Kidney International, Vol. 65 (2004), pp. 1272–1279

The roles of IGF-I and IGFBP-3 in the regulation of proximaltubule, and renal cell carcinoma cell proliferation

CATHERINE W. CHEUNG, DAVID A. VESEY, DAVID L. NICOL, and DAVID W. JOHNSON

Department of Medicine and Department of Surgery, University of Queensland, Brisbane, Queensland, Australia; and Departmentof Renal Medicine and Department of Urology, Princess Alexandra Hospital, Brisbane, Queensland, Australia

The roles of IGF-I and IGFBP-3 in the regulation of proximaltubule, and renal cell carcinoma cell proliferation.

Background. Insulin-like growth factor I (IGF-I), a potentproximal tubule cell (PTC) mitogen, has been implicated inthe progression of many human cancers. Our previous work onhuman renal tissues has suggested that IGF-I and several of itsbinding proteins (IGFBP-3 and -6) are up-regulated in clear cellrenal cell carcinoma (RCC).

Methods. To further elucidate the role of IGF-I and IGFBPsin RCC growth, immunohistochemistry, thymidine incorpora-tion, and Western analysis were performed in primary culturesof normal PTC (priPTC) and clear-cell RCC (priRCC), as wellas in SN12K1 cells (a cell line derived from metastatic RCC).

Results. By immunohistochemistry, IGFBP-3 and IGF-I wereprominently expressed in SN12K1 cells, and weakly expressedin priPTC and priRCC. Incubation with 100 ng/mL IGF-I sig-nificantly augmented DNA synthesis by priPTC (mean ± SD120.7% ± 19.7% of controls, P < 0.05), priRCC (238.7% ±279.9% of controls, P < 0.01), and SN12K1(120.0% ± 22.9%of controls, P < 0.05). Neutralizing antibodies to IGF-I andIGF-I receptor significantly suppressed SN12K1 growth (81.9%± 13.5% of control, P < 0.01 and 87.4% ± 16.2% of control,P < 0.05, respectively). Removal of endogenous IGFBP-3 byan anti-IGFBP-3 increased SN12K1 DNA synthesis (243.9%± 35.3% of control, P < 0.001), which was partially abrogatedby coincubation with exogenous IGFBP-3 (135.97% ± 5.9% ofcontrols, P < 0.001). Using Western analysis, IGFBP-3 expres-sion was enhanced in IGF-I–stimulated SN12K1 cells exposedto exogenous IGF-I. Coincubation with anti-IGFBP-3 furtherenhanced IGF-I–induced DNA synthesis.

Conclusion. RCC cells express IGF-I and IGFBP-3, andare responsive to exogenous IGF-I stimulation. Moreover, inSN12K1 cells (derived from metastatic RCC), autocrine IGF-I and IGFBP-3 actions, respectively, stimulated and inhibitedgrowth. These results suggest that IGF-I and IGFBP-3 may bepotential candidates for therapeutic manipulation in patientswith advanced RCC.

Key words: renal cell carcinoma, insulin-like growth factor I, insulin-likegrowth factor binding protein 3, DNA synthesis, immunohistochem-istry, Western blotting.

Received for publication June 3, 2003and in revised form September 26, 2003Accepted for publication November 20, 2003

C© 2004 by the International Society of Nephrology

Renal cell carcinoma (RCC) accounts for 2% of allmalignancies. It is derived from the proximal tubular ep-ithelium of the kidney and the major histologic subtypeis the clear cell/conventional RCC (about 75% of renalneoplasms in all surgical series) [1]. Although surgeryis the standard treatment for localized disease [2], ap-proximately one third of RCC patients present withunresectable or metastastic disease for which currentsystemic therapies are minimally effective [3, 4]. There-fore, development of novel treatment alternatives isimperative. Better understanding of the growth factorpathways in RCC progression may aid in the advance-ment of more effective therapeutic strategies.

Recent evidence suggests that members of the insulin-like growth factor (IGF) family may play a key rolein RCC progression, and hence, may represent a po-tential target for therapeutic manipulation [5, 6]. IGF-Iis a potent cell mitogen and antiapoptotic agent [5–7].High serum IGF-I levels have been associated with anincreased risk of developing RCC [8], while high IGF-I pretreatment levels in patients with established RCCstrongly predicted an impaired response to interleukin-2 (IL-2) therapy. In experimental mice, injection of agrowth hormone-releasing hormone antagonist reducedthe IGF-I–induced proliferation of Caki-I RCC cell line[9]. Furthermore, changes in the cellular transcription ofmembers of the IGF family have been observed in ferricnitrilotriacetate-induced RCC in rats [10].

IGF-I exerts its biologic effects through binding toIGF-I receptors (IGF-IR), and these effects can be mod-ulated by one or more stimulatory or inhibitory IGFbinding proteins (IGFBPs). IGFBPs can, under certaincircumstances, potentiate the stimulatory effects of IGF-Ion cells by enhancing IGF-I presentation to its receptor[11, 12], as well as by reducing the free ligand-accelerateddissociation of IGF-I from IGF-IR [13]. Alternatively,IGFBP-3 can also inhibit cell proliferation under certaincircumstances [14–16], either by sequestering IGF-I orby IGF-I independent pathways, which include regula-tion of downstream signaling pathways through specificIGFBP-3 receptors and secondary messengers [16–19].

1272

Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation 1273

Previous studies from our laboratory have shown thatIGF-I and IGFBP-3 mRNA and protein levels (but notother IGFBPs) were up-regulated in human RCC (ourunpublished results), raising the possibility that these IGFmembers may play a role in RCC development and/orprogression. The aim of the present study, therefore, wasto examine the expression of IGF-I and IGFBP-3, and theeffects of manipulating their actions in an in vitro modelof human RCC.

METHODS

Growth factors and antibody

IGF-I, mouse antihuman IGF-IR (clone 33255.111),and a mouse IgG1 isotype control (clone 11711.11) werepurchased from R&D Systems (Bioscientific, Gymea,Australia). Long R3-IGF-I, an IGF-I analog, which bindsto IGF-I receptor, but not to IGFBPs[20], was obtainedfrom Gropep (catalog number AU200; Gropep, Ade-laide, Australia). Neutralizing antibodies assays againsthuman IGF-I (clone Sm1.2) and human IGF-IR (clone33255.111), used in thymidine incorporation assays, werepurchased from Upstate Biotechnology (Lake Placid,NY, USA) and R&D Systems, respectively. The humananti-IGF-I antibody used for immunohistochemistry, wasa rabbit antibody from Gropep (catalog number PABCa).The human recombinant IGFBP-3 was from Sigma-Aldrich (Castle Hill, NSW, Australia). Neutralizingrabbit antihuman IGFBP-3 antiserum was generously do-nated by Professor Robert Baxter at the Kolling Insti-tute of Medical Research in Sydney, Australia. This anti-serum reacted with IGFBP-3 at a site other than the IGF-Ibinding site [21]. Rabbit polyclonal antihuman IGFBP-3antibody (clone H-98), obtained from Santa CruzBiotechnology (Santa Cruz, CA, USA), was used forWestern blotting and immunohistochemsitry.

Cells

Primary cultures of normal proximal tubular cell (PTC)(priPTC) and RCC (priRCC) were isolated from patientswho underwent radical nephrectomy for localized clearcell RCC. Informed consent was obtained from all pa-tients involved. The collection and use of tissues wereapproved by the Princess Alexandra Hospital HumanResearch Ethics Committee. The priPTCs were estab-lished from histologically normal segments of kidneysremote from the tumor site. Cell isolations for nor-mal PTCs were performed according to the method ofJohnson et al [22]. Briefly, tissues were decapsulated witha pair of jeweler forceps, then chopped into small frag-ments of about 1 cm3 using scalpels and razor bladesin a Petri dish (on ice). Fragments were washed threetimes in 1:1 mixture of Ham’s F12 and Dulbecco’s mod-ified Eagle’s media (DMEM/F12) (Trace Biosciences,

Melbourne, Australia), glucose concentration 17.5 mmol/L, and digested with 1 mg/mL collagenase type II (Wor-thington “Scimar;” Templestone, Victoria, Australia) for20 to 30 minutes at 37◦C. Digested fragments were sievedthrough an 80 to 100 lm mesh and the reaction washalted by adding ice cold Krebs-Henseleit (KHS)-I solu-tion (6.320 g/L NaCl, 0.363 g/L KCl, 0.389 g/L CaCl2.H2O,0.430 g/L NaH2PO4, and 2.350 g/L NaHCO3). Cells werewashed twice with DMEM/F12 media before suspendingin 45% Percoll in KHS-II (KHS-II is a double strength so-lution of KHS-I) and subjecting to high-speed centrifuga-tion (20,000 rpm for 30 minutes). Tissues were separatedinto four distinctive bands, F1 to F4. The lowermost band,F4 (rich in PTCs), was carefully removed for cultures.

For RCC isolation, tumor tissue was processed as de-scribed above, except that the “F1 layer,” where mostcells were enriched after Percoll separation, was col-lected for cultures. The cells were washed twice withDMEM/F12 and seeded in 75 cm3 tissue cultured flaskat 1 mg tissue/mL in appropriate tissue culture media.Primary cells (priPTC and priRCCs) were checked forappropriate markers (Cam 5.2 and vimentin) to confirmtheir identities. Experiments were performed on secondor third passage primary cells (priPTC and priRCC) whenthe growth rate of cells was most active and fibroblastcontamination was minimal (<0.5%).

Because priRCCs were nearly always obtained frompatients with localized disease, it was thought importantto also examine a cell line established from metastaticRCC, namely SN12K1. The SN12K1 cell line was a giftfrom Dr. Isaiah Fidler (Houston, TX, USA), and was ob-tained from a lung metastasis in a patient with dissemi-nated clear cell RCC.

Cell culture and experimental treatments for cellproliferation studies

SN12K1 RCC cells and priRCCs were maintained in10% fetal calf serum (FCS) in DMEM/F12. PriPTCs weregrown in “defined media” [(DMEM/F12 media supple-mented with 5 lg/mL human apotransferrin (Sigma),5 lg/mL bovine insulin (Sigma), 0.05 lmol/L hydrocor-tisone (Sigma), 10 ng/mL mouse epidermal growth fac-tor (Sigma), and 0.05 lmol/L synthetic prostaglandin E1

(Sigma)]. Cells were plated at a cell density of 4000cells/cm2 (100 lL volumes) in 96-well culture plates. Sub-confluent cells (70% confluent) were made quiescent byincubation for 24 hours in “basic media” (DMEM/F12media supplemented with transferrin). Quiescent cellswere then exposed to basic media containing growth fac-tors and/or antibodies for a further 24 hours prior toanalysis.

Immunohistochemistry

Cells were grown on round 13 mm glass cover-slips (Lomb Scientific, Taren Point, NSW, Australia) in

1274 Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation

24-well plates at a seeding density of 4000 cells/cm2. Qui-escent cells were then cultured in basic media for 24 hours.Media (1 mL) were collected, and centrifuged at 2000rpm to remove cell debris. The supernatants were keptfor Western blotting analyses. Cells were fixed in 4%paraformaldehyde in phosphate-buffered saline (PBS)for 20 to 30 minutes and permeabilized with 0.2% Triton-X-100 in Tris-buffered saline (TBS) at room temper-ature. Endogenous peroxidases were removed by 3%H2O2 in water and washed twice in TBS. Cells wereincubated with the Gropep rabbit antihuman IGF-I an-tibody and Santa Cruz rabbit antihuman IGFBP-3 anti-body in TBS for 30 minutes at 1/50 dilution (20 lg/mLfor IGF-I and 4 lg/mL for IGFBP-3) and were washedtwice with TBS with 3 minutes each. Secondary anti-body incubation was performed with Envision Plus-HRP(horseradish perioxidase) for rabbit antibody for 30 min-utes (Dako, Botany, NSW, Australia). A brown chro-mogenic signal was detected by diaminobenzidine plus(DAB+) (Dako). Cells were then treated with Mayershematoxylin (Sigma), washed once with water, bluedwith 0.25% sodium carbonate, dehydrated with gradedethanol (from 70% to 100%), cleared with xylene, andmounted in depex mounting media.

Thymidine incorporation

To ascertain the effect of IGF-I and IGFBP-3 onpriPTC, priRCC, and SN12K1 DNA synthesis, cellswere incubated in the presence of exogenous IGF-I,IGFBP-3, and/or their antibodies for 24 hours in 96-welltissue culture plates. Appropriate vehicle- and isotypecontrol antibody-treated cells were included for com-parison. After 20 hours of incubation, 10 lL of diluted3[H]-methyl-thymidine was added to the media to makea final concentration of 4 lCi/mL. Media were removed4 hours later at the end of the 24-hour incubation period.Cells were gently washed once by PBS, followed by three20-minute washes in 10% trichloroacetic acid (TCA).They were then washed by methanol, air-dried, and re-suspended in 100 lL of 0.3 mol/L NaOH/1% sodium do-decyl sulfate (SDS). Fifty microliters of the suspensionswere mixed thoroughly with 1 mL of scintillant (Packard,Melbourne, Australia). The radioactivity in samples wascounted in a b-scintillation counter.

3–4,5 dimethylthiazol-2,5 diphenyl tetrazolium bromide(MTT) assays

DNA synthesis results were verified by 3–4,5dimethylthiazol-2,5 diphenyl tetrazolium bromide(MTT) assays, which were performed according to themethods of Denizot and Lang [23] and Twentymanand Luscombe [24]. Following treatment in each of theexperimental groups, media were removed from cellsand 50 lL of 0.5 mg/mL MTT, prepared with phenol-redfree DMEM/F12 media, were added to the wells in

96-well culture plates. The plates were incubated at37◦C for 15 minutes. MTT solutions were removed and100 lL of dimethyl sulfoxide (DMSO) were added.The plates were agitated for 5 minutes to fully dissolvecrystals. Absorbance was measured at 540 nm, andbackground was subtracted at 690 nm by LabsystemsiEMS Reader MF spectrophotometric plate reader andAscent software, version 2.4.1 (Labsystems, Helsinki,Finland).

Western blotting

Western blotting was performed on media conditionedby priPTC, priRCC, and SN12K1 cells incubated in thepresence or absence of IGF-I or vehicle for 24 hours. Priorto Western blotting, protein assays were performed onthe conditioned media to ascertain similar protein lev-els would be loaded. The “microassay” procedure forthe Bio-Rad Protein Assay protocol (Bio-Rad, Hercules,CA, USA) was carried out according to the manufac-turer’s instructions. For Western blotting, samples wereseparated under reducing conditions on a Novex 4% to12% Bis-Tris gel (Invitrogen, Melbourne, Australia) for1 hour at 200 V using 3-[N-morpholino] propane sulfonicacid (MOPS) buffer. Proteins were then transferred ontopolyvinylidine difluoride (PVDF) membranes using asemidry blotting method according to the manufacturer’sinstructions. For immunoblotting, membranes were wet-ted with methanol briefly and rinsed in TBS/0.1% Tween.They were then blotted overnight at 4◦C with 5% skimmilk powder in TBS/0.1% Tween and washed twice withTBS/0.1% Tween. For incubation of primary antibody,rabbit antihuman IGFBP-3 antibody was incubated at1:400 dilution in TBS/0.1% Tween. Washes were per-formed in TBS/0.1% Tween. Secondary antibody incu-bations were performed at 1:3000 in TBS/0.1% Tween.Membranes were incubated in the enhanced chemilu-minescence (ECL Plus Substrate; Amersham Pharma-cia, Castle Hill, NSW, Australia), and then exposed toHyperfilm-enhanced autoradiographic film (AmershamPharmacia).

Statistical analysis

Results were expressed as mean ± SD. All studiesinvolving priPTC and priRCC were performed in sixreplicates from at least four different human donors. Ex-periments involving SN12K1 cells were performed in sixreplicates on at least two separate occasions. Each ex-periment contained internal controls originating fromthe same culture preparation. For the purpose of anal-ysis, each experimental result was expressed as a changefrom the control value, which was regarded as 100%,and analyzed independently. Statistical comparisons be-tween groups were made by one-way analysis of vari-ance (ANOVA). Pair-wise multiple comparisons for sig-nificant ANOVA results were made by Fisher protected

Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation 1275

SN12K1 priPTC priRCC

IGFBP-3

IGF-1

Negativeisotypeantibodycontrol

Fig. 1. Representative results of insulin-like growth factor-I (IGF-I) and IGF binding protein 3 (IGFBP-3) immunostainings on primary proximaltubular cells (priPTC), primary renal cell carcinoma (priRCC), and SN12K1 RCC. Positive stainings appeared as brown discoloration (magnificationswere 200×). IGF-I and IGFBP-3 were strongly expressed in SN12K1 RCC cell lines and weakly expressed in priPTC and priRCC. IGF-I stainingin (A) SN12K1, (B) priPTC,and (C) priRCC. IGFBP-3 staining in (D) SN12K1, (E) priPTC, and (F) priRCC. Negative isotype antibody controlsfor (G) SN12K1, (H) priPTC, and (I) priRCC. Arrows indicate cytoplasmic stainings (Ct), whereas arrowheads represent nuclear stainings (N).

least-significant differences test. Analyses were per-formed using the software package SPSS version 11.0(North Ryde, Australia). P values less than 0.05 were con-sidered significant.

RESULTS

IGF-I and IGFBP-3 were strongly expressed in SN12K1but weakly expressed in priPTCs and priRCC

Immunohistochemical staining of SN12K1 cellsshowed appreciable expression of IGF-I (Fig. 1A) andIGFBP-3 (Fig. 1D). Cell-associated IGFBP-3 and IGF-Iimmunostainings were mainly localized in the cytoplasmand occasionally in nuclei. In contrast, IGF-I and IGFBP-3 were only weakly expressed in priPTC (Fig. 1B and E)and priRCC (Fig. 1C and F).

IGF-I stimulates DNA synthesis in priPTC, priRCC, andSN12K1 RCC

IGF-I at 100 ng/mL significantly stimulated thymidineincorporation by priPTCs (120.7% ± 19.7% of control

values, P < 0.05), priRCCs (238.7% ± 279.9% of controlvalues, P < 0.01), and SN12K1 cells (120.0% ± 22.9% ofcontrol values, P < 0.05) (Fig. 2). The magnitude of IGF-I–induced DNA synthesis compared to vehicle-treatedcontrols was significantly greater in priRCC cells com-pared to priPTCs and SN12K1 RCC cells.

MTT assays were performed to verify the results ofthe thymidine incorporation experiments. SN12K1 cellgrowth was significantly stimulated when incubated withIGF-I for 24 hours, as shown by MTT assay (116.7% ±4.0% of vehicle control, P < 0.01). This result was sup-portive of the DNA synthesis data for SN12K1. The assaywas not performed on priPTC and priRCC due to the rel-atively limited supply of primary cell stocks.

IGF-I immunoneutralization suppresses SN12K1and priPTC growth but not priRCC growth

SN12K1 cells, which exhibited positive immunohisto-chemical staining for IGF-I, exhibited significant suppres-sion of thymidine incorporation in the presence of 10lg/mL anti-IGF-I (81.9% ± 13.5% of controls, P < 0.01)

1276 Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation

0

100

200

300

400

500

600

Thy

mid

ine

inco

rpor

atio

n,%

of c

ontr

ol

priPTC priRCC SN12K1

* *

**

Fig. 2. Thymidine incorporation of insulin-like growth factor-I (IGF-I) treatment in normal proximal tubular cells (PTCs) (N = 21), renalcell carcinomas (RCCs) (N = 30), and SN12K1 RCC (N = 18) cellline cultures. Results were normalized against their own internal ve-hicle controls and expressed as % of control (mean ± SD). Controlvalues (100%) were depicted by the dotted line. The average thymidineincorporation values of IGF-I vs. vehicle treatment (dpm/culture well)before correcting against vehicle controls were 8059 ± 1716 vs. 6600 ±2000 primary PTC (priPTC); 2063 ± 1891 vs. 1103 ± 792 primary RCC(priRCC); and 194515 ± 93368 vs. 160485 ± 54267 SN12K1. ∗P < 0.05vs. control; ∗∗P < 0.01 vs. control and other groups.

0

40

20

60

80

100

120

140

160

Thy

mid

ine

inco

rpor

atio

n,%

of c

ontr

ol

priPTC priRCC SN12K1

* ***

anti-IGF-I

anti-IGF-IR

Fig. 3. Thymidine incorporation of insulin-like growth factor-I (IGF-I)axis intervention. The effect of blocking autocrine IGF-I was examinedby incubation of primary proximal tubular cells (priPTC) (N = 24),primary renal cell carcinoma (priRCC) (N = 24), and SN12K1 RCC(N = 12) with anti-IGF-I or anti-IGF-I receptor (IGF-IR) alone. Thedotted line represented control values observed in isotype antibody-treated cells. Results were expressed as % of control (mean ± SD).∗P < 0.05; ∗∗P < 0.01 vs. isotype antibody control-treated cells.

(Fig. 3) or 10 lg/mL anti-IGF-IR (87.4% ± 16.2% of con-trols, P < 0.05).

Similarly, anti-IGF-I but not anti-IGF-IR significantlysuppressed thymidine incorporation by priPTC (89.7% ±8.3%, P < 0.05; and 93.4% ± 18.2% of controls, P = NS).However, no significant effects were observed in priRCCtreated with the same antibodies(105.0% ± 27.9% and105 ± 35.3% of controls, P = NS).

IGFBP-3 expression in SN12K1 RCC cellsis stimulated by IGF-I

To determine the effect of IGF-I on endogenous pro-duction of IGFBP-3, IGFBP-3 protein expression was

priPTC priRCC SN12K1IGF-I Vehicle IGF-I Vehicle IGF-I Vehicle

38-41 kDIGFBP-3

Fig. 4. The effect of insulin-like growth factor-I (IGF-I) treatment onIGF binding protein 3 (IGFBP-3) protein expression. Western blot ofmedia conditioned by primary proximal tubular cells (priPTC), primaryrenal cell carcinoma (priRCC), and SN12K1 cell incubated in the pres-ence of 100 ng/mL IGF-I or vehicle for 24 hours. IGFBP-3 appearedbetween 38 to 41 kD at reducing conditions of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) protein electrophore-sis. A representative sample and its vehicle-treated control from eachcell type were shown. Western analysis was performed in duplicate setsand both sets showed similar results.

examined in IGF-I–treated priPTC, priRCC, or SN12K1cells by Western blotting. As determined by protein as-says, no significant difference in total protein secretionwas present between IGF-I–treated and vehicle-treatedcells for all cell types. Protein levels between IGF-I–treated and vehicle-treated groups were 1.7 ± 0.9 lg/mLand 2.0 ± 0.03 lg/mL (priPTC, P = NS); 6.3 ± 0.7 lg/mLand 6.0 ± 0.5 lg/mL (priRCC, P = NS); and 10.7 ±7.2 lg/mL and 12.1 ± 8.5 lg/mL (SN12K1, P = NS). Com-pared with vehicle-treated controls, secreted IGFBP3levels were up-regulated 10.7-fold of control follow-ing 24-hour exposure to 100 ng/mL IGF-I (Fig. 4) forSN12K1 cells. IGFBP-3 in the conditioned media existedas 38 to 41 kD doublets under reducing electrophoresisconditions. There was no detectable secreted IGFBP-3for priPTC and priRCC under the same experimentalconditions.

IGFBP-3 inhibits the growth of SN12K1, but notpriPTC or priRCC

As IGFBP-3 might play a part in regulating DNA syn-thesis via IGF-I–dependent or IGF-I–independent ef-fects, we tested this possibility by treating SN12K1 RCC,priPTC, and priRCC with exogenous IGFBP-3 in thepresence or absence of a specific neutralizing antibodyagainst IGFBP-3 (Fig. 5). The amount of exogenousIGFBP-3 (300 ng/mL) used was based on previous datathat primary cultures of cortical fibroblast surroundingpriPTC produced around 266 ± 38 ng/mL/day of IGFBP-3 [25]. The anti-IGFBP-3 antisera concentration used inthis experiment (1:500 dilution) was previously found toneutralize 200 ng/mL of IGFBP-3.

Adding 300 ng/mL exogenous IGFBP-3 did not in-fluence SN12K1 DNA synthesis (Fig. 5). However,

Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation 1277

priPTCpriRCCSN12K1

0

50

100

150

200

300

250

Thy

mid

ine

inco

rpor

atio

n,%

of c

ontr

ol

IGF

BP

-3

anti-

IGF

BP

-3

anti-

IGF

BP

-3+I

GF

BP

-3

*

*#

Fig. 5. DNA synthesis on SN12K1 renal cell carcinoma (RCC) cellline, primary proximal tubular cells (priPTC), and primary renal cellcarcinoma (priRCC) with manipulation of insulin-like growth factorbinding protein 3 (IGFBP-3). SN12K1 cells (N = 6), priPTC (N = 42),and priRCC (N = 35) were incubated for 24 hours in the presence orabsence of 300 ng/mL recombinant human IGFBP-3 (h-IGFBP-3) withand without neutralizing anti-IGFBP-3 antibody (1:500). Results werecorrected against their own internal controls (vehicles) and expressedas % of control (mean ± SD). The dotted line depicts the control values(100%). ∗P < 0.001 vs. control; #P < 0.001 vs. anti-IGFBP-3.

0

50

100

150

200

300

250

Thy

mid

ine

inco

rpor

atio

n,%

of c

ontr

ol

+anti-IGFBP3–anti-IGFBP3

IGF-I Long R3-IGF-I

*** ***###

Fig. 6. Effect of insulin-like growth factor binding protein 3 (IGFBP-3)immunoneutralization on IGF-I- and Long R3-IGF-I–induced SN12K1DNA synthesis (N = 6). Thymidine incorporation was measured inSN12K1 cells exposed for 24 hours to 100 ng/mL IGF-I or Long R3-IGF-I in the presence or absence of anti-IGFBP-3. Both results werecorrected against their own vehicle-treated internal controls and werepresented as mean ± SD. ∗∗∗P < 0.001 vs. the absence of anti-IGFBP3 inIGF-I–treated or Long R3-IGF-I–treated cells; ###P < 0.001 vs. IGF-I+ anti-IGFBP-3.

immunoneutralization of endogenous IGFBP-3 with1:500 dilution of anti-IGFBP-3 significantly stimulatedthymidine incorporation (243.9% ± 35.5% of control,P < 0.001). These stimulatory effects were partially abro-gated by coincubation of cells with 300 ng/mL of exoge-nous IGFBP-3 (135.97% ± 5.9% of control, P < 0.001).Since previous experiments demonstrated that IGF-I in-duced endogenous IGFBP-3 production by SN12K1, thepossibility that IGF-I–induced DNA synthesis was mod-ulated by altered autocrine IGFBP-3 action was further

assessed by IGFBP-3 immunoneutralization. In Figure 6,IGF-I–induced DNA synthesis was further augmented bycoincubation with neutralizing anti-IGFBP-3 antibody.Very similar effects were observed in cells treated with100 ng/mL Long R3-IGF-I (an IGF-I analog which bindsto and stimulates the IGF-I receptor but does not bind toIGFBPs).

In contrast to SN12K1 cells, neither priPTC norpriRCC exhibited significant changes in DNA synthesisfollowing incubation with IGFBP-3 and/or anti-IGFBP-3(Fig. 5).

In keeping with the DNA synthesis data, coincuba-tion of anti-IGFBP3 with 100 ng/mL IGF-I or with100 ng/mL Long R3-IGF-I significantly increased cellgrowth as determined by the MTT assay (IGF-I + anti-IGFBP-3: 112.7% ± 1.9% of IGF-I alone control, P <

0.01; and Long R3-IGF-I + anti-IGFBP-3: 126.9 ± 1.0%of Long R3-IGF-I alone control; P < 0.01, respectively).

DISCUSSION

In this paper, the effects of IGF-I and IGFBP-3 wereevaluated in priPTCs, priRCC, and a metastatic RCC cellline (SN12K1). IGF-I and IGFBP-3 were strongly ex-pressed by SN12K1, but weakly expressed in both priPTCand priRCC. Exogenous IGF-I induced cell growth in allcell types. Furthermore, endogenous IGFBP-3 produc-tion in SN12K1 cells was stimulated by IGF-I and re-sulted in autocrine suppression of RCC growth. Takentogether, these results suggest that IGF-I and IGFBP-3may play important roles, respectively, in stimulating andregulating growth, particularly in its advanced stages.

The present study supports and extends the findings ofprevious in vitro and in vivo studies, which have shownthat IGF-I is a potent mitogen for RCC [5, 9]. IGF-I was found to play an important role in renal cellu-lar transformation induced by proto-oncogenes (such asc-myb) and viral DNA oncogenes (such as SV40 T-antigen) [26]. Moreover, alterations in the cellular tran-scription of IGF/IGFBP family members have beenfound to precede the development of RCC in rats treatedwith ferric nitrilotriacetate [10]. The risk of RCC is alsosubstantially increased in obese individuals [8], and it hasbeen speculated that this may be related to the higherserum levels of IGF-I observed in overweight patients.Biallelic von Hippel-Lindau (VHL) mutations, which oc-cur in more than 70% of sporadic clear cell RCCs, havealso recently been shown to enhance RCC growth and in-vasiveness by augmenting IGF-I signaling [5]. BlockingIGF-I or IGF-IR signaling abolished or delayed the pro-gression of a variety of tumors in animal models [27], andLissoni et al [28] have demonstrated that high pretreat-ment serum levels of IGF-I in patients with clear cell RCCare associated with a substantially impaired response to

1278 Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation

IL-2 immunotherapy. A subcutaneously injected growthhormone releasing hormone (GH-RH) antagonist,MZ-4–71, has been shown to inhibit the IGF-I–inducedproliferation of the Caki-I RCC cell line in mice [9].

All cell types tested in the present study showed re-sponsiveness to exogenous IGF-I stimulation, as indi-cated by significant increases in DNA synthesis. TheIGF-I–induced DNA synthesis in SN12K1 cells was IGF-I–specific and IGF-IR–mediated since it was completelyabrogated by both anti-IGF-I and anti-IGF-IR anti-bodies. Endogenous IGF-I expression, as evidenced byimmunohistochemical staining, was more pronouncedin SN12K1 cells than priRCC. Moreover, IGF-I im-munoneutralization suppressed growth of SN12K1 butnot priRCC. This suggests that autocrine IGF-I actionplays a significant role in SN12K1 growth stimulation,but not in priRCC. The disparity in results between thesetwo cells may relate to the transformation of SN12K1cells, or alternatively, may be due to their difference intumor stage since priRCCs were developed from patientswith localized disease and SN12K1 cells were establishedfrom metastatic RCC.

IGFBP-3, which exerts both IGF-I–dependent andIGF-I–independent actions, has also been implicated inRCC development. Recent studies using cDNA microar-ray, suppression subtractive hybridization and Northernblot analyses have identified IGFBP-3 as one of the com-monly up-regulated genes in pathologic samples of clearcell RCC and established RCC cell lines [29, 30]. Thefunctional consequences IGFBP-3 in normal cell sys-tems and malignancies are complex and could poten-tially involve both stimulatory and inhibitory effects [6,20, 31, 32]. In the present study, IGFBP-3 was clearlyup-regulated in SN12K1 cells and appeared to act in anautocrine action to suppress cell growth.

The significance of IGFBP-3 in the regulation of RCCprogression is not well understood, but we hypothesizedthat its up-regulation may have represented a compen-satory inhibitory mechanism in response to unrestrainedIGF-I–induced cell growth. In SN12K1 cells, IGF-I stimu-lated DNA synthesis concurrently with IGFBP-3 produc-tion (Figs. 2 and 4). Similar studies by Grill and Cohick[33] have also shown that IGF-I–induced production ofIGFBP-3 in an immortalized bovine mammary epithelialcell line maintained a regulatory growth loop. The authorsclaimed that an increase in IGFBP-3 enhanced cellularresponsiveness to IGF-I, as reflected by “mock cells” (de-void of IGFBP-3) having less intense IGF-I–stimulatedcell growth than IGFBP-3–expressing cells. While we ob-served similar increases in DNA synthesis and IGFBP-3 production with IGF-I incubation (Figs. 2 and 4), theIGF-I–enhanced DNA synthesis was accentuated whenIGFBP-3 was removed (Fig. 6). To further evaluate if thiseffect was dependent upon extracellular sequestration ofIGF-I, Long R3-IGF-I, an analog of IGF-I, was coincu-

bated with neutralizing IGFBP-3 antibody. IGFBP-3 im-munoneutralization significantly augmented the stimula-tory effects of both IGF-I and Long R3-IGF-I on SN12K1cells DNA synthesis, suggesting that the inhibitory au-tocrine effects of IGFBP-3 were predominantly mediatedby IGF-I–independent mechanisms. Nevertheless, LongR3-IGF-I exerted a marginally greater mitogenic effectthan an equipotent dose of IGF-I, indicating that a smallcomponent of the antiproliferative effects of IGFBP-3may have involved extracellular sequestration of IGF-I.

Nevertheless, it is difficult to determine whether theIGFBP-3 effects demonstrated in this paper were solelyIGF-I–dependent or IGF-I–independent (or a combina-tion of both) as SN12K1 cells do express IGF-I, whichmay interfere with IGFBP-3. The classical theories ofIGF-I–dependent effects proposed for IGFBP-3 includepotentiation of IGF-I action by IGFBP-3 [11, 12] and in-hibition of IGF-I action by sequestration of extracellularpeptides [20]. In addition, IGFBP-3 has been reported toexert a number of IGF-I–independent effects, includingthe inhibition of IGF-IR autophosphorylation and tyro-sine kinase activity [34], signaling via cell surface typeV transforming growth factor-b (TGF-b) receptors forIGFBP-3 [14, 35], mitogen-activated protein (MAPK) ac-tivation [36], up-regulation of kinase phosphatases [17],activation of growth inhibitory TGF-b , and Smad path-ways [18, 19].

In the present study, IGF-I and IGFBP-3 immunore-activity were present in the cytoplasm and nuclei, a find-ing that is similar to that in lung cancer [37, 38]. Thesignificance of this in RCC is uncertain. However, nu-clear transport studies by Li et al [39] indicate that twosignaling pathways may be possible. First, cotransporta-tion of IGF-I and IGFBP-3 complex across the plasmamembrane could lead to nuclear accumulation of the pep-tides, resulting in direct nuclear transcription of factorsinvolved in cell growth. Alternatively, IGF-I and IGFBP-3 could be internalized into the endosomal compartmentsvia receptor-mediated pathways, where their subsequentaction may be mediated by cytoplasmic secondary mes-senger systems.

CONCLUSION

IGF-I and IGFBP-3 were most highly expressed inSN12K1 cells, a cell line developed from RCC metas-tasis. In these cells, IGF-I acted in an autocrine fashionto stimulate the growth of renal cell carcinomas and theirsecretion of IGFBP-3. IGFBP-3 in turn served to recipro-cally inhibit cellular growth, acting as a brake for endoge-nous IGF-I action. Further investigation of the potentialtherapeutic value of inhibiting IGF-I and/or promotingIGFBP-3 action in animal RCC models is warranted.

Cheung et al: IGF-I and IGFBP-3 regulate renal cell carcinoma proliferation 1279

ACKNOWLEDGMENTS

We thank Ms. Shirley Callaghan from the Tissue Tumor Bank of De-partment of Surgery, the University of Queensland, for collection ofrenal tissues, Dr. Isaiah Fidler for donating the SN12K1 RCC cell line,and Professor Rob Baxter for donating the anti-IGFBP-3 neutralizingantibody. We also extend our gratitude to Ms. Stephanie Sermon andMr. Dominic Poon for technical assistance. This work has beenfunded by the Princess Alexandra Hospital Research and DevelopmentFoundation and the National Health and Medical Research Council(NHMRC). Catherine Cheung is supported by a NHMRC Dora Lush(Biomedical) Postgraduate Research Scholarship (App ID 143114).

Reprint requests to Associate Professor David Johnson, M.D., Depart-ment of Renal Medicine, Princess Alexandra Hospital, Ipswich Road,Woolloongabba, Brisbane, Queensland, 4102, Australia.E-mail: david johnson@health.qld.gov.au

REFERENCES

1. KOVACS G, AKHTAR M, BECKWITH BJ, et al: The Heidelberg classifi-cation of renal cell tumours. J Pathol 183:131–133, 1997

2. RUSSO P: Localized renal cell carcinoma. Curr Treat Options Oncol2:447–455, 2001

3. TIAN GG, DAWSON NA: New agents for the treatment of renal cellcarcinoma. Expert Rev Anticancer Ther 1:546–554, 2001

4. STROHMAIER WL: New treatment modalities—The urologist’s view.Anticancer Res 19:1605–1609, 1999

5. DATTA K, NAMBUDRIPAD R, PAL S, et al: Inhibition of insulin-likegrowth factor-I-mediated cell signaling by the von Hippel-Lindaugene product in renal cancer. J Biol Chem 275:20700–20706, 2000

6. GRIMBERG A, COHEN P: Role of insulin-like growth factors and theirbinding proteins in growth control and carcinogenesis. J Cell Physiol183:1–9, 2000

7. BASERGA R, HONGO A, RUBINI M, et al: The IGF-I receptor incell growth, transformation and apoptosis. Biochem Biophys Acta1332:F105–F126, 1997

8. CHOW WH, GRIDLEY G, FRAUMENI JF, JARVHOLM B: Obesity, hy-pertension and the risk of kidney cancer in men. N Engl J Med343:1305–1311, 2000

9. JUNGWIRTH A, SCHALLY AV, PINSKI J, et al: Growth hormone-releasing hormone antagonist MZ-4–71 inhibits in vivo proliferationof Caki-I renal adenocarcinoma. Proc Natl Acad Sci USA 94:5810–5813, 1997

10. TANAKA T, KONDO S, IWASA Y, et al: Expression of stress-responseand cell proliferation genes in renal cell carcinoma induced by ox-idative stress. Am J Pathol 156:2149–2157, 2000

11. CONOVER CA: Potentiation of insulin-like growth factor (IGF) ac-tion by IGF-binding protein-3: Studies of underlying mechanism.Endocrinology 130:3191–3199, 1992

12. RAJAH R, KATZ L, NUNN S, et al: Insulin-like growth factor bindingprotein (IGFBP) proteases: Functional regulators of cell growth.Prog Growth Factor Res 6:273–284, 1995

13. HODGSON DR: Free-ligand accelerated dissociation of insulin-likegrowth factor 1 (IGF-1) from the type I IGF receptor is reduced byinsulin-like growth factor binding protein 3. Regul Pept 90:33–37,2000

14. LEAL SM, LIU Q, HUANG SS, HUANG JS: The type V transform-ing growth factor beta receptor is the putative insulin-like growthfactor-binding protein 3 receptor. J Biol Chem 272:20572–20576,1997

15. PRATT SE, POLLAK MN: Insulin-like growth factor binding protein3 (IGF-BP3) inhibits estrogen-stimulated breast cancer cell prolif-eration. Biochem Biophys Res Commun 198:292–297, 1994

16. LEE HY, CHUN KH, LIU B, et al: Insulin-like growth factor bindingprotein-3 inhibits the growth of non-small cell lung cancer. CancerRes 62:3530–3537, 2002

17. RICORT JM, BINOUX M: Insulin-like growth factor-binding protein-3 activates a phosphotyrosine phosphatase. Effects on the insulin-like growth factor signaling pathway. J Biol Chem 277:19448–19454,2002

18. FANAYAN S, FIRTH SM, BAXTER RC: Signaling through the Smad

pathway by insulin-like growth factor- binding protein-3 in breastcancer cells. Relationship to transforming growth factor-beta 1 sig-naling. J Biol Chem 277:7255–7261, 2002

19. FANAYAN S, FIRTH SM, BUTT AJ, BAXTER RC: Growth inhibition byinsulin-like growth factor-binding protein-3 in T47D breast cancercells requires transforming growth factor-beta (TGF- beta) and thetype II TGF-beta receptor. J Biol Chem 275:39146–39151, 2000

20. GRIMES RW, HAMMOND JM: Insulin and insulin-like growth factors(IGFs) stimulate production of IGF-binding proteins by ovariangranulosa cells. Endocrinology 131:553–558, 1992

21. MARTIN JL, BAXTER RC: Antibody against acid-stable insulin-like growth factor binding protein detects 150,000 mol wt growthhormone-dependent complex in human plasma. J Clin EndocrinolMetab 61:799–801, 1985

22. JOHNSON DW, SAUNDERS HJ, BAXTER RC, et al: Paracrine stimula-tion of human renal fibroblasts by proximal tubule cells. Kidney Int54:747–757, 1998

23. DENIZOT F, LANG R: Rapid colorimetric assay for cell growth andsurvival. Modifications to the tetrazolium dye procedure giving im-proved sensitivity and reliability. J Immunol Methods 89:271–277,1986

24. TWENTYMAN PR, LUSCOMBE M: A study of some variables in a tetra-zolium dye (MTT) based assay for cell growth and chemosensitivity.Br J Cancer 56:279–285, 1987

25. JOHNSON DW, SAUNDERS HJ, BREW BK, et al: Human renal fibrob-lasts regulate proximal tubule cell growth and transport via the IGF-I axis. Kidney Int 52:1486–1496, 1997

26. PORCU P, FERBER A, PIETRZKOWSKI Z, et al: The growth-stimulatoryeffect of Simian virus 40 T antigen requires the interaction of insulin-like growth factor 1 with its receptor. Mol Cell Biol 12:5069–5077,1992

27. LEE CT, WU S, GABRILOVICH D, et al: Antitumor effects of an aden-ovirus expressing antisense insulin-like growth factor I receptor onhuman lung cancer cell lines. Cancer Res 56:3038–3041, 1996

28. LISSONI P, BARNI S, ARDIZZOIA A, et al: Clinical efficacy of can-cer subcutaneous immunotherapy with interleukin-2 in relation tothe pretreatment levels of tumor growth factor insulin-like growthfactor-1. Tumorigenesis 81:261–264, 1995

29. STASSAR MJ, DEVITT G, BROSIUS M, et al: Identification of humanrenal cell carcinoma associated genes by suppression subtractivehybridization. Br J Cancer 85:1372–1382, 2001

30. TAKAHASHI M, RHODES DR, FURGE KA, et al: Gene expression pro-filing of clear cell renal cell carcinoma: Gene identification and prog-nostic classification. Proc Natl Acad Sci USA 98:9754–9759, 2001

31. BAXTER RC: Insulin-like growth factor (IGF)-binding proteins: In-teractions with IGFs and intrinsic bioactivities. Am J Physiol En-docrinol Metab 278:E967–E976, 2000

32. JONES JI, CLEMMONS DR: Insulin-like growth factors and their bind-ing proteins: Biological actions. Endocr Rev 16:3–34, 1995

33. GRILL CJ, COHICK WS: Insulin-like growth factor binding protein-3 mediates IGF-I action in a bovine mammary epithelial cell lineindependent of an IGF interaction. J Cell Physiol 183:273–283, 2000

34. RICORT JM, BINOUX M: Insulin-like growth factor (IGF) bindingprotein-3 inhibits type 1 IGF receptor activation independently ofits IGF binding affinity. Endocrinology 142:108–113, 2001

35. LEAL SM, HUANG SS, HUANG JS: Interactions of high affinity insulin-like growth factor-binding proteins with the type V transforminggrowth factor-beta receptor in mink lung epithelial cells. J BiolChem 274:6711–6717, 1999

36. LEE DY, YI HK, HWANG PH, OH Y: Enhanced expression of insulin-like growth factor binding protein-3 sensitizes the growth inhibitoryeffect of anticancer drugs in gastric cancer cells. Biochem BiophysRes Commun 294:480–486, 2002

37. PILICHOWSKA M, KIMURA N, FUJIWARA H, NAGURA H: Immunohis-tochemical study of TGF-alpha, TGF-beta1, EGFR, and IGF-1 ex-pression in human breast carcinoma. Mod Pathol 10:969–975, 1997

38. JAQUES G, NOLL K, WEGMANN B, et al: Nuclear localization of insulin-like growth factor binding protein 3 in a lung cancer cell line. En-docrinology 138:1767–1770, 1997

39. LI W, FAWCETT J, WIDMER HR, et al: Nuclear transport of insulin-like growth factor-I and insulin-like growth factor binding protein-3in opossum kidney cells. Endocrinology 138:1763–1766 1997