New insights into the molecular mechanism of the Rab GTPase Sec4p activation
© Rinaldi et al. 2015
Received: 6 May 2015
Accepted: 8 July 2015
Published: 12 August 2015
Sec4p is a small monomeric Ras-related GTP-binding protein (23 kDa) that regulates polarized exocytosis in S. cerevisiae. In this study we examine the structural effects of a conserved serine residue in the P-loop corresponding to G12 in Ras.
We show that the Sec4p residue serine 29 forms a hydrogen bond with the nucleotide. Mutations of this residue have a different impact than equivalent mutations in Ras and can form stable associations with the exchange factor allowing us to elucidate the structure of a complex of Sec4p bound to the exchange factor Sec2p representing an early stage of the exchange reaction.
Our structural investigation of the Sec4p-Sec2p complex reveals the role of the Sec2p coiled-coil domain in facilitating the fast kinetics of the exchange reaction. For Ras-family GTPases, single point mutations that impact the signaling state of the molecule have been well described however less structural information is available for equivalent mutations in the case of Rab proteins. Understanding the structural properties of mutants such as the one described here, provides useful insights into unique aspects of Rab GTPase function.
Rab GTPases comprise the largest family of proteins among the small Ras-like GTPase superfamily functioning as critical regulators of a variety of membrane transport processes . Sec4p is among the 11 Rab GTPases identified in S. cerevisiae and acts on the surface of membranes to regulate transport between the Golgi apparatus and the plasma membrane [2–4]. The localization pattern of Sec4p depends upon its guanine nucleotide bound state. In the GTP bound state Sec4p exhibits an active conformation capable of interaction with targets on the plasma membrane. After GTP hydrolysis, the protein becomes inactive for effector recruitment and is then recycled from the plasma membrane and delivered back to the donor membrane. To complete the cycle, the protein is reactivated by switching the guanine nucleotide state from GDP to GTP in a process stimulated by guanine nucleotide exchange factors (GEFs). The GTPase cycle is controlled by interactions with regulators and effectors and is dependent on structural changes caused by the guanine nucleotide bound state of the protein [5–8]. Members of the GTPase family are known to share high overall structural homology. However, the differences in sequence identity of key domains known as Switch I, Switch II and the P-loop are sufficient to establish the interaction with specific regulators and effectors [5, 9–12]. Mutations in equivalent regions in Ras are commonly present in oncogenic cells and are known to perturb interactions with regulators leading to imbalance of the GTPase cycle. For example, Ras mutation Gly12Val, which is located in the P-loop region, confers reduced intrinsic GTPase activity and insensitivity to the action of GTPase Activating Protein (GAP) . As a result, this Ras variant preferentially remains in the GTP-bound conformation leading to constitutive signaling and cell transformation [14, 15].
In this study we investigate the activation mechanism of Sec4p by making use of the Ser29Val mutant. Both intrinsic and stimulated nucleotide exchange reactions of Sec4pV29 show decreased rates similar to those observed previously for Rab3a . In addition, we have solved the crystal structures of Sec4pV29 in the GDP-bound state and in complex with Sec2p in order to analyze the structural properties of the mutant. The structural refinement for the complex Sec4pV29.Sec2p reveals an intermediate state of the nucleotide dissociation reaction with the GDP bound to Sec4p that we postulate is the result of the reduced rate of GDP release from the mutant V29. We have also performed nucleotide exchange reactions using different truncations of Sec2p in order to determine the minimal region of Sec2p necessary for its GEF activity. We found that amino acids 142–160 of Sec2p are essential for full exchange activity, despite their distal location from Sec4p binding region. Even after much study, many aspects of Rab function remain enigmatic in terms of the structural mechanisms used for their signaling output. A complete understanding of the structural properties of Sec4p, together with functional analysis of interactions with Sec2p and other regulators provides important insights into the structure/function relationships of Rab proteins during the transport process.
Cloning expression and purification
Bacterial strains used in this study
Sec2p (51–182) mutant 153–157Ala
Sec2p (51–182) mutant 142–147Ala
Sec2p (51–182) mutant F109A
Sec4p (19–187) mutant E80A and R81A,
Sec2p (51–142) K140C,
Preparation of Sec4p.GDP.Sec2p complex
Sec4p 19 − 187 S29V and Sec2p51–142 were first purified separately and then a 1.8-fold molar excess of Sec2p51–142 was mixed with Sec4p 19 − 187 S29V in buffer containing 20 mM Tris-HCl pH8.0 100 NaCl, 5 mM MgCl2 for 2 h at 4 °C. After incubation, the protein mixture was concentrated and applied into a size exclusion column Superdex200 (10/300) (GE Healthcare) to isolate the protein complex. Complex formation was evaluated by SDS-PAGE.
Crystallization and data collection
Data collection and refinement statistics
Data collection statistics
116.900 119.280 122.860
a, b, c (Å)
31.720, 75.450, 66.230
Resolution range (Å)
No. reflections/no. unique reflections
Resolution range (Å)
no. of reflections (working data/test data)
Mean B value/Wilson B value (protein, all atoms) (Å2)
Mean B value (ligand) (Å2)
rmsd bond length (Å)
rmsd bond angle (°)
Crystallization trials for the ternary complex Sec4p 19 − 187 S29V .GDP.Sec2p51–142 were performed at a protein concentration of 15 mg/ml. Crystals were grown by mixing protein complex with a reservoir solution containing 24 % PEG3350 0.2 M Sodium Citrate in a 1:1 ratio and with the addition of 10 mM GDP to the drop. Crystals were soaked in a cryoprotectant solution containing the mother liquor (14 % PEG4000, 50 mM Zinc Acetate) supplemented with 20 % glycerol prior to flash-cooling in 100 K nitrogen gas stream. The crystallographic data was collected at the Brookhaven National Laboratories (BNL) National Synchrotron Light Source (NSLS). The crystal diffracted up to 2.9 Å resolution and belong to space group I222 with one ternary complex per asymmetric unit (% solvent).
Structural determination and refinement
All the data sets were processed using Mosflm  and Scala [20, 21]. The crystal structure for the mutant S29V of Sec4p was solved by molecular replacement with the program Phaser  using the wild-type structure of Sec4p.GDP as a model (accession code: 1G16) . The crystal structure for the ternary complex Sec4p 19 − 187 S29V .GDP.Sec2p51–142 was also solved with Phaser  using a previously solved structure of the complex Sec4p.Sec2p bound to a phosphate molecule (accession code: 2EQB) . The structure refinement was performed by alternating automatic refinement using Refmac , Phenix  and manual inspection using Coot .
Preparation of Sec4p.mantGDP and enzyme assays
In order to perform the nucleotide exchange experiments all the GTPases used in this work were preloaded with mantGDP. The procedure consists of incubating the protein sample (buffer: 20 mM Tris–HCl pH8.0 100 mM NaCl, 2 mM MgCl2) with a 2-fold excess of mantGDP over the protein concentration and 5 mM EDTA. The mixture was incubated at room temperature for 10 min and the reaction stopped with the addition of 10 mM MgCl2. The unbound nucleotide was separated from the sample using a spin column (Micro Bio-Spin 6, Bio-Rad) and the protein eluted in buffer containing 20 mM Tris-HCl pH8.0, 100 mM NaCl, 5 mM MgCl2. The protein concentration was measured by Bradford .
Fluorescence measurements were performed in buffer containing 20 mM Tris–HCl pH8.0 100 mM NaCl, 5 mM MgCl2 at room temperature. All the measurements were carried out in a QuantaMaster™ 40 fluorescence spectrometer (PTI). The mantGDP was excited at 360 nm and emission detected at 450 nm. For the intrinsic nucleotide exchange assay, time-based fluorescence was performed by collecting the fluorescence intensity in intervals specified in the plots. The data were normalized and plotted using Kaleidograph software (Synergy Software).
The fast kinetic experiments were performed in buffer containing 20 mM Tris–HCl pH8.0 100 mM NaCl, 5 mM MgCl2 at room temperature. The experiments were carried out using a stop flow apparatus coupled to a QuantaMaster™ 40 fluorescence spectrometer (PTI). The mantGDP was excited at 360 nm and emission was detected at 450 nm. The data were normalized and plotted using Kaleidograph software (Synergy Software).
In vivo assay for Sec2p function
In order to assess if various Sec2 constructs could act as the only copy of Sec2 in the cell transformants of a Sec2 tester strain were streaked to 5-FOA media to select for loss of the URA3 SEC2 plasmid. Yeast expressing wild type SEC2 can survive equivalently on the 5-FOA containing media, while a control plasmid with no insert cannot.
Availability of supporting data
The data set supporting the results of this article are available in the Protein Data Bank (PDB) repository Accession Codes 4ZDW and 4Z8Y.
Sec4pV29 is only partially activated by Sec2p
Crystal structure of Sec4pV29
Our data shows that the mutation S29V on Sec4p is responsible for the reduction in both intrinsic and stimulated nucleotide dissociation. This mutation does not appear to cause any gross conformational differences in the three-dimensional structure of the protein. Therefore the increased affinity between Sec4pV29 and the nucleotide is probably due to loss of intramolecular electrostatic interactions caused by the presence of the valine non-polar side chain. A similar issue has been observed previously for the oncogenic mutant of Ras GTPase.
Crystal structure of the ternary complex Sec4pV29.GDP.Sec2p
Structural and functional analysis of the coiled coil domain of Sec2p
Enhancement of dimerization with disulfide bond increases activity of the truncation Sec2p51–142
Discussion and Conclusions
In this work we have investigated the nucleotide exchange mechanism of the Rab GTPase Sec4p and how its GEF Sec2p stimulates nucleotide dissociation, making use of the Sec4p mutant S29V corresponding to Gly12 in Ras proteins. Ras Glycine 12 has been extensively linked to the formation of cancer cells caused by the imbalance in the GTPase cycle of Ras. Since mutations on the same residue in Rab GTPases have been suggested to display the opposite phenotype, mutating the Ser29 to Val in Sec4p offered us a tool to study Sec4p activation giving us insights into the basic differences between members of the Ras GTPase superfamily.
We have determined that the mutation S29V clearly reduces the intrinsic nucleotide dissociation rate of Sec4p in comparison to the wild type protein. In order to investigate the structural implications of the mutation S29V we solved the structure of Sec4pV29 in the GDP bound state. The structure of Sec4pV29.GDP presents a very similar fold to the wild type protein indicating that the decrease in the intrinsic nucleotide dissociation rate is not due to large conformational changes. This is similar to observations of equivalent mutations in other GTPase family members. The structures of the mutant G12V of p21ras and Cdc42 for example do not exhibit explicit conformational changes due to the G12V mutation when compared with the wild type equivalents [36–39].
The difference in nucleotide dissociation rate is most likely explained by changes in the protein-nucleotide environment created by the hydrophobic side chain of the Val29. The mutation from the polar residue serine to the just slightly larger hydrophobic valine creates a more hydrophobic local environment. We can predict that the energy cost to solvate the active site of the mutant Val29 would be larger than for the wild type protein and could indirectly result in the increased binding affinity of the protein for the nucleotide. Furthermore, Sec2p stimulates the nucleotide dissociation from Sec4pV29 at a slower rate than the rate observed for Sec4pwt. During activation Sec2p binds to Sec4p causing conformational changes leading to the loss of magnesium and decreased affinity for the nucleotide. For the mutant, an increase in the affinity for the nucleotide due to the presence of Val29 would decelerate nucleotide release and explain the observed decrease in the intrinsic nucleotide dissociation rate. Another interesting property of the mutant is that Sec2p can stably associate with the GDP bound state of Sec4pV29 even in the presence of magnesium. In contrast, Sec2p preferentially associates with the nucleotide free state of wild type Sec4p. We analyzed the crystal structure for the complex Sec4pV29 and Sec2p and determined that the complex was bound to a GDP molecule. This complex most likely represents the intermediate state of the nucleotide dissociation just before nucleotide release. Prior to this study two structures for the complex of Sec4p.Sec2p have been described. These structures represent the free form of the complex  and the phosphate bound form of the complex , and together explain the nucleotide exchange mechanism in molecular detail. Superposition between the two structures that represent the intermediate state of the reaction (Sec4pV29.GDP.Sec2p and Sec4pwt.phosphate.Sec2p) reveals that besides the overall similarities, a small shift in the P-loop results in a slightly tighter active site. The distance between the P-loop and Switch I, more specifically between the residues Val29 and Ile50 is shortened in comparison with the structure Sec4pwt.phosphate.Sec2p. The slight alteration in the active site, places the Ile50 residue very close to the magnesium binding site. It is thought that the Ile50 residue plays a critical role preventing magnesium rebinding to the protein after complex formation . In the free state of the complex Sec4p.Sec2p, the Ile50 residue is placed within 0.9 Å from the magnesium binding site. The shift in the P-loop observed in our structure would place the whole active site closer to the SWI and more specifically to the Ile50 residue. We can predict that this difference would make magnesium rebinding to the intermediate state of the complex more difficult and together with the increase in the affinity for the nucleotide it could explain why the GDP bound state of Sec4pV29 can form a stable complex with Sec2p.
During our nucleotide dissociation studies we discovered that the truncation Sec2p51–142 stimulates nucleotide release from Sec4p at a much slower rate than the truncation containing the 18 residues in region 142–160. This was somehow surprising because the Sec4p binding site of Sec2p lies between residues 100–120, and the absence of the region 142–160 would not directly affect the protein protein interaction. However, looking closely, this region possesses several leucine residues forming a classical coiled-coil packing interface that could be important for dimer formation and protein stability. Mutation of the leucine residues participating in the coiled coil interaction to alanines reduced protein activity, most likely through a disruption of the dimer coiled-coil structure.
To further validate the idea that disruption of dimerization causes reduced Sec2p activity, we created the mutant K140C on the truncation of Sec2p51–142. This truncation of Sec2p contains all the elements necessary for GEF activity however, has a slower activity than a protein which contains the dimer interface region between residues 142–160. The idea is that the formation of a disulfide linkage between the equivalent cysteine residues from both monomers would recapitulate dimerization in the absence of residues 142–160. The results of the nucleotide exchange assays show that in the absence of reducing agent the nucleotide dissociation rate for Sec2 S1 ‐ 142 K140C increased drastically in comparison to the wild type truncation Sec2p51–142, to a level similar to the truncation containing the coiled coil region 142–160 of Sec2p. These data indicates that although the region 142–160 is not directly involved in the binding to Sec4p this region is important for protein dimerization and indirectly affects protein activity. In addition, we determined that the first 50 residues in the Sec2p NH2-terminus do not affect GEF activity, but are important for full Sec2p function in vivo. This study extends the determination of a minimal GEF region for Sec4 and Sec2p interaction and establishes a minimal region of Sec2p containing the fully active enzyme, which we determined to be between residues 51–160. The intrinsic nucleotide dissociation rate of Sec4p is high relative to other Ras-related GTPases and the biological necessity for the potent GEF activity of Sec2p remains a mystery. A complete understanding of the mechanism of Sec2p action, together with functional analysis and interactions with other protein partners will provide important insights into the role of Sec4p and related Rab GTPases during exocytosis.
Guanine nucleotide exchange factor
GTPase activating protein
Phosphate binding loop
Protein Data Bank
We thank members of the Collins, Sondermann and Whittaker lab for helpful discussions. This work was supported by a grant from the National Institutes of Health (5R01GM069596) to R.N.C. We thank the staff at MacCHESS for help with data collection.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Hutagalung AH, Novick PJ. Role of Rab GTPases in membrane traffic and cell physiology. Physiol Rev. 2011;91:119–49.PubMed CentralPubMedView ArticleGoogle Scholar
- Novick P, Field C, Schekman R. Identification of 23 complementation groups required for post-translational events in the yeast secretory pathway. Cell. 1980;21:205–15.PubMedView ArticleGoogle Scholar
- Salminen A, Novick PJ. A ras-like protein is required for a post-Golgi event in yeast secretion. Cell. 1987;49:527–38.PubMedView ArticleGoogle Scholar
- Novick P, Ferro S, Schekman R. Order of events in the yeast secretory pathway. Cell. 1981;25:461–9.PubMedView ArticleGoogle Scholar
- Dong G, Medkova M, Novick P, Reinisch KM. A catalytic coiled coil: structural insights into the activation of the Rab GTPase Sec4p by Sec2p. Mol Cell. 2007;25:455–62.PubMed CentralPubMedView ArticleGoogle Scholar
- Ignatev A, Kravchenko S, Rak A, Goody RS, Pylypenko O. A structural model of the GDP dissociation inhibitor rab membrane extraction mechanism. J Biol Chem. 2008;283:18377–84.PubMedView ArticleGoogle Scholar
- Sato Y, Fukai S, Ishitani R, Nureki O. Crystal structure of the Sec4p.Sec2p complex in the nucleotide exchanging intermediate state. Proc Natl Acad Sci U S A. 2007;104:8305–10.PubMed CentralPubMedView ArticleGoogle Scholar
- Stroupe C, Brunger AT. Crystal structures of a Rab protein in its inactive and active conformations. J Mol Biol. 2000;304:585–98.PubMedView ArticleGoogle Scholar
- Brondyk WH, McKiernan CJ, Burstein ES, Macara IG. Mutants of Rab3A analogous to oncogenic Ras mutants. Sensitivity to Rab3A-GTPase activating protein and Rab3A-guanine nucleotide releasing factor. J Biol Chem. 1993;268:9410–5.PubMedGoogle Scholar
- Eathiraj S, Pan X, Ritacco C, Lambright DG. Structural basis of family-wide Rab GTPase recognition by rabenosyn-5. Nature. 2005;436:415–9.PubMed CentralPubMedView ArticleGoogle Scholar
- Itzen A, Pylypenko O, Goody RS, Alexandrov K, Rak A. Nucleotide exchange via local protein unfolding--structure of Rab8 in complex with MSS4. EMBO J. 2006;25:1445–55.PubMed CentralPubMedView ArticleGoogle Scholar
- Ostermeier C, Brunger AT. Structural basis of Rab effector specificity: crystal structure of the small G protein Rab3A complexed with the effector domain of rabphilin-3A. Cell. 1999;96:363–74.PubMedView ArticleGoogle Scholar
- Scheffzek K, Ahmadian MR, Kabsch W, Wiesmuller L, Lautwein A, Schmitz F, et al. The Ras-RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants. Science. 1997;277:333–8.PubMedView ArticleGoogle Scholar
- Seeburg PH, Colby WW, Capon DJ, Goeddel DV, Levinson AD. Biological properties of human c-Ha-ras1 genes mutated at codon 12. Nature. 1984;312:71–5.PubMedView ArticleGoogle Scholar
- Barbacid M. ras genes. Annu Rev Biochem. 1987;56:779–827.PubMedView ArticleGoogle Scholar
- Li G, Liang Z. Phosphate-binding loop and Rab GTPase function: mutations at Ser29 and Ala30 of Rab5 lead to loss-of-function as well as gain-of-function phenotype. Biochem J. 2001;355:681–9.PubMed CentralPubMedView ArticleGoogle Scholar
- Sato Y, Shirakawa R, Horiuchi H, Dohmae N, Fukai S, Nureki O. Asymmetric coiled-coil structure with Guanine nucleotide exchange activity. Structure. 2007;15:245–52.PubMedView ArticleGoogle Scholar
- Matthews BW. Solvent content of protein crystals. J Mol Biol. 1968;33:491–7.PubMedView ArticleGoogle Scholar
- Leslie AGW, Powell HR. Processing diffraction data with MOSFLM. Evolv Methods Macromol Crystallogr. 2007;245:41–51.View ArticleGoogle Scholar
- Evans P. Scaling and assessment of data quality. Acta Crystallogr D Biol Crystallogr. 2006;62:72–82.PubMedView ArticleGoogle Scholar
- Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR, et al. Overview of the CCP4 suite and current developments. Acta Crystallogr D Biol Crystallogr. 2011;67:235–42.PubMed CentralPubMedView ArticleGoogle Scholar
- McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007;40:658–74.PubMed CentralPubMedView ArticleGoogle Scholar
- Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr. 1997;53:240–55.PubMedView ArticleGoogle Scholar
- Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr. 2010;66:213–21.PubMed CentralPubMedView ArticleGoogle Scholar
- Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004;60:2126–32.PubMedView ArticleGoogle Scholar
- Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;72:248–54.PubMedView ArticleGoogle Scholar
- Itzen A, Rak A, Goody RS. Sec2 is a highly efficient exchange factor for the Rab protein Sec4. J Mol Biol. 2007;365:1359–67.PubMedView ArticleGoogle Scholar
- John J, Sohmen R, Feuerstein J, Linke R, Wittinghofer A, Goody RS. Kinetics of interaction of nucleotides with nucleotide-free H-ras p21. Biochemistry. 1990;29:6058–65.PubMedView ArticleGoogle Scholar
- Neal SE, Eccleston JF, Hall A, Webb MR. Kinetic analysis of the hydrolysis of GTP by p21N-ras. The basal GTPase mechanism. J Biol Chem. 1988;263:19718–22.PubMedGoogle Scholar
- Walch-Solimena C, Collins RN, Novick PJ. Sec2p mediates nucleotide exchange on Sec4p and is involved in polarized delivery of post-Golgi vesicles. J Cell Biol. 1997;137:1495–509.PubMed CentralPubMedView ArticleGoogle Scholar
- Klebe C, Prinz H, Wittinghofer A, Goody RS. The kinetic mechanism of Ran--nucleotide exchange catalyzed by RCC1. Biochemistry. 1995;34:12543–52.PubMedView ArticleGoogle Scholar
- Goldberg J. Structural basis for activation of ARF GTPase: mechanisms of guanine nucleotide exchange and GTP-myristoyl switching. Cell. 1998;95:237–48.PubMedView ArticleGoogle Scholar
- Renault L, Kuhlmann J, Henkel A, Wittinghofer A. Structural basis for guanine nucleotide exchange on Ran by the regulator of chromosome condensation (RCC1). Cell. 2001;105:245–55.PubMedView ArticleGoogle Scholar
- Worthylake DK, Rossman KL, Sondek J. Crystal structure of Rac1 in complex with the guanine nucleotide exchange region of Tiam1. Nature. 2000;408:682–8.PubMedView ArticleGoogle Scholar
- Lupas AN, Gruber M. The structure of alpha-helical coiled coils. Adv Protein Chem. 2005;70:37–78.PubMedView ArticleGoogle Scholar
- Krengel U, Schlichting I, Scherer A, Schumann R, Frech M, John J, et al. Three-dimensional structures of H-ras p21 mutants: molecular basis for their inability to function as signal switch molecules. Cell. 1990;62:539–48.PubMedView ArticleGoogle Scholar
- Rudolph MG, Wittinghofer A, Vetter IR. Nucleotide binding to the G12V-mutant of Cdc42 investigated by X-ray diffraction and fluorescence spectroscopy: two different nucleotide states in one crystal. Protein Sci. 1999;8:778–87.PubMed CentralPubMedView ArticleGoogle Scholar
- Tong LA, de Vos AM, Milburn MV, Kim SH. Crystal structures at 2.2 a resolution of the catalytic domains of normal ras protein and an oncogenic mutant complexed with GDP. J Mol Biol. 1991;217:503–16.PubMedView ArticleGoogle Scholar
- Wittinghofer A, Franken SM, Scheidig AJ, Rensland H, Lautwein A, Pai EF, et al. Three-dimensional structure and properties of wild-type and mutant H-ras-encoded p21. Ciba Found Symp. 1993;176:6–21. discussion 21–7.PubMedGoogle Scholar
- Schultz J, Milpetz F, Bork P, Ponting CP. SMART, a simple modular architecture research tool: identification of signaling domains. Proc Natl Acad Sci U S A. 1998;95:5857–64.PubMed CentralPubMedView ArticleGoogle Scholar