| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Section for Physiology, Department of Biomedicine, University of Bergen, Bergen, Norway
Submitted 19 August 2005 ; accepted in final form 3 November 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Vi/Pif) of 0.012 ml·g wet wt–1·mmHg–1. Interstitial fluid was isolated, and interstitial fluid colloid osmotic pressure averaged 10.4 mmHg (SD 1.2), which is 64% of the corresponding level in plasma. We conclude that Pif drops during secretion and, thereby, increases the net transcapillary pressure gradient, a condition that favors fluid filtration and increases the amount of fluid available for secretion. The reduction in Pif is most likely induced by contraction of myoepithelial cells and suggests an active and new role for these cells in salivary secretion. The relatively low interstitial compliance of the organ will enhance the effect of the myoepithelial cells on Pif during reduced Vi.
interstitial fluid volume; micropuncture; pilocarpine
3 to 30 ml·min–1·100 g–1 during maximal parasympathetic stimulation (22). During salivation, gland volume, as well as interstitial fluid pressure (Pif), has been reported to decrease (15, 22). The decrease in tissue volume was suggested to reflect expulsion of preformed saliva from the salivary ducts caused by contraction of the myoepithelial cells and/or reduction of cellular and/or interstitial fluid volume (Vi) in the initial phase of secretion after onset of parasympathetic stimulation (15). The 10-fold increase in fluid flux across the capillaries during activation of parasympathetic nerves to the salivary glands demands dramatic changes in the Starling forces that regulate fluid flow across the capillary walls.
Our aim was to obtain a better quantitative estimate of the factors involved in interstitial fluid and saliva transport, and this led us to study the Starling forces, with our focus on the interstitial side. Transcapillary fluid transport is described by the so-called Starling equation
 | (1) |
is osmotic reflection coefficient to plasma proteins, COPp and COPif represent colloid osmotic pressure (COP) in plasma and interstitium, respectively, and L is lymph flow. For the salivary gland, many of these parameters of the Starling equation have been determined (for review see Ref. 21); recently, we used micropipettes to measure Pif in the salivary gland. COPif is, however, unknown in this organ, because the interstitial fluid is difficult to access. Recently, we showed in tumors, another tissue where interstitial fluid is not readily available, that interstitial fluid can be isolated by exposure of the tumor tissue to centrifugation at less than 400 g (25). Exclusion of wicks and microdialysis as potential methods to gain access to salivary gland interstitium because of the more traumatic nature of these techniques led us to test whether a centrifugation method similar to that used in our recent experiment (25) could also be applied to isolation of salivary gland interstitial fluid and, thereby, enable us to measure COPif.
Pif, another determinant of transcapillary fluid flux, is influenced by the interstitial compliance (C) of the organ, defined as the change in Vi divided by the corresponding change in Pif, i.e., C = Vi/Pif. In a low-compliant organ, a change in vascular volume (Vv) or/and Vi will induce a change in Pif. In a previous study, we observed immediate changes in Pif after changes were induced in vascular perfusion in rat submandibular gland (4). The changes induced in Vv and the concomitant change in Pif were interpreted to be a consequence of a relatively low compliance in the submandibular gland. Because the compliance may be important for fluid secretion, our aim was to determine Vi and Pif in secreting and nonsecreting glands. In nonsecreting glands, we changed the forces acting across the capillary wall by altering the tissue hydration systemically, whereas we administered pilocarpine, which is known to act via muscarinic receptors in the organ, to induce secretion locally in the salivary gland. We were especially interested in the possible mechanistic role of Pif in the initial phase of saliva secretion. Here we demonstrate that a reduction in Pif contributes significantly to the net filtration pressure, leading to salivation, and that samples representative for interstitial fluid can be isolated by centrifugation of the salivary gland. We also propose a novel mechanistic role for the myoepithelial cells in saliva secretion.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Fluid Distribution Volumes and Pif
The rats were studied in a supine position. A femoral vein was catheterized for injection of supplemental anesthesia and a femoral artery for continuous systemic blood pressure [arterial pressure (PA)] recordings with a Gould pressure transducer and recorder. The submandibular gland was isolated and placed in a cup for immobilization just before Pif measurements. In the cup, the salivary gland was flushed with saline at 37°C to keep the surface moist. Fluid was drained from the bottom of the cup.
Micropuncture Measurements of Pif
Under stereomicroscopic guidance, Pif was measured with sharpened glass micropipettes (2- to 6-µm-diameter tip) filled with Evans blue-dyed 0.5 M NaCl. The micropipette was connected to a servo-controlled counterpressure system, as first described by Wiederhielm et al. (23a), and advanced 0.5–1 mm into the salivary gland tissue with a Leitz-Wetzlar micromanipulator. The pipette was inserted through the intact thin capsule of the frontal/ventral aspect of the submandibular gland. Pif was recorded with a transducer (model 1280C, Hewlett Packard) connected to a Gould amplifier and recorder. The transducer was calibrated before each experiment, and zero pressure at the level of the submandibular gland was checked repeatedly by placement of the micropipette in the saline covering the gland. A recording of Pif was accepted when the following criteria were fulfilled: 1) the feedback gain of the system could be altered without interference with the measured pressure; 2) application of pump suction increased the resistance across the glass pipette, indicating fluid communication between the pipette and the interstitium; 3) the zero pressure level remained unchanged during a recording; and 4) a small amount of Evans blue dye was observed in the tissue after measurement, showing that the pipette was not placed intravascularly.
Fluid Volume Measurements
To eliminate interindividual differences, volume was measured in the same animal in the control situation in one submandibular gland and after changes in hydration or after pilocarpine infusion (see below) in the contralateral gland. The selection of right or left submandibular gland for control or experimental condition was alternated. After anesthesia and placement of catheters, both kidney pedicles were ligated via flank incisions, and 51Cr-EDTA (60–70 µCi) was injected intravenously for measurement of Vi. After 90 min of tracer equilibration, control Pif was measured. In rats where hydration changes were induced, a 100-µl blood sample was withdrawn from the femoral artery into a heparinized syringe; then the vessels to the submandibular gland were ligated, and the control gland was removed.
After an equilibration period following the induced change in hydration or 2–3 min before pilocarpine infusion (see Experimental Protocol), 3–4 µCi of 125I-human serum albumin (HSA) were injected and allowed to circulate for 5 min. A final 0.5- to 0.7-ml blood sample was obtained from the arterial catheter, and the rat was killed with an overdose of intravenously administered anesthetic. The remaining submandibular gland was removed. Tissue samples were placed in tared, covered vials, weighed, and then counted in a gamma counter (model 1282 Compugamma, LKB) with window settings of 15–75 keV for 125I and 290–350 keV for 51Cr. Counts were corrected for background and spillover.
Measurement of Vv. Our experimental protocol using each animal as its own control required special attention regarding measurements of plasma volume. Labeled 125I-HSA will gradually leak from the vascular compartment, and tracer injected at the end of the control period cannot be used to measure Vv at the end of the experimental period. Because Vv is small relative to Vi and may be assumed to be fairly stable under control conditions, we measured Vv (and Vi) in a separate series of experiments and used these volumes to estimate Vi in the control situation in the groups of animals subjected to over- or dehydration. Thus we used no intravascular tracer for the control situation in the latter animals and included 10 rats as controls for fluid distribution volume measurements in the submandibular gland (basal group). The mean Vv as percentage of total extracellular fluid volume (Vx) in the submandibular gland estimated in the basal group of rats was used for calculation of Vv in the control situation before over- or dehydration.
Possible Transport of Isotopes to Saliva
In three animals, a polyethylene catheter was inserted into the submandibular gland duct and ligated before infusion of 51Cr-EDTA to test whether the tracer was transported across the epithelial layer in the gland and into saliva. The tubes were removed after the animals were killed and placed in vials for counting.
Experimental Protocol
Systemic dehydration (group 1).
Dehydration was induced by peritoneal dialysis using a hypertonic glucose solution in seven rats. After measurements of Pif and sampling of tissue in the control situation, 10–15 ml of 20% glucose in Ringer acetate (Fresenius-Kabi, Oslo, Norway; pH 6.0) that had been preheated to 37°C were instilled via a catheter introduced into the peritoneal cavity from the lumbar region. After an equilibration period of 45 min, the dialyzing fluid was withdrawn and the dialysis was repeated. This procedure was used to remove a net volume of 5–15 ml from each rat. During dialysis, repeated intravenous infusions of 2 ml of preheated 10% HSA in Ringer acetate solution (
6 ml) were given in an attempt to prevent hypotension and induce additional interstitial dehydration. Measurements of Pif were started in the remaining experimental submandibular gland 1 h after withdrawal of the last sample of dialyzing fluid and were finished within the following hour. Then 125I-HSA was injected, and blood and tissue were sampled for volume determination as described above.
Systemic overhydration (group 2). Previous studies of the volume-pressure relation in skin and muscle of rats suggested a linear relation between Vi and Pif at low degrees of overhydration followed by no increase in Pif as the overhydration increases (20, 28). A gradual overhydration was therefore induced by volume loading. After measurements of Pif and sampling of tissue in the control condition, we infused 20, 25, and 35 ml of Ringer acetate solution that had been preheated to 37°C through the femoral vein catheter. The infusions lasted 15, 30, and 60 min, respectively. Measurement of Pif was initiated 90 min after the end of infusion and was completed within 1 h. After pressure measurements, 125I-HSA was injected for plasma volume determination, and plasma and tissue were sampled for volume measurements as described above.
Active secretion (group 3). Pilocarpine (1 mg/kg iv dissolved in 0.1 ml saline) was administered to 11 rats as bolus infusions immediately after injection of 125I-HSA. Pif was measured continuously in the experimental submandibular gland before, during, and after pilocarpine administration. We chose this agent to induce salivation because of its ability to induce maximal secretion even without activation of all the available receptors (3). In three of the rats, saliva produced during stimulation was collected from the oral cavity with micropipettes, placed in vials, and assayed for radioactivity. Submandibular glands collected for fluid volume measurements were dried after gamma counting to determine wet-to-dry weight ratio.
In four of the rats in this group and in one rat not given isotopes, local changes in red blood cell flux were measured (as an index of blood flow in the submandibular gland) during the experimental procedure with a laser-Doppler flowmeter (Periflux model 4001 Master, Perimed, Järfalla, Sweden) equipped with a needle probe (model PF 415:10; 125-µm fiber diameter with 500-µm separation). The laser probe was positioned with a micromanipulator above the area for Pif recordings and rotated to the position at which the largest resting blood flow signal [in arbitrary perfusion units (PU)] was measured. Calibration was performed using a motility standard giving a signal output of 250 PU. Zero blood flow was determined as the value recorded with the probe positioned at the submandibular gland after heart arrest. The flowmeter’s time constant was set at 0.03 s, and upper and lower bandwidths were 20 kHz and 20 Hz, respectively.
Isolation of Interstitial Fluid and Measurements of COP
Centrifugation technique.
In this subset of experiments (n = 12 rats), our aim was to isolate interstitial fluid from the salivary gland to enable us to estimate COPif using an approach similar to that described previously for tumors and skin (25). After anesthesia, blood was sampled by cardiac puncture and the rats were killed by an intravenous injection of saturated KCl. Immediately after euthanasia, with the capsule remaining intact, the submandibular glands were isolated and transferred to 2-ml centrifuge tubes provided with a nylon mesh (15- to 25-µm pore size) basket, with the apical part of the gland orientated downward in the tube. To estimate potential contamination in the centrifugation of fluid from cells, we determined the concentration ratio of the extracellular tracer 51Cr-EDTA in submandibular gland fluid to that in plasma isolated from six rats. Furthermore, we used the vascular tracer 125I-HSA to determine the plasma contribution to the sample. After anesthesia and placement of a PE-50 catheter in a femoral vein, both kidney pedicles were ligated via flank incisions, and 240 µCi of 51Cr-EDTA (Nycomed-Amersham, Buckinghamshire, UK) were injected intravenously for distribution in the interstitial fluid. After 90 min of equilibration, 3–4 µCi of 125I-HSA were injected and allowed to circulate for 5 min before blood was sampled for isolation of plasma, and the rat was killed. The submandibular glands were removed, transferred to centrifuge tubes, and handled as described above. Fluid isolated from the submandibular gland in these experiments was collected in microcapillaries for exact volume measurements and transferred to vials for gamma counting.
Initially, using a procedure described previously, we explored centrifugation rates of 68–955 g (800–3,000 rpm) for 10 min. At 
663 g (2,500 rpm), massive hemolysis and blurring of the centrifugate occurred; these samples were discarded. At 663 g, the isolated fluid was, with a few exceptions, clear; however, in all experiments using a centrifugation time of 10 min, we found an increase in the concentration of interstitial fluid, as evidenced by a 51Cr-EDTA tissue fluid-to-plasma concentration ratio significantly >1.0, suggesting that the tissue fluid became concentrated during the isolation procedure (see below). We then turned to an alternative centrifugation protocol, in which the tissue was subjected to 424 g for 5 min. Because we suspected that swelling of the metabolically active salivary gland cells could increase the concentration of substances dissolved in the extracellular fluid, we avoided all possible delay in the isolation procedure before the centrifugation. Thus <10 min elapsed from the time the animal was killed until fluid was isolated.
Fluid that accumulated at the bottom of the tubes after centrifugation was collected in graduated glass microcapillary tubes for volume measurement in a humidity chamber. The fluid was removed from the microcapillary tubes for COP measurements or diluted in buffer for HPLC. Samples that were visually contaminated with blood (<5% of total) were discarded.
To determine whether the connective tissue sheath ("capsule") surrounding the submandibular gland represented a barrier to plasma protein, leading to sieving of interstitial fluid and an underestimation of COPif, we added a series of experiments (n = 3) in which fluid was isolated after "decapsulation" of the submandibular gland. After anesthesia, one gland was isolated with the capsule intact as described above, whereas in the contralateral gland the thin connective tissue sheath surrounding the gland was removed by careful blunt dissection. Both glands from one animal were centrifuged simultaneously, and the fluid was isolated and handled as described above.
HPLC of Interstitial Fluid
The distribution of macromolecules in fluid isolated by centrifugation from the submandibular gland and in plasma was determined by HPLC using a Superose 12 HR 10/30 exclusion column (Pharmacia-Biotech, Uppsala, Sweden) with an optimal separation range of 10–300 kDa, as described in detail elsewhere (25). The pherograms were compared with relevant plasma protein standards.
COP
COPif and COPp were measured in a colloid osmometer designed for submicroliter samples (26) using membranes with a 30-kDa cutoff.
Calculations
Distribution volumes.
Distribution volumes were calculated as the plasma equivalent distribution volumes of the tracers [counts given as corrected counts per minute (cpm)] with the assumption that 51Cr-EDTA will distribute in the extracellular fluid phase and labeled HSA will distribute only in plasma. Vv in a tissue sample was calculated as the 5-min 125I-HSA distribution volume
 | (2) |
Because 125I-HSA has been in the animal for only 5 min, extravasation can be assumed to be negligible.
Tissue Vx was calculated as the distribution volume of 51Cr-EDTA before and after experimental intervention
 | (3) |
Vi was the difference between Vx and plasma volume (Vv)
 | (4) |
Interstitial compliance. Interstitial compliance was calculated from the relation between Vi and Pif during changes in hydration. In regular regression analysis, it is assumed that there are errors only in the independent variable; therefore, the coefficient of regression will depend on which parameter is chosen as the independent variable. Because we cannot know which parameter is independent, the regression coefficients are given as the geometric mean obtained when Vi is used as the independent, as well the dependent, parameter in the regression analysis, as described by Brace (5).
Statistics
Values are means (SD), and differences within a group were tested with paired t-tests. P < 0.05 was considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The volumes and pressures obtained in the basal group were used for determination of Vv under control conditions, i.e., no stimulation or secretion or change in hydration. Vx averaged 0.26 ml/g wet wt (SD 0.05) with a corresponding Vv of 0.03 ml/g wet wt (SD 0.02), resulting in an average Vi of 0.23 ml/g wet wt (SD 0.01). The mean Pif in this group was 3.0 mmHg (SD 1.3). Neither 51Cr-EDTA nor 125I-HSA was detected in saliva sampled with cannulas in the submandibular gland ducts, showing that no tracer crossed the epithelial layers in the glands.
Volume and Pressure During Systemic Changes in Hydration
Peritoneal dialysis (group 1), which was performed to reduce Vi, reduced mean Vx and Vv to 70% of control values (Table 1). As expected, we observed a significant drop in mean systemic blood pressure averaging 20 mmHg (P < 0.05; Table 1) and a variable increase in heart rate (data not shown), despite infusions of hyperoncotic HSA. The amount of fluid that was removed by dialysis showed large variations between the rats: 8–20.5 ml (median 15 ml). The net volume removed was independent of dialysis volume; therefore, the results from 15 and 20 ml of dialysis volume have been pooled (Table 1).
|
Corresponding values of local Vi and Pif in the control condition and after over- and dehydration, as well as changes in volume and pressure in the control condition, are shown in Fig. 1. The induced gradual reduction in Vi resulted in a gradual reduction in Pif, and there was a linear relation between the reduction in Vi and Pif. The maximal reduction in Pif was 4 mmHg after Vi was reduced by 0.08 ml/g. Opposite changes in Pif were observed when the Vx was increased by infusion of Ringer acetate solution. An increase in Vi exceeding 20% of the control volume did not result in a further rise in pressure (Fig. 1A). The maximal increase in Pif was 4.5 mmHg after an increase in Vi of 0.11 ml/g. There was a linear relation between volume and pressure during overhydration for the first 20% increase of Vi above control as well as for dehydration (Fig. 1B).
|
Volume and Pressure During Active Saliva Secretion
Pilocarpine bolus infusions caused immediate responses in systemic pressures. Thus the mean systemic pressure dropped from 99 mmHg (SD 21) in the control condition to 51 mmHg (SD 30) shortly after infusion before it stabilized at an elevated level [121 mmHg (SD 39)] within 40 s after infusions (Fig. 2). The increase in pressure was interpreted as a baroreceptor effect caused by the initial drop in systemic pressure. In the subgroup of rats where blood flow was measured (n = 4), pilocarpine resulted in an increase from 132 PU (SD 30) to a maximal level of 284 PU (SD 25), corresponding to a 97% (SD 2) increase in flow. The blood flow increase was immediate and remained elevated until the submandibular gland was removed for fluid volume measurements. In one rat, the glands were not removed, and measurements were monitored until blood flow and Pif returned to control levels after 15 min.
|
|
The wet-to-dry weight ratio was 4.23 (SD 0.17) for the experimental submandibular glands (n = 5) and 4.02 (SD 0.16) for the control glands (P = 0.11). These data, along with the data for Vi (see above), show that salivation did not result in changes in interstitial or cellular hydration at the time of gland collection after pilocarpine infusion. To ensure that saline flushing during Pif measurements did not influence the distribution volume in the experimental submandibular gland, we compared the Vx values from the contralateral control side (without Pif measurements) in this group of animals with those in the basal group (with Pif measurements) and found that they were not significantly different (P = 0.44).
Isolation of Interstitial Fluid and COP Measurements
As described in METHODS, we measured the recovered centrifugate tracer (51Cr-EDTA) that had equilibrated in extracellular fluid after exposure of the tissue to 424 g for 5 min. At 424 g for 5 min, the 51Cr-EDTA fluid-to-plasma concentration ratio was 0.99 (SD 0.07, n = 6), which is not significantly different from 1.0, indicating no concentration or dilution of the interstitial fluid. The corresponding ratio for 125I-HSA was 0.07 (SD 0.04). No other combinations of centrifugation speed and time resulted in a ratio for 51Cr-EDTA closer to 1.0; therefore, this combination of speed and time was chosen for isolation of interstitial fluid. The isolated fluid was clear and straw colored, and the volume was usually 2 µl.
COPif averaged 10.4 mmHg (SD 1.2), and COPp was 16.7 mmHg (SD 2.0, n = 6). Thus COPif corresponded to 64% of COPp.
The connective tissue sheath surrounding the submandibular gland did not result in any sieving of plasma proteins on centrifugation, as evidenced by an average COP in the centrifugate isolated from decapsulated glands of 11.0 mmHg (SD 0.6, n = 3), which is not different from the corresponding pressure of 11.2 mmHg (SD 0.6, n = 3) in fluid isolated from glands with intact capsule. COPp in these rats averaged 20.0 mmHg (SD 0.6, n = 3).
HPLC
An elution pattern from HPLC of interstitial fluid from a submandibular gland is shown in Fig. 4B; the elution pattern of rat plasma is shown for comparison in Fig. 4A. The pattern for interstitial fluid closely resembles that of plasma. There was, however, a lower fraction of high-molecular-weight substances in interstitial fluid, suggesting a higher albumin-to-globulin ratio. There was also a small fraction of low-molecular-weight substances in interstitial fluid that was not found in plasma. The HPLC also showed that the interstitial fluid was not contaminated with saliva, which would be expected to give a large peak in the fractions eluting before albumin, represented by amylase.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Methodological Aspects
Pif is a central parameter in this study. Pif can be described as the pressure in a saline-filled column brought in contact with the interstitium (1, 9). We chose the micropipette technique to measure Pif, because this method is practically atraumatic and has been validated in a number of studies (for references see Ref. 24). As pointed out previously (4), we cannot rule out the possibility that the micropipette tip could be placed intracellularly in some recordings. Nevertheless, if some of our recordings should reflect intracellular pressure, rather than Pif, the pressure will likely equal Pif, because hydrostatic pressure gradients are unlikely across the compliant plasma membrane of cells (18). The absence of a gradient between interstitial fluid and parenchymal cells is of particular importance during peritoneal dialysis, which will result in cellular dehydration of most organs, including the salivary gland, because of the high blood glucose (27). The wick-in-needle approach has also been used to measure Pif (23), but this method is obviously more traumatic than the micropipette technique. The pressure in the subcapsular space of canine submandibular gland has been measured by introduction of a microtipped transducer into this space (17), but whether this represents the pressure in the interstitial fluid per se or whether an element of total tissue pressure is included (9) is not known. In our opinion, the pressure recorded by micropipettes will give the best estimate of Pif and, thereby, be the relevant parameter for fluid balance studies in control conditions as well as during acute perturbations.
We were also able, for the first time, to isolate fluid representative of submandibular interstitial fluid and, thereby, estimate another previously unknown parameter in the Starling equation for this organ. On the basis of previous experience (25), we exposed the salivary gland to increased centrifugal force to isolate fluid and were able to validate that it is representative of interstitial fluid. One of the assumptions used in this validation is that a tracer equilibrated in the extracellular fluid is not diluted or concentrated compared with plasma. In our interpretation, a reduced concentration would indicate that fluid without tracer (i.e., intracellular fluid from erythrocytes or other cells or tracer-free saliva) had been added to the interstitial fluid and, conversely, that the tracer had been concentrated as a result of cell swelling. Indeed, using a protocol with consecutive centrifugations similar to a procedure used in a previous study (25), we found a higher tracer concentration in the isolated fluid than in the plasma, indicating that cell swelling had occurred. Assuming that such swelling occurred as a linear function of time, we reduced the time available for potential swelling by reducing the isolation time to a minimum and using one centrifugation speed. We then found an interstitial fluid-to-plasma ratio of extracellular tracer of 0.99, which is not significantly different from 1.0, suggesting no increase in concentration or dilution of the isolated fluid. Another potential problem would be sieving of plasma proteins by the thin connective tissue sheath surrounding the submandibular gland, but such sieving was ruled out by the experiments showing similar COPif from intact and decapsulated glands. The pherograms showed a higher albumin-to-globulin ratio for salivary gland interstitial fluid than for plasma, which could have been the result of sieving of proteins during passage of fluid through the gland tissue. This observation is, however, most likely a consequence of the known sieving occurring at the capillary wall, because a similar increased albumin-to-globulin ratio has been found also in fluid isolated using wick techniques in skin and muscle interstitial fluid (25, 30). Taken together, these observations suggest that we have presented a method to isolate salivary gland interstitial fluid and, thereby, determine COPif in this organ. Another implication of our observations is that the concentration of signaling substances, e.g., inflammatory mediators, can be measured locally in the fluid bathing the glandular cells.
Vi-Pif Curve in Passive Alteration of Salivary Gland Hydration
The Vi-Pif relation in large tissues, such as skin and muscle, is important for whole body fluid distribution, but the same cannot be said for this relation in the submandibular gland. Our reason for determining interstitial compliance in this organ was the suggestion of a previous study that the salivary gland was a low-compliant organ and that this phenomenon could influence transcapillary fluid transport (4). In the control situation, we found a positive Pif in the salivary gland, in agreement with previous data from the salivary gland (4) as well as other oral tissues (11, 13, 14). Finding an average compliance 36% of that found during corresponding changes in hydration in rat skin and muscle (20, 28), we were able to verify our original hypothesis. This means that changes in hydration induced from the vascular side will be more strongly counteracted in the salivary gland than in the two latter organs. This effect is, however, even more pronounced in other organs with a stronger capsule, such as brain (29), rat tail (2), bone marrow (12), and dental pulp (10). Nevertheless, the development of a substantial hydrostatic counterpressure during increased transcapillary filtration is important when we consider fluid dynamics in the salivary gland, in that it enhances the effect of the changes in transcapillary filtration induced by active secretion.
Vi-Pif Relation During Active Secretion
A novel and surprising finding was that activation of muscarinic receptors inducing salivation led to a significant drop in Pif without a concomitant drop in Vi or total tissue water content. During systemic changes in hydration, the change in pressure was a consequence of an altered hydration, requiring an alternative explanation for the reduction in Pif during salivation. In this context, it is of interest to consider observations by Lung (17), who reported that subcapsular pressures in canine submandibular glands dropped during stimulation of parasympathetic, as well as sympathetic, nerves to the glands. These observations were interpreted as a receptor-mediated activation of myoepithelial cells, causing a contraction of the acinar cells and intercalated ducts that moved the tissue away from the capsule. The subcapsular pressure dropped, despite increased submandibular gland blood flow, after stimulation of parasympathetic nerves. When ductal occlusion was induced during stimulation, the decline in subcapsular pressure was enhanced, indicating that the myoepithelial cells can contract more when they are distended and, even more important, that the pressure dropped even when no fluid was removed from the organ.
It is possible that the total water content in the experimental salivary glands in our study was reduced immediately after onset of secretion and gradually returned to baseline, as reported in cat submandibular gland (16). However, in salivary glands collected for fluid volume measurements, when Pif had stabilized at a lower level (after 4–6 min), there was no change in total fluid content of the submandibular glands.
Lung (17) measured pressure by introduction of a miniature transducer in the subcapsular space until it abutted the parenchyma and, thereby, separated the capsule from the submandibular gland parenchyma, whereas our pressure measurements were done in the interstitial fluid. Despite differences in methodology and the possibility that the pressure recorded by Lung may partially reflect solid tissue pressure (9), subcapsular pressure recorded by Lung during rest (3.0 mmHg) is close to our Pif measured during similar conditions (3.3 mmHg). The pressure drop recorded by Lung when acetylcholine was given systemically (about –4.2 mmHg) is similar in magnitude to that observed by us after pilocarpine infusions (about –3.8 mmHg), suggesting that there is no pressure gradient between the subcapsular space and interstitial fluid during salivation, and it is reasonable to assume that our measurements are comparable. We have, however, documented for the first time that the pressure drop takes place also in the interstitial space surrounding the secreting acinar cells and without changes in the Vi and that this effect is important for the formation of saliva.
The observed change in Pif during pilocarpine stimulation is the sum of temporary changes in Vv and Vi and possible conformational changes. We propose that the myoepithelial cells have an important role in the lowering of Pif and, thereby, in increasing the filtration pressure gradient leading to saliva formation. When the myoepithelial cells contract as a result of receptor activation, the interstitial tissue that surrounds the acinar cells and intercalated ducts undergoes a change in conformation that may lead to a reduction of hydrostatic tissue pressure. This may be considered to parallel changes in skin, where the integrins play an active role in control of Pif. After the 
1-integrin is blocked, the transcapillary fluid transport is increased by a reduction in Pif followed by edema formation (19). The lowering of pressure (Pif) causes a dramatic increase in the net filtration pressure. The myoepithelial cells may thus have an active and important role in reducing Pif and increasing transcapillary filtration. Also, the myoepithelial cells have an important role in reducing the luminal capacity and increasing the secretory pressure in the ducts and, thereby, accelerating the outflow of saliva (8). Even though we found no increase in Vi during secretion, the fluid flow through the epithelium may exceed the transport into the interstitium from the capillaries during brief periods of secretion, resulting in reduction in Vi. If this is the case, the reduction of Pif will be enhanced, because we can assume that the contraction of myoepithelial cells, as well as the volume reduction, will lead to a lowering of Pif. The latter is a result of a relatively low compliance in the organ as measured in this study. Our explanation is only partly consistent with that offered Smaje (21), who proposed that such a pressure reduction could be a consequence of a reduced Vi.
We may estimate the quantitative importance of a reduction in net transcapillary fluid pressure at about –4 mmHg with respect to transcapillary fluid flow. COPif and COPp were 10.3 and 16 mmHg, respectively. Because we measured no changes in Vi during lowered Pif and has been shown to remain relatively unchanged during secretion (0.79) (22), we can assume that there was no change in COPif when we measured the lowered Pif and collected the salivary glands for fluid volume distribution. Using these assumptions and the Starling equation, we can calculate the transcapillary fluid flow during secretion when applying all known parameters and, furthermore, estimate the contribution of the drop in Pif. The composite CFC, determined in the isolated perfused rabbit submandibular gland, is 1.35 ml·min–1·100 g–1·mmHg–1 and does not change in the presence of acetylcholine (7). The Pc in the submandibular gland has been estimated to 31 mmHg during secretion (7). If we calculate the transcapillary fluid flow during secretion with and without a drop in Pif, the respective values are 36.9 and 31.8 ml·min–1·100 g–1. Accordingly, the drop in Pif in this situation will increase the transcapillary fluid flow 16% without a reduction in Vi. To achieve a drop in Pif of the magnitude we measured after pilocarpine infusion without involvement of the myoepithelial cells, a reduction in Vi of 25% would have been needed, as calculated from our value for interstitial compliance, showing the significant importance of the myoepithelial cell effect for filtration.
In conclusion, we have shown that the salivary gland has a low interstitial compliance that may influence the secretion process. During active secretion, Pif drops, increasing the net filtration pressure and, thereby, contributing to the salivation. This drop in Pif is most likely mediated via the myoepithelial cells, suggesting an active and new role for these cells in salivary secretion.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
1-integrins in skin causes edema through lowering of interstitial fluid pressure. Circ Res 71: 978–983, 1992.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
, No Pif data because of tissue movement. 
, Start of infusion. Traces are from an animal with salivary glands intact for fluid volume measurements.
