Blood and Marrow Transplant Program, Chonnam National University Medical School, Kwangju, Korea ... ing hematopoietic stem cell transplantation (HSCT).
Bone Marrow Transplantation, (1999) 24, 411–415 1999 Stockton Press All rights reserved 0268–3369/99 $15.00 http://www.stockton-press.co.uk/bmt
Telomere length changes in patients undergoing hematopoietic stem cell transplantation JJ Lee, H Kook, IJ Chung, HJ Kim, MR Park, CJ Kim, JA Nah and TJ Hwang Blood and Marrow Transplant Program, Chonnam National University Medical School, Kwangju, Korea
Summary: Telomere length indicates the replicative history of cells, serving as a molecular measure of the replicative potential remaining in cells. To investigate telomere length changes in hematopoietic stem cells, patients undergoing hematopoietic stem cell transplantation (HSCT) were evaluated. Fifteen patients after allogeneic bone marrow transplantation (allo-BMT group), seven patients after autologous peripheral blood stem cell transplantation (auto-PBSCT group), and 39 healthy controls were studied. Telomere length was measured in peripheral mononuclear cells by Southern blot hybridization. There was no significant difference between the allo-BMT and the auto-PBSCT groups. In the allo-BMT group, the mean telomere length of recipients was 2.01 kb shorter than that of their donors (P = 0.008), and was 1.59 kb shorter than that of agematched putative normal controls (P = 0.002). Telomere shortening in the allo-BMT group was equivalent to 41.4 years of aging in the donors, and to 52.4 years of aging in the normal controls. The mean telomere length in the auto-PBSCT group was 2.36 kb shorter than that of the age-matched putative controls (P = 0.043), which was equivalent to 61.5 years of aging in normal controls. The extent of telomere shortening in the allo-BMT group showed a trend to negative correlation with the number of mononuclear cells infused. These findings suggest that hematopoietic stem cells after HSCT lose telomere length and these shortened telomeres may result in a higher incidence of clonal disorders later in life. Keywords: telomere; hematopoietic stem cell transplantation
Telomeres, located at the ends of eukaryotic chromosomes, are composed of specific proteins/tandem repeat DNA, and are essential for the stability of chromosomes and genes.1,2 In humans, telomeric DNA contains 5 to 15 kb of TTAGGG repeats which cap the chromosome ends. Telomeric DNA decreases by 50–100 bp with each somatic cell division, and progressively shortens with aging both in vitro Correspondence: Dr H-J Kim, Department of Internal Medicine, Chonnam National University Medical School, 8 Hak-Dong, Dong-Ku, Kwangju, 501–757, South Korea Received 4 February 1999; accepted 29 March 1999
and in vivo.3–9 Likewise, the mean telomere length of blood cells including both early progenitors and mature cells decreases with age.6–11 Therefore, telomere length indicates the replicative history of cells,8 serving as a molecular measure of the remaining replicative potential.12,13 As stem cells at the time of transplantation and during the engraftment undergo increased replicative proliferation, accelerated telomere shortening in patients after hematopoietic stem cell transplantation (HSCT) should ensue.14–17 Reductions in telomere length of reconstituting hematopoietic stem cells may cause chromosomal instability as well as replicative senescence,12,13,17 leading to the increased frequency of age-related hematologic disorders later in life.14,15,18–20 In this study, we investigated telomere length changes in patients undergoing HSCT and attempted to correlate several clinical factors with the extent of telomere shortening (⌬TEL). Materials and methods Patients Twenty-two patients who underwent HSCT from February 1995 to June 1998 at Chonnam University Hospital were included in this study. Among them, 15 patients underwent allogeneic bone marrow transplantation (allo-BMT group) (Table 1), and seven patients autologous peripheral blood stem cell transplantation (auto-PBSCT group) (Table 2). Their ages ranged from 7 to 45 years (mean 24.6) in the allo-BMT group and from 5 to 56 years (mean 30.7) in the auto-PBSCT group. In the allo-BMT group, engraftment was confirmed by cytogenetic analysis, FISH for sex mismatched, the RBC phenotyping, and/or variable number of tandem repeats by PCR amplification. All patients studied were engrafted and remained in CR at the time of study. Normal controls Blood samples were collected from 39 normal individuals ranging from newborn to 72 years of age. Fifteen controls were healthy donors for allogeneic bone marrow transplants. Cell separation and DNA extraction All peripheral blood samples from patients and normal controls were collected after obtaining informed consent, and
Telomere length changes after HSCT JJ Lee et al
412
Table 1 Case No.
Patient characteristics and telomere changes in allo-BMT group Recipient
Disease status Conditioning at HSCT regimen
Age Sex (years) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16 20 36 26 31 45 28 25 40 36 21 17 11 7 10
F M M F F M F M M M M M F M F
Donor
Period after HSCT (months)
Age Sex (years) AML/CR1 SAA AML/CR1 SAA SAA AML/CR1 AML/CR1 CML/CP AML/CR1 ALL/CR1 SAA SAA AML/CR1 WAS SAA
BU/CY CY/ATG/PCZ TBI/CY CY/ATG/PCZ CY/ATG/PCZ TBI/CY BU/CY BU/CY BU/CY TBI/CY CY/ATG/PCZ CY/ATG/PCZ BU/CY BU/CY/ATG CY/ATG/PCZ
Mean 24.6 ⫾ 11.5 ⫾ s.d.
22 22 34 22 29 36 37 30 38 32 20 17 6 5 12
F F M M F F F F F M F M F M F
24.7 29.0 13.7 27.6 12.8 1.6 35.7 20.3 11.8 24.9 15.2 19.5 24.7 42.8 25.5
24.1 ⫾ 10.8
21.8 ⫾ 10.2
Mean telomere length (kb) Recipient
Donor
9.25 7.60 6.20 11.40 6.80 6.80 7.12 10.0 8.08 6.07 9.41 7.49 9.08 10.59 9.09
9.40 10.01 9.08 13.25 9.20 8.76 10.74 12.07 9.95 8.25 10.51 8.16 10.84 14.16 10.77
8.3 ⫾ 1.64 10.34 ⫾ 1.74
⌬TEL (kb)
Infused MNCs (⫻ 108/kg)
0.15 2.41 2.88 1.85 2.40 1.96 3.62 2.07 1.87 2.18 1.10 0.67 1.76 3.57 1.68
2.73 4.30 5.97 3.30 2.24 2.46 3.50 1.85 2.84 3.52 4.32 3.37 2.73 3.76 5.1
2.01 ⫾ 0.94 3.82 ⫾ 1.61
MNC = mononuclear cell; SAA = severe aplastic anemia; WAS = Wiskott–Aldrich syndrome; CP = chronic phase; CRI = first complete remission. Conditioning regimens for transplantation were: BU/CY = busulphan 4 mg/kg ⫻ 4 days, cyclophosphamide 60 mg/kg ⫻ 2 days; CY/ATG/PCZ = cyclophosphamide 50 mg/kg ⫻ 4 days, antithymocyte globulin 30 mg/kg ⫻ 3 days, procarbazine 12.5 mg/day ⫻ 3 days; TBI/CY = 12 Gy in six fractions on 4 consecutive days, cyclophosphamide 60 mg/kg ⫻ 2 days.
Table 2 Case No.
Patient characteristics and telomere changes in auto-PBSCT group Age (years)
Sex
Disease status at HSCT
Conditioning regimen
Period after HSCT (months)
Mean telomere length (kb)
CR
A1 A2 A3 A4
42 5 6 56
M M M F
BEAM BEAM Ara-C/BU/VP-16 CVP
3.4 28.7 11.5 40.8
7.20 8.98 8.25 6.31
Yes Yes Yes Yes
A5 A6 A7
10 41 55
F F M
HD (relapsed) NHL (relapsed) AML (1CR) Small cell lung cancer AML (1CR) NHL (relapsed) NHL (relapsed)
Ara-C/BU/VP-16 BEAM BEAM
14.8 24.7 20.1
7.10 8.44 4.96
Yes Yes Yes
20.6 ⫾ 12.3
7.32 ⫾ 1.39
Mean ⫾ s.d.
30.7 ⫾ 22.9
HD = Hodgkin’s disease; NHL = non-Hodgkin’s lymphoma. Conditioning regimens for transplantation were: BEAM = BCNU 300 mg/m2 ⫻ 1 day, VP-16 200 mg/m2 ⫻ 4 days, cytosine arabinoside 200 mg/m2 ⫻ 4 days, melphalan 140 mg/m2 ⫻ 1 day; Ara-C/BU/VP-16 = cytosine arabinoside 100 mg/m2 ⫻ 6 days and 3.0 g/m2 bid ⫻ 2 days, busulphan 4 mg/kg ⫻ 4 days, VP-16 500 mg/m2 ⫻ 2 days; CVP = cyclophosphamide 1.75 g/m2 ⫻ 3 days, cisplatin 60 mg/m2 ⫻ 3 days, VP-16 400 mg/m2 ⫻ 3 days.
stored at ⫺80°C until DNA extraction. Mononuclear cells (MNCs) from peripheral blood were isolated by Ficoll– Hypaque density gradient centrifugation. High molecular weight DNA was extracted from 1 to 3 ⫻ 107 MNCs using 1 ml of DNAZOL reagent (GIBCO-BRL, NY, USA). The concentration of genomic DNA was measured at 260 nm by spectrophotometry. The integrity of extracted DNA was confirmed by 1% agarose gel electrophoresis. Mean telomere length measurement Four g of extracted DNA were completely digested with 40 U of RsaI for 12 h at 37°C. Electrophoresis of digested genomic DNA was performed in 0.6% agarose gels in 1 ⫻ TAE buffer (0.04 mol Tri-acetate, 0.001 mol EDTA, pH
8.0) at 40 V. After electrophoresis, gels were depurinated in 0.25 mol HCl for 30 min, and denatured in 0.4 N NaOH for 30 min. DNA was transferred on to a nylon membrane (Hybond N⫹; Amersham International plc, Amersham, UK) by capillary transfer method for Southern blotting. The filters were prehybridized in a hybridization buffer (TeloQuant; PharMingen, San Diego, CA, USA) for 1 h at 65°C and then hybridized with a biotinylated telomere probe (TTAGGG)4 (TeloQuant) in hybridization buffer overnight at 65°C. The filters were washed twice in 2 ⫻ sodium saline citrate (SSC)/0.1% sodium dodecyl sulfate (SDS) for 5 min at room temperature and then washed twice in 0.2 ⫻ SSC/0.1% SDS for 15 min at 42°C. The filters were blocked for 1 h at room temperature by the blocking buffer (TeloQuant) and were shaken for 1 h at
Telomere length changes after HSCT JJ Lee et al
Statistical analysis The Mann–Whitney U tests were used in the comparison of telomere length between groups. Linear regression analyses were performed to assess correlation between ⌬TEL and either donor age, elapsed time after the HSCT, or the number of MNCs infused. Results Mean telomere length in healthy normal controls The mean (⫾ s.d.) telomere length in peripheral blood MNCs from normal controls was 9.68 ⫾ 2.08 kb (range 5.65–14.40 kb) (Figure 1). There was progressive shortening of telomere length with age. The telomere length of normal controls was plotted by the following equation: T = 10.86 ⫺ 0.0384 ⫻ A, in which T = telomere length in kb and A = age in years (Figure 2).
413 16
T =10.86417 – 0.03843 x A (r = 0.383) 14
Telomere length (kb)
room temperature in a working solution of streptavidinehorseradish peroxide (⬍25 ng/ml) diluted in the blocking buffer. The filters were washed four times in 0.1% Tween20 (Boehringer Mannheim GmbH, Mannheim, Germany) in PBS for 10 min. After mixing with equal volumes of stable peroxide and luminol/enhancer (TeloQuant), the hybridized probe was shaken in this mixing solution for 5 min and was detected by the chemiluminescence method according to the manufacturer’s recommendations (TeloQuant). The filters were exposed to X-ray films (TMAT; Kodak, USA) for 30 s to 1 min. The telomere lengths were assessed quantitatively by densitometric analysis of autoradiographs using the transmitter scanning videodensitometer (Zenith Video Densitometer; Biomed, Fullerton, CA, USA). The mean telemere length in each sample was then identified as the peak intensity of the telomere length in kb by densitometry.
12 10 8 6 4
0
20
40
60
80
Age (years) Figure 2 Telomere length of healthy normal controls plotted against age in years. The regression line is shown as a solid line: T = 10.86417 ⫺ 0.03843 ⫻ A. P = 0.016. T = telomere length (kb); A = age (years).
(Figure 3). There was no significant difference between the two recipient groups (P = 0.217). To evaluate the effect of HSCT, the telomere length in each recipient group was compared to their donor or agematched controls. In allo-BMT group (Table 1), the mean telomere length of recipients was 2.01 kb shorter than that of their respective donors (10.34 kb) (P = 0.006), and was 1.59 kb shorter than that of age-matched putative normal controls (9.92 kb) (P = 0.003) (Figure 3). The mean telomere length of auto-PBSCT group (Table 2) was 2.36 kb shorter than that of age-matched putative normal controls (9.68 kb) (P = 0.004) (Figure 3). By applying linear regression to the normal control group, the telomere shortening of the allo-BMT group was equivalent to 41.4 years of aging in the donors and 52.4 years of aging in the age-matched putative normal controls. Likewise, the telomere shortening of auto-PBSCT group
Telomere shortening in patients who underwent HSCT D
11
N 10
Telomere length (kb)
The mean (⫾ s.d.) telomere length in the patients following HSCT was 8.01 ⫾ 1.60 kb (range 4.96–11.40), and was significantly shorter than that of normal controls (Figure 1). The mean (⫾ s.d.) telomere length in the allo-BMT recipients was 8.33 ⫾ 1.64 kb (range 6.07–11.40) and the autoPBSCT recipients 7.32 ⫾ 1.39 kb (range 4.96–8.98)
10.34 kb
N
9.92 kb 9.68 kb
9
P = 0.003
P = 0.006 P = 0.004
8.33 kb 8
Allo-BMT (n = 15)
7.32 kb
7
Auto-PBSCT Figure 1 A representative Southern blot analysis of telomere length in peripheral MNCs obtained from healthy normal controls (N), allo-BMT recipient (R)–donor (D) pairs, and auto-PBSCT patients (A). The vertical axis shows the telomere length (kb). The lanes of healthy normal controls denote individuals of different age: N1, cord blood; N2, 3 years; N3, 8 years; N4, 19 years; N5, 35 years; N6, 55 years; N7, 70 years.
(n = 7) 6
Figure 3 Comparison of mean telomere length between the HSCT group and control groups. D = donors of Allo-BMT group; N = putative normal controls.
Telomere length changes after HSCT JJ Lee et al
was equivalent to 61.5 year aging in the age-matched putative normal controls. In the allo-BMT group, we compared ⌬TEL with some clinical factors relevant to HSCT. ⌬TEL in the allo-BMT group ranged from 0.15 to 3.62 kb (mean 2.01 kb). ⌬TEL in the allo-BMT group shows a trend towards negative correlation with the number of infused MNCs (per kilogram of recipient body weight) at the time of HSCT (r = ⫺ 0.331; P = 0.229) (Figure 4). However, there was no correlation between ⌬TEL and donor age (r = 0.181; P = 0.519) or between ⌬TEL and time elapsed after the HSCT (r = 0.174; P = 0.534). Discussion This study showed that significant telomere shortening took place in reconstituting hematopoietic stem cells in patients undergoing HSCT, either allo-BMT or auto-PBSCT, and that the extent of telomere shortening had a trend towards negative correlation with the number of infused MNCs in the allo-BMT group but showed no correlation with the elapsed time after the HSCT or donor age. These results were consistent with recent reports from other investigators.14–16 Notaro et al14 reported similar results from 11 recipient–donor pairs and suggested that the loss of telomere might occur as a consequence of the in vivo replicative stress to normal hematopoietic stem cells. However, there was variation in telomere length among healthy normal controls. Inherited telomeric heterogeneities among individuals might contribute to the variation.21 Telomeres, the nucleoprotein complex at the ends of eukaryotic chromosomes, consist of tandem arrays of TTAGGG repeats bound to specific proteins and have the functions of stabilization of the chromosome ends and protection from enzymatic end-degradation.1,2,5 Telomere length is maintained by a balance between processes that lengthen and shorten telomeres.22 Because telomeric loss is caused by an end-replication problem during progressive 4
3
∆TEL
414
2
1
0
2
4
6
8
10
Infused cells x 108/kg Figure 4 Correlation of the change of telomere length (⌬TEL) with infused MNC dose in allo-BMT group (r = 0.331). The straight line shows the regression: ⌬TEL = 2.8627 ⫺ 0.1875 ⫻ D. ⌬TEL = difference in telomere length (kb); D = infused cell dose (⫻ 108/kg).
cell divisions,23 the telomere shortening is considered as a molecular clock that counts the number of cell divisions and determines the onset of cellular senescence.8,12 Thus, the length of telomeres is an indicator of the replicative history of the cells and may be the molecular counterpart of the replicative potential remaining in cells.8,12,13 In normal somatic cells, progressive telomere shortening is observed at a rate of 50–100 bp per cell division, eventually leading to shortened telomeres, and then to a stage of growth arrest. Critically shortened telomeres, when cells are no longer able to protect the ends of the chromosomes, cause chromosome fusion and massive genomic instability, and may contribute to age-related clonal disorders.15,17,20 Interestingly, the telomere length in primitive hematopoietic stem cells decreases despite telomerase activity.24 Akiyama et al16 suggested that telomerase activity in stem cells after HSCT might be too weak to overcome the shortening of telomere length. However, they did not measure telomerase activity in stem cells after HSCT. Thus, further study is needed to elucidate the relationship between telomere length and telomerase activity following HSCT. In addition, memory T cells after BMT have shorter telomeres than naive T cells, and a shift from naive T cells to memory T cells could potentially also account for this shortening.25 One of the aims of this study was to determine whether ⌬TEL observed in HSCT patients correlated with the number of infused MNCs at the time of HSCT. Notaro et al14 reported that ⌬TEL has an inverse correlation with the number of nucleated cells infused, and suggested that the fewer the stem cells transplanted to the recipient, the more cell divisions were needed for reconstitution of hematopoiesis. Thus, the use of a larger number of stem cells, such as megadose transplantation, might result in less telomere shortening and might reduce the incidence of clonal disorders such as myelodysplastic syndrome or cancer, later in life after HSCT.14,15,17,26–29 In this study, ⌬TEL showed a tendency towards negative correlation with the number of MNCs infused. However, a further study of larger numbers of patients is needed to draw a definitive conclusion. Wynn et al15 showed that the average reduction of telomere length after HSCT was equivalent to roughly 15 years’ aging, assuming a constant rate of loss in the normal population. The present study showed greater extent of telomere shortening, which was three to four times higher than reported by Wynn et al.15 The reason for this discrepancy is not clear but the possible differences in the infused stem cell numbers, small numbers of normal controls as a reference, or other factors should be contemplated. No correlation between ⌬TEL and the elapsed time after HSCT was found in this study, in agreement with the previous results.14 This finding suggests that telomere length after HSCT shortens rapidly during the engraftment period and the rates of telomere shortening after engraftment might be similar to the aging change of normal individuals. In addition, there was no correlation between ⌬TEL and donor age. A recent report by Frenck et al30 showed that telomeric repeats were lost rapidly in children, followed by an apparent plateau between age 4 years and young adulthood, and by gradual attrition later in life. Because the donor’s ages of allo-BMT patients in this study were chil-
Telomere length changes after HSCT JJ Lee et al
dren and adults ranging from 5 years to 38 years, the telomere length change for the donors might be minimal. In conclusion, our findings suggest that hematopoietic stem cells after HSCT lose telomere length and that their telomere shortening may be secondary to increased replicative proliferation. The monitoring of telomere length after HSCT may be needed as shortened telomere may result in a higher incidence of clonal disorders later in life. Acknowledgements The authors greatly appreciate Dr TF Tisdale, NHLBI, NIH, USA for advice and critical review of the manuscript. This work was supported in part by a grant from the Research Institute of Medical Sciences, Chonnam National University.
References 1 Blackburn EH. Structure and function of telomeres. Nature 1991; 350: 569–573. 2 Zakian VA. Telomeres: beginning to understand the end. Science 1995; 270: 1601–1607. 3 Moyzis RK, Buckingham JM, Cram LS et al. A highly conserved repetitive DNA sequence, (TTAGGG)n, present at the telomeres of human chromosomes. Proc Natl Acad Sci USA 1998; 85: 6622–6626. 4 Rhyu MS. Telomeres, telomerase, and immortality. J Natl Cancer Inst 1995; 87: 884–894. 5 Blackburn EH. Telomeres: no end in sight. Cell 1994; 77: 621–623. 6 Hastie ND, Dempster M, Dunlop MG et al. Telomere reduction in human colorectal carcinoma and with ageing. Nature 1990; 346: 866–868. 7 Vaziri H, Dragowska W, Allsopp RC et al. Evidence for a mitotic clock in human hematopoietic stem cells: loss of telomeric DNA with age. Proc Natl Acad Sci USA 1994; 91: 9857–9860. 8 Harley CB. Telomere loss: mitotic clock or genetic time bomb? Mutat Res 1991; 256: 271–282. 9 Engelhardt M, Kumar R, Albanell J et al. Telomerase regulation, cell cycle, and telomere stability in primitive hematopoietic cells. Blood 1997; 90: 182–193. 10 Vaziri H, Schachter F, Uchida I et al. Loss of telomeric DNA during aging of normal and trisomy 21 human lymphocytes. Am J Hum Genet 1993; 52: 661–667. 11 Slagboom PE, Droog S, Boomsma DI. Genetic determination of telomere size in humans: a twin study of three age groups. Am J Hum Genet 1994; 55: 876–882. 12 Shay JW, Werbin H, Wright WE. Telomeres and telomerase in human leukemias. Leukemia 1996; 10: 1255–1261.
13 Allsopp RC, Vaziri H, Patterson C et al. Telomere length predicts replicative capacity of human fibroblasts. Proc Natl Acad Sci USA 1992; 89: 10114–10118. 14 Notaro R, Cimmino A, Tabarini D et al. In vivo telomere dynamics of human hematopoietic stem cells. Proc Natl Acad Sci USA 1997; 94: 13782–13785. 15 Wynn RF, Cross MA, Hatton C et al. Accelerated telomere shortening in young recipients of allogeneic bone-marrow transplants. Lancet 1998; 351: 178–181. 16 Akiyama M, Hoshi Y, Sakurai S et al. Changes of telomere length in children after hematopoietic stem cell transplantation. Bone Marrow Transplant 1998; 21: 167–171. 17 Shay JW. Accelerated telomere shortening in bone marrow recipients. Lancet 1998; 351: 153–154. 18 Schulz VP, Zakian VA, Ogburn CE et al. Accelerated loss of telomeric repeats may not explain accelerated replicative decline of Werner syndrome cells. Hum Genet 1996; 97: 750–754. 19 van Steensel B, de Lange T. Control of telomere length by the human telomeric protein TRF1. Nature 1997; 385: 740–743. 20 Curtis RE, Rowlings PA, Deeg HJ et al. Solid cancers after bone marrow transplantation. New Engl J Med 1997; 336: 897–904. 21 Iwama H, Ohyashiki K, Ohyashiki JH et al. Telomeric length and telomerase activity vary with age in peripheral blood cells obtained from normal individuals. Hum Genet 1998; 102: 397–402. 22 de Lange T. Length control of human telomeres. Cancer J Sci Am 1998; 4: (Suppl. 1): S22–S25. 23 Watson JD. Origin of concatemeric T7 DNA. Nat New Biol 1972; 239: 197–201. 24 Norrback KF, Roos G. Telomeres and telomerase in normal and malignant haematopoietic cells. Eur J Cancer 1997; 33: 774–780. 25 Rufer N, Dumont-Girard F, Barbey C et al. Analysis of telomere length dynamics in human lymphocyte subpopulations after allogeneic bone marrow transplantation. Blood 1998; 92 (Suppl. 1): 26 (Abstr. 99). 26 Witherspoon RP, Deeg HJ, Storb R. Secondary malignancies after marrow transplantation for leukemia or aplastic anemia. Transplantation 1994; 57: 1413–1418. 27 Deeg HJ, Socie G. Malignancies after hematopoietic stem cell transplantation: many questions, some answers. Blood 1998; 91: 1833–1844. 28 Lowsky R, Lipton J, Fyles G et al. Secondary malignancies after bone marrow transplantation in adults. J Clin Oncol 1994; 12: 2187–2192. 29 Bhatia S, Ramsay NK, Steinbuch M et al. Malignant neoplasms following bone marrow transplantation. Blood 1996; 87: 3633–3639. 30 Frenck RW Jr, Blackburn EH, Shannon KM. The rate of telomere sequence loss in human leukocytes varies with age. Proc Natl Acad Sci USA 1998; 95: 5607–5610.
415