YBCMD-01804; No. of pages: 10; 4C: Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Blood Cells, Molecules and Diseases journal homepage: www.elsevier.com/locate/bcmd

The RNA in reticulocytes is not just debris: It is necessary for the final stages of erythrocyte formation EunMi Lee a, Hye Sook Choi a, Jung Hye Hwang b, Jeong Kyu Hoh b, Youl-Hee Cho c, Eun Jung Baek a,⁎ a b c

Department of Laboratory Medicine, College of Medicine, Hanyang University, Seoul, Republic of Korea Department of Obstetrics and Gynecology, College of Medicine, Hanyang University, Seoul, Republic of Korea Department of Genetics, College of Medicine, Hanyang University, Seoul, Republic of Korea

a r t i c l e

i n f o

Article history: Submitted 1 October 2013 Revised 4 February 2014 Available online xxxx (Communicated by B. Bull, M.D., 06 February 2014) Keywords: Reticulocyte mRNA Red blood cell Protein biosynthesis Erythropoiesis Ribonuclease

a b s t r a c t Reticulocytes contain both RNA and micro-organelles and represent the last stage of erythropoiesis before full maturation to red blood cells (RBCs). Even though there is continuing synthesis of hemoglobin and membranebound proteins in reticulocytes, the small amount of RNA that they contain has been regarded as nonfunctional residual material. Here we show that this residual RNA is both functional and essential for further reticulocyte maturation. Reticulocytes from which the remnant RNA had been removed by exposure to RNase did not survive or mature into RBCs in either humans or mice. Conversely, reticulocytes treated with an RNase Inhibitor were able to form normal biconcave cells. Similarly, poor survival was also seen in reticulocytes in which protein synthesis had been blocked. To identify the signaling pathways involved we isolated RNAs in reticulocytes versus those present in fully matured erythroblasts cultured from hematopoietic stem cells. RNAs found in erythroblasts were related to exocytosis, metabolism, and signal transduction all of which are critical for maturation through reticulocyte and into a fully mature, biconcave erythrocyte. Our results suggest that the mRNA in reticulocytes has to be translated into novel proteins that act to preserve mitochondria and maintain cell membrane integrity as reticulocytes mature. These results enhance our understanding of the final stage of erythropoiesis and may clarify why in vitro-generated reticulocytes for transfusion purposes survive poorly. © 2014 Elsevier Inc. All rights reserved.

Introduction Reticulocytes represent the last stage of erythropoiesis in human and other mammalian species before full maturation to red blood cells (RBCs). Young reticulocytes still contain ribonucleic acid (RNA), mitochondria, ribosomes, and lysosomes. As they mature, they expel residual subcellular micro-organelles via degeneration, autophagy, and mitoptosis, a mitochondrial death program [1]. In addition, a variety of structural and biochemical changes occur such as loss of the ability to synthesize DNA and, eventually, the complete cessation of protein synthesis. For about two to five days following enucleation in the bone marrow, discoid-shaped reticulocytes become biconcave RBCs by vesiculation and endocytosis. In the process they lose 20% of their membrane surface, along with tubulin and cytosolic actin [1–4]. The amount of RNA and remnant organelles in a reticulocyte is small and has generally been thought as debris, limited to possibly providing

⁎ Corresponding author at: Department of Laboratory Medicine, College of Medicine, Hanyang University, 153, Gyeongchun-ro, Guri-si, Gyeonggi-do, ZIP: 471-701, Republic of Korea. Fax: +82 31 560 2585. E-mail address: [email protected] (E.J. Baek).

supportive functions for several genes active in hemoglobin synthesis and organelle degradation [5]. In the same context, even though some protein is known to be produced, whether or not this newly-made protein is essential for the survival and maturation of reticulocytes to RBCs remains unclear. A progressive decrease in RNA content is characteristic of terminal erythropoiesis. At this stage, RNA degradation is the sole means of altering the cellular content of RNA, because the mammalian reticulocyte is unable to synthesize RNA [6]. Even though reticulocytes are no longer active in DNA and RNA syntheses [7], the small amount of RNA that remains might still be active in the final maturation process. Interestingly, the disappearance of RNA takes place simultaneously with the loss of subcellular structures such as the mitochondria, ribosomes, and endosomal vesicles. However, the ribosomes remaining in reticulocytes still have the ability to synthesize protein, as has been proven in cell free systems utilizing polyribosomes extracted from reticulocytes [8,9]. The residual molecular factors controlling transformation of discoid reticulocytes to biconcave cells are poorly understood. We speculated that if this transformation is not just a result of mechanical and physical phenomena following the disappearance of non-heme proteins and RNA, then discrete and specific signaling events might be guiding this

http://dx.doi.org/10.1016/j.bcmd.2014.02.009 1079-9796 © 2014 Elsevier Inc. All rights reserved.

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

2

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

transformation. Furthermore, there is no information about which specific kinds of RNA remain in reticulocytes or how their spatial distribution changes during the two days when reticulocytes are still confined to the bone marrow and are in the process of expelling their micro-organelles. Finally, it is not known whether the disappearance of all the RNA is necessary to facilitate the disappearance of other micro-organelles. Confirming whether the remaining RNA is critical for maturation, or in contrast, must be removed, would help to clarify the processes of terminal RBC maturation, autophagy and organelle disappearance. Elucidation of the role of RNA is also important for the in vitro production of RBCs for transfusion purposes, since it would clarify whether low reticulocyte viability is due to degeneration of RNA by RNases or inhibition of RNA removal by RNase Inhibitors under these culture conditions. A better understanding of the RNA function under these circumstances might lead to the development of more efficient generation of RBCs from stem cells for human transfusion. Our results show that the small amount of remnant RNA is functional and is not just debris or waste material but must be translated into proteins if the reticulocyte is to survive and mature. Also, we note that, contrary to the clumped material typically seen in supra-vital stains, RNA is diffusely distributed in the cytoplasm of viable reticulocyte, and the up-regulated RNAs are primarily directed toward signal transduction, metabolism, and transport. The removal of this RNA blocks the maturation of reticulocytes and leads to loss of membrane integrity and expulsion of mitochondria.

Approximately 300–500 μl of blood was then procured by retro-orbital puncture on days 1 and 3. On day 5, reticulocyte rich blood samples were obtained by cardiac puncture or by vena caval puncture after anesthesia. Thereafter, reticulocytes were isolated and leukocytes removed as described above for human reticulocytes. These experiments were approved by the Institutional Review Board of Hanyang University. Reticulocyte culture in the presence of RNase or RNase Inhibitor Reticulocytes were incubated at 5 × 105 cells/ml in Iscove's modified Dulbecco's medium (Gibco BRL; Invitrogen, Carlsbad, CA) containing 100 IU/ml penicillin, 100 mg/ml streptomycin, and 10% fetal bovine serum (FBS, Gibco) in a 5% CO2 incubator (Sanyo, Tokyo, Japan) at 37 °C for 24–48 h. Dose-dependent damage was measured to determine the required concentrations of RNase and RNase Inhibitor. The RNA in reticulocytes was destroyed by treating them with RNase A (Sigma) at 20 ng/μl. In other experiments, RNA degradation by RNase was blocked by RNase Inhibitor (10 IU/μl; Sigma). Each experiment was done in duplicate in 24-well plates, and cells were counted in duplicate after Trypan blue staining. RNA quantification after reticulocyte culture Reticulocytes incubated as described above were washed with cold phosphate buffered saline. Then, RNA was isolated with Trizol (Invitrogen) and measured in duplicate with a Nanodrop (BioSpec Nanospectrophotometer, Shimadzu, Japan).

Materials and methods mRNA expression microarrays Isolation of reticulocytes from human peripheral blood Use of all samples was approved by the Institutional Review Board of Hanyang University Kuri Hospital. Fresh peripheral blood samples were collected in EDTA bottles. Reticulocyte counts in high reticulocyte specimens (6–13%) were performed both manually and by CBC analyzer (Beckman Coulter, Miami, FL). The samples were examined visually to confirm that the RBC morphology was normal. After the removal of plasma by centrifugation, the reticulocytes were concentrated by density gradient centrifugation [10,11]. Following centrifugation, approximately 200 μl of the cells in the top layer of the gradient was collected and used in experiments. White blood cells were removed with a leukocyte reduction filter, specified to remove more than 99.99% of the leukocytes. Reticulocyte counting by vital RNA staining As the reticulocyte counting by supravital staining is known to be accurate for blood samples with at least 5% reticulocytes [12], the isolated reticulocytes were counted after brilliant cresyl blue (BCB) staining: blood was incubated with a 0.5% solution of new methylene blue, which precipitates RNA and stains it blue. In each experimental preparation cells were counted by two experts at least 3 times. There was a visual check on each sample to ensure that white blood cells and platelets had been removed. Induction of reticulocytes in mice The fraction of reticulocytes in patients with anemia rarely exceeds 15%. For reasons of patient safety, the amount of whole blood available from anemic patients is essentially limited to the 1–2 ml of a typical blood draw; therefore, other approaches had to be used when higher reticulocyte counts were required. This was done by bleeding animals as previously described by Joiner et al. [13] with slight modifications. In brief, eight-week-old male BALB/c mice were injected intraperitoneally with 1 ml of normal saline, and a pain reliever (ibuprofen, 7.5 mg/kg; Kwang Dong Pharma, Seoul, Korea) was administrated orally.

To identify candidate genes expressed in the residual RNA in reticulocytes, we compared the mRNA expression profile of reticulocytes with that of orthochromatic erythroblasts. The reticulocytes were isolated from fresh umbilical cord blood (CB). The orthochromatic erythroblasts represent the stage of erythropoiesis just before enucleation and were cultured from hematopoietic stem cells (HSC) isolated from human umbilical cord blood samples [14]. After obtaining informed consent from healthy pregnant women, CD34 + cells were then separated from cord blood samples using an EasySep CD34 isolation kit (StemCell Technologies, Vancouver, BC, Canada). The culture methods used are described in previous reports [15,16]. Briefly, several cytokines were added to stroma- and serum-free medium in a stage-specific manner to induce CD34+ cells to differentiate into a purely erythroid lineage in the absence of feeder cells and serum/plasma. Viable cells were counted by Trypan blue staining. Samples from three draws were each analyzed in triplicate using a NimbleGen human whole genome mRNA expression microarray (Roche, Indianapolis, IN). Quantitative RT-PCR Reticulocyte RNA was isolated before and after incubation using Trizol (Invitrogen) according to the manufacturer's instructions. It was reversetranscribed using SuperScript III Reverse Transcriptase (Invitrogen). Levels of mRNAs were then measured in triplicate by qRT-PCR using SYBR Premix ExTaq (Takara-Bio, Shiga, Japan). The sequences of the primers used were as follows (forward/reverse), and other primer sequences used were described previously [10]: Homo sapiens ribosomal protein L23 (RPL23), 5′-AAGGGACGGCTG AACAGACT-3′/5′-TCGAATGACCACTGCTGGAT-3′; Homo sapiens vesicleassociated membrane protein 2 (VAMP2), 5′-GCCCCTCTAGTCCAGTTT GC-3′/5′-ACCCCATGGGAATGTCAGTT-3′; Homo sapiens potassium large conductance calcium-activated channel (KCNMB4), 5′-GAGA TTGGTTCCCAGCCATT-3′/5′-CAGGACCACAATGAGAACGC-3′; Homo sapiens synaptophysin-like 1 (SYPL1), 5′-AGGTCCTCGAGTGGATTG CT-3′/5′-TACACCTGGAGGTGGCTGAA-3′; Homo sapiens calreticulin (CALR), 5′-GTGGTGTGGAGAAGCCACAG-3′/5′-GAAAGGGAGGGGTG

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

3

Fig. 1. Effects of RNase and RNase Inhibitor on in vitro maturation of reticulocyte and changes of residual RNA. Human (A) and mouse (B) peripheral blood samples containing many reticulocytes were incubated with RNase or RNase Inhibitor (RNase Inhibitor) for 24 h in a 37 °C incubator, and reticulin was stained with brilliant cresyl blue. RNA precipitates in reticulocytes are shown as blue granules (arrows). The photographs are representative of at least 8 human and mouse samples, respectively. Magnification, ×400; scale bars, 50 μm. (C) After removing leukocytes using a leukocyte filter, possible damage caused by the RNase and RNA inhibitor and not related to RNA effects was assessed using human adult peripheral blood erythrocytes from healthy donors; the results showed the poorest survival in RNase Inhibitor-containing medium, followed by RNase containing media (n = 2). (D) The RNA remaining in reticulocytes after treatment with RNase or RNase Inhibitor for 24 h was compared to 0 h reticulocytes. (E) Comparison of the amounts of RNA in each condition. Data are expressed as mean ± SEM. (n = 11; ANOVA and Bonferroni's correction for multiple comparisons), *: P b 0.05; **: P b 0.01. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

AAATGA-3′; Homo sapiens GrpE-like 1, mitochondrial (GRPEL1) 5′CAGACGTTCTGGAGAAGGCA-3′/5′-GGTTCAACTTGAGCAAGCCA-3′.

RNase and RNase Inhibitor toxicity tests Fresh RBC preparations containing less than 1% reticulocytes were incubated with RNase and RNase Inhibitor for 24 h. Then viable RBCs were counted by Trypan-blue staining.

Spatial distribution of reticulocyte RNA The spatial distribution of RNA was observed in real time by fluorescence microscopy using RNA tagging reagents. After incubation of reticulocytes in separate wells for 1 h, 2 h, 3 h, and 24 h in a CO2 incubator at 37 °C, 500 nM of SYTO RNA Select (Invitrogen) was added for 20 min. Then the distribution of RNA was observed and recorded. As a negative control, reticulocyte-free RBC preparations were used.

Inhibition of new protein synthesis in reticulocytes To confirm the presence of newly synthesized proteins and determine whether they had essential functions in RBC maturation, blood samples with high reticulocyte counts were placed in methionine-free medium and the survival of the reticulocytes was compared to that in control medium. Cells were enumerated in duplicate after 24 h incubation at 37 °C. To confirm that the dead cells were reticulocytes and not mature RBCs, blood samples were stained after incubation and inspected for the percentage of remaining reticulocytes. Statistical analysis Statistical analyses were performed using GraphPad InStat version 3.00 (San Diego, CA). Significant differences between responses obtained in each experimental condition were analyzed using two-tailed paired t tests unless otherwise specified. Where indicated, ANOVA (with Bonferroni's correction for multiple comparisons) was used to compare several groups. Differences were considered statistically significant at P values less than 0.05.

Transmission and scanning electron microscopy

Results

Reticulocytes were centrifuged and washed after incubation. The cells were then fixed and photographed with a transmission electron microscope (TEM; JEM 1011, Jeol, Eching, Germany) and a field emission-scanning electron microscope (SEM; Hitachi S-800, Tokyo, Japan). Images were recorded with a scanning microscope image analysis system (ESCAN-4000; Bummi Universe, Seoul, Korea) [16].

Reticulocyte isolation and concentration Fresh peripheral blood was collected from patients with severe, blood loss anemia. The causes of blood loss were mainly gastrointestinal hemorrhage, epidural hemorrhage, hepatic failure, and kidney diseases. The mean age of the patients (5 men and 5 women) was 54.6 years

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

4

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

Fig. 2. Time-lapse analysis of fluorescence-labeled RNA in reticulocytes. Blood containing 12% reticulocytes was filtered with a leukocyte removal filter and incubated in wells for 24 h at 37 °C. After collecting cells at successive times, the cells were stained with an RNA tagging reagent that does not trigger RNA precipitation, and we observed the distribution of the remnant RNA. RNA was diffusely spread throughout the cytoplasm rather than being restricted to microorganelles until it finally disappeared. The figures represent three independent experiments.

(40–69). To remove leukocytes, blood was filtered through a leukocyte removal filter and centrifuged to isolate the reticulocyte-rich layers. The gradient isolation method using Percoll [11] yielded not only a high level of reticulocytes but also a high level of platelets, which also contain RNA. Therefore, we used the Ficoll gradient method for most experiments [10]. With human samples, the purity of reticulocytes was up to 28% (mean 17.2%, n = 9). For the experiments in methionine-free medium, reticulocytes were 98% pure. To get higher levels and larger numbers of reticulocytes, a mouse anemia model was utilized which produced reticulocyte levels of 30–50% (mean 40%, n = 5, Figs. 1A and B). At least 1000 cells were counted per condition, and no residual leukocytes were detected. To observe the effect of RNA presence or absence, RNA in the reticulocyte-concentrated blood was either removed with RNase or maintained by using RNase Inhibitor to inhibit the action of native RNases. Few RNA precipitates were observed in the cells incubated with RNase for 24 h, while samples treated with RNase Inhibitor had similar or slightly higher percentage of reticulocytes and higher RNA amount than the control. These results demonstrate that RNase A removed RNA effectively (Figs. 1A and B).

Survival of reticulocytes after exposing RNA to RNase or RNase Inhibitor To investigate the possibility that the reticulocyte changes observed were due to direct cell membranes damage by RNase or RNase Inhibitor and not to effects on RNA, mature RBCs were treated with the two reagents. The immediate damage caused by the two treatments was similar, with the RNase Inhibitor causing more damage than the RNase (19.1% and 11.7%, respectively) (Fig. 1C). 24 h later, the total viable cell counts were affected in a similar way, again with more damage in the cell preparations exposed to the inhibitor (N = 11; Fig. 1D).

Effect of RNase and RNase Inhibitor activity on RNA levels The amount of remnant RNA remaining in cells incubated with RNase and RNase Inhibitor was quantified. The mean amount of RNA in 2 × 106 reticulocytes at 0 h was 983.9 ng/μl and decreased after 24 h to 695.4 ng/μl in the control, to 214.3 ng/μl in the RNase treated cells and to 440.0 ng/μl in the RNase Inhibitor condition where the RNase had been neutralized by RNase Inhibitor. These results show that the

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

5

Fig. 3. Role of RNA in reticulocyte maturation and membrane stabilization. (A) The pictures taken by scanning electron microscopy show that reticulocyte samples treated with RNase contain increased numbers of echinocytes (arrows), which have abnormal spiked cell membranes, as well as reduced numbers of discoid or flat-shaped reticulocytes. (B) Samples treated with RNase Inhibitor contained similar numbers of echinocytes to control samples. The figure and graph represent two independent experiments counting more than 500 cells.

reagents were effective in degrading and protecting RNA, respectively (n = 4, P = 0.02, ANOVA, Bonferroni's correction) (Fig. 1E). Reticulocyte RNA in is dispersed throughout maturation The spatial distribution of RNA is altered when it is precipitated by typical reticulocyte staining methods using brilliant cresyl blue or acridine orange, and the distribution of remnant RNA over the course of its removal from reticulocytes has not previously been described. We observed RNA distribution in real time by tagging the RNA with fluorescent tags, which aid not to cause RNA aggregation. RNA was diffusely spread throughout the cytoplasm of the reticulocytes without any increased concentration around the mitochondria, autophagic vacuoles, or membranes (Fig. 2). The fluorescent-tagged RNA was diffusely distributed until it had completely disappeared. The RNA fluorescence decreased rapidly over the first 6 h and by 24 h could be detected in only a few cells. The role of RNA in reticulocyte maturation to biconcave RBCs To investigate whether the remnant RNA affected the transformation from reticulocyte to biconcave RBCs, cells of each type were observed by SEM. In the control samples at 0 h, both discoid reticulocytes and biconcave RBCs were observed. In the controls at 24 h reticulocytes had decreased in number relative to the intact biconcave RBCs, suggesting that maturation had occurred. However, there was also a markedly

increased number of echinocytes. Echinocytes have spiked membrane projections and are seen when RBCs lack ATP or are placed in an albumin-free, alkaline diluent (Fig. 3). Reticulocytes treated with RNase did demonstrate a markedly increased number of echinocytes. In contrast, reticulocytes treated with RNase Inhibitor yielded fewer echinocytes but a slightly higher percentage of biconcave RBCs. Our results indicate that the enzymatic removal of residual RNA blocked the final maturation to normal biconcave RBCs and decreased the number of intact RBCs (Figs. 3A and B). Effect of RNA removal on mitochondria breakdown during reticulocyte maturation To investigate mitochondrial clearance and to study the changes in other remaining micro-organelles resulting from the manipulation of RNA, we observed the mitochondria by TEM (Fig. 4A). At 0 h cells treated with RNase demonstrated damaged plasma membranes with pore formation. Inside these cells the numbers of autophagic vacuoles containing poorly degraded organelles had increased. At 24 h of incubation with RNase intact mitochondria decreased in number and their inner mitochondrial membranes were blurry (Fig. 4B). Reticulocytes treated with RNase Inhibitor contained large autophagic vacuoles and double the concentration of the mitochondria as the 24 h control and RNase samples. This suggests that when enzymatic degradation of RNA is prevented, mitochondria are better preserved, and remaining intracellular organelles are finally cleared inside autophagic vacuoles.

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

6

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

Fig. 4. Reticulocyte mitochondrial changes following exposure to RNase or RNase Inhibitor. (A) Mouse reticulocytes were incubated in the presence of RNase or RNase Inhibitor (RNase Inhibitor) for 24 h and analyzed by transmission emission microscopy. More mitochondria and vacuoles remained in the RNase Inhibitor condition. (B) Numbers of mitochondria per cell were analyzed by counting more than 200 cells. Mitochondria (M), vacuoles (V). Scale bars, 2 μm.

The mRNAs in reticulocytes belong mainly to transport and signal transduction groups To explore which new genes are involved in reticulocyte maturation, CD34+ cells were allowed to differentiate and mature to erythroid cells and then mRNA microarray analysis was performed. Erythroid cell purity was more than 95% (Fig. 5A). On day 17–20 of culture, mature orthochromatic erythroblasts were collected and RNA was isolated. As the expression level of most RNAs decreases markedly from orthochromatic erythroblasts to reticulocytes, any relatively elevated mRNA persisting in reticulocytes might reasonably be suspected of playing an important role in erythropoiesis. Of the 44,000 human genes analyzed, 46 showed at least 2-fold increased expression in reticulocytes than in orthochromatic erythroblasts. Of the 46, 22 genes fell into three (overlapping) categories: metabolic (15 genes), transport (8 genes), and signal transduction (4 genes). Collectively, the processes specified by these categories might well be involved in expelling remaining micro-organelles from the cells and generating the energy required for this process from the mitochondria (Table 1). We examined further the transport-related genes, as their functions likely direct the final exocytotic activity of reticulocytes. To determine whether RNase and RNase Inhibitor differentially affected any particular gene type, the expression levels of 5 genes (SYPL1, CALR, VAMP2, RPL23, GRPEL1) were quantified by qPCR. Expression of all 5 genes decreased

during 24 h of incubation. The expression of some of these genes differed from that of the overall RNA, perhaps due to structural stability or because they were being translated. The expression of most other genes followed the overall decline in RNA levels in both the RNase and RNase Inhibitor conditions (Fig. 5B). Specifically, the CALR and GRPEL1 genes did not follow the general pattern of decreased expression following RNase treatment and greater stability following RNase Inhibitor treatment, suggesting that the stability or sensitivity of these genes to these agents was different (Fig. 5). However, there were wide variations in expression level depending on the particular transcript, especially at 0 h, and we failed to detect any statistically significant difference in transcript levels between the RNase and RNase Inhibitor treatment. Inhibition of new protein synthesis in reticulocytes leads to cell death If remnant mRNA has any essential function in the process of reticulocyte maturation, newly synthesized proteins would be expected to have some functions as well. To confirm that new protein is critical for maturation, reticulocytes were placed in methionine-free medium. Not surprisingly, in both human and mouse, hemolyzed cells increased markedly in this medium (Fig. 6A). As expected, the number of cells lost corresponded to the number of reticulocytes that disappeared, suggesting that the dead cells were mainly reticulocytes (Figs. 6B and C). Also, when the data from experiments involving human and mouse were

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

7

Fig. 5. Relative changes of mRNA in reticulocytes. (A) For the microarrays, mature erythroblasts were cultured from cord blood hematopoietic stem cells. Proerythroblasts on culture day 5, basophilic erythroblasts on day 8, and orthochromatic erythroblasts on day 17 are shown. Wright–Giemsa staining, ×400. (B) The extents of RNA clearance of individual mRNA following incubation at 37 °C for 24 h were different. Some mRNAs behaved like the total mRNA in response to RNase and RNase Inhibitor, but others did not, indicating that their responses and sensitivities were different (n = 5, qPCR in duplicate).

combined, the difference achieved statistical significance (P = 0.02, n = 5) (Fig. 6D). This result implies that the translation of remnant RNA is required for reticulocytes to survive and mature into RBCs.

Discussion Reticulocyte maturation to mature RBCs continues after enucleation of late stage erythroblasts, when new mRNA generation ceases. Reticulocytes, however, still contain residual RNA. Even though there have been reports of ongoing hemoglobin synthesis in reticulocytes [5], it was not clear whether remnant RNAs and newly synthesized proteins were critical for reticulocyte maturation or just helpful in the process. In particular, the specific kinds of RNA that remain and the spatial distribution of this RNA prior to complete disappearance have not been characterized. A growing body of evidence has shown the potential of blood products produced in vitro, including reticulocytes and RBCs, to supplement or even replace limited supplies of donated blood. Many researchers have set up in vitro RBC generation systems and found that the reticulocytes produced are defective. Initially, the products of RBC generation systems are predominantly reticulocytes, and ideally these should mature to stable, fully functional RBCs after transfusion, while reticulocytes of poorer quality would likely have a shorter in vivo survival and be more fragile in vitro. Therefore, we speculated that elucidating the mechanism of RNA clearance and the role of newly synthesized proteins might clarify the rapid degradation and defective maturation of reticulocytes in RBC-generating systems and so improve RBC production in vitro.

We entertained two possible scenarios: 1. RNA naturally disappears within several days, therefore, delayed RNA clearance from reticulocytes might block the natural maturation process; and 2. Remnant RNA might still play an essential role in reticulocyte maturation and time is needed for it to be translated to proteins with essential functions before it is cleared. Regarding the hypothesis that RNA must be removed quickly for maturation to proceed, the RNase activity of erythroid cell lysates has been shown not to be energy dependent [9], and inhibition of RNase activity appeared to retard reticulocyte maturation [9]. Furthermore, the organisms that have been studied to date contain many RNases of different types, suggesting that RNA degradation is a very ancient and important process. Finally, RNases also have an important role in the normal turnover and maturation of erythroblasts during the maturation of erythroid cells. Like protein degradation, the loss of RNA is so fast that 20% of the initial RNA is lost in 4 h at 37 °C [17]. On the other hand, while loss of the remnant RNA and microorganelles might be necessary for RBC maturation, forcible premature removal of RNA might block functions necessary for terminal maturation. Ribonuclease activity is known to gradually decrease beginning in orthochromatic erythroblasts, and is 30–50 fold lower in reticulocytes than in erythroblasts [7]. The decrease in activity might protect the remnant RNA in reticulocytes and allow it to perform certain necessary functions [7]. Free RNase Inhibitor activity is not present in mononuclear cells, leukocytes, granulocytes, and plasma, but is present in erythrocytes [18]. Specifically, the relative RNase Inhibiting activity (calculated as total RNase Inhibiting activity / RNase) is more than 18, which is much higher than that of mononuclear leukocytes (0.58) or granulocytes (0.07) [18,19]. The decreased RNase activity and relatively

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

8

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

As RNase A could not completely remove mRNA, it could not prevent some translation, depending on the types of RNA. The commonly used protein synthesis inhibitor, cycloheximide, does not block translation in mitochondria. This continued protein synthesis might play a role in reticulocyte maturation. Therefore, we incubated reticulocytes in methionine-free medium since methionine is essential for protein synthesis. Protein synthesized in reticulocytes is largely globin comprising more than 85% of the protein produced [22]. The remaining proteins are not well characterized. As expected, reticulocytes failed to mature to intact RBCs in methionine-free medium, and echinocytes, indicative of a reduced ATP content, increased. Even though we could not clearly determine whether energy-dependent mechanisms were damaged or membranepreserving signals were destroyed, we confirmed that the maturation of reticulocytes and maintenance of their membranes were affected by blocking de novo protein production, leading to inhibition of RBC maturation. To identify any important signals that might be involved in these outcomes, we consulted previously published studies but could not find any relevant data. Therefore, we analyzed microarrays and performed qPCR to identify any RNAs that were more highly expressed in enucleated reticulocytes than in the final orthochromatic phase. As the maturation of reticulocytes to RBCs in vivo involves a selective loss of reticulocyte membrane [23], it would be possible to have a higher concentration of some mRNAs in reticulocytes than in orthochromatic erythroblasts. The latter cells seem to generate some necessary signals until the point at which RNA production ceases. We found that the candidate genes were mainly related to transport, metabolism, and signal transduction, which agrees with the changes in the reticulocytes. Transport involvement is also consistent with a recent report that exocytosis and endocytosis are critical to the final maturation of reticulocytes [24]. However, as the efficiency of penetration of siRNAs and the relative changes of newly synthesized proteins are low, the relevant genes and proteins could not be identified. Also, in attempts to observe changes in the levels of specific mRNAs relative to the change of total RNA, we

high RNase Inhibitor level, together with the residual ribosomes, suggest that the remnant RNA may act to facilitate RBC maturation by triggering the synthesis of important new proteins in reticulocytes. The results of our experiments would be of great interest if they could clarify whether the small amount of remnant RNA in reticulocytes is made up of final messenger RNAs that are produced and remain in the cell after the nucleus is expelled or whether the RNA is just debris. Unlike the situation in normal reticulocytes where RNA is cleared in 2 to 5 days [20], we removed RNA abruptly using RNase. RNase A is a relatively small protein (13.7 kDa), which easily enters the cytosol, cleaves RNA, and consequently terminates protein synthesis [21]. While this treatment kills most nucleated cells, reticulocytes would not be expected to die or to be interrupted in their maturation to RBCs if reticulocytes did not need the production of new proteins from the remnant RNA. Our results suggest strongly that new protein synthesis from the residual RNA is critical for survival. From the observation that the percentage of echinocytes increased when RNase was added, and that mitochondria were also more numerous and better preserved when RNase Inhibitor was added, we conclude that reticulocytes need the ATP produced by mitochondria for at least 24 h to maintain their membrane integrity and complete the assembly of sub-membrane structural proteins. Therefore, a lack of energy production due to defective mitochondrial function, or inadequate removal of RNA, could be the reason for the fragility of reticulocytes generated in vitro. The toxicity of RNase could be due to its ability to remove reticulocyte RNA or to nonspecific cell damaging effects as the drug passes through the membrane and enters the cell. Unfortunately it is not experimentally possible to directly counteract the toxicity of inserted RNase with an inhibitor. Therefore, we tested its toxicity on biconcave mature RBC membranes. These do not show toxicity due to RNA removal and are the cells with membranes most similar to reticulocytes. We also tested the toxicity effect of RNase Inhibitor. Both reagents showed toxicity to RBC but RNase Inhibitor was more toxic as shown in Fig. 1C. However, the percentage of damaged cells was not significantly different between controls, RNase group, and RNase Inhibitor group.

Table 1 mRNAs with at least 2-fold more expression in reticulocytes than in orthochromatic erythroblasts, as measured by microarray, and categorization of their functions. Gene symbol

Product

Function classification

RPL23

Ribosomal protein L23

O

O

COX4I2 ATP11C GRPEL1 CALR

Cytochrome c oxidase subunit IV isoform 2 ATPase, Class VI, type 11C isoform b GrpE-like 1, mitochondrial Calreticulin

O O O O

O O O O

Gene roles

Transport Metabolism Signal transduction

Vesicle-associated membrane protein 2 O O Large potassium conductance calcium-activated channel, subfamily M, beta member 4 SYPL1 Synaptophysin-like 1 O DDI2 DNA-damage inducible 1 homolog 2 DEFA1 Defensin, alpha 1 FKBP11 FK506 binding protein precursor HIST1H2AB H2A histone family, member M DEFA1 Defensin, alpha 1 preproprotein S100A9 S100 calcium binding protein A9 ETNK1 Ethanolamine kinase 1 isoform A LONRF2 LON peptidase N-terminal domain and ring finger 2 ADSS Adenylosuccinate synthase FAHD1 FAHD1 protein RASGEF1A RasGEF domain family, member 1A LOC339047 LOC339047 protein FSHB Follicle stimulating hormone, beta polypeptide precursor NMB Neuromedin B isoform 1

Ribosomes, the organelles that catalyze protein synthesis, consist of a small 40S subunit and a large 60S subunit Phospholipid-transporting ATPase IG The adenine nucleotide exchange factor of DnaK (Hsp70)-type ATPases A multifunctional protein that acts as a major Ca(2+)-binding protein and has a role in transcription regulation A member of the vesicle-associated membrane protein/synaptobrevin family An auxiliary beta subunit which slows activation kinetics, leading to greater calcium sensitivity

VAMP2 KCNMB4

O O O O O O O O O O O O O O

Important mRNA targets for further study are marked in bold. The explanations of gene function are summarized from the NCBI database.

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

9

Fig. 6. The effect of inhibition of protein synthesis in reticulocytes. (A) Blood samples with high reticulocyte numbers were cleared of leukocytes and incubated in methionine-free and control media to evaluate whether new protein synthesis is essential for reticulocytes. After 24 h, the remaining viable cells were observed by phase contrast microscopy. Dead cells are marked with arrows. Inset: representative cells stained with a vital RNA stain. Scale bars, 50 μm. (B, C) The loss of viable cells was similar to the loss of reticulocytes. (D) Overall viability decreased significantly in methionine-free conditions. *, P b 0.05.

failed to detect significant differences of individual mRNAs. This could be due to the low level of each mRNA. Conclusion Our present studies suggest that the small amount of remnant RNA in reticulocytes is not just inactive sub-cellular debris but must be translated into proteins for cells to survive. The RNA is diffusely distributed in the reticulocyte cytoplasm, and the up-regulated RNAs are primarily related to signal transduction, metabolism, and transport. Enzymatic destruction of the RNA blocks the maturation of reticulocytes and interferes with the maintenance of cell membrane and mitochondrial integrity as well. Additional studies to more exactly characterize the relevant signals would help to elucidate these pathways. Also, a clearer understanding of the role of residual RNA may lead to improvements in in vitro cultures of erythroid cells for transfusion purposes. Conflict of interest The authors report no conflict of interest. Acknowledgments This work was supported by the Medical Research Center (grant 2010-0029508, 2011-0028271) funded by the National Research Foundation (NRF) of the Ministry of Science, ICT, and Future Planning, Republic of Korea.

References [1] C. Geminard, A. de Gassart, M. Vidal, Reticulocyte maturation: mitoptosis and exosome release, Biocell 26 (2002) 205–215. [2] J. Liu, X. Guo, N. Mohandas, J.A. Chasis, X. An, Membrane remodeling during reticulocyte maturation, Blood 115 (2010) 2021–2027. [3] R.E. Waugh, Y.S. Huang, B.J. Arif, R. Bauserman, J. Palis, Development of membrane mechanical function during terminal stages of primitive erythropoiesis in mice, Exp. Hematol. 41 (2013) 398–408. [4] D.C. Wraith, C.J. Chesterton, Cell-surface remodelling during mammalian erythropoiesis, Biochem. J. 208 (1982) 239–242. [5] K. van Leyen, R.M. Duvoisin, H. Engelhardt, M. Wiedmann, A function for lipoxygenase in programmed organelle degradation, Nature 395 (1998) 392–395. [6] P.A. Marks, E.R. Burka, D. Schlessinger, Protein synthesis in erythroid cells, I. Reticulocyte ribosomes active in stimulating amino acid incorporation, Proc. Natl. Acad. Sci. U. S. A. 48 (1962) 2163–2171. [7] S.A. Hulea, H.R. Arnstein, Intracellular distribution of ribonuclease activity during erythroid cell development, Biochim. Biophys. Acta 476 (1977) 131–148. [8] M. Brigotti, P.G. Petronini, D. Carnicelli, et al., Effects of osmolarity, ions and compatible osmolytes on cell-free protein synthesis, Biochem. J. 369 (2003) 369–374. [9] E.R. Burka, RNase activity in erythroid cell lysates, J. Clin. Invest. 48 (1969) 1724–1732. [10] K. Brasch, G.H. Adams, J.M. Neelin, The morphology of erythroid cells separated by density gradient centrifugation through Ficoll, Cell Tissue Res. 176 (1977) 373–387. [11] B. Russell, R. Suwanarusk, C. Borlon, et al., A reliable ex vivo invasion assay of human reticulocytes by Plasmodium vivax, Blood 118 (2011) 2004–2011. [12] C. Brugnara, M.J. Hipp, P.J. Irving, et al., Automated reticulocyte counting and measurement of reticulocyte cellular indices. Evaluation of the Miles H*3 blood analyzer, Am. J. Clin. Pathol. 102 (1994) 623–632. [13] C.H. Joiner, R.S. Franco, M. Jiang, et al., Increased cation permeability in mutant mouse red blood cells with defective membrane skeletons, Blood 86 (1995) 4307–4314. [14] H.S. Choi, E.M. Lee, H.O. Kim, M.I. Park, E.J. Baek, Autonomous control of terminal erythropoiesis via physical interactions among erythroid cells, Stem Cell Res. 10 (2013) 442–453.

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

10

E. Lee et al. / Blood Cells, Molecules and Diseases xxx (2014) xxx–xxx

[15] E.J. Baek, H.S. Kim, J.H. Kim, N.J. Kim, H.O. Kim, Stroma-free mass production of clinical-grade red blood cells (RBCs) by using poloxamer 188 as an RBC survival enhancer, Transfusion (Paris) 49 (2009) 2285–2295. [16] E.J. Baek, J. You, M.S. Kim, S.Y. LEE, S.J. Cho, E. Kim, H.O. Kim, Enhanced production of red blood cells in suspension by electrostatic interactions with culture plates, Tissue Eng. 16 (2010) 1325–1334. [17] E.R. Burka, Characteristics of RNA degradation in the erythroid cell, J. Clin. Invest. 48 (1969) 1266–1272. [18] D. Nadano, T. Yasuda, H. Takeshita, K. Kishi, Ribonuclease inhibitors in human blood: comparative studies on the inhibitors detected in erythrocytes, platelets, mononuclear leukocytes and granulocytes, Int. J. Biochem. Cell Biol. 27 (1995) 971–979. [19] E.R. Burka, Hemin: An Inhibitor of Erythroid Cell Ribonuclease, Science 162, 1968.

[20] M.J. Koury, S.T. Koury, P. Kopsombut, M.C. Bondurant, In vitro maturation of nascent reticulocytes to erythrocytes, Blood 105 (2005) 2168–2174. [21] T.Y. Chao, R.T. Raines, Mechanism of ribonuclease A endocytosis: analogies to cell-penetrating peptides, Biochemistry (Mosc) 50 (2011) 8374–8382. [22] E.R. Burka, Protein synthesis by membrane-bound reticulocyte ribosomes, Ann. N. Y. Acad. Sci. 241 (1974) 191–203. [23] S.E. Zweig, K.T. Tokuyasu, S.J. Singer, Member-associated changes during erythropoiesis. On the mechanism of maturation of reticulocytes to erythrocytes, J. Supramol. Struct. Cell. Biochem. 17 (1981) 163–181. [24] R.E. Griffiths, S. Kupzig, N. Cogan, et al., The ins and outs of human reticulocyte maturation: autophagy and the endosome/exosome pathway, Autophagy 8 (2012) 7.

Please cite this article as: E. Lee, et al., Blood Cells Mol. Diseases (2014), http://dx.doi.org/10.1016/j.bcmd.2014.02.009

The RNA in reticulocytes is not just debris: it is necessary for the final stages of erythrocyte formation.

Reticulocytes contain both RNA and micro-organelles and represent the last stage of erythropoiesis before full maturation to red blood cells (RBCs). E...
2MB Sizes 0 Downloads 3 Views