Butyrate-stable monosaccharide derivatives induce maturationand apoptosis in human acute myeloid leukaemia cells

V. SANTINI, B. SCAP PINI, A. GOZ ZINI, A. GROSSI, P. VILLA,* G. RO NCO,* O. DOUILLET,* P. POUILLART,P. A. BERNABEI AND P. ROS SI FERRINI Department of Haematology, University of Florence, Italy,and *Laboratoire de Chimie Organique et Cinetique, University of Picardie, Amiens, France

Received 7 November 1997; accepted for publication 4 March 1998

Summary. The rapid degradation and subsequent lack ofefficacy of n-butyric acid in vivo has been improved by thesynthesis of monosaccharide stable pro-drugs of butyricacid. We studied the effects of D1 (O-n-butanoyl-2,3-O-isopropylidene-alpha-D-mannofuranoside), G1 (1-O-n-buta-noyl-D,L-xylitol), and F1 (1-O-n-butanoyl 2,3-O-isopropyli-dene-D,L-xylitol) on the maturation and proliferation of AMLcell lines HL 60 and FLG 29.1 and of purified blast cells from10 cases of de novo acute myeloid leukaemia (AML). AML cellmaturation was measured by surface antigen expression,morphology and cytochemistry. Toxicology in mice was alsoevaluated (DL50 1000 mg/kg). In HL 60 cells G1 and D1increased the expression of CD15 and CD11a (presenting

62% of promyelo-metamyelocytes), and in 7/10 cases ofprimary AMLs that of CD11a, CD11b, CD15, and myeloper-oxidase. D1, G1 and F1 induced a dose-dependent inhibitionof tritiated thymidine uptake. Apoptosis (evaluated by flowcytometry and agarose gel electrophoresis) was induced inAML blasts by D1 and F1 (79% and 94% respectively for HL60 cells) and, with less effect, by G1 (27%). The persistence ofmaturative and apoptotic activity in these new pro-drugs ofbutyric acid, hydrolysed only inside the tumour cell, suggestsa possible use in differentiation therapy of myelodysplasticsyndromes and AMLs.

Keywords: butyrate, myeloid leukaemia, maturation, apoptosis.

Acute myeloid leukaemia results from the accumulation ofimmature myeloid precursors unable to undergo bothmaturation and apoptosis. As traditional chemotherapydoes not seem to be effective in eradicating leukaemia, therehas been a renewed effort in searching for drugs endowedwith differentiative as well as apoptotic potentials and alsowith a lower general toxicity. The maturative effect of butyricacid has been known and studied for two decades (Leder &Leder, 1975). Butyric acid has been shown to induce geneexpression in several cell types, acting most probably viahistone hyper-acetylation (Riggs et al, 1977; Candido et al,1978; Boffa et al, 1981). Notwithstanding such long-knownobservations and numerous in vitro studies, sodium butyrateor other butyrate derivatives have been rarely employed inclinics. This infrequent application is not due to side-effectsor general toxicity, but to the extremely short half life ofbutyrate salt derivatives, which impairs any long-lastingeffect in vivo and compels a continuous intravenousadministration regimen to maintain acceptable plasmalevels. Most recently, preliminary results of a trial havebeen reported in which sodium phenylbutyrate was given to

patients affected by myelodysplastic syndromes (Gore et al,1996). Considering these experiences, we have analysed thein vitro anti-leukaemic effects of three newly synthesizedstable monosaccharide butyrate derivatives (Pouillart et al,1998). In these compounds a mannose molecule is bound asester to one to five butyrate moieties, conferring pharmaco-logical stability. We have now studied the possible efficacy ofthe three butyrate derivatives in inhibiting cell proliferationand in inducing programmed cell death of primary humanacute myeloid leukaemia cells and of human acuteleukaemia cell lines.

MATERIALS AND METHODS

Cell lines, patients and preparation of AML cells. The humanmyeloid leukaemia cell line HL 60 (Collins, 1977) and theosteoclastic cell line FLG 29.1, established from a humanmonoblastic leukaemia (Gattei et al, 1992), were maintainedin culture in RPMI 1640, supplemented with 10% fetal calfserum (FCS) and employed in experiments during theexponential growth phase.

Heparinized peripheral blood and bone marrow sampleswere obtained from 10 patients with AML, after informed

British Journal of Haematology, 1998, 101, 529–538

529q 1998 Blackwell Science Ltd

Correspondence: Dr V. Santini, Department of Haematology,Policlinico di Careggi, Viale Morgagni 85, 50134 Firenze, Italy.

consent had been obtained. FAB diagnosis (Bennett et al,1985) included three M2, five M4 and two M5. Tlymphocytes were removed after Ficoll-Isopaque centrifuga-tion by E-rosetting with sheep red blood cells and AML cells,then cryopreserved in dimethylsulphoxide (DMSO), aspreviously described (Santini et al, 1991). After thawingand washing AML blasts were depleted of adherent cellsfollowing incubation in serum-free medium (Salem et al,1988) in plastic culture flasks (250 ml, Greiner, Germany)and then cultured.

Chemicals. Toxicology in mice. We compared the effects ofthree stable butyrate monosaccharide esters (French patentFR 94/09348, PCT FR 95/00743) with those of equimolarconcentrations of sodium butyrate. These newly synthesizedmolecules are formed by a mannose or a xylitol bound asester to one to five butyrate moieties. Details of chemistry,synthesis and structure of these molecules are reportedelsewhere (Pouillart, 1998). Butyrate derivatives werefreshly prepared immediately before experiments and usedin cell cultures at the final concentrations of 0·1, 0·5 and1 mM. DL10 and DL50 were determined in pathogen-free,immunocompetent male or female Swiss mice, 19–20 d old,as follows. All mice were maintained under standardconditions without isolation. Non-substituted glucidic unitsD1, G1 and F1, together with the corresponding n-butyricderivatives, were dissolved in saline solution at roomtemperature immediately before use. Aliquots of 0·2 ml ofsolution with different concentrations of butyrates wereinjected intravenously (i.v.) as a bolus into groups of fivemice. Scoring of functional as well as behavioural effects wascarried out for a 24 h observation period in the entire animalpopulation. The lowest dose resulting in the death of either10% or 50% of the experimental population was defined asDL10 or DL50. Data were examined by a log-linear regressionanalysis.

Cell cultures. Cell lines were cultured in 250 ml plasticflasks (Greiner, Germany) in RPMI 1640, 10% FCS at 378C,5% CO2 and exposed to D1, G1 and F1 0·1, 0·5 and 1 mM for4 and 7 d. Culture conditions for AML blasts derived frompatients were alike, except that serum-free medium wasused, supplemented with IL-3 10 ng/ml, GM-CSF and G-CSF100 ng/ml.

Maturation. Prior to and after culture, cytospin prepara-tions of AML cells were stained with May-Grunwald-Giemsa(MGG). Morphology of cells was then examined by lightmicroscopy using ×100 lens and immersion oil. At least 200cells per slide were counted in duplicate and the percentageof blasts, promyelocytes, myelocytes, metamyelocytes, bandforms, granulocytes, monoblasts, monocytes and macro-phages were scored. Nucleated cell number and viability wasdetermined by counting with Turck solution and Tripan bluedye exclusion. Picnosis of nuclei, cytoplasmic and chromatincondensation were also scored as specific signs of apoptosis,accompanying maturation. At the same time, theappearance of membrane surface antigens specific ofgranulocytic and monocytic maturation was assessed bydirect immuno-fluorescence. Cytofluorimetric analysis wasperformed on a FACSscan (Becton Dickinson) by the use of apanel of FITC and PE conjugated monoclonal antibodies

(CD11a, CD11b, CD14, CD15, CD16 ) and myeloperoxidase(MPO).

Cell kinetics and apoptosis.. After incubation, AML cells,untreated and treated with butyrates, were washed thor-oughly, re-suspended in PBS/ethanol 1:3 vol/vol and keptovernight at 48C. After fixation, cells were incubated inPropidium iodide (PI) 50 mg/ml plus Nonidet 0·01%, RNAse62 mg/ml. DNA content was measured by a FACSscan flowcytometer (Becton Dickinson, San Jose, Calif.). Analysis ofthe data was accomplished by the use of CellFit program,applying MANL statistic program which quantifies theconsistence of the apoptotic peak ‘pre-G1’, present in theregion of channel 100.

DNA fragmentation. Cellular DNA was obtained from blastcells, cultured in the same conditions with or withoutbutyrates as for cell cycle determination (see above). DNAwas obtained by lysis of the nuclear membrane and then byphenol/chloroform extraction. Equal amounts of purifiedDNA were then loaded on to a 2% agarose gel andelectrophoresis was performed at low voltage (60 V) for5 h. The appearance of the characteristic DNA digestionfragments of 200 base pairs or multiple and the ladder-likeDNA smear were considered indicative of apoptosis.

Tritiated thymidine uptake. Blast cells were cultured asdetailed above, plated on to 96-micro-well plates at theconcentration of 1 × 105/ml in the presence of D1, G1 andF1 0·1, 0·5, 1 mM. Control cultures were carried out in theabsence of any drug. After 48 h of culture in theseconditions, cells were exposed for 18 h to tritiated thymidine([3H]TdR, 185 kBq/well, specific activity 74 GBq/mmol,Amersham (U.K.). Pulsed cells were harvested by anautomated cell harvester (Skatron, Norway) and incorpora-tion of [3H]TdR evaluated by beta scintillation counting.

Statistical analysis. Student’s t test (level of confidence95%; d.f.¼9) for paired data was applied to verify significantdifferences in the presence of mature cells and in theincorporation of tritiated thymidine between untreated andbutyrate-treated leukaemia cells.

RESULTS

Toxicology in miceAcute toxicity of butyrate derivatives and non-substitutedglucidic compound was evaluated. Non-substituted glucidicmoieties were devoid of toxic effects. DL10 for G1 and F1 was>1000 mg/kg and therefore DL50 was not determined. Onlythe mannose derivative D1 affected mice spontaneousactivity and severe behaviour disturbances were observed.DL10 for D1 was 900 mg/kg for male mice and 950 mg/kgfor female mice, with a DL50 of 1250 and 1000 mg/kg,respectively. Toxic death was observed in the D1 ester treatedgroup as caused by respiratory function defect.

MaturationMorphology. Exposure to butyrate ester derivatives

induced maturation in HL 60 cells, but not in FLG 29.1osteoblastic cell line. As clearly shown by the modificationsin HL 60 cellular morphology (Fig 1), all the molecules: D1(Fig 1b, c), G1 (Fig 1d, e) and F1 (Fig 1f) at 0·5 and 1 mM

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

530 V. Santini et al

531AML Cell Maturation and Apoptosis by Monosaccharide Butyrates

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

Fig 1. Morphological maturation of HL 60 cells treated with butyrate monosaccharide derivatives. (May-Grunwald-Giemsa staining). Cytospinswere prepared after 4 d of culture in the absence of butyrates (a), or in the presence of D1 0·5 and 1 mM (b, c); G1 0·5 and 1 mM (d, e); F1 0·5 mM

(g); sodium butyrate 1 mM (f ). Magnification ×100.

concentration, were able to promote granulocytic matura-tion and their maturative effect was more evident than thatof sodium butyrate (Fig 1g), used as comparative control. Theappearance of 16% of band forms and 28% metamyelocytesand myelocytes was documented after 4 d of culture in thepresence of D1 (Table I), whereas G1 was more effective after7 d. HL 60 cells showed 48% of mature granulocytic cells after4 d, but the prolongation of the exposure to the xylitolderivative of butyric acid induced the appearance of 74%mature granulocytic cells. The overall proportion of maturecells found in AML primary cultures, after 4 d of culture withD1 and G1 0·5 mM, was significantly higher than that ofcultures carried out without addition of butyrates (P<0·05)(Table I). Although with an evident heterogeneity, all AMLcases studied responded to butyrate derivatives, producingmaturing cells. No difference in terms of susceptibility tomature was noted among FAB subtypes, but this observationwas not significant because of the small cohort of casesstudied. The incubation in the presence of F1 provoked theformation of cytoplasmic and nuclear vacuolization, prob-ably as a sign of cellular toxicity, which sometimes impaireda complete morphological evaluation of maturation.

Phenotype. The expression of specific maturative antigenon cell surface was evaluated in HL 60 cells and in primaryAML cells exposed to monosaccharide butyrates andcompared to that of unstimulated samples (Table II). CD15expression was enhanced by G1 treatment in HL 60 cells, aswell as CD11a, although the latter was not significant. D1and G1 0·5 mM were very effective in determining asignificant increase in expression of CD11c (P<0·05) in 5/10 cases (nos. 4, 6, 7, 8 and 10). CD14 expression was lessaffected by the exposure to butyrates, but was still increasedin cases 5, 9 and 10. CD15 expression was significantly(P<0·05) enhanced in 3/10 cases (nos. 7, 8 and 9) byexposure to G1; cases 7 and 8 responded also to F1, D1 andsodium butyrate

ApoptosisAnalysis of DNA content. When DNA content was analysed

by FACSscan the appearance of the pre-G1, A0 apoptoticpeak was monitored. HL 60 cells were very sensitive toprogrammed cell death induced by monosaccharide buty-rates. Incubation with G1 produced 24% of apoptotic cellsafter 4 d, but exposure to D1 and F1 0·5 mM had induced

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

532 V. Santini et alTable I. Morphological maturation of acute myeloid leukaemia cells after exposure to butyrate monosaccharide esters.

HL60 1 2 3 4 5 6 7 8 9 10

NoneBlasts 73 85 68 90 80 96 78 100 95 100 85Mono 0 15 28 9 8 4 22 0 3 0 10Proþmeta 27 0 4 1 12 0 0 0 2 0 5N band 0 0 0 0 0 0 0 0 0 0 0

G1Blasts 50 50 10 61 29 35 32 48 67 67 60Mono 2 30 80 39 35 50 0 0 7 18 10Proþmeta 42 20 10 0 36 15 68 52 26 15 30N band 6 0 0 0 0 0 0 0 0 0 0

F1Blasts n n 16 85 18 35 45 73 71 60 32Mono n n 46 15 28 49 0 0 8 21 15Proþmeta n n 38 0 52 16 55 27 21 19 53N band n n 0 0 2 0 0 0 0 0 0

D1Blasts 56 50 22 50 20 35 47 50 65 67 25Mono 0 40 44 30 52 60 0 0 5 27 18Proþmeta 28 10 34 20 28 5 53 50 30 6 57N band 16 0 0 0 0 0 0 0 0 0 0

Na butyrateBlasts 48 47 40 45 42 43 54 51 63 65 45Mono 0 40 60 38 42 55 0 0 10 30 12Proþmeta 42 13 0 17 16 2 46 49 27 5 43N band 10 0 0 0 0 0 0 0 0 0 0

Purified AMl blasts were cultured for 4 or/and 7 d in the presence of G1, F1, D1 and sodium butyrate 0·5 mM, in serum-freemedium or 10% FCS (HL60). From recovered cells, cytospins were prepared and stained with May-Grunwald-Giemsa. Cells wereexamined by light microscopy, at least 100 cells were counted, and the percentage of blasts (immature cells to myelocytes),monocytes, promyelocytes and metamyelocytes, intermediate and mature granulocytic cells (N band) scored. n¼necrosis impairedevaluation of the sample.

533AML Cell Maturation and Apoptosis by Monosaccharide Butyrates

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

79% and 95% of the cells to undergo apoptosis within thattime (Fig 2). FLG 29.1 cells, although less sensitive than HL60 cells, were clearly committed towards apoptosis by F1 andmostly by D1 (61% of the cell population). The effects of allmonosaccharide derivatives were comparable to those ofsodium butyrate. Among all AML cases (data not shown),percentages of apoptotic cells ranged from 15% to 26% afterbutyrate treatment for 4 d, which were always significantwith respect to control (untreated) cells.

DNA fragmentation. DNA agarose gel electrophoresisdemonstrated in all AML cases and in both HL 60 and FLG29.1 cell lines clear DNA fragmentation in cells treated withG1, D1 and F1 (but also sodium butyrate) paralleling thedata obtained by FACSscan analysis. Fig 3 shows theelectrophoresis of DNA obtained from HL 60 cells culturedfor 4 d in the presence of stable monosaccharide butyratederivatives at the final concentration of 0·5 mM. The typicalpattern of DNA digestion, yielding fragments of 200 bp or

multiples of this, indicated that cell death was obtainedthrough apoptosis. In parallel, the presence of apoptotic cellswas always counter-checked and scored by morphologicalexamination (Fig 3).

Tritiated thymidine uptakeTritiated thymidine uptake by AML cell lines and primarycultures was inhibited by incubation with butyrateanalogues. The results obtained by the addition of butyrateanalogues at concentration of 0·5 mM to the cultures areshown in Figs 4A and 4B. D1, G1 and F1 were constantlyable to significantly decrease tritiated thymidine uptake in adose–response manner, and their effects were equal to ormore effective than those of equimolar sodium butyrate. HL60 cell proliferation (Fig 4A) was inhibited to 40% of controlby D1 after 7 d of culture. F1 was also very effective ininhibiting cell growth (25% of thymidine incorporation withrespect to controls), as expected by its toxic effects evaluated

Table II. Surface antigen modification in de novo acute myeloid leukaemia cells after exposure to butyratemonosaccharide esters.

1 2 3 4 5 6 7 8 9 10

NoneCD11a 0 1 0 0 0 0 0 12 0 20CD11c 35 46 51 36 22 43 3 10 15 32CD14 0 0 0 0 10 0 2 0 11 4CD15 26 10 69 35 nd 75 50 34 0 34MPO 2 43 nd nd nd 0 4 26 0 0

G1CD11a 0 8 0 0 50 25 12 23 0 35CD11c 21 46 57 53 0 60 54 25 27 55CD14 0 0 1 0 25 0 0 3 20 27CD15 4 14 32 40 nd 59 87 57 39 25MPO 0 78 nd nd nd 0 51 37 18 0

F1CD11a 0 3 0 0 43 30 19 25 0 33CD11c 18 45 55 47 1 58 34 25 30 45CD14 0 0 0 0 25 0 2 0 30 32CD15 13 3 35 36 nd 30 90 60 10 20MPO 0 52 nd nd nd 0 19 34 18 0

D1CD11a 0 7 0 0 42 22 13 27 2 40CD11c 23 45 37 63 0 59 44 22 28 43CD14 0 0 61 3 23 0 0 0 20 27CD15 9 18 0 47 nd 23 87 58 15 33MPO 1 79 nd nd nd 0 25 50 10 5

Na butyrateCD11a 0 9 0 0 43 24 15 30 4 48CD11c 23 47 33 5 0 58 49 28 29 45CD14 0 0 1 0 21 0 2 0 19 30CD15 7 16 63 34 nd 12 92 60 40 30MPO 0 76 nd nd nd 0 20 50 20 13

Purified AML blasts were cultured in the presence of G1, F1, D1 and sodium butyrate 0·5 M, in serum-free medium or10% FCS (HL60). Cells recovered from cultures were marked with FITC and PE conjugated antibodies recognizingmyeloid maturation surface antigens and myeloperoxidase (MPO). Samples were then analysed by FACSscan;percentages of positive cells were determined by subtracting the percentage of gated control cells stained with nospecifically with FITC and PE. nd: not determined.

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

534 V. Santini et al

Fig 2. Effect of butyrate monosaccharides on leukaemic cell kinetic and viability. After culture in the presence of monosaccharide butyrates,HL 60 cells, FLG 29.1 cells and primary AML cells were fixed and then stained with propidium iodide (PI); their DNA content was thencytofluorimetrically analysed. The percentage of events in the aneuploid A0 apoptotic peak (pre-G1 peak) is indicated in the graphs.

535AML Cell Maturation and Apoptosis by Monosaccharide Butyrates

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

in the morphological analysis. Moreover, G1 0·5 mM,although not sufficient to reduce tritiated thymidine uptakeof HL 60 during the first days of culture, inhibited upto 40%cell proliferation at day 7. After 7 d of culture, FLG 29.1 cells(Fig 4B) incubated in the presence of D1, F1 as well assodium butyrate, had a rate of tritiated thymidine incorpora-tion that was approximately 60% (57–63%) of that ofcontrol untreated cells. G1 was only able to reduce nucleo-tide incorporation to 95%. After 4 d of incubation withmonosaccharide butyrates, primary AML cell proliferationwas blocked heterogenously among cases, as could bepredicted by the usual variability in the biology of primaryhuman AMLs (data not shown)

DISCUSSION

The effects on leukaemic cell growth in vitro of three butyricmonosaccharide esters (D1, G1 and F1) and their acutetoxicity in mice were evaluated. The three molecules haddifferent activity in inducing maturation, although it wasalways promoted, to various extents. D1 (O-n-butanoyl-2,3-O-isopropylidene-alpha-D-mannofuranoside) seemed to be

extremely effective after 4 d of incubation, whereas G1 (1-O-n-butanoyl-D,L-xylitol) showed increasing maturative propertiesfor prolonged exposure times (day 7). F1 (1-O-n-butanoyl 2,3-O-isopropylidene-D,L-xylitol) was the only molecule thatinduced degenerative alterations in leukaemic cells. Thesedifferences in efficacy and in timing are to be attributed to thedivergent molecule conformation and strength of binding ofthe sugar moiety to the butyrate. Similarly to other butyratesugar derivatives (Pouillart et al, 1990), the serum concen-tration is significantly higher than that of butyrate salts,reaching zero after 72 h from infusion. Moreover, the diffusionphase is 1–2 h for butyrate monosaccharides and 1 min forbutyrate salts (Pouillart et al, 1992).

Together with maturation after G1, D1 and F1 derivativeincubation, we observed apoptosis in all the primary AMLcell cultures, and apoptosis was also provoked in the FLG29.1 cell line, in which no evident maturation was detected.Programmed cell death was paralleled by a significantinhibition of AML cell proliferation, similar to sodiumbutyrate, which was able to block specifically and reversethe cell cycle in G1 phase (van Wijk et al, 1981). It has beenreported recently that induction of apoptosis by sodium

Fig 3. (A) DNA fragmentation pattern of HL 60 cells. Incubation with monosaccharide butyrates provoked the endonucleosomic digestion ofDNA, resulting in the typical ladder-like image at agarose electrophoresis. Lane 1: DNA from untreated leukaemic cells; DNA of HL 60 cells after4 d of incubation in vitro with, lane 2: D1 0·5 mM; lane 3: after G1 0·5 mM; lane 4: after F1 0·5 mM; lane 5: after sodium butyrate 0.5 mM; lane 6:molecular weight marker II (Boeheringer Mannheim). (B) Apoptosis in HL 60 cells after incubation with butyrate monosaccharides.Representative photomicrographs of shrunk cells with picnotic nuclei and formation of apoptotic bodies in HL 60 cell culture after 4 d of exposureto G1 0·5 mM (2). Original magnification ×100; May-Grunwald-Giemsa staining.

butyrate in HL 60 cells was accompanied by phosphoryla-tion of proteins of molecular weight 37 and 97 kD. Theaddition of sodium vanadate, a phosphatase inhibitor, wasable to enhance the apoptotic effect of sodium butyrate(Chang & Yung, 1996). These observations give furthersupport to the notion that butyrates act specifically to induceapoptosis.

The contemporary occurrence of maturation and apopto-sis in leukaemic cells exposed to monosaccharide butyratederivatives may seem contradictory. In fact, the coupling ofthese events has been demonstrated in several other cellularmodels. Retinoic acid, when added to acute myeloidleukaemia cell line cultures was able to provoke bothterminal granulocytic differentiation and appearance ofapoptotic cells with decreased bcl2 expression (Benito et al,1995; Sanz et al, 1997).

Granulocyte colony-stimulating factor (G-CSF), the prin-cipal inducer of neutrophil maturation, has been shown topromote programmed cell death in myeloid cells bearingwild-type G-CSF-receptor, even in the presence of survivalfactors. In this case it was also demonstrated that the

carboxy-terminal cytoplasmic region of G-CSF-R wasresponsible for both programmed cell death and maturationsignals (Dong et al, 1996). Moreover, aberrant expression ofthe maturation-specific transcription factors TAL-1 and PU.1in different myeloid and erythroleukaemic cell lines(Condorelli, 1997; Tran Quang et al, 1997) disruptsregulatory mechanisms, promoting cellular proliferation,but at the same time blocking both maturation andapoptosis. These results evoke two possible explanations.Physiological induction of terminal granulocytic maturationis followed, with cell ‘ageing’, by apoptotic cell death (Savill etal, 1989; Martin et al, 1990). Such homeostatic mechanismmay take place also in vitro (Sanz et al, 1997) and thus inbutyrate-induced cell maturation. On the other hand,leukaemic cells may be induced to incomplete maturation,because of their intrinsic differentiative inability. Failure toachieve complete terminal maturation may switch on apremature cell death signal to eliminate a defective cellpopulation (Manfredini et al, 1993; Dong et al, 1996).

Sodium butyrate and other butyrate salts are known tohave a differentiative potential in vitro on acute leukaemia

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

536 V. Santini et al

Fig 4. Effects of butyrate monosaccharideson leukaemic cell growth. Tritiated thymidineuptake was evaluated in the HL 60 cell line (A)and in the FLG 29.1 cell line (B), in the absenceof any butyrate (control 100%); in thepresence of D1 (þ), G1 (V), F1 (X) or sodiumbutyrate (O) at the concentration of 0·5 mM.Data are expressed as percentage ofincorporation as compared with controluntreated cells. Experiments were carried outin triplicate.

537AML Cell Maturation and Apoptosis by Monosaccharide Butyrates

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

cell lines (Collins et al, 1978; Koeffler, 1983). Butyrates arecapable of inducing expression of differentiation genes suchas haemoglobin gamma chains in erythroid precursors andin erythroleukaemic cell lines (Perrine et al, 1987). Studieson in vitro activity of butyrates on neoplastic growth andmaturation have been performed for two decades (Leder &Leder, 1975; Schneider, 1976; Andersson et al, 1989).Nevertheless, applications in clinical trials have been scarce,hampered by an extremely short half life of the molecule.Novogrodsky et al (1983) treated for the first time a childaffected by AML with sodium butyrate 500 mg/kg/d for 10 dby continuous intravenous infusion, obtaining a shortpartial remission. From this report and from anotherunsuccessful trial performed with the same scheduling innine adult AML cases (Miller et al, 1987) it was observed thatplasma levels of butyrate rapidly decreased, with a half-life ofapproximately 6·1 6 1·4 min, and almost absent butyrateurine excretion. The extremely rapid metabolism mayaccount for the absence of relevant results in adults andpartial results in the paediatric case. Notwithstanding theseexperiences, the interest in butyrates has not declined, andrecent trials have explored (in phase I) the possibility oftreating high-risk myelodysplastic syndromes and AMLswith sodium phenylbutyrate (Gore et al, 1996). 11 patientswith MDS and 13 patients with AML were treated, but thetarget plasma levels for butyrate were not reached, eventhough escalating doses in i.v. continuous infusion wereused. 20 out of the total of 24 patients treated have had animprovement in some haematological parameter, showing asignificant clinical activity of sodium phenylbutyrate, withno toxicity. Based on this and other previous in vitro studies(Pouillart et al, 1991; Calabresse et al, 1993), we wereinterested in evaluating the anti-leukaemic activity in vitro ofnew butyrate derivatives, in the attempt to identify thecompound endowed with maximal in vitro efficacy, andhighest in vivo stability coupled to the lowest toxicity. Thenew molecules, similar to others previously proposed(Planchon et al, 1991), were obtained by covalent bindingof butyric acid with natural polyhydroxylated moieties suchas monosaccharides. The stability of the butyrate derivativeswas increased, and the metabolism accomplished inside theleukaemic cell (Pouillart et al, 1992). The direct comparisonof the in vitro effects of sodium butyrate and the butyratemonosaccharide derivatives has enabled definition of idealpro-drugs in terms of induction of leukaemic maturation andapoptosis, to be the object of further study for clinicalapplications.

Maturation therapy in AML is based on the attempt toforce the differentiation blockade of the leukaemic clone, acomplete different strategy of treatment when compared totraditional eradication chemotherapy. It may be argued thatmaturation therapy it is not the ultimate solution to AML,but it may well support, by raising the burden of theimmature myeloid clone, alone or in association withchemotherapy, a combined therapeutic approach whichhas been extremely successful in APL (Avvisati et al, 1994;Fenaux et al, 1994). Moreover, the therapeutic application ofmolecules devoid of paramount cytotoxic effects may proveof particular interest in treating AML in the elderly. Indeed,

further AML cases have to be tested in vitro for sensitivity tomonosaccharide butyrate derivatives, in order to definewhether a FAB or a molecular subtype of AML could beoptimally responsive to G1, D1 or F1. Eventually, themetabolic pathways elicited by butyrates to determinematuration and apoptosis in AML have not been ascertained.The stability of monosaccharide butyrates and their extremeefficacy can render these molecules useful tools in theanalysis of this particular biological aspect.

ACKNOWLEDGMENTS

This work was supported by MURST 40% and by the ItalianAssociation ‘Futuro senza Thalassemia’. We thank FabioCorti and Stefano Salimbeni for helping to prepare thefigures.

REFERENCES

Andersson, L.F., Jokmen, M.& Gahmberg, C.G. (1979) Induction oferythroid differentiation in the human leukaemia cell line K562.Nature, 278, 364–365.

Avvisati, G., Baccarini, M., Ferrara, F., Lazzarino, M., Resegotti, L. &Mandelli, F. (1994) AIDA protocol (all-trans retinoic acid þ

idarubicin) in the treatment of newly diagnosed acute promyelo-cytic leukemia (APL): a pilot study of the Italian cooperative groupGIMEMA. Blood, 84, (abstract supplement), 380a.

Benito, A., Grillot, D., Nunez, G. & Fernandez-Luna, J.L. (1995)Regulation and function of Bcl2 during differentiation-inducedcell death in HL-60. American Journal of Pathology, 146, 481–485.

Bennett, J.M., Catovsky, D., Daniel, M.T., Flandrin, G., Galton, D.A.,Grainick, H.R. & Sultan, C. (1985) Proposed revised criteria forthe classification of acute myeloid leukemia: a report of theFrench–American–British Cooperative Group. Annals of InternalMedicine, 103, 620–625.

Boffa, L.C., Vidali, G., Mann, R.S. & Allfrey, V.G. (1978) Suppressionof histone deacetylation in vivo and in vitro by sodium butyrate.Journal of Biological Chemistry, 253, 3364–3366.

Calabresse, C., Venturini, L., Villa, P., Chomienne, C. & Belpomme, D.(1993) Butyric acid and its monosaccharide ester induceapoptosis in the HL60 cell line. Biochemical and BiophysicalResearch Communications, 195, 31–38.

Candido, E.P.M., Reeves, R. & Davie, J.R. (1978) Sodium butyrateinhibits histone deacetylation in cultured cells. Cell, 14, 105–113.

Chang, S.T. & Yung, B.V. (1996) Potentiation of sodium butyrate-induced apoptosis by vanadate in human promyelocytic leukemiacell line HL 60. Biochemical and Biophysical Research Communica-tions, 221, 594–601.

Collins, S.J. (1977) The HL60 promyelocytic leukemia cell line:proliferation, differentiation and cellular oncogene expression.Blood, 70, 1233–1237.

Collins, S.J., Ruscetti, F.W., Gallagher, R.E. & Gallo, R.C. (1978)Terminal differentiation of human promyelocytic leukemia cellsinduced by dimethylsulfoxide and other polar compounds.Proceedings of the National Academy of Sciences of the United Statesof America, 75, 2458–2462.

Condorelli, G.L., Tocci, A., Botta, R., Facchiano, F., Testa, U., Vitelli,L., Valtieri, M., Croce, C.M. & Peschle, C. (1997) Ectopic TAL-1/SCL expression in phenotypically normal or leukemic myeloidprecursors: proliferative and antiapoptotic effects coupled with adifferentiation blockade. Molecular and Cellular Biology, 17, 2954–2969.

Dong, F., Pouwels, K., Hoefsloot, L.H., Rozenmuller, H., Lowenberg,B. & Touw, I.P. (1996) The carboxy-terminal cytoplasmic region ofthe human G-CSF receptor mediates apoptosis in maturation-incompetent murine myeloid cells. Experimental Hematology, 24,214–220.

Fenaux, P., Chastang, C., Castaigne, S., Archimbaud, E., Sanz, M.,Link, H., Guerci, A., Fegueux, N., Zittoun, R., Stoppa, A.M.,Travade, P., Lamy, T., Maloisel, F., Saduon, A., San Miguel, J., Veil,A., Rayon, C., Conde, E., Fey, M., Bordessuole, D., Ganzer, A.,Bowen, D., Dreyfus, F., Huguet, F., Tilly, H., Guy, H., Auzanneau,G., Chomienne, C. & Degos, L. (1994) Treatment of newlydiagnosed acute promyelocytic leukemia (APL) with all-transretinoic acid (ATRA) followed by intensive chemotherapy (CT):updated results of European group. Blood, 84, (abstract supple-ment), 379a.

Gattei, V., Bernabei, P.A., Pinto, A., Bezzini, R., Ringressi, A.,Formigli, A., Tonini, A., Attadia, V. & Brandi, M.L. (1992) Phorbolester induced osteoclastic like differentiation of a novel human cellline (FLG 29.1). Journal of Biological Chemistry, 116, 437–447.

Gore, S.D., Miller, C.B., Weng, L-J., Burks, K., Griffin, C.A., Chen, T-L.,Burke, P.J., Grever, M. & Rowinsky, E.K. (1996) Improvedhematopoiesis in patients with myelodysplastic syndromes(MDS) and acute myeloid leukemia ( AML) following administra-tion of the putative differentiating agent sodium phenylbutyrate(SPB). Blood, 88, 2315a.

Koeffler, H.P. (1983) Induction of differentiation of humanmyelogenous leukemia: therapeutic implications. Blood, 62,709–721.

Leder, A. & Leder, P. (1975) Butyric acid, a potent inducer oferythroid differentiation in cultured erythroleukemia cells. Cell, 5,317–322.

Manfredini, R., Grande, A., Tagliafico, E., Barbieri, D., Zucchini, P.,Citro, G., Zupi, G., Franceschi, C., Torelli, U. & Ferrari, S. (1993)Inhibition of c-fes expression by an antisense oligomer causesapoptosis of HL-60 induced to granulocytic differentiation. Journalof Experimental Medicine, 178, 381.

Martin, S.J., Bradley, J.G. & Cotter, T.G. (1990) HL-60 cells induced todifferentiate towards neutrophils subsequently die via apoptosis.Clinical and Experimental Immunology, 79, 448–456.

Miller, A.A., Kurschel, E., Osieka, R. & Schmidt, C.G. (1987) Clinicalpharmacology of sodium butyrate in patients with acuteleukemia. European Journal of Cancer and Clinical Oncology., 23,1283–1287.

Novogrodsaky, A., Dvir, A., Shkolnik, T., Stenzel, K.H., Rubin, A.L. &Zaizov, R. (1983) Effect of polar organic compounds on leukemiccells: butyrate-induced partial remission of acute myelogenousleukemia in a child. Cancer, 51, 9–14.

Perrine, S., Miller, B.A., Green, M.F., Cohen, R.A., Cook, N.,Shackleton, C. & Faller, D.V. (1987) Butyric acid analoguesaugment gamma-globin gene expression in neonatal erythroidprogenitors. Biochemical and Biophysical Research Communications,148, 694–700.

Planchon, P., Raux, H., Magnien, V., Rosco, G., Villa, P., Crepin, M. &Brouty-Boye, D. (1991) New stable butyrate derivatives alterproliferation and differentiation in human mammary cells.International Journal of Cancer, 48, 443–449.

Pouillart, P., Cerutti, I., Ronco, G., Villa, P. & Chany, C. (1991),Butyric monosaccharide ester-induced cell differentiation andanti-tumor activity in mice: importance of their prolongedbiological effect for clinical applications in cancer therapy.International Journal of Cancer, 49, 89–95.

Pouillart, P., Cerutti, I., Ronco, G., Villa, P. & Chany, C. (1992)Enhancement by stable butyrate derivatives of antitumor andantiviral actions of interferon. International Journal of Cancer, 51,596–601.

Pouillart, P., Douillet, O., Ronco, G., Villa, P., Santini, V., Scappini, B.,Gozzini, A., Grossi, A. & Rossi Ferrini, P. (1998) Regiospecificsynthesis and biological profiling of butyric and phenylalkylcar-boxylic esters derivated from D-mannose and xylitol: effects ofalkyl chain length on the acute toxicology on mice. EuropeanJournal of Pharmaceutical Science (in press).

Pouillart, P., Ronco, G., Cerutti, I., Chany, C. & Villa, P., (1990) Lowlevel toxicity and antitumor activity of butyric mono- andpolyester monosaccharide derivatives in mice. Journal of BiologicalRegulation and Hemostatic Agents, 4, 135–141.

Riggs, M.G., Whittaker, R.G., Neuman, J.R. & Ingram, V.M., (1977)N-butyrate causes histone modification in HeLa cells and Frienderythroleukaemia cells. Nature, 268, 462–464.

Salem, M., Delwel, R., Touw, I., Mahamud, L. & Lowenberg, B.(1988) Human AML colony growth in serum free medium.Leukemia Research, 12, 157–165.

Santini, V., Colombat, P., Delwel, R., Van Gurp, R., Touw, I. &Lowenberg, B. (1991) Induction of granulocytic maturation inacute myeloid leukemia by G-CSF and retinoic acid. LeukemiaResearch, 15, 341–350.

Sanz, C., Benito, A., Silva, M., Albella, B., Richard, C., Segovia, J.C.,Insunza, A., Bueren, J.A. & Fernandez-Luna, J.L. (1997) Theexpression of Bcl-x is downregulated during differentiation ofhuman hematopoietic progenitor cells along the granulocyte butnot the monocyte-macrophage lineage. Blood, 89, 3199–3204.

Savill, J.S., Wyllie, A.H., Henson, J.E., Walport, M.J., Henson, P.M. &Haslett, C. (1989) Macrophage phagocytosis of ageing neutrophilsin inflammation. Journal of Clinical Investigation, 83, 865–869.

Schneider, F.H. (1976) Effects of sodium butyrate on mouseneuroblastoma cells in culture. Biochemical Pharmacology, 25,2309–2317.

Tran Quang, C., Wesseley, O., Pironin, M., Beug, H. & Ghysdael, J.(1997) Cooperation of Spi.1/PU.1 with an activated erythropoie-tin receptor inhibits apoptosis and Epo-dependent differentiationin primary erythroblasts and induces their kit ligand-dependentproliferation. The EMBO Journal, 16, 5639–5653.

van Wijk, R., Tichonicky, L. & Kruh, J. (1981) Effect of sodiumbutyrate on the hepatoma cell cycle: possible use for cellsynchronisation. In Vitro, 17, 859–862.

q 1998 Blackwell Science Ltd, British Journal of Haematology 101: 529–538

538 V. Santini et al