Shwachman-Diamond Syndrome is not necessary for the terminal maturation of neutrophils but is important for maintaining viability of granulocyte precursors

doi:10.1016/j.exphem.2006.12.010Shwachman-Diamond Syndrome is not necessary for the terminal maturation of neutrophils but is
important for maintaining viability of granulocyte precursors
Masafumi Yamaguchia, Kingo Fujimurab, Hanae Togaa, Asim Khwajac, Naoki Okamuraa, and Rajesh Choprac
aLaboratory of Physiological Chemistry; bLaboratory of Clinicopathological Therapeutics,
Hiroshima International University, Kure-shi, Japan; cDepartment of Hematology, University College London, London, UK
(Received 27 June 2006; revised 20 December 2006; accepted 22 December 2006)
Objective. Shwachman-Diamond syndrome (SDS) is an autosomal-recessive disorder charac- terized by exocrine pancreatic insufficiency and bone marrow failure. The SDS disease locus was mapped to chromosome 7q11, and disease-associated mutations were reported in the Shwachman-Bodian-Diamond syndrome (SBDS) gene. SBDS is a member of a highly conserved protein family in diverse species including archaea and eukaryotes. It is widely expressed in many tissues, and its function is still unknown. To investigate the function of the SBDS protein, we undertook loss-of-function experiments in the 32Dcl3 cell line, which has the potential to differentiate to mature neutrophils.
Methods. SBDS gene was downregulated with lentivirus-based RNAi system. SBDS knock- down cells were analyzed for surface marker expression by flow cytometry and analyzed for the sensitivity to apoptosis-inducing stimuli.
Results. After culture in granulocyte colony-stimulating factor (G-CSF)-containing medium for 3 days, 32Dcl3 cells demonstrated normal proliferation but complete downregulation of SBDS protein expression. The SBDS RNAi knockdown cells did not proliferate in G-CSF- containing medium but after 7 days had the appearance of segmented neutrophils. The neutro- phil maturation markers were detected on these cells. Undifferentiated SBDS RNAi knockdown cells demonstrated increased apoptosis of undifferentiated cells. Notably, SBDS RNAi knockdown cells demonstrated normal proliferation in interleukin-3-containing medium.
Conclusion. We have established an SDS model cell line and have used this model to demon- strate that SBDS is not required for neutrophil maturation. However, SBDS knockdown cells were sensitive to apoptotic stimuli, indicating that SBDS acts to maintain survival of granulocyte precursor cells. 2007 International Society for Experimental Hematology. Published by Elsevier Inc.
Shwachman-Diamond syndrome (SDS) is a rare autosomal- recessive disorder characterized by short stature, exocrine pancreatic insufficiency, and hematologic defects, in partic- ular neutropenia and impaired chemotaxis [1]. Hematologic manifestations other than neutropenia include anemia, increased fetal hemoglobin levels, thrombocytopenia, and aplastic anemia. Approximately 20 to 25% of patients develop acute myelogenous leukemia (AML) [1].
Offprint requests to: Masafumi Yamaguchi, Ph.D., 5-1-1 Hirokoshingai,
Kure-shi, Hiroshima, 737-0112, Japan; E-mail: [email protected].
ac.jp
doi: 10.1016/j.exphem.2006.12.010
The syndrome is caused by mutations in the Shwachman- Bodian-Diamond syndrome (SBDS) gene on chromosome 7 [2–5]. The majority of patients with SDS were found to have recurring mutations resulting from gene conversion in about 90% of unrelated individuals with SDS, with 60% carrying two converted alleles [2]. The SBDS gene is mapped to chro- mosome 7q11 and is composed of five exons. The gene has a 1.6-kb transcript and encodes predicted protein of 250 amino acids, which is a member of a highly conserved pro- tein family in diverse species ranging from plants, Archaea, and yeast to vertebrate animals, suggesting that it may have a fundamental, conserved biochemical role [2]. In Archaea, the SBDS gene is located in an operon that encodes, among
Experimental Hematology. Published by Elsevier Inc.
580 M. Yamaguchi et al./ Experimental Hematology 35 (2007) 579–586
other proteins, RNA-processing enzymes. In yeast, the gene for the SBDS homologue (YLR022C) is clustered with RNA- processing enzymes in transcription profiling experiments [2,6–9]. SBDS is widely expressed throughout human tis- sues, and SBDS localizes to the nucleolus [10]. These reports suggest that the SBDS contributes rRNA processing [2]. However, little is yet known about SBDS protein function during neutrophil differentiation.
To investigate the function of the SBDS protein, we un- dertook loss-of-function experiments in the myeloid 32Dcl3 cell line, which has the potential to differentiate to mature neutrophils.
Material and methods
Cell lines 32Dcl3 and WEHI-3b cells were obtained from Riken Cell Bank (Tsukuba, Japan). WEHI-3b conditional medium was prepared as described standard procedure [11]. Ten percent WEHI-3b condi- tional medium was equivalent to 1 ng/mL murine interleukin 3 (mIL-3). 32Dcl3 cells were cultured in Iscove’s modified Dul- becco’s medium (IMDM) supplemented with 10% fetal calf serum (FCS) and 10% WEHI-3b conditional medium. For neutrophil differentiation, 1 105 cells/mL were resuspended in IMDM containing 10% FCS and murine granurocyte colony-stimulating factor (mG-CSF) (5 ng/mL; R & D systems, Oxford, UK).
SBDS inerfering RNA The SBDS interfering RNA (RNAi) construct was generated by using the lentivirus vector pLL3.7 [12], kindly provided by Dr. Luk Van Parijs (Massachusetts Institute of Technology, Cambridge, MA, USA). We modified pLL3.7 vector with replace- ment enhanced green fluorescent protein fragment to puromycin resistant gene. The target sense sequence of murine SBDS (mSBDS) are 50-GAGACCTTACACCGTTATC-30 (RNAi1) and 50-GAAGCTGAAGGAGAAGCTG-30 (RNAi2), and the shRNAi of mSBDS was engineered in the pLL3.7puro vector to express RNAi for knocking down the endogenous mSBDS. To produce lentivirus, 293T cells were cotransfected with the pLL3.7puro along with the constructs containing the gag/pol and the vesicular stomatitis virus G protein (VSV-G) envelope using calcium phosphate precipitation. Lentiviral supernatants were collected at 72 hours after cotransfection in IMDM medium (containing 10% fetal bovine serum), filtered through a 0.45-mm filter. The re- combinant lentivirus was used to transduce 32Dcl3. Successfully infected cells were selected in puromycin (5 mg/mL). To avoid off-target effects, control virus was prepared after transfection of pLL3.7puro empty vector and infected to 32Dcl3. We established and analyzed two independent shRNAi infected cells.
Flow cytometry Cells were resuspended in phosphate-buffered saline (PBS) containing 1% bovine serum albumin. Cells were stained with fluorescein isothiocyanate anti-Mac1a (BD PharMingen, San Jose, CA), anti Gr-1 (BD PharMingen), and granulocyte colony- stimulating factor (G-CSF) receptor (Santa Cruz Biotechnology, Santa Cruz, CA) for 30 minutes at 4C. Analysis was done with a fluorescein-activated cell sorting (FACS) Calibur flow cytometer.
Morphologic analyzes and superoxide production assay Sample of control 32Dcl3 and shRNAi infected cells treated with mG-CSF (5 ng/mL; R & D Systems) were prepared on glass slides using the cytospin method. Morphologic features were evaluated with use of May-Grunwald and Wright-Giemsa staining. For superoxide production assays, 2 104 cells were centrifuged, sus- pended in 200 mL Hank’s and PIPES buffer, and incubated at 37C for 15 minutes with 100 mM cytochrome c and 100 nM phorbol 12-myristate 13-acetate (PMA).
Semiquantitative reverse-transcriptase polymerase chain reaction RNA was extracted from cell line using the TRIzol reagent (Invitro- gen, Carlsbad, CA, USA), according to the manufacturer’s protocol. For reverse-transcriptase (RT) polymerase chain reaction (PCR), cDNA was synthesized from total RNA using SuperScript II (Invi- trogen), according to the manufacturer’s specifications. Primers for PCR amplification of murine NADPH oxidase genes were as follows: p22phox 50-GATGTGGACAGAAGTACCTG-30 and 50- ACTGGCATTGGGTTAACCTG-70; and gp91phox 50-GGAAA CCCTCCTATGACTTG-30 and 50-CCAGACAGACTTGAGAAT GG-30; and p47phox 50-AGAACAGAGTCATCCCACAC-30 and 50-CTTCTCGTAGTCAGCAATGG-30; and p67phox 50-GCAT- CAACAGAGACAAGCAC-30 and 50-ACAGCTTCTGCTTCCA- GATG-30. The mouse gelatinase gene primers were 50-TGCTATT GCTGAGATCCAGG-30 and 50-GATCCACCTTCTGAGACTTC- 30, and the mouse lactoferrin primers were 50-CTTGCTAACCA- GACCAGATC-30 and 50-TTCTTAGCCTCAGTCACAGG-30. The mouse SBDS gene primers were 50-ATGTCGATCTTCACCC CCACC-30 and 50-GGAAGGCGATGAGAAGTTTGAATGA-30. PCR was performed at 95C for 15 seconds, 51.7C for 15 seconds, and 72C for 60 seconds for 28 and 30 cycles. The data were analyzed with CS Analyzer (ATTO, Tokyo, Japan).
Apoptosis assays Annexin-V binding to 32Dcl3 cells was performed using Annexin-V (BD PharMingen, San Jose, CA, USA) and 7-amino-actinomycin D (7-AAD) (BD PharMingen). Cells (2.5 105) were washed with PBS(). Then cells were incubated with IMDM supplemented with 1 ng/mL of mIL-3 and 10% FCS or 1% FCS overnight. The cells were labeled with Annexin-V-PE and 7-AAD for 10 minutes at room temperature and analyzed by FACS Caliber. Results are reported as a percentage of Annexin-V-positive cells in early stages of apoptosis.
Western blotting The human SBDS fragment was introduced into the vector pGEX-6P-1. The resulting construct was transformed into strain BL21/DE3. The protein purification by glutathione s-transferase (GST)-affinity chromatography was carried out as described stan- dard procedure [13]. Cleavage of the SBDS/GST-SBDS on the glutathione sepharose-sepharose resin was achieved by overnight incubation with 20 units of Precisions protease (Amersham Phar- macia, Piscataway, NJ) in Tris-buffered saline (TBS) at 4C. Pure SBDS protein was collected from the supernatant after centrifuga- tion of the resin. Rabbits were immunized with this recombinant SBDS protein at Sheffield University according to standard proto- col [14].
Cells were lysed in extraction buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1% Triton-X 100, 0.1% sodium dodecyl
581M. Yamaguchi et al. / Experimental Hematology 35 (2007) 579–586
sulfate, 0.1% sodium deoxycholic acid, 100 mm orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 2 mM leupeptin, and 2 mM pepstatin A). Lysates were centrifuged at 12,000g for 15 minutes at 4C to remove debris, and protein concentrations were deter- mined with Lowry method. Of each cell extract, 50 mg was re- solved with 15% sodium dodecyl sulfate-polyacrylamide gels and transferred to Immobilon-P (Millipore, Billerica, MA). The membrane was blocked in 5% nonfat milk in TBS-T-hybridized sequentially with primary antibodies and horseradish peroxi- dase–conjugated anti-immunoglobulin secondary antibody (ICN, Eschwege). Anti-mouse p47-phox antiserum was as described pre- viously [15]. Bound antibodies were detected by enhanced chem- iluminescence (Amersham Pharmacia).
Results
G-CSF induction is associated with downregulation of SBDS in 32Dcl3 cells SBDS protein was highly expressed in the hematopoietic myeloid cells, 32Dcl3 and HL-60, and its molecular weight size was 32 kd in both murine and human cell lines (Fig. 1A). 32Dcl3 is a nonleukemic myeloid cell line that proliferates in the presence of mIL-3 and differentiated with G-CSF stimulation [16,17]. We found that G-CSF
Figure 1. SBDS was downregulated during G-CSF differentiation. (A):
32Dcl3 cells were washed and transferred from IL-3- to G-CSF-containing
medium, and cell lysates were prepared at the times indicated. Protein
extract (50 mg) was resolved on 15% sodium dodecyl sulfate-polyacryl-
amide gel and transferred to polyvinylidene difluoride. The membrane
was probed with antibody against SBDS and p47-phox. (B): 32Dcl3 cells
cultured in IL-3 and G-CSF. (C): 32Dcl3 cells were washed with PBS and
transferred to IMDM/10% FCS medium for 12 hours, then IL-3 was added
and cultured for another 12 hours.
clearly downregulated SBDS expression in 32Dcl3 cells within 24 hours and SBDS was completely undetectable after 3 days. This downregulation of SBDS was not ass- ociated with inhibition of neutrophil differentiation as p47- phox, a component of NADPH oxidase in neutrophils, was detected from day 3. This downregulation of SBDS was completely suppressed by costimulation with interleukin-3 (IL-3) and G-CSF (Fig. 1B).
To show whether SBDS is a downstream target of IL-3, IL-3 was withdrawn from culture for 12 hours, and the cells then restimulated with IL-3 for another 12 hours. As shown in Figure 1C, SBDS was not downregulated after IL-3 withdrawal. These results further confirm that G-CSF- induced downregulation of SBDS occurs in response to differentiation.
G-CSF-induced differentiation is unimpaired in SBDS knockdown cells Two independent shRNAi-infected 32Dcl3 cell lines were selected for investigation; both exhibited downregulation of the SBDS protein, although to different levels (Fig. 2A), and both lines proliferated in mIL-3 (Fig. 2B), although to a reduced saturation density compared with control cells. After withdrawal of IL-3 and culture in G-CSF only, the viability of the SBDS knockdown cells was reduced compared with controls; at 24 hours survival was 44% and 15% in control and SBDS knockdown cells, respectively, and at 48 hours, virtually all the knockdown cells were dead, whereas 16% of controls were still alive (Fig. 2B). However, all surviving cells showed mature segmented nuclei (Fig. 2C), and at day 7 after G-CSF treat- ment, there was no difference between control and SBDS knockdown cells with respect to expression of the neutro- phil maturation markers Mac1 and Gr-1, and G-CSF receptor (Fig. 2D). Furthermore, as shown in Figure 2E, expression of NADPH oxidase components were readily detected in G-CSF-induced control and SBDS knockdown cells by RT-PCR. Especially gp91phox was clearly induced after G-CSF treatment. In correlation with expression of NADPH oxidase components, superoxide anion production was same as control in SBDS knockdown cells (Fig. 3). To confirm complete neutrophil maturation of SBDS knock- down cells, we also tested for expression of the secondary granule gene lactoferrin and the third granule gene gelatinase. As demonstrated in Figure 2E, both control and SBDS knockdown cells demonstrated upregulation lactoferrin and gelatinase expression following induction. These data strongly suggest that SBDS knockdown cells differentiated into mature neutrophils and that SBDS is not rate limiting for neutrophil differentiation in our model. We also undertook RT-PCR for the SBDS (Fig. 2E), which clearly shows that after 28 cycles of amplification, there was a significant decrease in SBDS expression. This con- firms that we have not selected for cells that have escaped from the shRNAi effect.
582 M. Yamaguchi et al./ Experimental Hematology 35 (2007) 579–586
SBDS knockdown cells were sensitive in apoptotic inducing treatment It is reported that programmed cell death is central to the pathogenesis of myelodysplastic syndrome [18], and this
may also be the case for SDS [19]. As shown in Figure 4, differentiated SBDS knockdown cells demonstrated in- creased apoptosis compared with control cells. To inves- tigate whether SBDS knockdown cells are more sensitive
Figure 2. Characterization of SBDS knockdown cells. (A): SBDS shRNAi oligo (RNAi1 and RNAi2) was induced with lentivirus. Lysates of control virus-
infected 32Dcl3 and RNAi cells were obtained from cells growing in IL-3. Each protein extract was subjected to Western blot analysis. The membrane was
probed with against SBDS and tubulin. (B): Mean cell proliferation (from three independent experiments) of control infected 32Dcl3 cells (circle), RNAi1
cells (triangle), and RNAi2 cells (square). Cells were maintained in IL-3, then washed with PBS and transferred to medium containing mIL-3 (open symbol),
5 ng/mL mG-CSF (dotted line, solid symbol), or without cytokine (solid symbol). Viable cells were counted. Significance is marked with a cross when *p ! 0.05. (C): The morphologic characteristics of control 32Dcl3 and RNAi cells visualized using May Grunwald staining. Figure 2, continued. (D): FACS anal-
ysis of Mac1, Gr-1, and G-CSF receptor in control 32Dcl3 and RNAi cells cultured in IL-3 or treated with 5 ng/mL G-CSF for 7 days. (E): Gene expression
of the NADPH oxidase components gp91phox, p22phox, p47phox, and p67phox, and granule enzyme lactoferrin, and gelatinase in uninduced (day 0) and
differentiated 32Dcl3 cells (day 7). RNA samples were reverse-transcribed to generate cDNAs, which were then analyzed by PCR for amplifications, which were
performed for the indicated number of cycles. Also shown is hypoxanthine phosphosibosyl transferase, amplified from the same cDNAs for each condition.
583M. Yamaguchi et al. / Experimental Hematology 35 (2007) 579–586
Figure 2. (Continued).
than control cells to an extrinsic apoptotic signal, control and knockdown 32Dcl3 cells were cultured in 1 ng/mL IL-3 supplemented with 1% and 10% FCS and compared (Fig. 5). Whereas in control cells there was no difference between 1% and 10% FCS in terms of cell death rates
(137%), in the SBDS knockdown cells, there was in- creased evidence of cell death in 1% compared with 10% FCS (162% and 206%). These results suggest that SBDS is important for maintaining viability of granulo- cyte precursor cells.
584 M. Yamaguchi et al./ Experimental Hematology 35 (2007) 579–586
Discussion SDS is a rare autosomal-recessive disorder affecting multi- ple organs with a wide range of clinical severity. It is char- acterized by pancreatic insufficiency in children and bone marrow failure syndrome [1]. The syndrome is caused by mutations in the SBDS gene on chromosome 7 [2–5]. The SBDS gene function is presumed to function in RNA
Figure 3. Superoxide anion production by differentiated SBDS knock-
down cells in response to PMA. Control infected and SBDS knockdown
cells were induced with G-CSF for 7 days were stimulated with PMA.
Data represent the mean 6 SD value from three independent experiments.
processing [2,6,7], but no physiologic function has been ascribed to date.
We generated an antibody that recognized both murine and human SBDS protein and used this to undertake
Figure 4. Apoptosis rate of differentiated neutrophils. Control 32Dcl3 and
RNAi cells were differentiated with 5 ng/mL G-CSF for 7 days. Viable
cells were stained with Annexin-V and 7-AAD and analyzed by FACS.
Data represent the mean 6 SD value from three independent experiments.
Figure 5. Apoptosis rate of cultured in low serum concentration. Control 32Dcl3 cells and RNAi cells were transferred to medium containing 10% FCS or
1% FCS and supplemented with 1 ng/mL IL-3 overnight. Cells were stained with Annexin-V and 7-AAD and analyzed with FACS. (B): Apoptosis rate of
cultured cells from three independent experiments, obtained using an Annexin-V/7-AAD method. Annexin-V-positive cells were compared from 10% FCS
and 1% FCS treatment. Data represent the mean 6 SD value from three independent experiments. Significance is marked with a cross when *p ! 0.05.
585M. Yamaguchi et al. / Experimental Hematology 35 (2007) 579–586
functional analyses of SBDS. SBDS was downregulated by G-CSF in 32Dcl3 cells, but this downregulation could be overcome by the addition of IL-3. When cells were grown in the absence of IL-3, we did not observe downregulation of SBDS. These results indicate that SBDS is downregu- lated in response to differentiation and suggest that SBDS is not an IL-3-responsive gene.
G-CSF-induced neutrophil maturation was not impaired in SBDS knockdown cells. This is in accordance with SDS patient studies that have demonstrated that CD34þ
cells from patients lacking SBDS could differentiate to functional neutrophils [5,10,20,21]. Together, these data indicate that SBDS is not critical for neutrophil maturation.
It is reported that 32Dcl3 fails to produce superoxide an- ion due to the absent expression of gp91phox [22,23]. How- ever, Nakajima and Ihle reported that PMA-stimulated neutrophils, which were differentiated from 32Dcl3, showed nitro blue tetrazolium reduction activity [24]. Our data are in agreement of those of Nakajima and Ihle; gp91phox was clearly induced after G-CSF stimulation (Figs. 2 and 3). The observed discrepancies in these find- ings may be due to clonal variation of 32Dcl3 cells.
It was previously reported that bone marrow failure in SDS is attributable to increased apoptosis, mediated by highly expressed Fas [19]. The SBDS knockdown cells in 32Dcl3 did not show an increased rate of apoptosis in IL- 3 proliferation. However, SBDS knockdown cells were compromised for survival under conditions where the anti- apoptotic signal was less than optimal. These data indicate that SBDS knockdown cells have an enhanced sensitivity to extrinsic apoptotic signals, and it is likely that a significant number of neutrophil progenitors underwent apoptosis be- fore maturation. Indeed, an increased proportion of apopto- tic fragments and preapoptotic neutrophils were demonstrated in SBDS knockdown compared with control neutrophils. It has been reported that low expression of anti- apoptotic protein, Bcl-2, or Bcl-XL in myeloid cells causes severe congenital neutropenia [25–27], but no differences in Bcl-2 or Bcl-XL expression were demonstrated between control and SBDS knockdown cells (data not shown). After the gene responsible for SDS was identified, there was much speculation about its physiologic function. In the cur- rent study, we have demonstrated that SBDS is not required for granulocyte cell maturation but may act to maintain survival of neutrophil progenitor cells. Recently, it was reported that loss of SBDS results in early embryonic lethality [28]. The development of SBDS/ embryos was arrested early than embryonic day 6.5, leading to early le- thality with markedly muted epiblast development. It is likely that residual expression of SBDS in the SBDS knock- down cells in the current study provided protection against cell death.
In summary, this study reveals an essential role for SBDS in granulocyte precursor cell survival.
Acknowledgments This work was supported in special grant from Hiroshima Interna- tional University.
References 1. Dror Y, Freedman MH. Shwachman-Diamond syndrome. Br J Haema-
tol. 2002;118:701–713.
2. Boocock GR, Morrison JA, Popovic M, et al. Mutations in SBDS are
associated with Shwachman-Diamond syndrome. Nat Genet. 2003;33:
97–101.
3. Kawakami T, Mitsui T,…