Two phosphorylation sites on eIF-2 alpha

Protein synthesis in mammalian cells can be regulated through phosphorylation/dephosphorylation of the alpha subunit of initiation factor 2, eIF-2. Two specific kinases have been identified that apparently phosphorylate the same site(s). Controversy exists as to whether serine-48 is a phosphorylation site in addition to serine-51. A recent publication is discussed that, in this author's view, answers the question of the phosphorylation sites. It is suggested that phosphorylation proceeds sequentially with serine-51 being the first and serine-48 the second phosphorylation site. Phosphorylation of both sites is required for inhibition of protein synthesis.     

Phosphorylation of vertebrate nonmuscle and smooth muscle myosin heavy chains and light chains

In this article we review the various amino acids present in vertebrate nonmuscle and smooth muscle myosin that can undergo phosphorylation. The sites for phosphorylation in the 20 kD myosin light chain include serine-19 and threonine-18 which are substrates for myosin light chain kinase and serine-1 and/or-2 and threonine-9 which are substrates for protein kinase C. The sites in vertebrate smooth muscle and nonmuscle myosin heavy chains that can be phosphorylated by protein kinase C and casein kinase II are also summarized.Original data indicating that treatment of human T-lymphocytes (Jurkat cell line) with phorbol 12-myristate 13-acetate results in phosphorylation of both the 20 kD myosin light chain as well as the 200 kD myosin heavy chain is presented. We identified the amino acids phosphorylated in the human T-lymphocytes myosin light chains as serine-1 or serine-2 and in the myosin heavy chains as serine-1917 by 1-dimensional isoelectric focusing of tryptic phosphopeptides. Untreated T-lymphocytes contain phosphate in the serine-19 residue of teh myosin light chain and in a residue tentatively identified as serine-1944 in the myosin heavy chain.Abbreviations MLC myosin light chain - MHC myosin heavy chain - Tris tris(hydroxymethyl)aminomethane - EGTA [ethylenebis(oxyethylenenitrilo)]tetraacetic acid - EDTA ethylenediaminetetraacetate - TPCK N-tosyl-L-phenylalanine chloromethyl ketone - PMA phorbol 12-myristate 13-acetate     

Phosphorylation sites of bovine brain myelin basic protein phosphorylated with Ca2+-calmodulin-dependent protein kinase from rat brain

The phosphorylation sites of myelin basic protein from bovine brain were determined after phosphorylation with Ca2+-calmodulin-dependent protein kinase. Four phosphorylated peptides were selectively and rapidly separated by reversed-phase high-performance liquid chromatography. Partial sequencing of the phosphorylated peptides by automated Edman degradation revealed that Ca2+-calmodulin-dependent protein kinase phosphorylated serine-16, serine-70, and threonine-95 specifically, as well as serine-115, which is located on the experimental allergic encephalitogenic determinant of the protein. Of the four amino acid sequences determined, two sequences surrounding phosphorylated amino acids, -Lys-Tyr-Leu-Ala-Ser(P)16-Ala- and -Arg-Phe-Ser(P)115-Trp-Gly-, have both sides of each phosphoserine residue occupied by hydrophobic amino acids, and a basic amino acid, arginine or lysine, is located at the position 2 or 4 residues amino-terminal to the phosphoserine residue. In contrast, the two other sequences surrounding phosphorylated amino acids, -Tyr-Gly-Ser(P)70-Leu-Pro-Glu-Lys- and -Ile-Val-Thr(P)95-Pro-Arg-, have a basic amino acid at the position 2 or 4 residues carboxyl-terminal to the phosphoamino acid residue.     

In vitro protein kinase C phosphorylation sites of placental lipocortin

Human placental lipocortin is a high-affinity substrate for rat brain protein kinase C in vitro with phosphorylation occurring on serine and threonine residues in a ratio of approximately 2 to 1. Comparison of the ability of various N-terminal-truncated derivatives of lipocortin to serve as phosphorylation substrates, and direct analysis of the N-terminal peptides cleaved from 32P-labeled lipocortin, indicated that threonine-24, serine-27, and serine-28 were the phosphorylation sites. The possibility is discussed that a lysine residue near the carboxy side of the phosphorylation site was involved in lipocortin interaction with the catalytic site of protein kinase C.     

The substrate specificity of adenosine 3'':5''-cyclic monophosphate-dependent protein kinase of rabbit skeletal muscle.

The known amino acid sequences at the two sites on phosphorylase kinase that are phosphorylated by cyclic AMP-dependent protein kinase were extended. The sequences of 42 amino acids around the phosphorylation site on the alpha-subunit and of 14 amino acids around the phosphorylation site on the beta-subunit were shown to be: alpha-subunit Phe-Arg-Arg-Leu-Ser(P)-Ile-Ser-Thr-Glu-Ser-Glx-Pro-Asx-Gly-Gly-His-Ser-Leu-Gly-Ala-Asp-Leu-Met-Ser-Pro-Ser-Phe-Leu-Ser-Pro-Gly-Thr-Ser-Val-Phe(Ser,Pro,Gly)His-Thr-Ser-Lys; beta-subunit, Ala-Arg-Thr-Lys-Arg-Ser-Gly-Ser(P)-VALIle-Tyr-Glu-Pro-Leu-Lys. The sites on histone H2B which are phosphorylated by cyclic AMP-dependent protein kinase in vitro were identified as serine-36 and serine-32. The amino acid sequence in this region is: Lys-Lys-Arg-Lys-Arg-Ser32(P)-Arg-Lys-Glu-Ser36(P)-Tyr-Ser-Val-Tyr-Val- [Iwai, K., Ishikawa, K. & Hayashi, H. (1970) Nature (London) 226, 1056-1058]. Serine-36 was phosphorylated at 50% of the rate at which the beta-subunit of phosphorylase kinase was phosphorylated, and it was phosphorylated 6-7-fold more rapidly than was serine-32. The amino acid sequences when compared with those at the phosphorylation sites of other physiological substrates suggest that the presence of two adjacent basic amino acids on the N-terminal side of the susceptible serine residue may be critical for specific substrate recognition in vivo.     

Identification of the sites on rabbit skeletal muscle protein phosphatase inhibitor-2 phosphorylated by casein kinase-II

Inhibitor-2 was phosphorylated by casein kinase-II in vitro at a rate similar to that of glycogen synthase, a physiological substrate of this protein kinase. The major phosphorylation sites were identified as serines-86, -120 and -121, the peptide containing serines-120 and -121 being labelled about 2.5-fold more rapidly than that containing serine-86. The 13 residues C-terminal to serine-121 (SGEEDSDLSPEERE) contain seven acidic amino acids, while the six residues following serine-86 (SDTETTE) contain three. These results are consistent with the known specificity requirements of casein kinase-II. The three serines are C-terminal to the threonine (residue 72) whose phosphorylation by glycogen synthase kinase-3 is potentiated by prior phosphorylation with casein kinase-II. This reinforces the view that a C-terminal phosphoserine residue is important for the specificity of glycogen synthase kinase-3. Identification of the residues phosphorylated by casein kinase-II will facilitate further studies on the in vivo phosphorylation state of inhibitor-2.     

Alterations of cell membrane composition during prolonged fasting in lizards

Autophosphorylation of cyclic GMP-dependent protein kinase (GMP:protein phosphotransferase, EC 2.7.1.37) in the presence of cyclic AMP and Mg-ATP has already been shown to result in the incorporation of up to 2.6 mol phosphate per mol subunit and decrease the A0.5 for cyclic AMP approx. 10-fold. The major sites of autophosphorylation have now been identified as serine-50, threonine-58, serine-72 and threonine-84. Serine-1 and serine-64 are phosphorylated to a minor extent. Threonine-58, which is initially phosphorylated most rapidly, is also the major site that is phosphorylated in the presence of cyclic GMP and Mg-ATP. Since autophosphorylation in the presence of cyclic GMP does not decrease the A0.5 for cyclic AMP, phosphorylation of serine-50, serine-72, or threonine-84 must be responsible for this effect.     

Phosphorylation of Argonaute 2 at serine-387 facilitates its localization to processing bodies

Ago (Argonaute) proteins are essential effectors of RNA-mediated gene silencing. To explore potential regulatory mechanisms for Ago proteins, we examined the phosphorylation of human Ago2. We identified serine-387 as the major Ago2 phosphorylation site in vivo. Phosphorylation of Ago2 at serine-387 was significantly induced by treatment with sodium arsenite or anisomycin, and arsenite-induced phosphorylation was inhibited by a p38 MAPK (mitogen-activated protein kinase) inhibitor, but not by inhibitors of JNK (c-Jun N-terminal kinase) or MEK [MAPK/ERK (extracellular-signal-regulated kinase) kinase]. MAPKAPK2 (MAPK-activated protein kinase-2) phosphorylated bacterially expressed full-length human Ago2 at serine-387 in vitro, but not the S387A mutant. Finally, mutation of serine-387 to an alanine residue or treatment of cells with a p38 MAPK inhibitor reduced the localization of Ago2 to processing bodies. These results suggest a potential regulatory mechanism for RNA silencing acting through Ago2 serine-387 phosphorylation mediated by the p38 MAPK pathway.     

Protein kinase C-related kinase 2 phosphorylates the protein synthesis initiation factor eIF4E in starfish oocytes

Phosphorylation of eIF4E is required for protein synthesis during starfish oocyte maturation. The activity of protein kinase C-related kinase 2 (PRK2) increases prior to the phosphorylation of eIF4E (G. Stapleton et al., 1998, Dev. Biol. 193, 34-46). We investigate here whether eIF4E is activated by PRK2. A 3.5-kb eIF4E clone isolated from starfish cDNA is 57% identical to human eIF4E and contains the putative phosphorylation site serine-209. The serine-209 environment (SKTGS(209)MAKSRF) is similar to the consensus sequence of the phosphorylation site of protein kinase C and related kinases. A starfish eIF4E fusion protein (GST-4E) was phosphorylated in vitro by PRK2 in the presence of 1,2-diolyl-sn-glycerol 3-phosphate. In contrast, replacing the GST-4E serine-209 with an alanine significantly reduced this phosphorylation. Analysis by two-dimensional phosphopeptide mapping reveals a major phosphopeptide in trypsin-digested GST-4E, but not in its serine-209 mutant. Importantly, this major phosphopeptide in GST-4E corresponds to a major phosphopeptide of eIF4E isolated from (32)P-labeled oocytes. Thus, PRK2 may regulate translation initiation during oocyte maturation by phosphorylating the serine-209 residue of eIF4E in starfish. We also demonstrate that high levels of cAMP inhibit the activation of PRK2, eIF4E, and the eIF4E binding protein during starfish oocyte maturation, while PI3 kinase activates these proteins.     

Phosphorylation and activation of tryptophan hydroxylase 2: identification of serine-19 as the substrate site for calcium, calmodulin-dependent protein kinase II

Tryptophan hydroxylase (TPH) is the initial and rate-limiting enzyme in the biosynthesis of serotonin. TPH was once thought to be a single-gene product but it is now known to exist in two isoforms. TPH1 is found in the periphery and pineal gland whereas TPH2 is expressed specifically in the CNS. Both TPH isoforms are known to be regulated by protein kinase-dependent phosphorylation and the sites of modification of TPH1 by protein kinase A have been identified. While TPH2 is activated by calcium, calmodulin-dependent protein kinase II (CaMKII), the sites at which this isoform is modified are not known. Treatment of wild-type TPH2 with CaMKII followed by mass spectrometry analysis revealed that the enzyme was activated and phosphorylated at a single site, serine-19. Mutagenesis of serine-19 to alanine did not alter the catalytic function of TPH2 but this mutant enzyme was neither activated nor phosphorylated by CaMKII. A phosphopeptide bracketing phosphoserine-19 in TPH2 was used as an antigen to generate polyclonal antibodies against phosphoserine-19. The antibodies are highly specific for phosphoserine-19 in TPH2. The antibodies do not react with wild-type TPH2 or TPH1 and they do not recognize phophoserine-58 or phosphoserine-260 in TPH1. These results establish that activation of TPH2 by CaMKII is mediated by phosphorylation of serine-19 within the regulatory domain of the enzyme. Production of a specific antibody against the CaMKII phosphorylation site in TPH2 represents a valuable tool to advance the study of the mechanisms regulating the function of this important enzyme.     

Phosphorylation of conserved serine residues does not regulate the ability of mosxe protein kinase to induce oocyte maturation or function as cytostatic factor

Expression of the mosxe protein kinase is required for the normal meiotic maturation of Xenopus oocytes and overexpression induces maturation in the absence of other stimuli. In addition, mosxe functions as a component of cytostatic factor (CSF), an activity responsible for arrest of the mature egg at metaphase II. After microinjection of Xenopus oocytes with in vitro synthesized RNA encoding either wild-type mosxe or kinase-inactive mosxe(R90), both proteins are phosphorylated exclusively on serine residues and exhibit essentially identical chymotryptic maps. Since the phosphorylated kinase-inactive mosxe(R90) protein was recovered from resting oocytes that have not yet begun to translate endogenous mosxe, this indicates that the major phosphopeptides of mosxe(R90) are phosphorylated by a preexisting protein kinase present in resting oocytes, and are not the result of autophosphorylation. The results presented here also indicate that the mosxe protein does not undergo significant phosphorylation at unique sites during oocyte maturation. If the biological activity of mosxe were regulated by phosphorylation, a site of regulatory phosphorylation would most likely be conserved among mos proteins of different species. Site-directed mutagenesis was used to construct 13 individual serine----alanine mutations at conserved residues (3, 16, 18, 25, 26, 57, 71, 76, 102, 105, 127, 211, and 258). These 13 mutants were analyzed for their abilities to induce oocyte maturation and to function as CSF. Results obtained with the mosxe(A105) mutant revealed that serine-105 is required for both maturation induction and CSF activity, even though serine-105 does not represent a major site of phosphorylation. All of the remaining serine----alanine mosxe mutants induced oocyte maturation and exhibited CSF activity comparable with the wild type. These results demonstrate that none of the conserved serines examined in this study function as regulatory phosphorylation sites for these biological activities. Peptide mapping of the remaining mosxe mutants identified serine-3 as a major phosphorylation site in vivo, which is contained within the chymotryptic peptide MPSPIPVERF.     

Phosphorylation of bovine platelet myosin by protein kinase C

Bovine platelet myosin is phosphorylated by protein kinase C at multiple sites. Most of the phosphate is incorporated in the 20,000-dalton light chain although some phosphate is incorporated in the heavy chain. Phosphorylation of the 20,000-dalton light chain of platelet myosin is 10 times faster than the phosphorylation of smooth muscle myosin. Platelet myosin light chain is first phosphorylated at a threonine residue followed by a serine residue. Dominant phosphorylation sites of the 20,000-dalton light chain are estimated as serine-1, serine-2, and threonine-9. Prolonged phosphorylation by protein kinase C resulted in an additional phosphorylation site which, on the basis of limited proteolysis, appears to be either serine-19 or threonine-18. Phosphorylation by protein kinase C causes an inhibition of actin-activated ATPase activity of platelet myosin prephosphorylated by myosin light chain kinase. Inhibition of ATPase activity is due to a decreased affinity of myosin for actin, and no change in Vmax is observed. It is shown that platelet myosin also exhibits the 6S to 10S conformation transition as judged by viscosity and gel filtration methods. Mg2(+)-ATPase activity of platelet myosin is paralleled with the 10S-6S transition. Phosphorylation by protein kinase C affects neither the 10S-6S transition nor the myosin filament formation. Therefore, the inhibition of actin-activated ATPase activity of platelet myosin is not due to the change in the myosin conformation.     

Phospholipid-Sensitive Ca2+-Dependent Protein Kinase Preferentially Phosphorylates Serine-115 of Bovine Myelin Basic Protein

Phospholipid-sensitive Ca2+ -dependent protein kinase (PL-Ca-PK) and cyclic AMP-dependent protein kinase (A-PK) both preferentially phosphorylated serine residues of bovine myelin basic protein (MBP). Tryptic peptide maps of MBP phosphorylated by PL-Ca-PK or A-PK, however, revealed different phosphopeptides, suggesting a difference in the intramolecular substrate specificity for the two enzymes. Serine-115 of MBP, in the sequence (-Arg-Phe-Ser(115)-Trp-), was found to be a preferred and probably major phosphorylation site for PL-Ca-PK. Because serine-115 of bovine MBP corresponds to serine-113 of rabbit MBP, an in vivo phosphorylation site reported by Martenson et al. (1983), and PL-Ca-PK is present at a very high level in brain and myelin, it is suggested that the enzyme may be responsible for the in vivo phosphorylation of this and other sites in MBP.     

Structure-function relationships in cardiac troponin T

Regions of rabbit and bovine cardiac troponin T that are involved in binding tropomyosin, troponin C and troponin I have been identified. Two sites of contact for tropomyosin have been located, situated between residues 92-178 and 180-284 of troponin T. A cardiac-specific binding site for troponin I has been identified between residues 1-68 of cardiac troponin T, within a region of the protein that has previously been shown to be encoded by a series of exons that are expressed in a tissue-specific and developmentally regulated manner. The binding site for troponin C is located between residues 180-284 of cardiac troponin T. When isolated from fresh bovine hearts, cardiac troponin T contained 0.21 +/- 0.11 mol phosphate per mol; incubation with phosphorylase kinase increased the phosphate content to approx. 1 mol phosphate per mol. One site of phosphorylation was identified as serine-1; a second site of phosphorylation was located within peptide CB3 (residues 93-178) and has been tentatively identified as serine-176. Addition of troponin C to cardiac troponin T does not inhibit the phosphorylation of this latter protein that is catalysed by phosphorylase b kinase.     

A serine/threonine phosphorylation site in the ectodomain of a T cell receptor beta chain is required for activation by superantigen

The presence of consensus phosphorylation sites in the ectodomains of cell surface proteins suggests that such post-translational modification may be important in regulation of surface receptor activity. To date, the only cell surface receptor for which such ectodomain phosphorylation has been conclusively demonstrated is the clonally expressed T cell antigen receptor (TCR). Attempts to conclusively identify individual phosphorylated residues in TCR alpha and beta chains and determine their functional significance by biochemical approaches failed due to insufficient quantities of purified molecules. Here we present the results of an alternative approach where survey of phosphorylation sites in the TCR alpha and beta chains was accomplished using site-directed mutagenesis and retroviral vector expression, as well as in vitro phosphorylation of synthetic peptide substrates. All mutants studied directed the cell surface expression of normal amounts of TCR, and all transfectants could be stimulated to produce IL-2 in response to substrate-immobilized antibody to TCR. However, mutation of serine-88 in the protein kinase A phosphorylation site of the TCR beta chain resulted in a complete lack of response to the superantigen staphylococcal enterotoxin B (SEB). In addition, this mutation abolished TCR-associated tyrosine phosphorylation, consistent with the impairment of cell signaling. Reversion of the serine-88/alanine mutation with phosphorylatable threonine completely restored the SEB recognition by TCR. These results, interpreted in the context of the known three-dimensional structure of the complex of SEB and TCR, are consistent with the view that serine-88 is important for the contact of the TCR beta chain with SEB.     

cPLA2 phosphorylation at serine-515 and serine-505 is required for arachidonic acid release in vascular smooth muscle cells

Cytosolic phospholipase A(2) (cPLA(2)) is activated by phosphorylation at serine-505 (S505) by extracellular regulated kinase 1/2 (ERK1/2). However, rat brain calcium/calmodulin-dependent kinase II (CaMKII) phosphorylates recombinant cPLA(2) at serine-515 (S515) and increases its activity in vitro. We have studied the sites of cPLA(2) phosphorylation and their significance in arachidonic acid (AA) release in response to norepinephrine (NE) in vivo in rabbit vascular smooth muscle cells (VSMCs) using specific anti-phospho-S515- and -S505 cPLA(2) antibodies and by mutagenesis of S515 and S505 to alanine. NE increased the phosphorylation of cPLA(2) at S515, followed by phosphorylation of ERK1/2 and consequently phosphorylation of cPLA(2) at S505. The CaMKII inhibitor 2-[N-(2-hydroxyethyl)]-N-(4-methoxybenzene-sulfonyl)]amino-N-(4-chlorocinnamyl)-methylbenzylamine attenuated cPLA(2) at S515 and S505, whereas the ERK1/2 inhibitor U0126 reduced phosphorylation at S505 but not at S515. NE in cells transduced with adenovirus carrying enhanced cyan fluorescent protein cPLA(2) wild type caused phosphorylation at S515 and S505 and increased AA release. Expression of the S515A mutant in VSMCs reduced the phosphorylation of S505, ERK1/2, and AA release in response to NE. Transduction with a double mutant (S515A/S505A) blocked the phosphorylation of cPLA(2) and AA release. These data suggest that the NE-stimulated phosphorylation of cPLA(2) at S515 is required for the phosphorylation of S505 by ERK1/2 and that both sites of phosphorylation are important for AA release in VSMCs.     

Phosphorylation of Serine-880 in GluR2 by Protein Kinase C Prevents Its C Terminus from Binding with Glutamate Receptor-Interacting Protein

Phosphorylation of the glutamate receptor is an important mechanism of synaptic plasticity. Here, we show that the C terminus of GluR2 of the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptor is phosphorylated by protein kinase C and that serine-880 is the major phosphorylation site. This phosphorylation also occurs in human embryonic kidney (HEK) cells by addition of 12-O-tetradecanoylphorbol 13-acetate. Our immunoprecipitation experiment revealed that the phosphorylation of serine-880 in GluR2 drastically reduced the affinity for glutamate receptor-interacting protein (GRIP), a synaptic PDZ domain-containing protein, in vitro and in HEK cells. This result suggests that modulation of serine-880 phosphorylation in GluR2 controls the clustering of AMPA receptors at excitatory synapses and consequently contributes to synaptic plasticity.     

Phosphoserine as a recognition determinant for glycogen synthase kinase-3: phosphorylation of a synthetic peptide based on the G-component of protein phosphatase-1

Prior phosphorylation of its substrate has been shown to be important for substrate recognition by the protein kinase glycogen synthase kinase-3 (GSK-3). Phosphorylation of glycogen synthase by GSK-3 is known to be enhanced by the previous action of casein kinase II and the sequence -SXXXS(P)- was proposed as the minimal recognition determinant for GSK-3. The glycogen binding subunit of type 1 phosphoprotein phosphatase has been shown to be phosphorylated by cyclic AMP-dependent protein kinase at serine-13 in the sequence KPGFS(5)PQPS(9)RRGS(13)ESSEEVYV (F.B. Caudwell, A. Hiraga, and P. Cohen (1986) FEBS Lett. 194, 85-89). Inspection of the sequence revealed potential GSK-3 sites at residues 5 and 9. Using a synthetic peptide with the above sequence, we found that phosphorylation of serine-13 by cyclic AMP-dependent protein kinase permitted the recognition of serine-9 and serine-5 by GSK-3. The work provides another example of a substrate for GSK-3 and demonstrates that the action of GSK-3 is linked to the presence of phosphate in the substrate and not the action of any particular protein kinase. In the course of the analyses, a novel feature of trypsin cleavage of phosphopeptides was noted. In the sequence -SRRGS(P)- trypsin acted uniquely after the first arginine whereas in the sequence -S(P)RRGS(P)- it cleaved randomly at either arginine residue. The fact that GSK-3 could phosphorylate a peptide derived from a phosphatase subunit also raises the possibility that GSK-3 might be involved in controlling glycogen-associated type 1 phosphatase and, more generally, in mediating cyclic AMP control of protein phosphorylation in cells.     

Identification of phosphoproteins and their phosphorylation sites in the WEHI-231 B lymphoma cell line

A major goal of the Alliance for Cellular Signaling is to elaborate the components of signal transduction networks in model cell systems, including murine B lymphocytes. Due to the importance of protein phosphorylation in many aspects of cell signaling, the initial efforts have focused on the identification of phosphorylated proteins. In order to identify serine- and threonine-phosphorylated proteins on a proteome-wide basis, WEHI-231 cells were treated with calyculin A, a serine/threonine phosphatase inhibitor, to induce high levels of protein phosphorylation. Proteins were extracted from whole-cell lysates and digested with trypsin. Phosphorylated peptides were then enriched using immobilized metal affinity chromatography and identified by liquid chromatography-tandem mass spectrometry. A total of 107 proteins and 193 phosphorylation sites were identified using these methods. Forty-two of these proteins have been reported to be phosphorylated, but only some of them have been detected in B cells. Fifty-four of the identified proteins were not previously known to be phosphorylated. The remaining 11 phosphoproteins have previously only been characterized as novel cDNA or genomic sequences. Many of the identified proteins were phosphorylated at multiple sites. The proteins identified in this study significantly expand the repertoire of proteins known to be phosphorylated in B cells. The number of newly identified phosphoproteins indicates that B cell signaling pathways utilizing protein phosphorylation are likely to be more complex than previously appreciated.