Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Am J Obstet Gynecol. Author manuscript; available in PMC 2013 July 11.
Published in final edited form as:
PMCID: PMC3709016

Secreted protein acidic and rich in cysteine as a regulator of murine ovarian cancer growth and chemosensitivity



Secreted protein acidic and rich in cysteine (SPARC) influences the growth of several solid tumors. Our objectives were to determine the effect of SPARC on the growth and response to cisplatin therapy of platinum-resistant ovarian cancer.


SPARC expression was determined in 4 platinum-resistant ovarian cancer cell lines. The effect of increasing SPARC on cell proliferation was determined in vitro. The effect of host-derived SPARC on tumor growth and response to therapy was determined in vivo using the murine ovarian cancer cell line, OSEID8, which was injected into the peritoneum of wild-type (WT) and SPARC-null (SP−/−) mice.


Forced expression of SPARC decreased growth of platinum-resistant ovarian cancer cell lines in vitro. In vivo, tumor growth was more aggressive in the absence of host-derived SPARC resulting in decreased survival compared with WT mice (P = .005). Cisplatin did not improve survival of WT mice. In contrast, cisplatin therapy resulted in a significant survival advantage (P = .0048) and decreased tumor volume (P = .02) in SP−/− animals.


We conclude that SPARC is an important extracellular matrix protein that regulates the growth and chemosensitivity of ovarian cancer. In general, SPARC appears to control tumor cell growth but also impede the efficacy of cisplatin therapy. Therefore, selective inhibition of SPARC may provide an attractive strategy for increasing the efficacy of therapy in platinum-resistant ovarian tumors.

Keywords: Chemosensitivity, extracellular matrix, murine, ovarian cancer, secreted protein acidic and rich in cysteine

Secreted protein acidic and rich in cysteine (SPARC) is a 32-kDa calcium binding glycoprotein. It is antiadhesive and antiproliferative, and it binds to many extracellular matrix (ECM) proteins.1,2 In this capacity, SPARC has been shown to have tumor-suppressor effects and significant function in the host response to tumor development.2,3 SPARC expression is altered in many cancers. For instance, there is increased expression of SPARC in melanoma, glioma, colorectal, and breast carcinomas compared with their respective normal tissues.4 In these tumors, high levels of SPARC are often associated with enhanced invasion and metastasis.4 In contrast, expression of SPARC by pancreatic carcinoma cells is limited because of promoter hypermethylation, whereas stromal cells in the tumor express increased levels of SPARC.5 Whether produced by tumor cells or stromal (host) cells, SPARC is often found at tumor-stromal interfaces; in tumor capsules; and in areas of desmoplasia, angiogenesis, and vascular remodeling.6 Thus, the context of SPARC expression in the microenvironment is critical for understanding its influence on cancer growth.

Previously epithelial ovarian cancer cells have been shown to express low levels of SPARC, whereas adjacent stroma expresses higher levels.2,7,8 Mok et al9 published 1 of the earliest studies on the function of SPARC in ovarian cancer. They found that SPARC-expressing transfectants had reduced growth in vivo compared with control cells. Furthermore, Yiu et al2 demonstrated that forced expression of SPARC or addition of recombinant SPARC induced apoptosis of ovarian cancer cell lines in vitro. A more recent study by Said and Motamed3 evaluated the effect of host-derived SPARC on ovarian cancer growth, in vivo, demonstrating more rapid and aggressive tumor growth in SPARC-deficient animals.

SPARC has been shown to modulate response to chemotherapy. A recent study by Tai et al10 demonstrated that, in colorectal cancer, addition of exogenous SPARC and/or forced expression of endogenous SPARC improved tumor cell response to chemotherapy in vitro. The authors also found in vivo evidence to suggest that SPARC can sensitize colon cancer cells to chemotherapy.10

These findings were validated by demonstration of enhanced apoptosis in colon cancer cells and decreased blood vessel formation in tumors after exposure to or forced expression of SPARC.10 If and how SPARC affects chemosensitivity of ovarian cancer is unknown. However, given these findings, SPARC is well placed to participate in the host response to tumor progression and therapy of ovarian cancer.

Epithelial ovarian cancer is considered a chemoresponsive disease.11 Initial response rates to cytoreductive surgery followed by platinum therapy are as high as 80%; whereas, with advanced stage disease, survival is poor because of recurrence and drug resistance.11 Patients who progress or recur 6 months after primary therapy are considered platinum resistant.12,13 Standard second-line treatments are limited and include various single-agent chemotherapies.13 More recent clinical trials have looked at biologics and antiangiogenic therapies. Few clinical trials have looked at ways to overcome drug resistance to primary platinum therapy or considered host pharmacogenomic pathways.11

Our primary objective was to determine the effect of modulating the level of SPARC on the growth of known platinum-resistant ovarian cancer. Our secondary objective was to determine whether SPARC modifies chemosensitivity of platinum-resistant ovarian cancer.

Materials and Methods

Cell culture and reagents

Platinum-resistant human ovarian cancer cell lines SKOV3, ES2 (American Type Culture Collection, Manassas, VA), and HCC60 from the Hamon Center for Therapeutic Oncology Research (gift of Dr John D. Minna) and a platinum-resistant murine ovarian cancer cell line (OSEID8)14 obtained from Dr Paul F. Terranova (University of Kansas Medical Center, Kansas City, KS) were used. All cell lines were grown in RPMI 1640 (Invitrogen Corp, Carlsbad, CA) or Dulbecco’s modified Eagle’s medium (Invitrogen) supplemented with 5–15% fetal bovine serum (Gemini Bio-Products, Woodland, CA), and 1% antibiotic/antimycotic (Invitrogen) at 37°C (5% CO2). The mouse monoclonal anti-body (MAb) 303 was produced and purified as described previously.15 MAb303 has been shown to inhibit SPARC-induced activation of AKT in glioma cells.16

SPARC expression

Cells at 80–90% confluence were harvested along with the corresponding cell-conditioned media. Cell lysates were extracted from cell pellets using lysis buffer containing 2.5 mM of 1 M Tris at pH 7.4, 1% Nonidet P-40, 50 mM NaCl, 1 mM EDTA, 0.1% sodium dodecyl sulfate, protease inhibitors (Complete Mini; Roche Diagnostics, Indianapolis, IN), 0.5 M NaF, 20 mM Na3VO4, and 1 M β-glycerophosphate. Equal volumes of cell-conditioned media, consisting of supernatant obtained from plates with 80–90% confluence, was used to evaluate secreted SPARC levels. Western blotting was performed as described previously17 using MAb303 as primary antibody. Signal was developed by incubating with a horseradish peroxidase-conjugated anti-mouse secondary antibody and chemiluminescent substrate (Pierce, Rockford, IL) prior to exposure to X-ray film (Phenix Research Products, Hayward, CA). The level of β-actin was used as the loading control, and recombinant SPARC protein was used as a positive control.

Tumor-associated SPARC modulation

Transient expression of SPARC was induced by transfection with pCMV-SPORT6 plasmid (Invitrogen Corp) containing complementary deoxyribonucleic acid (cDNA) for human SPARC. Cells were also transfected with the empty vector as a control. Western blot was performed at 72 hours to determine SPARC expression in the transfected cells. After harvesting transfected cells, direct cell counts (at 24, 48, and 72 hours) and liquid culture colony formation assays (at 14 days) were performed to determine the functional impact of endogenous SPARC on ovarian cancer cell growth. Two independent experiments were performed in duplicate 6-well plate format.

Chemosensitivity testing with SPARC modulation

In vitro

To determine the effect of SPARC on cisplatin sensitivity, cells at 80–90% confluence were trypsinized, pelleted, resuspended in media, and counted. Cells were seeded at a density of 1–2 × 103 cells/well in 96-well plates and allowed to attach for 24 hours, after which they were treated with cisplatin, MAb303, or a combination of the 2, in the presence of 10% serum. Six days after treatment, MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfo-phenyl)-2H)- tetrazolium reagent] (Promega, Madison, WI) was added to each well.

After an incubation of 4 hours, absorbance was read using the Emax Precision microplate reader (Molecular Devices Corp, Sunnyvale, CA) at 490 nm. Cell growth was normalized to untreated cells. SoftMax Pro 4.7.1 (Molecular Devices) converted the absorbance values into cell concentration compared with control. The concentration of cisplatin resulting in 50% inhibition (IC50), in the presence or absence of MAb303, was calculated. All cell proliferation assays were performed in replicates of 8, using 96-well plates in 2 independent experiments.

In vivo

All experiments were carried out in accordance with University of Texas Southwestern Institutional Animal Care and Use Committee Policies and Guidelines. Then 5 × 106 OSEID8 cells14 were injected intraperitoneally into female SPARC null (SP−/−) C57BL/6 mice (n = 12), generated as described by Norose et al,18 and age-matched, 8-week old wild-type (WT) C57BL/6 female mice (n = 12; Charles River Laboratories, Wilmington, MA).

Both the SP−/− and the WT groups were divided equally into observation and treatment arms, resulting in a total of 4 arms with 6 animals in each arm. After observation for 41 days, the treatment group (n = 6/genotype) was given 5 mg/kg cisplatin, intraperitoneally, weekly for 4 weeks. Animals were observed for evidence of disease (ascites, body weight, performance status) and killed for decreased performance status or recurrent ascites. Tumor location and volumes (cubic millimeters) were determined after the animals were killed by direct visualization and measurements of length × width × thickness were compared. Mice were observed for 117 days and survival was documented as time to the animals were killed or death.

Statistical analysis

Statistical analysis included Student t test for growth and volume comparisons with alpha set at 0.05. Kaplan-Meier survival curves were compared using Mantel-Cox log-rank test with alpha set at 0.05. Calculations, figures, and tables were produced with Graph Pad Prism version 5.00 (GraphPad Software, San Diego, CA).


SPARC expression

SPARC was variably expressed in the cell lines tested. Two of the platinum-resistant human ovarian cancer cell lines (ES2 and HCC60) and the platinum-resistant murine ovarian cancer cell line (OSEID8) showed expression of cellular SPARC in cell lysates (Figure 1, A). When testing conditioned media, we found 2 of the cell lines (ES2 and OSEID8) unequivocally secreted SPARC (Figure 1, B). There was little or no detectable SPARC in the conditioned media from HCC60 cells and questionable SPARC secretion in SKOV3, as compared with the control media lane. Although MAb303 reacts preferentially with mouse and human SPARC, there is some cross-reactivity with bovine SPARC as shown in the media/fetal bovine serum lane (Figure 1, B). Despite performing 2 separate experiments in duplicate, we cannot conclusively state that SKOV3 is or is not a SPARC secretor.

SPARC is variably expressed by platinum resistant ovarian cancer cells in vitro

Tumor-associated SPARC modulation

To determine the effect of increasing SPARC expression in ovarian cancer, we transfected all the cell lines with a SPARC cDNA and empty vector plasmid. Direct cell counts showed that forced expression of SPARC had a varied effect on cell growth at 24 and 48 hours. At 72 hours, however, all cell lines showed decreased growth compared with controls transfected with empty vector, with a significant difference seen in OSEID8 and ES2 cells (Figure 2, A; P = .02 and P = .004, respectively). HCC60 and SKOV3 showed decreased growth after transfection with the SPARC plasmid; however the difference was not significant (P = .6 and P = .5, respectively), with error bars overlapping for HCC60. Western blot analysis confirmed transfection in all cell lines at 72 hours after transfection (Figure 2, B and data not shown). We con-firmed that forced expression of SPARC significantly reduced growth in the OS-EID8 and ES2 cells by the use of a colony-formation assay (Figure 2, C). In summary, transfection with SPARC cDNA decreased ovarian tumor cell growth.

Forced expression of tumor-associated SPARC reduces cell proliferation

Chemosensitivity testing with SPARC modulation

In vitro

To determine what effect altering SPARC had on the efficacy of cisplatin, we treated cells with cisplatin alone or in combination with MAb303. MAb303 has been shown to block SPARC-induced survival of glioma cells.16 Direct treatment with MAb303 produced no significant growth inhibition. Treatment with cisplatin confirmed the relative resistance of the cell lines, with IC50s ranging from 1.79 μM to 10.68 μM. Combining MAb303 with cisplatin resulted in no significant decrease in the IC50 for cell lines ES2, HCC60, and SKOV3. However, treatment of OSEID8 cells with MAb303 and cisplatin resulted in a significant (P = .002) decrease in the IC50 of cisplatin (Table).

In vivo

To determine the effect of host-derived SPARC on the chemosensitivity of OS-EID8 cells in vivo, we performed tumor studies in WT and SP−/− animals. Overall tumor progression was enhanced in untreated control SP−/− compared with WT animals as evidenced by increased tumor volume and reduced survival at time the animals were killed (P = .19 and P = .005, respectively; Figures 3 and and44).

Tumor growth and response to cisplatin therapy
Cisplatin improves survival in tumor-bearing SP−/− mice

Therapy with cisplatin was initiated on day 41 after tumor cell injection at which time 4 of the SP−/− but none of the WT mice presented with ascites (Figure 3, A). Ascitic fluid was removed aseptically from these animals before therapy was administered. The untreated control WT and SP−/− mice were observed and at last follow-up, 5 of 6 SP−/− and 3 of 6 WT mice had developed ascites (Figure 3, B). Cisplatin therapy had little effect on tumor burden or ascites development in WT mice (Figure 3, B and C), consistent with OSEID8 being a platinum-resistant cell line. In contrast, SP−/− animals responded to therapy. In SP−/− animals, cisplatin significantly reduced tumor volume (Figure 3, C; P = .02, SP−/− control vs treatment) and reduced the incidence of ascites from 83% (5/6) in the control animals to 50% (3/6) in the cisplatin-treated animals (Figure 3, B; P = .55, SP−/− control vs treatment).

The effect of the absence of host-derived SPARC on tumor progression and therapy was also evident by monitoring survival curves (Figure 4). Untreated SP−/− mice displayed a significant decrease in overall survival (P = .0048) compared with treated SP−/− mice (Figure 4, A). Treatment with cisplatin improved the mean survival time by approximately 30 days in SP−/− mice (Figure 4, A). In contrast, cisplatin therapy had no significant effect on the survival of WT mice (Figure 4, B). Additionally, untreated SP−/− mice had a significantly shorter survival than WTcontrol mice (P = .0005; Figure 4, C). Therefore, SPARC deficient mice with platinum-resistant ovarian tumors had worse survival and greater tumor burden compared to WT mice; however, they had significantly improved survival and decreased tumor burden, when treated with cisplatin therapy, compared with untreated SPARC-deficient mice.


The major findings to emerge from this study are: (1) increased expression of SPARC by ovarian tumor cells inhibits cell growth, in vitro; (2) host-derived SPARC may reduce tumor volume and improve survival in the absence of therapy; (3) cisplatin therapy is more efficacious in tumor-bearing SPARC-deficient animals than wild-type mice.

SPARC is a matricellular glycoprotein that influences tumor cell interaction with the extracellular matrix.1,2 Differentiating the function of tumor-associated and host-derived SPARC is a challenge given that SPARC can be secreted by and can act on many different cells in the tumor 4microenvironment. Some studies have found tumor-associated SPARC to be an important modulator of tumor progression and metastasis,19 whereas others have found host-derived SPARC to have a greater influence.1,20

We found that increasing tumor-associated SPARC diminished ovarian cancer cell growth in vitro. Even in cell lines already expressing SPARC, forcing additional expression, the results were a decrease in cell growth. This is consistent with previous findings by Yiu et al.2 Cell lines overexpressing SPARC have significantly decreased growth with forced induction of SPARC, suggesting a potential for further SPARC induction specifically in SPARC producing ovarian cancers.

Attempts to block SPARC activity in the murine cell line OSEID8 with MAb30315 resulted in improved response to cisplatin treatment, but it had little effect on improving the cisplatin response in the 3 human cell lines. The mechanism(s) underlying the enhanced effect of MAb303 and cisplatin is unclear. Shi et al15 found that MAb303 reduced SPARC-induced activation of AKT in glioma cells. This, however, is counterintuitive, given that increased expression of SPARC by OSEID8 cells reduces cell number rather than promotes cell survival as would be expected if AKT activation was the primary driver of SPARC effects in ovarian cells.

Another intriguing possibility is SPARC-induced modulation of cell shape or adhesion via inhibition of integrin-collagen interaction. It is possible that MAb303, by binding to SPARC, perturbs OSEID8 cell adhesion, resulting in enhanced cisplatin-mediated cell killing. This is supported by in vitro studies demonstrating that MAb303 can induce cell rounding and deadhesion.14 The use of additional platinum-resistant human ovarian cancer cell lines and other anti-SPARC antibodies such as AON-1, MAb255, or MAb29314 may provide different effects.

To evaluate the effect of host-derived SPARC on chemoresistant ovarian tumor growth and response to chemotherapy, we compared the growth of OSEID8 cells in WT and SP−/− animals. Our results are similar to the studies of Said and Motamed,3 the results of which showed increased tumor burden in the absence of SPARC. In particular, we found that SP−/− animals had a decreased time to formation of ascites and a larger tumor burden.

Importantly, we show for the first time, in vivo, that platinum-resistant ovarian tumors have increased sensitivity to chemotherapy in the absence of host-derived SPARC. The mechanism by which SPARC affected the sensitivity of these tumor cells remains to be elucidated. Interestingly, because SPARC is known to participate in ECM deposition, it is postulated that the delivery of chemotherapeutic agents to tumor cells themselves may be improved in the absence of SPARC.

The use of cell lines is controversial because of molecular changes that may occur between generations. Because our goal did not depend on the parent cell lines’ molecular characteristics, we did not search for alterations from the original generation. Additionally, although SPARC is a protein often found in the stroma, we chose to study cell lines to examine cell associated and secreted levels of SPARC and to determine the comparative platinum resistance between the murine cell line, OSEID8, and known platinum-resistant human OSE cell lines.

Additional limitations of this study include the use of MAb303, which can bind bovine serum in the media to a small extent, and the small cohort of animals used for the in vivo experiment. The use of a transient transfection protocol for inducing SPARC expression prevented evaluation of the effect of altered endogenous SPARC on in vitro chemosensitivity, which would require a transfection that could persist several days.

Questions needing further study include examining whether forced expression or knockdown of SPARC is functionally equivalent to the natural expression pattern of SPARC in ovarian cancer and whether SPARC expression correlates with chemosensitivity. Future studies may include transfection with a stable intracellular modifier of SPARC, development of alternative SPARC modulators, and inoculation with human ovarian cancer cell lines for the in vivo model.

Cisplatin resistance in ovarian cancer may be related to alterations in excision repair proteins, decreased drug accumulation because of alteration in cellular transport, and increased capacity to tolerate DNA damage.21,22 Although our study does not elucidate mechanisms by which chemosensitivity is induced in the absence of host SPARC, it does underscore the importance of incorporating the interplay between tumor and host when studying ovarian cancer growth and response to chemotherapy. Our results and those of others reinforce the fact that SPARC is an important factor in regulating ovarian tumor growth.

Our study suggests that SPARC may also modulate the chemosensitivity of ovarian cancer, either as the primary modulator or an effector of an undiscovered modulator. It would be interesting to determine the effect SPARC has on those factors already known to alter chemoresistance, such as tumor suppressor gene products or the cellular matrix action on drug uptake and distribution. Future therapeutic approaches may include immunotherapy directed at inactivating host-SPARC. Further studies are warranted to determine the mechanisms by which SPARC acts. Selective inhibition of SPARC may provide an attractive strategy for increasing the efficacy of chemotherapy in platinum-resistant ovarian tumors.


SPARC is an important factor in regulating ovarian tumor growth. SPARC may also modulate the chemosensitivity of ovarian cancer, either as the primary modulator or an effector of an undiscovered modulator. SelectiveinhibitionofSPARCmayprovide an attractive strategy for increasing the efficacy of chemotherapy in platinum-resistant ovarian tumors.

Combination of anti-SPARC MAb303 with cisplatin reduced the in vitro IC50 of cisplatin in OSEID8 cells


This study was supported in part by the Reproductive Scientist Development Program (00849HD12K5, to J.S.L.), the Effie Marie Cain Scholarship in Angiogenesis Research (to R.A.B.), and a training Grant from the National Institutes of Health (T32 GM007062, to S.A.).

We gratefully acknowledge the receipt of cell lines OSEID8 from Dr Paul Terranova (University of Kansas Medical Center) and HCC60 from Dr John Minna (University of Texas Southwestern Medical Center).


Presented in part at the 35th Annual Meeting of the Western Association of Gynecologic Oncologists, Lake Tahoe, CA, May 3l-June 3, 2006, and at the 38th Annual Meeting on Women’s Cancer of the Society of Gynecologic Oncologists, San Diego, CA, March 3–7, 2007.


1. Brekken RA, Puolakkainen P, Graves DC, Workman G, Lubkin SR, Sage EH. Enhanced growth of tumors in SPARC-null mice is associated with changes in the ECM. J Clin Invest. 2003;111:487–95. [PMC free article] [PubMed]
2. Yiu GK, Chan WY, Ng SW, et al. SPARC (secreted protein acidic and rich in cysteine) induces apoptosis in ovarian cancer cells. Am J Pathol. 2001;159:609–22. [PubMed]
3. Said N, Motamed K. Absence of host-secreted protein acidic and rich in cysteine (SPARC) augments peritoneal ovarian carcinomatosis. Am J Pathol. 2005;167:1739–52. [PubMed]
4. Framson PE, Sage EH. SPARC and tumor growth: where the seed meets the soil? J Cell Biochem. 2004;92:679–90. [PubMed]
5. Sato N, Fukushima N, Maehara N, et al. SPARC/osteonectin is a frequent target for aberrant methylation in pancreatic adenocarcinoma and a mediator of tumor-stromal interactions. Oncogene. 2003;22:5021–30. [PubMed]
6. Koukourakis MI, Giatromanolaki A, Brekken RA, et al. Enhanced expression of SPARC/osteonectin in the tumor-associated stroma of non-small cell lung cancer is correlated with markers of hypoxia/acidity and with poor prognosis of patients. Cancer Res. 2003;63:5376–80. [PubMed]
7. Brown TJ, Shaw PA, Karp X, Huynh MH, Begley H, Ringuette MJ. Activation of SPARC expression in reactive stroma associated with human epithelial ovarian cancer. Gynecol Oncol. 1999;75:25–33. [PubMed]
8. Paley PJ, Goff BA, Gown AM, Greer BE, Sage EH. Alterations in SPARC and VEGF Immunoreactivity in epithelial ovarian cancer. Gynecol Oncol. 2000;78:336–41. [PubMed]
9. Mok SC, Chan WY, Wong KK, Muto MG, Berkowitz RS. SPARC, an extracellular matrix protein with tumor-suppressing activity in human ovarian epithelial cells. Oncogene. 1996;12:1895–901. [PubMed]
10. Tai IT, Dai M, Owen DA, Chen LB. Genome-wide expression analysis of therapy-resistant tumors reveals SPARC as a novel target for cancer therapy. J Clin Invest. 2005;115:1492–502. [PubMed]
11. Markman M. The myth of measurable disease in ovarian cancer. J Clin Oncol. 2003;21:3013–5. [PubMed]
12. Brewer CA, Blessing JA, Nagourney RA, Morgan M, Hanjani P. Cisplatin plus gemcitabine in platinum-refractory ovarian or primary peritoneal cancer: a phase II study of the Gynecologic Oncology Group. Gynecol Oncol. 2006;103:446–50. [PubMed]
13. Mutch DG, Orlando M, Goss T, et al. Randomized phase III trial of gemcitabine compared with pegylated liposomal doxorubicin in patients with platinum-resistant ovarian cancer. J Clin Oncol. 2007;25:2811–8. [PubMed]
14. Roby KF, Taylor CC, Sweetwood JP, et al. Development of a syngeneic mouse model for events related to ovarian cancer. Carcinogenesis. 2000;21:585–91. [PubMed]
15. Sweetwyne MT, Brekken RA, Workman G, et al. Functional analysis of the matricellular protein SPARC with novel monoclonal antibodies. J Histochem Cytochem. 2004;52:723–33. [PubMed]
16. Shi Q, Bao S, Maxwell JA, Reese ED, et al. Secreted protein acidic, rich in cysteine (SPARC), mediates cellular survival of gliomas through AKT activation. J Biol Chem. 2004;279:52200–9. [PubMed]
17. Lea JS, Ashfaq R, Muneer S, et al. Understanding the mechanisms of FHIT inactivation in cervical cancer for biomarker development. J Soc Gynecol Investig. 2004;11:329–37. [PubMed]
18. Norose K, Clark JI, Syed NA, et al. SPARC deficiency leads to early-onset cataractogenesis. Invest Ophthalmol Vis Sci. 1998;39:2674–80. [PubMed]
19. Ledda MF, Adris S, Bravo AI, et al. Suppression of SPARC expression by antisense RNA abrogates the tumorigenicity of human melanoma cells. Nat Med. 1997;3:171–6. [PubMed]
20. Sangaletti S, Stoppacciaro A, Guiducci C, Torrisi MR, Colombo MP. Leukocyte, rather than tumor-produced SPARC, determines stroma and collagen type IV deposition in mammary carcinoma. J Exp Med. 2003;198:1475–85. [PMC free article] [PubMed]
21. Perez RP. Cellular and molecular determinants of cisplatin resistance. Eur J Cancer. 1998;34:1535–42. [PubMed]
22. Hoskins WJ, Perez CA, Young RC, Barakat RB, Markman M, Randall ME. Principles and practice of gynecologic oncology. 4. Philadelphia: Lippincott Williams & Wilkins; 2005. pp. 468–71.