INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE SECRETIONS THAT BATHE the mucosal epithelial surface contain an array of host defense factors, including secretory immunoglobulins, of which IgA is the major class. The polymeric immunoglobulin receptor (pIgR) is produced by secretory epithelial cells and transports polymeric IgA and IgM from the basolateral to the apical surface. The rat pIgR is a 118- to 120-kDa glycoprotein composed of an extracellular immunoglobulin-binding portion termed secretory component (SC), which contains five immunoglobulin-like domains, a membrane-spanning element, and a cytoplasmic carboxy-terminal domain (30, 50). Following transcytosis of the pIgR to the apical region, free pIgR and pIgR bound to immunoglobulin accumulate in an endosome compartment called the "apical recycling endosomes" (2). Fusion of these vesicles with the apical membrane and proteolytic cleavage of the ectoplasmic segment of pIgR results in the secretion of free SC (FSC) and SC-containing immunoglobulins (secretory Ig).

Transcytosis of the pIgR via transcytotic vesicles from basolateral endosomes to apical endosomes (2, 4) involves heterotrimeric GTP binding (G) proteins (5) and the phosphorylation of the cytoplasmic tail of pIgR (7) prior to secretion from the apical cell surface (48). Several hormones and pharmacological agents that mediate their effects via cAMP have been shown to affect SC synthesis and secretion in cultured rat lacrimal acinar cells in vitro (17) and in uterine secretions (53) and in rat bile (54) in vivo. In addition, agents that elevate cAMP levels (including cholera toxin, forskolin, or 8-bromo-cAMP) have been shown to stimulate apically directed pIgR transcytosis and secretion in both cultured rat lacrimal cells (17) and pIgR-transfected Madin-Darby canine kidney (MDCK) cells (16, 37). In contrast, another study using pIgR-transfected MDCK cells reported that activation of stimulatory G_s protein by cholera toxin inhibited the delivery of pIgR from apical endosomes to the apical surface (4). Thus the role of the G_s-adenylyl cyclase-cAMP signal pathway in the control of pIgR transport and secretion is unclear. Although the induction of pIgR mRNA expression by several cytokines has been studied in colon adenocarcinoma cells in vitro (29, 38), we are unaware of any investigations that have examined the regulation of pIgR mRNA expression in vivo.

The pIgR is expressed in many secretory epithelia and has been demonstrated in distal tubule kidney epithelial cells of humans (1) and rodents (40) both in the presence and absence of IgA staining (1, 40). Excretion of SC into the urine is increased in urinary tract infections both as FSC (15) or bound to IgA (15, 49), where it may function to prevent degradation of secretory immunoglobulins (24) and to inhibit adhesion of bacteria to cell membranes (13). Water intake appears to influence susceptibility to urinary tract infections, as diuresis has been shown to lower the concentration of bacteria in the kidney in some experimental models of retrograde (28) and hematogenous (23) pyelonephritis. It is unclear, however, what role water balance and urine flow rate may play in host defenses of the urogenital tract and specifically in the production of pIgR and secretion of FSC and sIgA.

In the present study, we investigated the effect of water intake on pIgR expression and the role of vasopressin in this process in the rat kidney in vivo. The effects of vasopressin were of interest since cAMP signaling pathways appear to play a role in the control of pIgR function in some in vitro systems. The antidiuretic action of arginine vasopressin is mediated through G_s-coupled V₂ vasopressin receptors and is linked to adenylyl cyclase and the production of cAMP (14, 25). These V₂ receptors are localized to the collecting duct and, in addition, to the medullary and cortical thick ascending limb (TAL) in the rat (32). We found that pIgR mRNA and intracellular protein are increased in the kidney with water deprivation (compared with water load) and with vasopressin administration. Secretion of the SC fragment of pIgR and sIgA, however, appears to be independent of vasopressin levels and dependent on urine flow. These results suggest that the hydration status, vasopressin level and urine flow rate may alter the mucosal immunity of the urinary tract by regulating polymeric immunoglobulin transport in the kidney.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. Male Sprague-Dawley rats (Harlan, Houston, TX) weighing between 250 and 350 g were used for all experiments. Rats were placed in metabolic cages for an equilibration period of 3-6 days prior to starting the study. Animals were pair fed (Formulab diet 5008; PMI Feeds, St. Louis, MO) and initially received ad libitum water for 3 days prior to the experiments. There were five study groups: 1) water-loaded rats, which received 0.6 M sucrose (Sigma) to drink for 48 h; 2) water-deprived rats, which received no water for 48 h; 3) ad libitum rats, which received water ad libitum for 48 h; 4) desmopressin (DDAVP)-injected rats, with DDAVP (desmopressin acetate-saline, Rhone-Poulenc Pharmaceuticals) dosing as described below; and 5) saline-injected rats, which served as vehicle control for the DDAVP group. After treatment, the animals were anesthetized with pentobarbital sodium (50 mg/kg) injected intraperitoneally, and both kidneys were harvested after ice-cold 0.9% saline perfusion in situ. mRNA was isolated from one kidney; immunohistology and protein quantification were performed on the other kidney.

Antibodies. Rat FSC was isolated from bile (22) by affinity chromatography on human IgM bound to a N-hydroxysuccinimide (NHS)-activated Sepharose column (HiTrap NHS-activated; Pharmacia Biotech, Piscataway, NJ). The bound FSC was eluted with 1 M potassium thiocyanate in 10 mM borate buffer (pH 7.4). After the eluted SC protein was shown to be pure by SDS-PAGE, 500 µg of purified protein were mixed with complete Freund's adjuvant and injected into a female New Zealand White rabbit (D & D Rabbitry, Tyler, TX). A booster immunization was performed 2 mo later with 225 µg purified FSC in incomplete Freund's adjuvant. IgG was isolated from the rabbit antiserum using a protein G column (HiTrap Protein G column, Pharmacia) and rendered monospecific for SC by passing over rat myeloma IgA (Zymed Laboratories, San Francisco, CA) bound to a NHS-activated Sepharose column (Pharmacia). Specificity of the antibody fraction was confirmed by detection of bands of appropriate size for rat secretory IgA and SC protein in rat bile by Western blotting and the absence of any bands in whole rat serum (Fig. 1). Aliquots of this IgG polyclonal antibody were conjugated with horseradish peroxidase (HRP, Sigma), per the described method (31), for use in ELISA and Western blotting. Polyclonal rabbit serum against the endodomain (cytoplasmic tail) of the pIgR was kindly provided by Lucian Saucan (44) (University of California, San Diego, CA).

View larger version (109K): [in this window] [in a new window]   Fig. 1.   Rabbit anti-rat secretory component (SC) antibody is monospecific for SC. Western blot using polyclonal anti-SC antibody recognizes the 78- to 80-kDa SC of the polymeric immunoglobulin receptor (pIgR) present in purified SC and in rat bile. There is no detectable SC in rat serum. FSC, free SC.

Monoclonal mouse anti-rat IgA antibody used in IgA quantification was obtained from Zymed Laboratories, and the HRP-conjugated sheep anti-rat IgA (-chain) was obtained from The Binding Site (San Diego, CA). Secretory IgA used as a standard in sIgA quantification was isolated from rat bile on an HPLC Superdex 200HR 10/30 column (SMART System Pharmacia Biotech, Piscataway, NJ). The purity of sIgA was confirmed by SDS-PAGE gels stained with Coomassie and by Western blots (not shown) and standardized against purified rat myeloma IgA (Zymed). Anti-rat Tamm-Horsfall (T-H) glycoprotein antiserum was kindly provided by Dr. John Hoyer (University of Pennsylvania, Philadelphia, PA).

DDAVP injection. Rats were injected with physiological doses (12, 26) of either 0.3 µg/kg (low range) or 3.0 µg/kg (high range) DDAVP subcutaneously 12 h apart (two doses). Additional rats were injected with 3.0 µg/kg every 12 h for 48 h (four doses). Control rats were injected with equal volumes of 0.9% saline (vehicle) at the same times as the DDAVP-treated animals.

Urine collection and osmolality. Urine was collected daily under mineral oil from metabolic cages into polypropylene tubes (Falcon 2059; Becton-Dickinson, Bedford, MA). Aprotinin (200 mM; Sigma) in sodium borate-buffered saline was added to each urine collection tube (200 µl /tube) to prevent proteolysis and microbial growth, respectively. Osmolality was measured using a vapor-pressure osmometer (Wescor, Logan, UT) and was corrected for the dilution factor and osmolality of aprotinin and sodium borate-buffered saline.

ELISA. Renal and urinary FSC levels were quantified in a sandwich ELISA modified from a previously described method (9). Briefly, 96-well polystyrene microtiter plates (Dynatech Laboratories, Chantilly, VA), coated with 3 µg/ml of purified rat myeloma IgA (Zymed Laboratories) in borate-buffered saline were used to selectively capture FSC. The polyclonal rabbit IgG anti-rat FSC, conjugated to HRP served as the detecting reagent. Bound conjugate was quantified using O-phenlylenediamine dihydrochloride (0.2 mg/ml, Sigma) and H₂O₂ (3%) as substrates, and the colorimetric product was read at 492 nm on an ELISA plate reader (Titertek Multiskan, Costa Mesa, CA). Rat FSC utilized as a standard was purified from bile and quantified by UV absorbance utilizing the extinction coefficient of 12.66 (19). Identical methods were used in an ELISA for urinary rat sIgA levels, except mouse anti-rat IgA (Zymed) in borate-buffered saline (3 µg/ml) was utilizing to capture and our purified sIgA served as a standard. Urine and tissue homogenates were tested at multiple dilutions. The concentration of FSC and sIgA in the samples was determined by linear regression, using at least two different dilution points for each sample.

mRNA isolation and Northern analysis. Whole kidney mRNA was prepared by rapid homogenization in guanidinium thiocyanate as previously described (8). Poly(A) RNA (5 µg/lane) was isolated using oligo(dT)-cellulose (type 3; Collaborative Biomedical Products, Bedford, MA), electrophoresed on formaldehyde 1% agarose gels, and transferred overnight to nitrocellulose (BAS-85; Schleicher & Schuell, Keene, NH) (43). The nitrocellulose blots were prehybridized at 45°C overnight with denatured salmon sperm DNA (0.1 mg/ml) in 5× SSC, 50% deionized formamide, and 0.02% Denhardt's solution (0.02% Ficoll, 0.02% polyvinylpyrrolidone, and 0.02% bovine serum albumin). Hybridization was carried out in the same solution containing 10% dextran sulfate with random-primed (New England Biolabs, Beverly, MA) cDNA labeled with [³²P]dCTP (Amersham Life Sciences, Arlington Heights, IL). We utilized the 1.4-kb Sac I digest of the full-length cDNA probe for rat pIgR mRNA, which was kindly provided by George Banting (3). Nonspecifically bound radioactivity was removed by extensive washing (43). After washing, the blots were air dried and exposed overnight to Kodak XAR film with intensifying screens at 70°C. Variability in loading of intact mRNA was assessed by reprobing for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA. Quantifications of mRNA for both pIgR and GAPDH mRNA were determined utilizing a phosphorimager (Molecular Dynamics, Sunnyvale, CA). The quantity of pIgR mRNA was expressed as a ratio to that of GAPDH in the same samples.

Immunohistology. Tissue sections were prepared from whole kidney previously fixed in 4% neutral buffered formaldehyde, rinsed in PBS and placed, in 20% sucrose-PBS prior to freezing in OCT compound (Fisher Scientific, Pittsburgh, PA). Cross sections were cut at 7 µm, labeled with the monospecific rabbit anti-SC antibody, after which biotinylated goat anti-rabbit antibody (Vector Laboratories, Burlingame, CA) and avidin-biotin peroxidase complex with 0.01% diaminobenzidine (VectaStain Elite ABC; Vector Labs) were utilized to detect the primary antibody staining. Nonspecific staining was blocked with 5% normal goat serum (Sigma) in 0.1% Triton X-100 (Sigma), and endogenous peroxidase was inactivated by 1% H₂O₂ in PBS:methanol (1:1). Double immunostaining for both pIgR and T-H was performed on the same tissue section by first staining for pIgR as described, followed by incubation with rabbit anti-rat T-H antiserum. The anti-T-H antibody was detected with biotinylated goat anti-rabbit antibody (Vector), followed by alkaline phosphatase-conjugated anti-biotin antibody (Vector), after which the alkaline phosphatase was developed with alkaline phosphate substrate (Alkaline Phosphate Substrate Kit 3, Vector). Additional (duplicate) tissue sections were stained with either the polyclonal rabbit anti-rat pIgR cytoplasmic tail antiserum (44), to confirm that our polyclonal rabbit anti-rat SC antibody was staining intact intracellular pIgR, or the HRP-conjugated sheep anti-rat IgA (The Binding Site), to demonstrate the location of rat IgA and confirm that at least some of the pIgR detected had cycled from the basolateral membrane.

Preparation of homogenates and subcellular fractions of kidney. Kidneys were removed after in situ perfusion with 0.154 M saline at 4°C. Membrane fractions were then prepared using differential centrifugation of whole kidney homogenates, per published methods (34). Briefly, tissue was rapidly diced, placed in isotonic sucrose (0.25 M) in 10 mM Tris · HCl homogenization solution (pH 7.5) in the presence of 1 µg/ml aprotinin (Sigma), 1 mM phenylmethylsulfonyl fluoride (Sigma), 1 µg/ml leupeptin (Sigma), and homogenized in a Potter-Elvehjem tissue homogenizer (Wheaton, Millville, NJ) for 15 strokes at 4°C. After the initial spin (1,000 g), subfractions were obtained by differential centrifugation to produce a membrane pellet (10,000 g for 30 min), microsomal pellet (100,000 g for 1 h), and the cytosolic fraction (supernatant of the microsomal fraction). Total protein per sample was quantified from solubilized centrifuge pellets (diluted in PBS plus 0.1% Tween 20) and the microsomal supernatant (Tween 20 added to 0.1%) using a biuret reaction with cuprous sulfate and bicinchoninic acid (46). Bovine serum albumin was used to generate a standard curve, and the absorption at 562-nm wavelength measured per the manufacturer's recommendations (Micro BCA Protein assay; Pierce, Rockford, IL). The concentrations of pIgR and SC in each subcellular fraction were quantified by ELISA (as previously described), and the results are expressed as micrograms per milligram of total protein.

Western blots. Equal amounts of protein from the plasma membrane and microsomal fractions were run on an 8.0% polyacrylamide gel and subsequently electrophoretically transferred to a supported nitrocellulose (S & S) in Tris-glycine electrophoresis buffer (pH 8.3) in a temperature-controlled tank (Trans-Blot Cell; Bio-Rad, Hercules, CA) overnight. Nitrocellulose membranes were blocked with 5% dried milk in 20 mM Tris-buffered saline-1% Tween 20 to prevent nonspecific binding. Following incubation with rabbit anti-rat FSC IgG diluted in Tris-buffered saline-1% Tween 20 for 2 h, membranes were washed, and the primary antibody was detected, using goat anti-rabbit IgG:HRP conjugate (Bio-Rad). The bound antibody:HRP conjugate was detected with 2.8 mM 4-chloro-1-naphthol (Sigma) in TBS/0.01% hydrogen peroxide.

Statistical analysis. Differences for two groups were assessed by unpaired t-test assuming equal variances. Analysis of variance (Microsoft Excel, Redmond, WA) was performed for experiments with multiple groups; when a significant difference was present, groups were analyzed by the Bonferroni adjustment of the Student's t-test (47). P < 0.05 was considered statistically significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Development of monospecific polyclonal anti-rat FSC antibody. We produced a monospecific, polyclonal rabbit anti-rat SC antiserum by immunizing with purified rat FSC isolated from rat bile (see METHODS). This antibody was shown to specifically stain the 78- to 80-kDa SC portion of the pIgR in Western blots of reduced rat bile and serum (Fig. 1). This polyclonal antibody also recognized the intact pIgR in tissue sections and in protein homogenates from whole kidney (Figs. 2-4, 6, and 10).

View larger version (119K): [in this window] [in a new window]   Fig. 2.   Immunolocalization of the pIgR in the rat kidney. Immunohistochemical staining of control rat kidneys reveals staining in the cortex (A) for pIgR in the basolateral aspect of the proximal tubule (arrows), throughout distal convoluted tubule (), but no staining of the glomeruli (g). pIgR is prominent along the apical domain of the thick ascending limb (TAL) cells () in both the outer stripe (B) and inner stripe (C) of outer medulla. There is no staining of the vasculature (v) or collecting ducts (cd) in outer or inner medulla (im). Magnification, ×160.

Localization of the pIgR and sIgA in the rat kidney. Immunohistology of normal rat cortex demonstrated pIgR staining of the basolateral surface of the proximal tubule epithelium and the distal convoluted tubule but no staining of the glomeruli (Fig. 2A). The outer medulla demonstrated prominent staining for the pIgR in the cells of the TAL, with no staining of the collecting ducts, vasculature, or inner medulla (Fig. 2, B and C). Comparison of receptor staining in the inner stripe of the outer medulla of rats on different water intakes showed an increased expression of pIgR protein in the TAL cells of the water-deprived vs. the water-loaded animals (Fig. 3, A and B). Confirmation that the epithelial staining for pIgR was predominantly in the TAL was obtained by costaining of the same section with anti-T-H polyclonal antibody, which demonstrates that pIgR localizes to TAL cells and not to the collecting tubules, vasculature, or interstitium (Fig. 3C). Furthermore, although pIgR staining is apparent throughout the TAL epithelium, the pIgR localized predominantly to the to the apical aspect of the T-H-positive TAL cells, especially in the water-deprived group. The predominant apical staining suggested that this protein may have accumulated in the subapical vesicles or an "apical recycling compartment" (2, 4) during low urine flow. Anti-pIgR cytoplasmic tail antiserum (44) demonstrated an identical pattern of pIgR staining in the apical portion of TAL epithelial cells (not shown), confirming that the apical staining was due to intact pIgR receptor and not cleaved SC. There was no staining of kidney sections by rabbit preimmune serum (not shown). IgA plasma cells, when present, were very rare in the normal rat. Staining for IgA was less frequent and of considerably less intensity than that of pIgR but also localized to the apical membrane of TAL epithelium (not shown), suggesting that the majority of pIgR in the TAL is not associated with IgA (1).

View larger version (88K): [in this window] [in a new window]   Fig. 3.   Water deprivation results in subapical accumulation of pIgR. Immunohistological staining of the inner stripe of the outer medulla of representative samples demonstrates greater staining for pIgR protein in water-deprived (A) compared with water-loaded (B) animals. Double staining of inner stripe of the outer medulla reveals concurrent staining for both pIgR (brown) and Tamm-Horsfall glycoprotein (blue) in the TAL (C). pIgR is predominantly in the apical aspect of the TAL epithelium (), as confirmed by Tamm-Horsfall glycoprotein staining (arrows). Distribution of pIgR suggests that pIgR accumulates in the subapical vesicles of the TAL in water-deprived animals and that there is no recruitment of expression in other cell types; n = 3 rats/group. Magnifications: A and B, ×170; C, ×340.

Water deprivation increased renal pIgR protein. Membrane fractions isolated from whole kidney homogenates were examined for both pIgR concentrations and total protein, which allowed pIgR to be expressed as a proportion of the total protein each fraction (µg pIgR/mg total protein). Quantification of the rat pIgR protein by ELISA demonstrated a significant increase in the pIgR-to-total protein ratio in the microsomal fraction of the water-deprived group compared with either the water-loaded or control groups (P < 0.05) (Fig. 4). This increased pIgR-to-total protein ratio was not due to differences in total protein concentrations between groups (P > 0.085).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Increased pIgR protein in kidneys of water-deprived animals. ELISA quantification of kidney reveals a twofold increase in pIgR protein in microsomal (vesicle) fraction in water-deprived (solid bars) compared with water-loaded (hatched bars) or ad libitum water (control, open bars) animals. There is no significant difference in pIgR protein levels between groups in the plasma membrane fraction. * P < 0.05; n = 5 rats/group.

Kidney pIgR mRNA levels are greater in water-deprived than in water-loaded rats. To determine whether the different levels of pIgR protein expression are related to differences in mRNA levels, we isolated mRNA from rats that had been water deprived or water loaded for both 24 h and 48 h. Northern blot analysis demonstrated that the ratio of pIgR to GAPDH mRNA is greater in the water-deprived group than in the water-loaded group at 48 h (Fig. 5). Densitometry of the blots indicated an approximately threefold increase in pIgR mRNA in the water-deprived group compared with the water-loaded group at 48 h (P < 0.03), although mRNA levels in the water-deprived group did not differ significantly from controls (ad libitum water) (Fig. 5). There is no significant difference in mRNA expression between groups at 24 h (not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Water deprivation increased kidney pIgR mRNA in water-deprived compared with water-loaded animals. A: representative lanes from Northern blot of whole kidney reveals increased pIgR mRNA levels in water-deprived compared with water-loaded animals at 48 h. Although the selected lanes suggest a greater pIgR-to-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA ratio in the water-deprived group, there was no significant difference from the ad libitum water (control) animals; n = 3-4 rats/group. B: densitometry of pIgR and GAPDH mRNA from water-loaded (hatched bars), water-deprived (solid bars), and ad libitum water (control) animals (open bars) reveals a threefold greater pIgR-to-GAPDH mRNA ratio in the water-deprived compared with the water-loaded group. pIgR mRNA in water-deprived animals is not significantly different from that of ad libitum water group. * P < 0.03; n = 3-6 rats/group.

DDAVP injection increases intracellular pIgR protein. We next sought to determine whether the effects of water deprivation on pIgR mRNA and protein levels could be reproduced by injection of DDAVP. Quantification of the pIgR-to-total protein ratios in the microsomal fraction by ELISA demonstrated increased pIgR levels in the DDAVP-injected vs. saline-injected (controls) at 24 h (P < 0.05) (Fig. 6). Immunohistology of the kidney 24 h after initiating DDAVP suggested an increase in intracellular pIgR content in the apical aspect of the TAL of the kidney (data not shown) as demonstrated with water deprivation. Although intracellular pIgR protein concentration remained higher in the DDAVP-injected rats at 48 h, the difference was no longer significant (Fig. 6).

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Desmopressin acetate (DDAVP) injection increased kidney pIgR protein levels: quantification of pIgR vs. total protein in the microsomal fraction isolated from whole kidney homogenates. Intracellular pIgR protein levels were increased in the DDAVP-injected animals (solid bars) at 24 h compared with saline-injected controls (hatched bars) and, although increased at 48 h, is no longer significant. * P < 0.05; n = 5 rats/group.

Increase in pIgR mRNA with DDAVP injection. In addition to its effect to increase pIgR protein levels, injection of DDAVP at physiological doses resulted in an increase in pIgR mRNA at 24 h (Fig. 7). Comparison of two different dose regimens of DDAVP injection at 24 h demonstrates that pIgR mRNA upregulation is dose dependent (Fig. 7). Quantification by densitometry revealed that pIgR mRNA expression was fourfold that of controls after 24 h of DDAVP injection at 3.0 mg/kg every 12 h compared with saline-injected controls (P < 0.01) (Fig. 8). This effect on mRNA appeared to be short lived, as the difference in mRNA levels were no longer present after 48 h despite continued injection of DDAVP (3.0 mg/kg every 12 h) (Fig. 8).

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Dose-dependent induction of pIgR mRNA with DDAVP injection. A: Northern blot of representative samples demonstrating a dose-dependent increase in pIgR mRNA by DDAVP injection at 24 h. GAPDH demonstrates equal amounts of mRNA are loaded per lane. B: quantification by densitometry of pIgR/GAPDH ratios reveals a fourfold increase in pIgR with 3.0 µg/kg dose (solid bar) compared with 0.3 µg/kg (hatched bar) or controls (open bar). * P < 0.05; n = 4-7 rats/group.

View larger version (29K): [in this window] [in a new window]   Fig. 8.   pIgR mRNA returns to control levels with continued DDAVP injection for 48 h. A: Northern blot of representative samples reveals that pIgR mRNA returns toward baseline levels with continued injection of DDAVP for 48 h. B: quantification by densitometry of pIgR/GAPDH mRNA reveals that pIgR mRNA levels from animals injected with DDAVP (3.0 µg/kg every 12 h) are increased to 400% of control levels at 24 h postinjection (hatched bar) but return to 70% of control levels (open bar) with continued DDAVP injection for 48 h (solid bar). * P < 0.01; n = 5-7 rats/group.

Urinary excretion of SC and sIgA varies with urine flow rate. The excretion of the cleaved fraction of the pIgR, the FSC, correlated directly with urine flow rate in both the water-loaded and water-deprived animals (Fig. 9; Table 1). The relationship between urine FSC excretion and urine volume is relatively linear for urine flows from 0.5 ml up to 10 ml/24 h (r² = 0.50) (Fig. 9, inset); hence, the urine concentration of FSC is relatively constant at normal (physiological) urine volumes. However, at urine flow rates greater than 15 ml/24 h there is no further increase in FSC excretion (Fig. 9). To determine whether secretion of SC was occurring from the ureters or bladder itself, we collected urine simultaneously from the renal pelvis (right kidney) and from the bladder (left kidney). Urine collected from the bladder had on average 8% higher levels of FSC than the urine collected from the renal pelvis. A timed infusion of saline through the ureter into the bladder at a rate proportional to the urinary volume of controls did reveal some addition of FSC from the ureter and bladder, although this would account for only 7% of the average total urinary FSC (not shown). Immunohistological staining of the bladder revealed intermittent staining of the epithelium (Fig. 10), whereas the intraluminal staining in the ureter suggested mucous layer adhesion of FSC (not shown). These results suggest that, although SC protein is present in the ureter and bladder epithelium and contributes some FSC to the urine, greater that 90% of the urinary FSC is of kidney origin in this model. The urinary excretion of sIgA also correlates with urine flow rate and paralleled the changes observed in FSC excretion in both the water-deprived and water-loaded animals (Fig. 11). Urinary sIgA secretion (µg/day) averaged 10 ± 5% of urinary FSC secretion (Fig. 11). DDAVP treatment did not appear to affect the secretion of FSC into the urine over a 24-h period (Table 1).

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Urine excretion of SC varies with urine flow. Comparison of 24-h excretion of FSC both before and after water loading (), water deprivation (), or water ad libitum () reveals that FSC excretion directly correlates with mean urine flows from 0.5 up to 15 ml per 24 h. Symbols represent mean urine flow per 24 h vs. mean FSC excretion per 24 h, error bars represent the standard error of the mean FSC excretion per 24 h, and broken line reveals best fit of the data; n = 9-11 rats/group. Inset: plot of FSC excretion vs. 24 h urine flow for all controls animals reveals a r2 value of 0.50 by linear regression; n = 109 samples.

View larger version (114K): [in this window] [in a new window]   Fig. 10.   Immunolocalization of pIgR is in intermittent bladder epithelial cells: immunohistology of the bladder demonstrates that the distribution of pIgR staining is intermittent in superficial transitional epithelial cells. Magnification, ×140.

View larger version (29K): [in this window] [in a new window]   Fig. 11.   Urine excretion of sIgA parallels changes in FSC excretion. Comparison of 24-h excretion rates of FSC (µg/day) and sIgA (µg/day) both before [d(1), hatched bars] and after (d+1, solid bars) water deprivation (top) or water loading (bottom) reveals parallel changes in urine FSC excretion and sIgA excretion. Secretory IgA levels average 10 ± 4% of FSC levels (µg/day). * P < 0.05 in paired t-test; n = 4 rats/group.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The results of this study suggest that hydration status and vasopressin may play a role in the mucosal immunity of the kidney by regulating epithelial cell production of the pIgR and secretion of secretory IgA. The increased intracellular pIgR protein concentration and mRNA expression with water deprivation, compared with that of water-loaded animals, were mimicked by intraperitoneal injection of DDAVP. We believe this is the first demonstration of regulation of pIgR mRNA expression in vivo. Furthermore, this is the first report that demonstrates that urinary excretion of FSC and sIgA seem to dissociate from pIgR production and correlate directly with urine flow.

Localization of pIgR and sIgA in the rat kidney. We demonstrated that the pIgR is preferentially expressed in the TAL of the outer medulla, as confirmed by T-H staining (Figs. 2 and 3C) (41). The preferential apical staining observed in the TAL cells (Fig. 3C), which is clearly evident in the water-deprived group compared with the water-loaded group (Fig. 3, A and B), suggests that the pIgR accumulates in subapical vesicles or the so-called "apical recycling compartment" (2, 4). Although the apical distribution of pIgR by immunohistology is consistent with localization in either apical membranes or subapical vesicles, the predominance of pIgR in the microsomal fraction by ELISA suggests that the apical accumulation is in subapical vesicles. Demonstration of apical staining with the anti-pIgR cytoplasmic tail antibody, which is specific for the intact receptor, suggests that the subapical staining is the intact pIgR and not cleaved SC. The different pattern of staining in the proximal tubules (basolateral) compared with the distal segments (apical) may represent an antidiuretic hormone (ADH) effect. Adenylate cyclase-mediated vesicle transport of pIgR to the apical membrane would be present only in the distal tubules where (in the rat) the ADH receptors are present (32). This effect would not occur in the proximal tubules, hence the vesicles would remain basolateral.

Detection of IgA at low levels in the TAL demonstrated that at least some of the pIgR detected had cycled from the basolateral membrane. It appears that the source of IgA that is transported across the renal tubule epithelia is predominantly from the serum, as our immunohistology sections of rat kidney stained with anti-IgA (-chain) detected only rare (<1 cell/high-power field) cells staining for IgA (not shown). Interestingly, the localization of the rat pIgR in the TAL could target secretory antibodies to the renal tubular lumen where pathogenic strains of Escherichia coli have been shown to bind in murine models of ascending pyelonephritis (33).

Water intake affects mRNA and protein expression. Our immunohistological data suggested that the pIgR protein is increased in water-deprived animals as a result of an accumulation of vesicles in the subapical compartment of the renal TAL cells (Fig. 3). We confirmed by ELISA that the increase in pIgR protein occurs predominantly in the microsomal fraction, consistent with the localization to intracellular vesicles in the kidney (Fig. 4) as previously described in the liver (50). The parallel increases in pIgR mRNA expression (Fig. 5) and intracellular pIgR protein levels (Fig. 4) in the kidneys of water-deprived, compared with water-loaded, animals suggests that the greater intracellular levels of pIgR protein may reflect both increased synthesis of pIgR and decreased apical secretion of SC into the renal tubule lumen during anti diuresis (Table 1).

DDAVP increases intracellular pIgR protein and mRNA. We next investigated the role of vasopressin in the production and/or transport of the pIgR in the rat kidney. As water deprivation elicits stress hormones in addition to vasopressin (18) and vasopressin has additional systemic effects (21), we chose DDAVP, which specifically activates the V₂ receptors localized in the kidney (21, 32). Similar to the effects of water deprivation, DDAVP injection increased intracellular pIgR protein and mRNA expression. The increase in intracellular pIgR protein was evident at both 24 and 48 h of DDAVP injection (Fig. 6). Although the increase in intracellular pIgR protein was significant only at 24 h, pIgR protein was increased in four of five animals at 48 h (P = 0.055). The dose-dependent increase in pIgR mRNA with DDAVP injection at 24 h (Fig. 7) suggests that the increase in pIgR mRNA expression with water deprivation, compared with that of the water-loaded animals, may be mediated by a vasopressin-coupled cAMP-responsive mechanism. The return of pIgR mRNA to control levels with repeated DDAVP injections for 48 h (Fig. 8) is consistent with reports that repeated injections of high doses of DDAVP induce desensitization of the rat TAL to vasopressin (12).

The increase in pIgR mRNA with high-dose DDAVP injection at 24 h (Fig. 7), compared with water deprivation at 24 h (not shown), suggests that additional stress hormones released by water deprivation (18) may modify the response to vasopressin and/or may reflect a difference in the circulating levels of arginine vasopressin within the first 24 h between the water-deprived and high-dose DDAVP injection groups. Whether there is an independent role of medullary interstitial or urinary osmolarity and vasopressin in the regulation of pIgR mRNA and protein expression is unclear. Both hyperosmolarity and vasopressin have been reported to affect protein levels and mRNA expression of several genes in kidney cells both in vitro (10, 51) and in vivo (11, 27, 55, 56). The fourfold increase in mRNA levels with DDAVP injection in the absence of a significant difference in urine osmolarity at 24 h (Table 1) suggests that vasopressin hormone plays a direct role. The lack of significant difference between water-deprived and control (ad libitum water) mRNA levels at 24 h and 48 h may be due to similar levels of endogenous vasopressin despite thirsting, as indicated by the similar urine osmolarities between these groups. Thus the control rats may be in a relative state of water conservation. Although reports of mean urine osmolarities with water deprivation vary in the literature, our results are consistent with those previously reported (6, 35, 36). It was not possible to change urine osmolarity independent of vasopressin levels long enough in vivo to determine the exact roles of medullary osmolarity and vasopressin in pIgR regulation. Other factors such as cytokines may be involved in pIgR regulation, since hyperosmolality may induce certain cytokines (45) and interferon- has been shown to induce pIgR mRNA expression in vitro (38). Furthermore, both osmolar stress and certain cytokines signal through identical p38 mitogen-activated protein kinase pathways (39). Hence, we are unable to conclude definitively in this in vivo model the relative roles of medullary hyperosmolarity, vasopressin levels, and cytokines on pIgR mRNA expression.

Urine flow is the primary determinant of pIgR and sIgA excretion. These studies are the first to report that the transport of SC into the urine varies with urine flow rate. FSC excretion was highly correlated with urine flow as excretion was 10-fold greater in the water-loaded group compared with the water-deprived group (Table 1), suggesting that urine flow is the primary factor in determining pIgR secretion. These results suggest that the observed change in intracellular pIgR protein levels may, in part, be due to different rates of urinary excretion, as the water-deprived or DDAVP-injected animals excreted protein at a slower rate and hence stored pIgR protein in apical vesicles. Conversely, the decreased intracellular levels of pIgR protein with water diuresis may be a result of the increased urinary FSC secretion rate from increased traffic of pIgR out of apical vesicles (Table 1) and presumably to the decreased levels of pIgR mRNA at 48 h (Fig. 6). Finally, apical transport of SC into the urinary space appears to be concentration dependent, as urinary FSC concentrations are relatively constant (1.03 ± 0.6 µg/ml) at normal (physiological) daily urine volumes below 15 ml (Fig. 9, inset). Although FSC protein is detectable in the bladder urothelium (Fig. 10), urinary FSC is predominantly of kidney origin, as less than 10% is derived from the ureter or bladder.

We do not know the factor(s) associated with diuresis that stimulate increased pIgR secretion. The decreased excretion of FSC in the water-deprived group could be due to decreased cleavage of pIgR at the apical surface and reuptake by endocytosis from the apical plasma membrane (2), possibly due to G protein-mediated inhibition of ligand exit from the apical transcytotic compartment (4). Similar to our findings in the kidney of flow-regulated apical vesicle transport of pIgR/SC into the urinary space, Larkin and Palade (20) demonstrated that low bile flow (cholestasis) in the liver decreased FSC excretion into the bile and resulted in increased vesicle staining for pIgR in the subapical bile canalicular region. Furthermore, Wira and Rossoll (54) reported that FSC excretion rates in rat bile correlated directly with bile flow in dexamethasone-treated rats and was not due to changes in FSC concentration. Urinary FSC excretion did not increase in our DDAVP-injected animals, despite the increase in pIgR mRNA and intracellular protein. Our results suggest that vasopressin controls pIgR expression but that flow independently controls FSC excretion.

The biological significance of maintaining constant FSC levels in the urine over a wide range of flow rates is unclear. Although transport of secretory Ig is thought to be the major function of SC, recent evidence suggests that FSC can inhibit bacterial adhesion to cells (13) and may function in humans to inhibit phospholipase A₂ (52). In addition to FSC, our data suggest that urinary sIgA secretion also varies with urine flow (~0.1 µg/ml) (Fig. 11) and that SC bound to IgA (sIgA) accounts for only 10% (µg/day) of SC secretion. This regulation of secretory Ig secretion may be important in host defense of the urinary tract, as there does appear to be an optimum sIgA concentration to prevent adhesion of E. coli to urinary tract epithelial cells (49), and patients with recurrent urinary tract infections have been reported to have low sIgA levels in the urine between their infections (42).

In conclusion, both water deprivation and DDAVP injection result in the upregulation of the pIgR mRNA and increased intracellular protein in the TAL of the rat kidney. Despite the intracellular accumulation of pIgR, FSC excretion into the urine appears to be controlled independently and correlated directly with urine flow rate. The combination of increased synthesis and decreased luminal secretion may explain the extensive accumulation of intracellular pIgR in the apical domain of the TAL cells during antidiuresis. The maintenance of constant FSC and sIgA concentrations in the urine may be important in preventing ascending urinary tract infection. The mechanism of flow dependence of secretion is not known, although factors other than those that regulate mRNA and intracellular protein levels must be involved.