Activation of Protein Kinase C-α and Src Kinase Increases Urea Transporter A1 α-2, 6 Sialylation

Abstract

The urea transporter A1 (UT-A1) is a glycosylated protein with two glycoforms: 117 and 97 kD. In diabetes, the increased abundance of the heavily glycosylated 117-kD UT-A1 corresponds to an increase of kidney tubule urea permeability. We previously reported that diabetes not only causes an increase of UT-A1 protein abundance but also, results in UT-A1 glycan changes, including an increase of sialic acid content. Because activation of the diacylglycerol (DAG)-protein kinase C (PKC) pathway is elevated in diabetes and PKC-α regulates UT-A1 urea transport activity, we explored the role of PKC in UT-A1 glycan sialylation. We found that activation of PKC specifically promotes UT-A1 glycan sialylation in both UT-A1-MDCK cells and rat kidney inner medullary collecting duct suspensions, and inhibition of PKC activity blocks high glucose-induced UT-A1 sialylation. Overexpression of PKC-α promoted UT-A1 sialylation and membrane surface expression. Conversely, PKC-α–deficient mice had significantly less sialylated UT-A1 compared with wild-type mice. Furthermore, the effect of PKC-α–induced UT-A1 sialylation was mainly mediated by Src kinase but not Raf-1 kinase. Functionally, increased UT-A1 sialylation corresponded with enhanced urea transport activity. Thus, our results reveal a novel mechanism by which PKC regulates UT-A1 function by increasing glycan sialylation through Src kinase pathways, which may have an important role in preventing the osmotic diuresis caused by glucosuria under diabetic conditions.

Urea and the urea transporters (UTs) play critical roles in the urinary concentrating mechanism. The major mechanism for delivering urea to the inner medullary interstitium is urea reabsorption from the terminal inner medullary collecting duct (IMCD).¹ The first UT cDNA was cloned in 1993,² and two mammalian UT genes have been reported (UT-A and UT-B).² Four UT-A isoforms (UT-A1, UT-A2, UT-A3, and UT-A4) have been identified in kidney medulla.³ UT-A1 is the largest and most complete form. The others are structurally truncated forms of UT-A1. UT-A1 is expressed in the apical membrane of epithelial cells in the IMCD and plays a dominant role in urea reabsorption and the development of the corticomedullary osmolarity gradient.^4,5 As proof that UT-A1 is important, mice that lack UT-A1 and UT-A3 have seriously impaired urea reabsorption, urine-concentrating ability,⁴ and hypertension.⁶

UT-A1 is a heavily glycosylated protein. In the inner medulla (IM), UT-A1 protein typically presents two different glycosylated bands at 97 and 117 kD; both are derived from an 88-kD nonglycosylated protein.⁷ The 117-kD glycoform is fully glycosylated and contains poly-N-acetyllactosamine, whereas the 97-kD glycoform is a hybrid form with a high level of mannose.⁸ A functional study using tubule perfusion showed that the increased 117-kD glycoform in the IM is associated with increased urea transport activity.⁹ This finding suggests that changes in the relative abundances of the 97- and 117-kD forms of UT-A1 may have important regulatory roles for UT-A1 function. Structurally, UT-A1 has four potential consensus N-linked glycosylation sites at Asn13, Asn279, Asn544, and Asn742. Only Asn279 and Asn742 in the two large extracellular loops are proven to be sites for N-linked glycosylation.¹⁰ Importantly, loss of glycosylation by mutating the two glycosylation sites significantly impairs the response of UT-A1 to vasopressin.¹⁰

Vasopressin is the primary hormone regulating UT-A1 transport activity in vivo. Vasopressin binds to V₂ receptors on the basolateral plasma membrane (PM) of the IMCD, stimulates adenylyl cyclase, generates cAMP, and activates protein kinase A (PKA), which in turn, phosphorylates UT-A1 proteins, mainly at serines 486 and 499, and regulates urea permeability.^11,12 In addition to the cAMP-PKA signaling pathway, UT-A1 is also found to be regulated by protein kinase C (PKC). This connection was based on the observation that a PKC-α knockout mouse has a urine-concentrating defect.¹³ Indeed, subsequent studies revealed that PKC, which is highly expressed in the IMCD,¹⁴ is also an important regulator of urea permeability.^14,15 PKC-dependent pathways particularly contribute to urea permeability induced by angiotensin II stimulation and hypertonicity.^14,15 Activation of PKA and PKC has synergistic stimulation effects on urea permeability in rat IMCDs.¹⁴

The PKC family is comprised of a group of serine/threonine-related PKs that plays a key role in many cellular functions and affects many signal transduction pathways.¹⁶ A cumulative body of evidence shows highly activated PKC pathways under diabetic conditions,¹⁷ which is presumably caused by increased de novo synthesis of diacylglycerol (DAG) from glycolytic intermediates.¹⁸ Typically, the conventional (PKC-α, -β1, -β2, and -γ) and novel (PKC-δ, -ε, -θ, and -η) PKC isoforms are activated by DAG. Clinical trials of PKC-β inhibitor have been conducted, with some promising results for diabetic neuropathy, diabetic retinopathy, and endothelial dysfunction.¹⁷

In diabetes, the kidney has elevated urea reabsorption activity, which plays a critical role in ameliorating the osmotic diuresis caused by glucosuria.¹⁹ We previously reported that diabetes not only causes an increase of UT-A1 protein abundance but also, results in UT-A1 glycan changes.⁸ UT-A1 glycan undergoes increased sialylation⁸; however, the mechanism of how UT-A1 is sialylated and the significance of increased UT-A1 sialylation remain unclear. In this study, we show that activation of PKC promotes UT-A1 sialylation. Inhibition of the PKC pathway blunts high glucose-induced UT-A1 sialylation. We also show that Src kinase, but not Raf-1 kinase, is involved in PKC-mediated UT-A1 glycan sialylation.

Results

Activation of PKC Promotes UT-A1 Glycan Sialylation

UT-A1 is hypersialylated under diabetic conditions.⁸ In addition, diabetes increases DAG by glycolysis, which results in a high level of PKC pathway activation.¹⁷ To evaluate whether UT-A1 sialylation is regulated by PKC, UT-A1-MDCK cells were treated by the PKC activator phorbol dibutyrate (PDBu), and total UT-A1 and sialylated UT-A1 were examined. Activation of PKC by PDBu slightly increased UT-A1 protein abundance (Supplemental Figure 1). However, activation of PKC dramatically increased sialylated UT-A1 by 3.5±1.1-fold (P<0.01; n=3) as pulled down by sambucus nigra lectin (SNA) lectin (Figure 1A). PDBu stimulation also facilitated UT-A1 glycan maturation, which was indicated by Tomato and datura stramonium lectins that bind poly-N-acetyllactosamine on the terminal ends of glycans and β-1,4–linked N-acetylglucosamine, respectively. We then examined native UT-A1 glycan changes using kidney IM tissue from Sprague–Dawley rats. IMCD suspensions were prepared and treated with PDBu for 6 hours. Because the cell PM UT-A1 is concentrated in lipid raft microdomains,⁸ we particularly examined cell membrane UT-A1 in lipid raft domains. As shown in Figure 1B, PDBu treatment did not change UT-A1 distribution in lipid rafts. We collected membrane lipid raft fractions (fractions 2–5) and performed a lectin pull-down assay. PDBu treatment significantly increased sialylation of UT-A1 by 2.6±0.7-fold (P<0.01; n=3) from kidney IM that is precipitated by SNA lectin.

Figure 1.

PKC activator increases UT-A1 glycan sialylation. (A) UT-A1 MDCK cells were treated with 2 µM PDBu for 24 hours. The cells were lysed in RIPA buffer, equal amounts of total lysates were incubated with agarose-conjugated lectins, and then, they were immunoblotted with antibody to UT-A1. (B) Kidney IMCD suspensions prepared from Sprague–Dawley rats were treated with 2 µM PDBu for 6 hours. The cell membrane lipid rafts were prepared by a 5%–40% sucrose gradient ultracentrifugation. Fractions 2–5 were collected for lectin pull-down assay with different lectins. The lectin-precipitated samples were then analyzed for UT-A1 protein expression by Western blot. The bar graph shows band densities as fold of control (means±SDs; n=3). Con A, concanavalin A; Ctrl, control; DSL, datura stramonium lectin; IB, immunoblot; PHA, phaseolus vulgaris leucoagglutinin; Tomato, lycopersicum esculentum lectin; WGA, wheat germ agglutinin. **P<0.01.

High Glucose Promotes UT-A1 Glycan Sialylation through a PKC Pathway

Hyperglycemia is one of the most prominent features of both types 1 and 2 diabetes.²⁰ To investigate whether increased glucose could stimulate UT-A1 glycan sialylation, UT-A1-MDCK cells were preincubated with low-glucose (5.5 mM) medium for 24 hours and then switched to high-glucose medium containing 25 mM glucose for 24 hours. To exclude the effect of osmolarity caused by 25 mM glucose, the same amount of L-glucose was used as a control. High glucose increased total UT-A1 protein abundance by 43%±7% (P<0.05; n=3) (Figure 2A). We were particularly interested in the alteration of UT-A1 glycan structure. Lectin pull-down assay showed that SNA-precipitated sialylated UT-A1 from high glucose-treated cells was increased by 2.2±0.5-fold (P<0.01; n=3) (Figure 2B). High glucose stimulates PKC-α protein expression in murine and human podocytes.²¹ To address the possible mechanism of high glucose-stimulated UT-A1 sialylation by the PKC pathway, 2 hours before high-glucose treatment, chelerythrine, a general PKC inhibitor, was applied to the cells. Inhibition of PKC activity blocked high glucose-stimulated UT-A1 sialylation by 4.9±1.9-fold (P<0.05; n=3) (Figure 2C). Galanthus nivalis lectin (GNL) binds to glycans that have a high amount of mannose. There was no difference in GNL-precipitated UT-A1 after PKC inhibitor treatment.

Figure 2.

High glucose-stimulated UT-A1 sialylation is blocked by a PKC inhibitor. (A and B) UT-A1-MDCK cells were preincubated with 5.5 mM glucose medium for 24 hours and then incubated with DMEM containing 5.5 or 25 mM glucose or 25 mM L-glucose for another 24 hours. Cells were lysed in RIPA buffer, equal amounts of total cell lysates were used for GNL and SNA lectin pull-down assay, and UT-A1 protein expression was analyzed by Western blot. (A) Total or (B) lectin-precipitated UT-A1 was examined by Western blot. (C) UT-A1 MDCK cells were treated as above, except for that one group had a 2-hour pretreatment with 10 µM chelerythrine (Chel). The bar graph shows band densities as fold of control (means±SDs; n=3). Ctrl, control; DSL, datura stramonium lectin; IB, immunoblot; PHA, phaseolus vulgaris leucoagglutinin; Tomato, lycopersicum esculentum lectin; WGA, wheat germ agglutinin. *P<0.05; **P<0.01.

PKC-α Is Involved in UT-A1 Glycan Sialylation

PKC-α is one of the most abundant PKC isoforms expressed in kidney IM.¹⁴ Strong evidence shows PKC-α involvement in kidney urea transport regulation.^14,15,22 To more specifically examine whether UT-A1 sialylation could be mediated by PKC-α, we cloned the PKC-α gene from kidney IM and generated pcDNA3-PKC-α. UT-A1-HEK293 cells were transfected with PKC-α. Figure 3A shows that PKC-α at a dose of 0.2 µg stimulates UT-A1 cell membrane expression as measured by biotinylation by 1.9±0.11-fold (P<0.01; n=4). To measure UT-A1 sialylation levels, cell lysates were incubated with agarose-conjugated SNA lectin and immunoblotted for UT-A1. As shown in Figure 3B, PKC-α stimulated UT-A1 sialylation (SNA-precipitated UT-A1). However, dosages over 0.8 µg do not increase and in fact, decrease UT-A1 sialylation. The increased UT-A1 sialylation caused by PKC-α was consistent with the change of UT-A1 cell membrane expression. GNL lectin pull-down assay was used as a control. PKC-α does not change GNL-precipitated, mannose-enriched UT-A1.

Figure 3.

PKC-α increases UT-A1 sialylation and cell membrane expression. (A) UT-A1 membrane expression. UT-A1-HEK 293 cells grown in six-well plates were transiently transfected with pcDNA3-PKC-α (0.2 µg/well) for 48 hours. Cells were then processed for cell surface biotinylation, and membrane UT-A1 proteins were examined by Western blot with UT-A1 antibody. (B) UT-A1 protein sialylation. HEK293 cells were transiently transfected with pcDNA3-PKC-α at different doses for 48 hours. The cells were lysed with RIPA buffer, and the lysates were applied for lectin pull-down assay. Total and lectin-precipitated UT-A1 were analyzed by Western blot with UT-A1 antibody. Transfected PKC-α was assessed by PKC- α antibody. The bar graph shows band densities as fold of control (means±SDs, n=3). Ctrl, control; IB, immunoblot. *P<0.05; **P<0.01.

Hyposialylated UT-A1 in PKC-α Knockout Mouse

To acquire in vivo evidence that PKC-α can affect UT-A1 glycan sialylation, we took advantage of a PKC-α knockout mouse that has decreased kidney urea transport activity.^14,22 Mouse kidney IM tissue was harvested and processed for membrane lipid raft preparation. Unlike rat UT-A1 expression in lipid rafts (Figure 1B), which shows both 97- and 117-kD glycosylated forms, mouse UT-A1 in membrane lipid rafts showed only the 117-kD UT-A1 glycoform (Figure 4A). The reason for the difference between these two species is currently unknown. Lipid raft fractions 2–5 were collected and used for lectin pull-down assay. As presented in Figure 4B, the SNA-precipitated sialylated UT-A1 was dramatically reduced in the PKC-α–deficient mouse, suggesting a critical role of PKC-α in UT-A1 glycan sialylation.

Figure 4.

Analysis of UT-A1 glycan sialylation in PKC-α knockout mouse. Kidney IM tissues were dissected from C57B6 wild-type (WT) or PKC-α knockout mice and processed for lipid raft isolation by 5%–40% sucrose gradient ultracentrifugation. Lipid raft fractions (fractions 2–5) were collected and used for a lectin pull-down assay. UT-A1 proteins from (A) cell membrane lipid rafts and (B) lectin-precipitated samples were examined by Western blot with UT-A1 antibody. Con A, concanavalin A; IB, immunoblot; MAA, maackia amurensis agglutinin; Tomato, lycopersicum esculentum lectin; WGA, wheat germ agglutinin.

Src Kinase as a Downstream Effector of PKC-α in Regulating UT-A1 Sialylation

To elucidate possible PKC signaling pathways, we profiled multiple PKC downstream factors with PKC-α knockout mouse kidney IM tissue. The expression of Src kinase is downregulated by 56%±10% (P<0.01; n=3) (Figure 5A), indicating that Src might serve as the downstream effector involved in PKC-α–induced UT-A1 glycan sialylation. To further investigate this possibility, UT-A1-HEK293 cells were treated with PDBu or PDBu with the Src kinase inhibitor 4-amino-5-(4-chlorophenyl)-7-(t-butyl) pyrazolo[3,4-d]pyrimidine (PP2) for 6 hours. PDBu-stimulated, PKC-mediated UT-A1 sialylation was blocked by Src inhibition by 53.1%±2.9% (P<0.01; n=4) (Figure 5B), indicating that Src is involved in PKC-induced sialylation. Raf-1 kinase is another major downstream effector of the PKC pathway. Unlike Src kinase, inhibition of Raf-1 kinase did not affect the PDBu-induced UT-A1 sialylation. There was no alteration of GNL-precipitated, mannose-enriched UT-A1 after PDBu or PP2 treatments.

Figure 5.

Effect of Src kinase in PDB-induced UT-A1 glycan sialylation. (A) Src kinase expression in PKC-α knockout mouse. Kidney IM tissues from wild-type (WT) or PKC-α knockout mice were lysed in RIPA buffer. Equal amounts of total lysates were analyzed for protein expression with relevant antibodies. (B) Involvement of Src in PDB-induced UT-A1 sialylation. UT-A1-HEK293 cells were pretreated with 2 µM PP2 or 2 µM Raf-1 inhibitor for 1 hour and then treated with 2 µM PDB for another 6 hours. The cell lysates were used for SNA lectin pull-down assay. UT-A1 from total lysates or SNA lectin-precipitated samples was analyzed by Western blot with UT-A1 antibody. (C) UT-A1-HEK293 cells were transfected with Src siRNA by Lipofactamine 2000. Scamble siRNA was used as a control. After 48 hours, the cells were collected. UT-A1 from total lysates or SNA lectin-precipitated samples was analyzed by Western blot with UT-A1 antibody. The bar graph shows band relative densities (means±SDs; n=3). Ctrl, control; IB, immunoblot. **P<0.01.

To rule out a possible off-target effect of using the Src inhibitor PP2, we also used siRNA to knock down Src in UT-A1-HEK293 cells. Src siRNA (sc-29228; Santa Cruz Biotechnology, Santa Cruz, CA) transfection reduced Src protein levels by 59.5%±7.9% (P<0.01; n=4) compared with control scramble siRNA. SNA-precipitated sialylated UT-A1 was reduced by 38.6%±9.1% (P<0.01; n=4) (Figure 5C).

UT-A1 Glycan Sialylation Increases Urea Transport Activity

Finally, we examined the functional consequence of sialylation on UT-A1 transport activity. UT-A1-MDCK cells grown on transwell inserts were treated with sialic acid for 24 hours, and UT-A1 activity was assessed by measuring transepithelial urea flux. The urea flux assay (Figure 6A) showed an increase in urea transport activity after sialic acid treatment. As a control, high glucose (25 versus 5.5 mM) treatment also increased urea transport activity. Because MDCK cells are resistant to plasmid transfection, to investigate the possible role of sialylation caused by PKC-α, we used a Xenopus oocyte expression system. UT-A1 cRNA was injected alone or coinjected with PKC-α cRNA, and urea transport activity was measured by ¹⁴C-urea uptake. Coinjection of PKC-α at a dosage of 2 ng/oocyte increased UT-A1 transport activity by 35.9%±7.9% (P<0.05; n=6) (Figure 6B). Oocyte cell PM was isolated, and cell membrane UT-A1 and SNA-precipitated UT-A1 from PM fractions were examined by Western blot. The enhanced urea transport activity by PKC-α is consistent with an increase of UT-A1 cell surface expression and UT-A1 sialylation as pulled down by SNA lectin (Figure 6B).

Figure 6.

Sialylation modification alters UT-A1 urea transport activity. (A) Addition of sialic acid increases UT-A1 transepithelial urea flux activity. UT-A1-MDCK cells were grown on transwell inserts for 3 days and then treated with 2 mM sialic acid or 25 mM glucose for 24 hours. Transepithelial urea flux was measured as described in Concise Methods. Each time point was a 3-minute flux measurement. The first four time points indicate the basal urea activity. The cells were then exposed to 10 µM forskolin (FSK). The last three time points were collected in the presence of dimethyurea (DMU) to control for any nonspecific leakage during the flux measurement. (B) Effect of PKC-α on UT-A1 activity, cell membrane expression, and protein sialylation. Oocytes were injected with cRNA encoding UT-A1 (5 ng/cell) alone or together with PKC-α (2 ng/cell) for 3 days. Urea transport activity was measured by [¹⁴C]-urea flux (n=6 oocytes/time point; means±SDs). Oocyte cell PMs were prepared, and UT-A1 from the PM- and SNA-precipitated samples was analyzed by Western blot with UT-A1 antibody. The shown Western blots represent two separate experiments. IB, immunoblot. *P<0.05.

Discussion

Diabetes has increased urea reabsorption mediated by UT-A1 in kidney.⁹ We recently discovered that UT-A1 is hypersialylated in streptozotocin (STZ)-induced diabetes,⁸ suggesting that glycan sialylation modification could contribute to the increased urea transport activity in diabetic kidney IMCD tubules. In this study, we have new data showing that PKC, with activity that is elevated in diabetes, promotes UT-A1 sialylation and increases UT-A1 urea transport activity. There are six major findings in this study. (1) Activation of PKC promotes UT-A1 glycan sialylation in both UT-A1-MDCK cells and IMCD suspensions. (2) High glucose increases UT-A1 sialylation, and this effect is blocked by inhibition of PKC. (3) PKC-α, the major conventional PKC isoform in kidney IM, directly stimulates UT-A1 sialylation and UT-A1 protein cell surface expression. (4) UT-A1 is hyposialylated when PKC-α is absent in a knockout mouse. (5) PKC-α–stimulated UT-A1 glycan sialylation is through a Src pathway. (6) Increasing UT-A1 sialylation by either sialic acid treatment in UT-A1-MDCK cells or coinjection of PKC-α into oocytes increases UT-A1 transport activity.

Glycosylation plays important roles in regulating membrane protein activity in many aspects. Reminiscent of its effect on several other membrane transport proteins, such as AQP2,²³ NKCC2,²⁴ and ENaC,²⁵ N-linked glycosylation is critical for UT-A1 membrane trafficking, stability, and activity.^8,10 Animal studies revealed a functional link between UT-A1 activity and its glycosylation state. In particular, the highly glycosylated 117-kD form of UT-A1 is increased in several conditions associated with decreased urine concentration, such as STZ-induced diabetes mellitus,¹⁹ a low-protein diet,²⁶ hypercalcemia,²⁷ water diuresis,²⁶ and furosemide administration.²⁶ By a tubule perfusion study, Pech et al.⁹ showed that the presence of increased 117-kD UT-A1 glycoform in diabetes is associated with enhanced urea transport activity. This finding conveys an important message that differential glycosylation of UT-A1 (117- versus 97-kD form) may determine UT-A1 function under certain conditions.

Glycosylation, an extremely complicated post-translational process, is initially started in the ER as early as during protein synthesis. The glycan is further matured, largely through processing in the Golgi complex, by trimming and adding different sugars, such as fucose, sialic acid, iduronic acid, xylose, etc. Accumulating evidence has revealed that appropriate oligosaccharide modification affects glycoprotein activation and function.^28–30 One of the most important terminal sugars added to the mature glycan is sialic acid (N-acetylneuraminic acid). The addition of sialic acid to the terminal galactose of N-linked polylactosamine chains could be in either a α2–6 or α2–3 linkage. Because sialic acid is a negatively charged large sugar that caps the terminal galactose in carbohydrate chains on the cell surface, sialylation modification often affects glycoproteins in many aspects, like changing the proteins overall conformation, ligand binding, and galectin binding. Sialylation has been linked to malignant transformation and tumor metastasis.³¹

Interestingly, UT-A1 is highly sialylated in diabetes.⁸ Because urea reabsorption in diabetic kidney IMCD is increased, we presume that the enhanced sialylation of UT-A1 could, at least partially, contribute to the increased urea transport activity. To explore the mechanism underlying the increased UT-A1 sialylation, we focused on a possible role for PKC. We chose this direction for two major reasons. (1) In diabetes, multiple biochemical pathways have been activated by hyperglycemia, of which the activation of the DAG-PKC pathway is one of the most significantly changed pathways.¹⁷ (2) PKC-α knockout mice have a urine-concentrating defect,¹³ and studies from our group show that PKC regulates urea permeability in the rat IMCD.^14,15,22 This study shows that the PKC activator PDBu stimulates UT-A1 glycan sialylation in both UT-A1 MDCK cells and isolated IMCD suspensions. We also found that high glucose increases sialylated UT-A1, and this effect is blocked by PKC inhibition, indicating that high-glucose stimulation of UT-A1 is mediated by a PKC pathway. High glucose does not alter kidney IMCD total PKC-α protein expression but increases phosphorylated PKC-α levels (Supplemental Figure 2).

PKCs can be subclassified into three groups: the conventional PKC isoforms (PKC-α, -βI, -βII, and -λ), which are activated by phosphatidylserine, calcium, and DAG or phorbol esters; the novel PKCs (PKC-δ, -ε, -η, -θ, and -µ), which are Ca²⁺-independent; and the atypical PKCs (PKC-ι/λ and -ζ) that are activated by only phosphatidylserine.¹⁷ To exclude a nonspecific effect caused by the PKC activator PDBu and more specifically target PKC-α, we cloned the PKC-α gene and transfected it into UT-A1-HEK293 cells. PKC-α at a low dose increased UT-A1 cell membrane expression as judged by cell surface biotinylation. Interestingly, UT-A1 sialylation stimulated by PKC-α is consistent with the increased amount of UT-A1 at the cell membrane (Figure 3), indicating the important role of sialylation in UT-A1 cell membrane accumulation and, therefore, the increased urea transport activity, particularly in diabetes. We also noticed that high-dosage PKC-α transfection does not increase and actually decreases UT-A1 sialylation and UT-A1 membrane expression. The reason is currently not clear. The high dose of PKC-α might activate some other target that, in turn, inhibits the sialylation process, or a high dose of PKC-α may deplete sialylation substrate/enzymes.

The effect of PKC-α on UT-A1 sialylation was further verified in a PKC-α knockout mouse. The deficiency of PKC-α results in defective UT-A1 sialylation (Figure 4B). To determine the possible downstream effectors of PKC-α that mediate UT-A1 sialylation, we examined PKC-α knockout mouse IM samples by Western blot and found that the Src kinase expression is significantly decreased. This finding suggests that Src might mediate PKC-stimulated UT-A1 sialylation. Src kinase has been shown to be activated by PKC. Nomura et al.³² reported that PMA treatment could rapidly induce activation of Src within 5 minutes in A172 glioblastoma cells. A similar PKC-dependent c-Src activation pathway also has been found in osteoblasts,³³ NCI-H292 epithelial cells,³⁴ the A7r5 smooth muscle cell line,³⁵ and SH-SY5Y neuroblastoma cells.³⁶ Here, we provide evidence that the Src kinase inhibitor PP2 prevents PKC stimulator-induced UT-A1 sialylation. In contrast, Raf-1 inhibition shows only a limited inhibitory effect. This finding indicates that the PKC/Src signaling pathway mediates UT-A1 sialylation.

An excellent study by Mo et al.³⁷ shows that the sialylation of N-glycans promotes membrane protein endolyn in apical biosynthetic delivery. PKC-stimulated sialylation may promote UT-A1 membrane trafficking and increase UT-A1 membrane expression. By using RNA-sequencing technology, we recently identified an increase of sialyltransferase ST6GalI in STZ diabetic kidney IM (the unpublished observation is from our lab). ST6GalI is a type II membrane protein localized in the trans Golgi network that catalyzes 2,6 sialylation of N-glycans. Overexpression of ST6GalI promotes UT-A1 sialylation and simultaneously increases UT-A1 membrane expression in HEK293 cells (Supplemental Figure 3). The increased ST6GalI activity under diabetic conditions may contribute to the sialylation of UT-A1 glycan.

In summary, the major finding of this study is that activation of PKC-α promotes UT-A1 glycan sialylation. This new finding merges the PKC-α, UT-A1 sialylation, and kidney urea reabsorption responses, which all are activated or increased under diabetic conditions. Our mechanistic hypothesis is that, in diabetes, hyperglycemia activates the DAG-PKC-Src pathway and subsequently promotes UT-A1 sialylation. Meanwhile, hyperglycemia causes an increase of glucose catalytic intermediates that serve as glycosylation substrates. UT-A1 sialylation, in turn, increases IMCD urea permeability. Therefore, the axis consisting of PKC–Src kinase–UT-A1 sialylation–tubular urea reabsorption and its activation may have an important role in ameliorating the osmotic diuresis caused by glucosuria and preventing more water and solute loss in diabetes.

Concise Methods

Chemicals

PKC stimulator PDBu, PKC blocker chelerythrine chloride, and N-acetylneuraminic acid were purchased from Sigma-Aldrich (St. Louis, MO). Src kinase inhibitor PP2 was purchased from Calbiochem (San Diego, CA). Raf1 kinase inhibitor I was from EMD Millipore (Billerica, MA).

Construction of PKC-α Plasmids

Total RNA was extracted from rat kidney IM by Trizol reagent (Invitrogen, Carlsbad, CA) and reverse transcripted into cDNA. The PKC-α coding region gene was obtained by RT-PCR and cloned into mammalian expression vector pcDNA or Xenopus expression vector pGH19. The PKC-α gene and cloning orientation were verified by DNA sequencing.

Cell Culture, Transfection, and Treatment

UT-A1-MDCK cells were described before.³⁸ HEK293 cells were purchased from Stratagene. Cell lines were cultured in DMEM (Invitrogen) supplemented with 10% FCS, 100 units/ml penicillin, and 100 µg/ml streptomycin. Low-glucose DMEM (5.5 mM glucose) was from Invitrogen. Control siRNA (sc-37007; Santa Cruz Biotechnology), c-Src siRNA (sc-29228; Santa Cruz Biotechnology), and plasmid (pcDNA3-PKC-α) transfection was carried out by LipofectAMINE 2000 (Invitrogen).

Animals and Tissue Collection

All animal protocols were approved by the Emory University Institutional Animal Care and Use Committee. Male Sprague–Dawley rats (Charles River Laboratories, Wilmington, MA) weighing 125–200 g were used in this study. C57B6 wild-type and PKC-α knockout mice (gift from Jeffrey Molkentin, University of Cincinnati³⁹) were used. Rats and mice were euthanized by asphyxiation. Kidneys were removed, and the IM was collected for tubule suspension preparation or lysed in RIPA buffer for Western blot.

IMCD Tubule Suspension Preparation and Treatment

Rat kidney IMCD suspensions were prepared by digestion with hyaluronidase and collagenase B (Sigma-Aldrich) as described before.⁸ IMCDs were treated with PDBu (2 µM) for 6 hours and then used for lipid raft isolation.

Lipid Raft Isolation

After treatment, UT-A1-MDCK cells or animal IMCDs were lysed in ice-cold 0.5% Brij96-TNEV buffer for 30 minutes on ice. Lipid raft isolation was performed with a 5%–40% sucrose discontinuous gradient as described.⁸ Equal sizes of fractions (approximately 400 µl) were collected from the top to bottom of the tube. Fractions 2–5 representing the lipid rafts⁸ were pooled and used for lectin pull-down assay.

Lectin Pull-Down Assay

Equal amounts of total lysates or membrane lipid raft fractions were incubated with 30 µl agarose-conjugated lectins at 4°C overnight. UT-A1 proteins precipitated by different lectins were detected by Western blot. Agarose-bound concanavalin A, wheat germ agglutinin, SNA, GNL, datura stramonium lectin, phaseolus vulgaris leucoagglutinin, and Tomato lectin (Lycopersicum esculentum lectin) were purchased from Vector Laboratories (Burlingame, CA). Maackia amurensis agglutinin was purchased from EY Laboratories.

Cell Surface Biotinylation Assay

Cell surface biotinylation assays were conducted as described previously.¹⁰

Cell Lysate Preparation and Western Blot

Cell pellets were solubilized in RIPA buffer and sonicated on ice. Extracts were clarified by centrifugation for 10 minutes at 10,000 rpm at 4°C. Protein concentration was determined by the Bradford method using the Bio-Rad protein assay (Bio-Rad, Richmond, CA). Proteins (approximately 20–50 µg/lane) were size separated by SDS-PAGE and then electroblotted to polyvinylidene difluoride membranes (Bio-Rad). Antibodies used in this study were UT-A1,⁸ actin (Sigma-Aldrich), PKC-α (Santa Cruz Biotechnology), Src (Santa Cruz Biotechnology), Raf-1 (Santa Cruz Biotechnology), horseradish peroxidase anti-rabbit IgG, and horseradish peroxidase anti-mouse IgG (Amersham BioSciences). Densitometry was performed using ImageJ.

Transepithelial Urea Flux Assay

UT-A1-MDCK cell urea transport activity was measured by transepithelial urea flux in Transwell inserts. Cells were grown in collagen-coated Costar Transwell inserts (Corning, Lowell, MA). Cells were treated with 2 mM sialic acid (Sigma-Aldrich) or 25 mM glucose for 24 hours. Epithelial barrier integrity was monitored before urea flux by transmembrane resistance measurement with an epithelial resistance meter. The method used for transepithelial ¹⁴C-urea flux has been described previously.³⁸

Oocyte Urea Flux Experiments

X. laevis oocytes were prepared and maintained as described in a previous report.⁸ Capped cRNAs were transcribed in vitro from linearized cDNAs with T7 polymerase using the mMESSAGE mMACHINE T7 Ultra Kit (Ambion, Austin, TX); 5 ng UT-A1 alone or together with 2 ng PKC-α was injected into each oocyte. Three days later, healthy oocytes were selected for functional study and protein expression. Urea transport activity was measured as described previously.⁸ Oocyte cell PM was prepared as reported by Leduc-Nadeau et al.⁴⁰ Urea flux data are expressed as means±SDs. A paired t test was used to assess statistically significant differences.

Disclosures

None.