PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jbtJBT IndexAssociation Homepage
 
J Biomol Tech. 2007 December; 18(5): 287–297.
PMCID: PMC2392993

Enrichment and Detection of Molecules Secreted by Tumor Cells Using Magnetic Reversed-Phase Particles and LC-MALDI-TOF-MS

Abstract

Tumor cells change their genetic expression pattern as they progress to states of increasing malignancy. Investigations at the DNA and RNA level alone cannot provide all the information resulting after the translation and processing of the corresponding proteins, which is one reason for a poor correlation between mRNA and the respective protein abundance. In diagnostics, differentially expressed peptides or proteins are important markers for the early detection of cancer. Unfortunately, tumor cells secrete peptides and proteins in only very low amounts, making mass spectrometric determination very difficult. In this publication, methods have been developed for the effective enrichment and cleanup of substances secreted by cultivated cancer cells. To obviate peptides from fetal calf serum used in cell culture, a serum surrogate was developed, which maintained growth of the cancer cells. After the binding of substances from cell-culture supernatants to custom-made magnetic reversed-phase particles, the substances were eluted and separated by capillary high-performance liquid chromatography. Fractions were spotted directly on a MALDI target, and MALDI-TOF mass spectrometric data acquisition was performed in automatic mode. This technology was used to detect substances secreted by two mammary carcinoma cell lines differing in their malignancy (MCF-7, MDA-MB231). Unequivocal differences in the peptide secretion patterns were observed. In conclusion, this system allows the sensitive investigation of peptides secreted by cancer cells in culture and provides a valuable tool for the investigation of cancer cells in different states of malignancy.

Keywords: cancer cells, serum surrogate, magnetic RP particles, LC-MALDI

Cells communicate through a network of different substances, including cytokines, interleukins, and hormones. The secretion of such substances reflects the functional state of the cells and is programmed by gene expression. In the case of tumor cells, the progression to states of increasing malignancy is accompanied by changes in gene expression. It has been shown by Affymetrix GeneChip technology that the transcriptional profiles of different mammary carcinoma cell lines revealed 86 genes up-regulated and 321 genes down-regulated in invasive versus non-invasive cells.1 A change in transcription will also alter the pattern of newly synthesized proteins and ultimately also protein secretion. However, there is no stringent correlation between the transcriptome and the proteome.25 A single gene can encode several mRNAs by differential splicing, leading to different versions of a protein, and a newly synthesized protein can be post-translationally modified in the Golgi apparatus, which could also alter its secretory behavior.

Peptides or proteins that are secreted by cancer cells in different states of progression could be possible candidates in the search for new biomarkers. In the past, differences in protein patterns have been analyzed in serum or plasma samples from cancer patients and compared with those of healthy volunteers. The most commonly used technology for this purpose is the separation of the different sera by 2D-gel electrophoresis with a subsequent analysis of differentially expressed proteins by mass spectrometric methods.6,7 In spite of numerous efforts during the last 10 years, the only presently existing marker specific for a tumor type is the prostate-specific antigen (PSA),8,9 an indicator of prostate cancer disease. Other cancer markers, such as CA 15-3, CEA, or CA 19-9, have been found not to be specific for a certain cancer type.

Due to the large variance in the concentrations of different proteins, the analysis of serum or plasma is still an arduous task. One constraining factor is serum albumin, which is present at a concentration 107-fold higher than other peptides or proteins, such as signaling molecules or hormones. The six most abundant proteins (albumin, IgG, IgA, haptoglobin, α-1-antitrypsin, and transferrin) represent 85 to 90% of the total protein amount in serum.10 Since these proteins hamper a successful analysis of low concentrated and potentially important proteins, their removal is required before protein analysis. One option for the removal of the most common proteins in serum is the use of affinity columns (depletion columns) loaded with multiple polyclonal antibodies.11 However, a removal of albumin, which serves as a transporter for certain substances, e.g., peptides or steroids, may also deplete these potentially important molecules from the probe to be analyzed.12,13

A different approach in the search for new biomarkers is the direct investigation of cancer cells. Cancer cells are the origin of biomarkers detectable in the blood stream. For example, in normal prostate tissue, PSA is secreted by the luminal epithelial cells mainly toward the glandular lumen. However, after transformation of the normal prostate epithelial cells to the adenocarcinoma state, PSA is secreted toward the basal compartment, thereby entering the blood vessels.14

Carcinoma cells secrete peptides or proteins in only very low amounts, and such low concentrations impede the investigation by mass spectrometric methods. Therefore, a large number of cells (108 to 109) is required to obtain sufficient material for successful mass spectrometric investigations.15 The cultivation of cancer cells, however, usually requires medium containing fetal calf serum (FCS) as a source of growth factors, and this leads to similar complications in protein analysis as when working with plasma or serum.

To avoid the presence of abundant proteins when analyzing cultured cancer cells by mass spectrometry, many investigators use FCS-free cell-culture media.1618 This approach has a major drawback: most of the cells need the supply of growth factors provided by the serum for substrate adhesion and growth. Without these growth factors, many cells undergo cell death.1921 Thus, under these conditions the protein pattern obtained will not represent the functional state of the cells.

Another difficulty is the efficient isolation and purification procedure for low-concentration substances with sufficient yield. The use of desalting procedures like dialysis with cut-off membranes (e.g., Slyde-A-Lyzer, Pierce) followed by concentration steps using solvent evaporation, or the use of ultrafiltration devices with molecular-weight cut-off membranes normally works best for high protein concentrations. Using these procedures, low-concentration proteins will tend to bind to the wall of the receptacles, resulting in a dramatic loss of analyte. One approach to avoid such a loss is the direct binding of secreted substances to protein chips. For this purpose, cell-culture supernatants have been applied directly to protein chips with different functionalized surfaces—strong anion exchangers (SAX), weak cation exchanges (WAX), or reversed phases. After the binding process, the surface-bound substances were ionized using the surface-enhanced laser desorption ionization (SELDI) process22 followed by the mass spectrometric detection of the substances.16,17 The limit of sensitivity using this system is about 300 nM, which makes the SELDI-TOF-MS technology applicable to profiling serum or plasma peptides,2325 but is not sufficient for the direct analysis of cell-culture supernatants. The sensitivity in the detection of peptides and proteins can be further increased by using microparticles, which have a much greater binding surface than planar protein chips. It has been reported that using Poros R2 beads with reversed-phase surfaces (20 μm) for the enrichment of proteins prior to mass spectrometric analysis, it was possible to identify proteins at concentrations of ≤100 nM.26 Unfortunately, the use of non-magnetic particles is not suitable for automatic procedures. Therefore, super-paramagnetic particles with reversed-phase surfaces are being used for the mass spectrometric profiling of serum samples.27,28

In this paper, we describe an improved method for the mass spectrometric analysis of low-concentrated substances. This method was applied to compare the secretion patterns of two different breast cancer cell lines (MCF-7 and MDA-MB231). These cells were chosen as a model for representing tumor cells with different degrees of malignancy. The estrogen-sensitive MCF-7 with epithelial-like morphology possesses receptors for estrogen, progesterone, and HerB2; is apoptosis competent; and shows little invasive potential. In contrast, the estrogen-insensitive MDA-MB231, with a more unspecific morphology, does not have estrogen or progesterone receptors, is apoptosis resistant, and highly invasive in animal tumor models. These cell lines are thus well characterized with respect to the effects of hormonal stimulation and chemosensitivity, and have been used in previous investigations from our laboratory.29,30 To avoid the drawbacks of serum-free culture as well as to reduce the protein contribution of FCS, a custom-made serum surrogate was developed that is free of low-molecular-weight peptides yet maintains the growth of the investigated cancer cells. The secreted substances of cells growing in medium with the serum surrogate were isolated and purified with custom-made magnetic reversed-phase particles. These substances were then separated and detected using LC-MALDI technology. Reproducible ion signal patterns were obtained with the two different cancer cell lines, demonstrating the applicability and sensitivity of the improved mass spectrometric technology.

MATERIAL AND METHODS

Chemicals and Devices

Unless specified otherwise, chemicals and peptides were purchased from Sigma-Aldrich Co. (Munich, Germany).

For cell cultivation, Dulbeccos Modified Eagle Medium + 4500 mg/L glucose (DMEM) and Fetal Calf Serum (FCS) were purchased from Invitrogen (Karlsruhe, Germany). Panexin H and Panserin PX10 were obtained from PAN-Biotech GmbH (Aidenbach, Germany). Sterile culture plastic ware was obtained from TTP AG (Trasadingen, Switzerland). The CASY cell-counter system and the Casyton buffer were from Schaerfe GmbH (Reutlingen, Germany).

For the analytics, vials and pipette tips were purchased from Eppendorf (Hamburg, Germany) and glass vials from Agilent Technologies (Boeblingen, Germany). Water was purchased from Fisher Scientific (Loughborough, UK) and acetonitrile from Baker (Deventer, Holland). Custom-made magnetic reversed-phase particles were obtained from Solaton GmbH (Stuttgart, Germany). The bovine serum surrogate was produced in our laboratory and is currently under patent application. Microcon Centrifugal Devices (10 and 30 kD) were obtained from Microcon (Schwalbach, Germany). Monolithic reversed-phase columns (PS-DVB, 200 μm) were purchased from Dionex (Idstein, Germany). The magnetic separator was from Invitrogen (Karlsruhe, Germany). Prespotted AnchorChips (384 spots per well, matrix α-cyano cinnamic acid) and commercially available magnetic reversed-phase particles (CLINPROT MB-HIC 8) were from Bruker Daltonics GmbH (Bremen, Germany).

The peptide mixture used for the evaluation of the mass spectrometric experiments was comprised of the peptides angiotensin II, angiotensin I, substance P, bombesin, renin substrate, and ACTH-CLIP 18-39.

Cell Cultivation

The human breast cancer cell lines MCF-7 and MDA-MB231 have been cultivated for many years in our laboratory without detectable changes in their morphological and biochemical characteristics. Cells were transferred weekly to new flasks with a titer of 2 × 104 cells/mL DMEM + 5% FCS. To test different growth conditions, cells were seeded at 105/2 mL DMEM with 5% FCS into six-well plates, and the culture medium was exchanged the next day by DMEM alone or DMEM supplemented with either 5% FCS, 5% bovine serum surrogate (produced in our laboratory; currently under patent application), or 2% Panserin PX10 (a serum additive). Alternatively, a medium surrogate Panexin was added alone.

For mass spectrometry experiments, cells (105/2 mL) seeded into six-well plates in DMEM with 5% bovine serum surrogate were incubated for 4 d until subconfluent. The respective cell-culture media incubated without cells served as control for mass spectrometric analysis. Following the change to fresh culture medium, with or without serum proteins added, the cell-culture supernatants and control media were removed at the times 0, 24, and 48 h, transferred to glass vials, and frozen immediately at –20°C (see below). The adherent cells remaining in the wells were then counted after preparing a nuclei suspension. The cells were first treated for 20 min with a hypotonic buffer (20 mM HEPES, pH 7.4; 1 mM MgCl2; 0.5 mM CaCl2) until swollen, and then lysed with a solution of 5% benzalkonium chloride in 10% acetic acid.28 The resulting suspension of cell nuclei was diluted in a standardized buffer (Casyton) and counted using an electronic cell counter (Casy). The cell-culture experiments were performed in triplicate.

Determination of Peptide-Binding Properties to Magnetic Reversed-Phase Particles

A defined mixture of seven peptides (angiotensin II, angiotensin I, substance P, bombesin, renin substrate P, ACTH-CLIP 1-17, and ACTH-CLIP 18-39, each at a concentration of 1 or 10 nM) was prepared in DMEM alone or supplemented with 5% FCS or 5% bovine serum surrogate. These mixtures were added to molecular-weight cut-off devices and centrifuged. Magnetic reversed-phase (RP) particles (1 mg; either commercial or custom made) were then mixed with 500 μL of the filtrates and the different particles were separated by applying a magnetic device. After washing the particles three times with 20 mM NH4HCO3, pH 7.2, the bound substances were eluted with acetonitrile:1% formic acid (1/1; v/v). From each eluate 0.5 μL was spotted on to a MALDI target. Spectra were acquired by summing up 1000 laser shots.

Binding Properties of Phenol Red to Different RP Particles

To determine the amount of phenol red (a constituent of DMEM) binding to RP particles, a calibration curve was generated by diluting DMEM with 20 mM NH4HCO3, pH 7.8 (1:1, 1:5, 1:10, 1:50, and 1:100) and measuring the absorption values at λ = 559 nm for each dilution. The binding capacity of phenol red to different RP beads was analyzed by incubating DMEM solutions each with 1 mg of the RP particles. After mixing, the particles were separated with a magnetic device and the solutions were discarded. The beads were washed three times with 20 mM NH4HCO3, pH 7.2, and the absorption value of each washing solution was determined at λ = 559 nm. The amounts of phenol red bound were calculated using the calibration curve.

Sample Separation Using Capillary LC and Direct Spotting to MALDI Targets

Eluates obtained after the purification procedure with the magnetic RP particles were injected into a capillary HPLC System (Agilent Series 1100, Agilent Technologies Deutschland GmbH, Boeblingen, Germany) equipped with a cooled autosampler, an eluent degasser, and a binary capillary pump. For separation, a monolithic PS-DVB column was used with an I.D. of 200 μm and a length of 5 cm. The temperature of the column was controlled by a custom-made column oven and was set to 60°C. The capillary outlet after the column was connected to a 215 Gilson Liquid Handler/Injector (Gilson International B.V., Bad Camberg, Germany), which served as LC-MALDI spotting robot system. Reversed-phase high-performance liquid chromatography (RP-HPLC) was performed with the mobile phase composed of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). After an equilibration time of 16.5 min with a constant percentage of 2% of the organic component, solvent B was increased to 35% until 45 min. The gradient was controlled by the Agilent Chemstation software for LC systems Version B. 01.03 (Agilent Technologies Deutschland, Boeblingen, Germany). The flow rate was 3 μL/min, and the fractions were directly applied to a prespotted AnchorChip (matrix: α-cyano cinnamic acid) via the robotic system as described above. The robotics system was controlled by the Gilson Software Unipoint Version 3.01 (Gilson International B.V.) and each run was automatically started by an outgoing signal from the HPLC system. After a defined time, the spotter started automatically. The fraction time was set to 30 sec and a total of 48 fractions were collected for each run.

Isolation, Separation, and Detection of Substances Secreted by Different Cancer Cell Lines

MCF-7 and MDA-MB231 cells were cultivated as described above. The cell-culture supernatants were collected, centrifuged at 14,000 rpm, and immediately frozen at –20°C. After thawing, the cell-culture supernatants were added to molecular-weight cut-off devices and centrifuged. The filtrates were mixed with 1 mg of the custom-made RP particles and further processed as described above. From each eluate, 0.5 μL was directly applied to a prespotted AnchorChip MALDI target. The remaining eluate was injected into a capillary HPLC system and separation was performed as described above.

Mass Spectrometric Measurements

Mass spectrometry was performed using a Ultraflex II MALDI-TOF mass spectrometer from Bruker Daltonics GmbH equipped with a nitrogen laser (λ = 337 nm) operating at a frequency of 50 Hz. The accelerating voltage used was 25 kV, and mass spectra were measured in the positive mode using the reflector mode. For sample preparation, the collected fractions were applied to prespotted AnchorChips, allowed to dry, and washed as described by the manufacturer. Mass spectra were collected by adding up the spectra from 1000 laser shots (20 × 50 laser shots) from different regions of each spot and processed automatically using the Bruker LC-Warp software. Spectra were externally calibrated using the pres-potted calibration spots of the AnchorChips. The mass accuracy of the instrument using this external calibration procedure was typically better then 50 ppm. The peptide ions detected in each fraction were visualized by the Warp-LC SurveyViewer (Bruker Daltonics GmbH). Additionally, a complete list of all received peptide ion masses and their corresponding retention times received from each HPLC run was generated using the program Warp-LC. After data transfer into Excel, the ion masses of the cell-culture experiments were compared. Each cell-culture experiment was performed in independent triplicates.

RESULTS AND DISCUSSION

Effects of Cell-Culture Supplements on Cell Growth

The detection of secreted peptides requires culture medium conditions fulfilling two requirements: (1) a reduced protein load, and (2) a sufficient peptide content to support cell growth. To develop alternatives to cultivating cells in serum-free medium, different growth supplements were tested, which would maintain cells in a viable and vital state. MCF-7 cells were first cultivated for 4 d in DMEM + 5% FCS. The medium was then changed to DMEM supplemented either with 5% FCS, 5% of a bovine serum surrogate, or 2% Panserin PX10. Parallel cultures received protein-free DMEM or Panserin H. After an additional incubation time of 2 d, morphology and cell growth were analyzed by microscopic imaging and nuclei count.

As shown in Figure 1, MCF-7 cells cultured in DMEM supplemented with 5% FCS (1A) or with 5% bovine serum surrogate (1B) grew as a monolayer of epithelial-like cells spread on the substratum. In contrast, when cultured in DMEM with 2% Panserin PX10 (1C) or in Panexin H (1D), cells aggregated into clusters.

FIGURE 1
Morphology of MCF-7 cells growing in DMEM supplemented with (A) 5% FCS, (B) 5% bovine serum surrogate, (C) 2% Panserin PX10, or (D) growing in a complete medium Panexin H. Cells were cultivated for 4 d.

The more malignant cell line MDA-MB231 was also investigated in different medium supplements. When cultivated under standard conditions, i.e., in DMEM with 5% FCS, their morphology differs from that of MCF-7 cells by being more flat in shape and giving rise to homogenous monolayers (Figure 2A). When these cells were cultivated in DMEM supplemented with 5% bovine serum surrogate, no morphological differences were observed (Figure 2B).

FIGURE 2
Morphology of MDA-MB231 cells growing in (A) DMEM + 5% FCS or (B) DMEM + 5% bovine serum surrogate. Cells were cultivated for 4 d.

The effects of different medium supplementation on cell growth are shown in Figures 3 and and4.4. When MCF-7 cells were cultivated in medium supplemented with 5% bovine serum surrogate, their rates of proliferation were as good as with medium containing 5% FCS (Figure 3). Experiments with MDA-MB231 cells and other cell lines, e.g., hybridoma cells, also confirmed that the bovine serum surrogate did not appear to alter growth behavior (data not shown).

FIGURE 3
MCF-7 cell number growing in DMEM + FCS or DMEM + 5% serum surrogate. Cells were incubated for 6 d either in DMEM + 5% FCS or DMEM + 5% bovine serum surrogate. The culture medium was refreshed on day 3, and cell nuclei were counted on days 0, 3, 4, 5, ...
FIGURE 4
MCF-7 and MDA-MB231 cell numbers growing in DMEM alone or with 5% serum surrogate. Cells were incubated for 4 d in DMEM + 5% FCS before medium was changed to either DMEM alone (serum-free conditions) or DMEM + 5% bovine serum surrogate for an additional ...

For mass spectrometric analyses, MCF-7 and MDA-MB231 cells were therefore cultivated in DMEM with 5% bovine serum surrogate until confluency before either refreshing the culture with the same medium or changing to serum-free DMEM. As Figure 4 shows, both cell lines continued to proliferate in medium with 5% bovine serum surrogate. When the medium had been changed to serum-free conditions, MDA-MB231 cells continued to proliferate, albeit at reduced rate, while MCF-7 cells stopped growing within a day and died within 48 h. On the basis of these results, cells were cultivated in DMEM with 5% bovine serum surrogate in experiments designed for the detection of secreted substances in culture supernatants.

Purification of Peptides from Cell-Culture Medium

For developing an efficient and rapid purification procedure for peptides from culture medium, the binding quality of commercially available and custom-made RP particles was compared. A mixture of seven peptides (see Material and Methods) with each peptide at a concentration of either 1 nM or 10 nM was added to serum-free DMEM. As the mass spectrometric analysis of the bead eluates shows (Figure 5), the custom-made RP particles yielded a higher sensitivity and a broader specificity for peptide binding than the commercial particles.

FIGURE 5
Mass spectra of different peptide mixtures purified with different RP-particles. The spectra received for a peptide concentration in DMEM of 10 nM (A/B) and 1 nM (C/D) are shown using the commercial (A/C) or custom-made (B/D) reversed-phase particles. ...

During the binding experiments with DMEM, which contains phenol red as a pH indicator, the washing solutions from commercial RP particles showed a strong red discoloration, indicating that phenol red from the medium had been bound. Washing solutions from the custom-made magnetic RP beads showed only a slight discoloration. Spectrometric measurements of the washing solution at λmax = 559 nm confirmed the presence of phenol red. To compare the binding efficiency of phenol red with respect to the protein content in the medium, bovine serum albumin (BSA) was added at concentrations of 0–500 μg/mL to DMEM and mixed with 1 mg of either commercial or custom-made magnetic RP beads. Following a brief incubation, the particles were separated and washed before measuring the absorption of the washing solution, as described in Material and Methods.

As Figure 6 shows, the amount of phenol red bound by the commercial RP beads was reduced only in DMEM containing BSA concentrations greater than 50 μg/mL. In contrast, only marginal amounts of phenol red were found in the washing solutions when using custom-made RP particles, irrespective of the BSA concentration in the medium. These experiments show that phenol red also binds to RP beads, which could diminish the binding capacity of the beads for peptides or proteins. This effect, however, is less pronounced with the custom-made than with the commercial RP particles, making the modified particles more suitable for the enrichment of peptides at pico- to nanomolar levels.

FIGURE 6
Binding of phenol red dye to different magnetic beads in the presence of different BSA concentrations in DMEM. ([filled square] = commercial RP-beads; ○ = custom-made RP-beads)

To investigate the influence of substances present in FCS on the peptide-binding capacity of the RP particles, a mixture of seven peptides (as above) was added to DMEM supplemented with either 5% FCS or 5% bovine serum surrogate. After the adsorption step, the magnetic RP particles were washed, and the bound substances were eluted as described in Material and Methods. From each eluate 0.5 μL was pipetted onto the MALDI target. The mass spectrum of the eluates obtained from probes in medium with 5% FCS shows numerous unknown ion peaks (Figure 7), and only three of seven reference peptides could be identified with low signal intensities. In contrast, good ion signals corresponding to the seven peptides could be identified in mass spectra when these peptides were in DMEM + 5% bovine serum surrogate. Only a few minor ion signals from unknown substances were observed.

FIGURE 7
Mass spectra after purification of a peptide mixture in culture medium containing (A) DMEM + 5% FCS or (B) DMEM + 5% bovine serum surrogate. The purification of the peptide mixture was performed using custom-made RP-beads. (* indicates undefined substances) ...

Enhancement of the Detection Limit Using LC-MALDI

The direct application of the eluates from the magnetic RP particles to the prespotted AnchorChips resulted in very complex and dense ion signal patterns (data not shown). An additional drawback in the direct application of the eluates to prespotted MALDI targets is the dissolution of the thin prespotted matrix film due to the high content of organic solvent in the eluate. After recrystallization, the matrix spots were not located at the identical position on the target, resulting in problems during the automatic spot detection. Therefore, eluates from the RP particles were fractioned using LC-MALDI.

The sensitivity of the LC-MALDI system was first tested with increasing amounts of a peptide mixture (0–10 nM) added to DMEM with 5% bovine serum surrogate. The peptides were isolated and purified using the custom-made magnetic RP particles. Each eluate with the isolated peptides was separated with LC-MALDI as described in Material and Methods. The collected fractions were allowed to dry on the matrix, and mass acquisition was performed automatically using Warp-LC. Each LC-MALDI run was visualized with the SurveyViewer. As shown in Figure 8, the ion signals of angiotensin I, substance P, and bombesin were observed at concentrations as low as 1 nM, while renin substrate was detected at 100 pM and the ion signal of ACTH-CLIP 18-39 was obtained even at the 10 pM level. In each LC-MALDI run, four unknown ion peaks were observed in a mass range between 3100 and 3400 Da with almost unvarying intensities. When cell-culture experiments were performed (see below), this ion peak cluster served as control for successful LC runs. Three independent LC-MALDI runs using equal amounts of known peptides were performed, resulting in a variation of the ion signal intensities for the peptides not exceeding 15% (data not shown).

FIGURE 8
LC-MALDI mass spectra of different peptide mixtures with concentrations between 0 and 10 nM. Seven different peptides in DMEM + 5% bovine serum surrogate were purified using custom-made RP-beads. The eluates were separated and detected by performing LC-MALDI-TOF-MS. ...

Determination of Secretion Patterns from Different Cancer Cell Lines

For the mass spectrometric determination of the substances from cell-culture supernatants, the described methods were combined. The workflow of the methods is shown in Scheme 1.

SCHEME 1
Flow chart for isolation, purification and mass spectrometric detection of substances secreted by cancer cells.

MCF-7 cells and MDA-MB231 cells were cultivated for 4 d before medium was changed to DMEM + 5% bovine serum surrogate. Cell supernatants were then collected after 0 and 48 h, and the substances were isolated with the custom-made magnetic RP particles. After washing the beads, the elution step and the following LC-MALDI runs were performed as described above. Figure 9 shows the LC-MALDI runs of the isolated substances from MCF-7 and MDA-MB231 cancer cells obtained 0 h and 48 h after the medium change. The culture medium with 5% bovine serum surrogate incubated without cells served as control.

FIGURE 9
Secretion peptide patterns from the mammary carcinoma cell lines MCF-7 and MDA-MB231. Cell-culture supernatants and culture medium (DMEM + 5% bovine serum surrogate) as control collected at 0 h and 48 h after medium change were purified and analyzed as ...

In each LC-MALDI run, an ion cluster at a mass range between 3100 and 3400 Da appeared after a defined retention time, which is indicative of a successful chromatographic run. At time 0 h, the ion signals obtained from the medium control and the MCF-7 and the MDA-MB231 cells showed almost identical ion peak distributions (Figure 9 A–C), as was to be expected. In contrast, the signals obtained at time 48 h (Figure 9 D–F) showed remarkable differences between the MDA-MB231 cells and the MCF-7 cells, and these were clearly different from those of the medium control. Supernatants from MDA-MB231 cells contained more ion signals than those of the MCF-7 cells.

Ion signals obtained from three independent cell experiments were averaged and summarized in an ion peak list (Table 1). Ion signals from the medium controls at 0 and 48 h, as well as cellular signals below 1000 Da, were omitted for clarity. As indicated in Table 1, a number of ion signals can be detected that are common to both cell types; however, there are also numerous ion signals that are specific for either MCF-7 or MDA-MB231 cells. These results thus represent the distinct patterns of molecules secreted by two different cancer cell lines.

TABLE 1
A list of Ion Signals Received from Substances Secreted by MCF-7 and MDA-MB231 Cells

CONCLUSIONS

A methodology has been developed that enables the detection of molecules secreted at very low concentrations by different cancer cells. The high sensitivity of the procedure was made possible by the combination of two different improvements: (1) the replacement of FCS by a bovine serum surrogate as a culture medium supplement, and (2) the use of custom-made magnetic RP particles, which allow a rapid and efficient isolation and purification of substances from cell-culture supernatants. The development of a serum surrogate became necessary to avoid the high amounts of low-molecular-weight substances from FCS, which hamper the detection of very low concentrations of secreted molecules. It also provides a more physiological growth condition than the use of serum-free culture medium. By using custom-made magnetic RP particles, which were designed for the efficient binding of peptides and proteins in cell-culture supernatants and which can be easily separated from the probe with a magnetic device, the isolation and purification of such secreted molecules can be performed from parallel probes in less than 1 h.

A further increase in ion intensity for the separation and detection of secreted molecules was achieved by using an LC-MALDI system. Exploiting the features of automatic MALDI-TOF-MS measurements by using repeated laser shots at different positions on a flat matrix surface, this technique is also a very attractive method for the quantitative determination of substances in the pM to nM range.

Using this improved technology, we were able to obtain stable and reproducible mass spectrometric ion patterns of substances secreted by different cancer cell lines into their culture supernatants. These ion mass distributions were found to be distinctly different for the low malignant cancer cell line MCF-7 and the highly malignant cancer cell line MDA-MB231. In future, this technology will be applied for the investigation of other cancer cell lines. However, this system can also be employed for other mass spectrometry–based systems in biomedical applications.

ACKNOWLEDGMENTS

This research was supported by the DFG Grant OT 92/3-1 and Roche Diagnostics GmbH, Penzberg, Germany. The authors wish to thank Ildiko Szabados for competent assistance in cell culture. There are no conflicts of interest in this manuscript..

REFERENCES

1. Evtimova V, Zeillinger R, Weidle UH. Identification of genes associated with the invasive status of human mammary carcinoma cell lines by transcriptional profiling. Tumour Biol. 2003;24(4):189–198. [PubMed]
2. Gygi SP, Rochon Y, Franza BR, Aebersold R. Correlation between protein and mRNA abundance in yeast. Mol Cell Biol. 1999;19(3):1720–1730. [PMC free article] [PubMed]
3. Gewartowski K, Tomecki R, Muchowski L, Dmochow-Ska A, Dzwonek A, Malecki M, et al. Up-regulation of human PNPase mRNA by beta-interferon has no effect on protein level in melanoma cell lines. Acta Biochim Pol. 2006;53(1):179–188. [PubMed]
4. Chen G, Gharib TG, Huang CC, Taylor JM, Misek DE, Kardia SL, et al. Discordant protein and mRNA expression in lung adenocarcinomas. Mol Cell Proteomics. 2002;1(4):304–313. [PubMed]
5. Tong D, Schneeberger C, Czerwenka K, Schmutzler RK, Speiser P, Kucera E, et al. Messenger RNA determination of estrogen receptor, progesterone receptor, pS2, and plasminogen activator inhibitor-1 by competitive reverse transcription-polymerase chain reaction in human breast cancer. Clin Cancer Res. 1999;5(6):1497–1502. [PubMed]
6. Jungblut PR, Zimny-Arndt U, Zeindl-Eberhard E, Stulik J, Koupilova K, Pleissner KP, et al. Proteomics in human disease: Cancer, heart and infectious diseases. Electrophoresis. 1999;20(10):2100–2110. [PubMed]
7. Alexander H, Stegner AL, Wagner-Mann C, Du Bois GC, Alexander S, Sauter ER. Proteomic analysis to identify breast cancer biomarkers in nipple aspirate fluid. Clin Cancer Res. 2004;10(22):7500–7510. [PubMed]
8. Peter JF, Unverzagt C, Krogh TN, Vorm O, Hoesel W. Identification of precursor forms of free prostate-specific antigen in serum of prostate cancer patients by immunosorption and mass spectrometry. Cancer Res. 2001;61(3):957–962. [PubMed]
9. Peter JF, Unverzagt C, Hoesel W. Analysis of free prostate-specific antigen (PSA) after chemical release from the complex with alpha(1)-antichymotrypsin (PSA-ACT) Clin Chem. 2000;46(4):474–482. [PubMed]
10. Anderson NL, Anderson NG. The human plasma proteome: History, character, and diagnostic prospects. Mol Cell Proteomics. 2002;1(11):845–867. [PubMed]
11. Brand J, Haslberger T, Zolg W, Pestlin G, Palme S. Depletion efficiency and recovery of trace markers from a multiparameter immunodepletion column. Proteomics. 2006;6(11):3236–3242. [PubMed]
12. Camerini S, Polci ML, Liotta LA, Petricoin EF, Zhou W. A method for the selective isolation and enrichment of carrier protein–bound low-molecular weight proteins and peptides in blood. Proteomics Clin Appl. 2007;1(2):176–184.
13. Lowenthal MS, Mehta AI, Frogale K, Bandle RW, Araujo RP, Hood BL, et al. Analysis of albumin-associated peptides and proteins from ovarian cancer patients. Clin Chem. 2005;51(10):1–13. [PubMed]
14. Castellón E, Venegas K, Sáenz L, Contreras H, Huidobro C. Secretion of prostatic specific antigen, proliferative activity and androgen response in epithelialstromal co-cultures from human prostate carcinoma. Int J Androl. 2005;28(1):39–46. [PubMed]
15. Kratchmarova I, Kalume DE, Blagoev B, Scherer PE, Podtelejnikov AV, Molina H, et al. A proteomic approach for identification of secreted proteins during the differentiation of 3T3-L1 preadipocytes to adipocytes. Mol Cell Proteomics. 2002;1(3):213–222. [PubMed]
16. Zwickl H, Traxler E, Staettner S, Parzefall W, Grasl-Kraupp B, Karner J, et al. A novel technique to specifically analyze the secretome of cells and tissues. Electrophoresis. 2005;26(14):2779–2785. [PubMed]
17. Sato K, Sasaki K, Akiyama Y, Yamaguchi K. Mass spectrometric high-throughput analysis of serum-free conditioned medium from cancer cell lines. Cancer Lett. 2001;170(2):153–159. [PubMed]
18. Sato K, Sasaki K, Tsao MS, Yamaguchi K. Peptide differential display of serum-free conditioned medium from cancer cell lines. Cancer Lett. 2002;176(2):199–203. [PubMed]
19. Warburton MJ, Poole B. Effect of medium composition on protein degradation and DNA synthesis in rat embryo fibroblasts. Proc Natl Acad Sci USA. 1977;74(6):2427–2431. [PubMed]
20. Mostafapour MK, Green H. Effects of withdrawal of a serum stimulus on the protein-synthesizing machinery of cultured fibroblasts. J Cell Physiol. 1975;86(2 Pt 2 Suppl 1):313–319. [PubMed]
21. Plas DR, Thompson CB. Cell metabolism in the regulation of programmed cell death. Trends Endocrinol Metabol. 2002;13(2):75–78.
22. Hutchens TW, Yip TT. New desorption strategies for the mass spectrometric analysis of macromolecules. Rapid Commun Mass Spectrom. 1993;7:576–580.
23. Davies H, Lomas L, Austen B. Profiling of amyloid beta peptide variants using SELDI ProteinChip arrays. Biotechniques. 1999;27(6):1258–1262. [PubMed]
24. Marshall FF. Serum protein profiling by SELDI mass spectrometry: Detection of multiple variants of serum amyloid alpha in renal cancer patients. J Urol. 2005;173(6):1919–1920.
25. Tolson J, Bogumil R, Brunst E, Beck H, Elsner R, Humeny A, et al. Serum protein profiling by SELDI mass spectrometry: Detection of multiple variants of serum amyloid alpha in renal cancer patients. Lab Invest. 2004;84(7):845–856. [PubMed]
26. Doucette A, Craft D, Li L. Mass spectrometric study of the effects of hydrophobic surface chemistry and morphology on the digestion of surface-bound proteins. J Am Soc Mass Spectrom. 2003;14(3):203–214. [PubMed]
27. Villanueva J, Shaffer DR, Philip J, Chaparro CA, Erdjument-Bromage H, Olshen AB, et al. Differential exoprotease activities confer tumor-specific serum peptidome patterns. J Clin Invest. 2006;116(1):271–284. [PMC free article] [PubMed]
28. de Noo ME, Tollenaar RA, Ozalp A, Kuppen PJ, Bladergroen MR, Eilers PH, et al. Reliability of human serum protein profiles generated with C8 magnetic beads assisted MALDI-TOF mass spectrometry. Anal Chem. 2005;77(22):7232–7241. [PubMed]
29. Otto AM, Faderl M, Schonenberger H. Dissociation of estrogenic and cytotoxic properties of an estrogen receptor–binding platinum complex in human breast cancer cell lines. Cancer Res. 1991;51(12):3217–3223. [PubMed]
30. Otto AM, Müller CS, Huff T, Hannappel E. Chemotherapeutic drugs change actin skeleton organization and the expression of β-thymosins in human breast cancer cells. J Cancer Res Clin Oncol. 2002;128(5):247–256. [PubMed]

Articles from Journal of Biomolecular Techniques : JBT are provided here courtesy of The Association of Biomolecular Resource Facilities