tension on the extracellular fiber network. In analogy with the effect of cytochalasin D to lower of P_IF in skin is that this substance also inhibits collagen gel contraction in vitro (19).

The control of P_IF in vivo shares characteristics with fibroblast-mediated collagen contraction in vitro (35, 43). Fibroblasts cultured in three-dimensional collagen gels send out long extensions and generate traction forces on the collagen fibers within the gel leading to the compaction of the latter (15, 18, 20). The actin-based microfilament system is highly dynamic, and the sending out of cell extensions is dependent on this turnover of actin filaments (21). The fact that cytochalasin D induces a lowering of P_IF indicates that a dynamic assembly and disassembly of actin filaments also take place in dermal cells in vivo. This phenomenon has to the best of our knowledge not been reported before. Our data support the concept that connective tissue cells generate traction forces, which are important for normal tissue integrity.

The edema-generating effect of subdermal injections of the different drugs, acting on the cytoskeleton, were tested in animals with an intact circulation. Cytochalasin D injected subdermally was the only drug that induced edema. Edema occurs when transcapillary fluid filtration exceeds lymph drainage (3). In rat skin visible edema requires IFV to increase by 50-100% above that of the control (51). Transcapillary fluid filtration flux (J_V) is the product of capillary hydraulic conductivity (CFC) and the net filtration pressure across the capillary (P) (3). Under steady-state conditions P has been calculated to be ~0.5 mmHg in peripheral tissues (3). Because IFV in the rat skin is turned over in 12-24 h (36), the appearance of edema in 10-15 min requires that J_V be increased by 50-100 times above control. During acute inflammatory conditions, CFC increases two to three times above that of the control, even in severe tissue injuries such as burn injuries (2, 14, 33) and is therefore not sufficient to explain the rapid edema formation under these conditions. The lowering of P_IF by 2.0 mmHg, within 20 min after injection of cytochalasin D, adds directly to a P of 0.5 mmHg (3) and increases the net filtration pressure by a factor of four above the control. Cytochalasin D also appears to raise the capillary permeability to macromolecules as protein accumulation (E_alb) increased significantly from 0.05 in control to 0.40 ml/g dry wt. Capillary permeability is quantitatively described by the osmotic reflection coefficient () and the permeability-surface area product PS (45). The reflection coefficient is estimated at high transcapillary fluxes when diffusion becomes negligible as  = 1  C_L/C_P (17) where C is protein concentration in lymph (L) and plasma (P). However, if this requirement is not fulfilled, another estimate can be obtained from the ratio of C_L/C_P; i.e., the ratio of albumin accumulated divided by the fluid accumulated (increase in TTW). From the data in Table 1, C_L/C_P after cytochalasin D is 0.35/0.73 = 0.48 when E_ALB increases sevenfold; i.e., from 0.05 to 0.40 ml/g dry wt in 25 min. However, C_L/C_P is normally 0.5 (3, 45), and after cytochalasin D, the unchanged value at seven times water flux increase clearly attests to an increased capillary protein permeability. The reflection coefficient also contributes to transcapillary fluid flux because the effective osmotic pressure across the capillary wall is a product of  and colloid osmotic pressure differences. By assuming that the transcapillary fluid flux is increased to an extent where diffusion is negligible,  equals 0.51 and raises the net filtration pressure by 4 mmHg, assuming a transcapillary oncotic pressure of 10 mmHg. The reduction of  is in agreement with experiments where cytochalasin D was given intravenously and produced a corresponding decrease in  (28). This suggests that the edema formation induced by cytochalasin D is related in part to an increased microvascular permeability (28). Thus the initial edema formation seen after subdermal injection of cytochalasin D is both due to a lowering of P_IF and an increased microvascular permeability. However, as edema develops and IFV increases, P_IF will eventually rise and counteract further edema formation, and maintenance of edema must rely on increased capillary pressure and permeability (3).

In summary, the present findings are important for two reasons. The observation that cytochalasin D induces lowering of P_IF in the dermal skin provides important and new information on the state of the cytoskeleton in vivo by demonstrating that actin monomers both associate and dissociate from the ends of actin filaments in vivo. These experiments also provide additional support to the hypothesis that interstitial matrix components in dermal tissue are normally held under tension by connective tissue cells (23), and that this tension must be generated by the contractile apparatus in the cells.