- Email: [email protected]
Desperately Seeking…Something
Previews 219
Local versus Global Activation of Receptor Tyrosine Kinases
the internalization of NGF and NGF receptors (TrkA) but rather could be mediated by lateral propagation or a “wave” of TrkA activation induced by NGF from the nerve terminal to the neuronal cell body (MacInnis and Campenot, 2002; Miller and Kaplan, 2002). The experiments presented by Sawano et al. (2002) are at odds with such a mechanism. Local stimulation of cell signaling by extracellular cues appears to play an important role in embryonic development in which local release of soluble or extracellular matrix-bound growth factors play an important role in the development of tissues and organs, such as the development of the peripheral and central nervous system as well as in limb development. The recent debate on the biological significance of the immune-synapse also underscores the importance of localized versus global signaling in the control of signaling in immune cells. Finally, the global activation of EGFR observed in cells that overexpress EGFR (Verveer et al., 2000; Sawano et al., 2002) or when endocytosis of EGFR is prevented (Sawano et al., 2002) may play a role in cell transformation induced by overexpression of EGFR in human tumors. Overexpression of EGFR or erbB2 caused by gene amplification or other mechanisms has been implicated in the development of many human cancers including malignant glioblastomas and mammary carcinomas, respectively (Blume-Jensen and Hunter, 2001). Global activation of EGFR or erbB2 in these tumors may exert a strong mitogenic response as well as other hallmarks of cell transformation such as changes in cell morphology and cell migration.
(A) Local dimerization and activation of receptor tyrosine kinases leads to localized signal transmission. (B) Locally activated receptor tyrosine kinases phosphorylate inactive, unoccupied, or occupied receptor monomers, triggering a chain reaction that culminates in global receptor activation and signal transmission.
Joseph Schlessinger Department of Pharmacology Yale University School of Medicine New Haven, Connecticut 06510 Selected Reading
induced production of hydrogen peroxide will not inhibit the activity of a variety of PTPases resulting in the activation of many receptor and cytoplasmic protein tyrosine kinases in response to EGF stimulation. The experiment described by Sawano et al. (2002) clearly demonstrates that, at physiological levels of EGFR expression, signaling via EGFR is confined to the ligand-stimulated region and that lateral propagation of the signal occurs only under conditions of EGFR overexpression. The fact that activation of RTKs is confined to the region of ligand binding is of significant interest as PTKs play a critical role in the control of many important biological processes. For example, it was recently reported that NGFinduced retrograde survival signal may not depend on
Desperately Seeking…Something Published online this week in Structure, Wisely et al. present a high-resolution X-ray crystallographic structure of the ligand binding domain of human hepatocyte nuclear factor 4 ␥ (HNF4␥). They find fatty acids filling the ligand binding pocket of this receptor long consid-
Blume-Jensen, P., and Hunter, T. (2001). Nature 17, 355–365. Heldin, C.H. (1995). Cell 80, 213–223. Lee, S.R., Kwon, K.S., Kim, S.R., and Rhee, S.G. (1998). J. Biol. Chem. 273, 15366–15372. Lemmon, M.A., and Schlessinger, J. (1994). Trends Biochem. Sci. 19, 459–463. MacInnis, B.L., and Campenot, R.B. (2002). Science 295, 1536–1539. Miller, F.D., and Kaplan, D.R. (2002). Science 295, 1471–1472. Sawano, A., Takayama, S., Matsuda, M., and Miyawaki, A. (2002). Dev. Cell 3, 245–257. Schlessinger, J. (2000). Cell 103, 211–215. Verveer, P.J., Wouters, F.S., Reynolds, A.R., and Bastiaens, P.I.H. (2000). Science 290, (567–570).
ered an orphan, but these “ligands” appear to be locked into the protein and not readily amenable to exchange. Not only does this present a new paradigm for nuclear receptors but it also provides new insights into their evolutionary origins. The nuclear receptor superfamily of transcription factors consists of over 100 different members from every metazoan organism. Membership in this elite club which reg-
Molecular Cell 220
Model of Nuclear Receptor Action (A) For nuclear receptors with known ligands, inactive receptor dimers bound to promoter regions recruit corepressor complexes which keep the transcription of the target gene turned off. Upon ligand (agonist) binding, the receptor undergoes a conformational change that puts the AF-2 in an active conformation which recruits coactivators and transcription ensues. Purple oval, DBD; blue rectangle/oval, LBD. (B) In contrast, the work by Wisely et al. (2002) shows that HNF4 constitutively binds fatty acids as structural cofactors, raising the question of how the switch is made from corepressor to coactivator complex.
ulates the transcription of genes involved in every aspect of physiology—from development to differentiation to disease—is dependent upon amino acid similarity in at least one of the two highly conserved domains: the DNA binding domain (DBD) or the ligand binding domain (LBD). The first members of the family were identified by binding to known ligands-steroids. Subsequent receptors were identified as genes that played functional roles in Drosophila or via low-stringency hybridization to the DBD or LBD of known receptors. With the exception of the steroid receptors, all the other receptors had to undergo a ligand hunt of varying duration before a ligand was identified. That hunt has taken more than a dozen years for HNF4, but as Wisely and coworkers report, the hunt may have finally come to an end. What they found, however, was not a traditional ligand. HNF4␣ was originally identified as a factor that bound a DNA response element critical to the expression of certain genes in the liver (Sladek et al., 1990). Therefore, in contrast to many other orphan receptors, HNF4␣ always had at least one ligand—the DNA elements to which it bound. These elements eventually led to the identification of more than 60 target genes which implicate HNF4␣ in a variety of human diseases, including atherosclerosis and cancer. Point mutations in the HNF4␣ binding sites of several blood coagulation factor genes have also linked HNF4␣ directly to certain types of hemophilia, while mutations in both the coding and the regulatory region of the HNF4␣ gene itself have been found in an inherited form of type II diabetes, MODY1 (Sladek and Seidel, 2001). Therefore, it is not too surprising that many labs worldwide, particularly those in the pharmaceutical industry, have been actively searching for a ligand for HNF4␣. After all, several very effective, and profitable, drugs that act on other nuclear receptors have been developed to treat diseases such as diabetes, cancer, inflammation, and osteoporosis, so why not HNF4␣? The problem has been that in every in vitro and in vivo system analyzed, HNF4␣ activates transcription in a constitutive fashion, i.e., in the absence of exogenously added ligand. Whereas there have been reports that fatty acyl CoA thioesters modulate the activity of HNF4␣ (Hertz et al., 1998), these compounds could not be shown to act as traditional ligands (Bogan et al., 2000).
The mantra in the nuclear receptor field is that bona fide ligands bind in a hydrophobic pocket in the LBD and that binding of ligand alters the position of one of 12 helices (the AF-2), thereby inducing a conformational change that kicks off corepressors and recruits coactivators (see Figure, panel A) (Glass and Rosenfeld, 2000). The fatty acyl CoA thioesters could not be shown to alter the conformation of HNF4␣, to decrease binding to a corepressor, or to increase binding to a coactivator (Bogan et al., 2000; Ruse et al., 2002). Now, the structure solved by the GlaxoSmithKline group suggests why. The pocket of bacterially expressed human HNF4␥ LBD contained a mixture of saturated and cis-monounsaturated C14-C18 fatty acids, in particular C16:0 (palmitic acid) and C16:1 (palmitoleic acid). (Although there is little functional data on HNF4␥, its LBD is 80% identical to that of HNF4␣.) This finding is perhaps not too surprising as other receptors expressed in bacteria have been found, upon crystallization, to bind fatty acids. However, the finding by Wisely et al. (2002) is unusual in that, short of completely denaturing the HNF4␥ (or HNF4␣) LBD, the fatty acid could not be removed from the protein. This and the fact that the purported ligand did not directly contact the AF-2 suggests that the fatty acids are not so much exchangeable ligands that induce allosteric changes in the receptor as they are essential cofactors that are critical to the structure of the protein, much like the zinc in zinc fingers. This work coincides very well with recent work showing that, both in vitro and in vivo, the switch between corepressor and coactivator binding to HNF4␣ can be made without changing the ligand status (Ruse et al., 2002) (see Figure, panel B). How then might HNF4’s activity be regulated? As Wisely et al. (2002) point out, posttranslational modification is one possibility. There are up to 13 potential phosphorylation sites in HNF4␣ (Jiang et al., 1997) as well as sites for acetylation (Soutoglou et al., 2000). Another possibility is that the nine naturally occurring splicing variations in HNF4␣ regulate its activity. The two most abundant forms in the adult liver have already been shown to differentially recruit coactivators (Sladek et al., 1999), and an embryo-specific form has been shown to possess distinct transactivation properties (TorresPadilla et al., 2001). Since the splicing variations are in the N- or the C terminus, both of which are implicated
Previews 221
in transactivation, one possibility is that these protein domains—and not a ligand—regulate the receptor’s function. Finally, there is the possibility that upon synthesis of the HNF4 protein, a different fatty acid can be introduced into the protein. This is not a completely farfetched idea, as an increasing number of both endogenous as well as exogenous compounds have been found to regulate the expression of the HNF4␣ gene (Sladek and Seidel, 2001). However, even if it can be shown that a ligand can be exchanged under some physiological, or therapeutic, condition, it must also be shown that different ligands have the ability to impart different activities on the protein. At this point, it is not clear how that would happen, since the AF-2 does not contact the ligand. Although, the Structure paper notes that the bacterially expressed HNF4␥ bound a fatty acid rarely found in E. coli (C17:1), suggesting that there might at least be a selection process for the ligand/cofactor. Finally, the findings of Wisely et al. (2002) may alter our thinking about the evolution of nuclear receptors. It was previously proposed that HNF4 and two other highly conserved family members were ancestral nuclear receptors that lacked ligands and that the ability to bind ligands arose independently more than once during evolution (Escriva et al., 2000). Now, however, one must consider the possibility that the original nuclear receptor did bind a lipophilic compound, but its role was as an integral component of the protein structure, not as an interchangeable element that induced allosteric and functional changes, as do the more modern ligands. The binding of a nonexchangeable cofactor makes it easier to understand conceptually why the LBD has been conserved during evolution and how the first true ligands arose, i.e., as a variation on a preexisting theme as opposed to a completely new characteristic. In any case, in this post-genomic era it is humbling to consider the possibility that perhaps at least part of what makes us
different from other organisms is not our genes or the proteins they encode but rather the ligands they bind. The author thanks M. Ruse and Y. Maeda for discussions and the National Institutes of Health for funding (R01 DK53892).
Cleave to Leave: Structural Insights into the Dynamic Organization of the Nuclear Pore Complex
The nuclear pore complex (NPC), weighing in at ⵑ100 MDa, is an example of such a large and dynamic supramolecular assembly. Located in the nuclear envelope, NPCs are the gatekeepers between the nucleus and cytoplasm. NPCs consist of approximately 30 different proteins, termed Nups. These multimerize to form an octagonal tube, from which fibers extend into both the cytoplasm and the nucleus (see Figure, panel B). Some soluble transport factors such as GTP-bound Ran power the process, and others such as karyopherins act as carriers, facilitating transport of their cargoes by interacting with particular sequence motifs in specific Nups. To understand its role as a transport machine, recent “holistic” approaches have studied the entire NPC or reconstituted significant constituent subcomplexes. In this issue of Molecular Cell, Hodel et al. (2002) have taken an alternative approach, showing how one can obtain important information about a large and dynamic complex like the NPC by instead looking in detail at key structural domains found in certain of its components. The approach Hodel et al. have taken includes an
A detailed understanding of the fine structure of the nuclear pore complex has remained elusive. Now, studies on a small protein domain have shed light on the dynamic organization of this massive assembly. In the new age of proteomics, determining the interactions between proteins in large, dynamic supramolecular assemblies is a major goal. In a cell, the number of interacting proteins is dauntingly large, with interactions changing depending on the subcellular localization of the partners or the state of the cell. Furthermore, varying affinities and expression levels often make detection of functional complexes a formidable challenge.
Frances M. Sladek Department of Cell Biology and Neuroscience University of California, Riverside Riverside, California 92521 Selected Reading Bogan, A.A., Dallas-Yang, Q., Ruse, M.D., Jr., Maeda, Y., Jiang, G., Nepomuceno, L., Scanlan, T.S., Cohen, F.E., and Sladek, F.M. (2000). J. Mol. Biol. 302, 831–851. Escriva, H., Delaunay, F., and Laudet, V. (2000). Bioessays 22, 717–727. Glass, C.K., and Rosenfeld, M.G. (2000). Genes Dev. 14, 121–141. Hertz, R., Magenheim, J., Berman, I., and Bar, T.J. (1998). Nature 392, 512–516. Jiang, G., Nepomuceno, L., Yang, Q., and Sladek, F.M. (1997). Arch. Biochem. Biophys. 340, 1–9. Ruse, M.D., Jr., Privalsky, M.L., and Sladek, F.M. (2002). Mol. Cell. Biol. 22, 1626–1638. Sladek, F.M., Zhong, W., Lai, E., and Darnell, J.E. (1990). Genes Dev. 4, 2353–2365. Sladek, F.M., Ruse, M.D., Nepomuceno, L., Huang, S.-M., and Stallcup, M.R. (1999). Mol. Cell. Biol. 19, 6509–6522. Sladek, F.M., and Seidel, S.D. (2001). In Nuclear Receptors and Genetic Diseases, T.P. Burris and E.R.B. McCabe, eds. (London: Academic Press), pp. 309–361. Soutoglou, E., Katrakili, N., and Talianidis, I. (2000). Mol. Cell 5, 745–751. Torres-Padilla, M.E., Fougere-Deschatrette, C., and Weiss, M.C. (2001). Mech. Dev. 109, 183–193. Wisely, G.B., Miller, A.B., Davis, R.G., Thornquest, A.D., Jr., Johnson, R., Spitzer, T., Sefler, A., Shearer, B., Moore, J.T., Miller, A.B., et al. (2002). Structure. Published online August 13, 2002. DOI:10.1016S0969212602008298.
Local versus Global Activation of Receptor Tyrosine Kinases
the internalization of NGF and NGF receptors (TrkA) but rather could be mediated by lateral propagation or a “wave” of TrkA activation induced by NGF from the nerve terminal to the neuronal cell body (MacInnis and Campenot, 2002; Miller and Kaplan, 2002). The experiments presented by Sawano et al. (2002) are at odds with such a mechanism. Local stimulation of cell signaling by extracellular cues appears to play an important role in embryonic development in which local release of soluble or extracellular matrix-bound growth factors play an important role in the development of tissues and organs, such as the development of the peripheral and central nervous system as well as in limb development. The recent debate on the biological significance of the immune-synapse also underscores the importance of localized versus global signaling in the control of signaling in immune cells. Finally, the global activation of EGFR observed in cells that overexpress EGFR (Verveer et al., 2000; Sawano et al., 2002) or when endocytosis of EGFR is prevented (Sawano et al., 2002) may play a role in cell transformation induced by overexpression of EGFR in human tumors. Overexpression of EGFR or erbB2 caused by gene amplification or other mechanisms has been implicated in the development of many human cancers including malignant glioblastomas and mammary carcinomas, respectively (Blume-Jensen and Hunter, 2001). Global activation of EGFR or erbB2 in these tumors may exert a strong mitogenic response as well as other hallmarks of cell transformation such as changes in cell morphology and cell migration.
(A) Local dimerization and activation of receptor tyrosine kinases leads to localized signal transmission. (B) Locally activated receptor tyrosine kinases phosphorylate inactive, unoccupied, or occupied receptor monomers, triggering a chain reaction that culminates in global receptor activation and signal transmission.
Joseph Schlessinger Department of Pharmacology Yale University School of Medicine New Haven, Connecticut 06510 Selected Reading
induced production of hydrogen peroxide will not inhibit the activity of a variety of PTPases resulting in the activation of many receptor and cytoplasmic protein tyrosine kinases in response to EGF stimulation. The experiment described by Sawano et al. (2002) clearly demonstrates that, at physiological levels of EGFR expression, signaling via EGFR is confined to the ligand-stimulated region and that lateral propagation of the signal occurs only under conditions of EGFR overexpression. The fact that activation of RTKs is confined to the region of ligand binding is of significant interest as PTKs play a critical role in the control of many important biological processes. For example, it was recently reported that NGFinduced retrograde survival signal may not depend on
Desperately Seeking…Something Published online this week in Structure, Wisely et al. present a high-resolution X-ray crystallographic structure of the ligand binding domain of human hepatocyte nuclear factor 4 ␥ (HNF4␥). They find fatty acids filling the ligand binding pocket of this receptor long consid-
Blume-Jensen, P., and Hunter, T. (2001). Nature 17, 355–365. Heldin, C.H. (1995). Cell 80, 213–223. Lee, S.R., Kwon, K.S., Kim, S.R., and Rhee, S.G. (1998). J. Biol. Chem. 273, 15366–15372. Lemmon, M.A., and Schlessinger, J. (1994). Trends Biochem. Sci. 19, 459–463. MacInnis, B.L., and Campenot, R.B. (2002). Science 295, 1536–1539. Miller, F.D., and Kaplan, D.R. (2002). Science 295, 1471–1472. Sawano, A., Takayama, S., Matsuda, M., and Miyawaki, A. (2002). Dev. Cell 3, 245–257. Schlessinger, J. (2000). Cell 103, 211–215. Verveer, P.J., Wouters, F.S., Reynolds, A.R., and Bastiaens, P.I.H. (2000). Science 290, (567–570).
ered an orphan, but these “ligands” appear to be locked into the protein and not readily amenable to exchange. Not only does this present a new paradigm for nuclear receptors but it also provides new insights into their evolutionary origins. The nuclear receptor superfamily of transcription factors consists of over 100 different members from every metazoan organism. Membership in this elite club which reg-
Molecular Cell 220
Model of Nuclear Receptor Action (A) For nuclear receptors with known ligands, inactive receptor dimers bound to promoter regions recruit corepressor complexes which keep the transcription of the target gene turned off. Upon ligand (agonist) binding, the receptor undergoes a conformational change that puts the AF-2 in an active conformation which recruits coactivators and transcription ensues. Purple oval, DBD; blue rectangle/oval, LBD. (B) In contrast, the work by Wisely et al. (2002) shows that HNF4 constitutively binds fatty acids as structural cofactors, raising the question of how the switch is made from corepressor to coactivator complex.
ulates the transcription of genes involved in every aspect of physiology—from development to differentiation to disease—is dependent upon amino acid similarity in at least one of the two highly conserved domains: the DNA binding domain (DBD) or the ligand binding domain (LBD). The first members of the family were identified by binding to known ligands-steroids. Subsequent receptors were identified as genes that played functional roles in Drosophila or via low-stringency hybridization to the DBD or LBD of known receptors. With the exception of the steroid receptors, all the other receptors had to undergo a ligand hunt of varying duration before a ligand was identified. That hunt has taken more than a dozen years for HNF4, but as Wisely and coworkers report, the hunt may have finally come to an end. What they found, however, was not a traditional ligand. HNF4␣ was originally identified as a factor that bound a DNA response element critical to the expression of certain genes in the liver (Sladek et al., 1990). Therefore, in contrast to many other orphan receptors, HNF4␣ always had at least one ligand—the DNA elements to which it bound. These elements eventually led to the identification of more than 60 target genes which implicate HNF4␣ in a variety of human diseases, including atherosclerosis and cancer. Point mutations in the HNF4␣ binding sites of several blood coagulation factor genes have also linked HNF4␣ directly to certain types of hemophilia, while mutations in both the coding and the regulatory region of the HNF4␣ gene itself have been found in an inherited form of type II diabetes, MODY1 (Sladek and Seidel, 2001). Therefore, it is not too surprising that many labs worldwide, particularly those in the pharmaceutical industry, have been actively searching for a ligand for HNF4␣. After all, several very effective, and profitable, drugs that act on other nuclear receptors have been developed to treat diseases such as diabetes, cancer, inflammation, and osteoporosis, so why not HNF4␣? The problem has been that in every in vitro and in vivo system analyzed, HNF4␣ activates transcription in a constitutive fashion, i.e., in the absence of exogenously added ligand. Whereas there have been reports that fatty acyl CoA thioesters modulate the activity of HNF4␣ (Hertz et al., 1998), these compounds could not be shown to act as traditional ligands (Bogan et al., 2000).
The mantra in the nuclear receptor field is that bona fide ligands bind in a hydrophobic pocket in the LBD and that binding of ligand alters the position of one of 12 helices (the AF-2), thereby inducing a conformational change that kicks off corepressors and recruits coactivators (see Figure, panel A) (Glass and Rosenfeld, 2000). The fatty acyl CoA thioesters could not be shown to alter the conformation of HNF4␣, to decrease binding to a corepressor, or to increase binding to a coactivator (Bogan et al., 2000; Ruse et al., 2002). Now, the structure solved by the GlaxoSmithKline group suggests why. The pocket of bacterially expressed human HNF4␥ LBD contained a mixture of saturated and cis-monounsaturated C14-C18 fatty acids, in particular C16:0 (palmitic acid) and C16:1 (palmitoleic acid). (Although there is little functional data on HNF4␥, its LBD is 80% identical to that of HNF4␣.) This finding is perhaps not too surprising as other receptors expressed in bacteria have been found, upon crystallization, to bind fatty acids. However, the finding by Wisely et al. (2002) is unusual in that, short of completely denaturing the HNF4␥ (or HNF4␣) LBD, the fatty acid could not be removed from the protein. This and the fact that the purported ligand did not directly contact the AF-2 suggests that the fatty acids are not so much exchangeable ligands that induce allosteric changes in the receptor as they are essential cofactors that are critical to the structure of the protein, much like the zinc in zinc fingers. This work coincides very well with recent work showing that, both in vitro and in vivo, the switch between corepressor and coactivator binding to HNF4␣ can be made without changing the ligand status (Ruse et al., 2002) (see Figure, panel B). How then might HNF4’s activity be regulated? As Wisely et al. (2002) point out, posttranslational modification is one possibility. There are up to 13 potential phosphorylation sites in HNF4␣ (Jiang et al., 1997) as well as sites for acetylation (Soutoglou et al., 2000). Another possibility is that the nine naturally occurring splicing variations in HNF4␣ regulate its activity. The two most abundant forms in the adult liver have already been shown to differentially recruit coactivators (Sladek et al., 1999), and an embryo-specific form has been shown to possess distinct transactivation properties (TorresPadilla et al., 2001). Since the splicing variations are in the N- or the C terminus, both of which are implicated
Previews 221
in transactivation, one possibility is that these protein domains—and not a ligand—regulate the receptor’s function. Finally, there is the possibility that upon synthesis of the HNF4 protein, a different fatty acid can be introduced into the protein. This is not a completely farfetched idea, as an increasing number of both endogenous as well as exogenous compounds have been found to regulate the expression of the HNF4␣ gene (Sladek and Seidel, 2001). However, even if it can be shown that a ligand can be exchanged under some physiological, or therapeutic, condition, it must also be shown that different ligands have the ability to impart different activities on the protein. At this point, it is not clear how that would happen, since the AF-2 does not contact the ligand. Although, the Structure paper notes that the bacterially expressed HNF4␥ bound a fatty acid rarely found in E. coli (C17:1), suggesting that there might at least be a selection process for the ligand/cofactor. Finally, the findings of Wisely et al. (2002) may alter our thinking about the evolution of nuclear receptors. It was previously proposed that HNF4 and two other highly conserved family members were ancestral nuclear receptors that lacked ligands and that the ability to bind ligands arose independently more than once during evolution (Escriva et al., 2000). Now, however, one must consider the possibility that the original nuclear receptor did bind a lipophilic compound, but its role was as an integral component of the protein structure, not as an interchangeable element that induced allosteric and functional changes, as do the more modern ligands. The binding of a nonexchangeable cofactor makes it easier to understand conceptually why the LBD has been conserved during evolution and how the first true ligands arose, i.e., as a variation on a preexisting theme as opposed to a completely new characteristic. In any case, in this post-genomic era it is humbling to consider the possibility that perhaps at least part of what makes us
different from other organisms is not our genes or the proteins they encode but rather the ligands they bind. The author thanks M. Ruse and Y. Maeda for discussions and the National Institutes of Health for funding (R01 DK53892).
Cleave to Leave: Structural Insights into the Dynamic Organization of the Nuclear Pore Complex
The nuclear pore complex (NPC), weighing in at ⵑ100 MDa, is an example of such a large and dynamic supramolecular assembly. Located in the nuclear envelope, NPCs are the gatekeepers between the nucleus and cytoplasm. NPCs consist of approximately 30 different proteins, termed Nups. These multimerize to form an octagonal tube, from which fibers extend into both the cytoplasm and the nucleus (see Figure, panel B). Some soluble transport factors such as GTP-bound Ran power the process, and others such as karyopherins act as carriers, facilitating transport of their cargoes by interacting with particular sequence motifs in specific Nups. To understand its role as a transport machine, recent “holistic” approaches have studied the entire NPC or reconstituted significant constituent subcomplexes. In this issue of Molecular Cell, Hodel et al. (2002) have taken an alternative approach, showing how one can obtain important information about a large and dynamic complex like the NPC by instead looking in detail at key structural domains found in certain of its components. The approach Hodel et al. have taken includes an
A detailed understanding of the fine structure of the nuclear pore complex has remained elusive. Now, studies on a small protein domain have shed light on the dynamic organization of this massive assembly. In the new age of proteomics, determining the interactions between proteins in large, dynamic supramolecular assemblies is a major goal. In a cell, the number of interacting proteins is dauntingly large, with interactions changing depending on the subcellular localization of the partners or the state of the cell. Furthermore, varying affinities and expression levels often make detection of functional complexes a formidable challenge.
Frances M. Sladek Department of Cell Biology and Neuroscience University of California, Riverside Riverside, California 92521 Selected Reading Bogan, A.A., Dallas-Yang, Q., Ruse, M.D., Jr., Maeda, Y., Jiang, G., Nepomuceno, L., Scanlan, T.S., Cohen, F.E., and Sladek, F.M. (2000). J. Mol. Biol. 302, 831–851. Escriva, H., Delaunay, F., and Laudet, V. (2000). Bioessays 22, 717–727. Glass, C.K., and Rosenfeld, M.G. (2000). Genes Dev. 14, 121–141. Hertz, R., Magenheim, J., Berman, I., and Bar, T.J. (1998). Nature 392, 512–516. Jiang, G., Nepomuceno, L., Yang, Q., and Sladek, F.M. (1997). Arch. Biochem. Biophys. 340, 1–9. Ruse, M.D., Jr., Privalsky, M.L., and Sladek, F.M. (2002). Mol. Cell. Biol. 22, 1626–1638. Sladek, F.M., Zhong, W., Lai, E., and Darnell, J.E. (1990). Genes Dev. 4, 2353–2365. Sladek, F.M., Ruse, M.D., Nepomuceno, L., Huang, S.-M., and Stallcup, M.R. (1999). Mol. Cell. Biol. 19, 6509–6522. Sladek, F.M., and Seidel, S.D. (2001). In Nuclear Receptors and Genetic Diseases, T.P. Burris and E.R.B. McCabe, eds. (London: Academic Press), pp. 309–361. Soutoglou, E., Katrakili, N., and Talianidis, I. (2000). Mol. Cell 5, 745–751. Torres-Padilla, M.E., Fougere-Deschatrette, C., and Weiss, M.C. (2001). Mech. Dev. 109, 183–193. Wisely, G.B., Miller, A.B., Davis, R.G., Thornquest, A.D., Jr., Johnson, R., Spitzer, T., Sefler, A., Shearer, B., Moore, J.T., Miller, A.B., et al. (2002). Structure. Published online August 13, 2002. DOI:10.1016S0969212602008298.