Physical Labeling of Papillomavirus-Infected, Immortal, and Cancerous Cervical Epithelial Cells Reveal Surface Changes at Immortal Stage

Abstract

A significant change of surface features of malignant cervical epithelial cells compared to normal cells has been previously reported. Here, we are studying the question at which progressive stage leading to cervical cancer the surface alteration happens. A non-traditional method to identify malignant cervical epithelial cells in vitro, which is based on physical (in contrast to specific biochemical) labelling of cells with fluorescent silica micron-size beads, is used here to examine cells at progressive stages leading to cervical cancer which include normal epithelial cells, cells infected with human papillomavirus type-16 (HPV-16), cells immortalized by HPV-16, and carcinoma cells. The study shows a statistically significant (at p <0.01) difference between both immortal and cancer cells and a group consisting of normal and infected. There is no significant difference between normal and infected cells. Immortal cells demonstrate the signal which is closer to cancer cells than to either normal or infected cells. This implies that the cell surface, surface cellular brush changes substantially when cells become immortal. Physical labeling of the cell surface represents a substantial departure from the traditional biochemical labeling methods. The results presented show the potential significance of physical properties of the cell surface for development of clinical methods for early detection of cervical cancer, even at the stage of immortalized, pre-malignant cells.

Introduction

Cervical cancer is the second most common cause of cancer death in women worldwide; infection with oncogenic human papillomavirus (HPV) is the most significant risk factor in its etiology [1]. The mortality rate is second only to that for breast cancer. Early detection of this cancer can substantially decrease mortality due to this disease. The demand to reduce the mortality related to cervical cancer has resulted in inter-disciplinary research involving physics, chemistry, molecular biology, engineering, and medicine [2]. Research efforts over the years have been directed toward the development of a universal, reliable, and consistent test to detect malignant cells at the early stages of cancer development, and in particular, premalignant cervical cells, also known as cervical intraepithelial neoplasia (CIN) [3].

HPV-associated cervical carcinogenesis is a multistep process leading to the transformation of the infected cells to CIN [3]. The cells associated with the lesions become atypical and mitotically active, and they are characterized by the inability to terminally differentiate [4]. Low grade CIN are typically limited to the lower portion of the epithelia, but they can progress to high-grade CIN and become evident throughout the entire thickness of the epithelia and eventually lead to cervical carcinoma. The multistep process leading to transformation of infected cells to CIN is associated with over expression of two early genes, E6 and E7 [5, 6]. These genes are consistently expressed in cervical cancer cell lines and involved in HPV-mediated carcinogenesis [7]. The transforming ability of the genes is partly due to their ability to interact with cellular tumor suppressor proteins. E6 binds the tumor suppressor p53 through its interaction with another cellular protein, E6-associated protein (E6-AP), leads to degradation of p53 via the ubiquitin-mediated protein degradation pathway [8, 9]. E7 binds and degrades the retinoblastoma tumor susceptibility gene product (pRb) and also the other pRb “pocket” protein family members, p107 and p130 [10, 11].

The Papanicolaou (Pap) smear, a cytological screening test has been one of the most successful methods of cervical cancer detection over the years [12]. Although the Pap smear is the most widely used cancer screening method in the world and its impact on the incidence of cervical cancer is well known from a historical perspective, recent reports suggest that the sensitivity of Pap smear screening is 50–80%, with the relative proportion of sampling to screening errors being 2:1 [13]. There are multiple benign mimics of neoplastic cells such as atypia of repair, atrophy, radiation change, effect of intrauterine device, and metaplasia. These cells are called atypical cells of undetermined significance due to the inability of optical microscopy to identify definite features of either benign or neoplastic cells. The tests may be further complicated by high-unsatisfactory rates, preparation artifacts, and unnecessary cost interventions. Each year in the United States alone approximately 3.5 million Pap smears cytological tests are classified as equivocal, out of which only 8% of women will have preinvasive (high-grade squamous intraepithelial) lesions, and 0.4% will have carcinoma as found in further testing that involves invasive tissue biopsies. DNA tests (detection of HPV, human papillomavirus infection) are rather accurate, but identify only a risk of cancer (the fact of HPV infection) but not cancer [44]. The economic constraints in developing countries have prompted alternative methods of screening for cancer including visual inspection after application of 3–5% of acetic acid [14] and Lugol’s iodine [15]. The major disadvantage of these tests is low specificity. Given the considerable variation in the way these tests are applied and interpreted in different settings, there is no standard universally accepted definition of the test results. It remains to be seen if the specificity can be improved by further developments in test definitions and training strategies. The inability of these test methods to unambiguously detect abnormal cells during the early stages, poses a major challenge in an effort to reduce mortality associated with cervical cancer.

Demands for the development of a universal testing and screening method have generated enormous scientific and industrial interest in the past. Several methods such as organic receptor molecules that undergo color changes [16], electrochemical, [17] magnetic [18], and fluorescent tags [19, 20] have been developed and reported. Optical techniques like fluorescent microscopy [21] confocal imaging [22] and optical coherence tomography (OCT) [23] have shown promise for cervical precancer detection as measured against the traditional methods such a Pap smear. The sensitivity and specificities of these optical techniques have been reported to reach 80–90% in small and moderate clinical trials [24].

Among the optical techniques, detection using fluorescence is an important non-isotopic method. The fluorescent biological labels have played an important role in the field of biomedicine, such as the detection of materials inside or outside of the cell [25, 26], hybridization and sequencing of nucleic acids [27], and clinical diagnosis at the early stage [28]. These conventional fluorescent techniques have the following limitations: reduction in the emission intensity due to photobleaching, the toxicity of some of the fluorescent dyes may be harmful for living cells, and detection sensitivity is not so high because only a few fluorophores can be coupled to one biomolecule in the conventional fluorescent label methods.

It was recently found [29] that a several micron silica particle attached to a cantilever of an atomic force microscope [30] (AFM) interacted with cancerous and normal cells quite differently. The main reason was attributed to a different topological structure of surface “brushes” covering cells, which consisted of microvilli, microridges, cilia, and glycocalyx molecules. Using that observation, a new method to identify cervical cancer cells was suggested [31, 32]. The method was based on the use of silica particles of similar size but fluorescent in nature. The difference in the cell brush resulted in the different adhesion of the silica beads-cell. Furthermore, a high-resolution multimodal AFM imaging in novel HarmoniX mode revealed virtually 100% difference between the surface of normal and cancerous cells [43]. This AFM method was very precise but rather slow. In addition, the detection of pre-malignant (immortal) cells is more interesting for potential clinical early diagnosis of cancer. Thus, it is important to test if the faster labeling method Ref. [31] could potentially be effective in the defection of immortal cells. From the fundamental point of view, it is informative to learn the stage of cancer progression when the cell surface starts changing.

In this study, we use the fast method of Ref. [31] to study cervical cells during the different stages of multistep transformation leading from normal to cancer. Specifically, we study normal, infected (with HPV-16), immortal, and cancerous cells. Here we confirm previous results for normal and cancerous cervical cells [31], and find that in the progression leading to cervical cancer, adhesion (or cell surface) changes significantly at the stage of immortal cells. We found that cancerous and immortal cells (the latter were derived from HPV-infected cells) adhere to silica beads in a statistically different manner than normal cells and HPV-infected cells.

Materials and Methods

Spectrofluorometric Measurements and Imaging of Cells

A scheme of the experimental setup is shown in Fig. 1. An Olympus BH2-UMA microscope, which works in both transmission and reflection mode, was connected to a JVC TK-1280U color video camera. The images of cells were captured using FlashBus MV version 3.91 software. To record the fluorescent emission, an optic fiber was used to connect the microscope to an spectrometer (USB 2000 by Ocean Optics Inc., FL). A Cyonics air-cooled argon ion laser was used as an excitation source for fluorescence. A narrow-band 488 nm blocking filter (Omega Optical) was utilized to filter the laser light. The spectra were recorded using OOIBase32 software. To avoid excessive focusing on individual florescent particles, and consequent collecting the emission spectra from excessively those particles, after the focusing, the focus knob was rotated three full turns in counter clockwise direction (defocusing the sample by increasing the sample-objective lens distance). This procedure was maintained constant for spectrofluorometric measurements of all samples. 5× objective was used in these measurements. This allowed collecting the fluorescent signal coming from a predefined area of approximately 4 mm in diameter (as was found by moving individual fluorescent beads).

Fig. 1.

(Color online) Schematic diagram of the optical measurement system

To decrease possible dependency of the fluorescent signal on the device used, the spectrometer was calibrated with the help of LS-1 Tungsten Halogen Light Source (USB 2000 by Ocean Optics Inc., FL) by using the procedure described by the manufacturer. To decrease the noise, each fluorescent signal used in this study was an integral of the fluorescent intensities (taken in the range of 560–680 nm).

Statistical analysis was done with the help of Origin 8 software (Origin Labs). The significance of statistical difference was analysed with ANOVA test using the confidence level of p of 0.01. Specifically, one-way ANOVA using the Tukey test of mean comparisons was used. Equal variances of distributions were verified using Levene’s tests.

Cell Cultures

Primary cultures of human cervical epithelial cells were prepared by a two-stage enzymatic digestion of cervical tissue as described [33] In brief, each tissue was digested for 16 h at 4°C in dispase and the layer of epithelial cells was removed from the underlying connective tissue by scraping. The sheet of epithelial cells was cut into 1 mm² pieces and digested in 0.25% trypsin at 37°C for 10 min. Trypsin was neutralized by adding fetal bovine serum and cells were collected by low-speed centrifugation. Cultures consisting of ≥95% epithelial cells were maintained in keratinocyte serum-free medium (Invitrogen) which prevents outgrowth of fibroblasts and other stromal cells. HPV-16 immortalized cell lines [41] and cervical carcinoma cell lines [42] were also maintained in KSFM and no evidence of contamination by fibroblasts or other stromal cells was observed.

All human tissue was obtained from the Cooperative Human Tissue Network. Informed consent was obtained from patients according to their published guidelines (http://chtn.nci.nih.gov/phspolicies.html). The transformation of normal cells in precancerous squamous intraepithelial lesions is associated with over expression of the E6 and E7 genes. HPV genes were introduced into cultured cervical cells by infection with recombinant retroviruses encoding HPV-16 E6/E7 genes inserted into the vector pLXSN, which contains the neomycin resistance gene [34]. Infection (MOI = 10) was performed for 3 h in medium with 10 ng/ml polybrene with rocking every 15 min. Subsequently, medium was changed and cells grew for 24 h before cultures were split 1:3. After 24 h, infected cells were selected by growth for 2 days in KSFM containing 200 μg/ml G418 and used immediately. Approximately 70–90% of cells were infected as determined by survival after G418 selection.

Normal cervical cells were used between 40 and 60 population doublings (PDs), and carcinoma cell lines were used at 90–120. The slightly higher number of PDs for cancer cell lines avoids potential confusion because any normal cells (epithelial cells or stromal cells) that may contaminate the cancer culture dishes would die out by that number of PDs. This allows us to avoid possible confusion between cancer cells and either normal epithelial cells or fibroblasts. All cells were plated in 60 mm tissue culture dishes and dishes were used for experiments when cells were 80–100% confluent.

Fluorescent Silica Beads

Recently a new one step self-assembly of nanoporous silica particles with encapsulated organic dyes has been developed, [35–37] in which the dyes are physically entrapped inside silica matrix, inside 2–4 nm in diameter nanochannels. It was found that the synthesized particles could be up to two orders of magnitude brighter than the micron-size particles assembled from aqueous dispersible quantum dots [38] encapsulated in polymeric particles (scaled to the same size). Comparing with the maximum fluorescence of free dye in the same volume, the particles can show fluorescence which is higher by a factor of ~5,000. This makes the particles the brightest tags presently available.

The synthesis of these particles is described in the corresponding references. In brief, it is a one step synthesis. Tetraethylorthosilicate (TEOS, 99.99+%, Aldrich), cetyltrimethylammonium chloride (CTACl, 25 wt% aqueous solution, Pflatz & Bauer), formamide (99%, Aldrich) and hydrochloric acid HCl (37.6 wt% aqueous solution, Safe-Cote), Rhodamine 640 (R40) dye (Sigma-Aldrige, Inc.) were used. All chemicals were used as received. The surfactant, acid, dye, formamide, and distilled water (Corning, AG-1b, 1 MΩ-cm) were stirred in a polypropylene bottle at room temperature for 2 h, after which TEOS was added and the solution stirred for ca. 5 min. The solution was then kept under quiescent conditions for 3 days. The molar ratio of H₂O:HCl:formamide:CTACl:R6G:TEOS was 100:7.8:9.5:0.11:0.01:0.13. The materials that formed were washed by centrifugation to avoid damaging of the surface, washed with copious amounts of water (till the leakage of the dye stops), and mixed with HBSS solution for the further use on the cell culture.

Detection of Affinity of Fluorescent Silica Beads to Cells

After carefully washing the particles to remove any traces of the synthesizing chemistry, the particles were added to Hanks balanced salt solution (HBSS) at a pH of 7.4 to form a colloidal dispersion (~2 g/l). The cells in the 60 mm culture dishes were washed twice with HBSS solution for 2 min each. The cells were then exposed to 1 ml of the colloidal dispersion for 2 min. During this 2 min period, the dishes were subjected to initial 30 s of shaking on a Boekel Scientific Ocelot 260300F at minimum speed, to insure uniform distribution of the colloidal solution and to minimize particle agglomeration. To remove excess particles that had not adhered to the cells, the cells were then washed with HBSS solution two times thoroughly using the same Boekel shaker. The culture dishes were then dried under ambient conditions at room temperature.

It is important to note the density of cells in the culture dish. The measurements were done on cells that just reached confluency in the cell culture dish bottom. This allows us to exclude the possible affinity of the particles to the culture dish bottom as well as dealing with different total areas of coverage of cells in different dishes. In this case, the processing of the results becomes simple because the amount of adherent fluorescent particles is linearly proportional to the fluorescent signal. The analysis of cells that are far from confluency is possible [30] although this requires more processing time or development of a special software.

Results

Figure 2 shows representative optical images (combined fluorescence and transmission microscopy) of particles that adhered to the surface of normal, HPV-16 infected, immortal, and cancerous cells. One can see higher amounts of particles adherent to cancer and immortal cell surfaces as compared to normal and infected cells. To describe this difference quantitatively, we measured fluorescence for each type of cell. On average, the fluorescent signal (measured as described in the Materials and Methods section) is proportional to the number of fluorescent silica beads adherent to the cell surfaces. Therefore, the fluorescent signal is a good measure of the difference in adhesion of fluorescent silica beads to the cells.

Fig. 2.

(Color online) 200×110 μm² optical images of normal, infected, immortal, and cancer cells with fluorescent silica particles attached. The cells are seen as an almost continuous (green) background, and the particles are the bright spots (red)

Figure 3 shows the fluorescent signals of silica beads measured on culture dishes containing either normal, or HPV-16 infected, or immortal, or cancerous cells. Individual fluorescent spectra and their integral used to obtain the data presented in Fig. 3 are shown in the Supplementary Materials (Figure S1 and Table S1). Each cell type was derived from tissues of three human subjects, and was tested using three culture dishes, 10–20 readings per dish (5–6 fluorescent spectra averaged over three measurements each). Figure 3a shows the total statistical data of fluorescence for all human subjects. The average and one standard deviation are shown. Figure 3b presents the statistical fluorescence data for each human subject separately. The average, one standard deviation and minimum/ maximum are shown (details of the data are provided in the Supplementary Materials).

Fig. 3.

The fluorescent signals collected from silica beads adhered to cells taken from three healthy individuals, three cell stains infected with HPV-16 virus (derived from cells of three healthy individuals), three immortal cell lines (derived from three cell stains infected with HPV-16 virus), and malignant cells from three cancer patients. a The averaged fluorescent signal for all human subjects and one standard deviation are shown in the graph. b The average (middle line in the box), one standard deviation (the box) and minimum/maximum (error-bar markers) are shown for each human subject separately

One can see an insignificant difference in the fluorescence signals from the silica beads adherent on normal and infected cells. Statistical analysis shows that the difference in the values of these intensities is statistically insignificant at the confidence level of 0.01.

Comparing the fluorescence data of the immortal cells with that of the normal and infected cells, one can see a difference. It is evident that the fluorescent intensity is stronger in the case of immortal cells, and getting closer to that of cancerous cells. Statistical analysis shows that the fluorescent signal collected from immortal cells is significantly different from the signals collected on either normal or infected cells at the confidence level p <0.01.

The same statistically significant difference at the confidence level p <0.01 was found for the cancerous cells compared to the normal, infected cells, and even immortal cells. However, the fluorescent signals coming from the silica beads that adhered to cancer cells were noticeably less different from that of the immortal cells when compared to that of either normal or infected cells (see, the values of probability and q-value in Table S4; further details of the statistical analysis are shown in the Supplementary Materials (Tables S2–S4).

To summarize, our results show a statistically significant (at p <0.01) difference between all types of cells except normal-infected consisting of normal and infected and a group of immortal and cancer cells. This implies that the cell surface changes substantially when cells become immortal in its progression to cancer.

Discussion

Here we study the change of the surface of human cervical epithelial cells during this progression to cancer, from normal, to HPV-infected, to immortal, and then to the carcinoma stage. Such a change has previously been reported when comparing normal and cancerous epithelial cervical cells [29, 31, 32], and was attributed to the change of microvilli content of the cellular brush. Physical adsorption of fluorescent silica beads is used to monitor the change of cell surface, the method described in Ref. [31]. Visual inspection of the optical images (Fig. 2) shows that the number of particles adsorbed on the cancer and immortal cell surfaces is typically more than that on the normal and infected cell surface. The quantitative measurements of fluorescent signal coming from the fluorescent silica beads adherent to the cell surface, Fig. 3a), confirms this observation statistically with the confidence level of 0.01.

Figure 3b shows a noticeable variability of the fluorescent signals collected from different human subjects. While this variability does not change the conclusion of the alteration of the cell surface starting from the stage of immortalization, this implies the need for studying more human subjects if this method is going to be used for diagnostic purposes.

To determine whether the presence of E6 and E7 genes alters the cell surface (and consequently, the number of silica particles adsorbed to the cell surface), we compared the fluorescence signal coming from the silica beads adhered to normal and infected cells. Because the fluorescence signals from normal and infected cells were statistically similar (the difference was statistically insignificant at the confidence level p <0.01), one could conclude that the infected cells have surface characteristics similar to normal cells. From this we can conclude that just the presence of E6 and E7 genes is not sufficient to change the cell surface.

It was reported that the two genes E6 and E7 have to act synergistically to induce cell immortalization [4]. It was speculated that E7 can prolong the lifespan of cells and therefore, supposedly induce overgrowth of cells. However, E7 can also induce apoptosis [39]. E6 inhibits the apoptosis induced by E7, accentuating the effects of E7 on immortalization [40]. Synergistic action of the oncogenes is a progressive time-dependent process. From this we can conclude that the synergistic action of both E6 and E7 genes alone is not sufficient to change the cell surface. One can see that the fluorescent signal is stronger in the case of immortal cells, and getting closer to cancerous cells. It is also in agreement with in vivo results of Ref. [41] in which it was shown that immortal cells of similar passages used here behaved more as cancerous cells. In addition, the observed difference between normal and cancer cells is in agreement with the data reported previously [29].

To amplify, we showed that the cell surface changes substantially at the stage of immortalization. This implies direct correlation between the increased adhesion and expression of both E6 and E7 genes. In the light of results of Ref. [29], it presumably means that the cellular brush (microvilli, microridges, glycocalyx, etc.) alters substantially when cells become immortal. Second, our results showed that the physical labeling of cervical epithelial cells with fluorescent silica beads could be suitable for discrimination of both immortal and cancerous cervical cells in a culture dish. This demonstrates the significance of physical properties of the cell surface for the development of clinical methods for early detection of cervical cancer, even at the stage of immortalized premalignant cells. Such detection would represent a substantial departure from the traditional biochemically specific labeling methods.