Search results

NOTES & TIPS Evaluation of the Efficiency of In-Gel Digestion of Proteins by Peptide Isotopic Labeling and MALDI Mass Spectrometry Anna Shevchenko and Andrej Shevchenko 1 MPI of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, 01307 Dresden, Germany; and European Molecular Biology Laboratory (EMBL), 69012 Heidelberg, Germany Received December 21, 2000; published online August 15, 2001 Visualization of proteins separated by one-dimensional or two-dimensional polyacrylamide gel electrophoresis, in-gel digestion of excised protein bands (spots), followed by identification of proteins by mass spectrometry underpin many proteome characterization strategies (reviewed in (1 4)). A multitude of protocols for staining of polyacrylamide gels (reviewed in (5)) and of enzymatic in-gel digestion of proteins (reviewed in (1, 6)) at the low picomole femtomole level has been reported. A combination of protein visualization and in-gel digestion methods designed for proteomic application is usually evaluated according to two major criteria. First, the lowest limit of reliable visualization of protein spot (band) is determined (5, 7, 8). Second, the sequence coverage of MALDI peptide mass fingerprints acquired from in-gel digests of the visualized protein spots and the sensitivity of mass spectrometric detection are evaluated (8 10). However, the latter criteria are particularly difficult to apply. The intensity of peptide signals detected in complex mixtures by MALDI MS strongly depends on the employed sample handling and probe preparation routines. Furthermore, unavoidable micro-heterogeneity of MALDI probes and inevitable presence of the residual amount of dyes, salts, and detergents result in significant shot-to-shot variation of peptide and matrix signals, requires consistent tuning of the laser fluence (11) and, consequently, does not allow quantitative comparison of spectra acquired in the separate experiments. Without direct measurement of the yield of peptide digestion products it is difficult to provide consistent evaluation of the efficiency of in-gel digestion protocols. Sequencing of proteins from silver-stained gels may serve as an example. Successful identification of silver stained proteins and high sequence coverage of MALDI peptide mass maps at the low femtomole level was reported (12), and almost identical peptide mass fingerprints were observed in the digests of silver and Coomassie stained protein bands (13). However, other authors reported substantially lower sequence coverage of peptide mass maps acquired from the digests of silver-stained proteins (8, 10, 14) that, however, could be improved by destaining of the bands prior to in-gel trypsinolysis (15, 16). We therefore set out to develop a method for direct and quantitative evaluation of the efficiency of in-gel cleavage of proteins in order to outline a rational procedure for comparison of in-gel digestion efficiency. The yield of digestion products was determined by MALDI MS using 18 O-isotopically labeled peptides as the internal standards (17, 18). We further examined whether conventional methods of staining of polyacrylamide gels (Coomassie staining, silver staining, and zinc-imidazole staining) might affect the in-gel digestion efficiency. Materials and Methods Materials and reagents. All major chemicals were purchased from Sigma (Sigma Chemicals, St. Louis, MO) and were of analytical grade. H 2 18 O (Cambridge Isotopic Laboratories, MA) was purified by microdistillation as described (19). Gel electrophoresis and visualization of protein bands. The aliquots containing 1 pmol of bovine serum albumin (Sigma Chemicals) in Laemmli buffer were loaded onto separate lanes of a one-dimensional polyacrylamide gel. Immediately after electrophoresis, the gel was cut. Separate parts of the gel slab each containing three lanes with the BSA standards were stained by various methods. Coomassie staining and silver staining were performed as described (13). Zinc imidazole staining (negative staining) was performed according to (20). In a separate experiment the gel was stained with silver. Four BSA bands were excised from the gel and then destained with potassium ferricyanide and sodium thiosulfate as described (15). Stained and Preliminary results were reported at the 48th ASMS Conference on Mass Spectrometry and Allied Topics, Long Beach, CA, 2000. 1 Corresponding author: E-mail: shevchenko@mpi-cbg.de. Analytical Biochemistry 296, 279 283 (2001) 279 doi:10.1006/abio.2001.5321 0003-2697/01 $35.00 Copyright 2001 by Academic Press All rights of reproduction in any form reserved. All articles available online at http://www.idealibrary.com on destained bands were further processed in parallel using conventional recipe. In-gel digestion, preparation of sample probes and MALDI analysis. To prepare a standard mixture of 18 O-labeled peptides, a solution of 0.14 pmol/ L BSA in 25 mM ammonium bicarbonate buffer in H 2 18 O was digested overnight at 37 C; enzyme:substrate ratio 1:10 (w/w). Protein bands (three for each method of staining) were in parallel in-gel digested with trypsin (unmodified, sequencing grade, Roche Diagnostics GmbH, Germany) as described (13, 19). Gel pieces were extracted with 5% formic acid and acetonitrile and the extracts were dried down in a vacuum concentrator. Tryptic peptides were redissolved in 10 L of 10% formic acid. A 2- L aliquot was withdrawn and mixed with 1 L of an 18 O-labeled mixture of peptides (internal standard) prepared as described above. Four 0.5- L aliquots of every mixed sample were analyzed in parallel by MALDI MS as described in (13, 21) on a modified REFLEX III mass spectrometer (Bruker Daltonics, Germany). The determined relative concentrations of peptides were averaged. The relative standard deviation of the concentrations in all series of measurements was better than 20%. Results and Discussion Quantification of peptides in in-gel tryptic digests. Upon digesting of a protein in the buffer, which contains H 2 18 O, tryptic peptides incorporate one or two 18 O-atoms into their C-terminal carboxyl groups (22). Comparison of the peptide mass maps acquired from the digests of various standard proteins revealed that peptides, which contain arginine residue at their Cterminus incorporate two 18 O atoms (2 18 O peptides) mostly, whereas peptides having C-terminal lysine residue incorporate one 18 O atom (1 18 O peptides) (Fig. 1). Incubation of 2 18 O peptides in 10% formic acid in H 2 16 O at room temperature resulted in a mixture of unlabeled, 1 18 O and 2 18 O forms. However, this process was slow and required several days before substantial alteration of the isotopic profile was detected (data not shown). Isotopically labeled peptides produced by digesting of a protein in H 2 18 O could be employed as internal standards for quantitative measurements by MALDI MS. A standard mixture of 18 O-labeled peptides was prepared by digesting BSA with trypsin in solution in the buffer containing H 2 18 O. Very similar profiles of tryptic peptides were detected in MALDI peptide FIG. 1. A part of the spectrum of the tryptic digest of BSA in the buffer containing H 2 18 O. Peaks in the spectrum are designated with corresponding peptide sequences and m/z calculated for the unlabeled monoisotopic ions. Blowouts demonstrate isotopic profiles typical for the peptide ions having arginine or lysine residues at their C-termini. The positions of the corresponding monoisotopic unlabeled ions are designated with unfilled arrows. 280 NOTES & TIPS maps of in-solution digests and of in-gel digests, although the relative intensity of peptide peaks was altered. Equal volume of the mixture of 18 O-labeled peptides was spiked into the aliquots withdrawn from the experimental in-gel digests that were performed in H 2 16 O and the samples were analyzed by MALDI MS. Relative concentration of digestion products was calculated as a ratio of the intensity of the monoisotopic peak of the unlabeled peptide and the intensity of the monoisotopic peak of the corresponding 2 18 O peptide standard (Fig. 2). Linearity of the calibration curve was tested by analyzing the series of samples obtained by successive diluting of the aliquot withdrawn from the in-gel digest of 1 pmol of BSA. The relative concentrations calculated for various peptides were found linear over 1:5 dilution range and were affected by chemical noise at larger dilution ratios (data not shown). The effect of gel staining on the recovery of tryptic peptides. This was examined by analyzing in-gel digests of the bands containing 1 pmol of BSA, which were stained with Coomassie, silver, and zinc-imidazole. A similar profile of tryptic peptides was detected TABLE 1 Relative Concentration of Tryptic Peptides of BSA in In-Gel Digests of Bands Stained by Various Methods Staining method Relative concentration a (%) m/z 927.49 m/z 1439.81 m/z 1479.80 m/z 1567.74 m/z 1639.94 Silver, with reduction and alkylation 100 100 100 100 100 Coomassie 105 136 74 95 121 Zn/Imidazole 117 61 74 79 100 a Relative concentrations of peptides were normalized to the concentrations in the digests of silver-stained bands. FIG. 2. Calculation of the relative concentration of peptides. A blowout of the isotopic cluster of the peptide peak with m/z 1439.93. The monoisotopic peak of the unlabeled peptide is designated with a filled arrow. The peak of the isotopicaly labeled peptide, which incorporated two 18 O atoms (2 18 O) was used as an internal standard. The relative concentration of the unlabeled peptide (R c ) was calculated as R c I p /I st , where I p stands for the intensity of the peptide peak and I st stands for the intensity of the peak of the standard. 281 NOTES & TIPS in each of those samples (data not shown) and relative concentrations determined for five most intense peptide ions were compared. We observed slight variation of the relative concentration of individual peptides. However, no one method of staining provided significantly better recovery of peptides compared to other methods in the test (Table 1). We further tested whether the recovery of peptides from silver stained gels could be improved by destaining of protein bands prior to in-gel digestion (15). To this end we compared the relative concentrations of peptides in the in-gel digests of the destained bands and of the bands treated according to the conventional protocol. We observed no significant increase in the number of detected peptide peaks as well as in the yield of peptides if destaining of bands was applied (Table 2). Similar conclusion was reached by Moertz et al. (12) on the basis of MALDI analysis of a large number of automatically processed samples. Reduction and alkylation steps did not influence the recovery of peptides, which do not contain cysteine residues and therefore for the purpose of protein identification those steps could, in principle, be omitted (23). Notably cysteine-containing peptides were missing if reduction and alkylation steps were skipped (Fig. 3). We therefore concluded that at the level of 1 pmol of protein starting material the method of protein visualization FIG. 3. Comparison of the peptide maps of the silver stained bands processed using destaining, reduction, and alkylation (the upper spectrum) and using only destaining (the lower spectrum). Peaks are designated with corresponding m/z; peptide sequences are presented in Fig. 1. Three intense peptide peaks (designated with unfilled arrows) having matching cysteine-containig peptides from BSA were additionally detected after reduction and alkylation. Corresponding peptide sequences are: m/z 1419.66 SLHTLFGDELCK; m/z 1539.79 LCVLHEKTPVSEK; m/z 1880.91 RPCFSALTPDETYVPK. C stands for cysteine-S-acetamide residue. TABLE 2 Relative Concentration of Peptides Recovered from Silver-Stained Gels Sample preparation method Relative concentration a (%) m/z 927.49 m/z 1439.81 m/z 1479.80 m/z 1567.74 m/z 1639.94 With destaining, reduction and alkylation 100 100 100 100 100 With destaining only 95 110 93 97 98 With reduction and alkylation 115 104 102 83 94 Untreated b 5 12 14 6 19 a Relative concentrations of peptides were normalized to the concentrations in the digests of destained, reduced and alkylated bands. b Predigestion washing, destaining, reduction and alkylation were skipped. 282 NOTES & TIPS does not have any noticeable impact on the recovery of tryptic peptides. We also observed that relative concentration of all peptides in the digests of silver-stained bands, which were directly treated with trypsin (i.e. washing steps as well as destaining, reduction and alkylation were omitted), was dramatically lower. Nevertheless, the number of detected peptide was always sufficient for unambiguous identification of BSA upon searching a database. We therefore speculate that the sequence coverage of MALDI peptide maps alone does not constitute an adequate measure of the digestion efficiency and should be complemented by direct quantification of peptide products. Thus we have demonstrated that application of isotopically labeled peptide standards and MALDI MS enabled direct and quantitative evaluation of the efficiency of in-gel digestion. The method paves the way for further optimization of sample processing routines, thus improving sensitivity and throughput of the characterization of proteomes by mass spectrometry. Acknowledgments. The authors are grateful for members of Protein and Peptide Group for experimental support and useful discussions. REFERENCES 1. Lahm, H. W., and Langen, H. (2000) Electrophoresis 21, 2105 2114. 2. Pandey, A., and Mann, M. (2000) Nature 405, 837 846. 3. Andersen, J. S., and Mann, M. (2000) FEBS Lett. 480, 25 31. 4. Anderson, N. L., Matheson, A. D., and Steiner, S. (2000) Curr. Opin. Biotechnol. 11, 408 412. 5. Rabilloud, T. (2000) Anal Chem. 72, 48A 55A. 6. Patterson, S. D., and Aebersold, R. (1995) Electrophoresis 16, 791 814. 7. Rabilloud, T. (1990) Electrophoresis 11, 785 794. 8. Lopez, M. F., Berggren, K., Chernokalskaya, E., Lazarev, A., Robinson, M., and Patton, W. F. (2000) Electrophoresis 21, 3673 3683. 9. Yan, J. X., Wait, R., Berkelman, T., Harry, R. A., Westbrook, J. A., Wheeler, C. H., and Dunn, M. J. (2000) Electrophoresis 21, 3666 3672. 10. Lauber, W. M., Carrol, J. A., Dunfield, D. R., Kiesel, J. R., Radabaugh, M. R., and Malone, J. P. (2001) Electrophoresis 22, 906 918. 11. Jensen, O. N., Mortensen, P., Vorm, O., and Mann, M. (1997) Anal. Chem. 69, 1706 1714. 12. Moertz, E., Krogh, T. N., Vorum, H., and Go rg, A. (2000) Proceedings, 48th ASMS Conference on Mass Spectrometry and Allied Topics, Long Beach CA, pp. 1115 1116. 13. Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996) Anal. Chem. 68, 850 858. 14. Scheler, C., Lamer, S., Pan, Z., Li, X. P., Salnikow, J., and Jungblut, P. (1998) Electrophoresis 19, 918 27. 15. Gharahdaghi, F., Weinberg, C. R., Meagher, D. A., Imai, B. S., and Mische, S. M. (1999) Electrophoresis 20, 601 605. 16. Sumner, L. W., White, S., Wolf-Sumner, B., and Asirvatham, V. S. (2001) Abstracts 49th ASMS Conference on Mass Spectrometry and Allied Topics, Chicago IL. 17. Shevchenko, A., Wilm, M., and Shevchenko, A. (2000) Proceedings, 48th ASMS Conference on Mass Spectrometry and Allied Topics, Long Beach CA, pp. 859 860. 18. Mirgorodskaya, O. A., Kozmin, Y. P., Titov, M. I., Korner, R., Sonksen, C. P., and Roepstorff, P. (2000) Rapid Commun. Mass Spectrom. 14, 1226 1232. 19. Shevchenko, A., Chernushevich, I., Wilm, M., and Mann, M. (2000) in Protein in Peptide Analysis (Chapman, J. R., Ed.), Vol. 146, pp. 1 16, Humana Press, Totowa, NJ. 20. Fernandez-Patron, C., Calero, M., Collazo, P. R., Garcia, J. R., Madrazo, J., Musacchio, A., Soriano, F., Estrada, R., Frank, R., and Castellanos-Serra, L. R. (1995) Anal. Biochem. 224, 203 211. 21. Jensen, O. N., Podtelejnikov, A., and Mann, M. (1996) Rapid Commun. Mass Spectrom. 10, 1371 1378. 22. Schno lzer, M., Jedrzejewski, P., and Lehmann, W. D. (1996) Electrophoresis 17, 945 953. 23. Borchers, C., Peter, J. F., Hall, M. C., Kunkel, T. A., and Tomer, K. B. (2000) Anal. Chem. 72, 1163 1168. Reutilization of Immunoblots after Chemiluminescent Detection Scott H. Kaufmann Division of Oncology Research, Mayo Clinic, and Department of Molecular Pharmacology, Mayo Graduate School, Rochester, Minnesota 55905 Received March 9, 2001; published online August 16, 2001 Immunoblotting is widely utilized to evaluate the presence of antigens of interest in various biological samples, monitor antigen purification, assess epitope retention after antigen degradation, or assay for the presence of antibodies of a particular specificity in biological fluids [reviewed in Refs. (1 4)]. Under certain circumstances, e.g., if a blot suggests an unexpected difference in antigen expression between two samples or the antigens being analyzed are derived from a precious source, it can be important to sequentially probe the same blot for the presence of multiple antigens. The present study demonstrates that treatment with sodium azide after detection of bound HRP 1 -coupled secondary antibodies results in inhibition of the reporter group, thereby facilitating sequential probing of blots if reagents raised in multiple species are available. A number of approaches have been previously proposed for the detection of multiple antigens on immu-1 Abbreviations used: HRP, horseradish peroxidase; PBS, calciumand magnesium-free phosphate-buffered saline. 283 NOTES & TIPS Analytical Biochemistry 296, 283 286 (2001) doi:10.1006/abio.2001.5313 0003-2697/01 $35.00 Copyright 2001 by Academic Press All rights of reproduction in any form reserved.

Synthesis of adenosine triphosphate (ATP) by the F(1)F(0) ATP synthase involves a membrane-embedded rotary engine, the F(0) domain, which drives the extra-membranous catalytic F(1) domain. The F(0) domain consists of subunits a(1)b(2) and a cylindrical rotor assembled from 9-14 alpha-helical hairpin-shaped c-subunits. According to structural analyses, rotors contain 10 c-subunits in yeast and 14 in chloroplast ATP synthases. We determined the rotor stoichiometry of Ilyobacter tartaricus ATP synthase by atomic force microscopy and cryo-electron microscopy, and show the cylindrical sodium-driven rotor to comprise 11 c-subunits.

The lactose transport protein (LacS) of Streptococcus thermophilus belongs to a family of transporters in which putative alpha-helices II and IV have been implicated in cation binding and the coupled transport of the substrate and the cation. Here, the analysis of site-directed mutants shows that a positive and negative charge at positions 64 and 71 in helix II are essential for transport, but not for lactose binding. The conservation of charge/side-chain properties is less critical for Glu-67 and Ile-70 in helix II, and Asp-133 and Lys-139 in helix IV, but these residues are important for the coupled transport of lactose together with a proton. The analysis of second-site suppressor mutants indicates an ion pair exists between helices II and IV, and thus a close approximation of these helices can be made. The second-site suppressor analysis also suggests ion pairing between helix II and the intracellular loops 6-7 and 10-11. Because the C-terminal region of the transmembrane domain, especially helix XI and loop 10-11, is important for substrate binding in this family of proteins, we propose that sugar and proton binding and translocation are performed by the joint action of these regions in the protein. Indeed, substrate protection of maleimide labeling of single cysteine mutants confirms that alpha-helices II and IV are directly interacting or at least conformationally involved in sugar binding and/or translocation. On the basis of new and published data, we reason that the helices II, IV, VII, X, and XI and the intracellular loops 6-7 and 10-11 are in close proximity and form the binding sites and/or the translocation pathway in the transporters of the galactosides-pentosides-hexuronides family.

C O M M E N T Evolution and homology of the nervous system: cross-phylum rescues of otd/Otx genes ANNA C . SHARMAN AND MICHAEL BRAND anna.sharman@urz.uni-heidelberg.de brand@sun0.urz.uni-heidelberg.de I NSTITUTE OF N EUROBIOLOGY , U NIVERSITY OF H EIDELBERG , I N N EUENHEIMER F ELD 345, 69120 H EIDELBERG , G ERMANY . The homeotic or Hox genes were the first gene family to be shown to act in similar, probably homologous, ways in insect and mammalian development. They are thought to form a combinatorial code specifying the identity of different segments. They are clustered in the genome, and expressed in a nested pattern along the anterior posterior axis in an order that is colinear with their chromosomal order 1 . Since then, perhaps a surprisingly high number of developmental gene families have also been shown to have conserved expression patterns over several phyla. One such family is the Otx genes, which like the Hox genes encode homeodomain-containing DNA-binding proteins 2,3 . Flies and amphioxus have a single gene in the family, called orthodenticle (otd ) in Drosophila, while vertebrates typically have two Otx genes. Mouse Otx1 and Otx2, together with another family of homeobox genes, Emx1 and Emx2, are expressed in nested patterns in the fore- and mid-brain 3 , while their fly homologues otd and ems are expressed in the most anterior segments 4 6 . It has been suggested 3,7 that the Otx and Emx genes specify segmental identity in insects and vertebrates, fulfilling a role for the anterior brain similar to that of the Hox genes in the hindbrain and spinal cord. This hypothesis is supported by knockout phenotypes and mutations 8 12 : in Otx2 +/ Otx1 / mice the midbrain and posterior diencephalon are completely missing (interpretation of single Otx mutant phenotypes is complicated in mice by variability, an early role for Otx2 in gastrulation and by redundancy between Otx1 and Otx2). It also appears that in mammals, Otx function influences development of the midbrain hindbrain boundary, known to be an organizer of cell fate in the midbrain and anterior hindbrain 8 13 . In Emx knockouts discrete parts of the forebrain are missing 14,15 . So expression and mutation of Otx/otd and Hox genes show that the anterior posterior (AP) patterning mechanisms in insects and vertebrates share many features. Does this mean that these mechanisms are homologous, that is, that the common ancestor had these mechanisms too? The fact that amphioxus, a primitive chordate, also has anterior Otx expression 2 indicates that the mechanism is quite widespread, although otd expression in echinoderms 16 is quite different and variable. It is first of all necessary to test whether the genes are acting in the same way in both groups of animals. This can be done experimentally by expressing vertebrate genes in flies mutant for their homologues, or vice versa, and asking whether the foreign gene can rescue the mutant phenotype. This has already been done for some Hox genes 17 19 , showing that they can function in an equivalent manner, but it has not until now been tried for the Otx/otd family. A second factor also needs to be considered. The nervous system of vertebrates has been compared to the early ectoderm of insects, at a stage long before the central nervous system (CNS) is formed. A question mark has hung over interpretations of AP patterning similarities: could these two systems really be homologous, when they referred to such different stages of development? Two responses have been given towards the latter question. One (called the auricularia hypothesis) was to propose that the nervous system of chordates is homologous to the entire outer ectoderm of insects while the insect nervous system and the chordate outer ectoderm have arisen independently 20 . This was supported by studies of the anatomy of larval echinoderms (particularly auricularia larvae) and urochordates, which are proposed to be intermediates in the transition between the arthropodand chordate-type body plans, and by the fact that insect Hox genes are expressed in the surface ectoderm, whereas they are not in vertebrates. The nervous systems of insects and vertebrates have long been assumed to have evolved independently, because they are organized differently and are on opposite sides of the embryo 21 . The other response grew out of recent work showing that the dorsoventral axis is in fact patterned by the same set of genes in similar relationships in insects and in vertebrates, but with the dorsoventral axis inverted 21,22 . Many developmental genes, such as the achaete/scute homologues, Nkx2, Msx and Hox homeobox genes and netrins are expressed in similar patterns in insect and vertebrate nervous systems 21,23 . It is a reasonable hypothesis that the dorsoventral axis inverted at some point in evolution, and thus that the nervous systems of insects and vertebrates are homologous (Fig. 1). Which of these two theories is correct? Four recent papers 24 27 address the question of Otx functional equivalence, and supply further evidence for homology of the insect and vertebrate nervous systems. Three of them 24 26 show that the Otx and otd genes have a conserved function, as well as conserved expression patterns. Previously, the expression of otd and ems in the fly nervous system was determined (Ref. 6 and Fig. 2), and the fourth paper 27 finds that Hox gene expression there is more similar to expression of vertebrate Hox genes than was previously suspected. Functional equivalence of Otx and otd genes Two papers describe the overexpression of human OTX1 and OTX2 genes in flies defective in otd function. Nagao et al. 26 found that the human OTX genes could rescue the ocelliless phenotype (oc is a regulatory mutation causing defective otd expression in the fly pupa) just as well as otd could rescue. Not only the final phenotype, but also gene expression TIG J UNE 1998 V OL . 14 N O . 6 211 Copyright 1998 Elsevier Science Ltd. All rights reserved. 0168-9525/98/$19.00 PII: S0168-9525(98)01488-7 C O M M E N T in the eye-antennal imaginal discs, is rescued. Leuzinger et al. 25 used a similar approach to drive transient ubiquitous expression of otd, OTX1 or OTX2, to look at the better-studied embryonic function of otd. According to several criteria, otd expression can rescue the otd mutant phenotype, though in a variable proportion of embryos. OTX2 also rescues, though at a lower frequency than otd, and OTX1 rescues less efficiently still. The third paper 24 goes in the other direction: the mouse Otx1 coding region was replaced by that of Drosophila otd. Many of the defects of Otx1 / mice were rescued by otd, for example several brain regions previously missing are restored. Interestingly, the rescue was less efficient nearer to the MHB, for example the mesencephalon was never completely normal, and cerebellar foliation remained abnormal. In the sense organs, most defects were rescued, but the lateral semicircular canal in the inner ear was not. Hox genes in the fly nervous system Hirth et al. 27 examine the expression and function of the Drosophila homeotic (Hox) genes in the CNS. In the mouse, the group 2 Hox genes have the most anterior expression boundaries, while group 1 Hox genes are expressed in restricted domains, one rhombomere posterior to the group 2 anterior boundary (Ref. 1 and Fig. 2). This lapse in the rule of colinearity has so far only been noted in vertebrates 28 ; in the Drosophila ectoderm, strict colinearity is adhered to. Previous descriptions of homeotic gene expression in the fly nervous system put the anterior boundary of labial (lab ; a group 1 Hox gene) and proboscipedia (pb ; group 2) at the same level, within the deutocerebrum 29 . However, Hirth et al. have re-examined this expression using neural segmental markers, and now show that lab is actually expressed only in the posterior of brain segment b3 (the tritocerebrum), posterior to the anterior boundary of pb expression in segment b2 (the deutocerebrum) (Ref. 27; Fig. 2). The same exception to colinearity therefore occurs in the fly CNS as well as in vertebrates. Homology of nervous systems It is important to realize that the ability of homologous genes to replace each other functionally does not necessarily show that the regions in which they are expressed are homologous. The Otx and otd gene products, for example, are transcription factors. It is quite possible that the ability of OTX proteins to bind a particular DNA sequence has been conserved, while the downstream genes regulated by this binding have altered. This explanation would still allow the proteins to replace each other, but they would be regulating different downstream genes in each animal. The formal possibility has not been ruled out that expression of otd and Otx in the anterior brain could have arisen independently in insects and chordates. However, it is now clear that the similar expression of Otx genes in the two nervous systems has functional relevance; the similarity cannot be explained by the gain of a few enhancer elements, and thus is less likely to have arisen by convergent evolution. These papers also show that, even though Hox and Otx genes are expressed in the early blastoderm in flies, this does not (as the auricularia hypothesis 20 suggests) mean that the ectoderm wall is homologous to the vertebrate neural tube, because the same genes are also expressed in the fly CNS, and in a very similar manner to vertebrates. Together with other evidence 17,22,30,31 , these results might be the final nail in the coffin for theories like the auricularia hypothesis that support an independent origin for the insect and vertebrate nervous systems. Significance of otd/Otx functional equivalence The fly otd gene can replace mouse Otx1 astoundingly well. We would have perhaps expected that Otx1 would have evolved new functions to do with vertebrate-specific developmental programs, that otd could not perform. It appears that only the homeobox and the regulatory elements, plus perhaps some less well conserved regions like the acidic activation domains, are sufficient for most of the functions of Otx1. New expression domains acquired since the arthropod/vertebrate split might account for much of the differences between otd and Otx1; these differences can only be detected experimentally by examining and possibly replacing regulatory elements rather than just coding regions. Replacement of regions of the gene outside the homeobox would test whether these regions, or only the homeobox, is important. Although otd can replace most functions of Otx1, some cannot be replaced. This could simply be a quantitative effect: Acampora et al. 9 have shown that in brain development Otx `dosage' is more important than which individual Otx gene is present, although Otx2 does appear to be more `potent' than Otx1, since the Otx2 mutant has a stronger phenotype. In the rescue experiments of Leuzinger et al. 25 , Otx1 is less effective at rescuing the otd phenotype than Otx2. otd could simply be less TIG J UNE 1998 V OL . 14 N O . 6 212 dpp dpp sog Ventral Dorsal Drosophila Ventral Dorsal Mouse msh vnd vnd ms h ms x Nkx2 Nkx2 msx chordin Bmp4 Bmp4 Midline, netrin + AS-C/ash F IGURE 1. Transverse sections through the fly and vertebrate central nervous system primordia, showing similar dorsoventral regulation of pattern by the sog (short gastrulation)/chordin, dpp (decapentaplegic)/BMP4, Msx/msh, Nkx2/vnd, AS-C (achaete-scute complex)/ash (AS-C homologues) and netrin gene families. (Redrawn from Ref. 23, with extra information from Ref. 22.) C O M M E N T `potent' in the mouse than either Otx1 or Otx2, as is supported by the fact that mice with two copies of Otx2 and no Otx1 show a weaker phenotype than mice 24 with one copy of each gene (Ref. 32). Some of the lack of rescue cannot be explained in this way, however, and could reflect truly novel functions of Otx1 that depend on its coding region. One example is the lateral semicircular canal phenotype of Otx1, which cannot be rescued by otd, suggesting that Otx1 has taken on a role in development of this canal in the inner ear after duplication. Significantly, this canal evolved around the time when the Otx genes are thought to have duplicated, at the origin of jawed vertebrates 33 (agnathan fish have only two semicircular canals, and lack the lateral one 24 ). The evolution of the midbrain hindbrain boundary A second way in which otd cannot replace Otx1 is shown in the graded level of rescue in the brain: the midbrain hindbrain boundary (MHB), midbrain and cerebellum, at the posterior of the Otx domain, are most sensitive to a low level of otd/Otx function and thus can be less easily rescued, while the telencephalon (at the anterior of the Otx domain) is least sensitive 24 . The vertebrate MHB is known to act as an organizer, producing signals that pattern the midbrain and anterior hindbrain 35,36 , but no such organizer has been described in the brain of Drosophila. Like the lateral semicircular canal, the organizer may have arisen early in vertebrate evolution, since lampreys and all jawed vertebrates, but not hagfish or cephalochordates, have an undeniable midbrain and cerebellum 34 . Thus the inability of Drosophila otd to rescue the midbrain hindbrain boundary phenotype of mouse Otx1 may be because the MHB is a new structure that evolved, or was at least greatly elaborated, early in the vertebrate lineage. The Otx genes themselves may have helped in the evolution of the MHB organizer; changing Otx expression patterns could have produced a gap between Otx and Hox expression domains (otd and Hox are adjacent in the fly; Fig. 2), which was then filled by MHB-specific transcription factors and signalling molecules such as EN, PAX2, PAX5, PAX8, WNT1 and FGF8 (Ref. 36). A thorough analysis of the development of this region in a range of vertebrates is clearly needed. Perspectives The papers reviewed here answer some questions while raising others. The list of known similarities between the vertebrate and arthropod nervous systems has been expanded significantly, suggesting that they are probably homologous structures. If this is so, has the auricularia hypothesis been disproved? Enteropneusts, a group of invertebrate chordates, have both a dorsal neural tube and a ventral nerve plexus, and it has been argued 37 that this shows that the vertebrate dorsal neural tube cannot be homologous to the invertebrate ventral nervous system. However, it is possible to argue that they are, if one of the enteropneust nervous systems is an independently-derived novelty, or if the two enteropneust nervous systems arose from a duplication of the single ancestral one. Investigation of expression patterns of developmental genes in enteropneusts may resolve the discrepancy. Additionally, it is still not resolved whether the functional equivalence of genes, as shown by their ability to replace each other, really demonstrates that the genes are doing the same thing in the two animals concerned. It would be interesting, for example, to test whether mammalian Hoxb2 can rescue the phenotype of Drosophila lab mutants (lab being a group 1 Hox gene, Hoxb2 a group 2). So far, genes in the same subfamily have mostly been tested for rescuing ability [otd and Otx1 (Refs 24 26), hh and shh (Ref. 30) or Hoxb1 and labial (Ref. 17)], but this ability may simply show that the important coding-region functions of a whole gene family have remained constant, rather than that two genes are the closest homologues within the family. The ability to rescue should be tested for a range of related genes, not just the ones suspected to be most closely related to each other. Certainly, these studies emphasize the need to compare regulatory elements as much as coding regions between species, since it appears that evolution has relied predominantly on regulatory changes. References 1 McGinnis, W. and Krumlauf, R. (1992) Cell 68, 283 302 2 Williams, N.A. and Holland, P.W.H. (1996) Nature 383, 490 3 Simeone, A. et al. (1992) Nature 358, 687 690 4 Younossi-Hartenstein, A. et al. (1997) Dev. Biol. 182, 270 283 TIG J UNE 1998 V OL . 14 N O . 6 213 1 2 3 4 5 7 6 M P2 P3 T Rhombomeres Mouse 8 P1 B1 B2 B3 S1 S2 S3 ems lab (Hox1) pb (Hox2 ) otd Drosophila Dfd (Hox4 ) Scr (Hox5 ) Antp (Hox6 ) Ubx (Hox7 ) Hoxb1 Hoxb2 Otx2 Otx1 Emx2 Emx1 Hoxb3 Hoxb4 Hoxb5 Hoxb6 Hoxb7 F IGURE 2. Anteroposterior gene expression in the fly and mouse central nervous systems showing Hox, Otx/otd and Emx/ems expression patterns. (Redrawn from Ref. 27, with extra information from Refs 6, 29, 38 and H. Reichert, pers. commun.) Arrow denotes the midbrain hindbrain boundary; B1 B3, brain segments (proto-, deuto- and trito-cerebrum, respectively); S1 S3, mandibular, maxillary and labial segments, respectively; P1 P3, prosomeres in the diencephalon; T, telencephalon; M, mesencephalon. C O M M E N T TIG J UNE 1998 V OL . 14 N O . 6 214 Copyright 1998 Elsevier Science Ltd. All rights reserved. 0168-9525/98/$19.00 PII: S0168-9525(98)01500-5 5 Finkelstein, R. and Boncinelli, E. (1994) Trends Genet. 10, 310 315 6 Hirth, F. et al. (1995) Neuron 15, 769 778 7 Holland, P., Ingham, P. and Krauss, S. (1992) Nature 358, 627 628 8 Matsuo, I. et al. (1995) Genes Dev. 9, 2646 2658 9 Acampora, D. et al. (1997) Development 124, 3639 3650 10 Ang, S.L. et al. (1996) Development 122, 243 252 11 Acampora, D. et al. (1995) Development 121, 3279 3290 12 Acampora, D. et al. (1996) Nat. Genet. 14, 218 222 13 Marin, F. and Puelles, L. (1994) Dev. Biol. 163, 19 37 14 Pellegrini, M. et al. (1996) Development 122, 3893 3898 15 Qiu, M. et al. (1996) Dev. Biol. 178, 174 178 16 Lowe, C.J. and Wray, G.A. (1997) Nature 389, 718 721 17 Lutz, B. et al. (1996) Genes Dev. 10, 176 184 18 Malicki, J., Schughart, K. and McGinnis, W. (1990) Cell 63, 961 967 19 Zhao, J.J., Lazzarini, R.A. and Pick, L. (1993) Genes Dev. 7, 343 354 20 Lacalli, T.C. (1994) Am. Zool. 34, 533 541 21 Arendt, D. and N bler-Jung, K. (1996) BioEssays 18, 255 259 22 DeRobertis, E.M. and Sasai, Y. (1996) Nature 380, 37 40 23 D'Alessio, M. and Frasch, M. (1996) Mech. Dev. 58, 217 231 24 Acampora, D. et al. Development (in press) 25 Leuzinger, S. et al. Development (in press) 26 Nagao, T. et al. Proc. Natl. Acad. Sci. U. S. A. (in press) 27 Hirth, F., Hartmann, B. and Reichert, H. Development (in press) 28 Prince, V.E. et al. (1998) Development 125, 393 406 29 Diederich, R.J. et al. (1989) Genes Dev. 3, 399 414 30 Krauss, S., Concordet, J.P. and Ingham, P.W. (1993) Cell 75, 1431 1444 31 Halder, G., Callaerts, P. and Gehring, W.J. (1995) Science 267, 1788 1792 32 Suda, Y., Matsuo, I., Kuratani, S. and Aizawa, S. (1996) Genes Cells 1, 1031 1044 33 Williams, N.A. and Holland, P.W.H. Mol. Biol. Evol. (in press) 34 Butler, A.B. and Hodos, W. (1996) Comparative Vertebrate Neuroanatomy, Wiley-Liss 35 Marin, F. and Puelles, L. (1995) Eur. J. Neurosci. 7, 1714 1738 36 Joyner, A.L. (1996) Trends Genet. 12, 15 20 37 Peterson, K.J. (1995) Nature 373, 111 112 38 Puelles, L. and Rubenstein, J.L. (1993) Trends Neurosci. 16, 472 479 Last month's issue of Trends in Genetics featured a review of the disease-related potential of the transcriptional cofactors CREB-binding protein (CBP) and the adenovirus E1A-associated protein, p300 (Ref. 1). Shortly after the publication of this review, the combined efforts of David Livingston's laboratory (Harvard Medical School, Boston, MA, USA) and Richard Eckner's laboratory (Univ. of Zu rich, Switzerland) appeared in Cell, describing the deleterious effects of inactivating one or both murine CBP and/or p300 alleles 2 . The results-at-a-glance are presented in Table 1. When a single p300 allele is inactivated, the resultant embryos suffer a significantly reduced viability (up to 55% died in utero, depending on genetic background), although heterozygotes that do survive do not suffer from further p300-insufficiency after birth. Mice homozygous for p300 mutations always die in utero, between days 9 and 11.5 of gestation. These nullizygous embryos are much smaller than their littermates and exhibit severe open neural tube and heart defects. Interestingly, cells removed from the p300 homozygous mutants displayed poor proliferation properties, implying that p300 is required for growth stimulation, an idea that is contrary to the general opinion that CBP and p300 are tumor suppressor proteins. Unlike p300, CBP heterozygous mutant mice, described earlier 3 , manifest skeletal abnormalities consistent with the human congenital Rubinstein Taybi syndrome, in which one CBP allele is inactivated 4 . CBP homozygous mutant mice, however, strongly resemble the p300 mutants, and also die in utero, between days 9 and 11.5 of gestation. Crossing the p300 and CBP heterozygous mutants produced double heterozygous CBP/p300 mutant embryos, which died in utero but otherwise shared phenotypic similarities to both CBP and p300 homozygous mutants. This remarkable result suggests that the two proteins exert certain common embryonic survival functions and that the combined dose of CBP and p300 is critical for mouse embryonic development. Although CBP and p300 are not completely redundant physiologically, these results suggest that a 25% drop in combined CBP/p300 levels (through the loss of one CBP or p300 allele) is enough to interfere seriously with embryonic development, while a 50% drop results invariably in embryonic death. References 1 Giles, R.H., Peters, D.J.M. and Breuning, M.H. (1998) Trends Genet. 14, 178 183 2 Yao, T.P. et al. (1998) Cell 93, 361 372 3 Tanaka, Y. et al. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 10215 10220 4 Petrij, F. et al. (1995) Nature 376, 348 351 Rachel H. Giles rachel@ruly46.medfac.leidenuniv.nl Department of Human Genetics, Leiden University Medical Center, Wassenaarseweg 72, 2333 AL Leiden, The Netherlands. Update CBP/p300 transgenic mice T ABLE 1. Mouse models for CBP and p300 mutations CBP p300 Phenotype Refs ++ ++ Normal + ++ Skeletal abnormalities 3 ++ Embryonic lethal 2, 3 ++ + Reduced viability 2 ++ Embryonic lethal 2 + + Embryonic lethal 2 Abbreviations: +, normal allele; , inactive allele.

37), while the strongly bound state time, t s , determines the velocity at which movement occurs (16 18). Thus, the elevated ATPase is not a reflection of a change in t s and thus is not relevant to velocity considerations. Importantly, we can conclude that none of the mutated myosins have been hampered in their ability to hydrolyze ATP, which we take as an indication that the mutations did not have generally deleterious effects on the myosin. The mutant and wild-type myosins were then subjected to an in vitro sliding filament motility assay (32, 33, 41). The sliding velocities increased with increasing number of light chain binding sites for wild-type and mutant myosins (Fig. 4 Left), consistent with the swinging neck-lever model. Most significantly, the 2xELCBS mutant form moved faster than the wild-type myosin (the range was 21 33% faster in four experiments), and the most straightforward interpretation of making the enzyme move faster is that the neck behaves like a lever arm. If one makes the further (undoubtedly oversimplified) assumptions that, first, all of the stroke derives from the movement of a relatively rigid lever arm that rotates about some fulcrum point, and, second, that the 2xELCBS mutant has a lever arm that is elongated by the linear insertion of one extra ELC binding domain, then one can extrapolate the points in Fig. 4 Left back to zero lever arm length. This ``fulcrum point'' in the structure is shown by the red dot in Fig. 1, and the sliding velocities are now proportional to the length of lever arm when the same set of data is replotted against the length measured from this putative fulcrum point (Fig. 4 Right). Milligan and colleagues (13, 14) provided complementary evidence for a fulcrum point in this region by comparing helically reconstituted actomyosin structures between the ADP-bound and rigor (no nucleotide) states. Interestingly, this putative fulcrum point is very near to what has been called the reactive thiol region in skeletal muscle myosin, which undergoes dramatic changes in structure during the ATPase cycle (38, 39). There are other, albeit more complicated, explanations for the velocity results shown in Fig. 4. For example, it is possible that there is another minor but independent mechanism to generate movement, such as a change in binding angle between actin and the myosin head at the actin myosin binding face, as has been long postulated (40). Thus, the fulcrum point of the swinging motion of the lever arm may be to the right of the red dot in Fig. 1 Upper, closer to the ELC binding domain. Another possibility is that t s is linearly related to the number of light chain binding sites, and this contributes to the changes in velocity since v d t s . This possibility can be tested in the future. For example, the feedback-enhanced laser trap assay (6) can be used to determine t s directly as a measure of the duration time of the myosin displacement. In summary, the linear relationship between sliding velocity and the neck length strongly supports the swinging neck-lever model. It is particularly noteworthy that we were able to create a mutant motor that moves faster than the wild type in a way the model predicts. An interesting point to consider (see Appendix) is that this lever arm of the S1 will have a certain bending stiffness and may be the structural equivalent of the elastic element that has long been known to be part of the actin myosin system, as elucidated by tension-transient experiments using muscle fibers (40). APPENDIX: Is the lever arm of myosin a molecular elastic element? J ONATHON H OWARD * AND J AMES A. S PUDICH *Department of Physiology and Biophysics, University of Washington, Seattle, WA 98195-7290; and Department of Biochemistry, Stanford University School of Medicine, Stanford, CA 94305 The classic experiments of Huxley and Simmons (40) defined an elastic element in muscle that has been attributed to the myosin molecule. They measured the tension drop when a stimulated muscle held at a fixed length is rapidly shortened through a small distance and found that a component of the system behaves like a linear spring. Such an elastic element is fundamental to force generation because it allows strain to develop within the motor prior to movement of the cargo; relief of this strain then drives the relative displacement of the motor and the track along which it moves. While diagrammatic representations often show this elastic spring as being part of the myosin rod beyond the light-chain binding domain of the molecule, we consider here that the elastic element is the light-chain binding domain itself and may account quantitatively for the cross-bridge stiffness observed in muscle experiments. The head domain of myosin, commonly called subfragment 1 or S1, is the only part of the myosin molecule required for movement in vitro (33) and for production of force similar to that seen in intact muscle (9, 42). An unusual structural feature of S1 is the 8-nm-long light-chain binding domain that is at F IG . 4. Sliding velocities of mutant and wild-type myosins. Bars indicate standard deviation. (Left) Sliding velocity as a function of the number of light chain binding sites. These data are representative of four independent experiments with different preparations of proteins over a period of a year. (Right) The same set of data is replotted against the length of the putative lever arm. The lever arm lengths for wild type and each mutant were measured from the fulcrum point shown as a red dot in Fig. 1 to the 90 bend at the C terminus of the long heavy chain -helix (shown in violet in Fig. 1) that makes up the neck domain--these lengths are 3-D computer-graphic measurements based on the crystal structure (2). 4462 Biophysics: Uyeda et al. Proc. Natl. Acad. Sci. USA 93 (1996) the C terminus of the S1 moiety (2, 3). It has been suggested that this region of the myosin head could serve as a lever arm to amplify smaller conformational changes elsewhere in the motor domain (5, 13 15, 19 21, 43, and this paper). Indeed, fluorescence polarization experiments have shown that the light-chain binding region changes orientation by a minimum of 3 relative to the filament axis in muscle in response to quick length changes and during the transitions between states of the cross-bridge cycle associated with active force production (15). While this angle change would appear to be too small to account for a unitary displacement of several nanometers (6), it is a minimum value for technical reasons, and two other complementary studies strongly support the lever arm hypothesis. First, electron microscopy of decorated actin filaments showed that a rotation of the light-chain binding domain through 23 accounts well for the two different conformations that S1 adopts depending on whether ADP is bound at the active site; the difference could account for as much as 3.5 nm of movement of the far C terminus of S1 (13, 14). Second, this paper used molecular genetic approaches to shorten, and importantly, to elongate the lever arm and demonstrate a linear relationship between the lever arm length and the velocity with which the myosin moves in vitro. We argue here that the lever arm could also be the elastic element referred to above, since the elasticity of the light-chain binding domain is expected to be comparable to that measured in the rapid shortening experiments. Furthermore, the nature of the light chains and their interaction with the 8-nm-long -helical stretch of the heavy chain at the C terminus of S1 may determine the spring constant of the light-chain binding domain and therefore affect the force that the molecular motor can produce. Consider a very simple model of the lever arm as a clamped beam of length L and flexural rigidity (the resistance to bending forces) equal to EI. If a transverse force F is applied at the free end, then this end will move through a distance x such that: F 3EI L 3 x (44). In other words, the beam has a stiffness 3EI L 3 3kTL p L 3 , where L p EI kT is the persistence length (45), k is the Boltzmann constant, and T is temperature. The light-chain binding domain has a length of 8 nm. It seems reasonable to consider that the lever arm, which has two light chains wrapped around the long -helix, has a rigidity similar to that of a coiled coil, which has two -helices wrapped around each other. The persistence length of a coiled coil is 100 nm (J.H., unpublished measurements derived from the coiled-coil myosin rod domain). For comparison, the L p of DNA, which has a dimension similar to these two protein structures, is 50 nm (46). Substituting L 8 nm, L p 100 nm, and kT 4 pN nm, we obtain 2 pN nm. On the other hand, the rapid shortening experiments indicate a muscle stiffness equal to 0.27 pN nm when normalized to the total number of myosin heads per half sarcomere [a shortening of 6 nm per half sarcomere drops the force from 1.6 pN per head to zero (47)]. Since only about half the compliance in muscle resides in the myosin heads and the other half resides in the actin filaments (e.g., see ref. 48), this value for the stiffness needs to be doubled to 0.5 pN nm per myosin head. If only a quarter of the myosin heads were attached during isometric contraction (duty ratio of 0.25; refs. 6 and 16), then the stiffness per attached head would be 2 pN nm, equal to that derived above! Clearly, this equality could be fortuitous given the large uncertainties in both the experimental and theoretical stiffnesses. The assumptions made, however, are not unreasonable, and the calculations do show that it is quite plausible that the elasticity of myosin resides within the light-chain binding domain, which corresponds to the lever arm. Indeed, one expects the light-chain binding domain to contribute some compliance to the myosin molecule. There are three interesting predictions that follow from the hypothesis that the lever arm is the elastic element. (i) The motor force should be inversely proportional to the square of the length of the lever arm. To see this, let the force-generating conformational change be a rotation, through an angle , of the insertion point of the lever into the motor domain. Thus, in the absence of a restoring force, the tip of the lever arm (the C terminus of S1) would move through a distance x L , On the other hand if there were a restoring force (F max ) that prevented the C terminus of the lever arm from moving, then F max 3kTL p L 3 x 3kTL p L 3 L 3kTL p L 2 . Since the angular change is independent of the length of the lever arm, it follows that the maximum force is proportional to L 2 . On the other hand, if the lever arm acted as a rigid rod and the elasticity were due to a pivotal spring (49) located at the point of insertion into the motor domain, then the maximum force would depend on L 1 . (ii) The maximum work should be inversely proportional to the lever length (L 1 ). To see this, note that if the restoring force (F o ) is less than the maximum force, then the tip will move through a distance x F o (the working stroke), and the amount of work done will equal W F o x F o F o x F o2 . The maximum work occurs when F o F max 2, and W max F max x 4 3 4 kTL p 2 L. That is, the maximum work is inversely proportional to the lever length. This leads to a paradox at the shortest lever arm lengths where the work might get so large as to exceed the theoretical maximum force. Presumably a motor with a very short lever arm will fail at high forces (the rotation through would not take place). (iii) The maximum force will depend on the stiffness of the lever arm. For example, if the link between the ELC and the catalytic domain of S1 and or the link between the ELC and RLC domains were flexible, we would expect a smaller stiffness and thus a smaller force. Thus, the properties of the light chains may affect the flexural rigidity of the lever arm, thereby regulating the force produced by a particular myosin isoform. The establishment of laser trap technologies to measure directly the force and work produced by a single myosin molecule (6, 50) and systems that allow genetic engineering of the molecular motor myosin to produce myosins with different lever arm lengths (this paper) should allow critical testing of whether force production is inversely proportional to the lever arm length squared, as predicted by the elastic lever arm model. The same approaches should allow testing of the concept that the nature of the light chains modulates the spring constant of the elastic lever arm and therefore the amount of force that can be produced by different isoforms of myosin, Biophysics: Uyeda et al. Proc. Natl. Acad. Sci. USA 93 (1996) 4463 which have different light chains. Indeed, even skeletal myosin binds two alternate forms of RLC, for reasons that have been unclear. Moreover, myosin light chains are altered by posttranslational modifications, such as phosphorylation in the case of smooth muscle myosin and Dictyostelium myosin (for a review, see ref. 51) and binding of Ca 2 in the case of scallop myosin (52). One goal then is to use molecular genetics and laser trap technology to gain detailed molecular information about the physiological relevance of altered myosin types. We thank members of the Spudich laboratory for stimulating discussions and advice, and K. Zaita for technical assistance. This work is supported by a Human Frontier Science Program short-term fellowship to T.Q.P.U. and National Institutes of Health Grant GM33289 to J.A.S. 1. Itakura, S., Yamakawa, H., Toyoshima, Y. Y., Ishijima, A., Kojima, T., Harada, Y., Yanagida, T., Wakabayashi, T. & Sutoh, K. (1993) Biochem. Biophys. Res. Commun. 196, 1504 1510. 2. Rayment, I., Rypniewski, W., Schmidt-Base, K., Smith, R., Tomchick, D., Benning, M., Winkelmann, D., Wesenberg, G. & Holden, H. (1993) Science 261, 50 58. 3. Xie, X., Harrison, D. H., Schlichting, I., Sweet, R. M., Kalabokis, V. N., Szent-Gyorgyi, A. G. & Cohen, C. (1994) Nature (London) 368, 306 312. 4. Cooke, R., Crowder, M. S., Wendt, C. H., Bamett, V. A. & Thomas, D. D. (1984) Adv. Exp. Med. Biol. 170, 413 427. 5. Vibert, P. & Cohen, C. (1988) J. Muscle Res. Cell Motil. 9, 296 305. 6. Finer, J. T., Simmons, R. M. & Spudich, J. A. (1994) Nature (London) 368, 113 119. 7. Ishijima, A., Harada, Y., Kojima, H., Funatsu, T., Higuchi, H. & Yanagida, T. (1994) Biochem. Biophys. Res. Commun. 199, 1057 1063. 8. Miyata, H., Hakozaki, H., Yoshikawa, H., Suzuki, N., Kinoshita, K., Jr., Nishizaka, T. & Ishiwata, S. (1994) J. Biochem. (Tokyo) 115, 644 647. 9. Molloy, J. E., Burns, J. E., Kendrick-Jones, J., Tregear, R. T. & White, D. C. S. (1995) Nature (London) 378, 209 212. 10. Tokunaga, M., Sutoh, K. & Wakabayashi, T. (1991) Adv. Biophys. 27, 157 167. 11. Wakabayashi, K., Tokunaga, M., Kohno, I., Sugimoto, Y., Hamanaka, T., Takezawa, Y., Wakabayashi, T. & Amemiya, Y. (1992) Science 258, 443 447. 12. Highsmith, S. & Eden, D. (1993) Biochemistry 32, 2455 2458. 13. Whittaker, M., Kubalek, E. M. W., Smith, J. E., Faust, L., Milligan, R. A. & Sweeney, H. L. (1995) Nature (London) 378, 748 751. 14. Jontes, J. D., Kubalek, E. M. W. & Milligan, R. A. (1995) Nature (London) 378, 751 753. 15. Irving, M., St Claire Allen, T., Sabido-David, C., Craik, J. S., Brandmeier, B., Kendrick-Jones, J., Corrie, J. E. T., Trentham, D. R. & Goldman, Y. E. (1995) Nature (London) 375, 688 691. 16. Uyeda, T. Q. P., Kron, S. J. & Spudich, J. A. (1990) J. Mol. Biol. 214, 699 710. 17. Harada, Y., Sakurada, K., Aoki, T., Thomas, D. D. & Yanagida, T. (1990) J. Mol. Biol. 216, 49 68. 18. Spudich, J. A. (1994) Nature (London) 372, 515 518. 19. Lowey, S., Waller, G. S. & Trybus, K. M. (1993) Nature (London) 365, 454 456. 20. Uyeda, T. Q. P. & Spudich, J. A. (1993) Science 262, 1867 1870. 21. Waller, G. S., Ouyang, G., Swafford, J., Vibert, P. & Lowey, S. (1995) J. Biol. Chem. 270, 15348 15352. 22. Ausubel, F., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K., eds. (1994) Current Protocols in Molecular Biology (Current Protocols, New York). 23. Egelhoff, T. T., Manstein, D. J. & Spudich, J. A. (1990) Dev. Biol. 137, 359 367. 24. Ruppel, K. M., Uyeda, T. Q. P. & Spudich, J. A. (1994) J. Biol. Chem. 269, 18773 18780. 25. Sussman, M. (1987) Methods Cell Biol. 28, 9 29. 26. Egelhoff, T. T., Titus, M. A., Manstein, D. J., Ruppel, K. M. & Spudich, J. A. (1991) Methods Enzymol. 196, 319 334. 27. Spudich, J. A. & Watt, S. (1971) J. Biol. Chem. 246, 4866 4871. 28. Margossian, S. S. & Lowey, S. (1982) Methods Enzymol. 85, 55 71. 29. Gordon, D. J., Yang, Y. Z. & Korn, E. D. (1976) J. Biol. Chem. 251, 7474 7479. 30. Bradford, M. M. (1976) Anal. Biochem. 72, 248 254. 31. Perrie, W. T. & Perry, S. V. (1970) Biochem. J. 119, 31 38. 32. Kron, S. J. & Spudich, J. A. (1986) Proc. Natl. Acad. Sci. USA 83, 6272 6276. 33. Toyoshima, Y. Y., Kron, S. J., McNally, E. M., Niebling, K. R., Toyoshima, C. & Spudich, J. A. (1987) Nature (London) 328, 536 539. 34. Maita, T., Yajima, E., Nagata, S., Miyanishi, T., Nakayama, S. & Matsuda, G. (1991) J. Biochem. (Tokyo) 110, 75 87. 35. Warrick, H. M., DeLozanne, A., Leinwand, L. A. & Spudich, J. A. (1986) Proc. Natl. Acad. Sci. USA 83, 9433 9437. 36. Stein, L. A., Schwarz, Jr., R. P., Chock, P. B. & Eisenberg, E. (1979) Biochemistry 18, 3895 3909. 37. Rosenfeld, S. S. & Taylor, E. W. (1984) J. Biol. Chem. 259, 11920 11929. 38. Burke, M. & Reisler, E. (1977) Biochemistry 16, 5559 5563. 39. Wells, J. A. & Yount, R. G. (1979) Proc. Natl. Acad. Sci. USA 76, 4966 4970. 40. Huxley, A. F. & Simmons, R. M. (1971) Nature (London) 233, 533 538. 41. Kron, S. J., Toyoshima, Y. Y., Uyeda, T. Q. P. & Spudich, J. A. (1991) Methods Enzymol. 196, 399 416. 42. Kishino, A. & Yanagida, T. (1988) Nature (London) 334, 74 76. 43. Rayment, I., Holden, H. M., Whittaker, M., Yohn, C. B., Lorenz, M., Holmes, K. C. & Milligan, R. A. (1993) Science 261, 58 65. 44. Feynman, R. P., Leighton, R. B. & Sands, M. (1964) The Feynman Lectures on Physics (Addison-Wesley, Reading, MA), Vol. 2. 45. Gittes, F., Mickey, B., Nettleton, J. & Howard, J. (1993) J. Cell Biol. 120, 923 934. 46. Hagerman, P. J. (1988) Annu. Rev. Biophys. Biophys. Chem. 17, 265 286. 47. Bagshaw, C. R. (1993) Muscle Contraction (Chapman & Hall, London), 2nd Ed. 48. Higuchi, H., Yanagida, T. & Goldman, Y. E. (1995) Biophys. J. 69, 1000 1010. 49. Howard, J. & Ashmore, J. F. (1987) Hearing Res. 23, 93 104. 50. Finer, J., Mehta, A. & Spudich, J. A. (1995) Biophys. J. 68, 291s 297s. 51. Tan, J. L., Ravid, S. & Spudich, J. A. (1992) Annu. Rev. Biochem. 61, 721 759. 52. Houdusse, A. & Cohen, C. (1996) Structure 4, 21 32. 4464 Biophysics: Uyeda et al. Proc. Natl. Acad. Sci. USA 93 (1996)