Sphingomyelinase and Membrane Sphingomyelin Content

Determinants of Proximal Tubule Cell Susceptibility to Injury

Abstract

Ceramides acutely accumulate in proximal tubules during injury. Pathogenic relevance of this change is suggested by observations that adding ceramide to tubular cells alters superimposed hypoxic and toxic attack. Ceramide accumulation during cell injury is thought to arise from sphingomyelinase (SMase)-mediated sphingomyelin (SM) hydrolysis ± decreased catabolism. Thus, ceramide addition to cells cannot precisely simulate pathophysiologic events. Therefore, this study assessed direct effects of SMase activity on tubular cell viability under basal conditions and during superimposed attack. Cultured human proximal tubule (HK-2) cells were exposed to differing SMase doses. Its effects on cell phospholipids, ceramides, proliferation rates, and susceptibility to injury (ATP depletion, Fe-mediated oxidant stress) were assessed. Because SMase reduces cell SM content, the effect of exogenous SM on membrane injury (intact cells, isolated vesicles) was also tested. Finally, because SM decreases membrane fluidity, the impact of a fluidizing agent (A₂C) on membrane injury (phospholipase A₂, lipid peroxidation) was addressed. SMase reduced HK-2 SM content by approximately 33%, but only modest ceramide increments resulted (suggesting high endogenous ceramidase activity). SMase, by itself, caused no cell death (lactate dehydrogenase release). However, it was mildly antiproliferative, and it dramatically predisposed to both ATP depletion- and Fe-mediated attack. SMase also predisposed isolated vesicles to damage, suggesting that its impact on intact cells reflects a direct membrane effect. Adding SM to intact cells (or vesicles) mitigated ATP depletion and Fe- and phospholipase A₂-induced damage. In contrast, A₂C rendered membranes more vulnerable to attack. SMase predisposes tubular cells to superimposed ATP depletion and oxidant injury. This may be explained by SM losses, and not simply cytotoxic ceramide gains, given that SM can directly decrease cell/membrane damage. The ability of SM to decrease membrane fluidity may explain, at least in part, its cytoprotective effect.

Sphingomyelin (SM) traditionally has been viewed simply as a structural building block of plasma membranes, accounting for approximately 10 to 15% of total cellular phospholipid content. It resides predominantly in the outer leaflet of the plasma membrane bilayer, forming tight hydrophobic interactions with cholesterol (^1,2,3,4). Over the past several years, there have been increasing suggestions that, in addition to its structural functions, SM serves as a source of potentially important signaling molecules (^5,6,7). According to this view, a variety of extracellular or intracellular stimuli (such as tumor necrosis factor, interleukins, nerve growth factor, vitamin D, chemotherapeutic agents, heat shock) (^{5,6,7,8,9,10,11,12}) activate a family of sphingomyelinases (SMase), which cleave off SM's polar head group (phosphorylcholine), producing ceramide. The potential relevance of this process is suggested by observations that addition of exogenous ceramide to otherwise unchallenged cells can affect their homeostasis; as examples, altered cell cycle progression, proliferation, differentiation, and necrosis/apoptosis may result (^{5,6,7,8,9,10,11,12}).

Each of these processes has potential implications for both the development of, and recovery from, acute renal failure (ARF). Therefore, several recent studies have focused on whether ceramide accumulation is a consequence of acute ischemic/hypoxic and toxic proximal tubular cell attack. Indeed, this appears to be the case, because within several minutes of inducing acute tubular damage, approximately 50 to 200% ceramide increments result (^{13,14,15,16,17}). These elevations persist for at least 24 h post-renal injury, indicating that they do not simply represent a transient biologic event. Multiple biochemical pathways can contribute to this ceramide accumulation, including de novo synthesis (¹⁷), decreased ceramide catabolism (¹³), and SMase-mediated SM hydrolysis. The latter may then potentially contribute to injury-induced decrements in proximal tubular brush border membrane SM content (^18,19,20).

These types of observations have stimulated investigation of the “SMase-ceramide pathway” as a determinant of acute cell injury and repair. The general experimental approach has been to subject cultured cells to either “chemical hypoxic” or toxic injury in the presence or absence of exogenous ceramide treatment. Depending on the type of injury model studied, both injury potentiation (e.g., hypoxia) (^16,17) and attenuation (e.g., iron-mediated oxidant stress; arachidonic acid toxicity (^13,16) have been observed. While these findings underscore that ceramide can, indeed, influence tubular injury responses, they cannot be readily extrapolated to SMase-mediated ceramide generation for a number of important reasons. First, exogenous ceramide treatment tends to increase, rather than decrease, cellular SM content because provision of ceramide provides a substrate for de novo SM synthesis. This is the opposite situation to SMase-mediated ceramide generation, in which SM decrements, rather than increments, result. Second, commercially available long chain (physiologic) ceramides gain minimal, if any, intracellular access. Hence, when they are added to cultured cells, they cannot recapitulate the same membrane orientation that results from membrane SM hydrolysis. Third, the term “ceramide” refers to a heterogeneous family of compounds, consisting of sphingosine joined by amide linkage to any one of a number of fatty acids. For example, we have recently observed that after ischemic or toxic renal injury in the mouse, at least nine distinct ceramides accumulate, possibly in a disease-specific manner (¹⁵). Because most physiologic ceramides are not commercially available, one cannot simply add exogenous ceramides to re-create these ceramide accumulation patterns.

In view of these considerations, the present study has sought to identify the impact of SMase activity, and not simply exogenous ceramide treatment, on the expression of evolving renal tubular cell damage. To this end, cultured human proximal tubular (HK-2) cells were exposed to exogenous SMase, producing ceramide generation in the setting of SM hydrolysis. The impact of these changes on the expression of superimposed acute toxic and “chemical hypoxic” cell injury was then assessed. In addition, SMase's effects on cellular proliferation, both under basal conditions and after tubular cell injury, were assessed. The results of these studies form the basis of this report.

Materials and Methods

HK-2 Cell Culture Experiments: Culture Conditions

An immortalized human proximal tubular cell line (HK-2) (²¹) was used for all cell culture experiments. They were grown (37°C; 5% CO₂) in T75 Costar flasks (Cambridge, MA) with keratinocyte serum-free medium (K-SFM; Life Technologies, Grand Island, NY) containing 1 mM glutamine, 5 ng/ml epidermal growth factor, 40 μg/ml bovine pituitary extract, 25 U/ml penicillin, and 25 μg/ml streptomycin. At near confluence, the cells were trypsinized (²¹) and transferred to either additional T75 flasks (for passage) or to 24-well Costar plates for the conduct of specific experiments, as described below. Experiments were conducted 8 to 48 h after passage, with the cells in a subconfluent state.

Effects of SMase on ATP Depletion/Ca²⁺ Ionophore-Induced Injury

HK-2 cells were grown within cluster plates under routine culture conditions for 24 to 48 h. Then, each 24-well plate was divided into six groups, 4 wells per group, as follows: (1) continued control culture conditions; (2) addition of 1 U/ml bacterial sphingomyelinase (SMase; from Staphylococcus aureus, catalogue no. SE 108, Biolmol; Plymouth Meeting, PA); (3) a combined challenge of ATP depletion (7.5 μM antimycin A/20 mM 2-deoxyglucose) + cytosolic Ca²⁺ overload (10 μM Ca²⁺ ionophore A23187), as described previously (²²); and (4) the ATP depletion/Ca²⁺ ionophore challenge (“CAD” = Ca²⁺ ionophore + Antimycin + Deoxyglucose) conducted in the presence of 1 U/ml SMase. Because the SMase is provided in a glycerol carrier (final concentration of 0.125% when added to the cells), this same glycerol concentration was added to those cell wells that were paired to SMase treatment groups. (These same conditions were used with all subsequent SMase experiments described below.) After an overnight incubation (approximately 14 h), the extent of cell injury was assessed by percentage of lactate dehydrogenase (LDH) release (¹³). Each of these treatments was conducted on at least 4 wells of cells on four separate occasions. (Note: The 1 U/ml dose of SMase was selected for these experiments because pilot studies demonstrated that it caused both SM reductions and ceramide increases without exerting an independent cytotoxic effect. See Results.)

Effect of SMase on Fe-Mediated Oxidative Stress

HK-2 cells were cultured in 24-well cluster plates for 24 to 48 h and then wells within each plate were treated as follows: (1) continued control incubation; (2) treatment with 1 U/ml SMase; (3) Fe-mediated oxidative stress (7.5 μM ferrous ammonium sulfate, complexed to 7.5 μM 8-hydroxyquinoline [HQ], which facilitates Fe's intracellular access) (^13,23); and (4) combined simultaneous treatment with SMase + FeHQ. After completing an overnight 14-h incubation, the extent of lethal cell injury was assessed by calculating percentage of LDH release. Each of these treatments was conducted on at least 4 wells of cells on four separate occasions.

SMase Effects on the Expression of Arachidonic Acid Cytotoxicity

Previous studies have demonstrated that exogenous ceramide addition to cells can antagonize arachidonic acid's (C20:4) cytolytic effects (^13,16). The following experiment explored whether a blunting of C20:4 toxicity can also be observed under conditions of SMase-mediated ceramide generation. To this end, HK-2 cells within 24-well cluster plates were divided into four equal groups (6 wells each) as follows: (1) control incubation; (2) incubation with 1 U/ml of SMase; (3) C20:4 addition (25 μM; in ethanol; final concentration, 0.125%); and (4) combined SMase + C20:4 treatment. After completing a 14-h exposure, cell injury was gauged by percentage of LDH release. This experiment was repeated on four separate occasions, n = 6 wells each.

SMase Effects on HK-2 Cell Phospholipid, Cholesterol, and Ceramide Content

The following experiments ascertained the impact of SMase treatment on HK-2 cell SM and ceramide concentrations. In addition, its effects on other major membrane phospholipid classes and total cholesterol levels were assessed. Cholesterol levels were determined because: (1) cholesterol can independently influence tubular cell vulnerability to injury (²⁴); and (2) SMase has been reported to decrease cell cholesterol content in selected cell lines (^4,25).

Phospholipids. HK-2 cells were grown in eight T75 Costar culture flasks. After reaching near confluence (3 d), the cells were treated for 14 h with either 1 U/ml SMase or with SMase vehicle (n = 4 each). The cells were harvested from the flasks by detachment with a rubber policeman followed by centrifugation, as described previously (¹³). The recovered cells underwent chloroform:methanol extraction, and the lipids were analyzed by two-dimensional thin-layer chromatography (TLC) (²⁶). Specifically, the recovered lipids were analyzed for SM, phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), and phosphatidylinositol (PI) content. The amount of each phospholipid was quantified by the amount of recovered phosphate in each phospholipid spot recovered from the TLC plate (²⁶). The values were reported as the percentage to which each individual phospholipid contributed to the sum of the total quantified phospholipid mass (sum of SM + PC + PE + PI + PS).

Cholesterol Content. Six T75 flasks were seeded with HK-2 cells, and when near confluence, they were subjected to either continued control incubation or to SMase treatment, as noted above (n = 3 each). After completing a 14-h SMase challenge, the cells were recovered from each of the flasks as noted above, and then they were assayed for total cholesterol content (cholesterol assay kit; No. 352; Sigma Chemical Co., St. Louis, MO), as described previously (²⁴). In brief, this assay subjects total cholesterol to treatment with cholesterol esterase (to deacylate cholesterol esters) plus cholesterol oxidase. The latter then converts cholesterol to cholesterol-4-en-3-one + H₂O₂, the latter being quantified by a colorimetric technique. Cholesterol concentrations were expressed as nmoles/nmole phospholipid phosphate in the recovered HK-2 cell extracts.

Ceramide Content. Six T75 flasks were treated as above, three with and three without SMase for 14 h, and then they were subjected to lipid extraction. The samples were analyzed for individual ceramides, as determined by their specific constituent fatty acid (C16:0, C16:1, C22:0, C22:1, C24:0, C24:1, C24:2, C24:3). These analyses were conducted by a recently described liquid chromatography/mass spectrometric technique developed in this laboratory (¹⁵). Individual and total ceramides were expressed as pmoles ceramide/nmole phospholipid phosphate in the cell extracts.

Exogenous SM Treatment: Impact on ATP Depletion and Fe-Mediated Cell Injury

To assess whether SMase might affect cell injury via changes in SM, and not simply by generating ceramides, the effect of exogenous SM on cell injury responses was assessed. Three plates of HK-2 cells were incubated with either exogenous SM (500 μM; from bovine brain; S7004; Sigma Chemical Co.) or with the SM carrier (ethanol; final concentration, 0.5%) within 1 h of seeding them into 24-well cluster plates. After an 8-h exposure, the cells were either left under these same conditions or they were challenged with the ATP depletion/Ca²⁺ ionophore (“CAD”) injury protocol, as described above. The extent of cell injury was assessed 14 h later by determining percentage of LDH release.

To further determine SM's impact on cell injury, the above protocol was repeated, substituting FeHQ for the CAD challenge. Once again, cell injury was determined 14 h later by LDH release.

Renal Cortical Vesicle Experiments: SM Effects on Membrane Injury

The following experiments were undertaken to directly assess SM's impact on membrane damage, independent of integrated intact cell injury responses (such as metabolic conversions of test reagents).

SM Addition Experiments. Renal cortices were collected from eight mice, and the tissues were subjected to lipid extraction, as described previously (^24,26). The total amount of recovered phospholipid was determined by phosphate assay (²⁷). The samples were dissolved at a concentration of 500 nmol phospholipid/100 μl chloroform:methanol (2:1). To this extract was added differing amounts of exogenous SM (from bovine brain in ethanol; see above), creating 5, 10, and 20% additions (by phosphate content). The samples were redried under N₂, and then 800 μl of Hanks' balanced salt solution (with Ca and Mg) were added, followed by sonication for 30 min, forming vesicles.

Phospholipase A₂-Mediated Membrane Damage. Vesicles with 0, 5, 10, and 20% SM supplementation were incubated with Naja phospholipase A₂ (PLA₂; 0.05 U/ml; No. P 7778; Sigma Chemical Co.). After completing the incubations, the samples were reextracted in chloroform:methanol and subjected to two-dimensional TLC (²⁶). PLA₂-mediated phospholipid breakdown was gauged by quantifying phosphatidylcholine (PC) reductions and reciprocal lyso-phosphatidylcholine (LPC) increments. The values were expressed as PC/LPC ratios, with falling ratios indicating increasing PLA₂ activity. These experiments were repeated 6 times on different occasions.

Fe-Mediated Lipid Peroxidation Experiments. Vesicles that had been formed with the 0, 5, 10, and 20% SM additions were incubated at 37°C with 50 μM FeHQ. The extent of lipid peroxidation was assessed after completing 30-min incubations by determining malondialdehyde (MDA) content (²⁸). The results were compared with those observed in vesicles that received no FeHQ treatment.

Partial SM Depletion Experiments: Effects on PLA₂-Mediated Phospholipid Degradation. In these experiments, vesicles were reconstituted without exogenous SM. Rather, they were coincubated with SMase (0.25 U/ml) to partially deplete SM. The impact of the SMase-mediated SM reductions (approximately 40%) on PLA₂ (0.05 U/ml)-mediated PC degradation was assessed (coincubations ± one or both of the enzymes; n = 6 experiments assessing each treatment).

Effects of Membrane Fluidity on Fe- and PLA₂-Mediated Damage. SM, like cholesterol, decreases membrane fluidity. To assess the impact of the latter on the expression of membrane damage, vesicles were treated with a membrane-fluidizing agent, and then its impacts on PLA₂-mediated PC degradation and Fe-mediated lipid peroxidation were addressed, as follows.

PLA₂ Activity. Vesicles were incubated with PLA₂ as noted above either in the presence or absence of the membrane-fluidizing reagent A₂C [2-(2-methoxyethoxy)ethyl 8-(cis-2octyl-cyclopropyl) octanoate; M-9010; Sigma Chemical Co.) (²⁴). The A₂C was solubilized in DMSO, added to achieve a final 0.5 mM A₂C concentration. After completing 30-min incubations, PLA₂ activity was assessed by determining PC/LPC ratios, as noted above. Non-PLA₂-treated vesicles that were coincubated with and without A₂C/A₂C vehicle provided control PC/LPC values (n = 6 determinations each).

Fe-Mediated Lipid Peroxidation. The above experiment was repeated 4 times, substituting 50 μM FeHQ for the PLA₂ challenge. At the completion of 30-min incubations, the extent of Fe-mediated lipid peroxidation was assessed by determining MDA concentrations, as noted above.

SMase Effects on HK-2 Cell Proliferation

Basal Conditions. HK-2 cells were seeded at low density (0.75 × 10⁵ cells/well) for 24 h in 24-well cluster plates. They were then exposed to either 0, 0.1, 0.25, 0.5, or 1.0 U/ml SMase for an additional 48 h. At the completion of these incubations, cell proliferation/cell number was gauged by a tetrazolium dye uptake assay (methylthiotetrazole [MTT]), as described previously (²¹). These experiments were repeated 4 times on two separate occasions.

As a second indicator of SMase effects on HK-2 cell proliferation, two additional 24-well plates, prepared as noted above, were incubated for 48 h with 0, 0.25, or 0.5 U/ml SMase. Cell outgrowth was assessed by the crystal violet staining technique (²⁹), as described previously (³⁰).

Cell Proliferation after Fe-Mediated Injury

The following experiment assessed the impact of SMase on cell proliferation of sublethally damaged HK-2 cells. To this end, eight 24-well plates were cultured for 8 h and then exposed to a sublethal FeHQ dose (2.5 μM FeHQ for 18 h). After this treatment, the cells were cultured for an additional 48 h with 0.5 U/ml SMase. Cocultured cells subjected to either continuous control culture conditions, 48-h treatment with SMase but without a prior FeHQ challenge, or the FeHQ challenge without subsequent SMase treatment were used for comparison. After completing these protocols, relative cell numbers were assessed by MTT assay.

Statistical Analyses

All values are presented as means ± 1 SEM. Statistical comparisons were made by either paired or unpaired t test. If multiple comparisons were made, the Bonferroni method was applied.

Results

Effects of SMase on ATP Depletion/Ca²⁺ Ionophore Injury

Incubating HK-2 cells with SMase did not diminish cell viability, as assessed by a lack of increase in baseline percentage LDH release compared with coincubated control cells (Figure 1). However, the extent of CAD-induced cell injury was significantly accentuated by concomitant SMase treatment (Figure 1).

Figure 1: Effect of sphingomyelinase (SMase) treatment on HK-2 viability in the absence and presence of a combined ATP depletion/Ca²⁺ ionophore challenge (CAD = Calcium ionophore A23187 + Antimycin + Deoxyglucose). Cell injury was assessed after 14 h of treatment by percentage of lactate dehydrogenase (LDH) release. SMase had no independent effect on cell viability. However, it significantly exacerbated CAD-mediated cell death.

Effects of SMase on Fe-Mediated Cell Injury

As shown in Figure 2, the FeHQ challenge caused approximately 15% LDH release. The extent of this cytotoxicity was increased two- to threefold when the Fe challenge was conducted in the presence of SMase (Figure 2). Once again, SMase exerted no independent cytolytic effect.

Figure 2: Effect of SMase treatment on HK-2 viability under control conditions and in the presence of Fe (FeHQ)-mediated oxidant stress. In the absence of Fe, SMase had no effect on LDH release. However, it more than doubled the extent of LDH release induced by the Fe challenge.

Effects of SMase Treatment on the Expression of Arachidonic Acid Cytotoxicity

Incubating HK-2 cells with exogenous arachidonic acid (C20:4) caused approximately 65% LDH release (Figure 3). Unlike the results obtained with the CAD and FeHQ challenges, SMase decreased, rather than increased, this cytotoxicity.

Figure 3: Effect of SMase on arachidonic acid (C20:4) cytotoxicity. HK-2 cells were incubated with or without C20:4 (25 μM) for 14 h in the presence or absence of 1 U/ml SMase. C20:4 induced marked cytotoxicity, a result that was attenuated by concomitant SMase treatment. In the absence of C20:4, SMase did not affect LDH release.

SMase Effects on HK-2 Cell Phospholipid, Cholesterol, and Ceramide Content

Phospholipids. Incubating HK-2 cells for 14 h with SMase caused a 36% decline in absolute SM content (from 22.3 ± 2.2 to 14.2 ± 2.9 nmol/sample; P < 0.005). This was reflected by a relative decrease in its overall contribution to total plasma membrane phospholipid composition (from 11.7 to 8%) (Table 1). The percent decrement in SM was, in large part, matched by a reciprocal increase in percentage of PC in the SMase-treated cells.

Table 1: Phospholipid profiles in HK-2 cells subjected to 14-h incubations with either 1 U/ml SMase or under control conditions^a

Cholesterol. Exposing the HK-2 cells to 1 U/ml SMase for 14 h induced no change in their total cholesterol content (controls, 0.257 ± 0.03; SMase treatment, 0.259 ± 0.02; nmol cholesterol/nmol phospholipid phosphate; NS).

Ceramide. HPLC/MS analysis demonstrated C16 and C24, but not C22 ceramides (¹⁵) within HK-2 cells (Table 2). Varying degrees of fatty acid unsaturation were apparent within both the C16 and the C24 ceramide pools. Treatment with exogenous SMase caused surprisingly modest total ceramide increments (approximately 35 to 50% greater than that observed in control cells). The bulk of this increase was due to increases in the C16:0, C16:1, and C24:1 pools.

Table 2: Ceramide concentrations in HK-2 cells after a 14-h incubation with 1 U/ml bacterial SMase or under control conditions^a

Effect of Exogenous SM Treatment on HK-2 Cell Injury

As shown in Figure 4, incubating HK-2 cells with exogenous SM alone did not affect cell viability. The CAD challenge induced approximately 66% cell death. Approximately two-thirds of this injury was eliminated by the SM treatment (Figure 4, left). SM also attenuated FeHQ-mediated cell injury (P < 0.0005), albeit to a lesser degree (Figure 4 right). (Note: The greater cell injury induced by the CAD and FeHQ challenges in these experiments versus the SMase addition experiments reflects the difference in time the cells were in culture [8 to 9 h versus 24 to 48 h] before applying the cytotoxic challenges. In general, HK-2 cells are more vulnerable to injury if challenged 8 h versus 24 to 48 h after establishing the cultures. The shorter time frame was used in the SM addition experiments to increase cell injury, thereby allowing a more accurate quantitative assessment of SM's cytoprotective effect.)

Figure 4: Effect of exogenous SM pretreatment on the expression of ATP depletion/Ca²⁺ ionophore (CAD) and Fe-mediated HK-2 cell injury. Cells were pretreated with SM for 8 h and then subjected to the CAD and Fe challenge. In both instances, SM pretreatment mitigated the extent of cell death. In the absence of the CAD or the Fe challenge, SM had no effect on LDH release.

Isolated Vesicle Experiments

Impact of SM Additions on PLA₂-Mediated PC Degradation. Control vesicles had no discernible LPC present (and hence, no PC/LPC ratios are depicted in Figure 5). The PLA₂ challenge reduced vesicle PC content by 50 ± 4% (P < 0.001). A reciprocal increase in LPC resulted, producing a PC/LPC ratio of approximately 1.0 (Figure 5A). Raising vesicle SM content by exogenous SM addition caused a significant blunting of PC degradation, as reflected by a significant preservation of the PC/LPC ratio (Figure 5A) (^*P < 0.05) during PLA₂ treatment. This result was not dose-dependent, because essentially identical ratios were obtained with 5, 10, or 20% SM addition.

Figure 5: (A) Impact of sphingomyelin (SM) supplementation on phospholipase A₂ (PLA₂)-induced phosphatidylcholine (PC) degradation in isolated membrane vesicles. PLA₂ activity was quantified by assessing PC decrements and lysophosphatidylcholine (LPC) increments, with results expressed as the PC/LPC ratio. In the absence of PLA₂, no LPC was present, such that the baseline ratio = infinity. As depicted in Panel A, SM supplementations of 5, 10, and 20% each blunted the PLA₂-mediated PC/LPC ratio declines, indicating decreased PLA₂ activity. This was observed at all SM supplementation doses but in a non-dose-dependent manner. ^* P < 0.05 versus control vesicles that had no exogenous SM supplementation. (B) In this experiment, endogenous SM within vesicles was partially depleted (approximately 40%) by SMase treatment, and the effect of these SM reductions on PLA₂-mediated PC degradation was assessed. The SM reductions caused a significant increase in PLA₂-mediated PC degradation, as reflected by a significantly greater decline in the PC/LPC ratio. ^* P < 0.05. Thus, these data support those depicted in Panel A, indicating that SM mitigates PLA₂-induced PC degradation in isolated vesicles.

Impact of SM Depletion on PLA₂-Mediated PC Degradation. SMase treatment of the vesicles decreased their SM content by 40 ± 4% (P < 0.001). There was no associated change in PC, and hence, no LPC was apparent. When these partially SM-depleted vesicles were challenged with PLA₂, increased PC degradation/LPC generation resulted. This was reflected by a further lowering of the PC/LPC ratio, compared to that seen in vesicles subjected to PLA₂ treatment without SMase (Figure 5B).

Impact of SM Additions on Fe-Mediated Lipid Peroxidation. Baseline vesicle MDA concentrations were <1.0 μM. This result was unaffected by SM additions. Fe addition raised control vesicle MDA concentrations to 6.9 ± 1.1 μM. This result was significantly reduced by prior incorporation of SM into the vesicles (6.4 ± 0.9, 6.1 ± 1.1, and 5.7 ± 1.1 μM; with 5, 10, and 20% SM supplementation; P < 0.03 versus non-SM-supplemented vesicles).

Effects of Membrane Fluidity on Fe- and PLA₂-Mediated Damage. A₂C had no independent effect on vesicle PC content, and no LPC was observed. However, A₂C increased PLA₂-mediated PC deacylation, lowering the PC/LPC ratio by an additional 20% (Figure 6A).

Figure 6: Effect of the membrane-fluidizing agent, A₂C, on PLA₂-mediated vesicle deacylation (A) and on Fe-mediated lipid peroxidation, as gauged by malondialdehyde (MDA) concentrations (B). A₂C increased PLA₂-mediated reductions in the PC/LPC ratio (for details, see legend of Figure 5) and doubled Fe-mediated MDA increments. In the absence of PLA₂ or Fe, A₂C caused no LPC formation and it induced no change in MDA concentrations.

When only A₂C was added to the vesicles, no change in MDA concentrations resulted (Figure 6B). However, A₂C approximately doubled MDA generation in the presence of the Fe challenge (Figure 6B).

SMase Effects on HK-2 Cell Proliferation

Impact on Normal HK-2 Cells. When otherwise normal HK-2 cells were exposed to SMase treatment, a 10% suppression of HK-2 cell outgrowth was noted, as assessed by the MTT assay. This effect was maximal at the 1 U/ml SMase concentration (controls: 0.31 ± 0.01 MTT absorbance units; SMase at 0.1, 0.25, 0.5, and 1.0 U/ml: 0.31 ± 0.01, 0.31 ± 0.01, 0.30 ± 0.005, and 0.29 ± 0.01 absorbance units, respectively). These results were statistically significant at and above the 0.50 U/ml concentration (P < 0.05 to <0.0003). The crystal violet assay confirmed this approximate 10% decrease in HK-2 outgrowth in the presence of SMase (controls: 0.57 ± 0.02 absorbance units; 0.25 and 0.5 U/ml SMase: 0.48 ± 0.0.2 and 0.51 ± 0.02, respectively; P ≤ 0.025 with each concentration).

Impact on Cell Proliferation after Sublethal Fe-Mediated Injury. The sublethal Fe challenge, by itself, caused a 10 ± 1% decrease in cell outgrowth (MTT assay) compared with cocultured control cells (P < 0.001). SMase addition further decreased cell outgrowth in these post-Fe-damaged cells (to 79 ± 2% of controls; P < 0.001). However, because SMase alone caused a 10 ± 1% decrease in MTT uptake/cell proliferation (P < 0.01), the 21% decrease in cell outgrowth in the Fe + SMase-treated cells could be attributed to additive, rather than synergistic, Fe + SMase growth inhibitory effects.

Discussion

Several recent studies have demonstrated that ceramide accumulation, a hallmark of acute tubular damage (^{13,14,15,16,17}), can help determine tubular cell fate. For example, two laboratories have reported that exogenous ceramide addition to cultured proximal tubular cells exacerbates ATP depletion-mediated tubular cell injury (^16,17). Conversely, ceramide treatment attenuates Fe-mediated oxidative stress as well as arachidonic acid cytotoxicity (^13,16). As discussed previously, the relevance of results obtained with exogenous ceramide treatment, vis a[Combining Grave Accent] vis SMase-mediated ceramide generation, can be questioned because the latter, but not the former, decreases cell SM content. Because SM constitutes 10 to 15% of total plasma membrane phospholipid and serves as a critical modulator of membrane fluidity (^31,32), SMase-mediated ceramide generation in the setting of SM depletion could have different effects than exogenous ceramide treatment.

To evaluate the impact of SMase activity on evolving tubular injury, HK-2 cells were exposed to exogenous SMase under basal conditions and during superimposed ATP depletion/Ca²⁺ overload or iron-induced oxidant attack. That SM resides predominantly in the outer plasma membrane leaflet assures that exogenous SMase gains ready SM access. The use of SMase also assures that a full array of physiologic ceramides (i.e., with differing constituent fatty acids) is generated directly within the plasma membrane. It is noteworthy that most physiologic ceramides are not commercially available, and those that are undergo minimal cell uptake (³³). In the present studies, SMase-induced approximately 33% SM hydrolysis, decreasing it from approximately 12% to approximately 8% of the total cellular phospholipid mass. Despite this large SM reduction, only modest ceramide increments resulted. It should be recalled that with a normal 30:1 SM:ceramide ratio in HK-2 cells, 33% SM hydrolysis should yield an approximate 1000% ceramide increase. That total ceramide increased only by 33 to 50% indicates that the vast majority of generated ceramide was rapidly catabolized (presumably by ceramidases), thereby precluding marked ceramide increments. This finding of comparatively modest ceramide accumulation in the face of massive SM hydrolysis underscores our previous conclusion that the ceramide increments after acute tubular injury must stem, at least in part, from decreased ceramide catabolism, and not simply SM breakdown (¹³).

The present study demonstrates that SMase can substantially impact acute tubular cell damage. When HK-2 cells were challenged with ATP depletion/Ca²⁺ overload in the presence of SMase, an approximate doubling of LDH release resulted. SMase also exacerbated Fe-mediated cytotoxic death. These results cannot simply be explained by nonspecific SMase toxicity because SMase, by itself, caused no cell death. Furthermore, SMase attenuated, rather than exacerbated, arachidonic acid cytotoxicity, indicating the lack of a nonspecific “injury-provoking” effect. In sum, then, the present results indicate that SMase can differentially alter the evolution of acute tubular cell injury/cell death.

It is noteworthy that although SMase treatment generated only modest ceramide increments, it partially recapitulated results obtained with high (μM) dose exogenous ceramide additions: Both SMase and exogenous ceramide worsen ATP depletion/Ca²⁺ ionophore-induced injury, whereas each decrease arachidonic acid-mediated tubular cell death (^13,15,16,17) (present data). In contrast, SMase exacerbated Fe-mediated cytotoxicity (present data), whereas exogenous ceramide produced the opposite result (¹³). This latter difference illustrates that SMase's impact on evolving cell injury is not explained simply by ceramide generation. Rather, SMase-mediated SM decrements are likely involved, as discussed below.

A previous study from this laboratory indicates that plasma membrane cholesterol is a critical determinant of cellular resistance to injury (²⁴). This conclusion stems from observations that decreasing membrane cholesterol content (by decreasing synthesis or cell extraction) or its enzymatic modification (with cholesterol esterase or oxidase) dramatically increases tubular cell vulnerability to ATP depletion/Ca²⁺ ionophore or Fe-induced attack. A membrane-fluidizing reagent (A₂C) recapitulated this injury potentiation (²⁴), suggesting that cholesterol's influence on cell injury is mediated via its “anti-fluidity” effect. It is noteworthy that SM is tightly complexed to cholesterol within the plasma membrane, forming SM:cholesterol microdomains (^1,2,3,4,31). Because SMase-mediated SM reductions undoubtedly disturb these domains, it might then abrogate cholesterol's cytoprotective effect. Alternatively, if SM, like cholesterol, has a direct cytoprotective action, SMase could promote injury directly by reducing membrane SM content.

To explore this latter possibility, SM's impact on cell injury was assessed by incubating HK-2 cells with exogenous SM for 8 h (allowing for membrane uptake), and then cellular susceptibility to ATP depletion/Ca²⁺ ionophore and Fe toxicity were assessed. In both instances, substantial decrements in lethal cell injury resulted. It is noteworthy that in pilot studies conducted without this 8-h pretreatment period, SM failed to confer this same protective action. Thus, the SM-mediated attenuation of cell injury cannot simply be explained by inhibition of cellular toxin uptake. Because SMase may, under certain circumstances, decrease cellular cholesterol content (^4,25), it seemed possible that SMase's ability to exacerbate HK-2 cell injury could stem from reduction in cholesterol, rather than in SM. However, the fact that SMase did not alter HK-2 cell cholesterol content excludes this hypothesis. Finally, that SM supplementation of isolated vesicles directly blunted lipid peroxidation and PLA₂-mediated PC deacylation provides additional strong support for the concept that SM can, indeed, confer direct cytoprotection by stabilizing plasma membranes against attack. That A₂C exerted exactly the opposite effect as SM on PLA₂- and Fe-mediated vesicle damage strongly suggests that SM's membrane protective action, as with cholesterol, arises from an anti-fluidity effect.

It is conceivable that SM's ability to confer membrane, and hence cellular, resistance to injury could extend beyond the specific issue of SMase-induced SM hydrolysis. First, during acute tubular cell injury, SM moves from its concentrated position within the brush-border membrane to the basolateral membrane, a consequence of cytoskeletal disruption/loss of cell polarity (¹⁸) ± SMase activity. Given that brush-border damage is the earliest and most prominent morphologic consequence of ischemic cell injury (³⁴), it is tempting to postulate that these early SM losses contribute to this damage. Second, within 24 h of myohemoglobinuric renal injury, SM increments develop within renal cortex (²⁶). It is noteworthy that this coincides with the emergence of renal tubular resistance to superimposed attack (the so-called state of “acquired cytoresistance”) (³⁵). While multiple mechanisms contribute to the latter (e.g., increased heme oxygenase, ferritin, cholesterol, heat shock proteins, etc.) (^36,37,38), an increase in membrane SM content might also play a role. These two considerations underscore potential broad-based implications of altered SM expression on cellular resistance to superimposed attack.

In addition to altering tubular cell injury, SMase activity could potentially affect the evolution of ARF by retarding tubular cell proliferation after an ischemic or toxic insult. Noteworthy in this regard are reports that SMase can be antimitogenic, an effect linked to ceramide generation (^10,11). To our knowledge, this issue has not been addressed previously in proximal tubular epithelia. The present experiments indicate that SMase can, indeed, inhibit HK-2 cell outgrowth, but the effect appears quite mild (approximately 10% reductions), as gauged by two independent cell proliferation assays. It is tempting to postulate that this negative effect would be magnified in sublethally damaged tubular cells that already have a diminished proliferative capacity. However, this does not appear to be the case, given that SMase's antiproliferative activity was essentially identical in control cells and in cells that were sublethally damaged by prior Fe-mediated injury.

In conclusion, the present study demonstrates that SMase activity can potentially exacerbate both ATP/Ca²⁺ overload and Fe-mediated oxidant cell death. This action does not simply reflect nonspecific SMase “toxicity,” because the dose of SMase used, by itself, did not impact cell viability, and it decreased, rather than increased, arachidonic acid's cytotoxic effect. The ability of SMase to increase cell injury may stem from SM decrements, and not simply cytotoxic ceramide gains. This is based on observations that SM exerts membrane-protective effects, both in intact HK-2 cells as well as in isolated vesicles. Despite SMase's ability to evoke marked SM hydrolysis, surprisingly modest ceramide increments result. This suggests that high endogenous ceramidase activity within cells can markedly blunt ceramide accumulation, underscoring our previous assertion that the dramatic ceramide increments that develop during renal injury arise, in part, from ceramidase inhibition, and not simply SM hydrolysis. Finally, despite the fact that SMase can exert potent antiproliferative effects, its impact on proximal tubule cell growth appears quite modest, both under basal conditions and in the setting of sublethal tubular damage. If these results are relevant to in vivo renal injury, they suggest a greater potential for SMase to impact the induction, rather than the recovery, phase of ischemic and toxic ARF.

This work was supported by research grants from the National Institutes of Health (RO1 DK-38432; RO1 DK-54200). We thank Thomas Kalhorn for his assistance with the ceramide analysis.