Optical detection of structural properties of tumor tissues generated by xenografting of drug-sensitive and drug-resistant cancer cells using partial wave spectroscopy (PWS)

The quantitative measurement of structural alterations at the nanoscale level is important for understanding the physical state of biological samples. Studies have shown that the progression of cancer is associated with the rearrangements of building blocks of cells/tissues such as DNA, RNA, lipids, etc. Partial wave spectroscopy is a recently developed mesoscopic physics-based spectroscopic imaging technique which can detect such nanoscale changes in cells/tissues. At present, chemotherapy drug treatment is the only effective form of treatment; however, the development of drug-resistant cancer cells is a major challenge for this treatment. Earlier PWS analyses of prostate cancer cells, a 2D structure, have shown that drug-resistant cancer cells have a higher degree of structural disorder compared to drug-sensitive cancer cells. At the same time, structural properties of the metastasize tumor grown to 3D structure from drug-resistant and drug-sensitive cancer cells within the body is not well studied. In this paper, the structural properties of tissues from grown 3D tumors, generated from docetaxel drug-sensitive and drug-resistant prostate cancer cells xenografted into a mouse model, are studied. The results show that xenografted tumor tissues from drug-resistant cells have higher disorder strength than the tumor generated from drug-sensitive prostate cancer cells. Potential applications of the technique to assess chemotherapy effectiveness in cancer treatment are discussed.


Introduction
Elastic scattering, especially in the visible range of light, is an important method for probing structural morphologies of the biological cells/tissues. It is now shown that probing the structural alteration at nano to submicron scales enables the prediction of several properties of the physical conditions of cells/tissues. In addition to this, the progress of carcinogenesis results in nanoscale structural alteration due to the rearrangement of macro molecular components inside the cells/tissues. This nanoscale structural alteration is considered an important biomarker in the determination of cancer stages, as well as to assess the efficacy of a chemotherapy drug at the different levels of tumorigenicity [1][2][3]. However, the histopathological examinations of cells/tissues, conventionally, are based on a large degree of changes in the cellular architecture during the diseases process [4]. Also, the sensitivity of the existing optical microscopic techniques used to detect such nanoscale alterations are restricted by the diffraction limited resolution (>~200nm).
A recently developed spectroscopic microscopy technique, partial wave spectroscopy (PWS), which combines the interdisciplinary approaches of mesoscopic condensed matter physics and microscopic imaging, is used to quantify the change of the nanoscale structural disorder of weakly disordered biological medium like cells/tissues [3,5]. The statistical quantifications of the reflected intensities due to the nanoscale refractive index fluctuations of the biological cells/tissues are carried out using the PWS analysis. In the PWS technique, the backscattering signals from thin weakly disordered cell/tissue samples are divided into many parallel scattering quasi one-dimensional reflections to calculate the structural disorder strength of the samples [3,5]. Further, the spatial variation of the intracellular components such as DNA, RNA, lipids, and extracellular matrices (ECM) gives rise to spatial mass density fluctuations in terms of the refractive index fluctuations of the cells/tissues [6,7]. This spatial refractive index fluctuations can be quantified in terms of the degree of structural disorder. Therefore, it is necessary to explore early and effective diagnosis/treatment methods for prostate cancer.
At present, chemotherapy is the only way to treat metastasized prostate cancer, however, it is often found ineffective due to an individual patient's chemo-resistance that leads to tumor progression [10][11][12][13]. The PWS studies of cancer cell lines have shown some promising success in distinguishing the hierarchy and drug effectiveness based on the disorder strength Ld parameter, however, these studies were mainly focused on backscattering signals from isolated cancer cells where the cells were grown in 2D on glass slides. In reality, a metastasized cancer cell grows into a tumor with 3D structure when it grown within the body, and these tumor cells may have different structural properties due to its 3D growth into tissue structures. This leads to a demand for the development and characterization of 3D tumor tissues that are generated from the drug-sensitive and drug-resistant cancer cells. This could establish a correlation between the isolated cells grown in 2D and the cells grown in a tissue in 3D based on the structural parameter using the PWS technique. Human cancers have been studied by innumerable murine methods and the determinants responsible for malignant transformation, invasion and metastasis, as well as the examination of response to therapy is investigated by the aid of these murine models. At the same time, the structural disorder properties of cancer growth are not well studied by the xenografting of cancer cells.
Therefore, study of cancer stages and drug effect on cancer tissues that are grown into a 3D tumor from different types of metastasized cancer cells using PWS technique will provide a connection between 2D in vitro cell growth to 3D in vivo tissue growth of the same types of cells.
In this work, using the PWS technique, we explore the structural properties of the 3D tumor tissues obtained by xenografting drug-sensitive and drug-resistant human prostate cancer cells in a mouse model.
We have studied structural properties of tissue obtained by xenografting two types of human prostate cancer (PC) cell lines, namely DU145 and PC-3, whose drug-resistant and drug-sensitive structural properties were studied earlier by PWS technique using Ld parameter [3,5,10]. The earlier PWS results from prostate cancer cell lines showed an increase in the aggressiveness or tumorigenicity for drugresistant cancer cells, relative to its drug-sensitive cancer cells. The cells that grow on a slide are mainly 2D in nature, and have shown above trend. In particular, here we want to verify the structural properties of tissues based on disorder strength or Ld value when they grow in 3D structure by xenografting these cells, and to understand any relationship with their original 2D cells. Xenografting of human cancer cells in a mouse model is one of the most extensively used models to study the development of tumors from cells [14]. Cancerous human cells were subcutaneously injected in immunocompromised mice. Based on the number of cells injected, the tumors will develop over 1-8 weeks and reaction to the proper therapeutic regimes can be studied in vivo [14,15] or ex vivo. The disorder strength Ld of 4µm thickness excised tumor tissue were calculated, and the correlations of structural disorder values of these xenografted tissues corresponding to their original cells are compared. This study may provide the structural disorder difference of 2D cell lines and their corresponding growth of 3D structures and any correlations.
Therefore, this new PWS analysis of human tumor cell xenografting can be used to study the physical state and effectiveness of a chemotherapy drug in cancerous cells/tissues to improve chemotherapy treatment methods.

PWS Experimental Setup:
We perform the structural disorder measurement using a recently developed partial wave spectroscopy (PWS) experimental technique, with added further engineering of finer focusing. The partial wave spectroscopy (PWS) setup with a fine focus to measure the structural alterations at nanoscale level is as shown in Fig.1. Xenon Lamp (Newport, 150W), a source of stable broadband white light is used to illuminate tissue samples of micron thickness using Kohler Illumination. The white light is reflected towards the combination of lenses with silver coated mirror (Thorlabs, f=50.8mm). The combination of converging lenses (Thorlabs, f=50.8mm) along with the apertures (Newport) form a 4f system that helps to minimize the diffraction effect and preserve the high-frequency effect and hence enrich the sharpness in an image. This collimated light from the lens is reflected by a right-angle prism (Thorlabs, 25.4mm) and passed through the dichroic mirror (Thorlabs, 25.4mm) and then enters an objective lens (Newport, NA=0.65,40X). The low numerical aperture objective lens focuses the light in the sample within its working distance with the help of high-resolution 3D electronic motorized stage (Zebar Tech, 100nm in Z axis and 40nm in X-Y). This high-resolution motorized 3D stage is considered revolutionary to the microscopic setup for its extreme accuracy and finer focus. A finer focus is essential for correctly defining the effective scattering volume/length of a sample. The backscattered signal from the sample is passed through the objective which gets reflected into the thick collecting lens (Thorlabs, Φ=50.8mm) before the aperture. Finally, the collected backscattered signal passes through a liquid crystal tunable filter i.e. LCTF (Thorlabs, KURIOS-WB1) with spectral resolution of 1nm within the visible range (420-730nm) of the spectrum. A CCD camera (Retiga 3, 1460×1920) coupled with a LCTF controller records the filtered signal in the desired wavelength range. Here, the CCD camera and LCTF filter are coupled with the LCTF controller in such a way that for each 1nm increment in wavelength by the LCTF filter (resolution 1nm), the backscattered signals are recorded in the CCD within the visible range.

Calculation of the structural disorder or disorder strength Ld:
The backscattered images are recorded in the CCD camera at every wavelength () at the spatial pixel position (x,y) in the wavelength range 450-700nm, and the reflected data cube, R(x, y; ), is acquired by PWS system. Here, the data cube R(x,y;) includes the fluctuating part of the reflection coefficient over the visible wavelength regime due to the presence of a disordered medium including the high frequency noise. In a quasi-1D approximation, the collected backscattered data at each (x,y) from R(x,y;) is fitted with a polynomial of the 5 th order. The fitted polynomial is then extracted from the signal to remove the systematic errors. In the next step, the R(k) signal for each pixel position (x,y) is obtained after applying a fifth-order low-pass Butterworth filter with a suitable normalized cutoff frequency to remove the high frequency noise components from the reflected signal of micron size samples. From the extracted Where B is the normalization constant, k is the wave number ( = 2 / ).
For the Gaussian color noise of the refractive index at position r and r', <dn(r)dn'(r')>=dn 2 exp(-|r-r'|/lc), it can be shown that Ld=<dn 2 >lc [3,5], however, this form may change due to different situations.
The disorder strength quantifies the variability of local density of intracellular material within the samples, and hence the average and standard deviation of the Ld are calculated to characterize the system. excised from euthanized mice. The excised tumors were further paraffin embedded and sectioned using microtome of 4 μm thickness and placed on glass slides. Further, these slides were processed for antigen retrieval process as described previously [16]. The resultant tumor sections were subject for imaging studies.

Result
PWS detects the nanoscale structural alteration in the cells/tissues and can distinguish the different levels and effect of drug in the tumorous cells/tissues [7,10]. Among the different types of cancer, prostate cancer is a major concern of public health at present because of its low survival rate. The average Ld value of tumors obtained from drug-resistant cancer cells is 9% higher than the tumors obtained from drug-sensitive cancer cells and corresponding standard deviation std(Ld) is 8% higher. This result is in strong agreement with the disorder strength calculated for drug-resistant and drug-sensitive cell lines earlier [10]. The disorder strength of the prostate cancerous cell line calculated using PWS have shown that, the average and standard deviation of Ld values are higher in drug-resistant cells compared to the drug-sensitive for DU145 type cells. That means the xenografted tissue structure also have similar trends to original cell structures. confirms that the degree of disorder strength Ld can be used as a marker to detect the cancer stages or drug effects using 3D cancer tissues structure, similar to that of a cell line that is easy to study.

Conclusions
The result: In this work, we have reported the PWS study of the nanoscale structural changes in tissues generated from the xenografting of drug-sensitive and drug-resistant PC cells using a mouse model. This result in a higher structural alteration in drug-resistant PC cells than drug-sensitive ones. The different mechanisms such as: DNA damage repair, drug inactivation, alteration of drug targets, cancer cells/tissues heterogeneity, cells/tissue death inhibition, epithelial-mesenchymal transition and metastasis, etc. [17][18][19] are making the cells/tissues drug resistant [10]. Xenografted tumor tissues that are obtained from drugresistant PC cell lines, which survive due to one of the mechanisms mentioned above, have the same types of structural properties when they are grown into a 3D structure. This supports that the same hierarchy of structural disorder survives in the grown 3D structure. could be a reliable, easy, and quantitative approach to diagnose chemo resistance. This result seeks the potential application to monitor the effect of chemotherapy drugs on cancerous tissues and to study the different level of tumorigenicity which can be obtained both, in-vitro and in-vivo method. In summary, this method will help in understanding the drug-resistant and drug-sensitive cells that are grown within the body, by examining the cells only.