BCR-ABL1 mediates up-regulation of Fyn in chronic myelogenous leukemia

The management of chronic myelogenous leukemia (CML) has been revolutionized by kinase inhibitors that were developed in response to cues from biologic studies of the BCR-ABL1 oncogene. However two challenging problems persist: the progression of the disease to blast crisis and resistance to kinase inhibition.¹  Continued investigation of BCR-ABL1 kinase signaling will provide insight into these problems. Members of the Src kinase family, which regulate proliferation, differentiation, and motility,²  are known downstream targets of BCR-ABL1. In myeloid cell lines, BCR-ABL1 activates Lyn and Hck.^3,4  Several reports link growth, survival, and imatinib resistance of Philadelphia chromosome–positive (Ph⁺) leukemias to Lyn kinase expression and activation.^5,6  However, reports examining Fyn, a ubiquitously expressed Src family member, are sparse. Of note, phase-specific gene expression in CML using microarray analyses revealed that Fyn gene expression was linked to imatinib relapse.⁷  In addition, a separate study using combined systems biology and gene expression approaches in Ph⁺ acute lymphoblastic leukemia (ALL) specimens identified Fyn as a hub for signaling.⁸ 

Here we show that Fyn protein expression is increased in patients with blast-crisis CML compared with chronic-phase disease. By examining effects of silencing Fyn using shRNA, we find that Fyn transduces a mitogenic signal. Collectively, our results identify a novel effect of BCR-ABL1—up-regulation of Fyn—and delineate consequences of the observed up-regulation.

Patient specimens were used for this study and were collected after informed consent was obtained in accordance with the Declaration of Helsinki. The tissue microarray studies were initiated after approval from the University of Texas M. D. Anderson Cancer Center Institutional Review Board. Animal experiments were Institutional Animal Care and Use Committee–approved.

Antibodies were purchased from sources outlined in Document S1 (available on the Blood website; see the Supplemental Materials link at the top of the online article). Imatinib was kindly provided by Dr Elisabeth Buchdunger at Novartis Pharmaceuticals (Basel, Switzerland). Murine growth factor–dependent pro-B lymphoid BaF3 cell lines transformed with vector, wild-type BCR-ABL1, or imatinib-resistant mutant BCR-ABL1 were kindly provided by Dr Charles Sawyers⁹  and were cultured as previously described.⁹  K562 cells, TonB210 cells stably expressing a tetracycline-inducible BCR-ABL1 expression vector (kindly provided by Dr George Daley, Children's Hospital Boston, Harvard Medical School, MA),¹⁰  and mouse 32D and 32Dp210 cells were maintained in RPMI1640 medium with 10% FBS supplemented. Mouse 32D cells were supplemented with 10% WEHI-cultured conditioned medium as a source of interleukin-3 (IL-3) in addition to 10% FBS.

K562 cells were transfected with Fyn shRNA and control vectors (TranSilent human shRNA from Panomics, Redwood, CA) using the Nucleofector system kit V and transfection program T-16 (Amaxa Biosystems, Cologne, Germany). Lentiviral knockdown of Fyn and rescue design is detailed in Document S1. To generate the rescue construct, 4 nucleotides in shRNA target no. 1 and target no. 2 sequence regions in wild-type Fyn cDNA were replaced, thereby encoding the same amino acid as wild-type Fyn, but containing different nucleotide sequences in the Fyn shRNA target region.

Doubling time was measured either 24 hours or 48 hours after plating the indicated number of Fyn shRNA– or scrambled shRNA–containing cells; total cell numbers were counted using a Vi-Cell Viability Analyzer (Beckman Coulter, Fullerton, CA). Clonogenicity was measured using MethoCult GF H4434 “Complete” Methylcellulose medium with recombinant cytokines from StemCell Technologies (Vancouver, BC) and was used in accordance with the manufacturer's instructions. Flow cytometric analysis was performed to determine the percentage of subdiploid cells as described previously.¹¹ 

Standard polymerase chain reaction (PCR), Western blotting, and kinase assay procedures were used as described in Document S1.

Transplantation and shRNA in vivo experiments are described in Document S1.

Staining for Fyn was conducted in bone marrow samples as detailed in Document S1.

Although the phosphorylation status and contribution to leukemogenesis of the Src family members Lyn, Hck, and Fgr has been studied in CML,^3,4  little is known regarding Fyn. We examined Fyn protein expression in a tissue microarray containing bone marrow from healthy non-CML donors, and chronic-, accelerated-, and blastic-phase CML patients (Figure 1A). The average percentage of cells that stained positive for Fyn in blast crisis patients was higher than in chronic-phase or accelerated-phase patients (68.8% vs 18%; P = .007). In 11 of 16 blast crisis patients, 100% of the cells were positive for Fyn. In comparison, the majority of the combined chronic and accelerated specimens were negative for Fyn (11/18) and only one sample contained 90% Fyn-positive cells. In non-CML healthy donor bone marrow, 3 of 5 specimens were negative for Fyn and the average percentage of Fyn-positive cells was the lowest seen in any cohort (2.5%). Our results indicate that Fyn expression was lowest in non-CML marrow, slightly higher in chronic- and accelerated-phase CML, and significantly elevated in blast crisis patients. Furthermore, 2 patients in this cohort were evaluated for Fyn expression while in chronic phase and then after progression to blast crisis. In both patients, disease progression was accompanied by an increased percentage of cells staining positive for Fyn as shown by immunohistochemistry in one representative patient who progressed to myeloid blast crisis (Figure 1B).

To clarify the relationship between BCR-ABL1 and Fyn expression, we examined Fyn protein levels in numerous CML cell lines. Western blotting revealed a 4-fold increase in Fyn protein in BaF3 p210 cells compared with BaF3 vector–transduced cells (Figure 1C lanes 3,4). Imatinib-resistant point mutants of BCR-ABL1 (T315I and E255K)¹²  retained the up-regulation of Fyn (Figure 1C lanes 5,6). K562 cells, which also express p210 BCR-ABL1, were positive for Fyn protein expression. Imatinib-resistant versions of these cells (which do not carry point mutations)¹³  also retained Fyn protein expression (Figure 1C lanes 1,2). 32D cells stably transfected with p210 BCR-ABL1 similarly showed heightened Fyn protein expression compared with parental 32D cells. Confirmation of Fyn up-regulation was further provided by TonBp210 cells transfected with a tetracycline-inducible Bcr-Abl expression vector. After exposure to 1 μg/mL doxycycline, BCR-ABL1 and Fyn expression increased in a time-dependent manner (Figure 1D). The up-regulated Fyn was found to be active as a kinase by two methods. Phospho-Fyn was detected by immunoprecipitation with a Fyn-specific antibody and Western blotting with a phospho-Src antibody (Figure 1E top panel). Total Fyn in the immunoprecipitates was increased to the same degree in p210-expressing BaF3 cells. Total lysates, collected prior to immunoprecipitation, were used to measure actin levels, which were roughly equivalent across samples, indicating equal protein loading. A Fyn kinase assay was also performed in BaF3 vector and BaF3 p210 cells using Sam 68 as a substrate, and showed a 50% increase in p210-expressing cells (Figure 1E bottom panel). Autophosphorylation of Fyn was also seen as indicated by the lower band in the bottom panel of Figure 1E. The immunoprecipitation data show a more robust increase in phospho-Fyn than the kinase assay data, likely due to different amounts of protein lysate used for the assays. One milligram of protein was used for the kinase assay, whereas 100 μg was used for the immunoprecipitation. Since these are both semiquantitative methods, it is likely that the kinase assay signal was nearly saturated in the BaF3 vector cells, thus the difference between Fyn kinase activity in the vector versus p210-expressing cells is less obvious.

Fyn up-regulation was also evident in an in vivo CML model system. Bone marrow was harvested from donor mice and infected with MigR1-GFP-p210. Infected cells were injected intravenously into irradiated recipient mice. RNA was isolated from GFP-sorted peripheral blood cells of the recipient mice. GFP-positive cells expressed Fyn mRNA, whereas GFP-negative cells did not (Figure 1F lanes 1,2). Furthermore, mice that received a transplant of bone marrow infected with vector (Figure 1F, MigR1 vector, second panel) did not up-regulate Fyn mRNA. As a control, RNA from BaF3 vector– and p210-transduced cell lines was also subjected to reverse-transcription (RT)–PCR (Figure 1F lanes 3,4) and illustrate that in cell lines, Fyn mRNA is also increased. Together, these results indicate that Fyn up-regulation occurs in BCR-ABL1–expressing cells in vivo (Figure 1F) as well as in vitro (Figure 1C,D) at the mRNA and protein level.

To place all of these events within the context of BCR-ABL1 kinase activity, we measured Fyn protein levels after exposure to 0.25 μM imatinib for 24 hours (Figure 1G). Fyn protein up-regulation was blunted, demonstrating that BCR-ABL1 kinase activity contributes to heightened Fyn protein expression.

RNA interference-based knockdown of Fyn was conducted to determine the functional significance of increased Fyn levels. Lentiviral and plasmid shRNA approaches were used. Fyn protein (Figure 2A) and mRNA (Figure 2A) expression were inhibited by more than 80% in Fyn shRNA–transfected K562 cells compared with the control scrambled vector–transfected K562 cells. Similarly, lentiviral shRNA with 2 sequences (U613 and U67) to Fyn decreased protein expression, whereas a rescue construct overexpressing a nondegradable form of Fyn partially restored expression. Protein levels of neither Lyn nor c-abl were altered in the Fyn shRNA–expressing cells (Figure 2A). Interestingly, BCR-ABL1 kinase activity was significantly repressed as measured by phosphorylation of the substrate Crk. This finding is consistent with a recent report showing that Fyn is capable of phosphorylating the SH3-SH2 region of BCR-ABL1 in vitro.¹⁴ 

Proliferation of Fyn shRNA–containing cells versus scrambled shRNA–containing cells was assessed by plating equivalent cell numbers and comparing cell counts after 24 hours and 48 hours (Figure 2B left graph). Cells transfected with Fyn shRNA grew slower than counterparts transfected with scrambled shRNA (P < .01). A rescue construct, encoding a mutant Fyn not susceptible to shRNA degradation, stimulated cell growth (P < .04) compared with Fyn shRNA–expressing cells (Figure 2B right graph). The effects of Fyn knockdown on cell survival were assessed using clonogenic assays. Significantly fewer colonies were apparent after Fyn shRNA transduction (P < .05), whereas transduction with the rescue construct restored clonogenic growth (P < .05; Figure 2C). Fyn shRNA also sensitized K562 cells to imatinib treatment. Treatment with 5 μM imatinib for 48 hours elicited higher levels of DNA fragmentation in K562 cells expressing Fyn shRNA than nontargeted shRNA (P < .005) or Fyn rescue construct (P < .05; Figure 2D).

To test whether Fyn knockdown impacted CML growth in vivo, we injected severe combined immunodeficient (SCID) mice with K562 cells containing shRNA directed toward Fyn or toward an unrelated sequence. Four weeks later, peripheral blood cell counts were assessed. Percentages of segmented cells were lower in mice injected with Fyn shRNA–containing cells (P = .05; Figure 2E). There was a trend toward prolonged survival in mice carrying Fyn shRNA compared with those carrying nonspecific shRNA. However, statistical analyses of this small cohort indicated it did not meet statistical significance (P = .079; Figure 2F).

While existing therapies, such as dasatinib, target activation of Src family members and indeed inhibit Fyn kinase activity,¹⁵  our data suggest that increased Fyn protein expression is a feature of CML cells. Furthermore, our data corroborate those of others suggesting that Fyn expression may control BCR-ABL1 kinase activity thereby making Fyn an attractive target for CML therapies. Taken together, our data highlight the importance of Fyn expression in BCR-ABL1–expressing cells and have diagnostic implications for improved prediction of patients who might be transforming to blast crisis and require a therapeutic intervention.

We thank Gary Gallick, PhD, for sharing his expertise on Src family members and for thoughtful discussions and feedback.

This work was supported by grants from the National Institutes of Health (RO1 CA115811 to J.C) and from the Children's Leukemia Research Foundation (J.C.).

National Institutes of Health

Contribution: K.B., Y.G., X.L., and Q.L. performed research; H.M.A. and M.M. analyzed data; A.H. and C.M. performed research and analyzed data; R.B.A. and M.B.-E. contributed reagents; and J.C. designed research, analyzed data, and wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Joya Chandra, PhD, Box 853, Pediatrics Research, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030; e-mail: jchandra@mdanderson.org.