Predicting Efficacy of Cancer Cell Killing under Hypoxic Conditions

Article pubs.acs.org/ac

Predicting Efficacy of Cancer Cell Killing under Hypoxic Conditions with Single Cell DNA Damage Assay Yong Qiao and Liyuan Ma* NanoScience Technology Center, University of Central Florida, Florida ABSTRACT: The activity of anticancer drugs determined under normal conditions cannot accurately reflect true drug efficacy in a patient, as a tumor is often under low oxygen (hypoxia) conditions. In addition, patient responses to the same therapy can be drastically different due to tumor heterogeneity. This paper describes the use of single cell halo assay for detection and quantification of DNA damage induced by anticancer drugs or radiation under hypoxic conditions. By combining classical halo assay and state-of-theart microfabrication techniques, this single cell approach allows drug and radiation responses of cancer cells to be determined without population interference. The results from single cell assay indicate a diminished level of DNA damage at hypoxic conditions compared with those at normal conditions at the same drug concentrations or radiation dose, suggesting in vitro preclinical studies of drug and radiation activity can be performed under conditions that mimic physiological conditions of tumors and without population interference.

A

decreased within an affected tissue due to inadequate vascularization and perfusion under hypoxic conditions.28,29 Fourth, reduced oxygen tension at hypoxic conditions can lead to a decrease or complete arrest of the cell cycle, which limits drug effectiveness, because cancer therapeutics work well for rapidly dividing cells.30,31 Finally, hypoxia enhances genetic instability in cancer cells, allowing more rapid development of drug resistance in cells.32,33 All of these factors can affect a drug’s effectiveness in the hypoxic condition, suggesting that anticancer activity of a drug and radiation determined in standard conditions (20% O2) may not be a true indicator. We have developed a new single cell halo assay (SCHA) that can be used to quantify DNA damage level.34,35 This assay combines a classical halo assay and state-of-the-art microfabrication, in which surface patterns are used to capture single cells on silicon substrate via electrostatic attraction. The attached cells are embedded in agarose gel that provides an interconnected channel network, in which DNA fragments are released by alkaline lysis, and diffusion rates are inversely proportional to fragment sizes. After staining DNAs with ethidium bromide (EB), cells are imaged with fluorescence microscopy, where the intensity of the fluorescence signal is proportional to the amount of DNA. The level of DNA damage is quantified using the relative nuclear diffusion factor (rNDF) derived from the surface area of the nucleus and halo using rNDF = (R2-r2)/r2, where R and r are the radius of large and small circles, respectively. A unique feature of SCHA is its ability to detect DNA damage in individual cells, which permits detection of heterogeneity of cell response. This paper

big challenge for cancer therapy is tumor heterogeneity.1−3 Despite the availability of a panoply of drugs or therapeutics, patient response remains largely different even for cancers of identical tissue origin and histology.4−6 There exists clearly the need for strategies that can predict the response in individual cells prior to therapy.7−9 Although a number of assays can be used to profile tumor susceptibility, these assays have some issues. MTT assay relies on succinate dehydrogenase that can react with other chemicals, causing false positive responses;10,11 extreme drug resistance assay can only detect proliferating cells and cannot provide a reliable quantification;12,13 differential staining cytotoxicity assay is laborious and requires a skilled technician, though it can work with both proliferating and nonproliferating cells;14,15 adenosine triphosphate assay takes a long time (six days) and is not suitable for rapid clinical application.16−19 Most importantly, all the methods require a large amount of cells, and primary tissue samples especially those from patients with smaller tumors or hard to reach tumors will have to be extensively cultured prior to testing, which may not reflect events in a primary tumor. At last, these methods yield a population average readout of tumor response, which can yield misleading results since tumor samples from a biopsy display heterogeneity and contain both normal and tumor cells. Hypoxia is a feature of most tumors. Human solid tumors are typically hypoxic at 0−5% O2 compared to 0−14% in normal tissues.20−22 Hypoxia affects many aspects of tumors and their responses to therapy. First, hypoxia can be a direct cause of therapeutic resistance, because some drugs or radiation therapy require oxygen to be maximally cytotoxic.23−26 Second, the cellular metabolism of a drug can be altered under hypoxic conditions that can either increase or decrease a drug’s effectiveness on cancer cells.27 Third, drug distribution can be © XXXX American Chemical Society

Received: May 22, 2013 Accepted: June 18, 2013

A

dx.doi.org/10.1021/ac401543t | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 1. (A) Scheme of single cell halo assay; (B) an optical image of single cell array generated on silicon substrate, where amine modified areas attract cells for adhesion, and PEG modified areas repel cells; (C) fluorescence image of single cell array embedded in gel and stained with EthD-1/ Calcein AM in live/dead assay; (D) fluorescence image of cell array after single cell halo assay, where HeLa cells are treated with VP-16 (drug), and the red color is due to ethidium bromide stain; (E) an enlarged halo image, where a relative nuclear diffusion factor (NDF) can be derived from the radius of halo and nucleus.

stock solution was prepared by dissolving appropriate amounts in dimethyl-sulfoxide solvent to get the desired concentrations. All chemicals were used without further purification. The mixed gas (N2:CO2:O2 = 93%:5%:2%) was purchased from Air Liquide America Specialty (Orlando, FL). Gas regulator (30 pisg) and size 15D silicone rubber stoppers were from VWR (Atlanta, GA). Nalgene PMP Jars (1L) was from Thermo Scientific (Asheville, NC). Cell Culture. Two human cancer cell lines including HeLa (cervical cancer) and LNCaP (prostate cancer) were obtained from the American Type Culture Collection (ATCC Manassas, VA) and cultured in standard conditions (5% CO2 in air at 37 °C) in RPMI-1640 medium, supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) penicillin/streptomycin. HCT-116 cells and HCT-116-PARP1−/− cells were generous gifts from Dr. Eric Hendrickson at the Department of Biochemistry, Molecular Biology and Biophysics at University of Minnesota. Cells were cultured in McCoy’s 5A medium (Lonza, Walkers-ville MD), supplemented with 1% (v/v) penicillin/streptomycin, 1% (v/v) G418, and 10% (v/v) fetal bovine serum at standard conditions (5% CO2 in air at 37 °C). Cells were subcultured after they became confluent every 3−4 days, and all cells were used between passages 5−10. Cells were washed twice with D-PBS and resuspended in fresh new media at 1 × 105/well in 6-well plates that contain each of the different drugs (doxorubicin hydrochloride, etoposide, and irinotecan hydrochloride) at different concentrations (0.05, 0.1, 10, 50 μM) and incubated for 2.0 h at standard culture condition. After washing out the drugs, cells were loaded onto a patterned substrate to form cell array. Cell Array Formation. Silicon substrates (30 × 45 mm2) were cleaned by immersing in a 1:1 mixture of methanol and hydrogen chloride for 1.5 h, rinsed with deionized water 3 times, and transferred into concentrated sulfuric acid for 1.5 h, which generates a thin oxide film on silicon. After having been washed with water, the substrate was boiled in deionized water for 1 h, placed in an 80 °C oven for 3 h, and was ready for use. The cleaned substrate was modified with polyethylene glycol terminated silane (PEG-silane) by incubating with 3 mM of PEG-silane in toluene containing 1% triethylamine (v/v) for 1 h at 60 °C in a nitrogen environment. After removing unbound PEGsilane by sonication in toluene and ethanol, the substrate was washed with water and dried in a gentle flow of nitrogen.

describes the use of SCHA to measure DNA damages that are induced by an anticancer drug or radiation and determines their efficacy for cancer cell killing under hypoxic conditions versus normoxic conditions. The major advantages of this assay include the following: (a) the assay is carried out at single cell level and has single cell sensitivity; (b) adjacent cells and haloes will not overlap with each other; (c) live cell assay can be done to assess DNA damage; (d) spatially encoded cell locations allow simultaneous testing of multiple experimental conditions; (e) cells are located at the same height and focal plane, which saves time and greatly minimizes bias; (f) the clear boundary between nucleus and halo and symmetric shapes of halos and nuclei allow accurate determination of halo dimensions; (g) this method is fully compatible with automate imaging and robust statistical analysis with minimal level of user intervention; (h) the method detects DNA damage level with high sensitivity, accuracy, linearity, throughput, and technical readiness; (i) importantly, this platform can be further modified to a compact, point of care, and cost-effective technique suitable for clinical use.

■

MATERIALS AND METHODS Materials. Methoxy-polyethylenoxy propyl trichlorosilane (PEG-silane, 472-604 g/mol) was from Gelest (Tullytown, PA). Aminopropyltriethoxy silane (APTES), hydrogen peroxide, triethylamine, RPMI-1640 medium, EMEM medium, penicillin/streptomycin, polyvinyl alcohol, toluene, and ethanol were from Sigma-Aldrich (St. Louis, MO). Ten % (v/v) fetal bovine serum and EthD-1/Calcein AM (live/dead viability/ cytotoxicity kit) were from Invitrogen (Carlsbad, CA). Trypsin/ethylenediamine-tetraacetic acid (EDTA) solution was purchased from Cell Applications (San Diego, CA). Trypan blue, phosphate buffer saline (PBS), methanol, sodium hydroxide, hydrogen chloride, sulfuric acid, and agarose were from VWR (West Chester, PA). Silicon substrates (p-type boron-doped resistivity of 8−25 Ω cm−2) were from UniversityWafer (Boston, MA). Ethidium bromide (EB) was from Alfa Aesar (Ward Hill, MA). Three commercial anticancer drugs were used, doxorubicin hydrochloride (579.99 g/mol) from TOCRIS bioscience (Mllisvmoille, MN) and etoposide (VP-16, 588.6 g/mol) and irinotecan hydro-chloride (CPT-11, 623.14 g/mol) from Sigma-Aldrich (St. Louis, MO). The drug B

dx.doi.org/10.1021/ac401543t | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Figure 2. (A) A homemade hypoxic chamber is connected to a premade gas mixture for hypoxic culture (top) and normoxia culture (bottom); LNCaP cells cultured under hypoxic conditions without a drug for 24 h (B) and with drug treatment for 2 h (C) and 12 h (D); LNCaP cells cultured under normoxia without a drug for 24 h (E) and with drug treatment for 2 h (F) and 12 h (G).

in different cycle stages. The slide is incubated with 0.3 M NaOH for 30 min at room temperature and stained with 10 μg/mL EB for 10 min. During this period, DNA in treated cells diffuses out of the nucleus to form a larger circle relative to the compact nucleoid. The diffusion coefficient is inversely proportional to the size of the DNA fragment. Figure 1D shows a fluorescent image of a gel covered array of HeLa cells that are treated with 10 μM etoposide for 2 h. Note that some halos appear nonuniform, because cells are in different phases of cell cycle (topoisomerase II targeted agents cause more DNA damage in late S-G2 phase).36−38 The fluorescence intensity is proportional to the amount of DNA damage. Owing to a clear boundary, symmetric shapes of halos and nuclei and the nonoverlapping nature of adjacent cells/halos, a relative nuclear diffusion factor (rNDF) can be accurately derived using rNDF = (R2 − r2)/r2 (Figure 1E). The oxygen microenvironment regulates cell metabolism, proliferation, survival, and therapy response. We have cultured cells under hypoxic (2%) and normoxic conditions (Figure 2A). A low oxygen hypoxia chamber is built according to Woodring E Wright’s methods.39 The chamber consists of an airtight plastic jar with two holes drilled in the lid. Specimens for hypoxic culture are transferred to the chamber and purged with premixed gas (2% O2, 5% CO2, and 93% N2) for 3 min to obtain 2% oxygen tension inside. The chamber is regassed each time it is opened to keep a hypoxia environment. The chamber is checked for leakage by observing the color of medium (contained phenol red). If the chamber does not leak and 5% CO2 is retained, the medium maintains its orange-red color. Cells are cultured 24 h at hypoxia conditions before a drug is added and continuously cultured for a certain time. Figure 2B and 2E show LNCaP cells after seeding for 24 h under hypoxic and normoxic conditions, respectively. The LNCaP cells exhibited different morphologies, where cells are not tightly attached under hypoxic conditions and not spread. In contrast, under normoxic conditions, cells are tightly attached and spread. Figure 2C and 2D show LNCaP cells are treated with 10 μM doxorubicin under hypoxic conditions for 2 and 12 h, respectively, where cells are alive after 12 h exposure, meaning cells have adapted themselves to hypoxic stress. Figure 2F and 2G show LNCaP cells that are treated with 10 μM doxorubicin under normorxic conditions for 2 and 12 h, respectively, where cells are dead after 12 h exposure under normoxic conditions. These results indicate hypoxia has an inhibitory effect on DNA damage.

The PEG covered silicon substrate was patterned with deep ultraviolet laser lithography using 193 nm ArF excimer laser from Lambda Physik (Santa Clara, CA). The intensity of laser was homogenized by passing a beam through an in-line homogenizer. Micropatterns were written on a quartz plate in dark-field polarity, where patterned areas were transparent to laser, and the remaining areas covered with chromium were opaque. The PEG covered substrates were exposed at an intensity of 200 mJ/pulse and a frequency of 10 Hz for 1 min to obtain a patterned substrate. In order to form cell array, 1 mL of medium containing cells is added into each sample and cultured for 30 min at room temperature in the dark to allow adhesion complete. Nonadherent cells are removed by washing with medium. The whole fabrication process takes less than 1.5 h. DNA Damage Assay. DNA damage was measured with a modified alkaline halo assay as described below. Briefly, after agarose solidified, the slide was incubated with 0.3 M NaOH for 30 min at room temperature and stained with 10 μg/mL EB for 10 min. The slide was incubated in deionized water for 5 min. Fluorescent images were taken on an Olympus microscope (BX51M) with an Olympus ColorView CCD camera. An image analysis software (Image J) was used to obtain dimensions of halo and nucleus from collected images. An rNDF value of each cell was derived from radii of halo and nucleus. Untreated cells were used as negative controls. Statistical analysis of data is carried out with SPSS 19.0 (SPSS Inc., Chicago). Comparison between control group and treatment group is based on t test for two independent samples. A result was considered statistically significant difference when P ≤ 0.05.

■

RESULTS AND DISCUSSION Figure 1A shows the scheme of SCHA that is normally composed of three steps: (1) capturing single cells on a patterned substrate; (2) embedding patterned cells in the agarose gel; and (3) incubating cells in an alkaline solution to allow DNA fragments diffusion. Figure 1B shows an optical image of a typical single cell array generated on a silicon substrate vial in an electrostatic interaction. The patterned cell array is embedded in agarose gel by adding 1 mL of 1% lowmelting-point agarose (PBS, pH = 7.3) onto the substrate. The viability of patterned cells is tested with live/dead assay, where EthD-1/Calcein AM can stain viable cells to a green color. Figure 1C shows that arrayed cells inside the gel are alive after 3 h. The variation in cell sizes is likely due to the fact that cells are C

dx.doi.org/10.1021/ac401543t | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

Hypoxia also causes therapeutic resistance to radiotherapy. We have determined DNA damage induced by X-ray radiation under hypoxia and normoxia conditions. The low dose X-ray radiation is generated from a mini-X-ray generator (Amptek, Bedford, MA) operated at 40 kV and 100 μA using silver target. The X-ray dose rate is calibrated with a radiation meter and is evenly distributed over an area of 6 × 6 cm2. At a vertical distance of 5 cm, X-ray irradiation for 5 min provides 1.25Gy dose on the irradiated area. Figure 3A−D are typical halo

Figure 3. DNA damage induced by X-ray radiation under hypoxia and normoxia. Fluorescence images of LNCaP cells at different X-ray dose under normoxia, where halos become larger as increasing dose (from A to D) due to more DNA damage; dose dependent DNA damage by X-ray (E), where rNDFs at hypoxic conditions increase from 0 to 2.2 as the dose increases from 0 to 2.5Gy, which is less than those at normoxic conditions (rNDFs from 0 to 3.0). The star symbol indicates significant difference.

Figure 4. Drug concentration dependent DNA damage under hypoxia and normoxia. rNDF values of LNCaP (A) and HeLa (B) cells treated with three drugs under normoxic conditions, and those of LNCaP (C) and HeLa (D) treated under hypoxic conditions; rNDFs of four types of cells (HeLa, LNCaP, HCT-116 wild-type, and HCT-116 with PARP1 knock down) treated with 10 μM doxorubicin (E) and CPT11 (F) at normorxic and hypoxic conditions. The star symbol indicates significant difference.

images of single LNCaP cells after exposure to X-ray under hypoxia conditions. As the dose increases, halos become larger due to more damage to DNA. The fluorescence intensity of the core decreases, suggesting residual nuclei decrease progressively. The rNDF values of 50 randomly selected halos are averaged to quantify DNA damage. Figure 3E shows that, as the dose increases from 0 to 2.5Gy, the rNDFs at hypoxic conditions increase from 0 to 2.2, which is less than those at normoxic conditions (0−3.0). The results indicate that low oxygen supply at hypoxic conditions limits radiation to create oxygen free radicals for DNA damage.40,41 LNCaP and HeLa cell lines are used to test the ability of single cell halo assay to “predict” the response of three clinically approved anticancer drugs (CPT-11, VP-16, and doxorubicin) under hypoxic and normoxic conditions. Figure 4 shows rNDF values of LNCaP and HeLa cells under normoxic conditions (4A and 4B) and under hypoxic conditions (4C and 4D) treated with three drugs at different concentrations (relative to untreated control). The selected concentrations are close to the clinically used drug concentrations. Each experimental point is averaged from at least 50 cells, and the results are expressed as mean ± standard deviation using OriginPro 8.5. When drug concentration increases from 0 to 50 μM, rNDFs of LNCaP cells increase from 0 to 5.2, and those of HeLa cells increase from 0 to 5.9 at normoxic conditions; those of LNCaP cells increase from 0 to 4.4, and those of HeLa cells increase from 0 to 4.1 at hypoxic conditions. These results indicate that drug effect is reduced under hypoxic conditions. CPT-11 induces more DNA damage in LNCaP cells than doxorubicin and VP16; VP-16 induces more DNA damage in HeLa cells than doxorubicin and CPT-11. Thus, LNCaP cells and HeLa cells are more susceptible to CPT-11 and VP-16, respectively. Figure 4E and 4F show rNDFs of four types of cells (HeLa, LNCaP,

HCT-116, and HCT-116/PARP1−/−) treated with 10 μM doxorubicin and CPT-11 at normoxic and hypoxic conditions, where diminished DNA damage can be found in four types of cells at hypoxic conditions. The effects of hypoxia on response to chemotherapy depend on drug and cell line: the highest effect occurs in CTP-11 treated HeLa cells, while the lowest effect occurs in doxorubicin treated wild-type HCT-116 cells. These results suggest single cell halo assay can differentiate DNA damage level under hypoxic and normoxic conditions after drug exposure.

■

CONCLUSIONS The single cell halo assay allows determination of drug and radiation responses of cancer cells without population interference. Diminished level of DNA damage is observed at hypoxic conditions compared to normorxic conditions at the same drug concentrations or radiation dose. CPT-11 can induce more DNA damage in LNCaP cells than the other two drugs; VP-16 induces more DNA damage in HeLa cells than that of the other two drugs. These results suggest that single cell halo assay can differentiate DNA damage level under hypoxic and normoxic conditions after exposure to drug and radiation. Single cell halo assay has the potential to be a low cost point-of-care platform for personalized screening of therapeutics at clinics. D

dx.doi.org/10.1021/ac401543t | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

■

Article

(20) Atkuri, K. R.; Herzenberg, L. A. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 3756. (21) Ivanovic, Z. J. Cell. Physiol. 2009, 219, 271. (22) Vaupel, P.; Höckel, M.; Mayer, A. Antioxid. Redox Signaling 2007, 9, 1221. (23) Vaupel, P.; Thews, O.; Hoeckel, M. Med. Oncol. 2001, 18, 243. (24) Shannon, A. M.; Bouchier-Hayes, D. J.; Condron, C. M.; Toomey, D. Cancer Treat. Rev. 2003, 29, 297. (25) Colvin, O. M. Cancer Treat. Res. 1994, 73, 249. (26) Teicher, B. A. Cancer Metastasis Rev. 1994, 13, 139. (27) Teicher, B. A.; Waxman, D. J.; Holden, S. A.; Wang, Y.; Clarke, L.; Alvarez Sotomayor, E.; Jones, S. M.; Frei, E. Cancer Res. 1989, 49, 4996. (28) Teicher, B. A.; Holden, S. A.; Rose, C. M. J. Natl. Cancer Inst. 1985, 75, 1129. (29) Teicher, B. A.; Crawford, J. M.; Holden, S. A.; Cathcart, K. N. S. Cancer Res. 1987, 47, 5036. (30) Hockel, M.; Vaupel, P. J. Natl. Cancer Inst. 2001, 93, 266. (31) Brown, J. M.; Koong, A. J. Natl. Cancer Inst. 1991, 83, 178. (32) Hockel, M.; Vaupel, P. Semin. Oncol. 2001, 28, 36. (33) Baronzio, G.; Freitas, I.; Kwaan, H. C. Semin. Thromb. Hemostasis 2003, 29, 489. (34) Qiao, Y.; Wang, C.; Su, M.; Ma, L. Anal. Chem. 2012, 84, 1112. (35) Qiao, Y.; An, J.; Ma, L. Anal. Chem. 2013, 85, 4107. (36) Chow, K. C.; Ross, W. E. Mol. Cell. Biol. 1987, 7, 3119. (37) Corbett, A. H.; Osheroff, N. Chem. Res. Toxicol. 1993, 6, 585. (38) D’Arpa, P.; Beardmore, C.; Liu, L. F. Cancer Res. 1990, 50, 6919. (39) Wright, W. E.; Shay, J. W. Nat. Protoc. 2006, 1, 2088. (40) Hockel, M.; Schlenger, K.; Mitze, M.; Schaffer, U.; Vaupel, P. Semin. Radiat. Oncol. 1996, 6, 3. (41) Brown, J. M.; Giaccia, A. J. Cancer Res. 1998, 56, 4509.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Dr. James Hickman for allowing us to use the laser fabrication facility and Dr. Ming Su for insightful discussions. This work has been supported by a New Investigator Research Grant to L. Ma from Bankhead-Coley Cancer Research Program of Florida Department of Health and a seedling research grant on space radiobiology to L. Ma from Kennedy Space Center.

■

REFERENCES

(1) Anderson, K.; Lutz, C.; van Delft, F. W.; Bateman, C. M.; Guo, Y.; Colman, S. M.; Kempski, H.; Moorman, A. V.; Titley, I.; Swansbury, J.; Kearney, L.; Enver, T.; Greaves, M. Nature 2011, 469, 356. (2) Gerlinger, M.; Rowan, A. J.; Horswell, S.; Larkin, J.; Endesfelder, D.; Gronroos, E.; Martinez, P.; Matthews, N.; Stewart, A.; Tarpey, P.; Varela, I.; Phillimore, B.; Begum, S.; McDonald, N. Q.; Butler, A.; Jones, D.; Raine, K.; Latimer, C.; Santos, C. R.; Nohadani, M.; Eklund, A. C.; Spencer-Dene, B.; Clark, G.; Pickering, L.; Stamp, G.; Gore, M.; Szallasi, Z.; Downward, J.; Futreal, P. A.; Swanton, C. New Engl. J. Med. 2012, 366, 883. (3) Kirk, R. Nat. Rev. Clin. Oncol. 2012, 9, 250. (4) Sevin, B. U.; Perras, J. P. Am. J. Obstet. Gynecol. 1997, 176, 759. (5) Mercer, S. J.; Somers, S. S.; Knight, L. A.; Whitehouse, P. A.; Sharma, S.; Di Nicolantonio, F.; Glaysher, S.; Toh, S.; Cree, I. A. Anticancer Drugs 2003, 14, 397. (6) Harry, V. N.; Gilbert, G. J.; Parkin, D. E. Obstet. Gynecol. Surv. 2009, 64, 548. (7) Pusztai, L.; Rouzier, R.; Wagner, P.; Symmans, W. F. Drug Resist. Updates 2004, 7, 325. (8) Mi, Z.; Holmes, F. A.; Hellerstedt, B.; Pippen, J.; Collea, R.; Backner, A.; Bush, J. E.; Gallion, H. H.; Wells, A.; O’Shaughnessy, J. A. Anticancer Res. 2008, 28, 1733. (9) Dietz, A.; Boehm, A.; Horn, I.; Kruber, P.; Bechmann, I.; Golusinski, W.; Niederwieser, D.; Dollner, R.; Remmerbach, T. W.; Wittekind, G.; Dietzsch, S.; Hildebrandt, G.; Wichmann, G. Eur. Arch. Oto-Rhino-Laryngol. 2010, 267, 483. (10) Wu, B.; Zhu, J.-S.; Zhang, Y.; Shen, W.-M.; Zhang, Q. World J. Gastroenterol. 2008, 14, 3064. (11) Rubinstern, L.; Shoemaker, R.; Pacell, K.; Simon, R.; Tosini, S. J. Natl. Cancer Inst. 1990, 82, 1113. (12) Ellis, R. J.; Fabian, C., J.; Kimler, B. F. Breast Cancer Res. Treat. 2002, 71, 95. (13) Kern, D. H.; Weisenthal, L. M. J. Natl. Cancer Inst. 1990, 82, 582. (14) Bird, M. C.; Godwin, V. A.; Antrobus, J. H. Br. J. Cancer 1987, 55, 429. (15) Weisenthal, L.; Marsden, J.; Dill, P.; Macaluso, C. Cancer Res. 1983, 43, 749. (16) Andreotti, P. E.; Cree, I. A.; Kurbacher, C. M.; Hartmann, D. M.; Linder, D.; Harel, G.; Gleiberman, I.; Caruso, P. A.; Ricks, S. H.; Untch, M.; Sartori, C.; Bruckner, H. W. Cancer Res. 1995, 55, 5276. (17) Cree, I. A.; Kurbacher, C. M.; Untch, M.; Sutherland, L. A.; Hunter, E. M.; Subedi, A. M.; James, E. A.; Dewar, J. A.; Preece, P. E.; Andreotti, P. E.; Bruckner, H. W. Anticancer Drugs 1996, 7, 630. (18) Kangas, L.; Gronroos, M.; Nieminen, A. Med. Biol. 1984, 62, 338. (19) Ugurel, S.; Schadendorf, D.; Pföhler, C.; Neuber, K.; Thoelke, A.; Ulrich, J.; Hauschild, A.; Spieth, K.; Kaatz, M.; Rittgen, W.; Delorme, S.; Tilgen, W.; Reinhold, U. Clin. Cancer Res. 2006, 12, 5454. E

dx.doi.org/10.1021/ac401543t | Anal. Chem. XXXX, XXX, XXX−XXX