Systematic Assessment of Research on Autism Spectrum Disorder (ASD) and Mercury Reveals Conflicts of Interest and the Need for Transparency in Autism Research

Historically, entities with a vested interest in a product that critics have suggested is harmful have consistently used research to back their claims that the product is safe. Prominent examples are: tobacco, lead, bisphenol A, and atrazine. Research literature indicates that about 80-90% of studies with industry affiliation found no harm from the product, while only about 10-20% of studies without industry affiliation found no harm. In parallel to other historical debates, recent studies examining a possible relationship between mercury (Hg) exposure and autism spectrum disorder (ASD) show a similar dichotomy. Studies sponsored and supported by industry or entities with an apparent conflict of interest have most often shown no evidence of harm or no "consistent" evidence of harm, while studies without such affiliations report positive evidence of a Hg/autism association. The potentially causal relationship between Hg exposure and ASD differs from other toxic products since there is a broad coalition of entities for whom a conflict of interest arises. These include influential governmental public health entities, the pharmaceutical industry, and even the coal burning industry. This review includes a systematic literature search of original studies on the potential relationship between Hg and ASD from 1999 to August 2015, finding that of the studies with public health and/or industry affiliation, 86% reported no relationship between Hg and ASD. However, among studies without public health and/or industry affiliation, only 21% find no relationship between Hg and ASD. The discrepancy in these results suggests a bias indicative of a conflict of interest.

The relationship between mercury and autism: A comprehensive review and discussion

The brain pathology in autism spectrum disorders (ASD) indicates marked and ongoing inflammatory reactivity with concomitant neuronal damage. These findings are suggestive of neuronal insult as a result of external factors, rather than some type of developmental mishap. Various xenobiotics have been suggested as possible causes of this pathology. In a recent review, the top ten environmental compounds suspected of causing autism and learning disabilities were listed and they included: lead, methyl-mercury, polychorinated biphenyls, organophosphate pesticides, organochlorine pesticides, endocrine disruptors, automotive exhaust, polycyclic aromatic hydrocarbons, polybrominated diphenyl ethers, and perfluorinated compounds. This current review, however, will focus specifically on mercury exposure and ASD by conducting a comprehensive literature search of original studies in humans that examine the potential relationship between mercury and ASD, categorizing, summarizing, and discussing the published research that addresses this topic. This review found 91 studies that examine the potential relationship between mercury and ASD from 1999 to February 2016. Of these studies, the vast majority (74%) suggest that mercury is a risk factor for ASD, revealing both direct and indirect effects. The preponderance of the evidence indicates that mercury exposure is causal and/or contributory in ASD.

Thimerosal as discrimination: vaccine disparity in the UN Minamata Convention on mercury

When addressing toxins, one unmistakable parallel exists between biology and politics: developing children and developing nations are those most vulnerable to toxic exposures. This disturbing parallel is the subject of this critical review, which examines the use and distribution of the mercury (Hg)-based compound, thimerosal, in vaccines. Developed in 1927, thimerosal is 49.55% Hg by weight and breaks down in the body into ethyl-Hg chloride, ethyl-Hg hydroxide and sodium thiosalicylate. Since the early 1930s, there has been evidence indicating that thimerosal poses a hazard to the health of human beings and is ineffective as an antimicrobial agent. While children in the developed and predominantly western nations receive doses of mostly no-thimerosal and reduced-thimerosal vaccines, children in the developing nations receive many doses of several unreduced thimerosal-containing vaccines (TCVs). Thus, thimerosal has continued to be a part of the global vaccine supply and its acceptability as a component of vaccine formulations remained unchallenged until 2010, when the United Nations (UN), through the UN Environment Programme, began negotiations to write the global, legally binding Minamata Convention on Hg. During the negotiations, TCVs were dropped from the list of Hg-containing products to be regulated. Consequently, a double standard in vaccine safety, which previously existed due to ignorance and economic reasons, has now been institutionalised as global policy. Ultimately, the Minamata Convention on Hg has sanctioned the inequitable distribution of thimerosal by specifically exempting TCVs from regulation, condoning a two-tier standard of vaccine safety: a predominantly no-thimerosal and reduced-thimerosal standard for developed nations and a predominantly thimerosal-containing one for developing nations. This disparity must now be evaluated urgently as a potential form of institutionalised discrimination.

New science challenges old notion that mercury dental amalgam is safe

Mercury dental amalgam has a long history of ostensibly safe use despite its continuous release of mercury vapor. Two key studies known as the Children's Amalgam Trials are widely cited as evidence of safety. However, four recent reanalyses of one of these trials now suggest harm, particularly to boys with common genetic variants. These and other studies suggest that susceptibility to mercury toxicity differs among individuals based on multiple genes, not all of which have been identified. These studies further suggest that the levels of exposure to mercury vapor from dental amalgams may be unsafe for certain subpopulations. Moreover, a simple comparison of typical exposures versus regulatory safety standards suggests that many people receive unsafe exposures. Chronic mercury toxicity is especially insidious because symptoms are variable and nonspecific, diagnostic tests are often misunderstood, and treatments are speculative at best. Throughout the world, efforts are underway to phase down or eliminate the use of mercury dental amalgam.

Evidence supporting a link between dental amalgams and chronic illness, fatigue, depression, anxiety, and suicide

The purpose of this review is to examine the evidence for a relationship between mercury (Hg) exposure from dental amalgams and certain idiopathic chronic illnesses--chronic fatigue syndrome (CFS), fibromyalgia (FM), depression, anxiety, and suicide. Dental amalgam is a commonly used dental restorative material that contains approximately 50% elemental mercury (Hg0) by weight and releases Hg0 vapor. Studies have shown that chronic Hg exposure from various sources including dental amalgams is associated with numerous health complaints, including fatigue, anxiety, and depression--and these are among the main symptoms that are associated with CFS and FM. In addition, several studies have shown that the removal of amalgams is associated with improvement in these symptoms. Although the issue of amalgam safety is still under debate, the preponderance of evidence suggests that Hg exposure from dental amalgams may cause or contribute to many chronic conditions. Thus, consideration of Hg toxicity may be central to the effective clinical investigation of many chronic illnesses, particularly those involving fatigue and depression.

Thimerosal exposure and the role of sulfation chemistry and thiol availability in autism

Autism spectrum disorder (ASD) is a neurological disorder in which a significant number of the children experience a developmental regression characterized by a loss of previously acquired skills and abilities. Typically reported are losses of verbal, nonverbal, and social abilities. Several recent studies suggest that children diagnosed with an ASD have abnormal sulfation chemistry, limited thiol availability, and decreased glutathione (GSH) reserve capacity, resulting in a compromised oxidation/reduction (redox) and detoxification capacity. Research indicates that the availability of thiols, particularly GSH, can influence the effects of thimerosal (TM) and other mercury (Hg) compounds. TM is an organomercurial compound (49.55% Hg by weight) that has been, and continues to be, used as a preservative in many childhood vaccines, particularly in developing countries. Thiol-modulating mechanisms affecting the cytotoxicity of TM have been identified. Importantly, the emergence of ASD symptoms post-6 months of age temporally follows the administration of many childhood vaccines. The purpose of the present critical review is provide mechanistic insight regarding how limited thiol availability, abnormal sulfation chemistry, and decreased GSH reserve capacity in children with an ASD could make them more susceptible to the toxic effects of TM routinely administered as part of mandated childhood immunization schedules.

Thiol-redox antioxidants protect against lung vascular endothelial cytoskeletal alterations caused by pulmonary fibrosis inducer, bleomycin: comparison between classical thiol-protectant, N-acetyl-L-cysteine, and novel thiol antioxidant, N,N'-bis-2-mercaptoethyl isophthalamide

Lung vascular alterations and pulmonary hypertension associated with oxidative stress have been reported to be involved in idiopathic lung fibrosis (ILF). Therefore, here, we hypothesize that the widely used lung fibrosis inducer, bleomycin, would cause cytoskeletal rearrangement through thiol-redox alterations in the cultured lung vascular endothelial cell (EC) monolayers. We exposed the monolayers of primary bovine pulmonary artery ECs to bleomycin (10 µg) and studied the cytotoxicity, cytoskeletal rearrangements, and the macromolecule (fluorescein isothiocyanate-dextran, 70,000 mol. wt.) paracellular transport in the absence and presence of two thiol-redox protectants, the classic water-soluble N-acetyl-L-cysteine (NAC) and the novel hydrophobic N,N'-bis-2-mercaptoethyl isophthalamide (NBMI). Our results revealed that bleomycin induced cytotoxicity (lactate dehydrogenase leak), morphological alterations (rounding of cells and filipodia formation), and cytoskeletal rearrangement (actin stress fiber formation and alterations of tight junction proteins, ZO-1 and occludin) in a dose-dependent fashion. Furthermore, our study demonstrated the formation of reactive oxygen species, loss of thiols (glutathione, GSH), EC barrier dysfunction (decrease of transendothelial electrical resistance), and enhanced paracellular transport (leak) of macromolecules. The observed bleomycin-induced EC alterations were attenuated by both NAC and NBMI, revealing that the novel hydrophobic thiol-protectant, NBMI, was more effective at µM concentrations as compared to the water-soluble NAC that was effective at mM concentrations in offering protection against the bleomycin-induced EC alterations. Overall, the results of the current study suggested the central role of thiol-redox in vascular EC dysfunction associated with ILF.

Novel lipid-soluble thiol-redox antioxidant and heavy metal chelator, N,N'-bis(2-mercaptoethyl)isophthalamide (NBMI) and phospholipase D-specific inhibitor, 5-fluoro-2-indolyl des-chlorohalopemide (FIPI) attenuate mercury-induced lipid signaling leading to protection against cytotoxicity in aortic endothelial cells

Here, we investigated thiol-redox-mediated phospholipase D (PLD) signaling as a mechanism of mercury cytotoxicity in mouse aortic endothelial cell (MAEC) in vitro model utilizing the novel lipid-soluble thiol-redox antioxidant and heavy metal chelator, N,N'-bis(2-mercaptoethyl)isophthalamide (NBMI) and the novel PLD-specific inhibitor, 5-fluoro-2-indolyl des-chlorohalopemide (FIPI). Our results demonstrated (i) mercury in the form of mercury(II) chloride, methylmercury, and thimerosal induced PLD activation in a dose- and time-dependent manner; (ii) NBMI and FIPI completely attenuated mercury- and oxidant-induced PLD activation; (iii) mercury induced upstream phosphorylation of extracellular-regulated kinase 1/2 (ERK1/2) leading to downstream threonine phosphorylation of PLD(1) which was attenuated by NBMI; (iv) mercury caused loss of intracellular glutathione which was restored by NBMI; and (v) NBMI and FIPI attenuated mercury- and oxidant-induced cytotoxicity in MAECs. For the first time, this study demonstrated that redox-dependent and PLD-mediated bioactive lipid signaling was involved in mercury-induced vascular EC cytotoxicity which was protected by NBMI and FIPI.

Mercury toxicity presenting as chronic fatigue, memory impairment and depression: diagnosis, treatment, susceptibility, and outcomes in a New Zealand general practice setting (1994-2006)

In a group of 465 patients diagnosed as having chronic mercury toxicity (CMT), 32.3% had severe fatigue, 88.8% had memory loss, and 27.5% had depression. A significant correlation was found between CMT and the Apo-lipoprotein E4 genotype (p=0.001). An investigation into an additional 864 consecutively seen general practice patients, resulted in 30.3% having evidence consistent with CMT, and once again a significant correlation was found with the APO-E4 genotype (p=0.001). Removal of amalgam mercury fillings when combined with appropriate treatment resulted in a significant symptom reduction (p<0.001) to levels reported by healthy subjects.

Reduced levels of mercury in first baby haircuts of autistic children

Reported rates of autism have increased sharply in the United States and the United Kingdom. One possible factor underlying these increases is increased exposure to mercury through thimerosal-containing vaccines, but vaccine exposures need to be evaluated in the context of cumulative exposures during gestation and early infancy. Differential rates of postnatal mercury elimination may explain why similar gestational and infant exposures produce variable neurological effects. First baby haircut samples were obtained from 94 children diagnosed with autism using Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM IV) criteria and 45 age- and gender-matched controls. Information on diet, dental amalgam fillings, vaccine history, Rho D immunoglobulin administration, and autism symptom severity was collected through a maternal survey questionnaire and clinical observation. Hair mercury levels in the autistic group were 0.47 ppm versus 3.63 ppm in controls, a significant difference. The mothers in the autistic group had significantly higher levels of mercury exposure through Rho D immunoglobulin injections and amalgam fillings than control mothers. Within the autistic group, hair mercury levels varied significantly across mildly, moderately, and severely autistic children, with mean group levels of 0.79, 0.46, and 0.21 ppm, respectively. Hair mercury levels among controls were significantly correlated with the number of the mothers' amalgam fillings and their fish consumption as well as exposure to mercury through childhood vaccines, correlations that were absent in the autistic group. Hair excretion patterns among autistic infants were significantly reduced relative to control. These data cast doubt on the efficacy of traditional hair analysis as a measure of total mercury exposure in a subset of the population. In light of the biological plausibility of mercury's role in neurodevelopmental disorders, the present study provides further insight into one possible mechanism by which early mercury exposures could increase the risk of autism.

Creatine kinase in human retinal pigment epithelium

Retinal pigment epithelial ion transport activity, and consequent ATP consumption vary significantly as a function of photoreceptor activity. In a variety of cell types, ATP levels are maintained during high-energy usage by phosphocreatine hydrolysis, catalysed by the enzyme creatine kinase. The present work was designed to assess the importance of creatine kinase in retinal pigment epithelial cell metabolism. To this end, activity measurements, non-denaturing gel electrophoresis, Western blot analysis and immunohistochemistry were used to characterize creatine kinase in retinal pigment epithelium. Total creatine kinase activity in the retinal pigment epithelium is approximately 0.05 micromol ATP mg protein(-1) min(-1). The bulk of this activity was mediated by the B-CK isoform. However, by immunoblotting, non-denaturing gel electrophoresis and immunohistochemistry, the presence of the M-CK isoform of creatine kinase was also detected. The M-CK isoform was plasma membrane associated and predominately localized to the apical surface. Creatine kinase in the retinal pigment epithelium could function in a spatial energy shuttle that helps to sustain apical plasma membrane ion transport activity.

Orientation of GTP and ADP within their respective binding sites in glutamate dehydrogenase

Previous studies have identified the guanine and adenine binding domains of the GTP and ADP binding sites of GDH. In this study the peptide sequences within or near to the terminal phosphate-binding domains of the GTP and ADP binding sites of bovine liver glutamate dehydrogenase (GDH) were identified using photoaffinity labeling with the benzophenone nucleotide derivatives, [gamma-32P]GTPgammaBP and [gamma-32P]ATPgammaBP. Without activating light, GTPgammaBP exhibited inhibiting effects on the GDH reaction similar to GTP; ATPgammaBP, as expected, produced activating effects similar to those of ADP. Photoinsertion into GDH by both probes exhibited saturation effects in agreement with the respective kinetic effects. Specificity of labeling was supported by specific and effective reduction of photoinsertion of [gamma-32P]GTPgammaBP and [gamma-32P]ATPgammaBP into GDH by GTP and ADP, respectively. Using a combination of immobilized Fe3+-chelate affinity chromatography and reversed-phase HPLC, photolabeled peptides located within or near the phosphate-binding domains of the GTP and ADP sites were isolated. Sequence analysis showed that GTPgammaBP primarily modified a peptide near the middle of the GDH sequence, Asn135-Lys143 and Glu290-Lys295. However, ATPgammaBP modified a single peptide corresponding to the sequence Met411-Arg419 near the C-terminal domain. Using these results and the data from the previously identified base-binding domain peptides the orientation of GTP and ADP within their respective binding sites in the catalytic cleft of GDH is proposed and explained on the basis of a proposed three-dimensional schematic model structure derived from the bacterial enzyme.

ATP nucleotidylation of creatine kinase

Creatine kinase (CK) will autoincorporate radiolabel from [gamma32P]ATP and has thus been reported to be autophosphorylated. Also, in contrast to normal brain enzyme, CK in Alzheimer-diseased brain homogenate shows greatly decreased activity, abolished photolabeling with [32P]8N3ATP, and no detectable autoincorporation of radiolabel by [gamma32P]ATP. Surprisingly, our studies with both human brain and purified CK showed that [alpha32P]ATP, [gamma32P]ATP, [alpha32P]ADP, [2,8H3]ATP, [gamma32P]2',3'-O-(2,4, 6-trinitrophenyl)-ATP, and [gamma32P]benzophenone-gammaATP all autoincorporate radiolabel into CK with good efficiency. This demonstrates that the gamma-phosphate and the 2' and 3' hydroxyls are not involved in the covalent linkage and that all three phosphates, the ribose and base of the ATP molecule are retained upon autoincorporation (nucleotidylation). Treatment with NaIO3 to break the 2'-3' linkage effected total loss of radiolabel indicating that nucleotidylation resulted in opening of the ribose ring at the C1' position. Nucleotidylation with increasing [alpha32P]ATP at 37 degrees C gives an approximate k0.5 of 125 microM and saturates at 340 microM nucleotide. Modification of 8-10% of the copy numbers occurs at saturation, and CK activity is inhibited to approximately the same degree. Low micromolar levels of native substrates such as ADP, ATP, and phosphocreatine substantially reduce [alpha32P]ATP nucleotidylation. In contrast, AMP, GTP, GMP, NADH, and creatine did not effectively reduce nucleotidylation. When [alpha32P]ATP-nucleotidylated or [alpha32P]8N3ATP-photolabeled CK is treated with trypsin a single, identical radiolabeled peptide (V279-R291) is generated that comigrates on reverse phase HPLC and Tris-tricine electrophoresis. Nucleotidylation into this peptide was prevented 86% by the presence of ATP. We conclude that CK is nucleotidylated within the active site by modification at the C1'position and that autophosphorylation of this enzyme does not occur.

Abnormal properties of creatine kinase in Alzheimer's disease brain: correlation of reduced enzyme activity and active site photolabeling with aberrant cytosol-membrane partitioning

The report shows that Alzheimer's disease (AD) brain creatine kinase (CK) is modified such that the nucleotide binding site of CK is blocked and that abnormal partitioning of CK between the soluble and pellet fractions occurs. First, CK activity was 86% decreased in AD brain homogenates in comparison to age-matched controls. Secondly, over a 23.5 fold greater 32P photoincorporation of [alpha 32P]8N3ATP was observed into CK of control vs. AD samples. Also, a 7.4-fold increase of enzyme induced 32P incorporation was observed in controls vs. AD samples by incubation with [gamma 32P]ATP. Thirdly, Western blot analysis showed that CK copy numbers in the AD homogenate were decreased by less than 14% in comparison to controls. However, analysis showed that control supernatant and pellet fractions contained 10.3 and 0.4 times the CK copy number found in the corresponding AD fractions. 32P incorporation by both photolabeling and enzyme catalyzed incorporation of radiolabel followed CK activity and not CK copy number. Further, [alpha 32P]ADP and [gamma 32P]ATP incorporated 32P into control brain and purified brain CK equally well, indicating that a mechanism different from gamma-phosphoryl transfer is involved in the enzymatic incorporation of radiolabel. Also, the level of abnormal partitioning of CK into AD brain pellet correlated with the decreased [32P]8N3GTP photolabeling and abnormal partitioning of beta-tubulin, a protein known to be aberrantly modified in the AD brain. This indicates that a common chemistry is affecting both CK and tubulin in AD.

Mercury vapor inhalation inhibits binding of GTP to tubulin in rat brain: similarity to a molecular lesion in Alzheimer diseased brain

Hg2+ interacts with brain tubulin and disassembles microtubules that maintain neurite structure. Since it is well known that Hg vapor (Hg0) is continuously released from "silver" amalgam tooth fillings and is absorbed into brain, rats were exposed to Hg0 4h/day for 0, 2, 7, 14 and 28 d at 250 or 300 micrograms Hg/m3 air, concentrations present in mouth air of some humans with many amalgam fillings. Average rat brain Hg concentrations increased significantly (11-47 fold) with duration of Hg0 exposure. By 14 d Hg0 exposure, photoaffinity labelling on the beta-subunit of the tubulin dimer with [alpha 32P] 8N3 GTP in brain homogenates was decreased 41-74%, upon analysis of SDS-PAGE autoradiograms. The identical neurochemical lesion of similar or greater magnitude is evident in Alzheimer brain homogenates from approximately 80% of patients, when compared to human age-matched neurological controls. Total tubulin protein levels remained relatively unchanged between Hg0 exposed rat brains and controls, and between Alzheimer brains and controls. Since the rate of tubulin polymerization is dependent upon binding of GTP to tubulin dimers, we conclude that chronic inhalation of low-level Hg0 can inhibit polymerization of brain tubulin essential for formation of microtubules.

Identification of an adenine-nucleotide-binding site on interferon alpha2

Using 32P-labeled 2-azidoadenosine 5'-triphosphate (2N3ATP) and 8-azidoadenosine 5'-triphosphate (8N3ATP), we have identified a site on human interferon alpha2 (IFN-alpha2) that binds adenine nucleotides. The results from saturation and competition experiments demonstrated the specificity of the nucleotide interaction. Half-maximal saturation of IFN-alpha2 was observed at 10 microM 2N3ATP or 35 microM 8N3ATP. ATP effectively decreased photoinsertion of both photoaffinity analogs of ATP. Photoinsertion of 8N3ATP was enhanced by MgCl2, independent of the ionic strength, and exhibited an optimum pH between 7.0 and 7.5. Immobilized-Al3+ affinity chromatography and HPLC were used to purify the modified peptides from IFN-alpha2 that had been photolabeled with 8N3ATP and digested with trypsin or chymotrypsin. Overlapping-sequence analysis localized the sites of photoinsertion to the region corresponding to Lys121-Tyr135 in the amino acid sequence of IFN-alpha2, which almost perfectly overlaps a nuclear-localization signal (R120KYFQRITLYLKEKKY135).

A comparison of changes in nucleotide-protein interactions in the striatal, hippocampus and paramedian cortex after cerebral ischemia and reperfusion: correlations to regional vulnerability

[32P]Azido-purine analogs of ATP and GTP were used to detect changes in phosphorylation and nucleotide binding induced by ischemia and subsequent reperfusion in rat brain striatum, hippocampus and paramedian cortex (PM cortex) tissues. Major changes in phosphorylation were observed for a 130-kDa protein, tentatively identified as the Ca2+ transport ATPase, and calcium/calmodulin-dependent protein kinase II (CaM Kinase II) in all tissues. However, recovery of the phosphorylation of the 130-kDa protein occurred only in the PM cortex on reperfusion. A 200-300% increase in [32P]8N3ATP photoinsertions was observed in the striatum and hippocampus regions for a 43-kDa protein with an isoelectric point of 6.8. This protein was identified as glutamine synthetase (GS) and the increase in binding was found to be due to both increased copy number and activation by Mn2+. An increase in [32P]8N3GTP photoinsertion into a 55-kDa protein, identified as the beta-subunit of tubulin, was found only in the striatum and hippocampus. This indicates the depolymerization of microtubulin in these tissues. These changes correlate to the vulnerability of the striatum and hippocampus to ischemia-induced neuronal death.

Oxidatively induced structural alteration of glutamine synthetase assessed by analysis of spin label incorporation kinetics: relevance to Alzheimer's disease

The activity of the astrocytic enzyme glutamine synthetase (GS) is decreased in the Alzheimer's disease brain, which may have relevance to mechanisms of chronic excitotoxicity. The molecular perturbation(s) that results in GS inactivation is not known, although oxidative lesioning of the enzyme is one likely cause. To assess structural perturbation induced in GS by metal-catalyzed oxidation, a series of spin-labeling studies were undertaken. Ovine GS was oxidized by exposure to iron/hydrogen peroxide and subsequently labeled with the thiol-specific nitroxide probe MTS [(1-oxyl-2,2,5,5-tetramethyl-pyrroline-3-methyl)methanethiosulfonate]. The reaction of MTS with cysteine residues within GS was monitored in real time by electron paramagnetic resonance spectrometry. Structural perturbation of GS, manifested as decreased thiol accessibility, was inferred from an apparent decrease in the rate constant for the second-order reaction of MTS with protein thiols. A subsequent spin-labeling study was undertaken to compare the structural integrity of GS purified and isolated from Alzheimer's disease-afflicted brain (AD-GS) with that of GS isolated from nondemented, age-matched control brain (C-GS). The rate constant for reaction of MTS with AD-GS was markedly decreased relative to that for the reaction of spin label with C-GS. The kinetic data were partially corroborated by spectroscopic data obtained from circular dichroism analysis of control and peroxide-treated ovine GS. In an adjunct experiment, the interaction of GS with a synthetic analogue of the Alzheimer's-associated beta-amyloid peptide, known to induce free radical oxidative stress, indicated strong interaction of the enzyme with the peptide as reflected by a decrease in the rate constant for MTS binding to reactive protein thiols.

Identification of adenine binding domain peptides of the NADP+ active site within porcine heart NADP(+)-dependent isocitrate dehydrogenase

Photoaffinity labeling with [2'-32P]2N3NADP+ and [32P]2N3NAD+ was used to identify two overlapping tryptic and chymotryptic generated peptides within the adenine binding domain of NADP(+)-dependent isocitrate dehydrogenase (IDH). Photolysis was required for insertion of radiolabel, and prior photolysis of photoprobes before addition of IDH prevented insertion. Photoincorportion of 2N3NAD+ inhibited the enzymatic activity of IDH. Photolabeling of IDH with both [32P]2N3NAD+ and [2'-32P]2N3-NADP+ showed saturation effects with apparent Kds of 20 and 14 microM (+/-12%), respectively. The efficiency of photoincorporation at saturation of binding sites was determined to be about 50%. Also, photolabeling was observed with [32P]8N3ATP and [32P]2N3ATP but with saturation effects observed at lower affinity. With all radiolabeled probes reduction of photoinsertion was effected best by the addition of NADP+ followed by NAD+ and then ATP, indicating that photoinsertion with all the probes was within the NADP+ binding site. Isolation of [32P]2N3NAD+ and [2'-32P]2N3NADP+ photolabeled peptides by use of immobilized boronate and immobilized Al3+ chromatography, respectively, followed by HPLC purification resulted in the identification of overlapping peptides corresponding to Ile244-Arg249 and Leu121-Arg133 (tryptic fragments) and Lys243-His248 and Leu121-His135 (chymotryptic fragments). Trp125 and Trp245 were identified as the sites of photoinsertion based on these residues not being detectable on sequencing, the lack of chymotryptic cleavage at these residues, and the decreased rate of trypsin digestion at nearby Lys243 and Lys127. Sequence analysis of [32P]8N3ATP and [32P]2N3ATP photolabeled peptides gave essentially the same peptide regions being photolabeled but at much lower efficiency, indicating that the effects of ATP on IDH activity are dependent on competition for the same site.

Novel unconventional binding site in the variable region of immunoglobulins

The variable immunoglobulin (Ig) domains contain hypervariable regions that are involved in the formation of the antigen binding site. Besides the canonical antigen binding site, so-called unconventional sites also reside in the variable region that bind bacterial and viral proteins. Docking to these unconventional sites does not typically interfere with antigen binding, which suggests that these sites may be a part of the biological functions of Igs. Herein, a novel unconventional binding site is described. The site is detected with 8-azidopurine nucleotide photoaffinity probes that label antibodies efficiently and under mild conditions. Tryptic peptides were isolated from photolabeled monoclonal antibodies and aligned with the variable antibody domains of heavy and light chains. The structure of a variable Ig fragment was used to model the binding of the purine nucleotide to invariant residues in a hydrophobic pocket of the Ig molecule at a location distant from the antigen binding site. Monoclonal and polyclonal antibodies were biotinylated with the photoaffinity linker and used in fluorescence-activated cell sorter and ELISA analyses. The data support the utility of this site for tethering diagnostic and therapeutic agents to the variable Ig fragment region without impairing the structural and functional integrity of antibodies.

Identification of adenine binding domain peptides of the ADP regulatory site within glutamate dehydrogenase

Photoaffinity labeling with [alpha-32P]-8-azidoadenosine 5'-diphosphate (8N3ADP) and [beta-32P]-2-azidoadenosine 5'-diphosphate (2N3ADP) was used to identify overlapping tryptic and chymotryptic generated peptides within the adenine binding domain of the regulatory ADP site of bovine liver glutamate dehydrogenase (GDH). In the absence of UV irradiation, 8N3ADP was able to activate the reverse reaction catalyzed by GDH as well as ADP. Photoinsertion of both [alpha 32P] 8N3ADP and [beta 32P]2N3ADP was reduced best by ADP in comparison to other nucleotides. Photolabeling of GDH with [alpha 32P]8N3-ADP appeared to be biphasic, with saturation occurring near 80 and 130 microM, whereas [beta 32P]2N3ADP showed saturation near 50 microM. When 60 microM [alpha 32P]8N3ADP (below the first saturation value) was used to identify peptides within the ADP binding domain, peptides corresponding to residues G156-K200 and E175-K200 (tryptic) and I158-Y183 (chymotryptic) were photolabeled. However, when 160 microM [alpha 32P]8N3ADP (above the second saturation value) was used, the peptide D403-R418 was also photolabeled. Digestion with both trypsin and chymotrypsin resulted in isolation of peptides E175-Y183 and A184-I192. [beta 32P]2N3ADP at 90 microM also photolabeled tryptic peptides G156-K200 and C270-K289. C270-K289 was shown earlier to be within the NAD+ binding site [Kim, H., and Haley, B. (1991) Bioconjugate Chem. 2, 142-147]. These results are consistent with the residues E175-[192 being within the adenine binding domain of the ADP regulatory site.

Identification of the NADP(H) binding site of rat liver microsomal 5 alpha-reductase (isozyme-1): purification of a photolabeled peptide corresponding to the adenine binding domain

We have previously shown that [2'-32P]-2-azido-NADP+ is an effective probe of the NADP-(H) binding site of rat liver microsomal 5 alpha-reductase (5 alpha R-1) [Bhattacharyya et al. (1994) Steroids 59, 634-641]. PEG-fractionated (6.5%) detergent-solubilized preparations (40 mg) containing 5 alpha R-1 activity were UV-photolyzed with [32P]-2-azido-NADP+ and subjected to preparative gel electrophoresis on 8% SDS-PAGE. Fractions corresponding to the second major [32P]-labeled peak following the dye-front were analyzed by 10% SDS-PAGE and showed a single [32P]-labeled species with an apparent molecular mass of approximately 26 kDa (5 alpha R-1). TCA precipitation (13.6%) of the labeled fractions resulted in recovery of > 70% of the total radioactivity in the protein pellet. Trypsin digestion of the resuspended pellet followed by immobilized-Al3+ affinity chromatography indicated that > 90% of the radioactivity remained bound to the affinity column. The [32P]-2N3-NADP(+)-labeled peptide was eluted with potassium phosphate, concentrated, and further purified by reverse-phase (C8) HPLC. Sequence analysis of the purified peptide indicated that it consisted of 11 amino acids with the sequence N-L-R-K-P-G-E-T-G-Y-K, corresponding to residues 170-180 of the rat 5 alpha R-1 sequence [Andersson et al. (1989) J. Biol. Chem. 264, 16249-16255].

Identification of N-terminus peptide of human granulocyte/macrophage colony stimulating factor as the site of nucleotide interaction

The interaction of nucleotides with recombinant human granulocyte/macrophage colony stimulating factor (rhGM-CSF) has been investigated. Utilizing nucleotide photoaffinity probes [gamma 32P]-8N3ATP and [beta 32P]-8N3Ap4A, an analog of alarmone, the specificity of interaction was demonstrated by saturation of photoinsertion by these analogs and protection of photoinsertion by these analogs in the presence of natural nucleotide. The site of photoinsertion was tentatively determined to be Ser9. The photolabeled cytokine has lost most of its biological activity in a cellular proliferation assay, indicating a possible physiological role for this interaction.

Photoaffinity labeling of the ATP binding domain of Rubisco activase and a separate domain involved in the activation of ribulose-1,5-bisphosphate carboxylase/oxygenase

Photoaffinity labeling of Rubisco activase with 2- and 8-N3ATP was used to identify the adenine binding domain for ATP. Rubisco activase hydrolyzed both of these analogs of ATP and used their hydrolysis to support a low rate of Rubisco activation. When irradiated with ultraviolet light, these and other azido-substituted adenine nucleotides covalently modified Rubisco activase at two distinct binding sites. Competition binding experiments with ATP and ADP showed that one of the sites was the ATP binding domain. The other site was not a nucleotide binding domain per se but would bind adenine nucleotides if an azido moiety was present on the base. Tryptophan and other indoles prevented azidoadenine nucleotides from labeling this domain but afforded little protection to the ATP binding domain. The ability to selectively protect each of the two binding sites made it possible to localize the adenine binding domain for ATP to the region of Rubisco activase from N68-D74 and the other binding domain to a region near the N-terminus from Q10 to D14. Modification of the region from Q10 to D14 by photoaffinity labeling prevented Rubisco activase from promoting activation of Rubisco without affecting ATP hydrolysis. These data suggest that a specific region of Rubisco activase near the N-terminus may be a site of interaction with Rubisco. Binding of azidoadenine nucleotides in this region appears to be fortuitous and may involve base-stacking with the species-invariant Trp at position 16 and hydrogen bonding of the azido moiety.

Photoaffinity labeling of rat liver microsomal steroid 5 alpha-reductase by 2-azido-NADP+

Preincubation of female rat liver microsomal preparations with [2'-32P]2N3-NADP+ followed by photolysis with UV light (254 nm) and analysis by SDS-PAGE/autoradiography showed incorporation of 32P into at least 3 major protein bands in the molecular weight range of 14-97 Kd. Labeling of a 26 kD band, the apparent molecular weight of 5 alpha-reductase in liver microsomes, was accompanied by a loss of enzyme activity, consistent with its covalent modification. The inclusion of 20-fold excess NADP+ (100 microM) completely inhibited the incorporation of [2'-32P]2N3-NADP+ and preserved the enzyme activity, whereas excess NAD+ (100 microM) failed to protect 5 alpha-reductase (5 alpha R) activity. Similar results were obtained with the detergent-solubilized form of 5 alpha R. Polyethylene glycol (PEG) fractionation of detergent-solubilized preparations of 5 alpha R showed that all the 5 alpha R activity could be recovered in the 6.5% pellet with a 3-4-fold increase in the specific activity. photolysis of this fraction with [2'-32P]2N3-NADP+ resulted in approximately 2-fold increase in 32P labeling of the 5 alpha R band. Increasing photolysis time and concentration of the [2'-32P]2N3-NADP+ indicated that the half-life for photoincorporation and the apparent Kd were 1.0 min and 2 microM, respectively. These results suggest that 2N3-NADP+ is an effective probe of the NADP(H) binding site of 5 alpha R, and is a useful marker during purification of the enzyme.

Photoaffinity labeling of creatine kinase with 2-azido- and 8-azidoadenosine triphosphate: identification of two peptides from the ATP-binding domain

Two different analogs of ATP, [gamma-32P]2N3ATP and and [gamma-32P]8N3ATP, were used to photoaffinity label the MM and BB isoforms of rabbit cytosolic creatine kinase. Evidence that photoinsertion was within the ATP-binding domain was as follows: (1) Assays for creatine phosphate production demonstrated that [gamma-32]2N3ATP and [gamma-32P]8N3ATP are substrates for creatine kinase. (2) Enzymatic activity was inhibited by photolabeling with either analog. (3) Saturation of photoinsertion was observed for both analogs. Half-maximal saturation was observed at 5 microM [gamma-32P]2N3ATP or 12 microM (gamma-32P]8N3ATP. (4) Photoinsertion of both probes could be decreased by micromolar levels of ATP. Immobilized Al3+ affinity chromatography and HPLC were used to isolate the peptides modified by these probes. Overlapping sequence analysis of the isolated peptides from the tryptic and chymotryptic digests of the photolabeled MM isoform revealed that [gamma-32P]8N3ATP photoinserted into the peptide region corresponding to Val279-Arg291, whereas [gamma-32P]2N3-ATP photoinserted into Val236-Lys241. The corresponding peptide (Ile279-Arg291 and Val236-Lys241) from the BB isoform were shown to be selectively modified. We conclude that amino acid residues within the peptide regions 236-241 and 279-291 of rabbit cytosolic creatine kinase are localized within the binding domain for the adenine moiety of ATP. The results also demonstrate the effectiveness and selectivity of Al3+ as the chelating agent in immobilized metal affinity chromatography for the isolation of photolabeled peptides as well as its potential to enhance retention of radiolabel during HPLC.

Interaction of nucleotides with acidic fibroblast growth factor (FGF-1)

A wide variety of nucleotides are shown to bind to acidic fibroblast growth factor (aFGF) as demonstrated by their ability to (1) inhibit the heat-induced aggregation of the protein, (2) enhance the thermal stability of aFGF as monitored by both intrinsic fluorescence and CD, (3) interact with fluorescent nucleotides and displace a bound polysulfated naphthylurea compound, suramin, (4) reduce the size of heparin-aFGF complexes, and (5) protect a reactive aFGF thiol group. The binding of mononucleotides, diadenosine compounds (ApnA), and inorganic polyphosphates to aFGF is enhanced as the degree of phosphorylation of these anions is increased with the presence of the base reducing the apparent binding affinity. The nature of the base appears to have much less effect. Photoactivatable nucleotides (8N3-ATP, 2N3-ATP, 8N3-GTP, and 8N3-Ap4A) were employed to covalently label the aFGF nucleotide binding site. In general, Kd's in the low micromolar range are observed. Protection against 90% displacement is observed at several hundred micromolar nucleotide concentration. Using 8N3-ATP as a prototypic reagent, photolabeled aFGF was proteolyzed with trypsin and chymotrypsin and labeled peptides were isolated and sequenced resulting in the identification of 10 possible labeled amino acids (Y8, G20, H21, T61, K112, K113, S116, R119, R122, H124). On the basis of the crystal structure of bovine aFGF, eight of the prospective labeled sites appear to be dispersed around the perimeter of the growth factor's presumptive polyanion binding site. On residue (T61) is more distally located but still proximate to several positively charged residues, and another (Y8) is not locatable in crystal structures. Using heparin affinity chromatography, at least three distinct photolabeled aFGF species were resolved. These labeled complexes display diminished affinity for heparin and a reduced ability to stimulate mitogenesis even in the presence of polyanions such as heparin. In conclusion, nucleotides bind apparently nonspecifically to the polyanion binding site of aFGF but nevertheless are capable of modulating the protein's activity. Evidence for the presence of a second or more extended polyanion binding site and the potential biological significance of these results in terms of potential natural ligands of aFGF are also discussed but not resolved.

Identification of peptides within the base binding domains of the GTP- and ATP-specific binding sites of tubulin

Using gamma-32P-labeled 8-azidopurine nucleotide photoaffinity probes of GTP and ATP, the respective purine ring binding domain peptides of tubulin have been identified. First, the location of the GTP-specific binding site was shown to be on the beta-subunit, whereas the major ATP-specific binding site was on the alpha-subunit. Using a combination of anion-exclusion and immobilized Al3+ column chromatography, the respective photolabeled tryptic peptides of both nucleotide binding sites were isolated, further purified by reverse phase high performance liquid chromatography (HPLC) and sequenced. Chymotryptic peptides were also generated for the GTP binding site. High retention of the photoinserted radiolabel was observed with many of the peptides on reverse phase HPLC at low flow rates. The stability of the photoinserted radiolabel to HPLC varied with different peptides. However, certain peptides were easily distinguished as being within the base binding domains of the GTP and ATP binding sites of tubulin. Two beta-tubulin peptides containing the majority of photoinserted [gamma-32P]8-azidoguanosine 5'-triphosphate corresponded to N-terminal beta-tubulin amino acid residues 3EIVHIQAGQCGNQIGAK19 and 20FWEVISDEHGIDPTGS35. The peptide containing the majority of photoinserted [gamma-32P]8-azidoadenosine 5'-triphosphate corresponded to the C-terminal alpha-tubulin sequence 431DYEEVGVDSVEGEGEEEGEE450.

HgEDTA complex inhibits GTP interactions with the E-site of brain beta-tubulin

We have found that EDTA and EGTA complexes of Hg2+, which conventional wisdom has assumed are biologically inert, are potentially injurious to the neuronal cytoskeleton. Tubulin, a major protein component of the neuronal cytoskeleton, is the target of multiple toxicants, including many heavy metal ions. Among the mercurials, inorganic mercuric ion (Hg2+) is one of the most potent inhibitors of microtubule polymerization both in vivo and in vitro. In contrast to other heavy metals, the capacity of Hg2+ to inhibit microtubule polymerization or disrupt formed microtubules cannot be prevented by the addition of EDTA and EGTA, both of which bind Hg2+ with very high affinity. To the contrary, the addition of these two chelating agents potentiates Hg2+ inhibition of tubulin polymerization. Results herein show that HgEDTA and HgEGTA inhibit tubulin polymerization by disrupting the interaction of GTP with the E-site of brain beta-tubulin, an obligatory step in the polymerization of tubulin. Both HgEDTA and HgEGTA, but not free Hg2+, prevented binding of [32P]8N3GTP, a photoaffinity nucleotide analog of GTP, to the E-site and displaced bound [32P]8N3GTP at low micromolar concentrations. This complete inhibition of photoinsertion into the E-site occurred in a concentration- and time-dependent fashion and was specific for Hg2+ complexes of EDTA and EGTA, among the chelating agents tested. Given the ubiquity of Hg2+ in the environment and the widespread use of EDTA in foodstuffs and medicine, these mercury complexes may pose a potentially serious threat to human health and play a role in diseases of the neuronal cytoskeleton.

Photoaffinity labeling of human placental NAD(+)-linked 15-hydroxyprostaglandin dehydrogenase with [alpha-32P]2N3NAD+. Identification of a peptide in the adenine ring binding domain

Oxidation of many prostaglandins at C-15 results in the formation of 15-keto metabolites, which have reduced biological activity. This reaction is catalyzed by NAD(+)-dependent 15-hydroxyprostaglandin dehydrogenase. Using the photoaffinity analog of NAD+, [alpha-32P]nicotinamide-2-azidoadenine dinucleotide, we have identified a peptide in the adenine ring binding domain of the NAD+ binding site of 15-hydroxyprostaglandin dehydrogenase. The specificity of photolabeling was demonstrated by saturation and protection experiments. Saturation of photolabeling was observed at approximately 45-50 microM with an apparent Kd of 8-10 microM. Approximately 90% of photolabeling could be protected by 200 microM NAD+ when the protein was photolyzed in the presence of 10 microM probe. The photolabeled protein was digested with Staphylococcus aureus V8 or chymotrypsin, and the photolabeled peptides were purified by either boronate affinity chromatography or Fe+3 chelate chromatography followed by reverse phase HPLC. The photolabeled peptide region was identified to be Val32-Glu40.

Synthesis and application of bidentate photoaffinity cross-linking reagents. Nucleotide photoaffinity probes with two photoactive groups

Two "targeted bidentate" photoaffinity cross-linking reagents, the monoanhydride of 8-N3ADP with N-(4-(benzoyl)phenylmethyl)phosphoramide ([gamma-32P]8-N3ATP gamma BP) and the monoanhydride of 8-N3GDP with N-(4-(benzoyl)phenylmethyl)-phosphoramide ([gamma-32P]8-N3GTP gamma BP), were developed for studying the inter- and intramolecular interactions of nucleotide-binding proteins. Experiments using these bidentate reagents with two photoactive groups led to specific cross-linking: [gamma-32P]8-N3GTP gamma BP and [gamma-32P]8-N3ATP gamma BP showed intersubunit cross-linking of glutamate dehydrogenase and [gamma-32P]8-N3GTP gamma BP appeared to cross-link the alpha- and beta-subunits of tubulin. The non-azido "monodentate" versions of these reagents, the monoanhydride of ADP with N-(4-(benzoyl)phenylmethyl)-phosphoramide ([gamma-32P]ATP gamma BP) and the monoanhydride of GDP with N-(4-(benzoyl)phenylmethyl)-phosphoramide ([gamma-32P]GTP gamma BP), were also synthesized and characterized. The ability of these monodentate reagents with one photoactive group to serve as photoaffinity probes was established by photolabeling specifically the exchangeable GTP-binding domain of tubulin with [gamma-32P]GTP gamma BP and the ATP-binding domain of purified adenylate kinase and several nucleotide-binding proteins in human brain homogenate with [gamma-32P]ATP gamma BP.

Photoaffinity labeling of ribulose-1,5-bisphosphate carboxylase/oxygenase activase with ATP gamma-benzophenone. Identification of the ATP gamma-phosphate binding domain

The phosphate-binding domain of the ATP-binding site of tobacco Rubisco (ribulose-1,5-bisphosphate carboxylase/oxygenase) activase was elucidated by photo-affinity labeling with a monoanhydride of ADP with N-(4-(benzoyl)phenylmethyl)phosphoramide ([gamma-32P]ATP gamma BP). Covalent incorporation of [gamma-32P]ATP gamma BP into the 42-kDa Rubisco activase subunit was dependent upon irradiation with ultraviolet light. Photolabelling of Rubisco activase with ATP gamma BP exhibited saturation kinetics; the apparent Kd for photolabeling was 5 microM. Two lines of evidence showed that ATP gamma BP modified Rubisco activase at the ATP-binding domain. First, physiological concentrations of ATP and ADP afforded complete protection against photolabeling of Rubisco activase by ATP gamma BP. Second, photolysis of Rubisco activase in the presence of ATP gamma BP decreased both the ATPase and the Rubisco activating activities. Inactivation of enzyme activity was dependent on ATP gamma BP concentration and could be prevented by including ADP during photolabeling. The region of Rubisco activase that was modified by ATP gamma BP was identified by isolating photolabeled peptides. Sequence analysis showed that ATP gamma BP modified Rubisco activase in two distinct regions; one region, S117-A136, is adjacent to the P-loop and the other region, V223-T234, exhibits homology to a region of adenylate kinase that ligates the essential metal ion. Photolabeling of these two regions of Rubisco activase was consistent with modification of the ATP gamma-phosphate-binding domain of Rubisco activase with ATP gamma BP.

Identification and characterization of a nucleotide binding site of ovine prolactin with 2-azido-NAD

Photoaffinity labeling of ovine prolactin with the NAD+ photoaffinity analog [alpha-32P]nicotinamide-2-azidoadenine dinucleotide has been used to identify an NADH/NADPH binding site. Specificity of nucleotide interaction was demonstrated by saturation and protection of labeling at physiologically relevant concentrations. Saturation of photoinsertion was observed at approximately 100 microM probe with an apparent Kd of approximately 25 microM. Protection of photoinsertion was observed with NAD+ and NADH. The photoinsertion was decreased by 75% and greater than 95%, respectively, upon addition of 200 microM of the above-mentioned compounds. The protection obtained with NADP+ and NADPH was of the same order, respectively. The adenine ring binding domain of NADH/NADPH binding site was identified by trypsin and chymotrypsin digestion of the photolabeled prolactin and purification of the photolabeled peptide by boronate affinity chromatography and immobilized Fe3+ affinity chromatography. The peptide was identified to be Ala22-Tyr28. These studies demonstrate that prolactin contains an NADH/NADPH binding site which may be significant in the mechanism of action of this hormone.

Ischemic brain injury selectively alters ATP binding of calcium and calmodulin-dependent protein kinase II

Calmodulin and ATP affinity and total binding capacity were characterized for CaM kinase II isolated from control and ischemic animals. Ischemic CaM kinase II exhibited equivalent apparent affinity and total binding for calmodulin when compared to control enzyme. However, ischemic CaM kinase II exhibited a significant decrease in apparent affinity for ATP in saturation experiments. ATP binding was characterized using the ATP photoaffinity analog [alpha-32P] Azido-ATP. A significant decrease in total binding and binding affinity for ATP was observed for the alpha (50 kDa) and beta (60 kDa) subunits. The observation that ischemia induced an alteration of ATP binding without affecting calmodulin binding is consistent with the hypothesis that ischemia directly affects the ATP binding of CaM kinase II which results in subsequent inhibition of the enzyme.

Identification of a guanine binding domain peptide of the GTP binding site of glutamate dehydrogenase: isolation with metal-chelate affinity chromatography

Photoaffinity labeling with [alpha-32P]8N3GTP and [gamma-32P]8N3GTP was used to identify the guanine binding domain of the GTP regulatory site within glutamate dehydrogenase (GDH). Without photolysis, 8N3GTP mimicked the regulatory properties of GTP on GDH activity with 8N3GTP exhibiting a Ki of 5 microM while the Ki for GTP was about 0.6 microM. Under optimal photolabeling conditions saturation of photoinsertion with 1 microgram of GDH revealed an apparent Kd of 9 +/- 4 microM for [gamma-32P]8N3GTP. Photolabeling with this analog could be competitively inhibited with GTP with an apparent Kd of 12 +/- 2 microM. Other nucleotides such as ATP and NAD(P)H could not reduce the amount of photoinsertion as effectively as GTP. ADP could decrease photoinsertion, but only at much higher concentrations. NAD(P)+, GDP, AMP, and GMP had little effect on photoinsertion. Divalent cations Mg2+ and Ca2+ also reduced photoinsertion significantly while the monovalent K+ and Na+ ions had no effect. Aluminum(III)-chelate or iron(III)-chelate affinity chromatography and reversed-phase HPLC were used to purify photolabel-containing peptides generated with either trypsin or chymotrypsin. This identified a portion of the guanine binding domain within the GTP regulatory site as the region containing the sequence Ile439 to Tyr454. Photolabeling of this peptide was prevented 91% by the presence of 300 microM GTP during photolysis. Lys445 was not identified in sequence analyses of the photolabeled peptides. Also, trypsin was unable to cleave the photolabeled peptide at this site. These results suggest that Lys445 may be the residue modified by [alpha-32P]8N3GTP.

Role of calcium in inactivation of calcium/calmodulin dependent protein kinase II after cerebral ischemia

Transient cerebral ischemia demonstrates an increase in activated oxygen species in the brain that could lead to eventual neuronal cell death. Neuronal cells respond to oxygen free radicals through the restructuring of the cytoskeleton and membranes, mobilization of calcium and gene expression which play a role in cell injury. Ten min of bilateral carotid artery occlusion resulted in a decrease in calcium/calmodulin dependent protein kinase II (CaM kinase II) phosphorylation and activity detected in the brain immediately following ischemia and was partially restored within 24 h of reperfusion. Pretreatment of animals with an anesthetic dose of pentobarbital (40 mg/kg) resulted in partial protection of inactivation of CaM kinase II following ischemia. CaM kinase II activity was maintained following pretreatment of animals with alpha-phenyl N-tert-butyl nitrone (PBN), which traps oxygen free radicals. Infusion of superoxide dismutase or catalase prior to ischemia, blocked CaM kinase II inactivation. Blockage of calcium uptake with bepridil resulted in a marked protection of CaM kinase II inactivation. In addition, trifluoperazine, a calmodulin antagonist also diminished the inhibition of CaM kinase II phosphorylation in our model. These results suggest that ischemia and reperfusion injury results in the generation of activated oxygen and the mobilization of calcium which inactivate CaM kinase II. These results indicate that changes associated with protein kinase activity in the brain following an ischemic insult may have profound effects upon neurodegeneration and neuronal survival.

Identification and characterization of a nucleotide binding site on recombinant murine granulocyte/macrophage-colony stimulating factor

Granulocyte/macrophage-colony stimulating factor (GM-CSF) is a regulatory cytokine important in the proliferative and functional activation of hematopoietic cells. It belongs to a family of 20 kDa or less acidic glycoprotein molecules found in a broad range of cellular sources. On the basis of the previously reported nucleotide-binding properties of interleukin-2 (IL-2), atrial natriuretic factor (ANF), and glucagon, the interaction of GM-CSF with nucleotides was investigated. Using radiolabeled 8-azidoadenosine-containing photoprobes of ATP ([gamma-32P]-8N3ATP) and Ap4A, the putative biological alarmone ([beta'-32P]-8N3Ap4A), we have identified a nucleotide binding site on recombinant murine GM-CSF (rmGM-CSF). Specificity of binding was demonstrated by saturation and competition experiments. Saturation of photoinsertion by [gamma-32P]-8N3ATP and [beta'-32P]-8N3Ap4A occurs with apparent Kd's of 10 and 0.7 microM, respectively. Using an immobilized Fe3+ affinity chromatography technique, developed specifically for the isolation of photolabeled peptides, a single radiolabeled peptide was isolated. It was identified as amino acids 5-14 near the N-terminus of GM-CSF. This peptide region has been shown in previous studies to be critical for biological activity. Also consistent with this observation is our finding that the photolabeled GM-CSF has lost most, if not all, of its biological activity, as determined by a cellular proliferation assay.

NAD+ binding site of Clostridium botulinum C3 ADP-ribosyltransferase. Identification of peptide in the adenine ring binding domain using 2-azido NAD

C3 ADP-ribosyltransferase is an exoenzyme produced by certain strains of Clostridium botulinum types C and D, which specifically ADP-ribosylates rho proteins in eukaryotic cells. Using the photoaffinity probe [alpha-32P]nicotinamide-2-azidoadenine dinucleotide, we have identified the adenine ring binding domain of the NAD+ binding site. The specificity of labeling was demonstrated by saturation effects and protection by the natural compound at physiologically relevant concentrations. Saturation of labeling was observed at 50 microM. Protection experiments indicated an 80% protection of labeling by 100 microM NAD+ when protein was photolyzed in the presence of 10 microM probe. Trypsin or Staphylococcus aureus V8 protease digestion of the photolabeled protein, along with boronate affinity chromatography and immobilized metal affinity chromatography, was used to specifically isolate the peptide region photolabeled with the probe. The peptide corresponded to Phe9-Gly19 near the N terminus.

Identification of peptides from the adenine binding domains of ATP and AMP in adenylate kinase: isolation of photoaffinity-labeled peptides by metal chelate chromatography

Photoaffinity labeling with azidoadenine nucleotides was used to identify peptides from the ATP and AMP binding domains on chicken muscle adenylate kinase. Competition binding studies and enzyme assays showed that the 8-azido analogues of Ap4A and ATP modified only the MgATP2- site of adenylate kinase, whereas the 2-azido analogue of ADP modified the enzyme at both the ATP and AMP sites. The positions of the two nucleotide binding sites on the enzyme were deduced by isolating and sequencing the modified peptides. Photolabeled peptides were isolated by a new procedure that used metal chelate chromatography to affinity purify the photolabeled peptides prior to final purification by reverse-phase HPLC. The sequences of the peptides that were photolabeled with the 8-azido analogues corresponded to residues K28-L44, T153-K166, and T125-E135 of the chicken muscle enzyme. The residues that were present in both tryptic- and Staphylococcus aureus V-8 protease-generated versions of these peptides were assigned to the ATP binding domain on the basis of selective photoaffinity labeling with the 8-azidoadenine analogues. These peptides and an additional peptide corresponding to positions I110-K123 were photolabeled with 2-N3ADP. Since I110-K123 was photolabeled by 2-N3ADP but not by 8-N3Ap4A, it was assigned to the AMP binding domain.

Identification and characterization of serine/threonine protein kinase activity intrinsic to the L protein of vesicular stomatitis virus New Jersey

A photoaffinity analogue of ATP, 8-azido-adenosine 5'-triphosphate (8-N3ATP), was used to probe ATP-binding sites in native transcription complexes of vesicular stomatitis virus (VSV) (New Jersey serotype). The analogue was found to be a substrate for a serine/threonine protein kinase that phosphorylated both the NS and L proteins of native complexes. The analogue failed to interact with the RNA polymerase, another ATP-utilizing activity associated with the transcription complex. Kinetic analyses of both ATP and 8-N3ATP utilization by the protein kinase yielded biphasic saturation curves. Photolysis of 8-N3ATP in the presence of VSV transcription complexes resulted in selective labelling of the L protein. The photolabelling of L was saturable and apparently biphasic. Photolabelling of the L protein was significantly reduced by competition with ATP whereas other nucleoside triphosphates (GTP, UTP and CTP) were ineffective competitors. The stoichiometry of photolabelling was 0.2 at 10 microM-8N3ATP and 1.3 at 100 microM-ATP. These data provide chemical evidence for a virus-encoded serine/threonine protein kinase which resides on the L protein.

Synthesis and characterization of 5-azido-UDP-glucuronic acid. A new photoaffinity probe for UDP-glucuronic acid-utilizing proteins

A new active site-directed photoaffinity analogue, [beta-32P]5-azido-UDP-glucuronic acid (UDP-GlcA), was enzymatically synthesized from [beta-32P]5-N3UDP-Glc using UDP-glucose dehydrogenase. The product was characterized by its mobility on ion exchange and two thin-layer chromatographic systems, by its UV absorbance at 288 nm, and the loss of this absorbance after UV irradiation of the compound. Photoincorporation of [beta-32P]5-N3UDP-GlcA into bovine liver UDP-Glc dehydrogenase (EC 1.1.1.22) was saturable with an apparent Kd of 12.5 microM, and was inhibited by the known active-site effectors UDP-GlcA, UDP-Glc, and UDP-xylose. When human liver microsomes with known UDP-glucuronosyltransferase (EC 2.4.1.17) activities were photolabeled with [beta-32P]5-N3UDP-GlcA, major photolabeled bands of 35-37 and 50-54 kDa were detected. When rat liver microsomes from phenobarbital-injected rats were photolabeled with [beta-32P]5-N3UDP-GlcA, there was a marked increase in photoincorporation of a 51-kDa protein as compared with control animals. Evidence is presented which suggests that the photolabeled 51-54-kDa proteins in the liver microsomes from both tissues are UDP-glucuronosyltransferase and that [beta-32P]5-N3UDP-GlcA represents a new alternative approach in the study of UDP-glucuronosyltransferase and other UDP-GlcA-utilizing enzymes.

Photoaffinity labeling of atrial natriuretic factor analog atriopeptin III with [gamma-32P]8N3GTP

We have determined that atrial natriuretic factor (ANF) analog atriopeptin III (AP III) is a guanosine 5'-triphosphate (GTP) binding protein as shown by photoaffinity labeling with the GTP-analog [gamma-32P]8N3GTP (azido-GTP). The photo-insertion was saturable at 40 microM azido-GTP. Half-maximal protection from photo-insertion was observed using 40-microM concentrations of the native nucleotide GTP. The photolabeling was inhibited approximately 70% by addition of micromolar concentrations of the divalent cations MgCl2, MnCl2 or CaCl2. The related purine-containing nucleotide adenosine 5'-triphosphate (ATP) was able to elicit half-maximal protection of AP III from photo-insertion with 20 microM azido-GTP at a concentration of 120 microM. No significant protection from photoinsertion was observed with other compounds such as GDP, GMP, cGMP, UTP and pyrophosphate. Additionally, a carboxy-terminal F-R-Y deleted analog of ANF, AP I; reduced and carboxymethylated AP III; and resact, a peptide released from sea urchin eggs, were not photolabeled by azido-GTP. These results demonstrate that the AP III analog of ANF contains a functional GTP binding site that would be saturated at physiological concentrations of GTP. At present, the role of this interaction in mediating the biological effects of ANF is unknown, but we speculate that direct GTP interaction with AP III may be involved in the mechanism of action of this peptide.