© The Author 2006. Published by Oxford University Press.
EDITORIAL |
Can Mutations in 
-Actin Modulate the Toxicity of Microtubule Targeting Agents?
Affiliation of author: Medicine Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD
Correspondence to: Tito Fojo, MD, PhD, Medicine Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Building 10, Room 12N226, 9000 Rockville Pike, Bethesda, MD 20892 (e-mail: tfojo{at}helix.nih.gov).
So, how do microtubule targeting agents work, anyway? It is widely taught that stabilizing agents, such as paclitaxel, bind and stabilize microtubules, whereas destabilizing agents, such as vincristine, bind 
/
-tubulin heterodimers or the ends of microtubules and prevent microtubule polymerization (1). However, Jordan and Wilson have long argued that, when used at the low concentrations that kill cells, both stabilizing and destabilizing agents suppress microtubule dynamics, and that it is this change in microtubule dynamics that is important (1,2). Under either scenario, mitotic arrest ensues and is then followed by cell death. In this issue of the Journal, Verrills et al. (3) argue that 
-actin is essential for the efficacy of microtubule targeting agents and that, by affecting the activity of microtubule targeting agents, alterations in -actin constitute a novel mechanism of resistance to microtubule targeting agents. Can this be so? Can we have been unaware of this for so long? As with most novel findings, time will tell, but the manuscript raises some interesting questions.
Since vincristine was first introduced to clinical oncology in the late 1950s, basic scientists have made enormous strides in clarifying the distinct roles of the microtubule and actin cytoskeletons in a variety of processes, including maintenance of cell shape, cell division, and the intracellular movement of protein cargo and organelles. Such diverse functions are possible because both cytoskeletons can be assembled rapidly into polarized polymers via the polymerization of monomers, in the case of actin, and heterodimers, in the case of microtubules; both can be disassembled rapidly; and both are able to interact with myriad proteins and molecular motors that help in their assembly, movements, and functions.
Because of the widespread use of microtubule targeting agents in the therapy of cancer, most of us are familiar with basic microtubule biology (1). Microtubules are ubiquitous proteins that are composed of /-tubulin heterodimers that self-assemble into long protofilaments; 13 protofilaments assemble in turn into the structure that we recognize as a microtubule. In most cells, microtubules are organized into a single array that originates near the nucleus at the microtubule organizing center, and they fill the cytoplasm as they participate in numerous cellular processes. Although oncologists often focus on the role of microtubules in mitosis, these structures are involved in numerous other cellular processes. In addition to their important roles in cell shape and motility and intracellular trafficking, microtubules also provide an enormous surface for protein–protein interactions in a nondividing interphase cell. This surface has been estimated at approximately 1000 µm2 of protein surface—as large as the plasma membrane and 10 times the size of the nuclear envelope (4).
The actins are also ubiquitous proteins that are involved in a variety of cellular functions, including cell motility, adhesion, shape and muscle contraction (5). Like tubulins, the actins are highly conserved, with more than 90% amino acid sequence homology between the single actin gene in yeast and the most distantly related -actin gene in humans (6). Six highly conserved actin isoforms have been identified in vertebrates: three alpha isoforms, one beta isoform, and two gamma isoforms. Four isoforms (the three alpha isoforms and one of the two gamma isoforms) are muscle specific and are involved in contractile movement. In most nonmuscle cells, the single beta isoform and the non–muscle-specific gamma isoform coexist as components of the cytoskeleton and act as mediators of internal cellular motility. -Actin is the predominant isoform in most cells and is thus commonly used as a control in scientific experiments. -Actin predominates in intestinal epithelial cells and in auditory hair cells, where it is found in all actin-containing structures (i.e., stereocilia, cuticular plates, and adherens junctions). Mutations in -actin have been identified in families with hearing loss, as discussed below. The -actin and -actin proteins differ by only four amino acids near the amino terminus but have different roles in cellular architecture. In most cells, the ratio of -actin to -actin is tightly regulated, and any perturbation of this ratio has consequences (7–9).
Although the actin and microtubule cytoskeletons have distinct roles within the cell, scientists have long suspected that that these two seemingly independent systems interact. In 1970, Vasiliev et al. (10) showed that the microtubule targeting agent colcemid adversely affects locomotory behavior by disrupting the polarized distribution of actin-dependent protrusions at the leading edge of a migrating fibroblast. The coincidence of the impaired locomotion and the disrupted actin cytoskeleton following colcemid-induced microtubule depolymerization suggested that the microtubule system is crucial for the proper distribution of actin and for actin-mediated membrane contractions. Since this original observation, we have come to understand that complex interactions between the microtubule and actin systems are responsible for many cellular processes and properties, including a cell's shape, direction of movement, and overall intracellular organization (11,12). Because both the microtubule and actin systems can be rapidly assembled and disassembled, these cellular properties are dynamic and can change rapidly.
The interactions between actin and microtubules can be classified as either regulatory or structural (12). Regulatory interactions are those that allow the microtubule and actin systems to control each other by affecting signaling cascades. The best example of a regulatory interaction involves the Rho family of small GTPases (13). The activity of the Rho proteins is regulated by both microtubules and actin (14,15). In turn, RhoA mediates the formation of contractile actin structures, such as stress fibers, and the stabilization of a subpopulation of microtubules (16,17). By contrast, structural interactions are those that physically link actin and microtubules. Such interactions can be static or dynamic (12). Given the extensive interactions between these two systems and their ability to regulate each other's function, it is perhaps not surprising that Verrills et al. (3) have found that alterations in -actin can alter the cell's sensitivity to microtubule targeting agents.
Microtubule targeting agents are arguably the most effective and widely used chemotherapeutic agents in clinical oncology. They are used in the treatment of most major cancers, including breast, ovarian, lung, esophageal, and head and neck cancers, as well as in the treatment of lymphomas and leukemias. Although the development of new therapeutics is eagerly awaited, novel targeted agents are unlikely to replace microtubule targeting agents in the foreseeable future. Consequently, it is important to understand what impairs the effectiveness of microtubule targeting agents and whether we can discriminate between tumors that may respond to such agents and those that will be resistant to them. Given this, can we agree with the interpretation of the data advanced by Verrills et al. (3)—more specifically, can we expect that the reduced levels of -actin they observed in clinical samples might confer reduced sensitivity to vincristine or other microtubule targeting agents in patients with acute lymphoblastic leukemia? Are these changes meaningful, or are they fortuitous and inconsequential?
Although a definitive answer will come only when others have had a chance to perform similar analyses, I found the evidence presented by Verrills et al. to be persuasive. The extent of reduction in -actin RNA in clinical samples obtained from patients with childhood leukemia at the time of a relapse mirrored the extent of reduction observed in neuroblastoma cells transfected with siRNA directed against -actin. And although the reduction in -actin RNA in the siRNA-transfected SH-EP neuroblastoma cells was—by the authors' own admission—modest, this reduction was sufficient to increase resistance to the microtubule targeting agents paclitaxel, vinblastine, and epothilone B and to reduce paclitaxel-mediated microtubule polymerization. The authors argue that a larger reduction in -actin could not be achieved in siRNA-transfected cells because "the -actin gene is an essential gene and cells cannot tolerate excessive silencing." This assessment is probably correct. In addition, it should also be noted that, despite the modest size of the reduction in -actin expression in siRNA-transfected cells, these cells had more pronounced actin stress fiber networks and were larger than mock-transfected or control siRNA–transfected cells. This observation is consistent with prior findings that increasing -actin results in larger cells, whereas increasing -actin reduces cell size (8).
But the authors also leave us with questions that are either not answered at all or not answered satisfactorily. The first—and admittedly the most difficult—question concerns the basic mechanism responsible for the altered drug sensitivity. Their initial observation was the existence of mutations in -actin; subsequent experiments examined the effect of -actin levels. However, the authors' suggestion that the P98L and V103L substitutions in the mutant -actins that they isolated from CCRF-CEM leukemia cell lines resistant to microtubule targeting agents could reduce sensitivity to microtubule targeting agents by altering myosin binding is unsatisfactory. The myosin–actin interaction is complex, and there is no evidence that this region would in turn be involved in engaging tubulin (18–20). Furthermore, although the authors are cautious in their interpretation of the data overall, I do not agree with their contention that wild type -actin expression may be required for disruption of microtubule structures induced by antimicrotubule drugs. It is important to remember that microtubule targeting agents can "disrupt" (polymerize or depolymerize) microtubules quite readily in vitro using purified tubulin that lacks any actin.
It is also unclear why two mutant forms of -actin were found in each cell line, as evidenced by the sequence analysis. It seems highly unlikely that a cell containing either of the two pairs of mutations preexisted in the population—much less two such cells because different clones were isolated in the course of these independent selections. Never mind that the double substitutions did not appear to have an additive effect. We are taught that mutations in structural proteins are often dominant, and for -actin, such mutations have been identified in families with autosomal dominant hearing loss (21,22). The apparent need to either inactivate or alter the second -actin gene in both CCRF-CEM drug selections reported by Verrills et al. is reminiscent of similar observations in paclitaxel and epothilone selections, in which drug binding was impaired (23), but it is unlike the findings in hemiasterlin-selected cells, in which mutant tubulins with altered structural stability were found to be coexpressed with a wild-type allele (24). The need to mutate the second allele suggests that the mutations, rather than affecting a structural property of -actin, have an effect on an enzymatic property of the protein. Given the increasing number of properties attributed to the actins, an explanation other than a straightforward disruption of a structural interaction should be considered.
Finally, we should consider what the data might be telling us indirectly. First, they underscore the complexity of microtubule targeting agent drug action and remind us that, almost 50 years after the first use of vincristine, we still do not fully understand how such drugs kill cells. Although both basic and clinical scientists are routinely challenged with complex signal transduction pathways, all too often we think linearly about microtubule targeting agents: drug binds target 
mitotic arrest occurs cell dies. Verrills et al. remind us again that, although apparently simple in their action, these drugs are mechanistically complex.
A second and subtler point is the implication that this complex mechanism involves more than mitotic arrest. Both -actin and -actin are often referred to as cytoplasmic actins. Although it has been suggested that -actin may have a role in the regulation of cell growth and there is ample evidence that both actin and myosin have important roles in the readout of genetic information, the results presented herein do not support a role for actin in promoting mitotic arrest (25,26). Instead, they lend credence to the thesis that the activity of microtubule targeting agents is not confined to dividing cells. The thesis that even cells not actively dividing can be killed by microtubule targeting agents can explain the observation made by every clinical oncologist: that fractional cell kill often exceeds the mitotic fraction by a large factor. There is no doubt that, in vitro, with cells dividing every 10–24 hours, the addition of a microtubule targeting agent is followed within 24 hours by arrest in G2/M (actually mitotic arrest) and subsequent cell death. But fortunately most tumors that an oncologist encounters are not dividing every 10–24 hours. Indeed, the majority have mitotic indices of 1% or less. Given this, how then can one explain shrinkage of 50% or even more after two cycles of chemotherapy, each administered over a few hours? A possible antiangiogenic effect aside, it can be argued that this shrinkage occurs because microtubule targeting agents target not only dividing but also nondividing cells. Any cell that relies on the microtubule network for the transport of cargo and for signal transduction as do cancer cells and neurons is likely to suffer, and if the disruptive effect of the agent is severe enough, it can lead to cell death (4). What evidence do we have that this happens? The most straightforward evidence is provided by the principal toxicity of microtubule targeting agents: neurotoxicity. Neurons are damaged and often killed by microtubule targeting agents despite their inability to divide in adults. The finding by Verrills et al. that a protein not directly linked to mitosis (-actin) is important in the modulation of the activity of microtubule targeting agents provides indirect support for the thesis that microtubule targeting agents are active in cell cycle phases other than M.
Studies attempting to confirm the report by Verrills et al. should follow quickly. If the findings reported in this issue of the journal are confirmed, we will be humbled by the fact that it took us nearly 50 years to uncover the fact that -actin is essential for the activity of microtubule targeting agents. But we will also be encouraged that there is yet much to learn that can hopefully lead us to better therapies for those who fight cancer so vigorously.
REFERENCES
(1) Jordan MA, Wilson L. Microtubules as a target for anticancer drugs. Nat Rev Cancer 2004;4:253–65.[CrossRef][ISI][Medline]
(2) Jordan MA, Toso RJ, Thrower D, Wilson L. Mechanism of mitotic block and inhibition of cell proliferation by taxol at low concentrations. Proc Natl Acad Sci U S A 1993;90:9552–6.
(3) Verrills NM, Po'uha ST, Liu MLM, Liaw TYE, Larsen MR, Ivery MT, et al. Alterations in -actin and tubulin-targeted drug resistance in childhood leukemia. J Natl Cancer Inst 2006;98:1363–74.
(4) Gundersen GG, Cook TA. Microtubules and signal transduction. Curr Opin Cell Biol 1999;11:81–94.[CrossRef][ISI][Medline]
(5) Kaksonen M, Toret CP, Drubin DG. Harnessing actin dynamics for clathrin-mediated endocytosis. Nat Rev Mol Cell Biol 2006;7:404–14.[CrossRef][ISI][Medline]
(6) Zhu M, Yang T, Wei S, DeWan AT, Morell RJ, Elfenbein JL, et al. Mutations in the gamma-actin gene (ACTG1) are associated with dominant progressive deafness (DFNA20/26). Am J Hum Genet 2003;73:1082–91.[CrossRef][ISI][Medline]
(7) Otey CA, Kalnoski MH, Bulinski JC. Identification and quantification of actin isoforms in vertebrate cells and tissues. J Cell Biochem 1987;34:113–24.[CrossRef][ISI][Medline]
(8) Schevzov G, Lloyd C, Gunning P. High level expression of transfected beta- and gamma-actin genes differentially impacts on myoblast cytoarchitecture. J Cell Biol 1992;117:775–85.
(9) Lloyd C, Schevzov G, Gunning P. Transfection of nonmuscle beta- and gamma-actin genes into myoblasts elicits different feedback regulatory responses from endogenous actin genes. J Cell Biol 1992;117:787–97.
(10) Vasiliev JM, Gelfand IM, Domnina LV, Ivanova OY, Komm SG, Olshevskaja LV. Effect of colcemid on the locomotory behaviour of fibroblasts. J Embryol Exp Morphol 1970;24:625–40.[ISI][Medline]
(11) Goode BL, Drubin DG, Barnes G. Functional cooperation between the microtubule and actin cytoskeletons. Curr Opin Cell Biol 2000;12:63–71.[CrossRef][ISI][Medline]
(12) Rodriguez OC, Schaefer AW, Mandato CA, Forscher P, Bement WM, Waterman-Storer CM. Conserved microtubule-actin interactions in cell movement and morphogenesis. Nat Cell Biol 2003;5:599–609.[CrossRef][ISI][Medline]
(13) Wittmann T, Waterman-Storer CM. Cell motility: can Rho GTPases and microtubules point the way? J Cell Sci 2001;114:3795–803.
(14) Ren XD, Kiosses WB, Schwartz MA. Regulation of the small GTP-binding protein Rho by cell adhesion and the cytoskeleton. EMBO J 1999;18:578–85.[CrossRef][ISI][Medline]
(15) Waterman-Storer CM, Worthylake RA, Liu BP, Burridge K, Salmon ED. Microtubule growth activates Rac1 to promote lamellipodial protrusion in fibroblasts. Nat Cell Biol 1999;1:45–50.[CrossRef][ISI][Medline]
(16) Etienne-Manneville S, Hall A. Rho GTPases in cell biology. Nature 2002;420:629–35.[CrossRef][Medline]
(17) Cook TA, Nagasaki T, Gundersen GG. Rho guanosine triphosphatase mediates the selective stabilization of microtubules induced by lysophosphatidic acid. J Cell Biol 1998;141:175–85.
(18) McKane M, Wen KK, Meyer A, Rubenstein PA. Effect of the substitution of muscle actin-specific subdomain 1 and 2 residues in yeast actin on actin function. J Biol Chem 2006. [Aug 1 Epub ahead of print.]
(19) Siddique MS, Miyazaki T, Katayama E, Uyeda TQ, Suzuki M. Evidence against essential roles for subdomain 1 of actin in actomyosin sliding movements. Biochem Biophys Res Commun 2005;332:474–81.[ISI][Medline]
(20) Holmes KC, Angert I, Kull FJ, Jahn W, Schroder RR. Electron cryo-microscopy shows how strong binding of myosin to actin releases nucleotide. Nature 2003;425:423–7.[CrossRef][Medline]
(21) van Wijk E, Krieger E, Kemperman MH, De Leenheer EM, Huygen PL, Cremers CW, et al. A mutation in the gamma actin 1 (ACTG1) gene causes autosomal dominant hearing loss (DFNA20/26). J Med Genet 2003;40:879–84.
(22) Zhu M, Yang T, Wei S, DeWan AT, Morell RJ, Elfenbein JL, et al. Mutations in the gamma-actin gene (ACTG1) are associated with dominant progressive deafness (DFNA20/26). Am J Hum Genet 2003;73:1082–91.[CrossRef][ISI][Medline]
(23) Giannakakou P, Sackett DL, Kang YK, Zhan Z, Buters JT, Fojo T, et al. Paclitaxel-resistant human ovarian cancer cells have mutant beta-tubulins that exhibit impaired paclitaxel-driven polymerization. J Biol Chem 1997;272:17118–25.
(24) Poruchynsky MS, Kim JH, Nogales E, Annable T, Loganzo F, Greenberger LM, et al. Tumor cells resistant to a microtubule-depolymerizing hemiasterlin analogue, HTI-286, have mutations in alpha- or beta-tubulin and increased microtubule stability. Biochemistry 2004;43:13944–54.[CrossRef][Medline]
(25) Masibay AS, Qasba PK, Sengupta DN, Damewood GP, Sreevalsan T. Cell-cycle-specific and serum-dependent expression of gamma-actin mRNA in Swiss mouse 3T3 cells. Mol Cell Biol 1988;8:2288–94.
(26) Grummt I. Actin and myosin as transcription factors. Curr Opin Genet Dev 2006;16:191–6.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||