Special article

Vitamin D analogs, a new treatment for

retinoblastoma: The first Ellsworth Lecture1
Ophthalmic Genetics 1381-6810/02/
US$ 16.00
Daniel M. Albert1
Ophthalmic Genetics – 2002, Vol. 23
No. 3, pp. 137–156
Robert W. Nickells1
© Swets & Zeitlinger 2002 David M. Gamm1
Accepted 21 February 2002 Michele L. Zimbric1
Cassandra L. Schlamp1
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

Mary J. Lindstrom2
Isabelle Audo1
1
Department of Ophthalmology & Visual Sciences, School of
Medicine, and 2Department of Biostatistics, University of
Wisconsin, Madison, WI, USA
For personal use only.

Abstract Purpose: This lecture honors the memory of Dr. Robert Correspondence and reprint
M. Ellsworth, an important figure in the development of current treat- requests to:
ments of retinoblastoma (RB), and reviews our studies of vitamin D Daniel M. Albert, MD, MS
analogs as treatments for retinoblastoma in two experimental mouse Dept of Ophthalmology & Vis
Sciences
models. We identified vitamin D receptors in retinoblastoma and
F4/334 Clinical Science Center
examined the effectiveness and mechanism of action of these analogs.
600 Highland Avenue
Methods: Reverse-transcriptase polymerase chain reaction (RT-PCR) Madison WI 53792-3284
amplification was used to detect vitamin D receptor mRNAs in human USA
and mouse retinoblastomas. The effectiveness and toxicity of vitamin Tel: (608)263-9797
D2, calcitriol, and synthetic analogs were studied in the athymic/Y-79 Fax: (608)263-1466
xenograft and transgenic mouse models of RB. Dosing was 5X/week E-mail: dalbert@facstaff.wisc.edu
for five weeks. Dose-response studies focused on tumor inhibition;
toxicity studies investigated survival and serum calcium. The mecha- 1
The Inaugural Ellsworth Lecture
nism of action of vitamin D was investigated using terminal transferase delivered at the Combined
dUTP labeling 3¢-overhang ligation to measure apoptosis; immunohis- International Society for Genetic
tochemistry measured p53-dependent gene expression and cell prolif- Eye Disease (ISGED) and
eration. Result: Vitamin D receptor mRNAs were detectable in Y-79 International Symposium on RB
RB cells, LHb-Tag tumors, and human RB specimens using RT-PCR. Meeting in Fort Lauderdale, Florida,
May 4, 2001.
Calcitriol inhibited cell growth in vitro. Calcitriol and vitamin D2 inhib-
ited in vivo growth in xenograft and transgenic models, but therapeu- Acknowledgements:
tic levels were toxic due to hypercalcemia. Two analogs, 16,23-D3 and This research was supported by the
1a-OH-D2, inhibited tumors in animal models of RB with reduced tox- Research To Prevent Blindness
icity. The mechanism of action appears related to increased p53-related (RPB) and NIH/NEI RO1 grant
gene expression resulting in increased apoptosis. Conclusion: 16,23-D3 EYO1917. 16,23-D3 was graciously
and 1a-OH-D2 are effective in tumor reduction in two mouse models provided to us by Ilex Corporation.

Treatment of retinoblastoma with vitamin D analogs 137

 1a-OH-D2 and supplemental funds
for animal care costs were graciously
provided to us from BoneCare
International, Inc. The authors also
wish to thank these physicians and
their patients for donating human
RB samples for vitamin D receptor
analysis: David H. Abramson, MD,
Cornell Medical College, New York
NY; Thomas Lee, MD, New York
Presbyterian Hospital; J. William
Harbour, MD PhD, Washington
University Medical School; Jerry
Shields, MD, and Carol Shields, MD,
Wills Eye Hospital; Timothy Murray,
MD, Bascom Palmer Eye Institute;
Thaddeus Dryja, MD, Massachusetts
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14
Eye & Ear Infirmary; Joan O’Brien,
MD, University of California, San
Francisco; A. Linn Murphree, MD,
Anita Fisher, PhD, Children’s
Hospital, Los Angeles; Monte Mills,
MD, Children’s Hospital of
Philadelphia, and Monsoor
Movaghar, MD, Davis Duehr Dean
Eye Clinic, Madison WI.
For personal use only.
Fig. 1. Robert M. Ellsworth, M.D.
Courtesy of D. Jackson Coleman,
M.D.
138
of RB with low toxicity. These results warrant initiating phase 1 and
phase 2 clinical studies in children.
Key words Retinoblastoma; vitamin D analogs; p53; p21; apoptosis
Introduction Dr. Robert M. Ellsworth (1928–1994) (Figure 1) was
described as “the world’s leading authority on tumors of the eye, par-
ticularly RB . . .”1 Dr. Ellsworth made the treatment of RB the focus
of his nearly forty years of ocular oncology practice, and is credited
with playing an important role in improving the survival rate of RB
patients. Dr. Ellsworth was born in eastern Pennsylvania, the son of a
physician, and attended Princeton University and Columbia College of
Physicians and Surgeons. He came under the tutelage of Dr. Algernon
B. Reese at the Edward S. Harkness Eye Institute at Columbia Pres-
byterian Medical Center, where he specialized in the treatment of eye
cancer. He moved to New York Hospital-Cornell Medical Center in
1979. All who attended his clinics were impressed with his effectiveness
as on ocular oncologist and compassion for his patients. He was a col-
orful, high-spirited man, and a loyal and caring friend.
At the time of Dr. Ellsworth’s death, external beam radiation was the
principal means of treatment of RB, and through it survival rates of
better than 90 percent were achieved.2 This treatment, however, was
associated with a 35 percent or higher risk of secondary cancers in
patients with bilateral RB during a 30-year period.3–5 Since Ellsworth’s
death, more extensive use of traditional chemotherapies have been
employed, but many of these are also mutagenic and may result in an
increased risk of secondary cancers.2,6 Hence, there remains a need for
improved methods of treatment for RB.
In 1966, Frederick C. Verhoeff suggested the possibility that RB cells
may be sensitive to vitamin D because this tumor sometimes under-
goes calcification and spontaneous regression.7 Due to the toxicity that
D.M. Albert et al.
 OH
Fig. 2. Vitamin D compounds and
OH
analogs.

HO OH
HO OH HO OH HO

Vitamin D2 Calcitriol (1,25-dihdroxy-D3) 16,23-D3 1a-hydroxy-D2

such treatment would have caused, and because there were no clinical
or experimental data at the time to indicate a role for vitamin D
therapy in the treatment of any human malignancy, implementation of
Verhoeff’s suggestion was deferred. Until the 1970’s, all therapeutic
studies involving RB were carried out in patients. With our establish-
ment of the Y-79 human cell line of RB in 1974, tissue culture studies
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

were utilized for drug response and other investigations.8 Hetero-

transplantation of RB into the athymic ‘nude’ mouse provided an
animal model that still is extremely useful.9 In 1990, we demonstrated
that transgenic mice expressing the SV40 large tumor antigen (Tag)
gene in the retina developed RB.10 This model resembles human RB in
its morphology and clinical behavior.11
The two predominant natural forms of vitamin D are vitamin D2
(ergocalciferol) and vitamin D3 (cholecalciferol). Vitamin D2 comes
strictly from diet, and vitamin D3 both from diet and synthesis in the
skin via a photochemical reaction.12 As vitamin D2 and D3 are hydrox-
For personal use only.

ylated, first in the liver and again in the kidney, they become more
potent.13 In our experiments, we used the hydroxylated form of vitamin
D3, calcitriol (1,25-dihydroxycholecalciferol); nonhydroxylated vitamin
D2 (ergocalciferol); a synthetic analog of calcitriol, 1,25-dihdyroxy-16-
ene-23-yne-vitamin D3 (16,23-D3); and a synthetic analog of vitamin D2,
1a-hydroxyvitamin D2 (1a-OH-D2) (Figure 2). The availability of the
Y-79 cell line and the xenograft and transgenic models of RB have
made possible the systematic study of these compounds as a treatment
for RB, and have made possible investigation of their mechanism of
action.

Materials and methods

Comment The description of methodology that follows, represents

that being currently used. These techniques are specific for data pre-
sented here for the first time, namely the demonstration of vitamin D
receptor mRNAs (VDR) as determined by PCR, and the characteri-
zation of the mechanism of the arrest of tumor cell growth. Some
details regarding methods used in the studies of effectiveness and
toxicity of vitamin D compounds (which have been carried out over a
fifteen-year period and are previously reported) may vary from those
given below. Readers are referred to the original references for exact
details.
All research utilizing mouse models of RB conformed to the guide-
lines set by the Research Animal Resources Center (RARC) of the
University of Wisconsin and the ARVO statement for the Use of

Treatment of retinoblastoma with vitamin D analogs 139

 Animals in Ophthalmic and Vision Research. Collection and investi-
gation of human tissue samples for vitamin D receptor analysis con-
formed to the guidelines of the Declaration of Helsinki and were
approved by the Human Subjects Committee office at the University
of Wisconsin.

sa mple acquisition for vitamin d receptor studies for

human and transgenic mouse rb Human RB samples were
collected from tumor-containing eyes following enucleation. Shortly
after removal, unfixed globes were transected and a representative
portion of tumor was frozen at -80° Centrigrade to await further analy-
sis. The remaining tumor sample was fixed and subjected to histopatho-
logic examination to confirm the clinical diagnosis of RB. Cultured Y-79
RB cells were also stored at -80° Centrigrade to be used in later
PCR experiments. In addition, intraocular tumors and extraorbital
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

(neck) metastases from LHb-Tag mice were removed and frozen.

Whole kidneys grossly free of tumor were obtained from the LHb-Tag
mice to serve as controls for the murine vitamin D receptor (VDR)
polymerase chain reaction (PCR) experiments. Lastly, HL60 human
leukemia cells were propagated in culture and harvested in order to
serve as a control for the human VDR PCR experiments.

reverse transcription pcr of vdr cdna Individual tumor,

tissue, and cell culture samples were thawed, and total RNA was iso-
lated using the method of Chomczynski and Sacchi.14 Contaminating
For personal use only.

genomic DNA was eliminated by incubating each sample with RNase-

free DNase I, followed by phenol : chloroform extraction and precipi-
tation of the remaining total RNA.
Complementary DNA (cDNA) was synthesized from messenger
RNA (mRNA) using reverse transcriptase along with random oligonu-
cleotide primers and the required deoxynucleosides triphosphates.
Afterwards, the samples were incubated with RNase to digest the RNA
templates.
Complementary DNA strands encoding the human or murine VDR
were specifically amplified using species-specific sets of oligonucleotide
primers recognizing either the human or mouse VDR coding sequence,
respectively. The sequences of the primers were as follows: human
forward primer, 5¢ T(856)GGACCTGTGGCAACCAAGA(875) 3¢;
human reverse primer, 5¢ G(958)TCCCACCTGGAACTTGATG(939)
3¢; murine forward primer, 5¢ T(949)GGGACTGTGGCAGC
CAAGA(968) 3¢; murine reverse primer, 5¢ G(1051)CCCCACCTG
GAACTTTATG(1032) 3¢. To control for genomic DNA contamina-
tion, these primer pairs were specifically chosen to amplify sequence
regions that span an intron. Amplification was carried out in a Perkin-
Elmer thermocycler in the presence of Taq DNA polymerase, the
aforementioned primers, and the deoxynucleoside triphosphates. The
predicted size of both the human and mouse VDR cDNA PCR
products in these experiments were 103 base pairs (bp), as analyzed
by agarose gel electrophoresis. To ensure that the observed 103 bp band
was derived from VDR mRNA, the amplified PCR product from a
single human RB sample was subcloned and sequenced.

140 D.M. Albert et al.

 cells a n d cell c ultur e Y-79 RB cells that had been maintained
in nude mice were cultured at 37° Centrigrade in an atmosphere of 5%
carbon dioxide and 95% air utilizing Ham’s F-12 media supplemented
with 15% fetal bovine serum (v/v) and were used to establish the Y-79
cell culture stocks. Details of culture techniques have been previously
described.8,15

vitamin d receptor assay for y79 rb cells Quantitative

measurement of VDR was made using a tritiated calcitriol binding
technique described elsewhere.15 Tritiated calcitriol was added to a Y-
79 RB cell suspension and the lysate (cytosol) separated from the cell
nuclear membranes. The specifically and non-specifically bound triti-
ated calcitriol was then measured. Scatchard analysis of the data was
performed and the dissociation constant of the calcitriol receptors and
the quantity of the receptors/cell calculated.15
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

xenograft model treatment Athymic ‘nude’ mice (four to six

weeks old) were injected subcutaneously with 1 ¥ 107 Y-79 human RB
cells suspended in a 1 : 1 of Iscove culture medium supplemented with
20% fetal bovine serum and basement membrane matrix suspension.
Five days after tumor injection, the mice were randomized into treat-
ment and control groups. Animals receiving calcitriol, vitamin D2, and
16,23-D3 received intraperitoneal injections of the vitamin D com-
pound in mineral oil or mineral oil alone adminstered intraperitoneally
five times a week. Animals receiving 1a-OH-D2 were given the com-
pound dissolved in 0.1 ml coconut oil by oral gavage while the control
For personal use only.

animals received coconut oil alone. Administration was continued for

five weeks, during which tumor size was measured using calipers, and
individual mouse weights were recorded three times/week. Toxicity was
assessed by survival, animal body weight, and clinical appearance. The
mice were then killed and the size, tumor volume, and tumor weight
were determined. Serum calcium was measured in blood taken from
the subclavian artery just prior to time of death. From representative
animals in each group, kidneys were evaluated for calcification. These
techniques are described in detail elsewhere.16,17

transgenic model treatment LHb-Tag mice, a well-

characterized model of RB,10 were used to determine drug effective-
ness in inhibiting tumor growth. Eight- to ten-week-old mice were
randomized by sex and litter into treatment groups and control groups.
The presence of the transgene was confirmed by PCR.10 Animals
receiving calcitriol, vitamin D2, and 16,23-D3 received intraperitoneal
injections of the vitamin D compound in mineral oil, while control
animals received intraperitoneal injections of mineral oil (vehicle)
alone.Treatment was administered five times per week.Animals receiv-
ing 1a-OH-D2 were given the compound dissolved in 0.1 ml coconut
oil by oral gavage, and control animals received coconut oil alone. The
treatment schedule was five weeks. Toxicity was assessed and serum
calcium determined in the manner described for the xenograft model.
At the termination of the treatment period, animals were killed and
the eyes were enucleated, fixed in 10% neutral buffered formalin, and
subjected to routine histological processing. Five-micron thick paraffin

Treatment of retinoblastoma with vitamin D analogs 141

 sections through the anatomical center of the globe (the pupil-optic
nerve plane) were stained with H&E and examined. The outline of the
tumor in each eye was traced from a digitized image. The area of the
trace was determined by Optimas software (Bioscan, Edmonds, WA),
and the average cross-sectional area for each eye was calculated (see
below). The histopathologic appearance of the tumors was character-
ized by scoring the degree of involvement of the retina and other ocular
structures, the severity of necrosis, degree of tumor differentiation,
number of mitotic figures, and degree of calcification. The extent and
character of the inflammation was also graded.18,19
The measure of tumor size used is the average cross-sectional tumor
area for both eyes. Typically, one-way ANOVA was used to assess the
effect of dose followed by paired t-tests when a significant effect was
found. Also typically, the tumor areas were transformed to the log scale
before analysis to obtain uniform variability. Please see specific manu-
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

scripts for exact methods.19–21

analysis of tumor cell death Histological methods of detect-

ing DNA fragmentation were used to identify dying cells in tumors.
Terminal transferase dUTP end labeling (TUNEL) was carried out as
described previously22–24 on paraffin sections of xenograft tumors har-
vested five weeks after treatment. Basically, the 3¢ ends of nicked DNA
were labeled by the addition of biotin-conjugated dUTP using Termi-
nal deoxynucleotide Transferase (TdT). Labeled DNA was identified
by reacting the sections with avidin peroxidase ABC kit (Vector Lab-
For personal use only.

oratories, Burlingame, CA) followed by staining with diaminobenzi-

dine. Since TUNEL labels degraded nuclei in both apoptotic and
necrotic nuclei, we also stained sections using the ligation method of
Didenko and Hornsby.25 This method specifically labels double-strand
DNA strand breaks that have 3¢ overhanging ends, typical of apoptotic,
but not necrotic cells. A 200 base pair DNA fragment, corresponding
to the multiple cloning site of the plasmid vector pBK-CMV (Strata-
gene, La Jolla, CA) was generated by Taq DNA polymerase and PCR
using Texas Red-conjugated nucleotides. DNA fragments generated by
Taq also contain 3¢ overhanging ends, which can be ligated to DNA
fragments in apoptotic nuclei. Labeled DNA was then layered over
tissue sections and ligated to nicked DNA using T4 DNA ligase enzyme
(Promega, Madison, WI). After extensive washing, unligated DNA was
removed and labeled nuclei were visualized using a Zeiss Axiophot flu-
orescent microscope. With each method, the tissues were viewed using
Nomarski interference optics and were not counter-stained.

immunohistochemistry Paraffin embedded tissue sections were

labeled with antibodies to human p53 (monoclonal DO-1, 0.8 mg/mL,
Oncogene Research Products, Boston, MA), human p21 (monoclonal
Ab-6, 0.5 mg/mL, Oncogene Research Products), human bax (poly-
clonal S-18, 0.5 mg/mL, Santa Cruz Biotechnology, Santa Cruz, CA)
and the Ki-67 antigen (monoclonal MIB-1, 2 mg/mL, Immunotech, Inc.,
Westbrook, ME). All antibodies were used with essentially the same
protocol described previously.24 Sections were deparaffinized and sub-
jected to antigen retrieval by incubating them in 100 mM TRIS buffer

142 D.M. Albert et al.

 Fig. 3. Two separate gels were run to
assay PCR amplification of vitamin
D receptor (VDR) cDNA that was
obtained from human and mouse
tissue samples. Lane 1 contains VDR
mRNA from a representative section
of fresh human retinoblastoma. Lane
2 contains mRNA from human
retinoblastoma cell line Y-79. Lane 3
(pH 9.5) at 90°C for 10 minutes (p21, bax, and MIB-1) or 30 minutes contains mRNA from cultured HL60
(p53), followed by 60 minutes in the same buffer at room temperature. human leukemia cells. Lane 4
After incubation with primary antibody, the sections were washed and contains a representative sample of
incubated with appropriate biotinylated secondary antibody and murine intraocular retinoblastoma.
stained with the avidin peroxidase ABC kit. Sections were viewed with Lane 5 shows a predicted 103 bp
Nomarski optics and not counterstained. band for extraorbital metastasis in
the mouse model. Lane 6 contains a
representative sample of mouse
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

Results
kidney, which is known to express
VDR mRNA.
the presence of vitamin d receptors in y-79 rb cells In
studies carried out in 1988,15 Scatchard analysis of the receptor assay
showed a concentration of 94 fmol of calcitriol receptors per one
million Y-79 RB cells, or 56,000 receptors/cell. These receptors have a
dissocation constant of 1.18 nmol/liter.

pcr amplification of vdr cdna from human rb samples A

total of 23 RB samples taken from the enucleated eyes of different
For personal use only.

patients were analyzed by PCR for the original presence of VDR

mRNA. In every sample, examination of the PCR product by agarose
gel electrophoresis revealed a predicted 103 bp band corresponding to
a targeted portion of the human VDR coding sequence (representa-
tive sample shown in Figure 3, lane 1). The PCR primer pairs used in
both the human and mouse experiments span an intron in order to
control for residual genomic DNA. Using the same oligonucleotide
primers, an identical 103 bp band was amplified from Y-79 RB (Figure
3, lane 2) and cultured HL60 human leukemia cells, a cell line known
to express VDR26,27 (Figure 3, lane 3). DNA sequencing of the PCR
product from one human RB sample confirmed that the 103 bp band
encoded the expected VDR region. While all 23 samples were positive
in this study, there remains a possibility that a future sample may not
be positive. This potential variability is reflected as a 95% confidence
interval of (0.85, 1) for the estimated probability of any RB sample
being positive. This interval is calculated assuming binomially dis-
tributed outcomes.

pcr amplification of vdr cdna from a mouse model of

retinoblastoma Retinoblastoma samples from five LHb-Tag
mice, along with one metastatic tumor sample and one whole kidney,
were processed and analyzed essentially as described for the human
RB samples. Mouse renal tissue is known to express VDR,28 and there-
fore serves as a positive control for these experiments. A predicted
103 bp band was visualized following agarose gel electrophoresis for all
samples containing intraocular RB (representative sample shown in

Treatment of retinoblastoma with vitamin D analogs 143

 Fig. 4. Relative counts of
retinoblastoma cells expressed as
percentage of control; 10–9 mol/L of
calcitriol caused a 15% decrease in
cell growth by day 9.
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

Figure 3, lane 4), extraorbital metastasis (Figure 3, lane 5), and whole
kidney (Figure 3, lane 6). Analogous to the human samples, the size of
the band corresponds to a portion of the murine VDR coding sequence
targeted by the murine-specific oligonucleotide primers used during
PCR.

in vitro effects of calcitriol Concentrations of 10-9 and

For personal use only.

10-6 mol/L of calcitriol were effective at inhibiting the growth of the RB

cells in vitro (Figure 4).15 This inhibition was statistically significant at
10-6 mol/L of calcitriol on day 6 and 10-9 mol/L of calcitriol on day 9 (p
< 0.05). The 10-12 mol/L of calcitriol did not statistically significantly
inhibit cell growth. The inhibition of cell growth was dose-dependent
on days 3 and 6.

effectiveness and toxicity of ergocalciferol in the

athymic xenograft model Mice were treated with a high dose
(7.8 mg/kg) of ergocalciferol and a low dose (2.8 mg/kg) of ergocalcif-
erol. The geometric mean diameter (the cube root of length ¥ width ¥
height) was measured 33 days after treatment was initiated. The effect
of ergocalciferol on tumor size is shown in Figure 5a. The survival data
is summarized in Figure 5b. Histopathologic examination showed a
dose-dependent increase in tumor necrosis and calcification in the
treated animals. A more detailed description of the results has been
published.18

effectiveness and toxicity of calcitriol treatment in

the athymic mouse xenograft model In this experiment, the
treated group received 500 ng/kg of calcitriol. Tumor size was judged
on the basis of geometric mean diameter (Figure 6a). The results after
30 days of treatment are shown in Figure 6b. Histopathologic exami-
nation showed no difference in calcification or necrosis between the
control tumors and those in the treated animals. Further details regard-
ing these results can be found in the original report.29

144 D.M. Albert et al.

 9 Fig. 5A. Effect of ergocalciferol on
8.2
GEOMETRIC MEAN DIAMETER

growth of Y-79 retinoblastoma

8
xenografts after 28 days of
7 treatment. The low dose is most
6 effective with p < 0.05 compared to
calcitriol. Tumors were measured in
5 terms of geometric mean diameter.
4 3.5
2.8
3

0
Control High Dose (7.8 mg/kg) Low Dose (2.8 mg/kg)
(42% of control) (34% of control)
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

100
100

90

80
PERCENT SURVIVAL

70

60
50 46
For personal use only.

40

30 25

20
10 Fig. 5B. Survival of control and
0 ergocalciferol-treated Y-79
Control (15/15 surviving) High Dose 7.8 mg/kg (3/12 Low Dose 2.8 mg/kg (6/13 retinoblastoma xenograft mice after
surviving) surviving) 33 days of treatment.

effectiveness and toxicity of calcitriol in the lh b-tag

transgenic model In studying the antineoplastic effect of cal-
citriol in LHb-Tag mice, toxicity studies were done using calcitriol doses
of 0.05 mg, 0.1 mg, and 0.2 mg that were compared to controls. In the dose-
response studies, a high dose (0.05 mg) and a low dose (0.025 mg) were
administered. Tumor size was measured by the mean largest cross-
sectional area. The results after five weeks of treatment are shown in
Figure 7a. The combined survival results for both the dose-response
and toxicity studies following five weeks of treatment are summarized
in Figure 7b.
Histopathologic studies showed no increased calcification in the
tumors of treated animals. There was, however, greater cell death in the
tumors in the low dose group (p < 0.02). More differentiation was noted
in the low dose group (p < 0.003), which was manifested by increased
numbers of Homer-Wright rosettes and a smaller proportion of the
tumor appearing undifferentiated. Additional information regarding
these results is given in the original article.19

Treatment of retinoblastoma with vitamin D analogs 145

 Fig. 6A. Effect of ergocalciferol on 30
growth of Y-79 retinoblastoma

MEAN GEOMETRIC DIAMETER

26
xenografts after 28 days of 25
treatment. The low dose is most
effective with p < 0.05 compared to
20
calcitriol. Tumors were measured in
terms of geometric mean diameter.
15

10

5
5

0
Control 500 ng/kg (19% of control tumor size)
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

100
90
90

80
PERCENT SURVIVAL

70

60

50
40
40
For personal use only.

30

20
Fig. 6B. Survival of control and
10
ergocalciferol-treated Y-79
retinoblastoma xenograft mice after 0
33 days of treatment. Control (9/10 surviving) 500 ng/kg (4/10 surviving)

effectiveness and toxicity of 16,23-d 3 in the athymic

xenograft mouse model Preliminary experiments determined
the most efficacious dose of 16,23-D3 to be 0.5 mg. The effectiveness
and toxicity of this drug was evaluated using a negative control that
received mineral oil alone, and a positive control that received 0.5 mg
of calcitriol. Results of tumor size (tumor volume) are given as of the
32nd day of treatment, because after that day there were insufficient
calcitriol-treated mice surviving to permit analysis. The results after five
weeks of treatment are compared in Figure 8a. Survival for the three
groups following five weeks of treatment is shown in Figure 8b. On
histopathologic examination, no qualitative morphologic difference
was found among the three groups with regard to calcification, necro-
sis, or differentiation. A more detailed description of these results can
be found in the original report.16

the effectiveness and toxicity of 16,23-d 3 in the lhb-

tag transgenic mouse model In an initial study of the effec-
tiveness of 16,23-D3,20 an extremely low dose of this vitamin D analog

146 D.M. Albert et al.

 Fig. 7A. Effect of calcitriol on
MEAN CROSS-SECTIONAL AREA (MM3)
8
growth of LHb-Tag transgenic
6.9 retinoblastomas after 5 weeks of
7
treatment. The difference in tumor
6 size was significant between the high
5 dose and control (p < 0.008) and the
5
low dose and control (p < 0.014).
4
3
3

0
Control High Dose (0.05 ug) 43% Low Dose (0.025 ug) 72%
of control tumor of control tumor
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

100
100 93
90 83
80
PERCENT SURVIVAL

70
60
50 50
50

40
For personal use only.

30
20

10

0 Fig. 7B. Survival of control and

Control (19/19 0.025 ug (14/15 0.05 ug (19/23 0.1 ug (5/10 0.2 ug (4/8 calcitriol treated LHb-Tag transgenic
surviving) surviving) surviving) surviving) surviving) mice after 5 weeks of treatment.

was compared to a mineral oil-treated control. The results are

shown in Figure 9a. All animals survived the five-week treatment
schedule.
Subsequent toxicity studies indicated that considerably larger doses
of 16,23-D3 were well tolerated by the transgenic mice.21 Consequently,
a second experiment was performed in which the treatment arms
included 0.35 mg, 0.5 mg, and 0.75 mg doses. These treatment groups were
compared to a control group receiving only mineral oil. Tumor size was
measured as the average of the square root of the right and left eye
tumor area. The results are presented in Figure 9b. Survival data is
shown in Figure 9c. On histological examination of the tumors, no dif-
ference in appearance was found between the control and treated
animals. Further details of results can be found in the original
reports.20,21

the effectiveness and toxicity of 1 a-oh-d 2 in the

athymic xenograft mouse model Following five weeks of

Treatment of retinoblastoma with vitamin D analogs 147

 Fig. 8A. Effect of 16,23-D3 on 4
growth of Y-79 retinoblastoma 3.45
3.5
xenografts compared to control and

TUMOR VOLUME (CM3)

calcitriol following 5 weeks of 3
treatment. 16,23-D3-treated mice had
significantly smaller tumor size than 2.5
controls (p = 0.02) but were not
2
significantly different from the 1.55
calcitriol-treated mice. 1.5 1.26

0.5

0
Control Calcitriol (0.05 ug) 37% of 16,23-D3 (0.5 ug) 52% of
control control
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

100
100
90
90
PERCENT SURVIVAL

80
70
60
50
40
40
For personal use only.

30
20
Fig. 8B. Survival of 16,23-D3-treated
Y-79 retinoblastoma xenograft mice 10
compared to control and calcitriol- 0
treated mice after 5 weeks of Control Mice (10/10 Calcitriol (0.05 ug) (4/10 16,23-D3 (0.5 ug) (9/10
treatment. surviving) surviving) surviving)

treatment with 1a-OH-D2, the size of tumors in animals receiving

this analog in 0.1 mg, 0.2 mg, 0.3 mg, and 0.6 mg doses were compared
to controls that received coconut oil via oral gavage. The results
are shown in Figure 10a. Survival data is summarized in Figure
10b.17

the effectiveness and toxicity of 1 a-oh-d 2 in the lhb-

tag transgenic mouse model These studies are now being
completed. Results thus far indicate this analog to be similar in effec-
tiveness and toxicity to that of 16,23-D3 (Dawson et al., manuscript in
preparation).

characterization of the mechanism of arrest of tumor

growth by vitamin d analogs Specimens of Y-79 xenografts
taken from athymic mice treated for five weeks with either 16,23-D3 or
calcitriol were compared with tumors from control animals at the same
stage. Paraffin sections of tumors (two mice per treatment) were ana-
lyzed for cell death using Terminal Transferase dUTP-nick end label-

148 D.M. Albert et al.

 Fig. 9A. Effect of an extremely low
AVG CROSS-SECTIONALAREA MM3 dose (0.05 mg) of 16,23-D3 on growth
1.12
of LHb-Tag transgenic
1.2
retinoblastoma after 5 weeks of
1 0.88 treatment. The difference in tumor
0.8 size was significant (p = 0.02).
0.6

0.4

0.2

0
Control 0.05 ug16,23-D3 (79% of control
size)

500
AVERAGE TUMOR AREA*

457
450
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

400
350 318 Fig. 9B. Effect of 16,23-D3 on
300 263 growth of LHb-Tag transgenic
250 206 retinoblastoma after 5 weeks of
200 treatment. The difference between
150 the 0.35 mg treated group and the
100 controls was statistically significant
50 (p = 0.0056). *Tumor size was
0 calculated as the average of the
Control 0.75 ug (70% 0.5 ug (58% 0.35 ug (45% square root of the right and left eye
of control) of control) of control) tumor area.
For personal use only.

PERCENT SURVIVAL

95
100 89 89 89

80
60
40
20
0
Control 0.75 ug 0.5 ug (16/18 0.35 ug Fig. 9C. Survival of control and
(17/18 (16/18 surviving) (16/18 16,23-D3 treated LHb-Tag transgenic
surviving) surviving) surviving) mice after 5 weeks of treatment.

ing (TUNEL) and cell proliferation using the monoclonal antibody

MIB-1. Figure 11 (sections A–C) shows that control tumors and tumors
treated with calcitriol exhibit some TUNEL-labeling (approximately
0.13 to 0.64% of the cells, respectively), but the level of cell death is
more than six times greater in tumors treated with 16,23-D3 (approxi-
mately 4.2% of the cells). In addition to an examination of the rate of
cell death in these tumors, we also looked for evidence of the type of
cell death (apoptosis or necrosis). The morphology of nuclear frag-
mentation was examined from sections of tumors stained with hema-
toxylin and eosin (Figure 11, panel D). All tumors, but especially those
treated with 16,23-D3, contained highly pyknotic and fragmented
nuclei, both morphologically consistent with apoptotic cell death.30,31

Treatment of retinoblastoma with vitamin D analogs 149

 Fig. 10A. Effect of 1a-OH-D2 on 3
2.71
growth of Y-79 retinoblastoma
xenografts compared to controls 2.5

TUMOR VOLUME (CM3)

following 5 weeks of treatment. The
0.3 mg group and 0.2 mg group were 2 1.77
statistically significantly smaller than
the control group (p < 0.003 and p <
1.5
0.004, respectively). 1.18 1.22
0.99
1

0.5

0
Control 0.1 ug (65% of 0.2 ug (37% of 0.3 ug (44% of 0.6 ug (45% of
control) control) control) control)
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

100

90 84
PERCENT SURVIVAL

80
68
70 65
61
60
47
50
40
For personal use only.

30
20

Fig. 10B. Survival of 1a-OH-D2 10

treated xenograft mice compared to 0
control mice after 5 weeks of Control 0.1 ug (25/37 0.2 ug (39/60 0.3 ug (34/55 0.6 ug (21/45
treatment. surviving) surviving) surviving) surviving)

Dying cells in these tumors also stained using the 3¢ Overhang ligation
technique (Figure 11, panel E). In addition to these results obtained
with Y-79 xenografts, similar features of apoptotic cell death were
observed in dying cells found in spontaneous tumors of transgenic mice
treated with vitamin D analogs (data not shown). Thus, it is important
to note that vitamin D-induced cell death in these tumors appears to
be apoptotic and therefore controlled by genes being expressed in the
dying cells.
Conversely, all three tumor specimens studied had relatively uniform
labeling with the MIB-1 antibody (Figure 12 – 2.3%, 2.6%, and 3.3%
for control, 16,23-D3 and calcitriol, respectively) suggesting that the
rate of cell proliferation is unaffected. The combined observations for
tumors treated with calcitriol suggest that the mechanism of tumor
growth arrest for this analog is also mediated by an increase in cell
death. It is possible, however, that the time point examined in this pre-
liminary study was too late to observe the period of peak TUNEL
activity for the calcitriol treatment.

150 D.M. Albert et al.

 Fig. 11. Cell death in Y79 xenograft
tumors after 5 weeks of vitamin D
treatment demonstrated by TUNEL
staining in (A) untreated tumors, (B)
tumors treated with 16,23-D3, and
(C) tumors treated with calcitriol.
Panels A, B, and C exhibit signs of
active cell death. 16,23-D3 tumors
had TUNEL staining which was six
times greater than control or
calcitriol tumors at this stage. (D)
Higher magnification of an H&E
stained section of a 16,23-D3-treated
tumor. Nuclear morphology of the
cells is consistent with apoptotic cell
Morphological and histochemical methods suggest that vitamin D death, including highly pyknotic
nuclei and cells with extensive
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

analogs activate apoptotic cell death in retinoblastoma. Since this

nuclear fragmentation. (E)
mechanism of cell death is genetically regulated, the expression pat-
Fluorescence micrograph of dying
terns of three genes associated with cell death were investigated by
cells in a tumor treated with 16,23-
immunohistochemistry in sections of five-week-treated tumors. The D3 and labeled using the 3¢ overhang
gene products examined included the tumor suppressor protein p53, ligation technique (Didenko and
and the gene products of two genes regulated by p53 (p21waf-1/cip-1 and Hornsby, 1996). DNA fragments are
bax). apoptotic, but not necrotic, nuclei
Nuclear p53 staining was present in all three treatment groups, containing a high proportion of
although it was most abundant in tumors treated with 16,23-D3. Simi- similar 3¢ overhanging nucleotides
larly, both p21 and Bax immunoreactivity were strongest in the 16,23- can be covalently linked to biotin-
For personal use only.

D3 groups. Since the expression of p53, and its downstream target labeled DNA. Ligated DNA is
genes, correlates with retinoblastoma cell death,24 it is not surprising to visualized using streptavidin
conjugated to Texas Red.
find elevated expression levels in the treatment group undergoing the
Morphological evidence combined
greatest rate of cell death.
with this labeling technique indicates
that these tumor cells are dying by
Discussion As noted above, alternatives to the current methods of apoptosis.
RB therapies are needed.32 Vitamin D analogs hold the promise of ful-
filling this need.We initially studied ergocalciferol and calcitriol (Figure
2) in the athymic Y-79 RB xenograft mouse model. Although these
compounds showed impressive reductions in tumor growth as com-
pared to controls, the doses required for this effect caused mortality
with rates ranging from 25% to 46%.19,29 It should be noted, however,
that the immunocompromised athymic mouse model has an extremely
high sensitivity to calcitriol, vitamin D2, 16,23-D3, and 1a-OH-D2, and
that immunocompetent mice, such as the transgenic strain, are a better
indicator of actual drug toxicity. In an experiment in which calcitriol
was administered to athymic mice in a dose required to reduce tumor
growth to 36% of the control (i.e., 0.05 mg), there was still only a 40%
survival rate of treated animals.16 By withholding doses of ergocalcif-
erol from sick animals, the survival rate could be improved to 75%.19
In all of the experiments, the toxicity appeared to be related to the
hypercalcemia induced by these compounds. Consequently, we con-
cluded that ergocalciferol and calcitriol were excessively toxic, and
were not suitable for treatment of children with RB.
We then turned our attention to a synthetic analog of calcitriol, 16,23-
D3 (Figure 2) which has an antineoplastic effect similar to calcitriol but

Treatment of retinoblastoma with vitamin D analogs 151

 Fig. 12. Immunohistochemical
labeling of Y-79 tumors grown in
athymic mice after five weeks of
vitamin D treatment. Sections of
untreated tumors, 16,23-D3-treated
tumors, and calcitriol-treated tumors
were stained with the following
proteins: p53 (monoclonal DO-1),
p21 (monoclonal Ab-6), bax
(polyclonal Ab-1), and antibodies
directed against the Ki-67 nuclear
antigen (MIB-1). Each panel is a
Normarski interference image taken
near the center of the tumors
because they exhibited uniform cell
density. p53 and p53-regulated genes
(p21 and bax) are upregulated in
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

tumors treated with 16,23-D3. This

change in gene expression mirrors
the increased apoptotic activity seen
in Figure 10. MIB-1 staining is
relatively equal among all the
treatments suggesting that cell
proliferation is unaffected.
For personal use only.
with less hypercalcemic activity. Pure crystalline 16,23-D3 was provided
by Dr. Milan Uskokovich (Hoffman-LaRoche, Nutley, NJ) and by Ilex
Corporation (San Antonio, Texas). In our initial experiments with this
compound, animals treated with 0.05 mg of 16,23-D3 were found to have
a significantly smaller tumor cross-sectional area when compared to
corresponding controls (p = 0.02).20 All animals survived five weeks of

treatment. In subsequent studies using higher doses (0.2 mg to 0.75 mg),

the drug was effective in reducing tumor growth and the survival rate
was approximately 90%.16,21 This appears to be a promising drug for
use in treatment of RB children. It was not approved by the Food &
Drug Administration (FDA) until March 1999 for investigational use
in human cancer patients.This led us to investigate an analog of vitamin
D2, 1a-OH-D2.
1a-OH-D2 (BoneCare International, Inc.) is a vitamin D analog that
was approved by the FDA in 1999 for oral use in the treatment of sec-
ondary hyperparathyroidism due to renal failure. This compound was
approved earlier for investigational use in human tumor treatment in
1996 and is currently being used in Phase 2 human clinical trials of
prostate cancer (George Wilding, personal communication). Similar to
16,23-D3, 1a-OH-D2 is known to induce low levels of hypercalcemia
while providing effective systemic serum drug levels for tumor treat-
ment.33,34 In a recent study,17 we found 1a-OH-D2 limited tumor growth
of the human Y-79 RB cell line that was subcutaneously injected in
athymic ‘nude’ mice. We reported a dose-response efficacy curve with
minimal toxicity. Our results indicate that 1a-OH-D2, a compound not
known to be a mutagen, can limit tumor growth in a non-toxic dose
range. In comparing this study17 to a similar study of 16,23-D3 and
calcitriol in same mouse model,16 1a-OH-D2 appears to be similar in
tumor reduction capability at a 0.3 mg dose level when compared to
0.5 mg of 16,23-D3 or 0.05 mg of calcitriol.
It is probable that tumor cells must contain VDR in order for vitamin
D analogs to be effective in limiting their growth. In the present report,

152 D.M. Albert et al.

 we describe examination of 23 consecutively received RB samples with
PCR amplification for the presence of VDR mRNA. Message encod-
ing receptors were present in all specimens, providing convincing evi-
dence that these tumor cells express the receptor protein. The 95%
confidence interval for the probability of any RB being positive is (0.85,
1), based on the binomial distribution.
Calcitriol and its analogs have been demonstrated to inhibit cellular
proliferation in other malignant cell lines besides RB, including
leukemic, breast, colon, renal, and lung carcinomas.35–38 It is hypothe-
sized that the antiproliferative effect of these compounds is mediated
by a vitamin D receptor-linked mechanism, although exceptions
exist.39–42 Considerable evidence exists that the antineoplastic and
differentiating effects of vitamin D compounds affect fundamental
cellular processes of proliferation, differentiation, and apoptosis. The
resulting key biochemical events are related to activation of cyclin-
dependent kinase inhibitors, such as p21waf-1/cip-1, and there are some
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14

reports that activated VDR can directly mediate the expression of this
gene.43–47
In the present report, we present data showing that tumor growth
attenuation of Y-79 xenografts in athymic mice following treatment
with 16,23-D3 is due to apoptotic cell death. Calcitriol and vitamin D
analogs can also induce apoptosis in leukemic (HL60) cells as well as
human breast cancer and colon cancer cell lines.48–50 Recent studies
have shown that human RB and RB cell lines are extremely suscepti-
ble to p53-mediated apoptosis and elevated p21waf-1/cip-1 expression.24,47
For personal use only.

The results shown in this report indicated that the treatment of

xenograft tumors elicits the increased expression of both p53 and p53-
regulated genes. These data are consistent with the role of bound VDR
in stimulating the activation of apoptotic genes and may account for
the decreased toxicity of these compounds without compromised effi-
cacy. It is likely that 1a-OH-D2 has a similar mechanism of action
against RB tumors in the athymic mouse.

References 5 Abramson DH, Frank CM. Second

1 Jackson Coleman D. RM Ellsworth
obituary. In: New York Times. Jan. 5,
1994.
2 Ferris FL, Chew EY. A new era for
the treatment of RB. Arch
Ophthalmol. 1996;114:1412.
3 Eng C, Li FP, Abramson DH,
Ellsworth RM, Wong FL, Goldman
MB, Seddon J, Tarbell N, Boice Jr.
JD. Mortality from second tumors
among the long-term survivors of
RB. J Natl Cancer Inst. 1993;85:
1121–1128.
4 Roarty JD, McLean IW,
Zimmerman LE. Incidence of
second neoplasms in patients with
bilateral RB. Ophthalmology. 1988;
95:1583–1587.
nonocular tumors in survivors of
bilateral RB. A possible age effect
on radiation-related risk.
Ophthalmology. 1998;105:573–580.
6 Gallie BL, Budning A, DeBoer G,
Thiessen JJ, Koren G, Verjee Z, Ling
V, Chang HS-L. Chemotherapy with
focal therapy can cure intraocular
RB without radiotherapy. Arch
Ophthalmol. 1996;114:1321–1328.
7 Verhoeff R. RB undergoing
spontaneous regression: Calcifying
agents suggested in treatment of
RB. Am J Ophthalmol. 1966;62:
573–574.
8 Reid TW, Albert DM, Rabson AS,
Russell P, Craft J, Chu ES, Tralka TS,
Wilcox J. Characteristics of an

Treatment of retinoblastoma with vitamin D analogs 153

 established cell line of RB. J Natl
Cancer Inst. 1974;53:347–360.
9 Gallie BL, Albert DM, Wong JJ,
Buyukmihci N, Puliafito CA.
Heterotransplantation of RB into
the athymic ‘nude’ mouse. Invest
Ophthalmol Vis Sci. 1977;16:256–299.
10 Windle JJ, Albert DM, O’Brien JM,
Marcus DM, Disteche CM, Bernards
R, Mellon PL. RB in transgenic
mice. Nature. 1990;343:665–669.
11 O’Brien JM, Marcus DM, Bernards
R, Carpenter JL, Windle JJ, Mellon
P, Albert DM. A transgenic mouse
model for RB. Arch Ophthalmol.
1990;108:1145–1151.
12 Koshy KT. Vitamin D: An Update. J
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14
Pharm Sci. 1982;71:137–152.
13 Jones G, Byrnes B, Palma F, Segev
D, Mazur Y. Displacement potency
of vitamin D2 analogs in competitic
protein-binding assays for 25-
hydroyxvitamin D3, 24,25-
dihydroyvitamin D3, and
1,25-idhydroxyvitamin D3. J Clin
Endocrinol Metab. 1980;50:773–775.
14 Chomczynski P, Sacchi N. Single-
For personal use only.
step method of RNA isolation by
acid guanidinium thiocyanate-
phenol-chloroform extraction. Anal
Biochem. 1987;162:156–159.
15 Saulenas A, Cohen S, Key L, Winter
C, Albert DM. Vitamin D and RB:
The presence of receptors and
inhibition of growth in vitro. Arch
Ophthalmol. 1988;106:533–535.
16 Sabet SJ, Darjatmoko SR,
Lindstrom MJ, Albert DM.
Antineoplastic effect and toxicity of
1,24-dihydroxy-16-ene-23-yne-
vitamin D3 in athymic mice with Y-
79 human RB tumors. Arch
Ophthalmol. 1999;117:365–370.
17 Grostern RJ, Bryar PJ, Zimbric ML,
Darjatmoko SR, Lissauer BJ,
Lindstrom MJ, Lokken JM,
Strugnell SA, Albert DM. Toxicity
and dose-response studies of 1a-
hydroxyvitamin D2 in a
retinoblastoma xenograft model.
Arch Ophthalmol. 2002, in press.
18 Albert DM, Saulenas AM, Cohen
SM. Verhoeff’s query: Is vitamin D
effective against RB? Arch
Ophthalmol. 1988;106:536–540.
154 D.M. Albert et al.
19 Albert DM, Marcus DM, Gallo JP,
O’Brien JM. The antineoplastic
effect of vitamin D in transgenic
mice with RB. Invest Ophthalmol Vis
Sci. 1992;33:2354–2364.
20 Shternfeld IS, Lasudry JGH,
Chappell RJ, Darjatmoko SR,
Albert DM. Antineoplastic effect of
1,25-dihydroxy-16-ene-23-yne-
Vitamin D3 analogue in transgenic
mice with RB. Arch Ophthalmol.
1996;114:1396–1401.
21 Wilkerson CL, Darjatmoko SR,
Lindstrom MJ, Albert DM. Toxicity
and dose-response studies of 1,25-
(OH)2-16-ene-23-yne vitamin D3 in
transgenic mice. Clin Cancer
Research. 1998;4:2253–2256.
22 Gavrieli Y, Sherman Y, Ben-Sasson
SA. Identification of programmed
cell death in situ via specific labeling
of nuclear DNA fragmentation. J
Cell Biol. 1992;119:493–501.
23 Quigley HA, Nickells RW, Kerrigan
La, Pease ME, Thibault DJ, Zack
DJ. Retinal ganglion cell death in
experimental glaucoma and after
axotomy occurs by apoptosis. Invest
Ophthalmol Vis Sci. 1995;36:774–786.
24 Nork TM, Poulsen G, Millecchia LL,
Jantz RG, Nickells RW. p53
regulates apoptosis in human RB.
Arch Ophthalmol. 1997;115:213–219.
25 Didenko VV, Hornsby PJ. Presence
of double-strand breaks with single-
ase 3¢ overhangs in cells undergoing
apoptosis but not necrosis. J Cell
Biol. 1996;135:1369–1376.
26 Reitsma PH, Rothberg PG, Astrin
SM, Trial J, Bar-Shavit Z, Hall A,
Teitelbaum SL, Kahn AJ. Regulation
of myc gene expression in HL-60
leukaemia cells by a vitamin D
metabolite. Nature. 1983;306:
492–494.
27 Mangelsdorf DJ, Koeffler HP,
Donaldson CA, Pike JW, Haussler
MR. 1,25-dihydroxyvitamin D3-
induced differentiation in a human
promyelocytic leukemia cell line
(HL-60): Receptor-mediated
maturation to macrophage-like cells.
J Cell Biol. 1984;98:391–398.
28 Colston RW, Feldman D.
Demonstration of a 1,25-
 dihydroxycholecalciferol cytoplasmic
receptor-like binder in mouse
kidney. J Clin Endocrinol Metab.
1979;49:798–800.
29 Cohen SM, Saulenas AM, Sullivan
CR, Albert DM. Further studies on
the effect of vitamin D on RB. Arch
Ophthalmol. 1988;106:541–543.
30 Kerr JFR, Wylli AH, Currie AR.
Apoptosis: A basic biological
phenomenon with wide-ranging
implications in tissue kinetics. J Cell
Biol. 1972;26:239–257.
31 Nickells RW, Zack DJ. Apoptosis
in ocular disease: A molecular
overview. Ophthalmic Genetics.
1996;17:145–165.
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14
32 O’Brien JM. Alternative treatment
in RB. Ophthalmology.
1998;105:571–572.
33 Knutson JC, Hollis BW, LeVan LW,
Valliere C, Gould KG, Bishop CW.
Metabolism of 1a-hydroxyvitamin
D2 to activated dihydroxyvitamin D2
metabolites decreases endogenous
1a-dihydroxyvitamin D3 in rats and
monkeys. Endocrinology. 1995;136:
For personal use only.
4749–4753.
34 Strugnell S, Byford V, Makin HLJ,
Moraiarty RM, Gilardi R, LeVan
LW, Knutson JC, Bishop CW, Jones
G. 1a,24S-dihydroxyvitamin D2: A
biologically active product of 1a-
hydroxyvitamin D2 made in the
human hepatoma, HEP3B. Biochem
J. 1995;310:233–241.
35 Walters M. Newly identified actions
of the vitamin D endocrine system.
Endocr Rev. 1992;13:719–764.
36 Brenner RV, Shabahang M,
Schumacker LM, Nauta RJ,
Uskokovic MR, Evans SR, Buras
RR. The antiproliferative effect of
vitamin D analogs on MCF-7 human
breast cancer cells. Cancer Lett.
1995;92:77–82.
37 Schwartz GG, Oeler TA, Uskokovic
MR, Bahnson RR. Human prostate
cancer cells: Inhibition of
proliferation by vitamin D analogs.
Anticancer Res. 1994;14:1077–1082.
38 Pakkala S, deVos S, Elstner E, Rude
RK, Uskokovic M, Binderup L,
Koeffler HP. Vitamin D3 analogs:
Effect on leukemic clonal growth,
Treatment of retinoblastoma with vitamin D analogs
differentiation, and serum calcium.
Leuk Res. 1995;19:65–72.
39 Inaba M, Okuno S, Koyama H,
Nishizawa Y, Morii H. Dibutyryl
cAMP enhances the effect of 1,25-
dihydroxyvitamin D3 on a human
promyelocytic leukemia cell, HL-60,
at both the receptor and the
postreceptor steps. Arch Biochem
Biophys. 1992;29:181–186.
40 Kawa S, Yoshizawa K, Tokoo M,
Imai H, Oguchi H, Kiyosawa K,
Homma T, Nikaido T, Furihata K.
Inhibitory effect of 22-oxa-1,25-
dihydroxyvitamin D3 on the
proliferation of pancreatic cancer
cell lines. Gastroenterology.
1996;110:1605–1613.
41 Bhatia M, Kirkland JB, Meekling-
Gill KA. Monocytic differentiation
of acute promyelocytic leukemia
cells in response to 1,25-
dihydroxyvitamin D3 is independent
of nuclear receptor binding.
J Biol Chem. 1995;270:15.962–
15.965.
42 Brown AJ, Dusso A, Slatopolsky E.
Selective vitamin D analogs and
their therapeutic applications. Semin
Nephrol. 1974;14:156–174.
43 Frey MR, Zhao X, Evans SS, Black
JD. Activation of protein kinase C
(PKC) isozymes inhibits cell cycle
progression, modulates
phosphorylation of the RBs (Rb)
protein, and induces expression of
p21waf1/cip1 in intestinal epithelial
cells. Proc Am Assoc Cancer Res.
1996;37:51.
44 Wali R, Bissonnette M, Khare S,
Aquino B, Niedziela S, Sitrin M,
Brasitus T. Protein kinase C
isoforms in the chemopreventive
effects of a novel vitamin D3
analogue in rat colonic
tumorigenesis. Gastroenterology.
1996;111:118–126.
45 Millard SS, Koff A. Cyclin-
dependent kinase inhibitors
in restriction point control,
genomic stability, and tumorigenesis.
J Cell Biochem. 1998;30/31:37–
42.
46 Brugarolas JCC, Gordon JI, Beach
D, Jacks T, Hannon CJ. Radiation-
155
 induced cell cycle arrest
compromised by p21 deficiency.
Nature. 1995;377:552–557.
47 Kondo Y, Kondo S, Liu JB, Haqqi T,
Barnett GH, Barna BP. Involvement
of p53 and waf1/cip1 in gamma-
irradiation-induced apoptosis of
retinoblastoma cells. Exp Cell Res.
1997;251:51–56.
48 Elstner E, Linker-Israeli M, Umiel
T, Le J, Grillier I, Said J, Shintau IP,
Krajewski S, Reed JC, Binderup L,
Koeffler HP. Combination of a
potent 20-epi-vitamin D3 analogue
(KH 1060) with 9-cis-retinoic acid
irreversibly inhibits clonal growth,
Ophthalmic Genet Downloaded from informahealthcare.com by University of Bristol on 11/06/14
For personal use only.
156 D.M. Albert et al.
decreases bcl-2 expression, and
induces apoptosis in HL-60
leukemic cells. Cancer Res.
1996;56:3570–3576.
49 Vandewalle B, Hornez L, Wattez N,
Revillion F, Lefebvre J. Vitamin-D3
derivatives and breast-tumor cell
growth: Effect on intracellular
calcium and apoptosis. Int J Cancer.
1995;61:806–811.
50 Skarosi S, Abraham C, Bissonnette
M, Scaglione-Sewell B, Sitrin MD,
Brastius TA. 1,25-dihydroxyvitamin
D3 stimulates apoptosis in CaCo-2
cells. Gastroenterology.
1997;112:A608.