Biophysical and biochemical characteristics as complementary

Subscriber access provided by IDAHO STATE UNIV

Article

Biophysical and biochemical characteristics as complementary indicators of melanoma progression Justyna Anna Bobrowska, Kamil Awsiuk, Joanna Pabijan, Piotr Bobrowski, Janusz Lekki, Katarzyna M Sowa, Jakub Rysz, Andrzej Budkowski, and Malgorzata Lekka Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01542 • Publication Date (Web): 16 Jul 2019 Downloaded from pubs.acs.org on July 18, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Biophysical and biochemical characteristics as complementary indicators of melanoma progression Justyna Bobrowska1, Kamil Awsiuk2, Joanna Pabijan1, Piotr Bobrowski3, Janusz Lekki1, Katarzyna M. Sowa2, Jakub Rysz2, Andrzej Budkowski2, Małgorzata Lekka1* 1 Institute of Nuclear Physics, Polish Academy of Sciences, PL-31341 Kraków, Poland, 2 M. Smoluchowski Institute of Physics, Jagiellonian University, Łojasiewicza 11, PL-30-348, Kraków, Poland, 3 Institute of Metallurgy and Materials, Polish Academy of Sciences, Reymonta 25, 30-059 Kraków, Poland. ABSTRACT: The multi-step character of cancer progression makes it difficult to define a unique biomarker of the disease. Interdisciplinary approaches, combining various complimentary techniques, especially those operating at a nanoscale level, potentially accelerate characterization of cancer cells or tissue properties. Here, we study a relation between the surface and biomechanical properties of melanoma cells, measured by mass spectrometry (ToF SIMS) and atomic force microscopy (AFM). In total, seven cell lines have been studied. Six of them were melanoma cells derived from various stages of tumor progression: (1) WM115 cells derived from a 55 year old female skin melanoma at a vertical growth phase (VGP) in the primary melanoma site, (2) WM793 cells established from the vertical growth phase (VGP) of a primary skin melanoma lesion, (3) WM266-4 cells established from a cutaneous skin metastasis detected in the same patient as WM115 cells, (4) WM239 cells derived from a cutaneous skin metastasis, (5) 1205Lu cells originated from a lung metastasis diagnosed in the same patient as WM793 cells, (6) A375P – cells were derived from a solid malignant tumor located in the lung. As a reference cell line, human epidermal melanocytes from adult skin (primary cell line HEMa-LP) was used. Cellular deformability reveals low, medium and large deformability of melanoma cells originating from vertical growth phase (VGP), skin and lung metastasis, respectively. These changes were accompanied by distinct outcome from principal component analysis (PCA). In relation to VGP melanoma cells, cells from skin and lung metastasis reveal similar or significantly different surface properties. The largest deformability difference observed for cells from VGP and lung metastasis was accompanied by the largest separation carried out based on unspecific changes in their surface properties. In this way, we show the evidence that biomechanical and surface biochemical properties of cells change in parallel, indicating a potential of being used as nanobiophysical fingerprints of melanoma progression.

Cutaneous melanoma is one of the most aggressive and lethal cancer of the skin. It originates from melanocytes that, due to oncogenic mutations, transform through nevus stages to a flat tumor, growing horizontally (radial growth phase, RGP). Further development of tumorigenic phenotype results in a tumor, which invades to deeper layers of the skin (vertical growth phase, VGP), leading to melanoma metastasis1,2. Despite advances in the melanoma treatment3, the number of deaths caused by melanoma is still increasing. Thus, searching for new biomarkers or their combination is still a great challenge because, so far, there is no single biomarker having an ability to detect cancer with high specificity and sensitivity4. That is why interdisciplinary approach combining complimentary techniques and the convergence of diverse disciplines induces, potentially, an acceleration in cancer diagnosis and therapy5,6. Here, we combine two potential biomarkers of cancer progression that stem from distinct manifestations of oncogenic transformation, both independently occurring in cancer

cells, including melanoma. They are, namely, alterations in biomechanical and surface properties of cells. The former is quantified by Young’s (elastic) modulus, while the latter is described with the use of principal component analysis (PCA) applied to mass spectrometry data. ToF-SIMS and AFM techniques have been applied here to characterize and compare properties in the seven cell lines of melanoma cells and melanocytes. In past decades, biomechanical properties of cells measured at the single cell level gained large significance. The manifestation of various diseases has been shown to be strongly linked with mechanical properties of cells7. Among others, biomechanical alterations occur in osteoarthritis8, asthma9, malaria10,11, and cancer12,13. Atomic force microscopy (AFM) is a widely applied and wellestablished technique in a quantitative characterization of biological samples, ranging from single proteins to living cells 14,15. It measures a surface topography with a very good spatial resolution and, also, it enables to analyze mechanical properties of various materials, including

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 15

living cells. While the majority of solid tumors appears to be more rigid during palpation examination, it has been already demonstrated that individual cancer cells are more deformable than their normal counterpartners 16–18. Various studies reported increased cellular deformability for bladder, prostate, breast, thyroid cancers 19–22, connected with poorly developed cytoskeleton, mainly actin filament network22,23. Thus, biomechanical characterization enables to get an insight into mechanisms underlying the cancer progression24. The large variability of alterations at the molecular level is frequently observed during cancer progression.25 Mass spectrometry can be successfully applied to evaluation of overall changes in the biochemical/biomolecular composition of the cells. One of mass spectrometry techniques is the time-of-flight secondary ion mass spectrometry (ToF-SIMS), characterized by high transmission and high mass resolutions 26–29. So far, gathered research report on the applicability of the ToFSIMS in detection of overall cancer-related (bio)chemical changes in breast30,31, prostate32, and bladder33 cancers, however, the complexity of the analysis increased with the number of cell lines considered. In our studies, we demonstrated that biomechanical and surface biochemical properties of melanoma cells alter in the course of cancer progression. We have conducted the AFM and ToF-SIMS measurements and the analysis for six melanoma cell lines divided into three groups of cells, namely: cells originating from VGP (primary tumor site) and cells derived from skin and lung metastasis. Obtained results show that the degree of cellular deformability and variability in surface properties increases with the advancing stage of cancer progression. Compared to the human melanocytes, our data constitutes a complementary set of nanobiophysical fingerprints that could be applied subsequently to differentiate between the stages of melanoma progression.

The HEMa-LP melanocytes required a special, dedicated medium, i.e. MEDIUM 254 (GIBCO).

EXPERIMENTAL SECTIONS Cell culture. For the presented study seven cell lines

Sample preparation for ToF-SIMS measurements. As a support for cell cultures, commercially available silicon (Si-Mat, Germany) 1 cm2 square dices were used. They were cleaned with pressurized nitrogen and sterilized with a UV lamp for 1 hour on both sides. When cells were seeded, silicon slices were moved into Petri dishes (Sarstedt), filled with the corresponding culture medium and next cultured for 48 h in the CO2 incubator (Nuaire), providing 95%/5% air/CO2 atmosphere. After 48 h of culture, silicon substrates with cells were washed with phosphate buffered solution (PBS, Sigma), followed by fixation in 3.7% solution of paraformaldehyde (Fluka) in PBS for 20 minutes, at room temperature (RT). Next they were rinsed in subsequent PBS solutions and dehydrated by alcohol according to protocol described in our previous papers 33,41. Finally, samples were air dried for 20 minutes in RT.

were chosen. Six of them were melanoma cell lines from ESTAB Melanoma Cell Bank. Cells were derived from three different stages of tumor progression: (1) WM115 cells derived from a 55 year old female skin melanoma at the vertical growth phase (VGP) in the primary melanoma site, (2) WM793 cells established from the vertical growth phase of a primary skin melanoma lesion, (3) WM266-4 cells established from a cutaneous skin metastasis detected in the same patient as WM115 cells, (4) WM239 cells derived from a cutaneous skin metastasis, (5) 1205Lu cells originated from a lung metastasis diagnosed in the same patient as WM793 cells, (6) A375P – cells were derived from a solid malignant tumor located in the lung. As a reference cell line, human epidermal melanocytes from adult skin (primary cell line HEMa-LP, kind gift of P.Hinterdorfer and L.Chtcheglova, Linz University, Austria) was used. All melanoma cell lines, used in this study, were cultured in the RPMI-1640 medium (Sigma), supplemented with 10% fetal bovine serum (FBS, Sigma).

Sample preparation for AFM measurements. For elasticity measurements using AFM, cells were seeded on glass coverslips placed in the Petri dishes (Sarstedt) filled with the corresponding culture medium. The culture time was set to 48 hours. Afterwards, glass coverslips with cells were immersed into the AFM liquid cell setup, filled with the corresponding fresh culture medium and placed on the AFM piezo-scanner. AFM-based elasticity measurements and data analysis. All measurements were carried out using an AFM model XE120 (Park Systems, Korea) working in a force spectroscopy mode in liquid conditions. V-shaped silicon nitride cantilevers were applied, characterized by the nominal spring constant of 0.01 N/m and the open angle of 21⁰ (MLCT C, Bruker). Elasticity measurements were carried out for living cells in a following manner. The force curves (dependencies of cantilever deflection registered as a function of relative scanner position) were recorded over a scan area varying from 5 x 5 μm2 up to 10 × 10 μm2. A grid of 8 × 8 or 7 × 7 or 6 × 6 points was set over the scanned region. On average, 36 – 64 force curves for each single cell were automatically recorded at the approach and retract velocity of 8 μm/s. The total number of measured cells for each studied cell line was not less than 50 cells. Force versus indentations curves were determined by a subtraction of calibration curves (recorded on a stiff, non-deformable sample) from the curves recorded on cells. The resulting curves were analyzed using the Hertz-Sneddon model, delivering Young’s (elastic) modulus 35. The details of the procedure are presented in the Suppl. Mat. Note 1.

ToF SIMS measurements. The ToF-SIMS 5 (ION-TOF, GmbH, Munster), equipped with a 30 keV bismuth liquid ion gun was used to collect mass spectra of dried cells,


Page 3 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


previously cultured on silicon substrates. High-resolution m/z mass spectra (m/Δm > 6500) of positive secondary ions, emerging from samples bombarded with the Bi3+ primary ion beam, were acquired with a time-of-flight mass spectrometer. The primary beam was rastered over the area of 150 μm × 150 μm with single cell in the center observed using microscope camera. To maintain the static conditions (< 1012 ions/cm2 for primary ion beam), the time of acquisition and the primary beam current were adjusted to remain the sample surface intact. The total dose per one measurement was of the order of 108 of primary ions, what yields the dose density of 1012/cm2. Collected mass spectra were further analyzed using PCA in order to find alterations among studied cell lines. The details of the procedure are presented in the Suppl. Mat. Note 2. Statistics. All AFM measurements were performed for at least 50 cells per each cell line, in at least three independent samples. A map of force curves (at least 5 × 5) was collected per each cell. AFM data are presented as the Young’s modulus distributions of all collected force curves. Statistical significances of AFM data were verified using non-parametric Mann-Whitney test (at the significance level 0.05), performed by means of OriginLab 8 software. ToF-SIMS measurements were performed for at least two independent samples. In total 10 mass spectra were collected for each investigated cell line. Mass spectra were analyzed using Principal Component Analysis PCA (Suppl. Mat. Note 2). Statistical significances of PCA results was verified using confidence ellipses calculated at three confidence levels of 90%, 95%, and 99% performed by means of Mathematica software.

RESULTS Cellular deformability increases with advancing melanoma progression. AFM working in the force spectroscopy mode was applied to quantify nanomechanical properties of single melanoma cells, analogously as published elsewhere34–36. The basis for the calculations were force curves, that are, shortly, dependencies between the load force and the relative sample position. Subtracting a calibration force curve recorded on stiff substrate (usually a glass slide) from that recorded on an individual cell (here a single melanoma cell) delivers the basis for the quantitative determination of elastic (Young’s) modulus, by employing widely used Hertz contact mechanics 37,38 (Fig. 1a, details are in the Suppl. Mat. Note 1). Each single measurement, carried out above a cell center, was used to create a distribution of modulus (Fig. 1b). In our studies, six melanoma cell lines were grouped according to the stage of melanoma progression. Melanocytes were used as reference cells. The first group comprised melanoma cells originating from primary tumor site located in the skin. There were two cell lines (WM793 and WM115) derived from the vertical growth phase (referred here as VGP melanoma cells). The next two groups were cells originating from the metastasis of melanoma to the skin (cell lines WM239

and WM266-4) and lung (1205Lu and A375-P cells). To compare the elasticity between the studied cell lines, Young’s modulus was calculated for the indentation depth of 600 nm (Fig. 1b&c) as, in such a case, the mechanical response contains large contribution from deeper part of the cells and can be attributed to overall elastic properties of cells39.

Figure 1. a) AFM-based elasticity measurements enable to compare force curves recorded on a stiff non-deformable substrate (a glass or Petri dish) with that acquired on a soft melanoma cell. The difference between these curves is a basis for Young’s modulus determination. b) Results were divided into four groups: melanocytes, VGP melanoma cells (from primary melanoma site), skin and lung metastasis. For each group, moduli distributions were created and fitted with lognormal function. It is the largest for melanocytes, then for VGP cells followed by two melanoma metastases to skin and lung. Differences between all considered groups of data are statistically significant based on Mann-Whitney test at the level of 0.05. c) Young’s modulus is expressed by a mean ± standard deviation determined by fitting a lognormal function to moduli distribution for each cell line separately.

Histograms of Young’s modulus are not symmetric and follow a log-normal type of distribution (Fig. 1b). The narrowest distribution was observed for melanoma cells from the lung metastasis whereas the broadest one was observed for the melanocytes. The obtained results shows that melanocytes (E = 14.3 kPa, logarithmic standard deviation, log SD = 0.80) were more rigid, as compared to all other studied melanoma cell lines, what agrees with the already reported data on cancer cells deformability19– 22,38. VGP melanoma cells were slightly more deformable as compared to melanocytes. Their modulus is 9.91 kPa, log SD = 0.55. Regardless of the type, metastatic cells are characterized by smaller Young’s modulus, indicating their larger deformability, namely E = 7.8 kPa (log SD = 0.61) and E = 5.7 kPa (log SD = 0.61), respectively for skin and lung metastases. As it has been mentioned above, two distinct cell lines per each group were



measured, both revealing similar level of the elastic properties (Fig. 1c). Distributions of the Young’s modulus per each cell line are presented in the Suppl. Fig. S1 while Young’s modulus with left and right values of the standard deviation are presented in the Table S1. Variability in surface biochemical properties of cells increases with advancing stage of cancer progression. To elaborate whether alterations in cell deformability are accompanied by changes in the surface biochemical properties of cells, ToF-SIMS mass spectra were recorded for each cell line separately within the molecular mass range from 0 to 500 Da (Fig. 2a). Bi3+ ions were used to bombard the surface of fixed and dehydrated cells. ToF SIMS has been chosen as its surface sensitivity using Bi3+ primary ions is characterized by escape depth of ~3 nm for molecular secondary ions40. Therefore, the sampled surface properties correspond to the outermost regions of the cell surface.

Figure 2. a) Using Bi3+ clusters to bombard the cell surface results in mass spectra with molecular mass range from 0 to 500 Da. Complexity of cell surface biochemical properties translates into a complexity of mass spectra (b), thus, principal component analysis (PCA) has to be employed to differentiate among the studied cell lines. PCA outcome presented as a 3D scores plot (c) shows a clear grouping related to stages of melanoma progression. Each dot represents a single mass spectrum acquired on a single cell. The variability, described by each orthogonal principal component, is as follows: PC1 = 37.64%, PC2 = 20.21%, and PC3 = 11.42% (here, ellipsoids represent double standard deviation range along the PC1, PC2, and PC3 axes). To find better differentiation between various stages of melanoma development, 2D plots of PC components are presented (d). For each data set, confidence ellipses at confidence levels of 90%, 95%, and 99% were calculated.

The information on differences allowing to identify the specific stage of melanoma progression is hidden in the recorded mass spectra (Fig. 2b, see Suppl. Fig. 2). To uncover it, a principal component analysis (PCA) with autoscaling as the pre-processing method was applied, according to the protocol presented in our earlier papers33,41,42 (details of PCA are briefly presented in the

Page 4 of 15

Suppl. Mat. Note 2). 3D score plots of PCA enable to visualize how particular groups of cells overlap (Fig. 2c). In our case, each dot in the plot corresponds to a single mass spectrum recorded for a single cell. Ellipsoids represent double standard deviation range along chosen principal component (PC) axes. Despite large overlapping of PCA results, one can see that four stages of melanoma progression differ noticeably (each stage of melanoma progression is represented by two cell lines). Percentage of the total variance accounting for the melanoma group variability, explained by each orthogonal principal component, are as follows: PC1 = 37.64%, PC2 = 20.21%, and PC3 = 11.42% (Fig. 2c, loadings plots are presented in the Suppl. Fig. S3). To find the best differentiation between melanoma stages in relation to a composition of culture media, detailed analysis of 2D plots of PC components (e.g. PC2 versus PC1, Fig. 2c) was performed. The obtained PCA outcome shows the largest separation between studied groups of cell lines for PC2 plotted versus PC1. Importantly, a region corresponding to the composition of culture media used for the experiment (either 254 or RPMI 1640) does not overlap with cells’ regions. Several characteristic features can be obtained from the plot. Firstly, PCA outcome related to melanocytes shows their smaller variability in comparison with melanoma cells. Melanocytes are well-separated from cells (1205Lu and A375-P) derived from lung metastasis while they overlap with cells originating from VGP cells and skin metastasis. PCA outcome for lung metastasis only partially overlaps with the metastasis to skin (represented by WM239 and WM266-4 cell lines). The largest variability is observed for VGP cells encompassing all studied groups attributed to a given stage of melanoma progression. PCA better distinguishes lung than skin metastases from VGP melanoma cells. Identification of closely related cells, possessing similar morphology and properties43 is not easy. Melanocytes overlap fully with VGP melanoma cells what indicate that their surface in biochemically similar. Next, we ask question whether by reducing complexity (here removing mass spectra of melanocytes), it is possible to enhance the difference between cells from VGP, skin and lung metastases. Thus, PCA has been repeated in relation to culture medium composition treated here as a control for PCA calculation as these data should always separate from cells (Fig. 3). Two groups containing cells of similar origins were compared, i.e. VGP melanoma cells/skin metastasis and VGP melanoma cells/lung metastasis. PCA outcomes for both groups of melanoma cells show clear separation of data containing information about the chemical composition of the culture media. VGP melanoma cells poorly separate from cells derived from skin metastasis. Only PCA outcome obtained for WM793 cell line is located away from the PCA points obtained for the other groups of cells (Fig. 3c, loading plots are presented in Suppl. Fig. S4a). The other VGP cell line, i.e. WM115, overlaps with cells originating from skin metastasis


Page 5 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(WM239). PCA of both cell lines representing skin metastasis demonstrates large variability of WM266-4 cells and much smaller one for WM239 ones. WM266-4 cell line separates well from VGP WM793 melanoma cells and partially overlaps with WM115 (VGP melanoma) and WM239 (skin metastasis).

Figure 3. 3D score plots of PCs visualize variability of VGP melanoma cells (WM115 & WM793) in relations to the origin of metastasis, i.e. (a) skin (WM239 & WM266-4) and (b) lung (1205Lu & A375-P) metastasis. Culture media composition did not affect melanoma stage identification as all melanoma cells were cultured in the same medium type and composition. 2D score plots of PC2 plotted versus PC1 show the best separation, statistically verified by confidence ellipses calculated at the confidence levels of 90%, 95%, and 99%. PCA outcome for mass spectra recorded for skin metastasis (c) reveals cell-dependent separation while PCA outcome for lung metastasis (d) presents clearer separation of the studied cell lines.

Obtained percentage of the total variance, accounting for the melanoma group variability explained by each orthogonal principal component, are as follows: PC1 = 46.73%, PC2 = 19.37%, and PC3 = 8.68%. The comparison of VGP cells with lung metastasis shows well-separated cell lines (Fig. 3d, loading plots are shown in Suppl. Fig. S4b). The latter ones (1205Lu & A375-P) partially overlap. Their PCA outcome is located away from that obtained for VGP melanoma cells. Obtained percentages of the total variance, accounting for the melanoma group variability explained by each orthogonal principal component, are as follows: PC1 = 39.13%, PC2 = 26.52%, and PC3 = 14.59%. In conclusion, it seems that PCA of ToF-SIMS mass spectra distinguished the lung metastasis better than the skin one. PCA differentiates between primary and secondary tumor (melanoma) sites. Among cell lines studied by us, there are two pairs of cells that are derived from the same patients from two tumor sites, i.e. from primary (skin) and secondary (metastatic, either skin or lung) tumor sites. The first pair is WM115 and WM266-4 and the other is WM793 and 1205Lu (Fig. 4). The percentages

of the total variance for WM115 and WM266-4 cell lines are PC1 = 43.12%, PC2 = 25.17% and PC3 = 9.30% (Fig. 4a, loadings plots are presented in the Suppl. Fig. S5a). Cells derived from lung metastasis (1205Lu) compared to VGP melanoma cells (WM793) are characterized by PC1 = 44.34%, PC2 = 30.55% and PC3 = 15.22%. (Fig. 4b, loadings plots are presented in the Suppl. Fig. S5b).

Figure 4. 2D score plot (PC2 versus PC1) shows that biochemical properties of the cell surface can differ in primary and secondary tumor sites i.e. between (a) VGP/skin metastasis (WM115 and WM266-4) and (b) VGP/lung metastasis (WM793 and 1205Lu). The confidence ellipses calculated at three confidence levels of 90%, 95%, and 99% statistically confirm the difference.

PCA outcome shows lower degree of cell lines differentiation in the case of skin metastasis. Mass spectra recorded for cells derived from the lung metastasis clearly separate from the corresponding primary tumor site.

DISCUSSION Among spectroscopic techniques analyzing either molecular masses or vibrational oscillations of chemical bonds, ToF-SIMS seems to be a valuable tool in the cancer research, as with its high specificity and nanoscale sensitivity, it enables to study cancer-related alterations at the single cell level31,32,41. This is of particular importance as it has a potential to be applied in the clinical practice for a large data collection. Changes in biochemical composition of a cell surface are not the only the manifestation of cancer. Large amount of literature data reveals that deformability can be used as a biomarker of oncogenic changes that has been shown to correlate with the cancer progression7,12,19–21. Thus, a combination of biochemical and biomechanical features of the studied biological samples may deliver a novel approach using independent manifestations of changes associated with the disease leading to better identification and diagnosis of cancer. At the current stage of knowledge, it is difficult to assume that the surface biochemical properties are the cause for altered biomechanics (or vice versa). It might be that these manifestations of the cancer related changes proceed independently, however they are to certain extend correlated. There are some prerequisites supporting the existence of a relation between deformability of cells and their surface biomechanical properties. In our first paper, we have demonstrated that



ToF SIMS spectrometry can be successfully applied to differentiate between few bladder cancer cell lines33. In the reported case, PCA revealed large differences in the chemical composition of bladder cancer cells, between non-malignant cell cancer of ureter (HCV29) and cancerous cells including three lines: T24 (transitional cell carcinoma), HT1376 and HTB-9 (the latter two being bladder carcinoma). The applied methodology of PCA has distinguished between various types of cancer bladder cells without a’priori determination of specific masses in ToF SIMS spectra that could facilitate the discrimination between cell lines. The PCA was sensitive enough to find substantial differences between cells based on analysis of the whole range of acquired spectra. By analyzing the molecular mass range dominating in the separation of non-malignant cells from the cancerous ones, it was possible to select peaks at molecular masses that influenced strongly the obtained separation. Based on literature data the proposed peak assignments pointed out molecular mass characteristic for lipids and for some amino acids, such as arginine or tryptophan. These findings show that the invasive phenotype of human bladder cancer cells can be correlated with the alterations in the chemical composition of the studied cells. These cells are also characterized by distinct deformability, as it has been determined earlier19,45. Results show that nonmalignant cells are significantly more rigid (E = 10.0 ± 3.0 kPa) as compared to cancer cells (placed within the range of E = 3.6 kPa – 3.8 kPa)45. In analogous research, prostate cancers delivered form lymph node (LNCaP) and bone (PC-3) metastases reveal clear separation from PNT2 cells (transfected normal prostate epithelial cells with genome of SV40 virus). Also, a range of lipids and amino-acids (proteins) related signals has been proposed to be responsible for the discrimination among prostate cells32. In independent AFM measurements, e.g. conducted by Faria et al.20, the deformability of these cells has been shown to be related to cancer progression, i.e. LNCaP cells are more deformable as compared to PC-3 and PNT2-C2 ones. Lekka et al. have shown that prostate cells differ in the elastic properties significantly, PC-3 cells are much stiffer (E = 1.97 ± 0.41 kPa) in comparison with LNCaP (E = 0.46 ± 0.17 kPa)14. Another example of studies showing a sort of a correlation between the surface chemistry and the cellular deformability is a breast cancer. Using ToF-SIMS, it was possible to differentiate between cells originating from estrogen receptor positive (ER+, MCF-7 and T47D cell lines) and negative (ER-, MDA-MB-231 cell lines) groups. Results obtained from cell extracts show the well-separated groups of points, but the separation has not been correlated with the invasive potential of the studied cell lines. When the analysis of individual cells was carried out, the results showed that the ER+ MCF-7 and T47D cells, which have a similar noninvasive phenotype, were closely located and wellseparated from the ER-, highly metastatic MDA-MB-231 cells46. In another study31, ToF-SIMS has been applied to investigate lipid-related metabolites in breast cancer cell lines classified according to TNBC (triple negative breast cancer) characteristics. TNBC cells are negative for

Page 6 of 15

estrogen (ER-), progesterone (PR-), and herceptin 2 (HER2-) receptors. Among studied cell lines, mass spectra were collected from four, triple negative classified cell lines, three estrogen and progesterone positive cell lines (ER+ and PR+), and one cell line that was triple positive (ER+, PR+, and HER2+). Results show that for the positive ions, PCA identified phosphocholine and cholesterol as the largest sources of variance for TNBC cells while monoacylglycerols and diacylglycerols contributed mostly in differentiation of receptor positive cell lines. In one of the earlier studies, it has been shown that malignant (MCF-7) breast cancer cells have Young’s modulus significantly lower, about 1.4–1.8 times, as compared to the non-malignant (MCF10A) counterpartners, indicating larger rigidity of the latter cells47. Based on these data, it is difficult to plot the direct correlation between ToF-SIMS and AFM results as function of cancer stage. However, gathered data on biomechanical properties of breast cancer cell lines allow us to postulate that large cancerrelated alterations in cellular deformability are accompanied by large changes in surface biomechanical properties as compared to the normal cells. Thus, the question arises what happens in the case of closely related stage of cancer progression. In our studies, chosen melanoma cell lines were derived from various stages of cancer progression, namely: from VGP cells, skin and lung metastasis, and next compared to melanocytes (Fig. 5).

Figure 5. Summary of surface and biomechanical changes in melanoma cells.

Biomechanical characteristics of melanoma cells show the increased cellular deformability (i.e. the Young’s modulus decreases) with advancing melanoma progression (Fig. 1). Melanocytes are characterized by the highest Young’s moduli while cells derived from lung metastasis are the most compliant ones. PCA of ToF-SIMS mass spectra shows that melanocytes separate clearly from all melanoma cells (Fig. 2), indicating distinct biochemical characteristics of melanocyte’s surface. This is not surprising as normal cells possess totally different phenotype compared to cancer ones. Assuming that


Page 7 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


biochemical and biomechanical properties of cells can have a potential to be independent biomarkers identifying cancer progression, it was interesting to analyze the PCA separation among melanoma cells without the presence of melanocytes (still we have preserved the medium as a control of our PCA approach). Results show larger separation between cells belonging to VGP and skin metastasis groups. Smaller separation level has been observed for cells derived from VGP and skin metastasis. The latter ones are very similar cells representing skin, thus, the overlapping indicates similarities in surface biochemical properties (Fig. 4). However, the deformability of skin metastatic cells was larger as compared to VGP cells (Fig. 1). Even more pronounced deformability change was observed for cells from lung metastasis that were well-separated by PCA of ToF-SIMS mass spectra. The elasticity change was accompanied by the clearest separation of these cells in PCA plot (see exemplary molecular masses contributing to separation between skin and lung metastasis in Table S2). Comparing elasticity of all studied cells, the following relation of cellular deformability can be plotted: VGP cells (less deformable i.e. more rigid) > cells originating from skin metastasis (medium deformability) > cells originating from lung metastasis (largest deformability). PCA outcomes in relation to the VGP melanoma cells can be summarized as follows: cells originating from VGP and skin metastasis possess similar surface biochemical properties (Fig. 3a) while cells derived from lung metastasis have significantly different surface (Fig. 3b).

cancer cells it is possible to obtain non-specific markers describing melanoma progression. These results can open new opportunities for assessing novel labelled-free biomarkers for melanoma staging that may support cancer detection and diagnosis.

It is worth to highlight that among the studied cell lines, there are two pairs of cells that are derived from the same patients. Namely, WM266-4 cells are the metastasis of WM115 cells to skin and 1205Lu are the metastasis of WM793 to lung (Fig. 4). Deformability of cells follows the general relation, Young’s modulus of WM115 (VGP) cells is not significantly different from the corresponding WM266-4 cells (skin metastasis). Its value determined for WM793 (VGP) cells is higher as compared to cells from 1205Lu cells (lung metastasis). Analogously, PCA outcomes show larger separation between cells derived from VGP and lung metastasis than for cells from VGP and skin metastasis. These findings follow the results for larger groups of cells (Fig. 3). They indicate that the variability stemming from individual patient-related changes can be negligible (at least for melanoma cells, for which PCA of four cells lines delivers similar outcome as for two cell lines).

* [email protected]

CONCLUSIONS Cancer is characterized by the aberrant growth of cells that have collected mutations in genes controlling cell proliferations and survival. Even if it diagnosed for a specific organ or tissue, it is characterized by a large degree of heterogeneity. High-resolution techniques working in nanoscale enable to identify a group of nonspecific changes are can help the diagnosis. In our studies, we have demonstrated that by combining biochemical and biophysical characteristics of

ASSOCIATED CONTENT Supporting Information File: Bobrowska_et_al_SI.pdf; TOC: 1. Note 1. The Young’s modulus determination 2. Figure S1. Elasticity of cells 3. Note 2. Details of PCA 4. Figure S2. Exemplary mass spectra recorded for all studied cell lines. 5. Figure S3. Scores and loading plots for PCA obtained for all cells 6. Figure S4. Scores and loading plots for PCA obtained for a) cells from VGP melanoma and skin metastasis, and b) cells from VGP melanoma and lung metastasis. 7. Figure S5. Scores and loading plots for PCA obtained between two pairs of cells: a) VGP melanoma and skin metastasis, and b) VGP melanoma and lung metastasis. 8. Table S1. Summary of elastic properties of melanoma cells regarding melanocytes. 9.

AUTHOR INFORMATION Corresponding Author

ORCID IDs: Justyna Bobrowska: 0000-0002-5382-1465 Małgorzata Lekka: 0000-0003-0844-8662 Kamil Awsiuk: 0000-0001-9058-4561 Joanna Pabijan: 0000-0003-4518-1389 Janusz Lekki: 0000-0001-9848-6448 Katarzyna M.Sowa: 0000-0001-5259-3213 Jakub Rysz: 0000-0003-1668-3398 Andrzej Budkowski: 0000-0001-5200-3199

Author Contributions JB & ML designed experiment. JB performed AFM and ToFSIMS measurements, KA participated in PCA analysis and measurements, JP was involved in sample preparation, PB wrote routine for 3D presentation of PCA results, JL was involved in AFM data analysis and interpretation, KMS wrote a routine for the confidence ellipse calculations, JR & AB were involved in ToF-SIMS analysis and data interpretation, ML contributed to the design of experiments and result analysis. JB & ML wrote the manuscript. All authors contributed to results discussion and manuscript writing. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Competing interest The authors declare no competing interests.



ACKNOWLEDGMENT The authors are grateful to prof. Peter Hinterdorfer and dr Lilia Chtcheglova (Linz University, Austria) for their kind gift of melanocyte cell lines. This work was supported by National Science Centre (Poland) project no UMO2013/11/N/ST4/01860. The JPK purchase has been realized under the project co-funded by the Małopolska Regional Operational Program, Measure 5.1 - Krakow Metropolitan Area as an important hub of the European Research Area for 2007-2013, project no project no MRPO.05.01.00-12-013/15. The research was carried out with the equipment (TOFSIMS) purchased thanks to the financial support of the European Regional Development Fund in the framework of the Polish Innovation Economy Operational Program (contract no. POIG.02.01.00-12-023/08).

REFERENCES (1)

(2) (3)

(4) (5) (6)

(7) (8) (9) (10) (11) (12)

(13) (14) (15) (16) (17)

Soo, J. K.; Mackenzie Ross, A. D.; Kallenberg, D. M.; Milagre, C.; Heung Chong, W.; Chow, J.; Hill, L.; Hoare, S.; Collinson, R. S.; Hossain, M.; Keith, W. N.; Marais, R.; Bennett, D. C. Pigment Cell Melanoma Res. 2011, 24 (3), 490–503. Mangoldc; Santiago-Walker, A.; Li, L.; Haass, N. K.; Herlyn, M. Skin Pharmacol. Physiol. 2009, 22 (2), 114–121. Chapman, P. B.; Hauschild, A.; Robert, C.; Haanen, J. B.; Ascierto, P.; Larkin, J.; Dummer, R.; Garbe, C.; Testori, A.; Maio, M.; Hogg, D.; Lorigan, P.; Lebbe, C.; Jouary, T.; Schadendorf, D.; Ribas, A.; O’Day, S. J.; Sosman, J. a; Kirkwood, J. M.; Eggermont, A. M. M.; Dreno, B.; Nolop, K.; Li, J.; Nelson, B.; Hou, J.; Lee, R. J.; Flaherty, K. T.; McArthur, G. a. N. Engl. J. Med. 2011, 364, 2507–2516. Wu, L.; Qu, X. Chem. Soc. Rev. 2015, 44, 2963–2997. Kawasaki, E. S.; Player, A. Nanomedicine Nanotechnology, Biol. Med. 2005, 1 (2), 101–109. Bevelacqua, V.; Bevelacqua, Y.; Candido, S.; Skarmoutsou, E.; Amoroso, A.; Guarneri, C.; Strazzanti, A.; Gangemi, P.; Mazzarino, M. C.; D ’amico, F.; Mccubrey, J. A.; Libra, M.; Malaponte, G. Oncotarget 2012, 33 (8), 882–892. Lee, G. Y. H.; Lim, C. T. Trends Biotechnol. 2007, 25 (3), 111– 118. Trickey, W. R.; Lee, G. M.; Guilak, F. J. Orthop. Res. 2000, 18 (6), 891–898. An, S. S.; Fabry, B.; Trepat, X.; Wang, N.; Fredberg, J. J. Am. J. Respir. Cell Mol. Biol. 2006, 35 (1), 55–64. Suwanarusk, R.; Cooke, B. M.; Dondorp, A. M.; Silamut, K.; Sattabongkot, J.; White, N. J.; Udomsangpetch, R. J. Infect. Dis. 2004, 189 (2), 190–194. Suresh, S.; Spatz, J.; Mills, J. P. P.; Micoulet, A.; Dao, M.; Lim, C. T. T.; Beil, M.; Seufferlein, T. Acta Biomater. 2005, 1 (1), 15–30. Lekka, M.; Gil, D.; Pogoda, K.; Dulińska-Litewka, J.; Jach, R.; Gostek, J.; Klymenko, O.; Prauzner-Bechcicki, S.; Stachura, Z.; Wiltowska-Zuber, J.; Okoń, K.; Laidler, P. Arch. Biochem. Biophys. 2012, 518 (2), 151–156. Cross, S. E.; Jin, Y.-S.; Tondre, J.; Wong, R.; Rao, J.; Gimzewski, J. K. Nanotechnology 2008, 19 (38), 384003. Lekka, M.; Pogoda, K.; Gostek, J.; Klymenko, O.; PrauznerBechcicki, S.; Wiltowska-Zuber, J.; Jaczewska, J.; Lekki, J.; Stachura, Z. Micron 2012, 43 (12), 1259–1266. Wirtz, D.; Konstantopoulos, K.; Searson, P. C. P. P. C. Nat. Rev. Cancer 2011, 11, 512–522. Hinterdorfer, P.; DufrÃane, Y. F.; Dufrene, Y. F. Nat. Methods 2006, 3 (5), 347–355. Seo, Y.; Seo, Y.; Jhe, W.; Jhe, W. Reports Prog. Phys. 2008, 71 (1), 16101.

(18) (19) (20) (21)

(22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) (43) (44) (45) (46) (47)

Page 8 of 15 Alibert, C.; Goud, B.; Manneville, J.-B. Biol. cell 2017, 1–36. Lekka, M.; Laidler, P.; Gil, D.; Lekki, J.; Stachura, Z.; Hrynkiewicz, A. Z. Eur. Biophys. J. 1999, 28 (4), 312–316. Faria, E. C.; Ma, N.; Gazi, E.; Gardner, P.; Brown, M.; Clarke, N. W.; Snook, R. D. Analyst 2008, 133 (11), 1498–1500. Rebelo, L. M.; de Sousa, J. S.; Mendes Filho, J.; Radmacher, M.; Search, H.; Journals, C.; Contact, A.; Iopscience, M.; Address, I. P.; Rebelo, L. M.; Sousa, J. de. Nanotechnology 2013, 24 (5), 055102. Grady, M. E.; Composto, R. J.; Eckmann, D. M. J. Mech. Behav. Biomed. Mater. 2016, 61, 197–207. Lindberg, U.; Karlsson, R.; Lassing, I.; Schutt, C. E.; Höglund, A. S. Semin. Cancer Biol. 2008, 18 (1), 2–11. Suresh, S. Acta Biomater. 2007, 3 (4), 413–438. Cavallo, F.; De Giovanni, C.; Nanni, P.; Forni, G.; Lollini, P.L. L. Cell 2011, 60 (3), 319–326. Colliver, T. L.; Brummel, C. L.; Pacholski, M. L.; Swanek, F. D.; Ewing, a G.; Winograd, N. Anal. Chem. 1997, 69 (13), 2225–2231. Chandra, S. Appl. Surf. Sci. 2003, 203–204, 679–683. Passarelli, M. K.; Winograd, N. Biochim. Biophys. Acta 2011, 1811 (11), 976–990. Roddy, T. P.; Cannon Jr., D. M.; Meserole, C. a; Winograd, N.; Ewing, a G. Anal Chem 2002, 74 (16), 4011–4019. Kulp, K. S.; Berman, E. S. F.; Knize, M. G.; al, et. Anal. … 2006. Robinson, M. a.; Graham, D. J.; Morrish, F.; Hockenbery, D.; Gamble, L. J. Biointerphases 2016, 11 (2), 02A303. Baker, M. J.; Gazi, E.; Brown, M. D.; Clarke, N. W.; Vickerman, J. C.; Lockyer, N. P. Appl. Surf. Sci. 2008, 255 (4), 1084–1087. Gostek, J.; Awsiuk, K.; Pabijan, J.; Rysz, J.; Budkowski, A.; Lekka, M. Anal. Chem. 2015, 87 (6), 3195–3201. Prabhune, M.; Belge, G.; Dotzauer, A.; Bullerdiek, J.; Radmacher, M. Micron 2012, Vol.43(12) (12), 1267–1272. Lekka, M. M. Bionanoscience 2016, 6 (1), 65–80. Swaminathan Iyer, K.; Gaikwad, R. M.; Woodworth, C. D.; Volkov, D. O.; Sokolov, I. Cell Biochem. Biophys. 63 (2), 109– 116. Lin, D. C.; Dimitriadis, E. K.; Horkay, F. J. Biomech. Eng. 2007, 129 (6), 904. Hayashi, K.; Iwata, M. J. Mech. Behav. Biomed. Mater. 2015, 49, 105–111. Pogoda, K.; Jaczewska, J.; Wiltowska-Zuber, J.; Klymenko, O.; Zuber, K.; Fornal, M.; Lekka, M. Eur. Biophys. J. 2012, 41 (1), 79–87. Muramoto, S.; Brison, J.; Castner, D. G. Anal. Chem. 2012, 84 (1), 365–372. Bobrowska, J.; Pabijan, J.; Wiltowska-Zuber, J.; Jany, B. R.; Krok, F.; Awsiuk, K.; Rysz, J.; Budkowski, A.; Lekka, M. Anal. Biochem. 2016, 511, 52–60. Bobrowska, J.; Moffat, J.; Awsiuk, K.; Pabijan, J.; Rysz, J.; Budkowski, A.; Reading, M.; Lekka, M. Analyst 2016, 141 (22), 6217–6225. Gostek, J.; Prauzner-Bechcicki, S.; Nimmervoll, B.; Mayr, K.; Pabijan, J.; Hinterdorfer, P.; Chtcheglova, L. A.; Lekka, M. Eur. Biophys. J. 2015, 44 (1–2), 49–55. Sarna, M.; Olchawa, M.; Zadlo, A.; Wnuk, D.; Sarna, T. Nanomedicine Nanotechnology, Biol. Med. 2017. Ramos, J. R.; Pabijan, J.; Garcia, R.; Lekka, M. Beilstein J. Nanotechnol. 2014, 5 (1), 447–457. Kulp, K. S.; Berman, E. S. F.; Knize, M. G.; Shattuck, D. L.; Nelson, E. J.; Wu, L.; Montgomery, J. L.; Felton, J. S.; Wu, K. J. Anal. Chem. 2006, 78, 3651–3658. Li, Q. S.; Lee, G. Y. H.; Ong, C. N.; Lim, C. T. Biochem. Biophys. Res. Commun. 2008, 374 (4), 609–613.


Page 9 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



9


Figure 1. a) AFM-based elasticity measurements enable to compare force curves recorded on a stiff nondeformable substrate (a glass or Petri dish) with that acquired on a soft melanoma cell. The difference between these curves is a basis for Young’s modulus determination. b) Results were divided into four groups: melanocytes, VGP melanoma cells (from primary melanoma site), skin and lung metastasis. For each group, moduli distributions were created and fitted with lognormal function. It is the largest for melanocytes, then for VGP cells followed by two melanoma metastases to skin and lung. Differences be-tween all considered groups of data are statistically significant based on Mann-Whitney test at the level of 0.05. c) Young’s modulus is expressed by a mean ± standard deviation determined by fitting a lognormal function to moduli distribution for each cell line separately. 248x217mm (150 x 150 DPI)


Page 10 of 15

Page 11 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. a) Using Bi3+ clusters to bombard the cell surface results in mass spectra with molecular mass range from 0 to 500 Da. Complexity of cell surface biochemical properties translates into a complexity of mass spectra (b), thus, principal component analysis (PCA) has to be employed to differentiate among the studied cell lines. PCA outcome presented as a 3D scores plot (c) shows a clear grouping related to stages of melanoma progression. Each dot represents a single mass spectrum acquired on a single cell. The variability, described by each orthogonal principal component, is as follows: PC1 = 37.64%, PC2 = 20.21%, and PC3 = 11.42% (here, ellipsoids represent double standard deviation range along the PC1, PC2, and PC3 axes). To find better differentiation between various stages of melanoma development, 2D plots of PC components are presented (d). For each data set, confidence ellipses at confidence levels of 90%, 95%, and 99% were calculated. 318x238mm (150 x 150 DPI)



Figure 3. 3D score plots of PCs visualize variability of VGP melanoma cells (WM115 & WM793) in relations to the origin of metastasis, i.e. (a) skin (WM239 & WM266-4) and (b) lung (1205Lu & A375-P) metastasis. Culture media composition did not affect melanoma stage identification as all melanoma cells were cultured in the same medium type and composition. 2D score plots of PC2 plotted versus PC1 show the best separation, statistically verified by confidence ellipses calculated at the confidence levels of 90%, 95%, and 99%. PCA outcome for mass spectra recorded for skin metastasis (c) reveals cell-dependent separation while PCA outcome for lung metastasis (d) presents clearer separation of the studied cell lines. 331x267mm (150 x 150 DPI)


Page 12 of 15

Page 13 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. 2D score plot (PC2 versus PC1) shows that bio-chemical properties of the cell surface can differ in primary and secondary tumor sites i.e. between (a) VGP/skin metastasis (WM115 and WM266-4) and (b) VGP/lung metastasis (WM793 and 1205Lu). The confidence ellipses calculated at three confidence levels of 90%, 95%, and 99% statistically confirm the difference. 321x168mm (150 x 150 DPI)



Figure 5. Summary of surface and biomechanical changes in melanoma cells.

267x212mm (150 x 150 DPI)


Page 14 of 15

Page 15 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Graphical Abstract 262x202mm (150 x 150 DPI)


Biophysical and biochemical characteristics as complementary

Recommend Documents