Editorials and Perspectives their name to legitimize something they cannot really Journals clearly have a role to play in identifying and cor- vouch for. Though, more often than not, authors get away recting the most egregious examples. And pharmaceutical with it, authors should be very wary of being guests. In a companies must accept that trying to hide ghost, or entice number of recent high profile cases of fraud some senior guest authors is not acceptable. But the primary responsi- authors have protested that they knew nothing about the bility for prevention lies much further back, within the alleged misconduct, only to have their protestations met institutions where authors work, and where medical aca- with incredulity.9 It is a forceful reminder that authorship is demics are trained. Authorship of a scientific or medical a responsibility not just a privilege and should be offered or paper must be returned to something that can be a source accepted with appropriate consideration. of pride, and which is deserved and earned – and declared. What can be done? If there is one lesson to be learned from all these cases of Virginia Barbour is Chief Editor at PLoS Medicine, the open- ghost, guests and other illegitimate authors it is that access general medical journal published by the Public Library of authorship has somehow slipped recently from something Science (PLoS). She trained in hematology before moving into to be earned through a specific, meaningful contribution to medical publishing in 1999. a superficial designation that can be traded. These slippery The author thanks Larry Peiperl and Gavin Yamey for very notions of authorship have not come out of a vacuum, and helpful comments on this editorial. were certainly not invented by pharmaceutical companies. The issue of guest authorship – of including the head of References department among the list of authors of a paper for no 1. Dunbar CE, Tallman MS. 'Ghostbusting' at Blood. Blood 2009; other reason than esteemed status – has been widespread 113(3):502-3. in academia for many years. I would argue that it is this 2. Flanagin A, Carey LA, Fontanarosa PB, Phillips SG, Pace BP, Lundberg culture that pharmaceutical companies have tapped into, GD, Rennie D. Prevalence of articles with honorary authors and rather than inventing a new type of author. But by flatter- ghost authors in peer-reviewed medical journals. JAMA. 1998; 280(3):222-4. ing academics into being guest authors, they have created, 3. Gøtzsche PC, Hróbjartsson A, Johansen HK, Haahr MT, Altman DG, and then filled, a need for ghost authors to actually write Chan AW. Ghost authorship in industry-initiated randomised trials. the papers. The academics accepting the apparent honor of PLoS Med. 2007;4(1): e19. 4. Ross JS, Hill KP, Egilman DS, Krumholz HM. Guest authorship and authorship thus provide cover – as accomplices or as dupes ghostwriting in publications related to rofecoxib: a case study of – for manipulative marketing practices. industry documents from rofecoxib litigation. JAMA. 2008;299(15): Some have interpreted the anxiety over ghost authors as 1800-12. 5. Psaty BM, Kronmal RA. Reporting mortality findings in trials of rofe- a call to remove all medical writers from papers but that is coxib for Alzheimer disease or cognitive impairment: a case study not the case. Medical writers do have a role to play in writ- based on documents from rofecoxib litigation. JAMA. 2008; ing papers, but somehow, as we have argued before,10 299(15):1813-7. without appropriate standards this legitimate role can be 6. http://www.plosmedicine.org/static/ghostwriting.action. Accessed Oct 28th 2009. turned into something that subverts and threatens medical 7. http://blogs.wsj.com/health/2009/08/20/caspper-glaxosmithklines- publishing more widely. friendly-ghostwriting-program/ Accessed Oct 28th 2009. It is clear then that the responsibility for addressing the 8. http://www.icmje.org/ethical_1author.html 9. http://www.nytimes.com/2006/02/11/science/11clone.html mismatch between what an author should be and what 10. The PLoS Medicine Editors. Ghostwriting: the dirty little secret of authorship has come to mean lies with many groups. medical publishing that just got bigger. PLoS Med. 2009;6: e1000156. Rac GTPases play multiple roles in erythropoiesis Peng Ji1 and Harvey F. Lodish1,2 1Whitehead Institute for Biomedical Research, Cambridge, MA; USA; 2Department of Biology, Massachusetts Institute of Technology Cambridge, MA, USA E-mail: [email protected]. doi:10.3324/haematol.2009.015511 he four members of the Rac family of GTPases – As shown by single- and double knock-outs of Rac Rac1, Rac2, Rac3 and RhoG – are members of the genes, the Rac GTPases play important roles in many TRho superfamily that regulates the organization, hematopoietic cells.7 Rac2 is specifically expressed in dynamics, and function of the actin cytoskeleton. Rac hematopoietic cells, and is directly involved in chemotaxis GTPases play significant roles in many cellular processes and superoxide production in neutrophils and macro- including migration, cytokinesis, lamellipodia formation, phages.3,8-11 In addition, Rac2, together with Rac1, mediates and cell polarity.1 Genetically modified mice deficient in B-cell receptor signaling pathways.12 T-cell activation is also each of the Racs are available;2-6 deficiency of Rac1 causes affected in Rac2-deficient mice13 and hematopoietic stem intrauterine death, whereas mice defective in Rac2, Rac3 or cells from Rac2-/- mice show defective long-term engraft- RhoG develop fairly normally. Rac proteins may have ment.14 redundant functions in certain types of cells and unique In contrast, Rac1 is ubiquitously expressed and plays functions in others. essential roles in several organ systems, including 2 haematologica | 2010; 95(1) Editorials and Perspectives hematopoietic cells. Using a hematopoietic cell-specific Rac GTPases (Rac1, Rac2, Rac3? RhoG?) knockout of Rac1, Gu et al. demonstrated that depletion of Rac1 blocks the ability of hematopoietic stem cells to ? mDia2 mDia2/? engraft irradiated recipient mice.15 When crossed with Rac2-/- mice, a hematopoietic-specific deletion of Rac1 results in a massive egress of hematopoietic stem cells into Differentiation of Late erythroblast Integration of erythrocyte erythroid progenitors enucleation cytoskeleton the blood from the bone marrow.15 Rac regulates signaling pathways downstream of integrins and c-kit in mast cells and hematopoietic stem cells15-17 and presumably these defects cause loss of adhesion of the hematopoietic stem cells to the bone marrow stroma. Rac GTPases play essential roles in erythropoiesis Figure 1. Rac GTPases play essential roles in erythrocytes. There is growing evidence that Rac GTPases are essen- tial for erythropoiesis. In this issue of the Journal, Kalfa et al. demonstrate that Rac1 and Rac2 are required for early stages of erythropoiesis in the bone marrow but – surpris- mice,4-6 it is likely that other Rac GTPases family proteins ingly – not in the spleen.18 Using hematopoietic tissue-spe- compensate for any possible negative effects of deletion of cific Rac1 knock-out mice in a total Rac2-/- background, these proteins. In this aspect, in vitro studies of cultured they showed that deficiency of both Rac1 and Rac2 blocks mouse erythroblasts could elucidate whether Rac3 and early stages of erythropoiesis in the bone marrow without RhoG have functions in erythropoiesis. affecting cell survival or proliferation. Abnormalities in the Kalfa et al. show that the major disorder of Rac1-/-;Rac2-/- morphology of erythroid burst-forming units resembled hematopoietic cells is the defective proliferation of myelo- alterations seen previously in Rac1-/-;Rac2-/- myeloid erythroid progenitor cells in the bone marrow, but there is colonies, and suggested an impairment in cell migration a normal or possibly increased survival and/or proliferation and/or proliferation. These results indicated that Rac1 and of these progenitors in the spleen. Presumably, as noted by Rac2 have redundant but essential roles in the early ery- the authors, this derives from differences in the bone mar- throid progenitor stages of erythropoiesis in the bone mar- row and splenic microenvironment, but we do not know row. which cytokines or other factors produced by stromal cells The same group previously reported that Rac1-/-;Rac2-/- may be responsible for this crucial difference. Bone mar- erythrocytes have an unstable cytoskeleton; deficiency of row transplantation studies of Rac1-/-;Rac2-/- hematopoietic Rac1 and Rac2 alters actin assembly and decreases erythro- progenitors into normal recipients, and vice versa, could cyte deformability, and generates a hemolytic anemia.19 begin to separate the hematopoietic cell autonomous func- Thus, it was not surprising that in these Rac1-/- Rac2-/- mice tions of Rac1 and Rac2 from those of the stromal popula- there was compensatory erythropoiesis in the spleen. tions. What was surprising was that splenic erythroid progeni- We would also like to know how Rac1 and Rac2 regulate tors somehow circumvented the deficiency of the Rac1 differentiation of erythroid progenitors. As discussed by and Rac2 GTPases. the authors, cytokine-mediated signaling pathways may While Kalfa’s study focused on early stages of erythro- be involved. In other cell types Rac GTPases regulate gene poiesis, we showed that Rac1 and Rac2 are required for expression and cell transformation in multiple pathways, enucleation of late stage erythroblasts.20 Deregulation of such as in the Jun N-terminal kinase21,22 and Ras signaling Rac GTPases during the late stages of erythropoiesis com- pathways,23 and these or other pathways could mediate pletely blocked enucleation of cultured mouse fetal ery- the effects of Rac GTPases in early stages of erythropoiesis. throblasts without affecting their normal proliferation and differentiation. The contractile actin ring that forms on the Dr. Lodish is a Professor of Biology at the Massachusetts plasma membrane of late-stage erythroblasts at the bound- Institute of Technology and a member of the Whitehead Institute ary between the cytoplasm and nucleus of enucleating cells for Biomedical Research.