Genetic complementation of cytokine signaling identifies central role of kinases in hematopoietic cell proliferation


Molecular evidence suggests a multistep process in the development of acute leukemia. Since inappropriate activation of cytokine signaling cascades is a recurring theme in human leukemia, we performed expression screens to identify genes that transform cytokine-dependent cells. Using retroviral cDNA libraries derived from peripheral blood mononuclear cells of patients with myeloproliferative disorders, we isolated numerous genes that genetically complement cytokine requirements for proliferation of BaF/3 and TF-1 cells. The majority of recovered genes represent members of the kinase family, including several previously linked to leukemogenesis. Our unbiased screen highlights the central role of kinase activation in hematopoietic cell proliferation and identifies a number of potential leukemic oncoproteins.


Significant insights into the molecular basis of human leukemias have been gained through cytogenetics and the subsequent characterization of aberrant protein products of chromosomal translocations. The BCR/ABL protein of human chronic myeloid leukemia (CML) is the quintessential translocation product – a fusion protein resulting in the constitutive activation of a tyrosine kinase (Groffen et al., 1984; Ben-Neriah et al., 1986). In murine models, the BCR/ABL oncogene is sufficient to induce a myeloproliferative syndrome (Daley et al., 1990; Li et al., 1999). The natural evolution of CML is the eventual development of blast crisis – an event resembling acute leukemia in which immature myeloid progenitors predominate. Presumably, this transition entails the accumulation of additional genetic lesions within the leukemic clone (Dash et al., 2002). This multistep process for the development of acute leukemias coincides with the prevailing paradigm that multiple genetic events are required for solid tumor carcinogenesis (Vogelstein et al., 1988).

Numerous translocations products of activated kinases have now been identified in bone marrow disorders, including HIP1-PDGFR in atypical CML, ZNF198-FGFR in atypical CML, TEL-PDGFR in chronic myelomonocytic leukemia, TEL-ABL in acute myeloid leukemia (AML), and TEL-JAK in T-cell acute lymphocytic leukemia. All these translocation products have similar structural motifs as the BCR/ABL oncogene: an N-terminal oligomerization domain and C-terminal constitutively activated kinase domain, suggesting that kinase activation is a common theme. Furthermore, a unifying principle for all of these activated kinases is the capacity to transform cytokine-dependent hematopoietic cell lines (Golub et al., 1994, 1996; Papadopoulos et al., 1995; Carroll et al., 1996; Reiter et al., 1998; Ho et al., 1999; Kulkarni et al., 2000). These observations imply that these activated kinases can functionally complement cytokine signaling (Koh and Daley, 2001). In addition to these activated kinases, Onishi et al. (1998) have demonstrated that a constitutively active transcription factor (STAT5-signal transducers and activators of transcription) is sufficient to transform cytokine-dependent hematopoietic cell lines.

Realizing the importance that inappropriate activation of cytokine signaling cascades has in the development of human leukemias, we performed genomewide screens to identify genes whose expression would transform cytokine-dependent cells. Speculating that cells derived from patients with myeloproliferative disorders (MPDs) would be enriched for expression of genes important for cytokine signal transduction, we generated several cDNA libraries in the pEYK retroviral vectors (Koh et al., 2002), using RNA isolated from peripheral blood mononuclear cells (PBMCs) of patients with MPDs. We identified a set of genes involved prominently in cytokine signaling pathways. The majority of the identified genes are kinases, some of which have already been implicated in the leukemogenic process, while others represent novel candidates for involvement in myeloproliferative disease.


Performance of the screen and recovery of candidates

Utilizing RNA from PBMCs of patients with either polycythemia vera (PV) or essential thrombocythemia, we created several cDNA libraries in pEYK1, pEYK2.1, and pEYK3.1 retroviral vectors. We have previously established the efficiency and selection parameters of these vectors for cDNA screening and gene isolation in cytokine-dependent cells (Koh et al., 2002). By cotransfecting 293T cells with the cDNA libraries and the packaging construct pCL-Eco (Naviaux et al., 1996), we generated retroviral supernatants and used them to infect the murine IL-3-dependent cell line BaF/3 (Palacios and Steinmetz, 1985) and the human GM-CSF-dependent TF-1 cell line (Kitamura et al., 1989). After 2 days, we imposed cytokine withdrawal on the infected cells. Robust cytokine-independent populations emerged within 10–14 days of growth in liquid culture. As a control for the background rate of induced and spontaneous cytokine independence, populations infected with a green fluorescent protein retrovirus were subjected to the same selection process. Under these conditions, factor-independent populations of cells were not observed within 14 days, but eventually emerged in all cultures after a prolonged period (>4 weeks), suggesting a modest but tolerable background of spontaneous factor independence in the screen (Koh et al., 2002).

In order to distinguish between spontaneously factor-independent cells and cells that had acquired a cDNA that would allow the cells to proliferate in the absence of cytokines, we employed an iterative process to verify that the cytokine-independent phenotype was a heritable trait due to a specific cDNA (Figure 1). In our initial experiments, we reisolated the c-FMS and c-FES genes numerous times in independent screens (Table 1). We hypothesized that these genes were conferring a high proliferative rate, and that our conditions of selection in bulk liquid culture biased our screen towards genes that would outcompete slower growing cells. To overcome this biased selection process, we modified our procedures to screen sublibraries that consisted of fewer independent cDNAs, and we imposed selection on small independent pools of cells. We size fractionated the cDNAs and created libraries representing sequences of 1–3 kb and greater than 3 kb. In addition, we plated cells in 96-well dishes immediately following infection, effectively providing for outgrowth of individual clones. Through these modifications, we were able to isolate a number of additional genes that conferred factor-independent growth (PRH, HLX, and TYK2, Table 1). As shown below, the frequency of detection in our screens corresponded to the relative proliferation rate of BaF/3 cells expressing individual genes.

Figure 1

(A) A schema for sequential iteration of the functional screen. The functional screens were initiated by infecting cytokine-dependent cells with retroviral supernatants. The screens were repeated 2–3 times through the generation of sublibraries for pEYK2.1 and pEYK3.1 vectors or the utilization of helper virus for pEYK1 as described (Koh et al., 2002). (B) Representative enrichment of cDNA candidates following screen iteration. (1) For pEYK1, three rounds of iteration enriched for a specific cDNA, as detected by PCR amplification. Lanes 1–3 represent PCR amplification of genomic DNA derived from infected cells from primary, secondary, and tertiary screens, respectively. Sequence analysis identified the recovered cDNA as the c-FES gene. (2) For pEYK2.1, restriction enzyme analyses of recovered proviruses after two rounds of screening yielded several clones. The clones from lanes 1 and 3 contain the c-FES cDNA, while the clone in lane 2 consists of a cDNA that was unable to confer cytokine-independent proliferation in subsequent screens. (3) For pEYK3.1, the restriction analyses of sublibraries from two rounds of screening resulted in the recovery of multiple cDNAs. Lanes L, 1, and 2 represent the primary library, the primary screen, and the secondary screen respectively. In the next panel, restriction enzyme analyses of recovered proviruses from the tertiary screen reveal the isolation of TYK2 in lane A. For all panels, Lane M represents the DNA molecular weight ladder

Table 1 Summary of the functions of the genes identified in the screen

Verification of cytokine-independent proliferation

To examine the proliferation-inducing capacity of each cDNA candidate that was isolated, cloned virus containing specific genes were introduced into murine BaF/3 and human TF-1 cells. All recovered genes (Table 1) generated factor-independent populations in both cell types. Utilizing BaF/3 cells expressing the BCR/ABL oncogene (B-BCR/ABL) as a positive control for cytokine-independent survival and proliferation (Daley and Baltimore, 1988), we assessed growth rates of the BaF/3 cells over 3 days (Figure 2). Of the genes isolated in our screen, the most rapid proliferation was conferred by c-FES and c-FMS, confirming our hypothesis about the bias in our initial screens.

Figure 2

Growth curve of BaF/3 cells expressing cDNAs that confer cytokine independence. Through retroviral-mediated gene transfer into BaF/3, factor-independent populations expressing individual cDNAs were generated: B-HLX, B-PRH, B-A-RAF-1, B-c-FES, B-c-FMS, B-COT, B-TYK2 B-MKK3b, and B-BCR/ABL. As a positive and negative control, BaF/3 cells were grown in the presence (5%) and absence (0%) of IL-3, respectively. Identical numbers of cells from each population in triplicates were plated on day 0 in the complete absence of IL-3 (except for the positive control). Viable cell counts were determined by trypan-blue exclusion on each of three subsequent days. For purposes of clarification, error bars were included only for the data corresponding to 72 h of growth

Cytokine-independent proliferation does not entail autocrine stimulation

Hematopoietic oncogenes, like BCR/ABL, have been linked to the induction of cytokine production for both autocrine and paracrine stimulation in certain cell lines and primary hematopoietic tissue (Hariharan et al., 1988; Koh and Daley, 2001; Peters et al., 2001). To ascertain if the above genes were generating autocrine or paracrine factors, conditioned media was isolated from BaF/3 cells expressing a specific cDNA. Parental BaF/3 cells were placed in 50% conditioned media isolated from each cell line. After 3 days, the viability was assayed by trypan-blue exclusion and fluorescence-activated cell sorter (FACS) analyses in triplicates. None of the conditioned media supported survival and proliferation of the parental BaF/3 cells (data not shown). These data establish that cytokine independence was not conferred through autocrine stimulation.

Genes recovered from screens

Our screens have identified several genes that figure prominently in cytokine receptor signaling (Table 1). We isolated the cytokine receptor c-FMS, two receptor-associated cytoplasmic tyrosine kinases, three MAP kinases, and two homeobox transcription factors.

Two of the identified genes had N-terminal truncations. As no genetic rearrangements of these genes were detected by Southern analyses of genomic DNA from the patients whose RNA had been used in library generation, and more 5′ gene sequences could be amplified from the cDNA preparations (data not shown), the truncations were most likely artifacts of incomplete cDNA syntheses during the generation of the libraries that emerged in the context of the functional screen. For the truncated form of the A-RAF-1 (MAP kinase kinase kinase (MKKKs or MEKKs)) gene, two different cDNAs were isolated that differed at the 5′ end but shared the same ATG, encoding a 218 amino-acid N-terminal truncation. This N-terminal truncation still encoded an intact kinase domain. Two independent screens from the same library identified the same N-terminal truncation of TYK2, a member of the Janus family of kinases. With 794 amino acids missing, this TYK2 truncation encoded an intact tyrosine kinase domain, but missing the adjacent kinase-like domain. Similar to the A-RAF-1 truncations, the isolated truncation of TYK2 would be predicted to result in a constitutively active kinase.

Both c-FMS (receptor tyrosine kinase (RTK)) and c-FES (non-RTK) cDNAs were isolated numerous times in independent screens with different libraries. The c-FMS and c-FES cDNAs isolated from the screen were sequenced. Other than mutations that were not consistent between independent clones and were thus likely introduced during PCR amplification, there were no differences from wild-type genes. Subsequently, we confirmed that overexpression of both wild-type c-FMS or c-FES genes allowed both BaF/3 and TF-12 cells to proliferate without the addition of cytokines (data not shown).

The COT kinase (TPL2, MKKKs or MEKKs) cDNA was isolated twice in two independent screens utilizing the same library. The two isolated cDNAs were identical at the 5′ untranslated region, suggesting recovery of the same cDNA synthesis product. The COT kinase cDNA isolated in the screen encoded the shorter 56 kDa isoform.

The MKK3b (MAP kinase kinase) gene was isolated twice from two different libraries. Sequence analyses of one of the isolated MKK3b cDNA revealed no point mutations or deletions. The MKK3b is a splice variant of MKK3, encoding an extra 29 amino acids at the N-terminus that is more efficient in activating downstream signaling targets (Han et al., 1997).

Unlike the other genes, which encoded kinases, the PRH and HLX cDNAs were identified as transcription factors that, upon overexpression, were able to render cytokine-dependent cell lines factor independent. These two transcription factors belong to the homeobox family of genes – a set of genes previously linked to developmental fate rather than survival and proliferative capacity (Magli et al., 1997; Cillo et al., 1999). Both cDNAs encoded full-length proteins, and sequence analyses of both genes corresponded to the reported wild-type sequences.

Screen for survival factors

All the previous genes provide for both cytokine-independent survival and proliferation. In order to identify genes that might confer only a survival signal, we modified the screen. In our initial endeavors with this screening strategy in BaF/3 and TF-1 cell lines, we were only able to reisolate genes that would fully transform cells to factor-independent proliferation. Subsequently, we utilized 32D cells (Greenberger et al., 1983), a murine IL-3-dependent hematopoietic cell line, because they were less prone to transformation to cytokine-independent proliferation when compared to BaF/3 and TF-1 cells. At 2 days after retroviral infection of the IL-3-dependent 32D cells, we withdrew cytokine for a 2-week period and then rescued surviving cells with IL-3. With this modification in the screen, we recovered BCL-XL, a known antiapoptotic protein (Table 1).


Utilizing retroviral cDNA libraries from patients with MPDs, we set out to identify genes whose expression would complement cytokine signaling. We identified several genes that figure prominently in cytokine receptor signaling pathways, including the c-FMS RTK, genes implicated in MAP kinase cascades (COT kinase, MKK3b, and N-terminal-truncated A-RAF-1), and genes that are linked to the JAK/STAT pathway (the nonreceptor-tyrosine kinases c-FES and the N-terminal-truncated TYK2) (Figure 3). We also recovered two homeobox genes (PRH and HLX) that have never been implicated in cytokine signal transduction but have been linked to transformation of hematopoietic cells (Magli et al., 1997; Cillo et al., 1999). The striking tendency to recover kinases in our global screen suggests a central role for this class of signaling proteins in stimulating hematopoietic cell survival and proliferation.

Figure 3

Role of recovered genes in cytokine signaling cascades. Genes isolated in the phenotypic screens (designated in red) act at various stages of signal transduction pathways – from cell-surface RTKs to transcription factors (figure adapted from Lewis et al., 1998)

Recent experiments have modeled the notion that two genetic lesions can suffice to induce acute leukemia in mice (Dash et al., 2002; Deguchi and Gilliland, 2002). One event entails expression of constitutively activated kinases, while the other event entails expression of a deregulated transcription factor. These lesions provide a proliferative stimulus and a block in differentiation, respectively, and cooperate to induce acute leukemia.

Many of the genes that we isolated have already been implicated in human bone marrow disorders (Table 1). Previously, point mutations in the c-FMS gene have been detected in human myelodysplastic syndromes and AMLs (Tobal et al., 1990). Several labs have demonstrated that the overexpression of wild-type c-FMS in the hematopoietic cell line FDC-P1 could either induce complete cytokine-independent growth or sensitize cells to suboptimal levels of GM-CSF, IL-3, or IGF-1 (Kato et al., 1989; Bourette et al., 1992; Baker et al., 1994). McArthur et al. (1994) retrovirally infected murine fetal liver cells with c-fms. In these experiments, they found a preferential expansion of erythroid colonies in the presence of CSF-1, highly reminiscent of the predominant outgrowth of the erythroid lineage found in PV, one of the MPD subtypes. Despite our efforts, however, we have been unable to link point mutation or overexpression of c-FMS to the pathogenesis of PV.

c-FES is a non-RTK expressed exclusively in the myeloid lineage (Yates and Gasson, 1996) and implicated in the JAK-STAT signaling pathway. Nelson et al. (1998) has shown that overexpression of c-FES can cause phosphorylation of STAT3 and activation of its DNA-binding activity. c-FES was first isolated in the context of a transforming retrovirus (Trus et al., 1982), but has not previously been linked to human leukemia. Given its capacity to drive hematopoietic cell proliferation, c-FES is an interesting candidate gene for inducing bone marrow proliferative disorders.

TYK2 is another isolated candidate linked to the JAK-STAT pathway. Of the four Janus kinases, TYK2 seems to play a critical role in the development and maturation of the immune system (Shimoda et al., 2000). Interestingly, Tabrizi and colleagues observed that the pathogenesis of familial hemophagocytic lymphohistiocytosis (an autosomal recessive disorder of histiocytosis) involves an inability to downregulate the activity of TYK2 by the protein tyrosine phosphatase SHP-1 (Tabrizi et al., 1998). Spiekermann and colleagues have further demonstrated that constitutive activation of the JAK-STAT pathway contributes to the pathogenesis of acute leukemia (Spiekermann et al., 2001). To date, activation of TYK2 has not been associated with human leukemia.

Three of our recovered genes are elements of MAP kinase cascades. Two of the candidate genes are MKKKs or MEKKs: A-RAF-1 and COT kinase (Tpl2) (Lewis et al., 1998). The RAF proteins are direct downstream targets of RAS that mediate proliferative signals from extracellular mitogens. Our isolation of N-terminal-truncated A-Raf-1 recapitulates a previous report that truncated forms of RAF proteins that could render TF-1 cells cytokine independent (McCubrey et al., 1998). Retroviral infection with a gag-A-RAF-1 fusion protein induced the development of sarcomas and erythroleukemias in mice (Beck et al., 1987). This specificity of transforming the erythroid lineage is again reminiscent of the predominant erythroid expansion in PV. Two groups identified COT kinase in a functional cloning strategy for genes that transform fibroblasts (Miyoshi et al., 1991; Chan et al., 1993). This gene has not previously been implicated in hematopoietic transformation. As a mediator of the SAPK pathways via p38, MKK3b is linked to apoptosis (Hanada et al., 1998). The role of MKK3b in cytokine signaling, however, remains elusive as is the mechanism of rendering cytokine-dependent cells factor independent.

The homeobox transcription factors (HOX genes) have traditionally been assigned a role in embryonic development (Magli et al., 1997; Cillo et al., 1999), but more recently have also been identified as mediators of proliferation and differentiation of the hematopoietic system (Sauvageau et al., 1994; Magli et al., 1997; Kyba et al., 2002). The proline-rich homeodomain protein PRH is expressed in the myeloid, erythroid, and B-cell lineages, but not in T cells (Crompton et al., 1992; Manfioletti et al., 1995), and is upregulated in certain murine B-cell leukemias (Hansen and Justice, 1999). The HLX gene is expressed in both myelomonocytic (macrophages and granulocytes) and B-cell lineages, but is not expressed in T cells, mast cells, or erythroid cells (Allen et al., 1991). Ectopic expression of the murine Hlx gene in the cytokine-dependent cell line FDC-P1 induced myeloid maturation (Allen and Adams, 1993). However, the ability of the HLX gene to induce cytokine-independent growth was neither examined nor reported. Interestingly, Deguchi et al. (1992) have demonstrated that overexpression of human HLX (HB24) gene in the human Jurkat T-cell line induced genes involved in T-cell growth and activation, including c-FOS, c-MYC, c-MYB, LCK, NF-κB, and interleukin-2 (IL-2). On the other hand, transgenic mice overexpressing the murine Hlx gene in lymphoid tissues had development defects in both T- and B-cell lineages, rather than a leukemogenic phenotype (Allen et al., 1995). The particular mechanisms linking PRH and HLX to cytokine-independent hematopoietic proliferation remain to be defined.

Using Southern hybridization to rule out gene rearrangement and gene sequencing to rule out mutation, we have not found evidence that any of the genes we recovered in our screen were mutated in the patients we surveyed, and thus we can only speculate whether these pathways are directly involved in the molecular pathogenesis of MPDs. While our screen might have yielded a similar set of genes had we generated our cDNA libraries from blood cells of normal individuals, it is likely that the abnormal cells from patients with MPDs overexpress multiple genes in the pathways most relevant to hematopoietic cell proliferation, and that the libraries thus represent an enriched resource.

Our identification of numerous genes that bypass cytokine signaling validates the utility of an expression cloning strategy for discovering genes that drive blood cell proliferation, and might therefore contribute to the development of acute leukemia. The predominance of kinases recovered in our screen reinforces the notion that overexpression or mutation of this class of signaling proteins plays a central role in bone marrow proliferative disorders. By introducing retroviral cDNA libraries in hematopoietic cells that can be induced to differentiate (e.g. HL-60, U937), and selecting for abrogation of cell differentiation, one could exploit cDNA expression cloning to identify genes responsible for blocking differentiation (Starr et al., 1997). Together, these two strategies should identify crucial genes in hematopoietic proliferation, differentiation, and leukemogenesis.

Materials and methods

Cell culture

Culturing of the murine IL-3-dependent BaF/3 and 32D cells has been described previously (Klucher et al., 1998). The human GM-CSF-dependent TF-1 cells (Kitamura et al., 1989) were engineered to express the ecotropic receptor (Baker et al., 1992), and the resulting cell line eco-TF-1 had similar infection rates as the BaF/3 cells (data not shown). The eco-TF-1 cells were grown in RMPI 1640 containing 10% fetal calf serum (FCS) and 4 ng/ml of human GM-CSF (Peprotech). 293 T cells (DuBridge et al., 1987) were grown in DMEM containing 10% FCS and penicillin–streptomycin (10 U/mL and 10 μg/mL). All cells were incubated at 37°C with 5% CO2.

Library construction

PBMCs were isolated from patients with a diagnosed MPD who presented to the blood bank at the Massachusetts General Hospital for either phlebotomy or therapeutic pheresis. Samples were treated with RNA-STAT-60 (Tel-Test, Friendswood, Inc., TX, USA) to isolate total RNA. PolyA RNA was isolated using the Oligotex kit (Qiagen). The Superscript Choice System (Invitrogen) was used for cDNA synthesis. BstX1/EcoR1 adaptors (Invitrogen) were ligated to the cDNA using T4 DNA ligase (New England Biolabs). The cDNA was then size fractionated on a low-melt agarose gel (American Bioanalytical) into two fractions: >3 and 1–3 kb. Each fraction was purified with agarase treatment according to the manufacturer's protocol (Roche). The cDNA was ligated into the nonpalindromic BstXI sites in the purified BstXI-digested pEYK vectors. The ligation mixture was purified with phenol–chloroform extraction and electroporated into Electromax DH10B (Invitrogen). For amplification of the library, the bacteria were plated on LB plates containing 100 μg/ml of ampicillin and/or 50 μg/ml of zeocin (Invitrogen). After incubating at 37°C for 16–18 h, the plates were scraped, and the plasmid DNA was isolated using Qiagen Megaprep columns (Qiagen).

Retroviral generation and infections

The generation of retroviral supernatants and subsequent infections with target cells have been described previously (Koh et al., 2002).

Selection and iterative process

The infected cells were allowed to grow for 48 h. For selection, the cells were washed twice with PBS and placed in media without growth factors (RPMI, 10% FCS). Genomic DNA was isolated from the factor-independent population, using standard procedures of SDS–proteinase K treatment (Maniatis et al., 1989). For the iterative process of pEYK1-based libraries, the factor-independent population was infected with wild-type Moloney virus (helper virus) in order to mobilize the viruses. After 2 days of infection, the medium was changed. The resulting viral supernatant was collected 24 h later. As a secondary screen, fresh uninfected cells were spin infected and selected 2 days later. The mobilization procedure was repeated for a tertiary and final screen. For pEYK1, PCR-based recovery was utilized to amplify specific cDNAs (Koh et al., 2002). For the pEYK2.1- and pEYK3.1-based libraries, sublibraries were generated from transformation of rescued proviruses as described previously (Koh et al., 2002). The sublibraries were then used to generate retroviral supernatants.



myeloproliferative disorders


fluorescence-activated cell sorter




  1. Allen JD and Adams JM . (1993). Blood, 81, 3242–3251.

  2. Allen JD, Harris AW, Bath ML, Strasser A, Scollay R and Adams JM . (1995). J. Immunol., 154, 1531–1542.

  3. Allen JD, Lints T, Jenkins NA, Copeland NG, Strasser A, Harvey RP and Adams JM . (1991). Genes Dev., 5, 509–520.

  4. Aoki M, Hamada F, Sugimoto T, Sumida S, Akiyama T and Toyoshima K . (1993). J. Biol. Chem., 268, 22723–22732.

  5. Baker BW, Boettiger D, Spooncer E and Norton JD . (1992). Nucleic Acids Res., 20, 5234.

  6. Baker DA, Glover HR and Dibb NJ . (1994). Leukemia, 8, 141–150.

  7. Beck TW, Huleihel M, Gunnell M, Bonner TI and Rapp UR . (1987). Nucleic Acids Res., 15, 595–609.

  8. Ben-Neriah Y, Daley GQ, Mes-Masson AM, Witte ON and Baltimore D . (1986). Science, 233, 212–214.

  9. Boudard D, Vasselon C, Bertheas MF, Jaubert J, Mounier C, Reynaud J, Viallet A, Chautard S, Guyotat D and Campos L . (2002). Am. J. Hematol., 70, 115–125.

  10. Bourette RP, Von Ruden T, Ballmer-Hofer K, Morle F, Blanchet JP and Mouchiroud G . (1992). Growth Factors, 7, 315–325.

  11. Carroll M, Tomasson MH, Barker GF, Golub TR and Gilliland DG . (1996). Proc. Natl. Acad. Sci. USA, 93, 14845–14850.

  12. Chan AM, Chedid M, McGovern ES, Popescu NC, Miki T and Aaronson SA . (1993). Oncogene, 8, 1329–1333.

  13. Cillo C, Faiella A, Cantile M and Boncinelli E . (1999). Exp. Cell. Res., 248, 1–9.

  14. Crompton MR, Bartlett TJ, MacGregor AD, Manfioletti G, Buratti E, Giancotti V and Goodwin GH . (1992). Nucleic Acids Res., 20, 5661–5667.

  15. Daley GQ and Baltimore D . (1988). Proc. Natl. Acad. Sci. USA, 85, 9312–9316.

  16. Daley GQ, Van Etten RA and Baltimore D . (1990). Science, 247, 824–830.

  17. Dash AB, Williams IR, Kutok JL, Tomasson MH, Anastasiadou E, Lindahl K, Li S, Van Etten RA, Borrow J, Housman D, Druker B and Gilliland DG . (2002). Proc. Natl. Acad. Sci. USA, 99, 7622–7627.

  18. Deguchi K and Gilliland DG . (2002). Leukemia, 16, 740–744.

  19. Deguchi Y, Thevenin C and Kehrl JH . (1992). J. Biol. Chem., 267, 8222–8229.

  20. DuBridge RB, Tang P, Hsia HC, Leong PM, Miller JH and Calos MP . (1987). Mol. Cell. Biol., 7, 379–387.

  21. Golub TR, Barker GF, Lovett M and Gilliland DG . (1994). Cell, 77, 307–316.

  22. Golub TR, Goga A, Barker GF, Afar DE, McLaughlin J, Bohlander SK, Rowley JD, Witte ON and Gilliland DG . (1996). Mol. Cell. Biol., 16, 4107–4116.

  23. Greenberger JS, Sakakeeny MA, Humphries RK, Eaves CJ and Eckner RJ . (1983). Proc. Natl. Acad. Sci. USA, 80, 2931–2935.

  24. Groffen J, Stephenson JR, Heisterkamp N, de Klein A, Bartram CR and Grosveld G . (1984). Cell, 36, 93–99.

  25. Hampe A, Laprevotte I, Galibert F, Fedele LA and Sherr CJ . (1982). Cell, 30, 775–785.

  26. Han J, Wang X, Jiang Y, Ulevitch RJ and Lin S . (1997). FEBS Lett., 403, 19–22.

  27. Hanada M, Kobayashi T, Ohnishi M, Ikeda S, Wang H, Katsura K, Yanagawa Y, Hiraga A, Kanamaru R and Tamura S . (1998). FEBS Lett., 437, 172–176.

  28. Hansen GM and Justice MJ . (1999). Oncogene, 18, 6531–6539.

  29. Hariharan IK, Adams JM and Cory S . (1988). Oncogene Res., 3, 387–399.

  30. Ho JM, Beattie BK, Squire JA, Frank DA and Barber DL . (1999). Blood, 93, 4354–4364.

  31. Kato JY, Roussel MF, Ashmun RA and Sherr CJ . (1989). Mol. Cell. Biol., 9, 4069–4073.

  32. Kitamura T, Tange T, Terasawa T, Chiba S, Kuwaki T, Miyagawa K, Piao YF, Miyazono K, Urabe A and Takaku F . (1989). J. Cell Physiol., 140, 323–334.

  33. Klucher KM, Lopez DV and Daley GQ . (1998). Blood, 91, 3927–3934.

  34. Koh EY, Chen T and Daley GQ . (2002). Nucleic Acids Res., 30, e142.

  35. Koh EY and Daley GQ . (2001). Chronic Myeloid Leukaemia: Biology and Treament, Carella AM, Daley GQ, Eaves CJ, Goldman JM and Hehlmann R (eds). Martin Dunitz: London, pp. 56–72.

    Google Scholar 

  36. Kulkarni S, Heath C, Parker S, Chase A, Iqbal S, Pocock CF, Kaeda J, Cwynarski K, Goldman JM and Cross NC . (2000). Cancer Res., 60, 3592–3598.

  37. Kyba M, Perlingeiro RC and Daley GQ . (2002). Cell, 109, 29–37.

  38. Lewis TS, Shapiro PS and Ahn NG . (1998). Adv. Cancer Res., 74, 49–139.

  39. Li S, Ilaria RL, Jr, Million RP, Daley GQ and Van Etten RA . (1999). J. Exp. Med., 189, 1399–1412.

  40. Magli MC, Largman C and Lawrence HJ . (1997). J. Cell. Physiol., 173, 168–177.

  41. Manfioletti G, Gattei V, Buratti E, Rustighi A, De Iuliis A, Aldinucci D, Goodwin GH and Pinto A . (1995). Blood, 85, 1237–1245.

  42. Maniatis T, Fritsch EF and Sambrook J . (1989). Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory: Cold Spring Harbor, NY.

    Google Scholar 

  43. McArthur GA, Rohrschneider LR and Johnson GR . (1994). Blood, 83, 972–981.

  44. McCubrey JA, Steelman LS, Hoyle PE, Blalock WL, Weinstein-Oppenheimer C, Franklin RA, Cherwinski H, Bosch E and McMahon M . (1998). Leukemia, 12, 1903–1929.

  45. Miyoshi J, Higashi T, Mukai H, Ohuchi T and Kakunaga T . (1991). Mol. Cell. Biol., 11, 4088–4096.

  46. Naviaux RK, Costanzi E, Haas M and Verma IM . (1996). J. Virol., 70, 5701–5705.

  47. Nelson KL, Rogers JA, Bowman TL, Jove R and Smithgall TE . (1998). J. Biol. Chem., 273, 7072–7077.

  48. Onishi M, Nosaka T, Misawa K, Mui AL, Gorman D, McMahon M, Miyajima A and Kitamura T . (1998). Mol. Cell. Biol., 18, 3871–3879.

  49. Palacios R and Steinmetz M . (1985). Cell, 41, 727–734.

  50. Papadopoulos P, Ridge SA, Boucher CA, Stocking C and Wiedemann LM . (1995). Cancer Res., 55, 34–38.

  51. Peters DG, Klucher KM, Perlingeiro RC, Dessain SK, Koh EY and Daley GQ . (2001). Oncogene, 20, 2636–2646.

  52. Reiter A, Sohal J, Kulkarni S, Chase A, Macdonald DH, Aguiar RC, Goncalves C, Hernandez JM, Jennings BA, Goldman JM and Cross NC . (1998). Blood, 92, 1735–1742.

  53. Sauvageau G, Lansdorp PM, Eaves CJ, Hogge DE, Dragowska WH, Reid DS, Largman C, Lawrence HJ and Humphries RK . (1994). Proc. Natl. Acad. Sci. USA, 91, 12223–12227.

  54. Shimoda K, Kato K, Aoki K, Matsuda T, Miyamoto A, Shibamori M, Yamashita M, Numata A, Takase K, Kobayashi S, Shibata S, Asano Y, Gondo H, Sekiguchi K, Nakayama K, Nakayama T, Okamura T, Okamura S and Niho Y . (2000). Immunity, 13, 561–571.

  55. Spiekermann K, Biethahn S, Wilde S, Hiddemann W and Alves F . (2001). Eur. J. Haematol., 67, 63–71.

  56. Starr R, Willson TA, Viney EM, Murray LJ, Rayner JR, Jenkins BJ, Gonda TJ, Alexander WS, Metcalf D, Nicola NA and Hilton DJ . (1997). Nature, 387, 917–921.

  57. Tabrizi M, Yang W, Jiao H, DeVries EM, Platanias LC, Arico M and Yi T . (1998). Leukemia, 12, 200–206.

  58. Tobal K, Pagliuca A, Bhatt B, Bailey N, Layton DM and Mufti GJ . (1990). Leukemia, 4, 486–489.

  59. Trus MD, Sodroski JG and Haseltine WA . (1982). J. Biol. Chem., 257, 2730–2733.

  60. Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC, Leppert M, Nakamura Y, White R, Smits AM and Bos JL . (1988). N. Engl. J. Med., 319, 525–532.

  61. Yates KE and Gasson JC . (1996). Stem Cells, 14, 117–123.

Download references


We acknowledge the efforts of Phil Waite and the members of the Blood Donor Center at the Massachusetts General Hospital in providing patient samples. This work was supported by grants from the NIH (CA76418 and CA86991). GQD is the Birnbaum Scholar of the Leukemia and Lymphoma Society of America. EYK was supported by NIH training Grants (GM07753-22 and CA09541).

Author information



Corresponding author

Correspondence to George Q Daley.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Koh, E., Chen, T. & Daley, G. Genetic complementation of cytokine signaling identifies central role of kinases in hematopoietic cell proliferation. Oncogene 23, 1214–1220 (2004).

Download citation


  • expression cloning
  • cytokine signaling
  • tyrosine kinase
  • myeloproliferative disease

Further reading


Quick links