All chemicals were purchased from Sigma (MO) unless otherwise stated. Soybean trypsin inhibitor was purchased from GIBCO (NY). Rabbit polyclonal antibody to total AMPK, rabbit polyclonal antibody to total alpha subunit of AMPK, and mouse monoclonal antibody to the phosphorylated alpha subunit of AMPK were purchased from Cell Signaling Technologies (MA). Rabbit polyclonal antibody to phosphorylated ACC1/2 was purchased from Upstate (NY). Rabbit polyclonal antibodies specific for the α1 and α2 isoforms of the catalytic domain of AMPK were purchased from Abcam (MA). Monoclonal antibody to β-actin was purchased from Chemicon (CA).

The generation and phenotype of whole body α1^-/- and α2^-/- KO mice have been described previously 19 20 . α1^-/- mice and their WT controls were bred on a mixed background (C57BL6J/129Sv), whereas α2^-/- KO mice and their WT controls were bred on a C57BL6J background. Heterozygous α1^-/- and –α2^-/- KO mice, used only for breeding, were obtained from the European Mouse Mutant Archive and bred in the Department for Laboratory and Animal Research at Stony Brook Hospital. Only homozygous KO mice, male and female, were used for experimentation. All studies were approved by the IACUC at Stony Brook Hospital. Total body weights and kidney weights were comparable amongst KO and WT mice (data not shown).

Techniques for the primary culture of MPT cells are well established in our laboratory 21 22 . Briefly, kidneys were harvested from mice at 4–6 weeks of age. Cortical tissue was finely minced and incubated in Hanks’s solution containing collagenase and soybean trypsin inhibitor. After collagenase digestion, 10% horse serum was added, and the tissue fragments were gently centrifuged. After two washes with DMEM, tissue fragments were suspended in growth medium, a serum-free, defined medium consisting of DMEM/Ham’s F-12 media (1:1) containing 2 mM glutamine, 15 mM HEPES (pH 7.2), 5 μg/ml transferrin, 5 μg/ml insulin, 50 mM hydrocortisone, 500 U/ml penicillin, and 50 mg/ml streptomycin. The tissue fragments, suspended in growth medium, were then plated on plastic tissue cultures plates and incubated in a humidified air/CO₂ incubator (5% v/v) at 37°C.

Fragments of tubules adhere to the culture dish, and MPT cells grow out of the tubules to form confluent monolayers over a 4–5 day period. MPT cells express megalin and other markers specific for proximal tubular cells. Five days after plating, the medium was changed to experimental medium, which consists of DMEM/Hams F12 containing HEPES (pH 7.2), 15 mM glutamine, 500 U/ml penicillin, and 50 mg/ml streptomycin. No insulin, hydrocortisone, or transferrin was added to this medium. Experiments were begun on the sixth day after plating.

MPT cells were exposed to metabolic stress using antimycin A, a mitochondrial inhibitor. We examined the effect of antimycin A in the presence of three concentrations of dextrose, 10, 5, and 2.5 mM. The severity of metabolic stress increases progressively, (i.e. cell ATP levels fall progressively), as the dextrose concentration in the medium is reduced from 10 to 2.5 mM. We refer to this model of metabolic stress as “graded” ATP depletion. We have used graded ATP depletion as a model of metabolic stress in previous publications 22 . Control (unstressed) MPT cells were incubated in 5 mM dextrose in the absence of antimycin.

ATP levels were determined using previously described methods 22 . Briefly, cell ATP content was measured by luciferase assay in cell lysates and normalized to total cellular protein, as assessed by bicinchoninic acid (BCA) protein assay (Pierce, Rockford, NY). Control ATP levels, obtained in cells incubated in dextrose-containing medium (5 mM) in the absence of antimycin, were expressed as nanomoles of ATP per milligram of cell protein (nM ATP/mg protein). ATP levels obtained during metabolic stress were expressed as a percentage of ATP levels under control conditions.

Immunoblotting was performed as previously described 23 . Briefly, cells were washed with ice-cold PBS, then lysed in ice-cold cell lysis buffer (20 mM Tris–HCl, pH 7.5, containing 130 mM NaCl, 1% v/v Triton X-100, 0.5% w/v deoxycholate, 0.1% w/v SDS, 1 mM DTT, 10 mM sodium pyrophosphate, 5 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, and 200 μM sodium orthovanadate). Lysates were centrifuged at 10,000 × g for 10 min at 4°C, and the supernatants were stored at -70°C. Protein samples, 20 μg per lane, as determined by BCA protein assay, were boiled in 6× reducing sample buffer, electrophoresed on SDS-polyacrylamide gels, and transferred to nitrocellulose membranes (BIO-RAD, Hercules, CA). Membranes were blocked with either 2.5% bovine serum albumin or 5% dry milk in TBS, before probing with primary antibody. After incubation with the appropriate secondary antibody, immunoreactive bands were visualized by the Western Lightning Chemiluminescence Reagent Plus (PerkinElmer, Boston, MA).

Cell viability was determined using the LIVE/DEAD Assay Kit purchased from Molecular Probes™ and used according to the manufacturer’s instructions. In brief, MPT cells were stained with ethidium homodimer-1 (EthD-1) and calcein AM. Live cells are identified by their ability to convert calcein AM, a non-fluorescent cell-permeant agent to calcein, an intensely fluorescent dye (excitation/emission wavelengths, ~495 nm/~515 nm) that is retained within live cells. Dead cells are identified by nuclear staining for EthD-1, which only enters cells with damaged plasma membranes and, upon binding to nucleic acids, undergoes a 40-fold enhancement of fluorescence (excitation/emission wavelengths, ~495 nm/~635 nm), thereby producing a bright fluorescence in dead cells. Since both dyes are essentially non-fluorescent before interacting with cells, background fluorescence is inherently low. Live and dead cells were quantitated using flow cytometry (FACScan, BD Biosciences), and data were analyzed using CELLQuestPro Version 3.3 (BD Biosciences). Cells were first analyzed by forward versus side scatter, and gated to remove debris, cell fragments, and cell aggregates. The proportion of live cells in each sample was expressed as a percent of the total number of cells analyzed (10,000/sample).

All data are presented as mean ± standard error (SE). Student’s t-test was used for comparing cell ATP levels and densitometry of immunoblots. The Bonferroni correction was applied when multiple comparisons were made. The viability of MPT cells cultured from KO versus WT mice and subjected to metabolic stress was compared by ANOVA for repeated measures using STATA® Data Analysis and Statistical Software. All p values <0.05 were considered statistically significant.

We determined the effect of graded ATP depletion, induced by exposing MPT cells to antimycin A and varying concentrations of dextrose, on cell viability, as assessed by flow cytometry. Cell viability was comparable in MPT cells from KO versus WT mice under unstressed control conditions (10 mM dextrose, no antimycin) (data not shown). In the presence of antimycin, the percentage of viable MPT cells from AMPK KO and WT mice decreased progressively as the concentration of dextrose was decreased (Figure 1). Nevertheless, at each dextrose concentration, the survival of MPT cells from α1^-/- or α2^-/- KO mice was no different than that of MPT cells from WT controls (Figure 1).

Cytosolic ATP levels in control unstressed MPT cells incubated in 10 mM dextrose without antimycin for 4 hrs were no different for KO versus WT controls (21 ±8 nM/mg protein versus 23 ± 6 nM/mg protein for α1^-/- mice; and 26 ± 9 nM/mg protein versus 21 ± 7 nM/mg protein for α2^-/- mice). Incubation of MPT cells in the presence of antimycin for 4 hrs led to a reduction in ATP levels. The severity of ATP depletion was inversely proportional to the concentration of dextrose in the medium (Figure 2). However, the severity of ATP depletion did not differ in MPT cells from KO versus WT mice at any dextrose concentration (Figure 2).

We used immunoblotting to determine the relative expression of the α1 and α2 isoforms of AMPK, as well as total α domain expression (α1 + α2 isoforms), in lysates obtained from whole kidney cortex (Figure 3) and confluent monolayers of MPT cells (Figure 4), both derived from AMPK KO mice and their WT controls. As expected, we detected no expression of the deleted α1 or α2 isoform in lysates from either whole kidney cortex (Figure 3) or MPT cell cultures (Figure 4) obtained from the α1^-/- and α2^-/- mice, respectively. Notably, however, expression of the non-deleted α-isoform was markedly increased in kidney cortex and MPT cell cultures from AMPK KO mice when compared to their WT controls (Figures 3 and 4). As a result, total alpha domain expression was comparable in kidney cortex and MPT cell cultures for both types of AMPK KO mouse versus its WT controls (Figures 3 and 4).

We next compared the degree to which AMPK is activated by metabolic stress in MPT cells from AMPK KO versus WT mice. Metabolic stress was induced by treatment of cells with antimycin in the presence of 5 mM dextrose. Activation of AMPK was assessed by immunoblotting for phosphorylation of Thr¹⁷² within the catalytic α domain of AMPK. Upon activation, AMPK phosphorylates a number of downstream targets 1 3 4 5 . As a further measure of AMPK activation, we evaluated the extent of phosphorylation of ACC at Ser⁷⁹, an event that inhibits ACC activity. We chose ACC, since it is one of the more thoroughly studied downstream targets of AMPK 1 .

Treatment with antimycin increased phosphorylation of both AMPK and ACC to a comparable extent in MPT cells from α1^-/- and α2^-/- mice versus their WT controls (Figure 5). These data suggest that the identity of the alpha domain isoform does not influence the degree to which AMPK is activated by antimycin-induced metabolic stress, and that each isoform can substitute for the absence of the other.

We examined the effect of compound C (CC), a pharmacologic inhibitor of AMPK, on AMPK activity during metabolic stress in MPT cells from α2^-/- and WT mice. Antimycin increased the phosphorylation of both AMPK and its downstream target, ACC, in MPT cells from α2^-/- and WT mice (Figure 6). CC comparably reduced the stress-induced increase in phosphorylation of AMPK and ACC in MPT cells from α2^-/- versus WT mice (Figure 6). Similar results, in the absence and presence of CC, were obtained in MPT cells from α1^-/- versus WT mice (data not shown).

We next examined the effect of CC on MPT cell viability during metabolic stress. MPT cells from α2^-/- mice and their WT controls were subjected to graded ATP depletion (antimycin in the presence of varying concentrations of dextrose) for 16–18 hrs, in the presence of either CC (20 μM) or its vehicle. Control cells were incubated in dextrose without antimycin. In the absence of CC, metabolic stress induced a comparable decrease of viability in MPT cells from α2^-/- versus WT mice (Figure 7). Although inhibition of AMPK with CC reduced MPT cell viability at all degrees of metabolic stress, the reduction in viability induced by CC was comparable in MPT cells from α2^-/- versus WT mice (Figure 7). Furthermore, regardless of the presence or absence of CC, ATP levels in metabolically stressed MPT cells did not differ for α2^-/- versus WT mice (data not shown). These data are consistent with those of immunoblotting (Figures 5 and 6).

We performed similar experiments using MPT cells from α1^-/- versus WT mice. In the absence of CC, metabolic stress (antimycin plus 5 mM dextrose) reduced cell viability to a similar extent in MPT cells from α1^-/- versus WT mice (64 ± 7% and 59 ± 8% of control, respectively). CC comparably exacerbated the stress-induced loss of cell viability in MPT cells from α1^-/- versus WT mice (35 ± 6% and 38 ± 7% of control, respectively). Thus, the effects of CC on the viability of stressed MPT cells was similar for α1^-/- and α2^-/- mice, as compared to their WT controls.

MPT cells from α1^-/- or WT mice were infected with either control (scrambled) shRNA or shRNA designed to knock down expression of the α2 isoform of AMPK. Infection with control shRNA did not alter expression of either the α2 isoform of AMPK or total α-domain AMPK in MPT cells from α1^-/- and WT mice (Figure 8). In contrast, infection with anti-α2 shRNA significantly reduced expression of the α2 isoform of AMPK as well as total alpha domain AMPK in MPT cells from both α1^-/- and WT mice (Figure 8).

We next determined the effect of knockdown of the α2 isoform of AMPK on metabolic stress-induced activation of the AMPK pathway. MPT cell monolayers from α1^-/- and WT mice were infected with either control or anti-α2 shRNA. Following infection, MPT cells were incubated in dextrose (5 mM) in the absence or presence of antimycin. In cells infected with control shRNA, metabolic stress induced marked phosphorylation of both AMPK and ACC (Figure 9). Infection with anti-α2 shRNA led to a marked but comparable reduction in phosphorylation of AMPK and ACC in MPT cells from α1^-/- versus WT mice (Figure 9).

Finally, we examined the effect of knockdown of the α2 isoform of AMPK on the viability of metabolically stressed MPT cells. MPT cells from α1^-/- and WT mice were infected with either control shRNA or anti-α2 shRNA, and then incubated in the absence or presence of antimycin plus varying concentrations of dextrose. After 16–18 hrs, the percentage of viable cells was determined by flow cytometry. As with pharmacologic inhibition of AMPK by CC, knockdown of the α2 domain comparably decreased the viability of metabolically stressed MPT cells from α1^-/- versus WT mice (Figure 10). Regardless whether cells were infected with control or anti-α2 shRNA, ATP levels did not differ in metabolically stressed MPT cells from α1^-/- versus WT mice (data not shown). Taken together, our studies on the effects of AMPK inhibition, accomplished either pharmacologically or molecularly, demonstrate antimycin-induced metabolic stress comparably activates the α1 and α2 isoforms of AMPK, and that either isoform can substitute for the other in ameliorating stress-induced cell death.