Breast cancer is the most commonly diagnosed malignant disease and the leading cause of cancer-related death among women worldwide [1]. Tumor microenvironment plays a crucial role in breast cancer progression, regulating tumor growth, vascularization, metastasis, and response to therapy [2-4]. Tumor-associated macrophages (TAM) are a major component of the innate immune system in breast cancer tissue; their role in breast cancer progression has been shown in animal models and in patients Larionova et al 2020 [5]. Both the overall number of macrophages and, in particular M2-like macrophages were found to indicate poor prognosis in most studies of breast cancer, which however do not account for intratumoral heterogeneity [2, 5, 6]. Our previous studies showed that macrophages can retain antitumor properties in certain intratumoral compartments, as supported by the negative correlation of breast tissue TAMs with local metastases [6]. Macrophages control tumor growth and metastasis by secreting cytokines, extracellular matrix components, and growth factors; the profile of secreted components is defined by programming on the transcriptional level as well as on the epigenetic and metabolic levels [2, 9–11].

The primary tumor is generally surgically removed without any additional exposure, since such exposure will not have any effect on the micrometastases. TAM metabolize chemotherapeutic agents and significantly increase the growth factors and proangiogenic factors secretion to stimulate tumor growth and further metastasis in response to chemotherapy [12]. The macrophage epigenetic program is the link between the genetic code and interaction with stressful factros, endogenous pathological factors and infectious agents. Macrophage pathological programming can create the conditions, in which even single-cell micro-metastases can develop into deadly secondary tumors resistant to chemotherapy and immunotherapy. New approaches for breast cancer treatment aim to edit the TAM epigenome, using advanced CRISPR/ dCas technology for targeted delivery of epigenome editors to target gene promoters. The authors of this article developed a model system of human primary macrophages ex vivo and demonstrated long-term programming of human macrophages using a new biomaterial-based cytokine release system [13].

Precise methods of metabolome analysis — in particular of lipids, the main energy component and building material of membranes, and of amino acids, the monomeric units of proteins — are required to study changes in immune system cells during reprogramming. A technique for metabolome analysis in model TAM and in monocytes of patients with tumors of varied localization will allow to study the molecular composition of monocytes in patients with tumors of varied localization and will contribute to the development of technology for clinical application of any discovered correllations. Data that can be obtained with such a technique is of interest for solving both fundamental and applied problems in biology and medicine; it may be particularly useful for studies of mechanism of pathogenesis in malignant neoplasm. Liquid chromatography combined with mass spectrometry (HPLC-MS/MS), the method this study employs, is currently the most effective approach to determining the molecular composition of biological samples [14]. The key problem of existing approaches is the lack of consensus on the minimum number of cells for analysis, which is especially relevant for monocytes and TAM of patients with cancer, since the complex procedure of obtaining monocytes and TAM requires a large amount of whole blood, long-term post-treatment and differentiation in in culture conditions ex vivo in the absence of cell proliferation.

Our study aimed to identify the minimum number of human immune cells (TAM and monocytes) necessary for the metabolomic profile analysis of these cells in oncological diseases using HPLC-MS/MS.

METHODS

Samples from healthy volunteers were used to determine the optimal number of cells to obtain results in metabolomic spectrum study. The inclusion criteria for healthy volunteers: no history of cancer; absence of chronic diseases in the acute phase. In the study of a group of patients, samples were obtained from four patients with breast cancer (BC) T1–4N0–3M0 (stages IIA–IIIB) luminal B-subtype undergoing treatment at the Oncology Research Institute clinic of the Tomsk National Research Medical Center in 2021. Selection criteria: age 45–55 years; no history of cancer of another localization; no hereditary predisposition to breast cancer (based on family history and genetic testing for the 5382insC, 4153delA, 185delAG mutations in the BRCA1 gene); perimenopause or menopause; morphological verification of the nonspecific invasive carcinoma diagnosis (according to the WHO 2019 classification); estrogen receptors (ER) expression ≥ 5 points; Ki-67 ≥ 40%; no hematogenous metastases. Exclusion criteria: age < 45 years or > 55 years; history of cancer; hereditary predisposition; premenopause; all histological subtypes except non-specific invasive carcinoma; ER expression < 5 points; Ki-67 < 40%; hematogenous metastases. An immunohistochemical study was performed before treatment to determine the tumor molecular subtype in accordance with the ASCO/CAP recommendations. BC stage was determined according to the AJCC TNM classification (8th edition, 2017).

We studied the TAM model system [13], blood monocytes from breast cancer patients before treatment and blood monocytes from healthy volunteers. Monocytes were isolated from the whole blood (patients) or from the buffy coat (donors). The CD14⁺ cell fraction was obtained by positive magnetic separation using the Miltenyi Biotec protocol. Model differentiated M2 macrophages were obtained by culturing monocytes from healthy volunteers at a concentration of 10⁶ cells per 1 ml of X-VIVO macrophage medium (Lonza; Germany) supplemented with 1 ng/ml M-CSF (Peprotech; Germany) and 10^–8 M dexamethasone (Sigma-Aldrich; Germany). Differentiation of macrophages into protumor M2 macrophages was induced by 10 ng/ml IL4 (Peprotech; Germany). Model TAM was obtained by adding conditioned media obtained from MCF7 breast carcinoma cells to the obtained macrophages in a volume of 10% of the total medium amount. Macrophages differentiated for 6 days at 7.5% CO₂ and were afterwards removed from the culture plates with a rubber scraper in the cold.

The resulting monocytes and macrophages were centrifuged in a solution of cold 0.3% ammonium acetate and 0.3% ammonium formate for 5 min at 311g thrice. After separation of the supernate, the precipitate was lyophilized at –197 °C using liquid nitrogen. Samples were stored at –80 °C until further analysis.

Sample preparation for HPLC-MS/MS analysis

Metalobome (lipid and amino acid composition) of TAM and monocytes from patients with breast cancer was determined by HPLC-MS/MS analysis of cell extracts. Lipid extraction was carried out by the modified Folch method [15–17] with double extraction: cells were mixed with 480 µl of chloroformmethanol mixture (2 : 1, v/v) with 20 µl of IS (internal standard) at 4 °C and subjected to ultrasonic treatment for 10 min; then 150 µl of water was added and the sample was centrifuged at 13,000 g for 5 min at room temperature; the lower organic layer containing lipids was taken and the extraction procedure was repeated (adding 250 µl of chloroform-methanol mixture (2 : 1, v/v, 10 min on a vortex; centrifugation at 13,000 g for 5 min); the water-methanol layer was taken into separate tubes; the organic layer was combined with the sample obtained in the first stage and lyophilized in a nitrogen stream. The dried lipid extracts (organic layer) were re-dissolved in 100 µl of isopropanol solution with acetonitrile (1 : 1, v/v). The extracts of water-soluble compounds (water-methanol phase) were used to analyze the amino acid profile: they were redissolved in 100 µl of water with acetonitrile (1 : 1, v/v) and 1% formic acid.

HPLC-MS/MS analysis

Lipid extracts were analyzed on a Dionex UltiMate 3000 liquid chromatograph (Thermo Scientific; Germany) connected to a Maxis Impact qTOF mass analyzer with an ESI ion source (Bruker Daltonics; Germany). Lipid separation by HILIC was performed according to a modified procedure described previously [18, 19]: 3 µL of the sample was injected into the system. Separation was performed on a Spherisorb Si column (150 — 2.1 mm, 5 µm; Waters, USA) at a flow rate of 50 µl/min using acetonitrile as solvent A and aqueous ammonium acetate solution (5 mmol/l) as solvent B (linear gradient from 6 to 23% (v/v) solvent B) for 25 min (column temperature 40 °C). Separation of lipids by reversed-phase chromatography was carried out on a Zorbax C18 column (150 — 2.1 mm, 5 µm; Agilent, USA) with a linear gradient from 30 to 90% of eluent B (solution of acetonitrile/isopropanol/water, 90/8/2 vol., with the addition of 0.1% formic acid and 10 mmol / l ammonium formate) for 20 minutes. A solution of acetonitrile/water (60/40 vol.) with the addition of 0.1% formic acid and 10 mmol/l ammonium formate was used as eluent A. The elution flow rate was 40 µL/min, the injected sample volume was 3 µL. Mass spectra were obtained in the positive and negative ion mode for 100–1700 m/z (capillary voltage — 4.1 kV; spray gas pressure — 0.7 bar; drying gas flow rate — 6 l/min; drying gas temperature — 200 °C). To identify lipids, tandem mass spectrometry was performed in dependent scan mode with a window width of 5 Da.

Amino Acids  LC-MS/MS Analysis Kit (Jasem; Turkey) (1-methyl-L-histidine, 3-aminoisobutanoic acid, 3-methyl-L-histidine, β-alanine, DL-5-hydroxylysine, DL-homocystin, ethanolamine, γ-aminobutanoic acid, L-2-aminoadipic acid, L-2-aminobutanoic acid, L-alanine, L-anserine, L-arginine, L-asparagine, L-aspartic acid, L-carnosine, L-citrulline, L-cystathionine, L-cystine, L-glutamic acid, L-glutamine, L-glycine, L-histidine, L-homocitrulline, L-lysine, L-norvaline, L-ornithine, L-phenylalanine, L-proline, L-serine, L-threonine, L-tryptophan, L-tyrosine, L-valine, sarcosine, taurine, trans-4-hydroxy-L-proline) was used for quantitative analysis of 38 amino acids in the cells. 40 µl of aqueous cell extract was mixed with 50 µl of internal standard solution and 300 µl of Reagent № 1 (Jasem; Turkey); the resulting mixture was thoroughly shaken for 2 min, then centrifuged at a rotation speed of 13,000 rpm for 10 min; 200 µl of the supernatant was taken into a vial with an insert, 15 µl was injected into an HPLCMS/MS chromatographic system consisting of an Agilent 6460 triple quadrupole mass spectrometric detector (Agilent; USA) and an Agilent 1260 II liquid chromatograph (Agilent; USA) with a column Amino acids-HPLC Column (Jasem; Turkey).

Conditions for chromatographic separation: temperature — 30 °C, eluent A — 0.1 vol.% formic acid in water; eluent B — acetonitrile 100%, gradient elution from 78% B to 20% B in 4.5 min; return to 78% B and isocratic mode. Amino Acid LC-MS/MS Analysis Kit (Jasem; Turkey) was used for amino acid analysis. Ion source parameters ESI: positive mode; drying gas temperature — 150 °C; capillary voltage — 2 kV. Mass spectrometry was performed in the multiple reaction monitoring (MRM) mode.

Lipid identification and analysis of GC-MS data

For preprocessing of chromato-mass spectrometric data, the msConvert programs from the Proteowizard 3.0.9987 package were used to convert files into the MzXml format, which contains information about the mass spectrum at any moment in time, and the ms2 format, which contains information about tandem mass spectra. In addition, the MzMine program was used to isolate the peaks and create a table containing information on the registered ions: the mass of the ion, the area of   its chromatographic peak, and the release time. To identify lipids, the LipidMatch program [20] was used, which took into account the exact mass (0.01 Da) of the corresponding ion, the mass of fragment ions from the tandem mass spectra of the identified ion falling within the specified mass window (3 Da), and the specified vicinity of the retention time (0.5 min). Semi-quantitative analysis of lipids was performed using previously developed R language scripts [11–23].

RESULTS

To develop an optimal methodology for studying human primary macrophage lipidome by LC/MS analysis of lipid extracts, several extraction/chromatography separation methods were compared, including reversed-phase chromatography and hydrophilic interaction liquid chromatography (HILIC), and the minimum required cell number was determined. The efficiencies for three extraction methods were compared in advance. One protocol, adapted from an earlier study [19], used a modified Folch method [2] with double extraction (see Methods). The second protocol was similar to another previously described approach [15], also Folch method-based [16], with single extraction. Shortly, the cells were mixed with 200 µL of ammonium acetate aqueous solution (155 mmol/L) and 990 µL of cold chroloform/methanol (10 : 1, v/v) and sonicated for 10 min; the sample was diluted with 150 µL of water and centrifuged at 13,000 g for 5 min at room temperature; the lipid-containing lower organic phase was dried in a nitrogen flow and re-dissolved in 200 µL of isopropanol/acetonitrile (1 : 1, v/v).

The third protocol employed the Bligh and Dyer extraction principle [24]. Shortly, the cells were mixed with 300 µL of chroloform/methanol (1 : 2); then 100 µL of chloroform were added and the sample was homogenized for 1 min; then 100 µL of water were added, the sample was homogenized for another 1 min and centrifuged at 1,000 g for 10 min at 4 °C. The organic phase was collected and the polar phase was re-extracted with 2 mL of cold chloroform; the organic phase was collected, mixed with the extract obtained at the previous step, dried in a nitrogen flow, re-dissolved in 4 mL of cold chloroform/methanol (1 : 1) with added 1.8 mL of NaCl aqueous solution (20 mmol/L) and double-filtered using 0.2 µm PFTE syringe filter.

The second method with a single Folch extraction showed low efficiency for weakly polar lipids (triglycerides, diglycerides and ceramides). The advantages of the third method include the possibility of successful extraction of cardiolipins on a par with other classes of  lipids. However, the approach is laborious and involves more stages than the first two protocols, which hampers the reproducibility and extraction yields. Overall, the first protocol turned out to be the most universally applicable and optimal in terms of reproducibility, as well as labor time costs, and accordingly was used as a method of choice in subsequent experiments. Importantly, the water-methanol phase (discarded in the course of lipid extraction) can be preserved for analysis of polar metabolites, notably amino acids, in the same sample.

In this study, we tested two principal techniques of chromatographic separation as applied to lipidome: reversed-phase chromatography and HILIC. Characteristic chromatograms and mass spectra of the studied lipid extracts are given on fig. 1 and fig. 2. In the positive ion mode, about 200 lipids were identified using characteristic tandem mass spectra and about 1000 lipids were identified by exact mass alone. The identified lipids were classified as mono, di- and triglycerides, ceramides and various modifications of phosphatidylcholines and phosphatidylethanolamines including oxidation products. HILIC advantages include default separation of lipids into classes; however, the MS peak intensity obtained by this technique is on average 30% lower compared to reversed-phase chromatography. In addition, HILIC method is of questionable suitability for non-polar lipids (notably triglycerides), poorly bound by the column and often lost (being eluted at a timepoint close to the 'dead time' of the column). Accordingly, the number of lipid identifications by HILIC was 25% lower. Subsequent tests were carried out using reversed-phase chromatography.

Table 1 (tab. 1) summarizes the lipidomic data for the analyzed samples. Extracts prepared from 2 × 10⁶ monocytes presented with 15% more lipids than 1 × 10⁶ cell extracts, whereas for 5 × 10⁶ vs 2 × 10⁶ and 10 × 10⁶ vs 5 × 10⁶ the differences constituted 21% and 65%, respectively. For TAM, 2 × 10⁶, 5 × 10⁶ and 10 × 10⁶ cells provided 28%, 10% and 6% more identifications than 1 × 10⁶, 2 × 10⁶ and 5 × 10⁶ cells, respectively. Thus, with limited access to biological material (primary human macrophages), it is necessary to analyze at least 2 × 10⁶ cells, as the 1 × 10⁶ to 2 × 10⁶ transition provides the maximal increment in the number of identifications. For monocytes, no corresponding increment dynamics were encountered.

The results of amino acid profiling of the monocyte and TAM samples are provided in tab. 2. The chromatographic peak areas correlate with cell number, with the exception of the smallest monocyte samples containing 1 × 10⁶ cells, which is apparently due to LC/MS sensitivity limits. In this regard, samples suitable for the analysis should contain at least 2 × 10⁶ cells.

CONCLUSIONS

We developed a HPLC-MS/MS-based method for monocyte and cultured TAMs metabolome analysis. The modified Folch method with double extraction was chosen as the optimal sample preparation method because it allows to simultaneously isolate the lipid extract (i.e. the organic phase) and the extract of water-soluble substances, in particular amino acids (i.e. the water-methanol phase). The most intense peaks in the mass spectra correspond to lipids and other metabolites, proving the effectiveness of this method of metabolomic profile obtainment. Within the framework of this study, use of both HILIC and reversed-phase chromatography was analyzed: the two methods yield different chromatographic separations for HPLC-MS/MS of lipid extracts of cells, with data yielded by reversed-phase chromatography proving more useful. A comparative analysis of lipid and amino acid composition of monocyte and macrophage samples containing 1 × 10⁶, 2 × 10⁶, 5 × 10⁶, and 10 × 10⁶ cells was carried out; lipids and amino acids were found in all studied samples, but the most effective identification can be performed for samples containing at least 2 × 10⁶ cells. The developed HPLC-MS/MS-based approach for cellular metabolome analysis will be used to determine the metabolomic composition of macrophages and monocytes, and to search for characteristic differences in the metabolomic profile of TAM and monocytes in patients with cancer.