Circulating tumor cells from a 4-dimensional lung cancer model are resistant to cisplatin

Accepted Manuscript Circulating Tumor Cells from 4D Lung Cancer Model are Resistant to Cisplatin Monika Vishnoi , PhD Dhruva K. Mishra , PhD Michael J. Thrall , MD Jonathan M. Kurie , MD Min P. Kim , MD, FACS PII:

S0022-5223(14)00671-0

DOI:

10.1016/j.jtcvs.2014.05.059

Reference:

YMTC 8643

To appear in:

The Journal of Thoracic and Cardiovascular Surgery

Received Date: 2 April 2014 Revised Date:

14 May 2014

Accepted Date: 21 May 2014

Please cite this article as: Vishnoi M, Mishra DK, Thrall MJ, Kurie JM, Kim MP, Circulating Tumor Cells from 4D Lung Cancer Model are Resistant to Cisplatin, The Journal of Thoracic and Cardiovascular Surgery (2014), doi: 10.1016/j.jtcvs.2014.05.059. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Circulating Tumor Cells from 4D Lung Cancer Model are Resistant to Cisplatin

Monika Vishnoi, PhD1,* , Dhruva K. Mishra, PhD1,*, Michael J. Thrall, MD2, Jonathan M. Kurie,

1

RI PT

MD3, and Min P. Kim, MD, FACS1,4

Department of Surgery and 2Pathology and Genomic Medicine, Houston Methodist Research

Institute, Houston, Texas; 3Thoracic/Head and Neck Medical Oncology, The University of Texas

College, Houston Methodist Hospital, Houston, Texas

SC

MD Anderson Cancer Center, Houston, Texas; 4 Department of Surgery, Weill Cornell Medical

M AN U

*MV and DKM contributed equally to this work as first author.

Funding: MPK received grant support from the Second John W. Kirklin Research Scholarship, the American Association for Thoracic Surgery Graham Research Foundation, and the Houston Methodist Foundation with a donation from J. Michael Jusbasche.

TE D

Conflict of Interest: MPK has received honoraria from Covidien and Ethicon for teaching VATS lobectomy course. MPK has applied for a patent related to this work. Corresponding Author:

EP

Min P. Kim, MD, FACS

6550 Fannin Street, Suite 1661

AC C

Houston, TX 77030 (p) 713-441-5177 (f) 713-790-5030

[email protected] Article word count: 3500

1

ACCEPTED MANUSCRIPT

Abstract Objective: To determine the effect of cisplatin on the circulatory tumor cells (CTC) and tumor nodules in the 4D lung cancer model.

RI PT

Methods: We treated the CTC from the 4D model seeded with H1299, A549 or H460 and respective cells that were grown on a 2D condition on a petri dish with 50µM cisplatin for 24 and 48 hours and determined the cell viability. We then treated the lung nodules in the 4D model

SC

with different continuous or intermittent doses of cisplatin and measured the nodule size, the number of CTC, and the matrix metalloproteinase (MMP) level.

M AN U

Results: Cisplatin led to a significant decrease in cell viability for tumor cells that were grown in the 2D condition (p<0.01) while cisplatin did not lead to a significant decrease in cell viability for CTC from the 4D model for both 24 hour and 48 hour time periods. Cisplatin led to tumor nodule regression for both the continuous or intermittent treatment. Moreover, there was a

TE D

significantly higher number of CTC per tumor area (p<0.05) and MMP-2 production per tumor area (p=0.007) for all human lung cancer cell lines grown in the 4D model when treated with cisplatin.

EP

Conclusion: The 4D lung cancer model allows for the isolation of CTC that are resistant to cisplatin treatment. The model may allow us to better understand the biology of cisplatin

AC C

resistance.

Word count: 225

2

ACCEPTED MANUSCRIPT

Ultramini Abstract: The 4D lung cancer model allows for the isolation of circulating tumor cells that are resistant to cisplatin treatment. Moreover, the treatment of tumor nodules in the 4D lung cancer model leads to a decrease in tumor nodule size and an increase in the number of

AC C

EP

TE D

M AN U

SC

RI PT

CTC and MMP-2 levels per tumor area.

3

ACCEPTED MANUSCRIPT

Introduction Lung cancer is the most common cause of cancer-related deaths in the United States1. The 5-year survival rate for lung cancer patients has increased only incrementally, from 13% in

RI PT

1975 to 16% in 20052. The reason for the high mortality and morbidity is that most lung cancer patients who present with metastatic disease develop resistance to chemotherapy1. One of the most effective treatments for lung cancer is platinum-based therapy, including cisplatin, that

SC

binds to and causes cross-linking, breaks, and mutations in the DNA chain-injuries that trigger

relapse due to cisplatin resistance3.

M AN U

apoptosis. It provides a very high initial response but the majority of cancer patients eventually

Overall, the mechanism of chemotherapeutic resistance is difficult to study due to the lack of preclinical models that mimic metastasis. Drug sensitivity data from a two-dimensional (2D) culture are often misleading4, and in vivo mouse studies, while the gold standard for

TE D

anticancer drug studies, are not always predictive of patient response5. We can often cure lung cancer in mice in such studies, but those results translate poorly to treating lung cancer patients. Efforts to improve the available models for anticancer drug research have led to the development

EP

of improved in vitro models.

Recently, we have developed a novel 4D lung model that forms perfusable nodules from

AC C

human non-small cell lung cancer cells grown on an acellular rat matrix. Unlike other in vitro models, it maintains a separation from its vascular or endothelial space and the epithelial space in the matrix6 allowing the tumor cells to grow in the 3-dimentional space and the media to flow through the vasculature and to create a perfusable nodule. The addition of flow makes this a unique 4D model. This 4D model mimics the human lung cancer histopathology and protease secretion pattern7, 8. Furthermore, the differential gene expression profiling shows that the gene

4

ACCEPTED MANUSCRIPT

signature of tumor cells grown in the 4D model predicts poor survival while the same cells grown in the 3D model or cells grown simply in a matrigel predict good survival9. This suggests that the 4D model allows tumor cells to behave in a more aggressive fashion.

RI PT

As the tumor nodules form in the 4D model, we have observed live tumor cells in the circulating media or circulating tumor cells (CTC). We postulated that the 4D model would show

SC

the impact of cisplatin on the primary tumor, the number of CTC, and protease secretion.

Materials and Methods

M AN U

The Institutional Animal Care and Use Committee at the Methodist Hospital Research Institute approved all the animal experiments in our study. Cell culture

We used the American type cell culture (ATCC) human non-small cell lung cancer cell

TE D

lines H1299, A549, H460, and HCC827. The H1299 cell line has a p53 deletion and was derived from lymph node metastasis; the A549 cell line was derived from the primary adenocarcinoma of the lung; the H460 cell line was derived from the pleural fluid of a patient with large cell lung

EP

cancer; and the HCC827 cell line has a mutation in the EGFR tyrosine kinase domain and was derived from primary lung adenocarcinoma. The cells were grown in cell culture flasks (BD

AC C

Biosciences) in complete media made from RPMI 1640 (HyClone) supplemented with 10% fetal bovine serum (Lonza) and antibiotics (100 IU/mL penicillin, 100 µg/mL streptomycin, and 0.25 µg/mL amphotericin B; MP Biomedicals) at 37°C in 5% CO2. 4D lung cancer model We created the 4D lung model as previously described7. We used decellularized lung matrices from 6-week-old Sprague Dawley rats (Harlan, Houston, TX, USA) and seeded 25

5

ACCEPTED MANUSCRIPT

million H129910, A54911, H460,10 or HCC82712 human lung cancer cells (ATCC) diluted in 50 mL of complete media in the trachea of the ex vivo 4D lung model. We set up a bioreactor with 200 mL of complete media as previously described7, 13. The media goes through the pulmonary

RI PT

artery of the acellular lung, perfuses the lung, exits the lung, and collects in a reservoir where the media is pumped in a closed circuit through a pump at 6 mL per minute and an oxygenator where it is then returned to the acellular lung. We collected the media every day from the reservoir,

SC

spun down the conditioned media at 500 g for 5 minutes and diluted the cells in 1 mL of complete media. Next, we counted the number of live tumor cells using an automatic cell counter

M AN U

(TC20 cell counter, Biorad). These cells were labeled as the circulating tumor cells (CTC). We replaced the media in the bioreactor with fresh complete media daily. Cisplatin treatment of CTC and 2D cells

We plated 10,000 CTC from the 4D model seeded with H1299, A549, or H460 or

TE D

respective cells that were grown on a petri dish (2D) in 96 well plates in 200 µl complete media. We excluded the HCC827 cells from this experiment since we could not reliably count the number of cells using the automatic cell counter due to cell clumping. Half of the plated CTC or

EP

2D cells were treated with 50 µM cisplatin (n=12) while the other half were used as a control group (n=12). We counted the number of live cells in the supernatant and live cells that were

AC C

adherent to the plate after trypsinization with 0.25% trypsin (Hyclone) at 24 (n=6) or 48 hours (n=6). We compared the number of total live cells (the sum of the live cells in the supernatant and the live cells adherent to the plate) between the CTC or 2D cells that were treated with cisplatin and the control group. Cisplatin treatment of tumor nodules grown on the 4D model

6

ACCEPTED MANUSCRIPT

We seeded the 4D lung cancer model with H1299 and grew it for 14 days (n=9). We treated the 4D model with no cisplatin as a control group (n=3), cisplatin at 10 µM (n=3), or 50 µM (n=3) starting on day 4 after seeding of the nodule. We examined the lung tissue of the 4D

RI PT

lung model for the presence of nodules every other day. We calculated the area per nodule and added them to obtain the total nodule size. On days 8, 11, and 14, we performed a right upper lobectomy, a right middle lobectomy, and a right lower lobectomy, respectively. The lung tissues

SC

from the ex vivo 4D lung model from the lobectomy were processed as described previously7 and evaluated by board certified pathologist (MT). To confirm the H&E results and to determine cell

M AN U

proliferation and cell death, we performed immunohistochemical analysis on lung tissue from day 8 in 2 groups (the control and 50-µM treatment). For the immunohistochemical analysis, we followed the standard protocol as previously described13. We determined the percentage of Ki-67 and Caspase-3 staining by examining 5 random high-power fields. We also isolated and counted

TE D

the number of CTC daily as described above. Next, we measured the CTC per tumor area by dividing the number of CTC by the size of the tumor from each given day. We also seeded the left side of the 4D lung cancer model with either H1299, A549,

EP

H460, or HCC827 and treated it with the physiological concentration of cisplatin observed in the plasma after treatment in patients14 of 10 µM given weekly for three doses starting day 4 for 21

AC C

days (n=4) or untreated as a control (n=4). We examined the 4D lung cancer model for the presence of nodules every five days. We calculated the area per nodule and added them to obtain the total nodule size. We also calculated the number of CTC per tumor area. We measured the area under the curve for the tumor size and CTC per tumor area for the model treated with cisplatin and the control group. We compared the area under the curve for the tumor size and CTC per tumor area between the control group and cisplatin-treated group.

7

ACCEPTED MANUSCRIPT

MMP levels in the media We performed MMP assay using the media collected from the 4D lung models treated with intermittent concentration of cisplatin (10 µM) as described above. We used a MILLIPLEX

RI PT

Human MMP Panel 2 kit (Millipore) to detect both active and inactive MMP-1, MMP-2, MMP7, MMP-9, and MMP-10 and followed the protocol as per the manufacturer’s instructions13. A Luminex 200TM instrument was used to read the plate. The results were further analyzed using

SC

MILLIPLEX Analyst Software (Millipore). We then calculated the MMP level per tumor area and determined the area under the curve for the model treated with cisplatin and the control

M AN U

group. We then normalized the value to the control group and determined the average MMP level per tumor area. We compared the normalized area under the curve of MMP per tumor area for all three cell lines. Statistical analysis

TE D

Independent 2-sample t-tests were performed to identify significantly different mean number of live cells after cisplatin treatment for CTC and 2D cells. In addition, a t-test was used to test for mean tumor size at different days for continuous cisplatin treatment and significantly

EP

different percentages of Ki-67 staining and percentages of Caspase-3 staining on day 8. Moreover, t-tests were done to test for the number of CTC per tumor area between the

AC C

continuous cisplatin treated (50 µM or 10 µM) 4D model to the control group and the area under the curve for the tumor size and the area under the curve for CTC per tumor area for the intermittent treatment of cisplatin (10 µM) compared to the control group. Finally, a t-test was used to compare the normalized area under the curve of MMP per tumor area. All analyses were performed using PRISM Version 5.0 software (GraphPad Software, La Jolla, CA). A p-value <0.05 was considered significant.

8

ACCEPTED MANUSCRIPT

Results: Circulating tumor cells from the 4D model

RI PT

There were significantly fewer total live tumor cells when tumor cell lines grown on petri dish (2D) were treated with 50 µM of cisplatin for 24 (Fig. 1A) or 48 hours (Fig. 1B) compared to the control group. There were significantly fewer A549 2D (p=0.0005), H1299 2D (p=0.001)

SC

and H460 2D (p=0.0006) when these cells were treated with cisplatin for 24 hours after plating on a petri dish as well as when A549 2D (p<0.0001), H1299 2D (p<0.0001) and H460 2D

M AN U

(p=0.02) were treated with cisplatin for 48 hours. In both time points the control 2D cells had a greater number of cells in the cultured condition than the cells that were plated (dotted line). On the other hand, there was no significant decrease in the total live tumor cell amount of CTC from the 4D model that were treated with 50 µM of cisplatin for 24 (Fig. 1A) or 48 hours

TE D

(Fig. 1B) on a petri dish as compared to the control group. For A549 CTC, there were significantly more live tumor cells when treated with cisplatin compared to the control group (p=0.03) at 24 hours while this difference disappeared at 48 hours (p=0.5). For H1299 CTC,

EP

there was no significant difference in the number of cells when treated with cisplatin for both 24 hours (p=0.3) and 48 hours (p=0.06) compared to the control group. For H460 CTC, there was

AC C

no significant difference in the number of live tumor cells after 24 hours of cisplatin treatment (p=1) but after 48 hours there were significantly more tumor cells with cisplatin treatment compared to the control group (p=0.02). Moreover, in both time points the control CTC cells had fewer cells in the cultured condition compared to the number of cells that were plated (dotted line). Continuous cisplatin treatment of cells in the 4D model

9

ACCEPTED MANUSCRIPT

In the 4D lung model, 97.6 ± 4.9% of all cells lines adhered to the decellularized lung matrix after 3 passes and incubation for 2 hours. The H1299 cells formed nodules on the 4D lung model by day 4 (Fig. 2). The tumor nodules enlarged without cisplatin treatment on day 4 (Fig. 2

RI PT

A), day 8 (Fig. 2C), and day 14 (Fig. 2E) while tumor nodule regressed with 50 µM of continuous daily cisplatin treatment starting on days 4 (Fig. 2B), 8 (Fig. 2D), and 14 (Fig 2F) as well as 10 µM of continuous daily cisplatin treatment (Fig. 2G). Prior to cisplatin treatment on

SC

day 4, the nodule size did not significantly differ between the groups (Fig. 2 A&B, P=0.2). On day 8, however, the nodules in the control group (506 ± 63 mm2) were significantly larger than

M AN U

those in the groups treated with 10 µM of cisplatin (112 ± 34 mm2, P=0.0007) and 50 µM of cisplatin (125 ± 55 mm2, P=0.0014) (Fig. 2). In both treatment groups, the nodules were significantly smaller on day 14 than they were on day 4 (P=0.001 for 10 µM of cisplatin, P=0.006 for 50 µM of cisplatin).

TE D

H&E analysis of the lobectomy specimen from days 8 (Fig. 2H), 11 (Fig. 2K), and 14 (Fig. 2N) from the control H1299 cells grown on the ex vivo 4D lung model showed tumor cells with cell-cell and cell-matrix interaction and with an intact vascular space without any signs of

EP

cell necrosis. However, most of the H1299 cells on the ex vivo model treated with 50 µM of continuous cisplatin were dead, resulting in either ghost cells without visible residual nuclear

AC C

material or cells with pyknotic nuclear remnants on days 8 (Fig. 2J), 11(Fig. 2M), and 14 (Fig. 2P). The outlines of the ghost cells showed that prior to treatment, the tumor cells had a growth pattern and degree of cellularity similar to those in the control group. The H1299 cells treated with 10 µM of continuous cisplatin showed viable tumor cells on day 8 (Fig. 2I); however, on day 11, about 50% of the cells were necrotic (Fig. 2L), and on day 14, most of the cells had died (Fig. 2O).

10

ACCEPTED MANUSCRIPT

We further analyzed these tissues for cell proliferation (Ki67) and apoptosis (CASP3). On day 8, the percentage of cells with Ki-67 staining (Fig. 3C) was much higher in the control (Fig. 3A, Control, 44.2 ± 3.6 %) H1299 cells grown on the ex vivo 4D lung model compared to

RI PT

the cells treated continuously with 50 µM of Cisplatin (Fig. 3B, 3.5 ± 0.4%, P<0.0001). Moreover, on day 8, most of the H1299 cells on the ex vivo model treated with 50 µM of continuous cisplatin treatment stained for Caspase-3 (Fig. 3E, 91.2 ± 0.9%), while only a few

SC

cells in the control group stained for Caspase-3 (Fig. 3D, 9.4 ± 1.4%). A significantly higher percentage of H1299 cells on the ex vivo 4D lung model with 50-µM continuously cisplatin-

M AN U

treated group stained positive for Caspase-3 than in the control group (Fig. 3F, P<0.0001). Further comparison of CTC from the untreated versus the treated 4D model with H1299 showed a significantly higher number of CTC per tumor area on day 6 with 50 µM cisplatin treatment and on day 8 and 10 with 10 µM cisplatin (p<0.05, Fig 3G) compared to the control group.

TE D

Intermittent cisplatin treatment of cells on the 4D model

We performed physiologic intermittent treatment of cisplatin (10 µM) of the 4D model seeded with H1299, A549, H460, or HCC827 cells at days 4, 11, and 18. The area under the

EP

curve for the tumor nodule size for the 4D model seeded with H1299 cells treated with cisplatin was smaller compared to the control group (Fig. 4A). The average of the area under the curve of

AC C

tumor size for all four cell lines treated with cisplatin was significantly smaller compared to the control group (p=0.01, Fig. 4B). The area under the curve for CTC per tumor area for the 4D model seeded with H1299 was larger compared to the control group (Fig. 4C). The average of the area under the curve of the CTC per tumor area for all four cell lines treated with cisplatin was significantly higher compared to the control group (p=0.02, Fig. 4D). MMPs in the media of the 4D model

11

ACCEPTED MANUSCRIPT

Luminex analysis detected MMP-1, MMP-2, MMP-7, MMP-9, and MMP-10 in the media of all four cells grown on the ex vivo 4D lung model. The area under the curve for MMP-1 per tumor area (Fig. 5A), MMP-2 per tumor area (Fig. 5B), MMP-9 per tumor area (Fig. 5D) and

RI PT

MMP-10 per tumor area (Fig. 5E) were higher for the 4D model seeded with H1299 treated with cisplatin compared to the control group while there was no large difference for MMP-7 per tumor area (Fig. 5C). The MMP-2 level per tumor area normalized to the control for the 4D

SC

model seeded with three different cell lines showed a significant increase compared to the control group (p=0.007) while there was no significant difference for other four MMPs among

M AN U

the three cell lines (Fig 5F).

Discussion

Drug resistance is a major hurdle in cancer treatments, which results in poor clinical

TE D

prognosis. In this study, we showed that CTC from our 4D model are resistant to cisplatin. For three different human non-small cell lung cancer cell lines, cisplatin treatment did not impact this unique group of cells while the same cells that were grown on a petri dish (2D) had significantly

EP

less viable tumor cells after cisplatin treatment. This is the first time that an in vitro model has isolated a resistant group of cells without using cisplatin. While there are several studies that

AC C

show the development of cisplatin resistant cell lines by gradually increasing the concentration and ultimately selecting the live population of cells15 to provide an understanding of the mechanism involved in drug resistance, our model improves the selection of resistant cells as they develop from a natural matrix environment as a natural phenomenon. Furthermore, the untreated CTC show no signs of proliferation. In the conditioned media, the CTC for all three-

12

ACCEPTED MANUSCRIPT

cell lines did not proliferate. Thus, a possible mechanism for cisplatin resistance is cell cycle arrest that makes cisplatin ineffective16. As expected, the continuous treatment of the tumor nodules formed in the 4D model

RI PT

seeded with H1299 cells showed a significant decrease in the primary tumor nodule size. This correlates with tumor regression in some patients with lung adenocarcinoma17. On a computed tomography scan, regression of lung cancer nodules is a hallmark response to lung cancer

SC

treatment17 and our 4D lung model also shows nodule regression in response to treatment. The analysis of the tissue confirms this with lower proliferation (Ki-67) and higher cell death

M AN U

(Caspase-3) with more immediate impact with 50 µM of cisplatin with most of the cells showing necrosis 4 days after treatment while it took 10 days with 10 µM of cisplatin. After two days of 50 µM of cisplatin treatment, there is a significant increase in live CTC per tumor area while there is an increase on 4 and 6 days after treatment with 10 µM of cisplatin, both of which

TE D

dissipated with necrosis of the primary tumor. We postulated that cisplatin might cause the cells in the nodule to become CTC. We tested this concept by an intermittent treatment of physiologic concentration of cisplatin. We found that for all four cell lines that were grown in the model

EP

there were significant decreases in the tumor size and significant increases in CTC per tumor area when the 4D model with human lung cancer cell lines were treated with cisplatin. This may

AC C

model the formation of resistant metastatic lesions in lung cancer patients. In lung cancer patients, cisplatin greatly impacts the primary tumor size. However, this may lead to the formation of CTC that are resistant to cisplatin treatment and allow for the development of metastatic lesions that are resistant to cisplatin. This increase in CTC production per tumor area is correlated with an increase in MMP-2, a protein involved in the breakdown of extracellular matrix. It is known to break down type IV

13

ACCEPTED MANUSCRIPT

collagen, which is the major structural component of basement membrane18. MMP-2 has been found at higher levels in the serum in lung cancer patients compared to the benign and control group19. It has also been shown as more sensitive predictor than MMP-9 for lung cancer

RI PT

progression, metastasis and survival 20. MMP-2 may be involved in the release of CTC from the primary tumor when it is treated with cisplatin.

The limitation of the current study is that we have yet to confirm these findings in lung

SC

cancer patients. We postulate that when lung cancer patients are given cisplatin, there will be a significant increase in CTC per tumor area and that these CTC will be more resistant to cisplatin.

M AN U

We plan to investigate our hypothesis as we improve our ability to isolate the CTC from lung cancer patients.

To our knowledge, the current study is the first to show the effect of cisplatin treatment on perfusable nodules and the selection of cisplatin-resistant CTC. The 4D lung model is not

TE D

likely to replace the current in vivo mouse models, but it will provide additional information through its ability to control many aspects of tumor growth and development that will allow an improved understanding of the mechanisms of action of various anticancer drugs..Ultimately, the

EP

4D model could potentially provide an alternative approach to study drug efficacy, mechanisms

AC C

of resistance, and molecular insights into the treatment of non-small cell lung cancer.

14

ACCEPTED MANUSCRIPT

References

6.

7.

8.

9.

10. 11.

12.

13.

14.

15.

RI PT

SC

5.

M AN U

4.

TE D

3.

EP

2.

Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA: a cancer journal for clinicians. 2013;63:11-30. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin. 2010;60:277300. Petrelli F, Coinu A, Cabiddu M, Ghilardi M, Ardine M, Barni S. Platinum rechallenge in patients with advanced NSCLC: a pooled analysis. Lung cancer. 2013;81:337-42. Shaw GL, Gazdar AF, Phelps R, Steinberg SM, Linnoila RI, Johnson BE, et al. Correlation of in vitro drug sensitivity testing results with response to chemotherapy and survival: comparison of non-small cell lung cancer and small cell lung cancer. J Cell Biochem Suppl. 1996;24:173-85. Vescio RA, Connors KM, Kubota T, Hoffman RM. Correlation of histology and drug response of human tumors grown in native-state three-dimensional histoculture and in nude mice. Proc Natl Acad Sci U S A. 1991;88:5163-6. Ott HC, Clippinger B, Conrad C, Schuetz C, Pomerantseva I, Ikonomou L, et al. Regeneration and orthotopic transplantation of a bioartificial lung. Nat Med. 2010;16:927-33. Mishra DK, Thrall MJ, Baird BN, Ott HC, Blackmon SH, Kurie JM, et al. Human lung cancer cells grown on acellular rat lung matrix create perfusable tumor nodules. The Annals of thoracic surgery. 2012;93:1075-81. Mishra DK, Sakamoto JH, Thrall MJ, Baird BN, Blackmon SH, Ferrari M, et al. Human lung cancer cells grown in an ex vivo 3D lung model produce matrix metalloproteinases not produced in 2D culture. PloS one. 2012;7:e45308. Mishra DK, Creighton CJ, Zhang Y, Gibbons DL, Kurie JM, Kim MP. Gene expression profile of A549 cells from tissue of 4D model predicts poor prognosis in lung cancer patients. International journal of cancer Journal international du cancer. 2013. Giaccone G, Battey J, Gazdar AF, Oie H, Draoui M, Moody TW. Neuromedin B is present in lung cancer cell lines. Cancer research. 1992;52:2732s-6s. Giard DJ, Aaronson SA, Todaro GJ, Arnstein P, Kersey JH, Dosik H, et al. In vitro cultivation of human tumors: establishment of cell lines derived from a series of solid tumors. Journal of the National Cancer Institute. 1973;51:1417-23. Girard L, Zochbauer-Muller S, Virmani AK, Gazdar AF, Minna JD. Genome-wide allelotyping of lung cancer identifies new regions of allelic loss, differences between small cell lung cancer and non-small cell lung cancer, and loci clustering. Cancer research. 2000;60:4894-906. Mishra DK SJ, Thrall MJ, Baird BN, Blackmon SH, Ferrari M, Kurie JM, Kim MP. Human Lung Cancer Cells Grown in an Ex Vivo 3D Lung Model Produce Matrix Metalloproteinases Not Produced in 2D Culture. PLoS ONE. 2012;7:e45308. Himmelstein KJ, Patton TF, Belt RJ, Taylor S, Repta AJ, Sternson LA. Clinical kinetics on intact cisplatin and some related species. Clinical pharmacology and therapeutics. 1981;29:658-64. Kashkin KN, Musatkina EA, Komelkov AV, Tonevitsky EA, Sakharov DA, Vinogradova TV, et al. Genes potentially associated with cisplatin resistance of lung cancer cells. Doklady Biochemistry and biophysics. 2011;438:147-50.

AC C

1.

15

ACCEPTED MANUSCRIPT

18. 19. 20.

RI PT

17.

Shen H, Perez RE, Davaadelger B, Maki CG. Two 4N cell-cycle arrests contribute to cisplatin-resistance. PloS one. 2013;8:e59848. Dillman RO, Herndon J, Seagren SL, Eaton WL, Jr., Green MR. Improved survival in stage III non-small-cell lung cancer: seven-year follow-up of cancer and leukemia group B (CALGB) 8433 trial. Journal of the National Cancer Institute. 1996;88:12105. Ortega N, Werb Z. New functional roles for non-collagenous domains of basement membrane collagens. Journal of cell science. 2002;115:4201-14. Ali-Labib R, Louka ML, Galal IH, Tarek M. Evaluation of matrix metalloproteinase-2 in lung cancer. Proteomics Clinical applications. 2014. Guo CB, Wang S, Deng C, Zhang DL, Wang FL, Jin XQ. Relationship between matrix metalloproteinase 2 and lung cancer progression. Molecular diagnosis & therapy. 2007;11:183-92.

SC

16.

M AN U

Figure Legend:

Figure 1: Cisplatin response of CTCs from the 4D model and 2D cells. (A–B) The bar diagram shows the number of live tumor cells found after 24 hours (A) or 48 hours (B) of 50µM cisplatin treatment for A549, H1299, and H460 2D cells and CTC cells. The CTCs showed no

TE D

significant decrease in cell viability as compared to the control group, while the 2D cells showed significantly fewer cells than the control group.

EP

Figure 2: Cisplatin leads to tumor nodule regression in the 4D model: (A–B) Photograph of tumor nodules formed on the 4D model prior to the start of treatment with no significant

AC C

difference in the size of the nodules on day 4. (C–D) Photograph on day 8 with continual tumor growth in the control group (C), while there is no growth of the nodule for the 4D model treated with cisplatin (50 µM) (D). (E–F) Photograph on day 11 that shows the continuous growth of a tumor nodule in the control 4D model (E), while there is a regression of the tumor nodule with cisplatin (50 µM) treatment (F). (G) The graph shows significant nodule regression with 50 µM cisplatin treatment compared to the control group. The total nodule size did not differ prior to

16

ACCEPTED MANUSCRIPT

treatment on day 4. RUL signifies the right upper lobe resection, RML signifies right middle lobe resection, and RLL signifies right lower lobe resection. (H–P) H&E staining of lung tissues obtained on days 8, 11, and 14 from the control group and cisplatin-treated (10 µM or 50 µM)

RI PT

cells grown on the ex vivo 3D lung model. (H–J) On day 8, healthy tumor cells grew in the control group (H) and cisplatin (10 µM) group (I), while all of the cells were necrotic in the cisplatin (50 µM) group (J). (K–M) On day 11, there are healthy cells in the control group (K)

SC

and necrotic cells in the cisplatin (50 µM) group (M) while 50% of the cells in the cisplatin (10 µM) group died (L). (N–P) On day 14, healthy tumor cells remain in the control group (N), while

M AN U

most of the cells in cisplatin (10 µM) group (O) and (50 µM) group (P) died.

Figure 3: Impact of cisplatin on tumor nodule and CTC production in the 4D model: (A–C) Immunohistochemical analysis showed a higher Ki-67 staining of the cells in the control group

TE D

(44%) compared to the cisplatin (50 µM) group (4%, P<0.001). (D–F) Immunohistochemical analyses show a higher percentage of cells in the cisplatin (50 µM) group (91%) stained for Caspase-3 than in the control group (9%, P<0.0001). (G) Cisplatin treatment with 10 µM and 50

EP

µM concentrations led to an increased number of CTCs per tumor area.

AC C

Figure 4: Tumor growth and CTCs production per tumor area in the 4D model with cisplatin treatment. (A) Graph of tumor nodule size on the 4D lung model seeded with H1299 cells treated on days 4, 11, and 18 with 10 µM of cisplatin compared to the control group. (B) The average tumor nodule size area under the curve for the 4D lung model seeded with H1299, A549, H460, or HCC827 cells treated with 10 µM cisplatin are significantly smaller compared to the control group (p=0.01). (C) Graph of the number of CTCs per tumor area of the 4D lung

17

ACCEPTED MANUSCRIPT

model seeded with H1299 cells. (D) The average CTC per tumor area, area under the curve for the 4D lung model seeded with H1299, A549, H460, or HCC827 cells show significantly more

RI PT

CTC per tumor area with 10 µM cisplatin treatment compared to the control group (p=0.02).

Figure 5: MMP production per tumor area in the 4D model with cisplatin treatment. (A–D) MMP-1 (A), MMP-2 (B), MMP-7 (C), MMP-9 (D), MMP-10 (E) level per tumor area for the 4D

SC

model seeded with H1299 cells treated with 10 µM cisplatin on days 4, 11, and 18 compared to the control group. (F) The normalized average area under the curve for MMP-2 production per

M AN U

tumor area was significantly higher for the 4D model seeded with A549, H1299, and HCC827 cells compared to the control group (p=0.007) while there was no significant difference for

AC C

EP

TE D

MMP-1, MMP-7, MMP-9, and MMP-10.

18

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT