Vol. 2, 379-387, February 1996 Clinical Cancer Research 379
Phenylbutyrate Induces Apoptosis in Human Prostate Cancer and Is More Potent Than Phenylacetate 1
Michael A. Carducci, 2 Joel B. Nelson, Kirk M. Chan-Tack, Sujatha R. Ayyagari, William H. Sweatt, Pearl A. Campbell, William G. Nelson, and Jonathan W. Simons The Johns Hopkins Oncology Center and The Brady Urological Institute, Baltimore, Maryland 21205
A B S T R A C T
Phenylbutyrate (PB), a novel lead compouad for pros- tate cancer therapy, has molecular activities distinct from its metabolite, phenylacetate (PA). Both PB and PA promote differentiation in human prostate cancer cell lines, yet little data exist comparing the cytotoxic effects of each drug. We found that PB is more potent than PA in vitro; PB is 1.5-2.5 times more active at inhibiting growth and inducing pro- grammed cell death than PA at clinically achievable doses against each human prostate cancer line studied. PB is equipotent to sodium butyrate, which induces apoptosis and differentiation through multiple mechanisms. Exposure of prostate cancer cell lines to PB reduces their DNA synthesis, leads to fragmentation of genomic DNA, and causes 50- 60% of cells to undergo apoptosis. These PB-induced effects are 2-10 times greater than those of the control or PA. The stereotypical changes of apoptosis can be seen with sodium butyrate at similar concentrations, but not with PA. Prostate cancer cell lines overexpressing P-glycoprotein or possessing heterogeneous molecular alterations, including p53 muta- tions, are also sensitive to the effects of PB. In vivo, Copen- hagen rats treated with oral PB had delayed growth of the androgen refractory Dunning R-3327 MAT-LyLu prostate cancer subline by 30-45% in a dose-dependent manner. These results demonstrate that PB induces cytotoxicity via apoptosis in human prostate cancer, in addition to its dif- ferentiating properties.
I N T R O D U C T I O N The phenyl fatty acids PA 3 and PB are potent differentiat-
ing agents (1). Samid and colleagues (2-6) have demonstrated differentiation and growth inhibition induced by PA or PB in malignant glioma, prostate cancer, leukemia, melanoma, and
Received 8/4/95; revised 10/11/95; accepted 10/17/95. Supported by NIH Prostate SPORE Grant CA-58236.
2 To whom requests for reprints should be addressed, at The Johns Hopkins Oncology Center and The Brady Urological Institute, 600 North Wolfe Street, Ross 359, Baltimore, MD 21205. Phone: (410) 614-3977; Fax: (410) 614-9006; E-mail: [email protected]. edu. 3 The abbreviations used are: PA; sodium phenylacetate; PB, sodium phenylbutyrate; BU, sodium butyrate; PSA, prostate-specific antigen; MDR, multidrug resistance; ICso, 50% inhibitory concentration.
neuroblastoma cell lines in vitro and in vivo. These differenti- ating agents also induce fetal hemoglobin production in children with sickle cell anemia and thalassemia (7, 8), are used in children with urea cycle disorders (9, 10), and seem to be nontoxic. In addition, these drugs have been used in adults with idiopathic hyperammonemia syndrome after high dose chemo- therapy without significant end organ toxicities (11).
PB is [3 oxidized in vivo to PA in mitochondria of the liver and kidney rapidly and provides a more tolerable delivery of PA (12-14). PB seems to have cellular and molecular effects dis- tinct from PA. In vitro studies provide insight to the unique effects of each agent, because PB is not [3 oxidized in vitro to PA in prostate cancer cell lines. 4 Wood et al. (15) reported that PA increases, whereas PB decreases, PSA protein production by the LNCaP prostate cancer cell line (15). Theoretically, differ- entiation could increase PSA secretion, whereas cytotoxicity leads to a decline in PSA production. On the basis of these data, we tested the hypothesis that PB, unlike PA, may promote apoptosis as well as differentiation. In this respect, PB may be analogous to BU, a short chain fatty acid, which is a differen- tiating agent and which also demonstrates cytotoxic effects in
vitro and in vivo (16-18). To investigate this hypothesis, PA, PB, and BU were tested
for cytotoxicity against human prostate cancer cell lines, includ- ing those with acquired p53 mutations and those selected for MDR1 overexpression. The potency of each agent was tested in the range of clinically achievable levels (1-5 mu) in humans. The ability of these agents to induce apoptosis in human prostate cancer was assayed in vitro directly. Dose-dependent cytotoxic effects were studied in vivo in the Dunning rat MAT-LyLu prostate cancer model.
M A T E R I A L S A N D M E T H O D S
Growth Inhibition and Cytotoxicity Assays Comparing Potency of PB, PA, and BU against Human Prostate Cancer Cell Lines
The human prostate cancer cell lines LNCaP-ATCC, LNCaP-GW, PC-3, PPC-1, and TSU-Adr 1000 (obtained from W. G. Nelson) and the Dunning rat R-3327 MAT-LyLu subline were plated uniformly in 96-well plates (5 × 103 cells/well). PB and PA (Triple Crown America, Perkasie, PA) and BU (Sigma Chemical Co., St. Louis, MO) were formulated in RPMI 1640 (GIBCO-BRL, Gaithersburg, MD) with 10% FCS (JRH Bio- chemicals, Lenexa, KS). After 96 h of drug exposure, treated cells were fixed and stained with 5% methanol and 0.5% crystal violet, air dried, and resolubilized with 1% SDS. Cell density was assayed by spectrophotometer at 590 nm (Bio-Rad 450 Micro plate reader; Bio-Rad, Hercules, CA). Each condition was expressed as
A . 1.2 F-I MLL PA [ ] MLL BU [ ] MLL PB [ B . Io TSU,', ~ ,SU ,U " 'SU '" I
! 1T- ~ T ~ T ~ 0.8
~ 1- O.6 !0" 0.2
- Pc3- - - - 3 - - Pc3 P---~]
0
C , 1.2- D, m-]PPCl PA NPPCl BU BBPPCl PB
1 T I I m
0.8 I
0.6
~ 0.4
!o,,. !
1
0.8
0.6
0.4
E 0.2
0
T . . b " - 1 - . z : l :z
- r - - I -
" 1 - - w
0 0.75 1.25 2.5 5 10 0 0.75 1.25 2.5 5 10
CONCENTRATION (mM) CONCENTRATION (mM) Fig. 1 Composite inhibitory potency of PB, PA, and BU against prostate cancer cell lines. Growth inhibition and cytotoxicity as proportionate survival (i.e., 0.4 = 40% inhibition) are plotted against varying concentrations of each agent. Inhibition after treatment with PA, BU, and PB is shown. Cell lines: A, MLL-MAT-LyLu (MLL) rat prostate cancer; B, TSU human prostate cancer; C, PC-3 human prostate cancer; D, PPC-1 human prostate cancer; E, LNCaP-ATCC human prostate cancer; and F, LNCaP-GW (p53 mutant) human prostate cancer.
Fig. 2 Morphological changes in PB-, PA-, and BU-treated TSU prostate cancer cells, a, untreated TSU cells; b, 2.5 mM PB-treated TSU cells; c, 2.5 mM PA-treated TSU cells; and d, 2.5 mM BU-treated TSU cells for 72 h, fixed and stained with hematoxylin and eosin (× 160).
an average of eight determinations for that concentration of the drug.
Phenotypic Changes of Human Prostate Cancer Cell Lines after Exposure to PB, PA, and BU
Each prostate cancer cell line was plated on eight-chamber glass slides (Nunc, Inc., Naperville, IL), allowed to adhere, and exposed to 2.5 and 5 m~ concentrations of the drugs. After 72
h of exposure, cells were fixed in 100% ethanol and stained with
hematoxylin and eosin.
Quantitation of DNA Synthesis after PB Treatment
The androgen-independent human prostate cancer cell line
TSU was exposed to 2.5 mM PB for varying time intervals from
30 rain to 96 h. After each time interval, cells were labeled with
1 i.LCi/ml [3H]thymidine (Amersham, Arlington Heights, IL).
Prostate cancer- precipitable radioactivity was determined using a liquid scintillation counter. The concentration of DNA for each sample was determined using diphenylamine as described
previously (19).
Cell Death Analysis Patterns of DNA damage were analyzed following drug
treatment. DNA Fragmentation Assay. Prostate cancer cell lines
were exposed to 2.5 and 5 mM PB or PA for 2, 4, and 5 days. All cells (adherent and floating) were harvested, and the sample was divided equally. Half of the cells were lysed with 10 mM Tris-1 mMEDTA (pH 7.4) and 0.2% Triton X-100 (Sigma), and the
remaining cells were processed for flow cytometry (see below).
Centrifugation, precipitation, and quantification of high and low
molecular weight DNA were performed as described previously
(19-22). The percentage DNA fragmentation was calculated
using the following equation: %DNA fragmentation = DNA in supernatant/DNA in supernatant + DNA in pellet or low mo-
lecular weight DNA/total DNA extracted.
Direct Fluorescence Detection of Digoxigenin-labeled Genomic DNA by Flow Cytometry. Using the ApopTag S71 l0 in situ apoptosis detection kit and protocol (Oncor, Inc., Gaithersburg, MD), samples were prepared for flow cytometry.
Using flow cytometry (FACStar Plus; Becton Dickinson, San
Fig. 3 Dose-related phenotypic changes in PC-3 prostate cancer cells after PB treatment, a, untreated PC-3 cells; b, 2.5 mM PB-treated PC-3 cells; and c, 5 mM PB-treated PC-3 cells for 72 h, fixed and stained with hematoxylin and eosin (× 160).
Jose, CA), populations of apoptotic and nonapoptotic nuclei were quantified. The ratio of apoptotic nuclei to all nuclei counted determined the apoptotic index. Dexamethasone (1 lXM)-treated human peripheral blood lymphocytes and termi- nal deoxynucleotidyltransferase enzyme-excluded samples acted as positive and negative controls, respectively (20).
DNA Ladder ing by Gel Electrophoresis. Prostate can- cer cell lines were exposed to 2.5 rnM PB for 2, 4, and 5 days. Cells were harvested after removal of media and nonadherent cells, lysed with 10 mM Tris (pH 8), 5 mM EDTA, and 0.5% Triton X-100 (Sigma), centrifuged, and incubated with 20 I~g/ml DNase-free RNase (Boehringer Mannheim, Indianapolis, IN; 500 Ixg/ml). The supernatant was then treated with 10% SDS and 300 lxg proteinase K/ml (Boehringer Mannheim; 16.4 mg/ml) for 1 h at 50°C. DNA was precipitated, washed, and resuspended. DNA was stained with ethidium bromide after 2% agarose gel electrophoresis.
Evaluation of Tumor Growth in Rats Exposed to PB in Their Drinking Water
Four groups of 10 male Copenhagen rats (Harlan-Sprague- Dawley, Indianapolis, IN), aged 12-16 weeks, were inoculated
Table 1 Percentage of DNA fragmentation of PB-treated TSU prostate cancer cells
Percentage of low molecular weight DNA fragmentation = low molecular weight DNA/low molecular weight DNA + high molecular weight DNA
with 1 × 10 4 Dunning R-3327 MAT-LyLu cells in the right
hind leg. Three days later, PB in final concentrations of 2, 50, and 100 mM was added to the drinking water. Rats in the control arm received sodium acetate at 50 rnM in their drinking water. On average, each rat consumed 6 - 8 ml water/day. Rats were evaluated twice weekly for tumor development and progression
by following the size of the tumor at the inoculum site. Inves- tigators measuring tumor size by calipers were blinded to treat- ment arm and PB dose.
Time (days) Fig. 4 Apoptotic assessment of PB-treated TSU prostate cancer cells. A, fold increase over controls of fragmented low molecular weight DNA after treatment with PB and PA at 2.5 or 5 mM drug concentrations. PB increases the percentage of fragmented DNA by 2-6-fold over PA- treated and untreated control TSU prostate cancer cells. B, fold increase in apoptotic nuclei after treatment with PB and PA at 2.5 or 5 mM drug concentrations. PB increases the percentage of apoptotic nuclei by 2-10-fold over PA-treated and untreated control TSU prostate cancer cells.
R E S U L T S
Comparison of PB with PA and BU for Cytotoxic Ef- fects. Growth inhibition assays were conducted using the Dunning rat R-3327 MAT-LyLu subline and the human prostate cancer cell lines TSU, PC-3, PPC-I, LNCaP-ATCC, and LN- CaP-GW. Twofold dilutions from 20 to 0.185 mM were tested for each drug.
Fig. 1 shows the composite data for each cell line, drug, and concentration after a 4-day exposure. The ICsos are 2.5 mM
PB in the TSU and LNCaP cell lines and 5 mM PB in the Mat-LyLu and PC-3 cell lines. The IC5o for BU is similar to that for PB and is 5 mM or less in the MAT-LyLu, TSU, and LNCaP-GW cell lines. This contrasts with PA, for which the ICso is 10 mM for Mat-LyLu and PC-3 cell lines, with the other cell lines requiring a far greater concentration than that which is tolerated clinically (10-20 mM). At 10 mM PB, all six of the cell lines tested have <50% growth inhibition. At clinically achiev- able levels, PB possesses greater potency against all tested prostate cancer cell lines compared with PA. BU is relatively equipotent to PB in every case in inducing growth inhibition or cell death•
The prostate cancer cell lines tested have heterogeneous p53, ras, and Rb mutations (23, 24), which may affect cellular and molecular responses to drug exposure. For example, the LNCaP-ATCC and LNCaP-GW lines differ; LNCaP-GW cells possess ap53 mutation (codon 273, arg --~ his; Ref. 23). Despite this, LNCaP-ATCC and LNCaP-GW cells are equally sensitive to the inhibitory effects of PB at 2.5 and 5 n ~ (Fig. 1). Overall, p53 status does not seem to play a major role in the activity or effect of these drugs, given the minimal difference in effect between PB and BU and no difference in terms of PA effect in the LNCaP variants.
An adriamycin-resistant TSU cell line (TSU-Adr 1000) was used to determine whether P-glycoprotein (MDR1) overex- pression confers PB resistance• Using the growth inhibitory assay, wild-type TSU and TSU-Adr 1000 cell lines were ex- posed to equivalent concentrations of PB, vinblastine, or etopo- side. Of note, both vinblastine and etoposide are used clinically in prostate cancer• Although the TSU-Adr 1000 cell line was resistant to vinblastine and etoposide, it was growth inhibited by PB (ICso, 2.5 mM). The MDR1 status of this cell line did not affect its response to PB, suggesting that PB may be useful in cancers with high levels of MDR1 expression.
Phenotypie Changes after PB, PA, and BU Treatment in Vitro. Hematoxylin and eosin stains demonstrated clear, dose-related, phenotypic changes for each agent and similar changes for PB and BU at the same doses. Phenotypic changes in the TSU cell line exposed to media or 2.5 mM PB, BU, and PA are shown in Fig. 2. Cell numbers reflect the growth inhib- itory effects of each agent with PB and BU, demonstrating dramatic inhibition (Fig. 1B). The treated cells demonstrate cellular flattening and extension of psuedopodia. The cytoplasm becomes vacuolized. In work by Samid et al. (3), these vacuoles are shown to contain lipid droplets after exposure to PA at a dose of 5 raM. In this study, the nuclei of PB-treated prostate cancer cells flatten as well, and nucleoli are well preserved. These histological changes in prostate cancer are consistent with cellular effects accompanying prostate cell apoptosis, as re- ported by Isaacs (25).
Fig. 3 demonstrates the dramatic phenotypic changes of the PC-3 line exposed to increasing concentrations of PB for 72 h. A marked increase in cells showing cytoplasmic extension, cytoplasm vacuolization, and nuclear disruption with multinu- cleated cells is apparent. These changes correlate with the growth inhibition assay for PC-3 shown in Fig. 1C. Of note, at 2.5 mM PB, the ICso has not been reached, yet the distinct changes related to drug exposure and associated with cellular injury are seen clearly. Similar data showing phenotypic
Fig. 5 Comparison of the ability of PB and PA to induce apoptosis. The plot compares the fold increase of low molecular weight DNA fragmentation over the fold increase in apoptotic nuclei of each prostate cancer cell line after treatment with PA or PB. PB-treated prostate cancer cell lines undergo apoptosis, as demonstrated by these tech- niques. PA does not induce significant apoptosis.
changes at doses lower than the IC5o were observed for the other cell lines: PPC-1, LNCaP-ATCC, and LNCaP-GW.
Inhibi t ion of DNA Synthesis. After treatment with PB, the TSU cell line demonstrates marked reduction in DNA syn- thesis by [3H]thymidine incorporation over the first 24 h and remains depressed, compared with controls, for as long as 96 h (data not shown). Samid et al. (3) has shown previously that DNA synthesis is inhibited after treatment with PA.
PB Induces Apoptosis. Low molecular weight DNA fragments are a marker of endonuclease activity accompanying apoptosis. PB induces apoptosis in the five prostate cancer cell lines tested (PC-3, TSU, LNCaP, DU-145, and PPC-1) at 2.5 mM (2-6-fold increase in the percentage of fragmented DNA over control at each time point) and at 5 mM (2.5-10 fold increase). PA induces DNA fragmentation above controls infre- quently, and in only two cell lines (LNCaP and PPC-1) does it increase DNA fragments 1.3-2 times control. Table 1 demon- strates the raw data for the TSU prostate cancer line.
The PB-induced DNA fragmentation data are confirmed by flow cytometric analysis of digoxigenin-labeled, genomic DNA. Fig. 4 demonstrates both fragmented low molecular weight DNA and antidigoxigenin-positive cells as measures of apopto- sis in the TSU cell line after exposure to PA and PB. The percentage of apoptotic nuclei after PB exposure increases 2-10 fold over controls, whereas PA does not induce apoptosis over controls as demonstrated by antidigoxigenin antibody fluores- cence. Fig. 5 demonstrates the correlation between the DNA fragmentation assay and the antidigoxigenin antibody staining (r = 0.688) and segregates PB from PA as an inducer of prostate cancer apoptosis clearly.
A B C D E F G
984
492
123
Fig. 6 Agarose gel electrophoretic patterns of DNA isolated from PB-treated and untreated TSU prostate cancer cells. Lane A, 123-bp marker; Lane B, k-HinDIII marker; Lane C, blank; Lane D, untreated TSU cells harvested at 96 h; Lane E, 2.5 mM PB-treated TSU cells for 48 h; Lane F, 2.5 mM PB-treated TSU cells for 96 h; and Lane G, 2.5 mM PB-treated TSU cells for 120 h.
Fig. 6 shows fragmentation of genomic DNA from PB- exposed TSU cells for 4 and 5 days. No DNA fragmentation of control TSU cells is seen at 4 days. Only adherent cells were harvested to represent the DNA fragmentation status of intact cells, thus underestimating the level of apoptosis seen in the previous methods potentially. Similar fragmentation ladders were generated for DU-145 and LNCaP-GW cell lines exposed to PB for 4 days at a 2.5 rnM concentration (data not shown). BU showed equal ability to induce apoptosis in the same cell lines.
The time course of the effect of PB on prostate cancer cells begins within the first 24 h of exposure, with a reduction in DNA synthesis. The morphological changes of cytoplasmic vacuolization and nuclear swelling occur after 48-72 h of continuous PB treatment. Finally, PB induces cell death via apoptosis as early as 48 h, continuing for at least 120 h of drug exposure.
In Vivo Evaluat ion of PB as a Cytotoxic Agent in Pros- tate Cancer. After identifying the time course of the effect of PB on human prostate cancer in vitro, we tested PB in vivo. The Dunning rat R-3327 MAT-LyLu prostate cancer line, which doubles every 24-36 h (26) and is androgen independent, was used. Fig. 7 shows a dose-related inhibition of tumor growth in animals with PB-supplemented drinking water. Rats with 100 mM PB-supplemented drinking water had a 30 -45% reduction in tumor size compared with controls. This difference was statistically significant (P < 0.026 by paired t test). PB at 100
with PB in rats with preestab- "- - 14 lished MAT-LyLu prostate t"q
cancer. Rats were inoculated E with 1 × 10 4 MAT-LyLu ~ 12 cells in the right hind limb. After 3 days, PB (or Na ace- ._ 1 0 tate) was added to the rats' CO drinking water. The average o 8 size of all tumors in each co- E hort is plotted against days ~.~ 6 since tumor implantation. Er- ror bars are shown for the 1:~ control group (50 mM Na ac- ~ 4 etate) and the smallest tumor group (100 mM PB) and rep- *~ 2 resent the range of tumor sizes in each cohort. The av- 0 erage tumor size was signifi- cantly smaller in the treated group (P = 0.027).
e50mMNaAc
• B 2 . 0 m M PB
÷50mM PB
~ 100mM PB
0 10 14 17 21
Time (Days Post Implant 1 x 10 4 cells)
2 3
mM given ad libitum in the drinking water can delay tumor progression in this model. Serum sampling for PB, PA, and phenylacetylglutamine demonstrated detectable levels of pheny- lacetylglutamine (0.1-1.6 mM) but not PB or PA in the animals receiving PB. No correlation could be established between an- fitumor effect and metabolite concentration.
D I S C U S S I O N
PB is an active differentiating agent and cytotoxic com- pound available for clinical testing in human prostate cancer therapy. In this report, we demonstrate for the first time that PB has cytotoxic effects and can induce apoptosis in human prostate cancer cell lines. Therefore, PB is an attractive candidate to develop as a novel, lead compound for the therapy of prostate cancer.
We found that PB is more potent at inhibiting the growth of the human prostate cancer cell lines than PA. PB inhibits pros- tate cancer growth at clinically achievable levels, whereas the concentration of PA required to inhibit cell growth in vitro is at the upper limits of the plasma concentrations produced by tolerated doses of the agent in patients. Morphological changes suggesting injury due to PB can be seen at doses below the IC5o. Also, PB induces its effect on prostate cancer cell lines inde- pendent of p53 mutational status or MDR1 overexpression. PB increases the apoptotic rate in prostate cancer cell lines over controls dramatically, 2-10 times the basal rate of apoptosis. Even a slight modification in apoptotic death rates may translate to the clinical significance of PB as an antitumor agent (27, 28). PB as monotherapy can also slow the growth of transplanted prostate tumors in rats when given in the rats' drinking water.
PA and PB have been reported to have many antitumor activities, including induction of tumor cytostasis and induction of tumor differentiation. These compounds enhance expression of MHC class I and class II, enhance expression of cellular adhesion molecules, reduce biosynthesis and secretion of trans- forming growth factor-132, reduce cell-associated urokinase plasminogen activity, and prevent carcinogenesis by 5-azo-de- oxy-cytidine (3, 29, 30). It has been proposed that PA and PB
affect gene expression by modification of lipid metabolism, hypomethylation of DNA, inhibition of protein isoprenylation, or glutamine starvation of cancer cells (30, 31).
In this report, we show that PB has potency equivalent to BU in inhibiting the growth of human prostate cancer cell lines. The structure of PB is similar to that of BU. BU itself has been reported to reduce anaerobic glycolysis, increase cyclic AMP, induce apoptosis, alter growth factor and hormone expression, increase histone acetylation, and alter chromatin conformation (17, 32-34). Nuclear swelling is quite dramatic and consistent in all the cell lines after exposure to either PB or BU. Like BU, PB can induce apoptosis, as demonstrated by fragmentation of genomic DNA. These effects associated with both BU and PB are primarily cytotoxic and may be related to the induction of a differentiated state. Our studies show that a pathway of PB drug activity includes signaling cancer cell apoptosis. This finding is consistent with the studies of Heerdt et al. (35), who have described the role of short chain fatty acids and BU in the differentiation and subsequent induction of apoptosis in human colon carcinoma cell lines.
PB is 13 oxidized in vivo to PA in the mitochondria of cells rapidly (12). Pharmacokinetic data from bolus human studies of PB suggest that PA and PB may have distinct pharmacological and pharmacokinetic properties (14). Thus, PB may induce an effect at the molecular level before it is metabolized to PA. Evaluation of bioactivity of PB in clinical trials for pure PB effects remains a challenge given its in vivo metabolism to PA. Determination of distinct in vitro PB versus PA effects on the expression of PSA, interleukin-6, endothelin- 1, and other mark- ers of prostate cancer morbidity may aid in identifying bioac- tivity unique to PB in vivo (36, 37).
PB is currently in Phase I testing for patients with refrac- tory solid tumors. Phase I testing will determine toxicities and will evaluate pharmacokinetics of both i.v. and oral formula- tions. Theoretically, differentiating agent therapies or even an- timetastatic therapies require continuous exposure of the agent to maintain the desired effect of manipulating established tumor sites. Preventive therapies are likely to require continuous ex-
posure as well. If PB has cytotoxic effects in addition to differ-
entiating effects, intermittent schedules in combination with
traditional chemotherapy agents, or with newer therapies such as
gene-modified tumor vaccines, may be more effective (3, 38).
PB is potentially a nontoxic adjunct to current therapies. Further
evaluation of its mechanism of action and range of activities is
needed.
ACKNOWLEDGMENTS We express appreciation to Saul Brusilow and O. Michael Colvin
for helpful discussion and to Jo Nell Foster for manuscript preparation.
REFERENCES 1. Samid, D., Shack, S., and Sherman, L. T. Phenylacetate: a novel nontoxic inducer of tumor cell differentiation. Cancer Res., 52: 1988- 1992, 1992.
2. Samid, D., Ram, Z., Hudgins, W. R., Shack, S., Liu, L., Walbridge, S., Oldfield, E. H., and Myers, C. E. Selective activity of phenylacetate against malignant gliomas: resemblance to fetal brain damage in phe- nylketonuria. Cancer Res., 54: 891-895, 1994.
3. Samid, D., Shack, S., and Myers, C. E. Selective growth arrest and phenotypic reversion of prostate cancer cells in vitro by nontoxic phar- macological concentrations of phenylacetate. J. C. I., 91: 2288-2295, 1993.
4. Samid, D., Yeh, A., and Prasanna, P. Induction of erythroid differ- entiation and fetal hemoglobin production in human leukemic cells treated with phenylacetate. Blood, 80: 1576-1581, 1992.
5. Liu, L., Shack, S., Stetler-Stevenson, W. G., Hudgins, W. R., and Samid, D. Differentiation of cultured human melanoma cells induced by the aromatic fatty acids phenylacetate and phenylbutyrate. J. Invest. Dermatol., 103: 335-340, 1994.
6. Wada, R., Han, G., Moore, T., Samid, D., and Sidell, N. Effects of phenylacetate and its interaction with retinoic acid on human neuroblas- toma differentiation. Proc. Am. Assoc. Cancer Res., 35: 419, 1994.
7. Fibach, E., Prasanna, P., Rodgers, G., and Samid, D. Enhanced fetal hemoglobin production by phenylacetate and 4-phenylbutyrate in ery- throid precursors derived from normal donors and patients with sickle cell anemia and beta-thalassemia. Blood, 82: 2203-2209, 1993.
8. Dover, G. J., Brusilow, S. W., and Charache, S. Induction of fetal hemoglobin production in subjects with sickle cell anemia by oral sodium phenylbutyrate. Blood, 84: 339-343, 1994.
9. Brusilow, S. W., Danney, M., Waber, L. J., Batshaw, M., Burton, B., Levitsky, L., Roth, K., McKeethren, C., and Ward, J. Treatment of episodic hyperammonemia in children with inborn errors of urea syn- thesis. N. Engl. J. Med., 310: 1630-1634, 1984.
10. Brusilow, S. W. Phenylacetylglutamine may replace urea as a vehicle for waste nitrogen excretion. Pediatr. Res., 29: 147-150, 1991.
11. Mitchell, R. B., Wagner, J. E., Karp, J. E., Watson, A. J., Brusilow, S. W., Przepiorka, D., Storb, R., Santos, G. W., Burke, P. J., and Saral, R. Syndrome of idiopathic hyperammonemia after high-dose chemo- therapy: review of nine cases. Am. J. Med. 85: 662-667, 1988.
12. Knoop, F. Der abbau aromatischer fettsanre tier korper. Beitr. Chem. Pysiol. Pathol., 6: 105-162, 1905.
13. Brusilow, S. W., Tinker, J., and Batshaw, M. L. Amino acid acylation: a mechanism of nitrogen excretion in inborn errors of urea synthesis. Science (Washington DC), 207: 659-661, 1980.
14. Piscitelli, S. C., Thibault, A., Figg, W. G., Tompkins, A., Headlee, D., Liebermann, R., Samid, D., and Myers, C. E. Disposition of phe- nylbutyrate and its metabolites, phenylacetate and phenylacetylglu- tamine. J Clin. Pharmacol., 35: 368-373, 1995.
15. Wood, C. G., Lee, C., Grayhack, J. T., and Kozlowski, J. M. Phenylacetate and phenylbutyrate promote cellular differentiation in human prostate cancer systems. Proc. Am. Assoc. Cancer Res., 35: 403, 1994.
16. Augeron, C., and Laboisse, C. L. Emergence of permanently dif- ferentiated cell clones in a human colonic cancer cell line in culture after treatment with sodium butyrate. Cancer Res., 44: 3961-3963, 1984.
17. Hague, A., Manning, A. M., Hanlon, K. A., Huschtscha, L. I., Hart, D., and Paraskeva, C. Sodium butyrate induces apoptosis in human colonic tumor cell lines in a p53 independent pathway: implications for the possible role of dietary fibre in the prevention of large bowel cancer. Int. J. Cancer, 55: 498-505, 1993.
18. Barnard, J. A., and Warwick, G. Butyrate rapidly induces growth inhibition and differentiation in HT-29 cells. Cell Growth. & Differ., 4: 495-501, 1993.
19. Burton, K. A study of the conditions and mechanism of the diphe- nylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem. J., 62: 315-323, 1956.
20. Wyllie, A. H. Glucocorticoid-induced thymocyte apoptosis is asso- ciated with endogenous endonuclease activation. Nature (Lond.), 284: 555-556, 1980.
21. Colotta, F., Re, F., Polentarutti, N., Sozzani, S., and Mantovani, A. Modulation of granulocyte survival and programmed cell death by cytokines and bacterial products. Blood, 80: 2012-2020, 1992.
22. Hanaoka, K., Fujita, N., Lee, S., Seimiya, H., Maito, M., and Tsuruo, T. Involvement of CD45 in adhesion and suppression of apop- tosis of mouse malignant T-lymphoma cells. Cancer Res., 55: 2186- 2190, 1995.
23. Carroll, A. G., Voeller, G. J., Sugars, L., and Gelmann, E. P. p53 oncogene mutations in three human prostate cancer cell lines. Prostate, 23: 123-134, 1993.
24. Isaacs, W. B., Carter, B. S., and Ewing, C. M. Wild-type p53 suppresses growth of human prostate cancer cells containing mutant p53 alleles. Cancer Res., 51: 4716-4720, 1991.
25. Isaacs, J. T., Landniu, P. I., Berges, R., Martikainen, P., Kyprianou, N., and English, H. F. Androgen regulation of programmed cell death of normal and malignant prostatic cells. J. Androl., 13: 455-484, 1992.
26. Isaacs, J. T., Isaacs, W. B., Feitz, W. F. G., and Scheres, J. Establishment and characterization of seven Dunning rat prostatic can- cer cell lines and their use in developing methods for predicting meta- static abilities of prostatic cancers. Prostate, 9: 261-281, 1986.
27. Isaacs, J. T. The timing of androgen ablation therapy and/or che- motherapy in the treatment of prostatic cancer. Prostate, 5: 1-18, 1984.
28. Isaacs, J. T. Antagonistic effect of androgen on prostatic cell death. Prostate, 5: 545-558, 1984.
29. Liu, L., Bar-Ner, M., Weber, J., Danielpour, D., Qian, S. W., Shearer, G. M., and Samid, D. Enhancement of tumor immunogenicity by phenylacetate and derivatives: changes in surface antigens and tumor derived immunosuppressive factors. Proc. Am. Assoc. Cancer Res., 35: 481, 1994.
30. Hudgins, W. R., Pineau, T., Sher. T., Gonzales, F. J., Myers, C. E., and Samid, D. Anticancer activity of phenylacetate and related aromatic fatty acids: correlation with lipophilicity and capacity to activate nuclear receptor. Proc. Am. Assoc. Cancer Res., 35: 391, 1994.
31. Samid, D., Hudgins, W. R., Shack, S., Miller, A. C., Liu, L., Prasanna, P., Walls, R., and Myers, C. E. Phenylacetate and derivatives: simple compounds with complex antitumor activities. Proc. Am. Assoc. Cancer Res., 35: 408, 1994.
32. Malvilio, F., Kreider, B. L., Valtieri, M., Naso, G., Shirsat, N., Venturelli, D., Reddy, E. P., and Rovera, G. Alteration of growth and differentiation factors response by Kirsten and Harvey sarcoma viruses in the IL-3-dependent routine hematopoietic cell line 32D C13. Onco- gene, 4: 301-308, 1989.
33. Norton, V. G., Imai, B. S., Yau, P., and Bradbury, E. M. Histone acetylation reduces nucleosome core particle linking number change. Cell., 57: 449-457, 1989.
34. deFazio, A., Chiew, Y. E., Donoghue, C., Lee., C., and Sutherland, R. L. Effect of sodium butyrate on estrogen receptor and epidermal growth factor receptor gene expression in human breast cancer cell lines. J. Biol. Chem., 267: 18008-18012, 1992.
35. Heerdt, B. G., Houston, M. A., and Augenlicht, L. H. Potentiation by specific short-chain fatty acids of differentiation and apoptosis in human colonic carcinoma cell lines. Cancer Res., 54: 3288-3294, 1994. 36. Twillie, D. A., Eisenberger, M. A., Carducci, M. A., Hseih, W. S., Kim, W. Y., and Simons, J. W. Interleukin-6: a candidate mediator of human prostate cancer morbidity. Urology, 45: 542-549, 1995.
37. Nelson, J. B., Hedican, S. P., George, D. J., Reddi, A. H., Pianta- dosi, S., Eisenberger, M. A., and Simons, J. W. Identification of endo-
thelin-1 in the pathophysiology of metastatic adenocarcinoma of the prostate. Nat. Med., 1: 944-949, 1995.
38. Figg, W. D., Walls, R. G., Cooper, M. R., Thibault, A., Sartor, O., McCall, N. A., Myers, C. E., and Samid, D. In vitro antitumor effect of hydroxyurea on hormone refractory prostate cancer ceils and its poten- tiation by phenylbutyrate. Anticancer Drugs, 5: 336-342, 1994.
1996;2:379-387. Clin Cancer Res M A Carducci, J B Nelson, K M Chan-Tack, et al. and is more potent than phenylacetate.Phenylbutyrate induces apoptosis in human prostate cancer
