Main

Bilirubin has long been known to cause brain toxicity(1, 2). Its effects on the brain in the jaundiced neonate range from mild reversible lethargy and changes in evoked potentials, to devastating permanent brain damage or death in the syndrome ofkernicterus. After the discovery that bilirubin inhibited respiration in a coarse rat brain homogenate(3), bilirubin has been found to modulate, most frequently in an inhibitory manner, most of the biologic systems in which it has been tested(4). The widespread nature of these bilirubin effects indicate that the search for mechanisms which could be responsible should be directed toward mechanisms involved in a variety of cellular processes. Reversible phosphorylation dephosphorylation of proteins has been found to be a key mechanism for the regulation of both neuronal and nonneuronal cellular functions(5).

We have previously shown that bilirubin inhibits phosphorylation of the synaptic vesicle-associated protein synapsin I in situ in isolated, intact nerve terminals (synaptosomes) in a time- and dose-dependent manner(6). Because phosphorylation of synapsin I is believed to be involved in the regulation of neurotransmitter release(7), these findings suggested that some of the effects of bilirubin on the brain, specifically inhibition of synaptic activation(8), could partly be caused by inhibition of protein phosphorylation. However, the mechanism whereby bilirubin interferes with protein phosphorylation in vitro(9–11), or in situ(6), where it might be interpreted as being partly secondary to inhibition of oxidative phosphorylation(3, 12), has remained unclear.

In the present study we demonstrate an inhibitory effect of bilirubin on the phosphorylation of a variety of substrate proteins catalyzed by purified protein kinases in vitro in assays devoid of ATP-generating organelles. Moreover, kinetic analysis of the effects of bilirubin on the activity of the isolated catalytic subunit of PKA, using [32P]ATP and a synthetic peptide (consisting of residues 57-72 of the protein termed“phospholemman”(13)) as substrates, indicate that bilirubin has a noncompetitive mechanism of action. Partial data from these studies have been presented in abstract form(14, 15).

METHODS

[32P]ATP was from ICN Biomedicals, Inc. (Irvine, CA). Ion exchange resin AG 1X-8 (acetate form) was from Bio-Rad (Richmond, CA). A synthetic peptide, comprising the extreme 15 carboxy-terminal amino acids (residues 58-72 of the canine protein, GTFRSSIRRLSTRRR in single letter code) of phospholemman(13) was from The Peptide Synthesis Facility, Forskningsparken, University of Oslo, Norway. Bilirubin, biliverdin, BSA, histone H1, histone 2A, and the catalytic subunit of PKA were from Sigma Chemical Co. (St. Louis, MO). Protein kinase C was purified from rat brain(16). Cyclic GMP-dependent protein kinase and Ca2+-calmodulin-dependent protein kinase II, as well as calmodulin, synapsin I, glycogen synthase, myosin light chain, and DARPP-32 were gifts from Dr. Angus C. Nairn, The Rockefeller University, New York, NY. Other reagents, of analytical grade or better, were from standard commercial suppliers.

Protein phosphorylation. Phosphorylation of proteins was performed in a medium containing 50 mM HEPES (pH 7.4) and 10 mM MgCl2 with [32P]ATP essentially as described elsewhere(17). Enzymes included the catalytic subunit of PKA, purified cGMP-dependent protein kinase holoenzyme in the presence of 10 μM 8-Br-cGMP, protein kinase C in the presence of 0.5 mM CaCl2 plus phosphatidylserine (50 μg/mL) and diolein (2 μg/mL), or Ca2+-calmodulin-dependent protein kinase II in the presence of 0.5 mM CaCl2 and 1.5 μM calmodulin (all enzymes at 10 nM final concentration). Bilirubin (1-320 μmol/L, dissolved in 0.1 N NaOH) and BSA(molar ratio 10:1 [bilirubin:BSA], 0.1-32 μmol/L BSA added to the control incubates) were added to the medium, and the reactions were initiated by addition of [32P]ATP (final concentration, 50 μM, 146 GBq/mmol). Substrate proteins included synapsin I, glycogen synthase, histone I, histone IIA, DARPP-32, G-substrate, and myosin (all proteins at a concentration of 100μg/mL). Only one substrate protein and one kinase were present in each assay. Samples were incubated in a shaking water bath at 30°C for 30 s, the reactions were terminated by addition of SDS-containing “stop solution”(18). Samples were subjected to SDS-PAGE using the buffers of Laemmli(18), and the phosphorylated proteins were located by autoradiography and cut out of the gels. Radioactivity was determined by liquid scintillation spectrometry.

Phosphorylation of synthetic peptide. Phosphorylation of the synthetic peptide was analyzed by incubations in a medium containing 20 mM Tris-Cl (pH 7.4) and 10 mM MgCl2, in the absence or presence of bilirubin (final concentrations 0-300 μM, solubilized in 0.1 N NaOH) or, in some experiments, biliverdin (50 μM) and BSA as shown in the figure legends. The assays were performed either with [ATP] kept constant (0.2 mM final concentration) and peptide concentrations being varied between 5 and 100μM or, alternatively, peptide concentrations being kept at 100 μM, whereas [ATP] was varied between 1 and 90 μM (final concentration). Reactions were initiated by addition of the catalytic subunit of PKA (final concentration, 2.5 nM). Samples were incubated in a shaking water bath at 30°C for 5 min, and the reactions were terminated by addition of acetic acid (final concentration, 10%, vol/vol)(19). Phosphorylated peptide was isolated and quantitated by anion exchange chromatography on AG 1X-8 resin(20) and Cerenkov counting of the eluates. Initial experiments verified that these incubations led to initial rate conditions(19). Kinetic constants were assessed by examination of Lineweaver-Burk plots of 1/v versus 1/[S], or Hanes-Woolf plots of [S]/v versus[S](21).

Miscellaneous methods. Protein was analyzed by the method of Smith et al.(22).

RESULTS

Bilirubin effects on protein phosphorylation. Bilirubin appears to affect the phosphorylation of a number of proteins in isolated nerve terminals in situ(6). We therefore examined whether bilirubin changed in vitro phosphorylations catalyzed by a number of distinct protein kinases, using substrate proteins with widely different characteristics. Table 1 shows that bilirubin was able to inhibit both cyclic nucleotide-dependent, Ca2+-calmodulin-dependent, and Ca2+-dependent protein kinases. Under the experimental conditions used, and using both basic and acidic proteins as substrates, the IC50 values could be determined to range between 20 and 125 μM bilirubin. Albumin is known to stabilize solutions in phosphorylation assays, and this was apparent in some of the experiments where the presence of higher concentrations of albumin resulted in increased incorporation of radioactive phosphate. We have corrected for this apparent stimulation by defining the result from the BSA-containing control incubate as 100% and correcting the results from the bilirubin-containing assays correspondingly (Table 1).

Table 1 Inhibitory effect of bilirubin on protein phosphorylation in vitro
Full size table

Bilirubin effects on phospholemman peptide phosphorylation. Given the complex domain and subunit structure of many protein kinases(23), and because some of the substrate proteins used have limited solubility, the exact mechanism of bilirubin inhibition was not determined by the approach described above. Rather, we examined in more detail the effects of bilirubin on a “minimal” phosphorylation system, consisting of the isolated catalytic subunit of PKA, which represents a“pure” catalytic kinase domain essentially devoid of regulatory components(23), and a water-soluble, synthetic peptide derived from the intracellular domain of the membrane protein termed phospholemman(13) as substrate(19). Kinetic studies and sequencing analysis have shown that this phosphorylation reaction, which occurs on a single serine residue(Ser-68), obeys Michaelis-Menten kinetics, with Lineweaver-Burk plots indicating an apparent Km value of 7 μM(19). The phospholemman peptide therefore appears to be a good substrate for analysis of PKA-catalyzed reactions.

Given the stimulatory effects of BSA seen in some of the experiments described in Table 1, and because BSA is routinely used to retard flocculation of bilirubin in aqueous solutions, we first examined whether BSA had any effect on peptide phosphorylation in the presence of bilirubin or biliverdin, the immediate (and apparently nontoxic) precursor of bilirubin. In the presence of biliverdin (50 μM final concentration), which by itself did not change the rate of peptide phosphorylation, BSA at 0.15μM had no effect, whereas at 5 μM concentration it increased the rate of the reaction (Fig. 1). Similarly, in the presence of bilirubin (50 μM final concentration), which by itself decreased the rate of peptide phosphorylation by approximately 30% when compared with control values (Fig. 1), BSA also increased phosphorylation of the peptide when added at concentrations above 5 μM. Indeed, despite the continued presence of 50 μM bilirubin, BSA at 50 μM concentrations induced an approximately 25% increase in peptide phosphorylation compared with that found in the absence of both BSA and bilirubin.

Figure 1
figure 1

BSA modulates the effects of bilirubin on the activity to the catalytic subunit of PKA in vitro. Phospholemman peptide (0.1 mM) was incubated with [32P]ATP (0.2 mM) in a medium containing 20 mM Tris-Cl (pH 7.4) and 10 mM MgCl2, in the absence or presence of bilirubin or biliverdin (both at final concentration 50 μM, solubilized in 0.1 N NaOH) and variable amounts of BSA. Reactions were initiated by addition of the catalytic subunit of PKA (final concentration, 2.5 nM), and samples were incubated in a shaking water bath at 30°C for 5 min. Reactions were terminated by addition of acetic acid (final concentration, 10%, vol/vol). Phosphorylated peptide was isolated and quantitated by anion exchange chromatography and Cerenkov counting of the eluates. Initial experiments verified that these incubations led to initial rate conditions. The results represent mean ± SD of triplicate determinations from one representative experiment, which was repeated twice. The results were analyzed with analysis of variance: F = 27.8, p < 0.0001. Theasterisks denote differences that were significant at p< 0.05 compared with control.

Full size image

Because the inhibitory effect of bilirubin could be reversed by high levels of BSA, presumably due to the well known bilirubin-binding capacity of BSA, we performed the remaining experiments with 0.15 μM (10 μg/mL) BSA. Under these conditions bilirubin induced a dose-dependent decrease in PKA-catalyzed phosphorylation of the phospholemman peptide, ultimately reaching approximately 50% of control values, and with half-maximal inhibition seen at 25-50 μM bilirubin (Fig. 2). Kinetic analysis of the reaction showed that when the concentration of the peptide substrate was varied, bilirubin decreased the maximal velocity of the reaction by 50-85%, without changing the apparent Km value (Fig. 3). Similarly, in experiments where the concentration of ATP was varied, bilirubin was also found to decrease the maximal velocity of the initial reaction, without changing the apparentKm value (Fig. 4). Hence, with phospholemman as substrate, bilirubin behaved as a noncompetitive inhibitor of the catalytic subunit of PKA. Graphical analysis of the data by means of Hanes-Woolf plots (not shown) gave similar conclusions(21).

Figure 2
figure 2

Bilirubin induces a dose-dependent inhibition of the activity of the catalytic subunit of PKA in vitro. Phospholemman peptide (50 μM) was incubated with [32P]ATP (0.2 mM) in a medium containing 20 mM Tris-Cl (pH 7.4), 10 mM MgCl2, and 0.15 μM BSA, in the absence or presence of bilirubin (final concentration, 1-100 μM, solubilized in 0.1 N NaOH). Reactions were initiated by addition of the catalytic subunit of PKA (final concentration, 2.5 nM), and samples were incubated in a shaking water bath at 30°C for 5 min. Reactions were terminated by addition of acetic acid (final concentration, 10%, vol/vol). Phosphorylated peptide was isolated and quantitated by anion exchange chromatography and Cerenkov counting of the eluates. The results represent mean ± SD of triplicate determinations taken from two independent experiments. The results were analyzed with analysis of variance: F= 48.0, p < 0.0001. The asterisks denote differences that were significant at p < 0.05 compared with control.

Full size image
Figure 3
figure 3

Bilirubin induces noncompetitive inhibition of the activity of the catalytic subunit of PKA with respect to peptide substratein vitro. Phospholemman peptide (5-100 μM) was incubated with[32P]ATP (0.2 mM) in a medium containing 20 mM Tris-Cl (pH 7.4), 10 mM MgCl2, and 0.15 μM BSA in the absence (B-) or presence(B+) of bilirubin (final concentration, 50 μM, solubilized in 0.1 N NaOH). Reactions were initiated by addition of the catalytic subunit of PKA(final concentration, 2.5 nM), and samples were incubated in a shaking water bath at 30°C for 5 min. Reactions were terminated by addition of acetic acid (final concentration, 10%, vol/vol). Phosphorylated peptide was isolated and quantitated by anion exchange chromatography and Cerenkov counting of the eluates. The results, which represent 32P incorporated into peptide substrate (cpm), denote mean ± SD of triplicate determinations from five independent experiments which gave similar results. They were fitted by linear regression and are presented as a plot of V (cpm)vs [S] (μM) (inset), and as a double reciprocal 1/V vs 1/[S] Lineweaver-Burk plot(21). Vmax = 0.225 μM/min/mg protein, Km = 6.8 μM.

Full size image
Figure 4
figure 4

Bilirubin induces noncompetitive inhibition of the activity of the catalytic subunit of PKA with respect to ATP substratein vitro. Phospholemman peptide (100 μM) was incubated with[32P]ATP (1-90 μM) in a medium containing 50 mM HEPES (pH 7.4), 10 mM 2-mercaptoethanol, 10 mM MgCl2, and 0.15 μM BSA in the absence or presence of bilirubin (final concentration 50 μM, solubilized in 0.1 N NaOH). Reactions were initiated by addition of the catalytic subunit of PKA(final concentration, 2.5 nM), and samples were incubated in a shaking water bath at 30°C for 5 min. Reactions were terminated by addition of acetic acid (final concentration, 10%, vol/vol). Phosphorylated peptide was isolated and quantitated by anion exchange chromatography and Cerenkov counting of the eluates. The results, which represent 32P incorporated into peptide substrate (cpm), denote mean ± SD of triplicate determinations from two independent experiments which gave similar results. They were fitted by linear regression and are presented as a plot of V (cpm) vs[S] (μM) (inset), and as a double reciprocal 1/V vs 1/[S] Lineweaver-Burk plot(21).

Full size image

DISCUSSION

The results presented in this study demonstrate that bilirubin interferes with several protein phosphorylation reactions in vitro. The fact that these effects were observed in the absence of respiratory enzymes or organelles suggests that the previously observed inhibitory effect of bilirubin on the phosphorylation of synapsin I in synaptosomes(6) was not solely due to the bilirubin-induced inhibition of oxidative phosphorylation described by Zetterström and Ernster(12), and that a direct effect of bilirubin on protein phosphorylation system might have been responsible. Our results further indicate that bilirubin may achieve this effect by direct interaction with one or more protein phosphorylation system components. Using a“minimal” phosphorylation system based on the readily available isolated catalytic subunit of PKA and a synthetic peptide in vitro, bilirubin was found to decrease the maximal velocities of the reaction without changing the apparent Km value for either the peptide substrate or for ATP. Thus, inhibition by bilirubin of PKA-catalyzed peptide phosphorylation appears to be mediated by a noncompetitive mechanism with respect to the two kinase substrates. Hence, bilirubin appears to be able to interact with domain(s) on catalytic protein kinase subunits distinct from the substrate-binding domains.

Our results partially explain previous, conflicting reports. Bilirubin has been stated to inhibit the PKA-catalyzed phosphorylation of histones, as well as the binding of cAMP to the regulatory subunit of PKA, by both competitive(9) and noncompetitive(11) mechanisms. Further, 50% inhibition of protein kinase C activity was achieved with 45-75 μM bilirubin in both cytosolic and particulate fractions from human skin fibroblasts(24), possibly through an apparently irreversible inhibitory effect of bilirubin on protein kinase C observed in in vitro studies(11). Our results suggest that part of these effects may be due to direct inhibition of protein kinase catalytic domains. This does not exclude additional effects on regulatory domains and substrates.

The extent to which the effects observed in this study are relevant for bilirubin toxicity remains to be determined. Inhibitory effects of bilirubin have been documented with energy metabolism(3, 12, 25–27), enzyme functions(28–35), carbohydrate uptake and metabolism(36–38), membrane functions(39–43), protein synthesis(44, 45), RNA/DNA synthesis and replication(46, 47), and immunologic mechanisms(48–50). Because protein phosphorylation plays a key role in the control and regulation of most, if not all, of these processes, the ability of bilirubin to inhibit phosphorylation reactions may contribute to bilirubin toxicity, both in vivo and in vitro.

The range of bilirubin concentrations, found to cause a 50% inhibition of protein phosphorylation reactions in the present study, is probably within the range found in strongly jaundiced foci in the brains of kernicteric newborns(51), and the effects of bilirubin we report here may therefore be directly involved in the neurotoxic responses seen in these patients. Brain bilirubin concentrations associated with apparently transitory effects such as lethargy and changes in evoked potentials are likely to be lower(52, 53). In the long run, however, cumulative effects of lower bilirubin concentrations could still cause significant protein kinase inhibition, with resulting decreases in a wide range of phosphorylation reactions and consequent interference with the regulation of cellular processes. Moreover, intracellular enrichment of bilirubin in hydrophobic compartments(39, 40, 45, 54–56) could expose specific protein phosphorylation systems to bilirubin concentrations far in excess of the overall level, with resulting toxic responses being determined by both subcellular localization and protein target specificity.

Bilirubin binding to proteins like BSA as well as ligandin appears to be on lysine residue(s)(57–59). The ATP-binding subdomain II of several catalytic protein kinase subunits contains an invariant lysine(23), as does the bilirubin-sensitive enzyme Na/K-ATPase(60). These observations indicate the possibility that binding to lysine represents a central mechanism for mediating the effects of bilirubin on enzymatic reactions.

In conclusion, bilirubin inhibits the phosphorylation of several proteins catalyzed by a number of protein kinases, possibly by a general noncompetitive mechanism. Given the central role of protein phosphorylation in regulation and control of cellular metabolic processes, it is conceivable that some of the toxic effects of bilirubin on the cell may be mediated through such inhibition of protein phosphorylation.