|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Prepublished online as a Blood First Edition Paper on September 19, 2002; DOI 10.1182/blood-2002-05-1519.
NEOPLASIA
From the Department of Adult Oncology, Dana-Farber
Cancer Institute, Boston, MA; the Department of Biostatistical Science,
Harvard School of Public Health, Dana-Farber Cancer Institute, Boston,
MA; the Department of Medicine, Harvard Medical School, Boston, MA; the
Department of Medicine, Indiana University School of Medicine,
Indianapolis; and the Biomedical Research Institute, Long Island Jewish
Health System, Manhasset, NY.
The physical interactions between B cells and stromal cells from
the lymphoid tissue microenvironment are critical to the survival of
normal and malignant B cells. They are principally mediated by
integrins expressed on B cells and counterreceptors on stromal
cells. Specifically, Integrins are Integrin crosslinking in normal and malignant human B cells induces
intracellular signaling, leading to protein kinase activation and
phosphorylation of various substrates.11,12 In this
report, we used high-density oligonucleotide microarrays to study the temporal profile of gene expression induced by Reagents
Generation of cRNA and microarray hybridization
Cell stimulation REH or Nalm-6 cells (2 pre-B ALL cell lines; 3 × 107) were cultured over time in RPMI 1640, with or without 10% fetal calf serum (FCS), on 10-cm plates or 6-well plates coated with Fn (15µg/mL) or PLL. For the condition without serum, the cells were also precultured in plain RPMI without serum for 3 hours at 37°C before stimulation. After 2 washes with phosphate-buffered saline (PBS), RNAs were extracted by adding Trizol directly on the plate or proteins were obtained with 400µL ice-cold NP-40 lysis buffer (containing 9.5 µg/mL aprotinin, 1 mM phenylmethylsulfonyl fluoride, 1mM pepstatin A, and 2.5 mM sodium vanadate), put on ice for 15 minutes, centrifuged, and then supernatants were collected. For transcription factor analysis, cells were lysed with boiling sample buffer, harvested, and boiled for an additional 5 minutes. For caspase detection, cells were lysed with Chaps buffer (Cell Signaling Technology) according to the manufacturer's instructions. Samples were analyzed by Western blot with specific antibodies, as previously described.14Apoptosis detection After stimulation with Fn or PLL, cells were detached with Trypsin (Life Technologies) and stained with annexin V-fluorescein isothiocyanate (FITC) and propidium iodide (PI) according to the manufacturer's instructions (BD Biosciences, San Jose, CA). Samples were then analyzed using a Coulter-EPICS-XL flow cytometer (Hialeah, Miami, FL).Data mining After scanning, the expression values for the genes were determined using Affymetrix GeneChip software. These data were plotted in order to validate the sample quality. Data mining was then performed on the expression values obtained using DNA-Chip Analyzer (D-Chip, Wong Lab, Dept of Biostatistics, Harvard School of Public Health, Boston, MA) with the *.CEL files obtained from GeneChip. As there were 4 time points in the temporal expression of the genes, we modeled the expression values as a quadratic function of time (see data plots in Figure 3 for an example). We also took into consideration the effects of treatment, cell line, and their interactions with time. This led to the following regression model: expression ~ cell + trtmt + cell*trtmt + tt + tt2 + cell*tt + trtmt*tt + cell*trtmt*tt + cell*tt2 + trtmt*tt2 + cell*trtmt*tt2, where "cell" stands for cell line, "trtmt" for treatment, "tt" for time, and "tt2" for time squared. In order to select potentially interesting genes, we applied 2 criteria: no obvious cell line-by-treatment interactions (the 3rd, 8th, and last terms in the right-hand side of the equation being "nonsignificant" as defined by a P greater than a certain threshold according to the F distribution) and a substantial treatment effect (the 2nd, 7th, and 10th terms in the right-hand side of the equation being "significant" as defined by a P less than a certain threshold). The P values here were used as a summary statistic of the corresponding effects and as a screening tool for selecting genes. To obtain the P values for the treatment effect, we used first the normality assumption of the expression data (ie, the F distribution of the statistic) and then a permutation method where the observed statistic was compared with the empirical distribution of the statistic from the permutations. The permutations were carried out within the same time points and the same cell lines, that is, the 2 labels each of Fn and PLL were permuted within the 4 replicates. This gave 6 possible permutations for each time point and each cell line. As there were 3 time points (excluding time 0 when there was no treatment given) and 2 cell lines, there were a total of 66 = 46 656 permutations. For our analysis we randomly chose 500 permutations, which would make computation feasible and give us a 95% confidence interval no wider than 0.1 for the estimated P value and, in particular, a 95% confidence interval no wider than 0.05 if the estimated P is around .1 ("Results"). These unadjusted (for multiple comparison) P values from either the F distribution or the permutation distribution provided useful ways to rank the genes. All data analyses were carried out using the statistical software package R (http://www.r-project.org/).Real-time quantitative PCR Analysis was performed with an ABI PRISM 7700 Sequence Detection System instrument using the SYBR Green I dye (PE Biosystems, Foster City, CA). Briefly, detection of PCR product was monitored by measuring the increase in fluorescence caused by the binding of SYBR Green I dye to dsDNA directly in the reaction tube. The sequence detector software calculates the threshold cycle number (CT) when signals reach 10-fold the standard deviation of the baseline. It was previously demonstrated that the calculated CT values are a quantitative measurement for the mRNA levels of various genes tested.15,16 The samples were treated as previously described.17 Briefly, DNase I-treated total RNAs were reverse-transcribed using the Advantage RT-for-PCR kit (Clontech, Palo-Alto, CA) following the manufacturer's instructions and cDNAs or water as control were amplified in triplicates by real-time PCR using the SYBR Green Master Mix reagent.For each sample, CT values for GAPDH and ribosomal protein S9 genes were generated for normalization purposes. A correction factor was calculated by dividing by the minimum GAPDH and ribosomal protein S9 values and averaging the adjusted control values for the sample. This value was used to correct each sample for differences in RNA content. Primer sequences The following primers were designed using Primer Express 1.0 software (PE Biosystems). 5' GJB3 CCCAACATCGTGGACTGCTA, 3' GJB3 GGCGCCCACCATGAAGT, 5' CORO1a TGCACCCAGACACGATCTACA, 3' CORO1a TGTCACGGCAGGAGGTACAA, 5' CCT4 CCTATCAGGACCGCGACAAG, 3' CCT4 AGCATCAGCAACCGCTTTG, 5' VAV2 TCGATGTGCGAGACTTTGGA. 3' VAV2 CTTTGTTCTGCGCGATGCT, 5' FRAP1 AGGCCGCATTGTCTCTATCAA, 3' FRAP1 GCAGTAAATGCAGGTAGTCATCCA, 5' EBP41L1 CACTGCGGAAACCCTCTCAA, 3' EBP41L1 TCTCAGAAAACCCTCCTTTCACA, 5' GAPDH CCACCCATGGCAAATTCC, 3' GAPDH GATGGGATTTCCATTGATGACA. To ensure the purity of the reaction, the PCR product for each gene was verified using dissociation curve software that illustrates a product as "pure" when it has a consistent melting temperature across replicates.
Sample quality To determine which genes were involved in the antiapoptotic effect induced by integrin ligation in B cells, REH or Nalm-6 cells were cultured in serum-free medium on plates coated with Fn, a ligand for 1 integrin, or PLL as control. We first analyzed the percentage of
apoptotic cells after 1, 3, or 5 hours of culture on PLL or Fn by
annexin-PI staining. Figure 1A shows a
representative experiment performed with Nalm-6 cells. Similar results
were obtained with REH cells (data not shown). Whereas an increase of
apoptotic cells (about 40%) was observed when cells were cultured on
PLL (P) or in media alone (control C), the number of apoptotic cells did not increase when the cells were cultured on Fn (F). Furthermore, the culture of cells on PLL in presence of serum prevented apoptosis (P+). This result indicates that under our conditions, Fn stimulation repressed the apoptosis of B cells induced by serum
starvation.
We previously reported that integrin ligation in B cells induces a
cascade of tyrosine phosphorylation events.18 We examined tyrosine (Tyr) phosphorylation of the total cellular lysates following Data mining The data were first processed using D-Chip software.21,22 Figure 2A shows the same replicate scatter plot as Figure 1B, but obtained after D-Chip processing and not Affymetrix's GeneChip software. We also plotted the cubic root of the values in order to evaluate the noise at the lower end of the expression values (right panel) and to apply a sufficient threshold to eliminate all low-expressed genes. We applied a threshold of more than 100 in expression level in at least one of the 28 experimental conditions to identify putative genes regulated by Fn or PLL treatment. Figure 2B shows the scatter plots of the average values (average of the 2 independent experiments) for both cell lines at 1, 3, and 8 hours. To select genes that were differentially expressed between the 2 treatments, we modeled the temporal expression profile of the genes as a quadratic regression function of time at 0, 1, 3, and 8 hours, plus the effects of treatment, cell line, and their interactions with time, as described in "Materials and methods." The selected genes were those that did not have obvious cell line-by-treatment interactions and those that had a treatment effect. We applied a threshold of P greater than .1 for no cell line-by-treatment interaction; this, together with the previous filter of at least one expression greater than 100, gave us a total of 4670 genes. To identify the differentially expressed genes, we first used the normality assumption of the expression data and a P cutoff of .1 in the F distribution (72 genes, data not shown) and then a permutation method where the observed statistic was compared with the empirical distribution of the statistic from 500 permutations.23 The Ps here did not carry the traditional meaning of pure statistical significance but allowed us to select the top 72 genes that were differentially expressed between the 2 treatments as summarized by the F statistic. The genes that were selected by the permutation method with a P cutoff of .1 (64 genes) are shown in Table 1. The permutation method does not rely on the normality assumption, although it is computationally intensive. The genes in Table 1 have been classified by biological functions. Forty-six of the 64 genes overlap with the 72 genes selected by assuming normality. We did not include in this analysis the additional 24-hour and 48-hour time-point experiments, since the 2 sets of experiments were performed independently and the second set does not have enough samples to permit permutation analysis. Furthermore, longer stimulation could have led to the regulation of the target genes of the first early responses genes. However, we did analyze these later time points independently, using a 2-sample t statistic, and we indicated with asterisks the genes that were identical or related to the ones selected from the first time-point experiments in Table 1.
The expression levels of these 64 genes obtained by D-Chip were then
plotted for each independent experiment as a function of time (not
shown). We also plotted the average expression level for each cell line
as a function of time. This allowed us to discard 26 genes that showed
very different levels of expression across the duplicates within a cell
line. Therefore, there were 38 genes, listed in Figure
3, which appeared to have a reproducible
pattern of expression as well as a high enough expression level. For
the purpose of clarity, we show plots only for Nalm-6 cells. However, similar patterns of expression were obtained for REH cells. (See the
Supplemental Figure link at the top of the online article, at the
Blood website.)
Regulation of proteins involved in intercellular communication, adhesion, and proliferation We analyzed the RNA expression levels of 6 of these 38 genes by RT-Q-PCR experiments24 using SYBR Green I, as previously described.17 We selected 3 genes induced by PLL (VAV2, CORO1A, EPB41L1) and 3 genes induced by Fn (CCT4, FRAP1, GJB3). These genes play a role in adhesion or proliferation and were therefore of potential relevance. We also selected GJB3, which is not known to be expressed by B cells. The results of one representative experiment for each gene are shown in Table 2. The CT values were normalized for each gene in all of the samples using GAPDH and S9 expression as described in "Materials and methods." A lower CT value reflects a higher mRNA content. The difference between the mean CT value for Fn and the mean CT value for PLL ( CT) was calculated for
each time point. No important differences were observed except for
VAV2, which exhibits a CT greater than 3. However, we were expecting low values for CT because no
considerable increase in fold change could be detected on the overall
scatter plot of Fn/PLL treatments shown in Figure 1B. Of the 6 genes
tested, 5 reproduced the results obtained from our data mining (shown on the left side of the table), showing an increased expression on PLL
for VAV2, CORO1A, and EBP41L1 (positive
CT) and an increased expression on Fn for CCT4,
FRAP1, and GJB3 (negative CT). No changes for FRAP1 could be observed for Nalm-6 cells when
the average expression was plotted as a function of time (Figure 3), but when plotted separately, one of the 2 experiments shows
amplification on Fn similar to the 2 REH experiments, whereas the other
one has the reversed effect (not shown). However, amplification on Fn
by RT-Q-PCR was detected with both cell lines. CORO1A
showed, as expected, an up-regulation after PLL stimulation but
then an up-regulation with Fn was detected that was not observed when the data were plotted as a function of time. Also, changes at 1 hour
could be detected by RT-Q-PCR, whereas no change was observed on
our plots.
FBI-1 induction in Fn-stimulated cells We investigated whether some of the genes found to be differentially expressed by microarrays could be validated at the protein level. We chose 5 genes up-regulated by either Fn or PLL and analyzed their expression by Western blotting experiments using specific antibodies.Several of the genes found by our models were transcription factors.
Among them, FBI-1 (also described as lymphoma-related factor
[LRF]) was increased by Fn stimulation. The protein encoded by this
gene has been described as a potential transcription
factor.25 It contains a POZ domain at its N terminus and 4 Kruppel-type zinc fingers at its C terminus. Furthermore, FBI-1 may
bind actin,26 and therefore likely plays a role in
adhesion. We investigated whether an up-regulation could also be
detected at the protein level. Nalm-6 cells were cultured on Fn- or
PLL-coated plates in the absence of serum and lysed after various
periods of time. Aliquots of each cell lysate were analyzed by Western
blot using anti-FBI-1 antibodies. The membrane was then stripped and
reprobed with anti-GAPDH antibodies to confirm that equivalent amounts of proteins were loaded in each lane. As depicted in Figure
4A, Fn stimulation induced up-regulation
of FBI-1, which was more obvious after 1 and 2 days of stimulation.
c-Fos and CD79A regulation Among transcription factors, we could also detect an induction of both c-Fos and Jun-D RNAs when the cells were cultured on PLL. The cellular proto-oncogenes encoded by these genes are members of the AP1 family of transcription factors, and they control proliferation, differentiation, and apoptosis.27 We analyzed whether c-Fos protein was modulated on Fn or PLL treatment. As shown in Figure 4B, an induction of c-Fos could be observed after 3 days of serum starvation but was repressed when the cells were cultured with Fn.An up-regulation of CD79A RNA could be detected using microarrays when the cells were cultured on both Fn and PLL, although a larger increase was observed on PLL. We confirmed this result by Western blot using Nalm-6 cells. CD79A protein was induced when the cells were stimulated by Fn or PLL. However, the induction was higher in the presence of PLL (Figure 4B). This result suggests that adherence of B cells to a substratum leads to an increased CD79A expression. Regulation of proteins controlling apoptosis by integrin stimulation Among the genes induced by PLL but suppressed on Fn, caspase 7 and requiem encode proteins regulating apoptosis. Caspase 7 belongs to a large family of cysteine proteases involved in apoptosis.28 These proteases exist as inactive proenzymes, which undergo proteolytic processing to produce 2 subunits that dimerize to form the active enzyme. Requiem is a zinc-finger protein with 2 C4HC3 domains that are highly homologous to the C4HC3 zinc fingers of MOZ, an acetyltransferase protein involved in the t(8;16) leukemia-associated breakpoint.29 It is required for the apoptotic response induced by interleukin 3 (IL-3) deprivation of murine myeloid cell lines.13 Since Fn treatment rescues cells from apoptosis, we investigated whether these 2 proteins were involved in apoptosis of human B cells, meaning up-regulated when the cells were cultured on PLL but repressed when cultured on Fn. Nalm-6 cells were cultured on Fn- or PLL-coated plates in the absence or presence of serum to induce or prevent apoptosis and were lysed after various periods of time. An aliquot of each cell lysate was then analyzed by Western blot using anti-requiem antibodies (Figure 5A) or anti-caspase 7 antibodies (Figure 5B). As shown in Figure 5A, an increase of requiem protein could be observed when the cells were cultured on PLL after 2 and 3 days in the absence of serum, whereas no induction was observed in presence of Fn. Furthermore, the presence of serum abrogates requiem induction observed with PLL. Similarly, a faint but reproducible increase of caspase 7 was detected after 4 and 5 days of culture on PLL but not on Fn. Furthermore, by using a specific antibody, we were able to detect the cleaved fragment of caspase 7 in PLL treated cells, which was greatly reduced when the cells were treated with Fn (Figure 5B) or in the presence of serum (data not shown). Therefore, caspase 7 was induced and activated in B cells cultured with PLL, whereas Fn stimulation inhibits its induction and further activation.
We then examined the potential role of other proteins regulating caspase 7 activity. We first analyzed whether caspase 3, which shares most of its biological functions with caspase 7, was also involved in this process. As shown in Figure 5C, caspase 3 was up-regulated during apoptosis but integrin engagement inhibited the induction. We also evaluated whether the expression of the antiapoptotic proteins XIAP and survivin, which inhibit caspase 3 and 7, were modified by integrin stimulation. Stimulation with Fn increased the expression of both proteins (Figure 5D). We could also detect an increased expression of both proteins in the presence of serum (data not shown), suggesting that the increased expression of both XIAP and survivin was not necessarily due to apoptotic conditions but was correlated with integrin engagement. Furthermore, in the same experiment, we found that XIAP expression inversely correlates with the level of caspase 3 activation. In the presence of Fn, XIAP was induced, and therefore no caspase 3 activation occurred. In contrast, XIAP was not induced on PLL, and we could detect the cleaved caspase 3 portion, indicating that caspase 3 was activated. Finally, we analyzed whether the integrin-mediated inhibitory effect on caspase 3 and 7 activation could be validated in another model of apoptosis. We cultured Nalm-6 cells on Fn or PLL with doxorubicin in order to induce apoptosis.30 After one day, cells were lysed and Western blots were performed using cleaved caspase 7 or 3 antibodies. As shown in Figure 5E, Fn stimulation clearly inhibited activation of both caspases.
The regulation of normal and neoplastic B-cell survival involves
integrin engagement. Precursor B leukemia cell survival is supported by
ligation of Among the 38 selected genes (Figure 3), a number play a role in adhesion. Half of these genes were up-regulated when cells were cultured on PLL. This indicates that the chemical adhesion of the cells on PLL was not neutral but rather up-regulated genes involved in the adherence of cells, such as ZYXIN, CORO1A, EBP41L1, and VAV2. These last 3 genes were confirmed by quantitative PCR. CORO1A (coronin) is an actin-binding protein involved in mitosis and cell motility.33,34 Erythrocyte membrane protein band 4 like 1 (EPB4L1), homologous to EPB4.1, mediates interactions between the erythrocyte cytoskeleton and the overlying plasma membrane and therefore plays also a role in cytoskeletal rearrangement.35 VAV2, the second member of the VAV oncogene family, is expressed in most tissues and has been shown to be required for cell spreading.36 Several genes identified provide links between integrin adhesion and
regulation of proliferation. We confirmed that Fn plating of the
precursor B cell lines led to an induction of CCT4, the One of the transcription factors up-regulated by Fn stimulation was
FBI-1 (or LRF). FBI-1 contains a POZ domain at its N terminus and 4 Kruppel-type zinc fingers at its C terminus. The C terminus is
sufficient for specific binding, and FBI-1 can form homomers through
its POZ domain. Furthermore, FBI-1 may bind actin. A recent study
reported that other POZ-containing proteins, the hcKrox family members
( Members of the transcription factor AP-1 complex c-Fos and Jun-D were highly overexpressed when the precursor B cell lines were cultured on PLL. These proteins are proto-oncogenes that control proliferation, differentiation, and apoptosis by propagating signals from the cell membrane to the nucleus. Interestingly, it has also been reported that inhibitors of microtubule assembly induce an increased phosphorylation and expression of specific Jun and Fos proteins, underlining a link between the cytoskeleton and AP-1 family members.45,46 Both c-Fos and Jun-D proteins have been linked to apoptosis, although the role of Jun-D has been described as antiapoptotic or proapoptotic depending on the cell type. c-Fos induces apoptosis in germinal center B cells, and its overexpression induces apoptosis of CD43+ pro-B cells.47,48 Jun-D also negatively regulates cell proliferation. Its constitutive overexpression inhibits proliferation of immortalized fibroblasts by increasing the number of resting cells,49 whereas Jun-D-deficient fibroblasts exhibit increased proliferation.50 The similar temporal inductions of both c-Fos and Jun-D suggest that this AP-1 complex plays a role in B-cell apoptosis and that Fn stimulation prevents this induction, hence protecting the cells from apoptosis. One major goal of this study was to try to understand the mechanisms by which adhesion to stromal cells promotes cell survival. We identified several proteins that play a role in apoptosis regulation. One of these was the cysteine protease caspase 7. These proteases are activated by a proteolytic cleavage cascade. We observed that caspase 7 was up-regulated when cells are cultured with PLL but suppressed by Fn. Of note, we observed that caspase 7 was cleaved and therefore activated, and that, similarly, Fn stimulation greatly reduced its activation. Moreover, we found that caspase 3 was also up-regulated and activated during apoptosis but that integrin engagement inhibited its induction. We show that integrin stimulation also protected the cells from apoptosis induced by chemotherapy drugs, indicating that this is a mechanism of rescue common to different apoptotic signals. We further analyzed whether, on the contrary, Fn stimulation could up-regulate proteins from the IAP (inhibitor of apoptosis proteins) family.51,52 We found that Fn stimulation increased XIAP and survivin expression. Both proteins directly inhibit at least caspase 3 and caspase 7,53,54 and XIAP has been described to inhibit serum deprivation-induced apoptosis.55 We also identified a gene up-regulated on PLL called REQUIEM that had been cloned after IL-3 deprivation in murine myeloid cells. This protein is required for apoptosis within these cells.13 We were able to detect an increase in the protein level of requiem, indicating that this protein is also likely involved in apoptosis of precursor B ALL cells induced by serum starvation. Furthermore, we show that Fn stimulation prevented the up-regulation of this protein in the absence of serum, making this protein an interesting target for future use in therapy. In conclusion, our results led to the following hypothesis, which is
illustrated in Figure 6: Upon stimulation
through
We thank S. Gullans, L. Hsiao, and R. Gentleman for helpful discussion. Many thanks to S. Chiaretti and K. Wang for their help with Affymetrix protocol. We are very grateful to N. Hernandez (CSHL, NY) for the generous gift of FBI-1 Ab. We thank H. Husson for his critical reading of the manuscript.
Submitted May 23, 2002; accepted September 4, 2002.
Prepublished online as Blood First Edition Paper, September 19, 2002; DOI 10.1182/blood-2002-05-1519.
Supported by National Institutes of Health grants CA81494 and CA06516, the Leukemia and Lymphoma Foundation of America, and the Norman Hirschfield Foundation. A.L.A. was partly supported by CNRS, France.
A.L.A. and R.X. contributed equally to this work.
The online version of the article contains a data supplement.
The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked "advertisement" in accordance with 18 U.S.C. section 1734.
Reprints: Arnold Stephen Freedman, Dana-Farber Cancer Institute, 44 Binney St, Boston, MA 02115; e-mail: arnold_freedman{at}dfci.harvard.edu.
1. van der Flier A, Sonnenberg A. Function and interactions of integrins. Cell Tissue Res. 2001;305:285-298[CrossRef][Medline] [Order article via Infotrieve]. 2. Dedhar S. Integrins and signal transduction. Curr Opin Hematol. 1999;6:37-43[CrossRef][Medline] [Order article via Infotrieve]. 3. Bendall LJ, Makrynikola V, Hutchinson A, Bianchi AC, Bradstock KF, Gottlieb DJ. Stem cell factor enhances the adhesion of AML cells to fibronectin and augments fibronectin-mediated anti-apoptotic and proliferative signals. Leukemia. 1998;12:1375-1382[CrossRef][Medline] [Order article via Infotrieve].
4.
Manabe A, Coustan-Smith E, Behm FG, Raimondi SC, Campana D.
Bone marrow-derived stromal cells prevent apoptotic cell death in B-lineage acute lymphoblastic leukemia.
Blood.
1992;79:2370-2377 5. Campana D, Coustan-Smith E, Manabe A, et al. Human B-cell progenitors and bone marrow microenvironment. Hum Cell. 1996;9:317322.
6.
Manabe A, Murti KG, Coustan-Smith E, et al.
Adhesion-dependent survival of normal and leukemic human B lymphoblasts on bone marrow stromal cells.
Blood.
1994;83:758-766
7.
Mudry RE, Fortney JE, York T, Hall BM, Gibson LF.
Stromal cells regulate survival of B-lineage leukemic cells during chemotherapy.
Blood.
2000;96:1926-1932 8. Sarkar S, Svoboda M, de Beaumont R, Freedman AS. The role of AKT and RAFTK in b1 integrin mediated survival of precursor B acute lymphoblastic leukemia cells. Leuk Lymphoma. 2002;43:1663-1671 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
4
subunit of
the cytosolic chaperonin containing TCP-1.
), regulate multiple extracellular matrix
genes.
,
cell death.