# Dynamic Phenotypic Switching and Group Behavior Help Non-Small Cell Lung Cancer Cells Evade Chemotherapy

^{1}

^{2}

^{3}

^{4}

^{5}

^{6}

^{7}

^{8}

^{9}

^{10}

^{*}

^{†}

^{‡}

## Abstract

**:**

## 1. Introduction

## 2. Results

#### 2.1. Cisplatin-Sensitive and Tolerant Cells Demonstrate Different Behaviors in Monotypic and Heterotypic Cultures

#### 2.2. Sensitive Cells Suppress Growth of Tolerant Cells in the Absence of Drug

#### 2.3. Sensitive Cells Secrete a Factor(s) that Retards the Growth of Tolerant Cells

#### 2.4. Intermittent Therapy Can Sustain a Population of Cisplatin-Sensitive Tumor Cells while Attenuating the Proliferation of Resistant Cells

#### 2.5. Epigenetic Modulation Can Distinguish Drug Sensitivity, Tolerance and Resistance in Lung Cancer

#### 2.6. Modeling Cancer Group Behavior Using Experimentally Derived Growth Curves

- Sensitive cells suppress the proliferation of the tolerant cells by secreting diffusible factors (can be overcome by increasing the frequency of tolerant cells)
- The suppressive effect is only prominent after co-culture of the two cell types for three weeks, but not if the cells are mixed and monitored immediately
- Competition by the sensitive cells is eliminated in the presence of cisplatin
- Epigenetic modifier SAHA can switch the tolerant cells to be drug-sensitive through non-genetic means, implying that these cells can switch their phenotypes in response to the environment.

- Sensitive cells generate one or more products that affect the proliferation of the tolerant cells (and possibly their own as well). We call this hypothetical product(s) ‘stress’ (we explain its significance in Supplementary Text Section S2).
- Since the cohabitation of the cells appear to change their phenotypes (e.g., stronger suppression of tolerant cells by the sensitive cells after three weeks of co-culture), and we do not have enough information to model this phenotypic change as function of the cohabitation conditions, every system (i.e., monotypic, heterotypic-12 h and heterotypic-3 weeks) must be treated as distinct with their own phenotypic parameters.
- Through mutual cooperation the tolerant cells can mitigate or neutralize the ‘stress’ generated by the sensitive cells in a frequency-dependent manner.
- Due to stochastic phenotypic switching (sensitive $\rightleftharpoons $ tolerant) by the two cellular species, a state of equilibrium exists between the two phenotypes at any point of time, where the equilibrium constant depends on the stress. As stress increases in the system, the equilibrium shifts to the right to increase the fraction of the tolerant phenotype.
- To keep the model conceptually tractable, we make the simplifying assumption that in relation to the observed growth dynamics, there is no significant difference between the true tolerant phenotype and the one produced through phenotypic switching of the sensitive cells. Likewise, the true sensitive phenotype and the one from phenotypic switching are identical as well (see caveats mentioned in the discussion).

_{GS}and K

_{GT}are the stress-dependent effective growth rates (incorporating both proliferation and cell death) of the sensitive and tolerant cells respectively. K

_{a}and K

_{b}are the rate of switching from sensitive to the tolerant phenotype and vice versa and K is the equilibrium constant of phenotypic switching (Equation (3)) [23]. We assume K

_{GS}, K

_{GT}, K to be linearly dependent on stress and K

_{b}to be fixed, although the exact functional forms that map these quantities to stress is less important, as long as a monotonic relationship is maintained. Notably, we also fit the PSMSR model assuming sigmoidal as opposed to linear relationships of the above rate parameters with stress (Supplementary Figure S6), without significant worsening of fitting error (Supplementary Figure S7). For details, see Supplementary Section S3.

_{Str}and K

_{Str,d}are the rates of stress generation and removal, respectively.

#### 2.7. PSMSR and Cisplatin Response

_{5}and AUC

_{95}represent the AUC values where 5% and 95% of the cisplatin death effect are achieved, respectively. The SCALE, AUC

_{5}and AUC

_{95}parameters are specific for the sensitive and the tolerant cell types.

#### 2.8. PSMSR in Monotypic and Heterotypic Cultures

_{0}and K

_{b}(the parameters for phenotypic switching), K

_{Gt}

_{0}(growth rate for the tolerant phenotype) and K

_{s}(rate of stress generation). One interesting observation is that the growth rate for the tolerant phenotype in monotypic and 3-week co-cultures is 7–10 times smaller than that of the sensitive phenotype (Supplementary Figure S10J), reminiscent of a persister-like trait seen in microbial systems (please see “Discussion”).

#### 2.9. Effects of Phenotypic Switching and Stress Give Rise to Diverse Game-Theoretical Strategies in Mixed Cell Populations

#### 2.10. PSMSR Model Demonstrates the Effectiveness of the Intermittent Cisplatin Therapy

## 3. Discussion

_{Gt}

_{0}in Table 1), so as to not compete for limited resources (altruism). This altruistic behavior is even more beneficial in the crowded environment of a tumor, where nutrients and oxygen could run low. The tolerant cells elucidated the evolutionary strategy of bet-hedging where they display low evolutionary fitness under normal conditions, but high fitness under stressful conditions, such as in the presence of cisplatin [39]. The intermittent therapy simulations show that the PSMSR model reproduces this behavior of the tolerant cell population (Figure 5E,F). Comparing the growth data with and without cisplatin in conjunction with the PSMSR model, we also find that the phenotypic switching (and the consequent altruism) may be turned off at high stress, when cisplatin is administered. Therefore, it appears that the altruistic stress removal benefit by the tolerant cells could be effective under normal conditions in the tumor, where a small tolerant cell population benefits the drug-sensitive cells to sustain proliferation. However, under the high stress of chemotherapy, such stress removal mechanisms may be insufficient to sustain the sensitive cell viability. It is therefore prudent to turn off phenotypic switching under such situations and allow the sensitive cells to become extinct and the tolerant cells to proliferate. While bet-hedging strategies by drug-tolerant phenotypes are well discussed in the literature [39], altruism by such phenotypes has hitherto been unexplored. There is overwhelming evidence that such tolerant persister phenotypes exist in the tumor in small proportions, even in non-drug-resistant disease. However, it is not clear whether they have a specific ecological role other than adding to the tumor heterogeneity, although recent evidence indicates that persisters can facilitate escape from drug induced toxicity by reversibly switching to slow cycling phenotypes [18].

## 4. Materials and Methods

## Supplementary Materials

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Alvarez-Arenas, A.; Podolski-Renic, A.; Belmonte-Beitia, J.; Pesic, M.; Calvo, G.F. Interplay of Darwinian Selection, Lamarckian Induction and Microvesicle Transfer on Drug Resistance in Cancer. Sci. Rep.
**2019**, 9, 9332. [Google Scholar] [CrossRef][Green Version] - Salgia, R.; Kulkarni, P. The Genetic/Non-genetic Duality of Drug ‘Resistance’ in Cancer. Trends Cancer
**2018**, 4, 110–118. [Google Scholar] [CrossRef] [PubMed] - Wu, A.; Liao, D.; Tlsty, T.D.; Sturm, J.C.; Austin, R.H. Game theory in the death galaxy: Interaction of cancer and stromal cells in tumour microenvironment. Interface Focus
**2014**, 4, 20140028. [Google Scholar] [CrossRef] - Kaznatcheev, A.; Peacock, J.; Basanta, D.; Marusyk, A.; Scott, J.G. Fibroblasts and alectinib switch the evolutionary games played by non-small cell lung cancer. Nat. Ecol. Evol.
**2019**, 3, 450–456. [Google Scholar] [CrossRef] [PubMed] - Staňková, K. Resistance games. Nat. Ecol. Evol.
**2019**, 3, 336–337. [Google Scholar] [CrossRef][Green Version] - Mirnezami, R.; Nicholson, J.; Darzi, A. Preparing for precision medicine. N. Engl. J. Med.
**2012**, 366, 489–491. [Google Scholar] [CrossRef] [PubMed][Green Version] - Cunningham, J.J.; Brown, J.S.; Gatenby, R.A.; Staňková, K. Optimal control to develop therapeutic strategies for metastatic castrate resistant prostate cancer. J. Theor. Biol.
**2018**, 459, 67–78. [Google Scholar] [CrossRef][Green Version] - Stanková, K.; Brown, J.S.; Dalton, W.S.; Gatenby, R.A. Optimizing Cancer Treatment Using Game Theory: A Review. JAMA Oncol.
**2019**, 5, 96–103. [Google Scholar] [CrossRef][Green Version] - Zhang, J.; Cunningham, J.J.; Brown, J.S.; Gatenby, R.A. Integrating evolutionary dynamics into treatment of metastatic castrate-resistant prostate cancer. Nat. Commun.
**2017**, 8, 1816. [Google Scholar] [CrossRef] - Gatenby, R.A.; Silva, A.S.; Gillies, R.J.; Frieden, B.R. Adaptive therapy. Cancer Res.
**2009**, 69, 4894–4903. [Google Scholar] [CrossRef] [PubMed][Green Version] - Hata, A.N.; Niederst, M.J.; Archibald, H.L.; Gomez-Caraballo, M.; Siddiqui, F.M.; Mulvey, H.E.; Maruvka, Y.E.; Ji, F.; Bhang, H.-e.C.; Krishnamurthy Radhakrishna, V.; et al. Tumor cells can follow distinct evolutionary paths to become resistant to epidermal growth factor receptor inhibition. Nat. Med.
**2016**, 22, 262–269. [Google Scholar] [CrossRef] [PubMed] - Venkatesan, S.; Swanton, C.; Taylor, B.S.; Costello, J.F. Treatment-Induced Mutagenesis and Selective Pressures Sculpt Cancer Evolution. Cold Spring Harb. Perspect. Med.
**2017**, 7, a026617. [Google Scholar] [CrossRef] - Bottery, M.J.; Pitchford, J.W.; Friman, V.-P. Ecology and evolution of antimicrobial resistance in bacterial communities. ISME J.
**2020**, 15, 939–948. [Google Scholar] [CrossRef] [PubMed] - Grolmusz, V.K.; Chen, J.; Emond, R.; Cosgrove, P.A.; Pflieger, L.; Nath, A.; Moos, P.J.; Bild, A.H. Exploiting collateral sensitivity controls growth of mixed culture of sensitive and resistant cells and decreases selection for resistant cells in a cell line model. Cancer Cell Int.
**2020**, 20, 253. [Google Scholar] [CrossRef] - Marusyk, A.; Tabassum, D.P.; Altrock, P.M.; Almendro, V.; Michor, F.; Polyak, K. Non-cell-autonomous driving of tumour growth supports sub-clonal heterogeneity. Nature
**2014**, 514, 54–58. [Google Scholar] [CrossRef][Green Version] - Mohanty, A.; Nam, A.; Pozhitkov, A.; Yang, L.; Srivastava, S.; Nathan, A.; Wu, X.; Mambetsariev, I.; Nelson, M.; Subbalakshmi, A.R.; et al. A Non-genetic Mechanism Involving the Integrin beta4/Paxillin Axis Contributes to Chemoresistance in Lung Cancer. iScience
**2020**, 23, 101496. [Google Scholar] [CrossRef] - Boumahdi, S.; de Sauvage, F.J. The great escape: Tumour cell plasticity in resistance to targeted therapy. Nat. Rev. Drug Discov.
**2020**, 19, 39–56. [Google Scholar] [CrossRef] [PubMed] - Mullard, A. Stemming the tide of drug resistance in cancer. Nat. Rev. Drug Discov.
**2020**, 19, 221–223. [Google Scholar] [CrossRef] - Babbs, C.F. Predicting success or failure of immunotherapy for cancer: Insights from a clinically applicable mathematical model. Am. J. Cancer Res.
**2012**, 2, 204–213. [Google Scholar] - Momeni, B.; Xie, L.; Shou, W. Lotka-Volterra pairwise modeling fails to capture diverse pairwise microbial interactions. Elife
**2017**, 6, e25051. [Google Scholar] [CrossRef] [PubMed] - Vet, S.; de Buyl, S.; Faust, K.; Danckaert, J.; Gonze, D.; Gelens, L. Bistability in a system of two species interacting through mutualism as well as competition: Chemostat vs. Lotka-Volterra equations. PLoS ONE
**2018**, 13, e0197462. [Google Scholar] [CrossRef] - Li, X.Y.; Pietschke, C.; Fraune, S.; Altrock, P.M.; Bosch, T.C.G.; Traulsen, A. Which games are growing bacterial populations playing? J. R. Soc. Interface
**2015**, 12, 20150121. [Google Scholar] [CrossRef][Green Version] - Carvalho, G.; Guilhen, C.; Balestrino, D.; Forestier, C.; Mathias, J.-D. Relating switching rates between normal and persister cells to substrate and antibiotic concentrations: A mathematical modelling approach supported by experiments. Microb. Biotechnol.
**2017**, 10, 1616–1627. [Google Scholar] [CrossRef][Green Version] - Ozawa, S.; Sugiyama, Y.; Mitsuhashi, Y.; Kobayashi, T.; Inaba, M. Cell killing action of cell cycle phase-non-specific antitumor agents is dependent on concentration--time product. Cancer Chemother. Pharm.
**1988**, 21, 185–190. [Google Scholar] [CrossRef] - Nagai, N.; Ogata, H. Quantitative relationship between pharmacokinetics of unchanged cisplatin and nephrotoxicity in rats: Importance of area under the concentration-time curve (AUC) as the major toxicodynamic determinant in vivo. Cancer Chemother. Pharm.
**1997**, 40, 11–18. [Google Scholar] [CrossRef] - El-Kareh, A.W.; Secomb, T.W. A mathematical model for cisplatin cellular pharmacodynamics. Neoplasia
**2003**, 5, 161–169. [Google Scholar] [CrossRef][Green Version] - Goldberg, D.E. Genetic Algorithms in Search, Optimization and Machine Learning; Addison-Wesley Longman Publishing Co., Inc.: Boston, MA, USA, 1989. [Google Scholar]
- Scrucca, L. GA: A Package for Genetic Algorithms in R. J. Stat. Software
**2013**, 53, 37. [Google Scholar] [CrossRef][Green Version] - Kareva, I.; Karev, G. Natural Selection Between Two Games with Applications to Game Theoretical Models of Cancer. Bull. Math. Biol.
**2019**, 81, 2117–2132. [Google Scholar] [CrossRef] [PubMed] - Bomze, I.M. Lotka-Volterra equation and replicator dynamics: A two-dimensional classification. Biol. Cybern.
**1983**, 48, 201–211. [Google Scholar] [CrossRef] - Bhattacharya, S.; Mohanty, A.; Achuthan, S.; Kotnala, S.; Jolly, M.K.; Kulkarni, P.; Salgia, R. Group Behavior and Emergence of Cancer Drug Resistance. Trends Cancer
**2021**, 7, 323–334. [Google Scholar] [CrossRef] [PubMed] - You, L.; Brown, J.S.; Thuijsman, F.; Cunningham, J.J.; Gatenby, R.A.; Zhang, J.; Staňková, K. Spatial vs. non-spatial eco-evolutionary dynamics in a tumor growth model. J. Theor. Biol.
**2017**, 435, 78–97. [Google Scholar] [CrossRef][Green Version] - Swierniak, A.; Krzeslak, M.; Borys, D.; Kimmel, M. The role of interventions in the cancer evolution-an evolutionary games approach. Math. Biosci. Eng. MBE
**2018**, 16, 265–291. [Google Scholar] [CrossRef] - West, J.; Robertson-Tessi, M.; Luddy, K.; Park, D.S.; Williamson, D.F.K.; Harmon, C.; Khong, H.T.; Brown, J.; Anderson, A.R.A. The Immune Checkpoint Kick Start: Optimization of Neoadjuvant Combination Therapy Using Game Theory. JCO Clin. Cancer Inform.
**2019**, 3, 1–12. [Google Scholar] [CrossRef] [PubMed] - Wu, A.; Liao, D.; Austin, R. Evolutionary game theory in cancer: First steps in prediction of metastatic cancer progression? Future Oncol.
**2015**, 11, 881–883. [Google Scholar] [CrossRef] - Mooney, S.M.; Jolly, M.K.; Levine, H.; Kulkarni, P. Phenotypic plasticity in prostate cancer: Role of intrinsically disordered proteins. Asian J. Androl.
**2016**, 18, 704–710. [Google Scholar] [CrossRef] [PubMed] - Goldman, A.; Majumder, B.; Dhawan, A.; Ravi, S.; Goldman, D.; Kohandel, M.; Majumder, P.K.; Sengupta, S. Temporally sequenced anticancer drugs overcome adaptive resistance by targeting a vulnerable chemotherapy-induced phenotypic transition. Nat. Commun.
**2015**, 6, 6139. [Google Scholar] [CrossRef][Green Version] - Gupta, P.B.; Fillmore, C.M.; Jiang, G.; Shapira, S.D.; Tao, K.; Kuperwasser, C.; Lander, E.S. Stochastic state transitions give rise to phenotypic equilibrium in populations of cancer cells. Cell
**2011**, 146, 633–644. [Google Scholar] [CrossRef][Green Version] - Jolly, M.K.; Kulkarni, P.; Weninger, K.; Orban, J.; Levine, H. Phenotypic Plasticity, Bet-Hedging, and Androgen Independence in Prostate Cancer: Role of Non-Genetic Heterogeneity. Front. Oncol.
**2018**, 8, 50. [Google Scholar] [CrossRef] - Brauner, A.; Fridman, O.; Gefen, O.; Balaban, N.Q. Distinguishing between resistance, tolerance and persistence to antibiotic treatment. Nat. Rev. Microbiol.
**2016**, 14, 320–330. [Google Scholar] [CrossRef] [PubMed] - Dawson, C.C.; Intapa, C.; Jabra-Rizk, M.A. “Persisters”: Survival at the cellular level. PLoS Pathog.
**2011**, 7, e1002121. [Google Scholar] [CrossRef][Green Version] - Vallette, F.M.; Olivier, C.; Lézot, F.; Oliver, L.; Cochonneau, D.; Lalier, L.; Cartron, P.F.; Heymann, D. Dormant, quiescent, tolerant and persister cells: Four synonyms for the same target in cancer. Biochem. Pharmacol.
**2019**, 162, 169–176. [Google Scholar] [CrossRef][Green Version]

**Figure 1.**

**Schematic summarizing the experiment and the source of data used in developing the theoretical cell growth models.**(

**A**) The schematic representation of the different incubation duration, ratios, and treatments used for generating the data to develop the mathematical model. (

**B**) schematic describing the principles on which the mathematical model PSMSR was developed; (

**C**) panel representing the functional form of PSMSR (please see the main text for further details).

**Figure 2.**

**Behavior of cisplatin-sensitive (S) and tolerant (T) NSCLC cells in 2D co-culture.**(

**A**) Schematic representation of the experimental design of co-culturing S and T cells in a ratio of 1:1 and collection of data points. Proliferation of sensitive (red) and tolerant (green) cells under different culture conditions in the absence or presence of cisplatin. Two-way ANOVA test (multiple comparison) showing statistical significance **** p < 0.0001. (

**B**) Sensitive and tolerant cells were plated in increasing T:S ratios and cultured for three weeks. Proliferation rate of sensitive cells (red) and tolerant cells (green) in heterotypic culture over the course of 144 h. (

**C**) Fold change in cell count of sensitive cells (red) and tolerant cells (green) in heterotypic culture was measured after 144 h for ratios 1:1, 2:1, 4:1 and 8:1. Two-way ANOVA was used for calculating statistical significance **** p < 0.0001, ns—not significant. (

**D**) Fold change in cell count of sensitive cells (purple) and tolerant cells (blue) in heterotypic culture was measured after 144 h in presence of cisplatin for ratios 1:1, 2:1, 4:1 and 8:1. Two-way ANOVA was used for calculating statistical significance *** p < 0.0001, ns—not significant. (

**E**) Change in tolerant/sensitive cells ratio with (orange) and without (black) 5 μM cisplatin over the course of 144 h was measured. Statistical significance * p ≤ 0.05, **** p < 0.0001, ns—not significant. (

**F**) Schematic representation of the conditioned medium experiment. (

**G**) The left line graph representing the effect of tolerant cell conditioned medium on sensitive cells, and the right line graph representing the inhibitory effect of tolerant cell conditioned medium on sensitive cells growth. (

**H**) Schematic representation of conditioned medium experiment to correlate the stoichiometry between cell number and inhibitory effect secreted by sensitive cells. (

**I**) The bar graph representing the inhibitory effect of condition medium on different cell number of tolerant or sensitive cells. Statistical significance information can be found in Supplementary Tables S2 and S3.

**Figure 3.**

**Tolerant cells reversibly switch their phenotype to become sensitive with intermittent therapy.**(

**A**–

**C**) Bar graph showing the ratio of tolerant versus sensitive cell population over a period of 10 days. The cell ratio for the “Continuous” group wherein the cells were continuously treated with cisplatin is shown in blue and the ratio for the “Intermittent” group wherein the cells were treated with cisplatin for two days and released in fresh medium (intermittent) is shown in black. (

**D**–

**F**) Media from “Intermittent—2 cycles” group was removed after four days of cisplatin treatment and replaced with fresh medium and the cells were allowed to grow until confluent. These cells were monitored in real-time to determine the ratio of tolerant versus sensitive over the course of 25 days. Similarly, the cells that only received cisplatin once (“Intermittent—1 cycle”) throughout the experiment were also followed for 25 days. (

**G**) Sensitive (S, red fluorescence) and tolerant (T, green fluorescence) cells were mixed at S:T ratio of 4:1 and microinjected into the perivitelline space of zebrafish larvae 48-h post fertilization (hpf). Twenty-four hours after microinjection, larvae were randomly divided into three groups: Group 1 received no drug treatment (Untreated), Group 2 received cisplatin 20 µM for three days and released with no drug for two days (Intermittent), and Group 3 received cisplatin 20 µM continuously for five days (Continuous). Ratio of tolerant versus sensitive cells was determined by measuring fluorescence intensity. One way ANOVA was used for calculating statistical significance ** p < 0.05, *** p < 0.001, ns—not significant (

**H**–

**J**) Effect of suberoylanilide hydroxamic acid (SAHA) on cisplatin-sensitive (H23), -tolerant (H2009), and -resistant (H1993) cells, demonstrating that SAHA can reverse the phenotype of H2009 from tolerant state to sensitive state. Statistical significance information can be found in Supplementary Table S4.

**Figure 4.**

**Cooperativity and stress response as described by the PSMSR model.**(

**A**) Schematic describing the PSMSR model; initially, the sensitive and the tolerant cells proliferate independently; as stress builds up, sensitive cells switch their phenotype to tolerant cells and vice versa; tolerant cells remove stress and maintain a small population, while enabling the sensitive cells to proliferate. (

**B**,

**C**) Fitting of the phenotype-switch model to the cellular growth curves of sensitive and tolerant cell populations, where the cells were mixed at different proportions and counting was started immediately; the colors represent the growth curves from different initial seeding proportions, as indicated in the legend (sensitive to tolerant cell seeding ratios); (

**D**–

**F**) predicted evolution of phenotypic switching and stress in monotypic cultures; (

**D**,

**E**) populations of sensitive and (switched) tolerant phenotypes with time, when seeded with sensitive cells only; (

**F**) stress as function of time; (

**G**–

**I**) predicted evolution of switched phenotypes and stress in heterotypic culture experiments, where cell growth was monitored immediately after mixing; (

**G**) fraction of sensitive cells that have switched to the tolerant phenotype, as function of time; (

**H**) fraction of tolerant cells that have switched to the sensitive phenotype, as function of time; (

**I**) stress with time; colors are according to the initial seeding ratio of sensitive to tolerant cells as shown in the legend; the total cell population in each case was close to 5000; (

**J**,

**K**) evolving game strategy landscape of cellular population due to stress and phenotypic switching; the heatmaps of time varying payoff values representative of inter-species competition/cooperation are shown as function of the sensitive-to-tolerant seeding ratio; payoff values are derived by fitting the PSMSR model to the competitive Lotka–Volterra equations; orange areas in the maps represent competitive behavior, green areas represent cooperative behavior; (

**J**) α

_{12}representing the effect of tolerant cells towards the sensitive cells; (

**K**) α

_{21}representing the effect of sensitive cells towards the tolerant cells.

**Figure 5.**

**Mathematical model for cisplatin resistance.**(

**A**) Schematic demonstration of AUC and cellular death rate as function of AUC; (

**B**,

**C**) fitting of the experimental growth data where the cells were co-cultured for three weeks; (

**B**): sensitive cells; (

**C**): tolerant cells; circles and lines represent the experimental and fitted trends respectively; (

**D**) SCALE parameter as measure of cisplatin sensitivity for the sensitive and the tolerant cells; the error bars represent 95% confidence limits (

**E**,

**F**) simulation of intermittent and continuous cisplatin treatment according to the protocols described in Figure 3; the initial sensitive to tolerant cell ratio was set to 4:1 with a total cell population of 50,000. (

**G**) An illustrative model depicting the presence (and absence) of group behavior among sensitive and tolerant cells under varying conditions of stress and effects of continuous versus intermittent therapy.

**Table 1.**Model parameters and parameter search ranges for PSMSR, including the 95% confidence limits.

Condition | K | K_{b} | K_{Gs}_{0} | K_{Gt}_{0} | K_{str} | K_{str,d} | a | b | g | S_{drug} | AUC_{s}^{5} | AUC_{s}^{95} | SCALE_{s} | AUC_{t}^{5} | AUC_{t}^{95} | SCALE_{t} |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|

Heterotypic | 0.049 ± 0.0003 | 3.57 ± 0.02 | 0.713 ± 0.001 | 0.687 ± 0.004 | 7.14 × 10^{−4} ± 3 × 10 ^{−6} | 5.64 × 10^{−3} ± 4 × 10 ^{−5} | 0.046 ± 0.0004 | 0.038 ± 0.0002 | 0.018 ± 0.0006 | |||||||

Heterotypic, 3 weeks | 0.052 ± 0.0007 | 2.6 ± 0.05 | 0.708 ± 0.002 | 0.189 ± 0.016 | 6.74 × 10^{−4} ± 8 × 10 ^{−6} | 5.52 × 10^{−3} ± 1 × 10 ^{−4} | 0.05 ± 0.001 | 0.038 ± 0.0006 | 0.02 ± 0.0006 | |||||||

Heterotypic, cisplatin 5 μM | 0.033 ± 0.0017 | 0 | 1.033 ± 0.046 | 0.976 ± 0.054 | 5.9 × 10^{−4} ± 2.8 × 10 ^{−5} | 0 | NA | 0.04 ± 0.003 | 0.034 ± 0.003 | 108 ± 14.5 | 239 ± 12 | 1153 ± 58 | 8.61 ± 0.5 | 233 ± 17 | 1201 ± 49 | 5.79 ± 0.6 |

Heterotypic, 3 weeks, cisplatin 5 μM | 0.033 ± 0.0015 | 0 | 0.966 ± 0.05 | 0.967 ± 0.05 | 5.8 × 10^{−4} ± 3.3 × 10 ^{−5} | 0 | NA | 0.039 ± 0.002 | 0.036 ± 0.003 | 96.8 ± 9.95 | 271 ± 14 | 1098 ± 44 | 8.91 ± 0.5 | 250 ± 14 | 1098 ± 47 | 6.89 ± 0.7 |

Parameter search range | 0–0.1 | 0–5 | 0–2 | 0–2 | 0–0.001 | 0–0.02 | 0–0.1 | 0–0.1 | 0–0.1 | 1–500 | 1–500 | 10–1500 | 0.1–20 | 1–500 | 10–1500 | 0.1–20 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Nam, A.; Mohanty, A.; Bhattacharya, S.; Kotnala, S.; Achuthan, S.; Hari, K.; Srivastava, S.; Guo, L.; Nathan, A.; Chatterjee, R.; Jain, M.; Nasser, M.W.; Batra, S.K.; Rangarajan, G.; Massarelli, E.; Levine, H.; Jolly, M.K.; Kulkarni, P.; Salgia, R. Dynamic Phenotypic Switching and Group Behavior Help Non-Small Cell Lung Cancer Cells Evade Chemotherapy. *Biomolecules* **2022**, *12*, 8.
https://doi.org/10.3390/biom12010008

**AMA Style**

Nam A, Mohanty A, Bhattacharya S, Kotnala S, Achuthan S, Hari K, Srivastava S, Guo L, Nathan A, Chatterjee R, Jain M, Nasser MW, Batra SK, Rangarajan G, Massarelli E, Levine H, Jolly MK, Kulkarni P, Salgia R. Dynamic Phenotypic Switching and Group Behavior Help Non-Small Cell Lung Cancer Cells Evade Chemotherapy. *Biomolecules*. 2022; 12(1):8.
https://doi.org/10.3390/biom12010008

**Chicago/Turabian Style**

Nam, Arin, Atish Mohanty, Supriyo Bhattacharya, Sourabh Kotnala, Srisairam Achuthan, Kishore Hari, Saumya Srivastava, Linlin Guo, Anusha Nathan, Rishov Chatterjee, Maneesh Jain, Mohd W. Nasser, Surinder Kumar Batra, Govindan Rangarajan, Erminia Massarelli, Herbert Levine, Mohit Kumar Jolly, Prakash Kulkarni, and Ravi Salgia. 2022. "Dynamic Phenotypic Switching and Group Behavior Help Non-Small Cell Lung Cancer Cells Evade Chemotherapy" *Biomolecules* 12, no. 1: 8.
https://doi.org/10.3390/biom12010008