Angiogenin Is Translocated to the Nucleus of HeLa Cells and Is
Involved in Ribosomal RNA Transcription and Cell Proliferation
Takanori Tsuji, Yeqing Sun, Koji Kishimoto, Karen A. Olson, Shumei Liu, Saori Hirukawa,and Guo-fu Hu
Center for Biochemical and Biophysical Sciences and Medicine, Department of Pathology, Harvard Medical School, Boston, Massachusetts
Abstract
Angiogenin is an angiogenic protein known to play a role inrRNA transcription in endothelial cells. Nuclear translocationof angiogenin in endothelial cells decreases as cell densityincreases and ceases when cells are confluent. Here we reportthat angiogenin is constantly translocated to the nucleus ofHeLa cells in a cell density–independent manner. Down-regulation of angiogenin expression by antisense and RNAinterference results in a decrease in rRNA transcription,ribosome biogenesis, proliferation, and tumorigenesis bothin vitro and in vivo . Exogenous angiogenin rescues the cellsfrom antisense and RNA interference inhibition. The resultsshowed that angiogenin is constitutively translocated into thenucleus of HeLa cells where it stimulates rRNA transcription.Thus, besides its angiogenic activity, angiogenin also plays arole in cancer cell proliferation. (Cancer Res 2005; 65(4): 1352-60)
Introduction
Angiogenin is a 14-kDa angiogenic protein originally isolatedfrom the conditioned medium of HT-29 human colon adenocar-cinoma cells based on its angiogenic activity (1). Angiogenin hasbeen shown to play a role in tumor angiogenesis (2, 3). Itsexpression is up-regulated in many types of cancers includingbreast (4), cervical (5), colon (6), colorectal (7), endometrial (8),gastric (9), kidney (10), ovarian (11), pancreatic (12), prostate (13),and urothelial (14) cancers, as well as astrocytoma (15), leukemia(16), lymphangioma (17), melanoma (18), Wilms tumor (19), andothers.Angiogenin undergoes nuclear translocation in endothelial cells,
which has been shown to be necessary for angiogenesis. Inhibi-tion of nuclear translocation of angiogenin (20) or mutagenesis atits nuclear localization sequence (21) both abolished the an-giogenic activity. Nuclear translocation of angiogenin in endothe-lial cells is rapid (22) and independent of microtubules andlysosomes (23), but is strictly dependent on cell density (22).It decreases as cell density increases and ceases when cells areconfluent. No nuclear angiogenin can be detected in normal,non–blood vessel cells regardless of the cell density. Consistently,angiogenin does not induce any detectable cellular response in
epithelial cells and fibroblasts. These results suggested that thenuclear function of angiogenin in normal cells is specific for en-dothelial cells and occurs only when cells are not confluent.Angiogenin interacts with endothelial cells to induce a wide
range of cellular response including migration (24), proliferation(25), and tube formation (26). All these are necessary steps in theprocess of angiogenesis. The activity of angiogenin is relatively lowwhen compared with that of classic endothelial cell mitogens suchas vascular endothelial growth factor and basic fibroblast growthfactor. However, angiogenin has a comparable angiogenic activityin various in vivo angiogenesis assays (1, 27). Moreover,angiogenin antagonists potently inhibited the establishment,progression, and metastasis of human cancer cells inoculated inathymic mice (2, 3). These results suggested that angiogenin has adistinct, yet to be uncovered role in angiogenesis and tumorigen-esis. Recently, we discovered that angiogenin is able to bind to therRNA gene and stimulate rRNA transcription (28). An angiogenin-binding element has been identified from rDNA and we haveshown that this DNA sequence has angiogenin-dependentpromoter activity (29), suggesting that the function of nuclearangiogenin is related to rRNA transcription.Cancer is characterized by sustained cell growth requiring
continuous protein synthesis that depends on a constant supply ofribosomes (30). Ribosome biogenesis is a multistep processinvolving assembly of ribosomal proteins and rRNA in an equalmolar ratio. During tumorigenesis, the transcription of ribosomalproteins is known to be up-regulated through the Akt-PI3K-mTOR-S6K pathway. However, it is less clear how rRNA transcription isup-regulated. Here we report that angiogenin is continuouslytranslocated into the nucleus of HeLa cells in a cell density–independent manner. Down-regulation of angiogenin expressioninhibited rRNA transcription, ribosome biogenesis, cell prolifera-tion, and tumorigenesis.
Materials and Methods
Cell Culture. Human umbilical vein endothelial cells (HUVEC) were
cultured in human endothelial serum-free basal growth medium + 5% fetal
bovine serum (FBS) and 5 ng/mL basic fibroblast growth factor. HeLa cells
were cultured in DMEM + 10% FBS. Cell viability was determined by trypan
blue exclusion method. Cell number was determined with a Coulter
counter.
Nuclear Translocation of Angiogenin. HUVEC and HeLa cells wereseeded at various densities on a coverslip for 24 hours, washed with serum-
free medium and incubated with 1 Ag/mL angiogenin at 37jC for 1 hour.
Cells were washed with PBS, fixed with methanol at �20jC for 10 minutes,
and washed again with PBS containing 30 mg/mL bovine serum albumin.The fixed cells were then incubated with 50 Ag/mL of anti-angiogenin
monoclonal antibody 26-2F for 1 hour, washed, and incubated with Alexa
488–labeled goat F(abV)2 anti-mouse IgG at 1:100 dilution for 1 hour. The
Note: T. Tsuji and Y. Sun contributed equally. Y. Sun is currently in the Institute ofEnvironmental Science and Systems Biology, Dalian Maritime University, Dalian116026, China. K. Kishimoto is currently in the Department of Oral and MaxillofacialSurgery, Okayama University Graduate Schools, Okayama 700-8525, Japan. K.A. Olsonis currently in Archemix Corp., 1 Hampshire Street, Cambridge, MA 02139.
Requests for reprints: Guo-fu Hu, Department of Pathology, Harvard MedicalSchool, 77 Avenue Louis Pasteur, MA 02115. Phone: 617-432-6582; Fax: 617-432-6580;E-mail: [email protected].
#2005 American Association for Cancer Research.
Cancer Res 2005; 65: (4). February 15, 2005 1352 www.aacrjournals.org
cells were then washed, mounted in 50% glycerol, and monitored witha Nikon Labphot epifluorescent microscope.
[3H]Thymidine and [3H]Uridine Incorporation. Cells were cultured
for 24 hours, serum starved for another 24 hours, stimulated with 10% FBS,
and continuously labeled with 1 ACi/mL [3H]thymidine or [3H]uridine for
24 hours. The cells were washed with PBS, precipitated with 10%
trichloroacetic acid, and solubilized with 0.2 N NaOH plus 0.2% SDS. After
neutralization with 0.2 volume of 1 N HCl, the radioactivity was determined
by liquid scintillation counting.
Northern Hybridization. The probe used for 45S rRNA (5V-GGTCGCCA-
GAGGACAGCGTGTCAG-3V) hybridizes with the first 25 nucleotides of the
45S rRNA (31). The probe used for actin (5V-CTCTGTGCTCG-
CGGGGCGGAC-3V) hybridizes with the 5V end of cytoplasmic h-actinmRNA. The probe used for angiogenin was a 569-bp DNA fragment
containing the complete 441-bp angiogenin coding sequence and 44 bp
from the 5V and 15 bp from the 3V noncoding regions.
ELISA Detection of Angiogenin. ELISA plates were coated with 1 Ag
26-2F and blocked with 5 mg/mL bovine serum albumin in PBS. Samples
were added in triplicates and the plates were incubated at 4jC overnight,
washed with PBS, and incubated with 100 AL/well anti-angiogenin
polyclonal antibody R112 (1:4,000) at room temperature for 2 hours. After
washing with PBS, an alkaline phosphatase–labeled goat anti-rabbit
antibody (1.25 Ag/mL) was added and incubated at room temperature
for 1 hour. The plate was washed with PBS and p-nitrophenyl phosphate
(5 mg/mL, 100 AL/well) dissolved in 0.1 mol/L diethanolamine containing
10 mmol/L MgCl2 (pH 9.8) was added. After 1-hour incubation at room
temperature, the absorbance was measured at 410 nm with a turbidity
reference at 630 nm. A standard curve (50-1,000 pg angiogenin per well)
was done each time on every plate.
Silver Staining of Nucleolar Organizer Region. Cells cultured on
a coverslip were fixed with methanol/acetic acid (9:1, v/v) at room
temperature for 10 minutes and incubated in 60 mmol/L NaAc (pH 4.1)
containing 0.8 g/mL AgNO3, 15% formaldehyde, and 3% methanol at 37jCin the dark for 30 minutes. The slip was washed with water, mounted on
a glass slide, and observed at 1,000� magnification. nucleolar organizer
region (NOR) dots were counted in 30 randomly selected nuclei.
[32P] Labeling of Newly Synthesized RNA. Cells were pulse-labeled
with 25 ACi/mL [32P]Pi for 2 hours and washed with PBS. Total cellular RNA
was extracted with Trizol. Equal amounts of RNA were applied for agarose/
formaldehyde gel electrophoresis and transferred onto a nylon membrane.
Radiolabeled RNA was visualized by autoradiography.
Stable Transfection of HeLa Cells with Angiogenin Antisense Vector.
The entire coding region of the human angiogenin cDNA was amplified
from pAngC plasmid and cloned into the pCI-neo vector in the antisense
orientation. The pCI-Ang(�) plasmid was transfected into HeLa cells using
Lipofectin and stable transfectants were selected with 2 mg/mL G418.
Integration of the transfected gene into chromosome was confirmed by
genomic DNA PCR with the forward primer (5V-AGTACTTAATACGACT-
CACTATAGGC-3V) from the T7 sequence of the pCI vector and the reverse
primer (5V-ATGCAGGATAACTCCAGGTACAC-3V) from the inserted angio-
genin sequence. Transcription of the antisense human angiogenin mRNA
was confirmed by reverse transcription–PCR using the same set of
primers.
Stable Transfection of HeLa Cells with a Plasmid ContainingAngiogenin RNA Interference Cassette. Three regions corresponding
to the nucleotides 106 to 126, 122 to 142, 381 to 401, respectively, of the
angiogenin mRNA were originally selected for small interfering RNA
targeting. Double-stranded, 21-nucleotide-long RNA with 2-nucleotide
overhang at the 3V end were synthesized and transfected into HeLa cells
to test the efficiency of these small interfering RNA in inhibiting angiogenin
expression. ELISA showed that the third region (381-401) with the sequence
of 5V-GGTTCAGAAACGTTGTTGTTA-3V was most effective in reducing
angiogenin expression. This sequence was therefore used to construct an
angiogenin RNA interference (RNAi) plasmid in the pBS/U6 vector
according to the method of Sui et al. (32). This plasmid (pAng-RNAi) and
pBabe-puro were cotransfected into HeLa cells in the presence of Lipofectin
and the stable transfectants were selected with 0.5 Ag/mL puromycin for
2 weeks.
Soft Agar Assay. Cells were seeded at a density of 4 � 103 cells per
35-mm cell culture dishes in 0.33% agar and cultured for 7 days. Dishes
were stained with crystal violet solution (0.05%) overnight at 4jC. Colonieswere counted in 10 fields at �25 magnification. Two-tailed Student’s t test
was used to verify the differences between the groups.
Xenografic Growth of HeLa Cell Tumors in Athymic Mice. pCI or pCI-Ang(�) transfectants, 8 � 105 cells per mouse, were injected s.c. into the
left shoulders of male athymic mice (eight mice per group). Tumor sizes
were measured with a microcaliper and recorded in cubic millimeters(length � width2). Mice were sacrificed on day 13 and the wet weight of the
tumor was recorded.
Immunohistochemistry. Dako’s (Carpinteria, CA) Envision system was
used. Neovesssels were stained with an anti–von Willebrand’s factor IgG at
a 1:200 dilution. von Willebrand’s factor–possitive vessels in each tumor
were counted in five most vascularized areas at 200� magnification.Proliferating cells were stained with an anti–proliferating cell nuclear
antigen (PCNA) IgG. Angiogenin was stained with 26-2F as the primary
antibody at the concentration of 10 Ag/mL.
Results
Angiogenin Is Constitutively Translocated to the Nucleus ofHeLa Cells. Immunofluorescent staining showed that nucleartranslocation of exogenous angiogenin in HUVECs occurred onlywhen cells were not confluent. Bright nuclear/nucleolar stainingwas detected when cells were cultured at 5 � 103 cells/cm2.Virtually no nuclear angiogenin was observed when cells werecultured at a density exceeding 1 � 105 cells/cm2 (Fig. 1A).However, angiogenin was detected in the nuclei of HeLa cellscultured under the densities ranging from 5 � 103 to 2 � 105 cells/cm2 (Fig. 1A).The primary antibody (26-2F) used in this experiment is specific
for human angiogenin (33). It does not recognize angiogenin fromother species including bovine, porcine, rabbit, and mouse. X-raystructural analysis of angiogenin-antibody complex has shown that26-2F interacts with two segments consisting of residues 34 to 41and 85 to 91, respectively (34). These two regions are apart in theprimary but close in the 3-dimentional structures. No fluorescencewas observed when 26-2F was replaced by a nonimmune IgG (datanot shown). Western blotting analysis confirmed that comparableamounts of angiogenin protein were detected when equal amountsof nuclear proteins, extracted from HeLa cells cultured undervarious densities, were applied. However, in HUVECs, angiogeninwas detectable only in the nuclear proteins extracted fromnonconfluent cells (Fig. 1B). Fig. 1C shows that the nuclei isolatedfrom HeLa cells cultured under various densities responded toangiogenin in transcribing 45S rRNA in a nuclear run-on assay.We have also found that angiogenin is translocated to the nucleusin other types of cancer cells including MB231 breast, PC-3prostate, and HT-29 colon cancer cells (data not shown).These results suggested that nuclear translocation of angiogenin
occurs in HeLa cells regardless of the cell confluence status andthat angiogenin stimulates rRNA synthesis in HeLa nuclei. Theyalso indicated that HeLa cell nuclei are not yet saturated withendogenous angiogenin because they can still be stimulated tosynthesize rRNA by exogenous angiogenin. However, down-regulation of endogenous angiogenin expression did reduce thetranscription potential of HeLa nuclei. As shown in Fig. 1D , therRNA transcription activity of the nuclei isolated from angiogenin-underexpressing HeLa cells after angiogenin RNAi transfection was
Angiogenin Plays a Role in HeLa Cell Proliferation
www.aacrjournals.org 1353 Cancer Res 2005; 65: (4). February 15, 2005
lower than that from vector control transfectants. These resultsshowed that the rRNA transcription activity of the nuclei iscorrelated with angiogenin level.Transient Transfection of an Angiogenin Antisense Com-
pound Inhibited Cell Proliferation. To confirm that endogenous
angiogenin is indeed involved in rRNA transcription, we transiently
transfected HeLa cells with an angiogenin antisense oligo to inhibit
angiogenin expression and examined the resultant change in rRNA
transcription. Fig. 2A shows that transfection of HeLa cells with
CT-1, a second-generation antisense oligo for angiogenin (28),
decreased the level of 45S rRNA but had no effect on that of
glyceraldehyde-3-phosphate dehydrogenase. Transfection with
CT-2, a scrambled control oligo, had no effect. A decrease in
angiogenin mRNA and protein levels was confirmed by Northern
blotting and ELISA, respectively (Fig. 2A and B).HeLa cell proliferation was inhibited by CT-1 transfection
(Fig. 2C). CT-2 had no effect on cell proliferation indicating thatthe inhibitory activity of CT-1 was not due to a nonspecific cytotoxic
effect. These results showed a positive correlation between
angiogenin level, rRNA transcription, and cell proliferation. CT-1also significantly inhibited the proliferation of MB231, PC-3, and
HT-29 cells, whereas CT-2 had no effect (data not shown).
Stable Transfection of an Angiogenin Antisense PlasmidInhibited rRNA Transcription. To further show that endogenousangiogenin is required for rRNA transcription and cancer cell
growth, we transfected HeLa cells with pCI-Ang(�), a plasmid
containing the antisense sequence of angiogenin cDNA, andselected stable transfectants with G418. Angiogenin level in the
antisense transfectants was reduced as determined by ELISA and
Western blotting analysis (Fig. 3A).The morphology of the cells changed dramatically in the
antisense transfectants (Fig. 3B). The steady-state level of the 45SrRNA was significantly lower in the antisense transfectants as
determined by Northern blotting (Fig. 3C , left ). Ribosome
biogenesis was measured by silver staining of NOR (Fig. 3C , right).
NORs are actively transcribing rDNA loops and reflect the ribosomebiogenesis status (35). They are associatedwith argyrophilic proteins
and can be visualized by silver staining. The size and number of
NORs reflect the capacities of the cell to transcribe rRNA and aretherefore indices for cell proliferation, transformation, and even
malignancy (36). The decrease in 45S rRNA level and in NOR
numbers in pCI-Ang(�) transfectants indicated that rRNA tran-scription and ribosome biogenesis were decreased when angiogenin
expression was inhibited.
Figure 1. Nuclear translocation and nuclear function of angiogenin in HeLa cells is independent of cell density. A, HUVEC and HeLa cells, seeded at thedensities indicated, were incubated with 1 Ag/mL angiogenin at 37jC for 1 hour. The cells were washed with PBS and fixed with methanol at �20jC for 10 minutes.Angiogenin was visualized with 10 Ag/mL 26-2F and Alexa 488–labeled goat anti-mouse IgG. B, nuclear proteins were extracted from HUVEC and HeLa cellscultured under various densities and analyzed by Western blotting (150 Ag per lane) with an anti-angiogenin polyclonal antibody (R112). C, HeLa cells were culturedfor 24 hours and nuclei were isolated by subcellular fractionation. Nuclear run-on transcription was carried out with 5 � 104 nuclei isolated from HeLa cells culturedunder various densities in the presence or absence of 1 Ag/mL angiogenin. D, nuclei were isolated from HeLa cells transfected with an angiogeninRNAi plasmid (pAng-RNAi) and the vector control (pBS/U6) and nuclear run-on assay was done.
Cancer Research
Cancer Res 2005; 65: (4). February 15, 2005 1354 www.aacrjournals.org
RNAi was also used to down-regulate angiogenin expression inHeLa cells. Very similar results were obtained as in antisensetransfection except that they were more significant due to agreater degree of inhibition in angiogenin expression (Fig. 3E).We used a DNA vector-based RNAi method (32) to construct apAng-RNAi plasmid, with the insertion DNA templatecorresponding to the nucleotides 381 to 401 of angiogeninmRNA. It was transfected into HeLa cells together with pBabe-puro (37) and those cells that stably express angiogenin RNAiwere selected by puromycin resistance. Stable transfection ofpAng-RNAi resulted in an 80% reduction of angiogeninexpression, as shown by ELISA and Western Blotting (Fig. 3D).The 45S rRNA level and NOR numbers were also decreased moresignificantly (Fig. 3F).The effect of down-regulating angiogenin expression on the
de novo synthesis of rRNA was determined by metabolic labeling
with 32P. Cells were pulse-labeled with [32P]Pi for 2 hours and thetotal cellular RNA was isolated with Trizol. Fig. 4A shows that thenewly synthesized 45S rRNA was significantly decreased in pAng-RNAi transfectants than that in pBS/U6 control transfectantswhen equal amounts of total RNA was loaded (Fig. 4B). In the2-hour period, some of the newly synthesized 45S rRNA has beenprocessed to 32S rRNA whose level was significantly lower inpAng-RNAi transfectants (Fig. 4A). Because of the large cytoplas-mic pool of stable ribosomes, the cellular levels of 28S and 18SrRNA were relatively stable (38). Thus, 28S and 18S rRNA stillserve as the best loading control. However, it should be notedthat the RNAi transfectants do have a reduced steady-state levelof 45S rRNA, which results in a decrease in the cellular level of 28Sand 18S rRNA. Therefore, use of 28S rRNA as a loading controlactually underestimated the difference in newly synthesized 45SrRNA per cell after antisense and RNAi transfection. In any event,these results indicated that the de novo synthesis of 45S rRNAdecreased when angiogenin expression was inhibited, confirmingthat the nuclear function of angiogenin is related to rRNAtranscription.Cell Proliferation Was Inhibited by Down-regulating Angio-
genin Expression. Because of the central importance of rRNA
transcription in cell growth, decreased rRNA transcription should
slow cell proliferation. This is confirmed by three different cell
proliferation assays (Fig. 5). Direct cell number counting (Fig. 5Aand B) showed that both antisense and RNAi transfectants have
reduced cell proliferation rates than that of their corresponding
vector control transfectants. However, the extent of inhibition was
greater in RNAi (Fig. 5B) than in antisense transfectants (Fig. 5A),
probably reflecting the fact that RNAi is more effective in
inhibiting angiogenin expression.The rate of DNA and RNA synthesis was determined by
[3H]thymidine and [3H]uridine incorporation, respectively. Cellswere continuously labeled for 24 hours and the incorporatedradioactivity was normalized to cell numbers. pCI-Ang(�) andpAng-RNAi transfectants had slower DNA (Fig. 5C and D) and RNA(Fig. 5E and F) synthesis rates than do the respective vectorcontrol transfectants.These results showed that endogenous angiogenin in HeLa cells
plays an important role in rRNA transcription, ribosomebiogenesis, and cell proliferation.Colony Formation in Soft Agar Was Inhibited by Angiogenin
RNAi. The anchorage-independent growth of the pBS/U6 andpAng-RNAi transfectants was analyzed by colony formation assayin soft agar (Fig. 6A-C). RNAi transfection decreased both colonynumber and size significantly. Colony number per 35-mm dish was1,631 F 52 and 908 F 42, and the average colony size was 115 F8 and 65 F 8 Am in vector control and in pAng-RNAi transfectants,respectively (Fig. 6A and B). A complete recovery was obtainedwhen exogenous angiogenin (0.1 Ag/mL) was added (Fig. 6C),indicating that the decrease in colony formation in pAng-RNAitransfectants was caused by reduced angiogenin expression. Theseresults showed that the tumorigenicity of HeLa cells is decreasedby down-regulating angiogenin expression.Angiogenin Down-regulation Inhibited Ecotopic Growth
of HeLa Cell Tumors in Athymic Mice. The effect of angiogeninantisense transfection on tumor growth was examined in athymicmice. Figure 7 shows that the appearance, establishment, andgrowth of HeLa cells inoculated in athymic mice were all sig-nificantly inhibited after angiogenin antisense transfection. There
Figure 2. The effect of angiogenin antisense oligonucleotide CT-1 on HeLacell growth. A, HeLa cells were treated with 1 Amol/L CT-1 or CT-2 in thepresence of 5 AL/mL Lipofectin. Total RNA from 2 � 106 cells wasisolated and analyzed by Northern blotting for 45S rRNA, glyceraldehyde-3-phosphate dehydrogenase, and angiogenin. B, HeLa cells, 5 � 104 cellsper 35-mm dish, were cultured in DMEM + 10% FBS for 24 hours. The cellswere washed by OptiMEM and transfected with 1 Amol/L CT-1 or CT-2in the presence of 5 AL/mL Lipofectin. The culture media were collected at 72hours and angiogenin levels were determined by ELISA and normalizedto cell numbers. C , cell numbers were determined at 48, 72, and 96 hours.Points, means of three experiments; bars, SD.
Angiogenin Plays a Role in HeLa Cell Proliferation
www.aacrjournals.org 1355 Cancer Res 2005; 65: (4). February 15, 2005
was a delay in tumor occurrence (Fig. 7A) and the progression wasslower (Fig. 7B) in angiogenin antisense-transfected HeLa cells.When tumors did develop, their average size was about half ofthose in the control group (Fig. 7B). These data confirmed theresults obtained in soft agar assays that the tumorigenicity of HeLacells decreased when angiogenin expression was down-regulated.Immunohistochemical staining with an anti–von Willebrand’s
factor antibody (Fig. 7C) showed that the HeLa tumor tissue grown
from pCI-Ang(�) transfectants had about half of the blood vessel
density (11 versus 21 vessels per field at 200�, peripheral regions
of the tumors) as compared with those grown from vector control
transfectants, indicating that angiogenesis was inhibited in
angiogenin antisense transfectants, in agreement with the known
role that angiogenin plays in tumor angiogenesis. More impor-
tantly, PCNA staining showed that cell proliferation in the pCI-
Ang(�) tumor was significantly lower than that in the control
(Fig. 7D), indicating that not only tumor angiogenesis but also
tumor cell proliferation per se was inhibited by angiogenin down-
regulation. Staining with 26-2F confirmed that angiogenin in the
nuclei of tumor cells decreased after antisense transfection
(Fig. 7E). These results further showed that angiogenin plays
a dual role in tumor growth by contributing both in angiogenesis
and in cancer cell proliferation.
Discussion
Angiogenin was isolated as a tumor angiogenic factor basedsolely on its angiogenic activity (1). Therefore, subsequent studieshad mainly focused on how it induces angiogenesis and how itsangiogenic activity can be intervened. Angiogenin has beenconsidered to interact only with endothelial cells and vascularsmooth muscle cells. Results presented in this article indicatedthat cancer cells are also targets for angiogenin.We have showed in this article that angiogenin is constitutively
translocated to the nucleus of HeLa cells where it plays a role inrRNA transcription. Because rRNA transcription regulates ribo-some production and, consequently, the translation potential of acell, it is conceivable that deregulation of rRNA transcription maybe an important determinant in neoplastic transformation.Continuous nuclear translocation of angiogenin in cancer cellscan certainly be one of the contributing factors. Indeed, inhibitingangiogenin expression reduced tumorigenicity and reversed themalignant phenotype of HeLa cells.
Figure 3. Down-regulation of angiogenin inhibits HeLa cell growth. HeLa cells were transfected with pCI-Ang(�) or pAng-RNAi and their corresponding controlvectors, pCI and pBS/U6. Stable transfectants were selected with 2 mg/mL G418 and 0.5 Ag/mL puromycin for pCI-Ang(�) and pAng-RNAi, respectively. Similarresults were obtained from all the individual clones and from the pooled populations. A-C , pCI-Ang(�) and pCI transfectants. D-F , pAng-RNAi and pBS/U6transfectants. A and D , angiogenin secretion levels determined by ELISA (left), and cellular angiogenin levels detected by Western blotting with actin as theloading control (right ). B and E, cell morphology changes in angiogenin-underexpressing cells. C and F, the steady-state level of 45S rRNA detected byNorthern blotting with 18S rRNA as the loading control (left ) and the numbers of AgNOR (right ).
Cancer Research
Cancer Res 2005; 65: (4). February 15, 2005 1356 www.aacrjournals.org
Several animal models have been used to examine the role
angiogenin plays in tumor growth (2, 39). Whereas the antitumor
activity of angiogenin antagonists was obvious, the mechanism by
which they prevented or delayed tumor appearance was not fully
understood. Based on the long-held assumption that angiogenin is
a tumor angiogenic protein, these data seemed to support the
proposition that the observed antitumor effects were due
to inhibition of angiogenin-induced tumor angiogenesis. However,
the tumors that eventually developed in angiogenin antagonist–
treated mice were substantially smaller than those in the control
group. It is therefore possible that other effects such as repression
of rRNA transcription of cancer cells also contributed to the
marked anticancer activity of angiogenin antagonists. Our results
shown in Fig. 7 showed that this is truly the case. Down-regulating
angiogenin expression in HeLa cells not only reduced tumor
angiogenesis (Fig. 7D) but also inhibited tumor cell proliferation
(Fig. 7E).The finding that angiogenin is involved in rRNA transcription in
cancer cells is significant for our understanding of cancer biologyand neoplastic transformation. One of the hallmarks of cancer issustained cell growth and this can only be achieved by increasedprotein synthesis. To accommodate this need, there must be anincrease in ribosome biogenesis. Up-regulation of ribosomalproteins and rRNA transcription is an important factor in cancertransformation (40). It has been reported that the proliferationeffects of estrogen in the rat induces pituitary tumors (41).However, the susceptibility to formation of such tumors is highlystrain dependent. The particularly susceptible Fisher 344 strain
Figure 5. Angiogenin-underexpressingHeLa cells have reduced growth rate. A,C, and E, antisense transfectants. B, D,and F, RNAi transfectants. A and B, cellproliferation was determined by direct cellnumber counting. C and D, DNA synthesisdetermined by [3H]thymidine incorporation.E and F, RNA synthesis determined by[3H]uridine incorporation. Cells werecultured for 24 hours, serum-starved foranother 24 hours, stimulated with 10% FBS,and continuously incubated with 1 ACi/mL[3H]thymidine (C and D ) or 1 ACi/mL[3H]uridine (E and F ) for 24 hours.
Figure 4. Angiogenin RNAi transfection inhibits the de novo synthesis of45S rRNA. pBS/U6 or pAng-RNAi transfectants were cultured in DMEM+ 10% FBS. [32P]Pi (25 ACi/mL) was added to each dish and incubated for2 hours. Total cellular RNA was isolated with Trizol and the amount determinedby absorbance at 260 nm. Equal amounts of RNA were loaded to each laneand subjected to agarose/formaldehyde electrophoresis. The newly synthesizedRNA were visualized by autoradiography (A ). 28S and 18S rRNA werestained by ethidium bromide (B). Representative experiment of three repeats.
Angiogenin Plays a Role in HeLa Cell Proliferation
www.aacrjournals.org 1357 Cancer Res 2005; 65: (4). February 15, 2005
develops tumors after 30 to 55 days of estrogen treatment.In contrast, the Sprague-Dawley strain is resistant to such tumors.The major difference between the two strains is the markedincrease (250%) of rRNA in the pituitaries of Fisher 344 rats(42). It is notable that the increase in rRNA accumulationoccurs within 3 days of estrogen treatment when the increase inthe number of lactotrophs is minimal or nonexistent. Theseresults indicated that estrogen induces rRNA synthesis in the
early stages of tumorigenesis before any tumor growth isdetectable.Androgen-dependent growth of the prostate has been well
documented in prostatic hyperplasia and prostatic carcinoma(43). It has been shown that androgens regulate the accumu-lation of rRNA during androgen-dependent cell growth (44) andthat androgen-stimulated rRNA synthesis is the mechanism bywhich androgens affect growth (45). Recently, it has been
Figure 6. Angiogenin RNAi transfectants have reducedtumorigenicity in soft agar. Anchorage-independent growth of pBS/U6 control (A ) and pAng-RNAi transfectants in the absence (B)or presence of 0.1 Ag/mL angiogenin (C ) was determined in softagar. The numbers of colonies in the entire 35-mm dish werecounted. Average colony size was determined by measuring thediameters of colonies in 10 microscope fields with a microcaliber.Pictures from representative areas.
Figure 7. Angiogenin antisense HeLa transfectants have reduced tumorigenicity in athymic mice. pCI vector control and pCI-Ang(�) antisense transfectants, 8 � 105
cells per mouse, were injected s.c. into the left shoulder of the mice. Eight mice were used per group. A, tumor appearance. Mice were examined for tumor growth thriceper week. B, tumor volume. Tumor sizes were measured with a caliper and recorded in cubic millimeters (length � width2). C, neovessels were stained with an anti–vonWillebrand’s factor IgG. Vessel density (vessels/field) was shown as mean F SD for each group. D, proliferating cells were stained by an anti-PCNA monoclonalantibody. PCNA-positive and total numbers of cells were counted in 10 randomly selected areas at 200� magnification. E, 26-2F was used to reveal nuclear angiogeninin HeLa tumors derived from pCI and pCI-Ang(�) transfectants. No staining was observed when a nonimmune IgG was used to replace 26-2F. Pictures from arepresentative animal of each group.
Cancer Research
Cancer Res 2005; 65: (4). February 15, 2005 1358 www.aacrjournals.org
reported that angiogenin is the most up-regulated gene inprostate intraepithelial neoplasia in the murine prostaterestricted Akt kinase transgenic mice (13). In these mice,expression of Akt1 in the prostate results in activation of thep70S6K pathway that up-regulates ribosomal proteins through theAkt-p70S6K-mTOR pathway. Because rRNA production has toincrease in an equimolar ratio to that of ribosomal proteins forribosome biogenesis, we are currently investigating whetherrRNA transcription in the prostate is stimulated by angiogeninand whether inhibition of angiogenin expression will inhibitprostate intraepithelial neoplasia development.The role angiogenin plays in rRNA transcription in cancer
cells suggests that up-regulation of angiogenin expression invarious cancer cells not only induces tumor angiogenesis butalso directly contributes to cell proliferation. Thus, inhibitorstargeting angiogenin will be more effective than those thatinhibit either angiogenesis or cancer cell proliferation alone. Wehave found that neomycin, an aminoglycoside antibiotic thatblocks nuclear translocation of angiogenin (20), significantlyinhibits both tumor angiogenesis and cancer cell proliferation inan ecotopic human tumor model in athymic mice.1
The mechanism by which angiogenin stimulates rRNA tran-scription is unclear at present. We believe that angiogenin acts incancer cell nucleus in a similar manner as it does in endothelial
cell nucleus. Angiogenin has been shown to undergo nucleartranslocation in endothelial cells through endocytosis and classicnuclear pore import (27). The detailed transportation steps acrossthe cytoplasm are unknown, but it seems to be lysosome andmicrotubule independent (29). More extensive work is needed tounderstand how angiogenin is translocated to the nucleus, how itinteracts with the RNA polymerase I machinery, and whether italso affects transcription catalyzed by RNA polymerase II. Wehave carried out a DNA array analysis (Atlas Human 1.2 Array I,II, III, Clontech, Palo Alto, CA) and found that the ribosomalproteins are universally down-regulated (19-80% of the control) inpAng-RNAi transfectants.2 All these results point to an importantrole angiogenin plays in ribosome biogenesis of cancer cells.
Acknowledgments
Received 6/7/2004; revised 11/9/2004; accepted 11/22/2004.Grant support: NIH grant CA91086 (G-F. Hu) and the Endowment for Research in
Human Biology, Inc.The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.
1 Hirukawa et al., unpublished data.2 Tsuji et al., unpublished data.
References1. Fett JW, Strydom DJ, Lobb RR, et al. Isolation andcharacterization of angiogenin, an angiogenic proteinfrom human carcinoma cells . Biochemistry1985;24:5480–6.
2. Olson KA, Fett JW, French TC, Key ME, Vallee BL.Angiogenin antagonists prevent tumor growth in vivo .Proc Natl Acad Sci U S A 1995;92:442–6.
3. Olson KA, Byers HR, Key ME, Fett JW. Prevention ofhuman prostate tumor metastasis in athymic mice byantisense targeting of human angiogenin. Clin CancerRes 2001;7:3598–605.
4. Montero S, Guzman C, Cortes-Funes H, Colomer R.Angiogenin expression and prognosis in primary breastcarcinoma. Clin Cancer Res 1998;4:2161–8.
5. Chopra V, Dinh TV, Hannigan EV. Circulating serumlevels of cytokines and angiogenic factors in patientswith cervical cancer. Cancer Invest 1998;16:152–9.
6. Li D, Bell J, Brown A, Berry CL. The observation ofangiogenin and basic fibroblast growth factor geneexpression in human colonic adenocarcinomas, gastricadenocarcinomas, and hepatocellular carcinomas.J Pathol 1994;172:171–5.
7. Shimoyama S, Shimizu N, Tsuji E, et al. Distribution ofangiogenin and its gene message in colorectal cancerpatients and their clinical relevance. Anticancer Res2002;22:1045–52.
8. Chopra V, Dinh TV, Hannigan EV. Serum levels ofinterleukins, growth factors and angiogenin in patientswith endometrial cancer. J Cancer Res Clin Oncol1997;123:167–72.
9. Shimoyama S, Kaminishi M. Increased angiogeninexpression in gastric cancer correlated with cancerprogression. J Cancer Res Clin Oncol 2000;126:468–74.
10. Wechsel HW, Bichler KH, Feil G, et al. Renal cellcarcinoma: relevance of angiogenetic factors. Antican-cer Res 1999;19:1537–40.
11. Barton DP, Cai A, Wendt K, et al. Angiogenic proteinexpression in advanced epithelial ovarian cancer. ClinCancer Res 1997;3:1579–86.
12. Shimoyama S, Gansauge F, Gansauge S, et al.Increased angiogenin expression in pancreatic canceris related to cancer aggressiveness. Cancer Res1996;56:2703–6.
13. Majumder PK, Yeh JJ, George DJ, et al. Prostateintraepithelial neoplasia induced by prostate restrictedAkt activation: the MPAKT model. Proc Natl Acad SciU S A 2003;100:7841–6.
14. Miyake H, Hara I, Yamanaka K, et al. Increasedangiogenin expression in the tumor tissue andserum of urothelial carcinoma patients is relatedto disease progression and recurrence. Cancer1999;86:316–24.
15. Eberle K, Oberpichler A, Trantakis C, et al. Theexpression of angiogenin in tissue samples of differentbrain tumours and cultured glioma cells. AnticancerRes 2000;20:1679–84.
16. Verstovsek S, Kantarjian H, Aguayo A, et al. Signifi-cance of angiogenin plasma concentrations in patientswith acute myeloid leukaemia and advanced myelodys-plastic syndrome. Br J Haematol 2001;114:290–5.
17. Park YW, Kim SM, Min BG, Park IW, Lee SK.Lymphangioma involving the mandible: immunohisto-chemical expressions for the lymphatic proliferation.J Oral Pathol Med 2002;31:280–3.
18. Hartmann A, Kunz M, Kostlin S, et al. Hypoxia-induced up-regulation of angiogenin in human malig-nant melanoma. Cancer Res 1999;59:1578–83.
19. Skoldenberg EG, Christiansson J, Sandstedt B, et al.Angiogenesis and angiogenic growth factors in Wilmstumor. J Urol 2001;165:2274–9.
20. Hu GF. Neomycin inhibits angiogenin-induced an-giogenesis. Proc Natl Acad Sci U S A 1998;95:9791–5.
21. Moroianu J, Riordan JF. Nuclear translocation ofangiogenin in proliferating endothelial cells is essentialto its angiogenic activity. Proc Natl Acad Sci U S A1994;91:1677–81.
22. Hu G, Xu C, Riordan JF. Human angiogenin is rapidlytranslocated to the nucleus of human umbilical veinendothelial cells and binds to DNA. J Cell Biochem2000;76:452–62.
23. Li R, Riordan JF, Hu G. Nuclear translocation ofhuman angiogenin in cultured human umbilicalartery endothelial cells is microtubule and lysosomeindependent. Biochem Biophys Res Commun 1997;238:305–12.
24. Hu G, Riordan JF, Vallee BL. Angiogenin promotesinvasiveness of cultured endothelial cells by stimulationof cell-associated proteolytic activities. Proc Natl AcadSci U S A 1994;91:12096–100.
25. Hu GF, Riordan JF, Vallee BL. A putative angiogeninreceptor in angiogenin-responsive human endothelialcells. Proc Natl Acad Sci U S A 1997;94:2204–9.
26. Jimi S, Ito K, Kohno K, et al. Modulation bybovine angiogenin of tubular morphogenesis andexpression of plasminogen activator in bovineendothelial cells. Biochem Biophys Res Commun1995;211:476–83.
27. Fu X, Roberts WG, Nobile V, Shapiro R, Kamps MP.mAngiogenin-3, a target gene of oncoprotein E2a-Pbx1,encodes a new angiogenic member of the angiogeninfamily. Growth Factors 1999;17:125–37.
28. Xu ZP, Tsuji T, Riordan JF, Hu GF. The nuclearfunction of angiogenin in endothelial cells is related torRNA production. Biochem Biophys Res Commun2002;294:287–92.
29. Xu ZP, Tsuji T, Riordan JF, Hu GF. Identification andcharacterization of an angiogenin-binding DNAsequence that stimulates luciferase reporter geneexpression. Biochemistry 2003;42:121–8.
30. Hanahan D, Weinberg RA. The hallmarks of cancer.Cell 2000;100:57–70.
31. Kuo BA, Gonzalez IL, Gillespie DA, Sylvester JE.Human ribosomal RNA variants from a single individualand their expression in different tissues. Nucleic AcidsRes 1996;24:4817–24.
32. Sui G, Soohoo C, Affar el B, et al. A DNA vector-basedRNAi technology to suppress gene expression inmammalian cells. Proc Natl Acad Sci U S A 2002;99:5515–20.
33. Fett JW, Olson KA, Rybak SM. A monoclonal antibodyto human angiogenin. Inhibition of ribonucleolytic andangiogenic activities and localization of the antigenicepitope. Biochemistry 1994;33:5421–7.
34. Chavali GB, Papageorgiou AC, Olson KA, et al. Thecrystal structure of human angiogenin in complex withan antitumor neutralizing antibody. Structure2003;11:875–85.
35. Ruschoff J, Plate K, Bittinger A, Thomas C. Nucleolarorganizer regions (NORs). Basic concepts and practicalapplication in tumor pathology. Pathol Res Pract1989;185:878–85.
36. Trere D, Pession A, Derenzini M. The silver-stainedproteins of interphasic nucleolar organizer regions as aparameter of cell duplication rate. Exp Cell Res1989;184:131–7.
Angiogenin Plays a Role in HeLa Cell Proliferation
www.aacrjournals.org 1359 Cancer Res 2005; 65: (4). February 15, 2005
37. Brummelkamp TR, Bernards R, Agami R. A systemfor stable expression of short interfering RNAs inmammalian cells. Science 2002;296:550–3.
38. Sugimoto M, Kuo ML, Roussel MF, Sherr CJ.Nucleolar Arf tumor suppressor inhibits ribosomalRNA processing. Mol Cell 2003;11:415–24.
39. Olson KA, Byers HR, Key ME, Fett JW.Inhibition of prostate carcinoma establishmentand metastatic growth in mice by an antiangio-genin monoclonal antibody. Int J Cancer 2002;98:923–9.
40. Ruggero D, Pandolfi PP. Does the ribosome translatecancer? Nat Rev Cancer 2003;3:179–92.
41. Phelps C, Hymer WC. Characterization of estrogen-induced adenohypophyseal tumors in the Fischer 344rat. Neuroendocrinology 1983;37:23–31.
42. Ying C, Gregg DW, Gorski J. Estrogen-inducedchanges in rRNA accumulation and RNA polymerase Iactivity in the rat pituitary: correlation with pituitarytumor susceptibility. Mol Cell Endocrinol 1996;118:207–13.
43. Montie JE, Pienta KJ. Review of the role of androgenic
hormones in the epidemiology of benign prostatichyperplasia and prostate cancer. Urology 1994;43:892–9.
44. Mainwaring WI, Derry NS. Enhanced transcriptionof rRNA genes by purified androgen receptor com-plexes in vitro . J Steroid Biochem Mol Biol 1983;19:101–8.
45. Kabler RL, Srinivasan A, Taylor LJ, et al. Androgenregulation of ribosomal RNA synthesis in LNCaP cellsand rat prostate. J Steroid Biochem Mol Biol1996;59:431–9.
Cancer Research
Cancer Res 2005; 65: (4). February 15, 2005 1360 www.aacrjournals.org
2005;65:1352-1360. Cancer Res Takanori Tsuji, Yeqing Sun, Koji Kishimoto, et al. ProliferationIs Involved in Ribosomal RNA Transcription and Cell Angiogenin Is Translocated to the Nucleus of HeLa Cells and
