PL247114B1 - Method of classifying MSC cells based on therapeutic properties - Google Patents

Abstract

The subject of the application is a method of classifying MSC cells with respect to therapeutic properties for use as a tool for matching a donor to a recipient. Additionally, the subject of the application is also a method of classifying MSC cells with respect to therapeutic properties supported by computer.

Description

Description of the invention

The subject of the invention is a method of classifying MSC cells with respect to therapeutic properties for use as a tool for matching a donor to a recipient. Additionally, the subject of the invention is also a computer-aided method of classifying MSC cells with respect to therapeutic properties.

Mesenchymal stem/stromal cells (MSCs) have been a subject of clinical research and a tool for regenerative therapy for over a generation. This is primarily due to their anti-inflammatory and immunomodulatory properties. Many studies have shown that MSC therapies can be helpful in treating a wide range of diseases and disorders, such as osteoarthritis, joint cartilage loss, congenital bone defect (OI), rheumatoid arthritis, graft-versus-host disease (GVHD), diabetes, Crohn's disease, multiple sclerosis (MS), amyotrophic lateral sclerosis (ALS), and other inflammatory diseases. The possibility of using MSCs to treat such a wide range of diseases is due to their biological activity, which consists of secreting paracrine factors such as cytokines, chemokines, and exosomes, which allow them to modulate the immune response by reducing inflammation. Moreover, MSC therapies are considered safe (1) (2) (3). Human MSCs can be obtained from a variety of tissues, and their therapeutic potential for treating diseases may depend on their origin. The most recognizable source is bone marrow (BM). Collection of BM from a donor is an invasive and painful procedure. In addition, the therapeutic properties of BM-derived MSCs, the number of isolated cells, and their differentiation potential are donor-dependent and decrease with age. Alternative sources of MSCs include adipose tissue, dental pulp, synovium, peripheral blood, dentin, endometrium, and fetal tissue such as Wharton's jelly from the umbilical cord, cord blood, amnion, amniotic fluid, and placenta. Most importantly, perinatal tissues are an accessible source of cells for cell therapy without the need for invasive collection procedures, as they are considered medical waste. In addition, scientists have proven that early fetal sources of MSCs enable the isolation of a greater number of cells with high multipotency and the ability to expand than from bone marrow from adult donors (4) (5) (6). The flagship and classification features of MSCs according to the International Society for Cell & Gene Therapy (ISCT), regardless of the origin and method of cell isolation, is the ability to attach to plastic. Their morphology is similar to fibroblasts and they share properties with them, such as the expression of CD105, CD73 and CD90 in the absence of CD11b or CD14, CD19, CD34, CD45, CD79 and the histocompatibility complex class II antigen (HLA-DR).

They demonstrate the ability to differentiate into three different cell types (osteoblasts, chondroblasts, and adipocytes) using specific culture media. However, recently the ISCT has called for improved characterization of MSCs, recommending the inclusion of functional assays for the cells (1) (7) (8).

Cell products manufactured in accordance with Good Manufacturing Practices (GMP) based on MSCs are currently designated by the European Medicines Agency (EMA) as Advanced Therapy Medicinal Products (ATMP). EMA guidelines during the release of the cellular end-product provide confirmation of the safety, quality and efficacy of cell therapy. This is confirmed by functional, cytogenetic, potency, sterility and purity tests, such as the absence of endotoxins or mycoplasma (2). In the event of any irregularities in the tests for the cell product, the product is disqualified. The main problem of cell therapy is poor characterization and divergence in therapeutic efficacy in vivo (1). ISCT guidelines, although widely accepted, do not guarantee the repeatability of the properties of MSC products. It should be emphasized that MSC cell products show significant heterogeneity in terms of proliferation dynamics, clonogenicity, secreted factors (secretome) or differentiation potential. This heterogeneity is not taken into account in routine cell characterization (5). Differences that have a proven impact on the cell product are different culture media, in particular their supplementation, cell seeding density, passage number or culture duration, and even the freezing method. In the case of medium supplementation, the presence of human platelet lysate (hPL) has been confirmed to affect not only cell proliferation but also their diverse potential and immunomodulatory abilities (3). There are differences in cell freezing protocols in different production areas and despite maintaining basic ISCT parameters after cell thawing, the immunomodulatory potential of cells can decrease by up to 50%. Due to the different immunomodulatory abilities of MSCs, they have different effects on the inhibition of T cell responses, maturation of antigen-presenting cells, cytotoxicity of resting NK cells, differentiation of monocytes into immature dendritic cells, or the MSC secretome (increased production of PGE2, TGFb and IL10), which helps in the induction and differentiation of cells into regulatory T cells (Treg) (9).

From document CN112748080 a method of characterizing immunoregulatory properties of mesenchymal stem cell therapy products is known, comprising culturing the product in a medium containing interferons to stimulate the cells, collecting the culture supernatant and subjecting it to a color reaction to measure the activity of indoleamine 2,3-dioxygenase 1 (IDO1) to obtain the amount of L-kynurenine produced by the test product. If the relative biological activity of the test product in relation to the reference cells is > 60%, it is determined that the test product meets the requirements for mesenchymal stem cell therapy products.

A similar approach, based on the rapid assay of indoleamine 2,3-dioxygenase 1 activity to determine the immunomodulatory properties of human mesenchymal stem cells is described in document CN113278674A.

In turn, the description of CA2876499A1 discloses the use of a series of over 20 markers for the detection of clinical grade hESMSC for the treatment of autoimmune diseases, in particular multiple sclerosis. The authors acknowledge that limited sources and variable quality of human tissues from different donors limit the research and use of MSCs and prevent standardization of MSCs as a therapeutic product for large-scale clinical use. MSCs derived from adult tissues are highly mixed cell populations and perhaps only a fraction of cells exert immunomodulatory effects. The method involves the determination of over 20 markers (including CD73, CD90, CD105, CD146, CD166, CD44, CD13, CD29, CD54, CD49E, CD45, CD34, CD31, SSEA4, OCT4, NANOG, TRA-1-60) divided into four groups.

In the article by Petrenko et al. (10). The authors compared the growth kinetics, immunophenotypic and immunomodulatory properties, gene expression and secretion profile of MSCs derived from adult bone marrow (BM-MSC), adipose tissue (AT-MSC) and Wharton's jelly (WJ-MSC) cultured clinically with an emphasis on neuroregenerative potential. The following markers were used: CD10, CD29, CD44, CD73, CD90, CD105, HLA-ABC, CD14, CD45, CD235a, CD271, HLA-DR, VEGFR2, CD34, CD133, CD146, SSEA-4, MSCA-1, CD271, HLA-DR. Similarly, in the article by Li et al. (11) mesenchymal stem cells derived from 4 different sources, human bone marrow, adipose tissue, Wharton's jelly, and placenta, were compared to determine which MSC population showed the strongest immunosuppressive effect on phytohemagglutinin-induced T cell proliferation and which had the greatest proliferative and differentiation potential. The expression of several immune-related genes (including MhC II, TLR4, TLR3, JaGi, NOTCH2, and NOTCH3) was analyzed by RT-PCR and RT-qPCR. The proliferative and differentiation potential of MSCs was determined by standard methods.

In the work of Ouafy et al. (12) the identification and validation of predictive markers associated with the immunomodulatory properties of WJ-MSCs were described. For this purpose, CD119, CD200, and Notch1 were selected and studied. The authors concluded that CD119 and Notch1 could potentially be used as markers of the immunomodulatory properties of WJ-MSCs, but further studies are needed.

A major problem in the field is the lack of standardization of the cell product manufacturing protocol (7). All manufacturers are unlikely to optimize the protocol for isolation and manufacturing of MSC-based cell products. The urgent need for a more rigorous method for characterizing MSCs for quality, safety, and efficacy of MSC products is addressed by the method of classifying MSCs by therapeutic properties according to the invention, which finds application as a standardized extended method for evaluating a cell product for clinical trials or medical experiments by dividing them into therapeutic clusters to eliminate batch variability. Therapeutic clusters may be necessary to avoid variability in the cell product but also to achieve predictable beneficial results of stem cell therapy in vivo. In order to assign a cellular medicinal product to a given cluster from the reference sample of the released medicinal product, additional tests were performed after thawing the sample: an extended cytometric panel with additional markers such as CD119, CD200, CD120b, CD72, CD273, CD274, CD39, HO, a colony forming unit (CFU) test, a molecular safety test (patent application No. P.433374 entitled Application of a set of genes to determine the teratogenic potential of mesenchymal cells, which obtained a patent on August 2, 2022), secretome analysis and a technology for co-culturing MSC from a cellular product with MNC (mononuclear cells) from a healthy donor was developed. The extended characterization and obtained results allowed the creation of an algorithm assigning a series of cellular products (WJ-MSC) to one of three therapeutic clusters. The series of cellular products assigned to therapeutic clusters based on the created algorithm will have similar biological activity. An additional application of the method according to the invention will be the possibility of selecting a cell product from a therapeutic cluster for a given disease area or even for a specific patient. This approach to the cell product and its quality assessment meets the strong need for standardization of the cell product production protocol.

THE ESSENCE OF THE INVENTION

The method of classifying MSC cells with respect to therapeutic properties according to the invention comprises the following steps:

a. measurement of the level of heme oxygenase presence, POU5F1 expression level, change in the share of Th lymphocytes under the influence of MSC cells; change in the share of CD69-positive Th lymphocytes under the influence of MSC cells and change in the share of naive Th lymphocytes under the influence of MSC cells, wherein the change in the share of Th lymphocytes under the influence of MSC cells, change in the share of CD69-positive Th lymphocytes under the influence of MSC cells and change in the share of naive Th lymphocytes under the influence of MSC cells is determined after co-culture of MSC cells with mononuclear cells in comparison to the control consisting of mononuclear cells that were cultured in identical conditions but without contact with MSC cells

b. classification of MSC cells as belonging to one of at least three therapeutic clusters: a pro-regenerative therapeutic cluster with regenerative potential and differentiation potential, or a hyperactive therapeutic cluster that shows no therapeutic effect, or an anti-stress therapeutic cluster with the potential to reduce oxidative stress based on the results of the measurements from step a.

Preferably, in step b:

- if the level of MSC cells expressing heme oxygenase is between 2% and 3.9% and the level of POU5F1 expression is between 0.5 CF and 1.01 CF and the change in the proportion of CD69-positive Th lymphocytes under the influence of MSC cells is between -17.83% and -13% and the change in the proportion of naive Th lymphocytes is between -0.35% and -0.15% and the change in the proportion of Th lymphocytes under the influence of MSC cells is between 0.05% and 0.23%, the MSC cells are classified into the overactive cluster.

- if the level of MSC cells showing the presence of heme oxygenase is from 11.4% to 17% and the level of POU5F1 expression is from 0.55 CF to 0.7 CF and the change in the share of CD69-positive Th lymphocytes under the influence of MSC cells is from -27.33% to -1.59% and the change in the share of naive Th lymphocytes is from -0.22% to 0.16% and the change in the share of Th lymphocytes under the influence of MSC cells is from 0.09% to 0.12%, the MSC cells are classified into the anti-stress cluster.

- if the level of MSC cells expressing heme oxygenase is from 3.4% to 10.1% and the level of POU5F1 expression is from 0.84 CF to 1.4 CF and the change in the proportion of CD69-positive Th lymphocytes under the influence of MSC cells is from -10.6% to 0.59% and the change in the proportion of naive Th lymphocytes is from 0.08% to 0.31% and the change in the proportion of Th lymphocytes under the influence of MSC cells is from -0.02% to 0.09%, the MSC cells are classified into the pro-regenerative cluster.

Preferably, in step a, the change in the share of Th lymphocytes under the influence of MSC cells, the change in the share of CD69-positive Th lymphocytes under the influence of MSC cells and the change in the share of naive Th lymphocytes under the influence of MSC cells are determined after co-culture of MSC cells with mononuclear cells in a quantitative ratio of 1:10, respectively.

Co-culture is most preferably carried out at a temperature of 37°C and an atmosphere of 5% CO2.

Co-culture can be continued for 4 days.

Preferably, the MSC cells are derived from Wharton's jelly or from bone marrow or adipose tissue, preferably Wharton's jelly.

Preferably, step b. is computer-assisted and includes the following steps of building the classifier:

a1. determining at least three therapeutic clusters: a pro-regenerative therapeutic cluster with regenerative potential and differentiation potential, or a hyperactive therapeutic cluster that does not demonstrate a therapeutic effect, or an anti-stress therapeutic cluster with the potential to reduce oxidative stress, wherein the determination comprises measuring MSC cell parameters comprising:

i. CFU test ii. determination of the phenotype of MSC cells by measuring the genetic markers CDKN24, CLDN1, DNMT3B, ESM1, HAND2, HEY1, MYO3B, NANOG, PDGFRA, OOU5F1, TDGF1, TWIST1, ZBTB16 iii. determination of the phenotype of MSC cells by measuring the genetic markers CD119, CD120b, CD274, CD273, CD200, CD72, CD39, HO iv. analysis of the secretome of MSC cells by measuring the secretion of BDNF, HGF, IL6, IL8, MCP1, RANTES, TGFb1, TGFb2

v. determining the immunomodulation of MSC cells by co-culture of MSC cells with mononuclear cells, wherein the co-culture is preferably carried out in a quantitative ratio of 1:10, respectively, at a temperature of 37°C and in a 5% CO2 atmosphere for 4 days, wherein after co-culture the number of mononuclear cells, their viability and the change in the % share of T lymphocytes, B lymphocytes, NK, NKT, Th, Tc, naive Th, memory Th, naive Tc, memory Tc, Th CD69+, Th CD25+CD69+, Th CD25+, non-activated Th, Tc CD69+, Tc CD25+Cd69+, Tc CD25+, non-activated Tc, Th1, Th2, Th17, internal cytokines for: Th1, Th2 and Th17, and T regulatory mononuclear cells are determined vi. division of MSC cells into at least three therapeutic clusters: hyperactive, anti-stress and pro-regenerative based on the results from stages i.-v., b1. building a classifier, which includes:

i. reducing the number of parameters to at least parameters by identifying significant parameters based on classification error coefficients or on the basis of a measure of variable significance based on a loss function, ii. performing correlation of parameters from stages a1. i.-v. and eliminating parameters that are correlated with each other, iii. determining the probability of membership in a specific cluster on the basis of comparing the classification results using the parameters selected in stage b1. i. with membership in the designated cluster in stage a1. vi.

and the steps of MSC classification, including:

c1. providing MSC cells, d1. measurement of selected significant parameters determined in step b1. i. of the MSC cells from step c1., e1. classification of the MSC cells using the classifier determined in step b1. as belonging to one of the therapeutic clusters on the basis of the measurement results from step d1., wherein steps a1. and b1. are performed once for a given population of MSC cells, step a1. vi., b1., e1. are performed by means of calculations in the computer system to which the data from the measurements in step a1. i.-v. or d1 are transferred.

Preferably, in step b1. i. only those parameters that have been indicated as significant are selected.

Preferably, in step a1. vi. the division into clusters is performed using hierarchical cluster analysis.

It is best if in stage b1. i. the identification is performed using the Boruta algorithm with a random forest classifier.

By using the method according to the invention, 14 MSC cell lines were classified into one of three therapeutic clusters:

- pro-regenerative cluster - a sub-population with increased pro-regenerative potential and a higher differentiation potential (including transdifferentiation) than in the other clusters, but without an increased risk of in vivo carcinogenesis.

- overactive cluster - a sub-population that does not produce the expected therapeutic effect.

- anti-stress cluster - a sub-population with a high potential to reduce oxidative stress (very often combined with inflammation) in the diseased area.

Classifying MSC cell lines into one of the therapeutic clusters listed above helps in selecting the appropriate cell product (MSC cell line) for a given disease area and can serve as a tool for matching donors to recipients. Reducing the number of measured parameters of MSC cell lines significantly shortens the time of analyses and their cost.

DESCRIPTION OF FIGURES

The subject of the invention is shown in the drawing in embodiments, in which:

Figure 1 shows the colony formation rate defined as the percentage of the number of colonies to the number of plated cells for 14 cell lines,

Figure 2 shows the mRNA expression of selected genetic markers for 14 cell lines based on the so-called molecular safety panel test. Expression was expressed in fold change to the reference line after normalization to the expression of the ACTB gene,

Figure 3 shows the correlation between the selected genetic markers from Example 2, where the size of the points denotes the absolute value of the correlation, the shape of a circle denotes a positive correlation, and the shape of a diamond denotes a negative correlation, while the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***,

Figure 4 shows the percentage distribution of selected genetic markers in the extended panel for 14 cell lines,

Figure 5 shows the correlation between the markers of the extended panel from Example 3, where the size of the points denotes the absolute value of the correlation, the circle shape denotes a positive correlation, and the diamond shape denotes a negative correlation, while the significance of the correlation at the 0.05 level is marked with *, at the 0.01 level with **, and at the 0.001 level with ***,

Figure 6 shows the secretome analysis of 14 cell lines,

Figure 7 shows the correlation between the analytes of Example 4, where the size of the points denotes the absolute value of the correlation, the shape of a circle denotes a positive correlation, and the shape of a diamond denotes a negative correlation, while the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***,

Figure 8 shows the inhibition of MNC cell expansion in co-culture by the immunomodulatory properties of 14 MSC cell lines (8A) and the viability of MNC cells (8B), with the colors of the bars indicating: black-D0, gray-D4 control, white-D4 co-culture,

Figure 9 shows the number and viability of MNC cells after co-culture with 14 MSC cell lines from Example 5,

Figure 10 shows the percentage of individual MNC cell subpopulations for activation markers: T cells (10b), B cells (10b), NK cells (10c), NKT cells (10d) in the case of co-culture with 14 MSC cell lines from Example 5, where the colors of the bars indicate: black-D0, gray-D4 control, white-D4 co-culture with MSC,

Figure 11 shows the percentage of individual MNC cell subpopulations including: Th (11a), Tc (11b), Th naïve (11c), Th memory (11d), Tc naïve (11e), Tc memory (11f) in the case of co-culture with 14 MSC cell lines from Example 5, where the colors of the bars indicate: black-D0, gray-D4 control, white-D4 MSC co-culture,

Figure 12 shows the percentage of individual MNC cell subpopulations: Th CD69+ (12a), Th CD25+ CD69+ (12b), Th CD25+ (12c), Th inactive (12d) in the case of co-culture with 14 MSC cell lines from Example 5, where the colors of the bars indicate: black-D0, gray-D4 control, white-D4 MSC co-culture,

Figure 13 shows the percentage of individual MNC cell subpopulations: Tc CD69+ (13a), Tc CD25+ CD69+ (13b), Tc CD25+ (13c), Tc inactive (13d) in the case of co-culture with 14 MSC cell lines from Example 5, where the colors of the bars indicate: black-D0, gray-D4 control, white-D4 MSC co-culture.

Figure 14 shows the percentage of individual MNC cell subpopulations for the MNC subpopulation marker: Th1 (14a), Th2 (14b), Th17 (14c), and their internal cytokines: IFN-γ (14d), IL4 (14e),

IL17 (14f), as well as the Treg subpopulation (14g) in the case of co-culture with 14 MSC cell lines from Example 5, wherein the colors of the bars indicate: black-D0, gray-D4 control, white-D4 MSC co-culture, Figure 15 shows the changes of MNC cell co-culture for individual activation markers and MNC subpopulations under the influence of co-culture with 14 MSC cell lines from Example 5,

Figure 16 shows the correlation between MNC cell subpopulations and their activation level, where the size of the points denotes the absolute value of the correlation, the shape of a circle denotes a positive correlation, and the shape of a diamond denotes a negative correlation, while the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***,

Figure 17 shows the division of MSC cell lines into therapeutic clusters using hierarchical cluster analysis,

Figure 18 shows the arrangement of cell lines in the new coordinates determined by principal component analysis.

EXAMPLES OF PERFORMANCE

Materials and methods

Sample preparation for characterization of WJ-MSC cells

Characterization is performed only on cell products released for marketing in accordance with ATMP (Advanced Therapy Medicinal Products). From the series from 14 different donors (ALS393, ALS439, ALS613, ALS558, ALS628, ALS634, ALS397, ALS432, ALS366, ALS410, ALS502, ALS633, ALS414, ALS487), hereinafter also referred to as "cell lines", one representative product (cell line WJ-MSC) manufactured by the Pharmaceutical Factory PB KM was thawed, the cell material of which was used to obtain extended characterization, enabling universal grouping of a given cell therapy product. After centrifugation of the cryoprotectant solution, the cells were suspended in the same complete medium that was used to manufacture the cell products, and then the number and viability of cells were determined. Immediately after thawing the cells, a colony forming unit (CFU) assay was performed. The remaining viable cells were plated onto culture dishes (27,000 cells/cm ² ) in the same complete medium and culture conditions used for the production of cell products. After 3 days, confluence was checked using an inverted microscope. Where confluence was less than 75%, all medium was replaced with fresh medium. In cases where confluence was greater than 75%, cells were detached using TrypLE™ Express Enzyme (1X) Gibco, Bleiswijk, The Netherlands). 10 million cells were plated onto culture dishes (13,000 cells/cm ² ) for the next passage. After another two days of culture, cells were detached using TrypLE™ Express Enzyme (1X) and cell numbers and viability were determined. 5 million cells were preserved as pellets and frozen at -80°C for further RNA isolation for molecular safety assessment using quantitative real-time PCR. Another 3.5 million cells were used for phenotype determination using the extended panel. 0.5 million cells were seeded into 3 wells of a 6-well plate (167,000 cells per well) to measure the secretome of 14 WJ-MSC cell lines. Another 0.5 million cells were seeded into 3 wells of a 6-well plate (167,000 cells per well) and placed in an incubator under standard conditions of 37°C and 5% CO2 for 48 h to prepare the co-culture assay with activated mononuclear cells (MNC cells).

Example 1

Cellular assay of colony forming units (CFU) from thawed cell products

For the colony forming unit assay, 14 WJ-MSC cell lines from thawed product cells for each of the 14 cell products were seeded in 6-well plates at 100 cells/well in triplicate. The cells were cultured for 14 days in the same complete medium and culture conditions used for the production of cell products as described above. After 14 days, to visualize cell colonies, the cells were washed with PBS (Gibco, Bleiswijk, The Netherlands) 4 ml per well, fixed with 4% paraformaldehyde solution (Sigma Aldrich, St. Louis, MO, USA) 1 ml per well, and incubated for 20 minutes at room temperature. For visualization, cells were stained with 0.5% toluidine blue (Sigma-Aldrich, St. Louis, MO, USA) 1 ml per well and incubated for 30 min on a rocker shaker at room temperature. The dye was washed out by washing twice with DPBS (Gibco, Bleiswijk, The Netherlands) 1 ml per well. The degree of colony formation was defined as the percentage of the number of colonies to the number of plated cells.

PL 247114 Β1

No colonies were observed in product lots ALS628 and ALS634. The remaining WJ-MSC showed a colony formation rate ranging from 7 to 19% (Figure 1). Product lots based on WJ-MSC differ in the percentage of colony-forming cells depending on the donor.

Example 2

Expression of selected genes for 14 WJ-MSC cell lines based on the so-called molecular safety panel test (patent application no. P.433374 entitled Application of a set of genes to determine the teratogenic potential of mesenchymal cells, which was granted a patent on August 2, 2022)

The first stage of the test was to isolate genetic material (RNA) from thawed pellets of 14 WJ-MSC cell lines using a column method on a silicon bed. Biological material was isolated and purified using Pure Link RNAase Mini Kit (Invitrogen) and Pure Link Dnase Kit (Invitrogen) according to the manufacturer's protocols. The nucleic acid (RNA) obtained in this way was the starting material for reverse transcription of RNA using the Maxima H Minus Kit to obtain high-quality cDNA devoid of genomic DNA. Finally, real-time polymerase chain reaction (RT-PCR) was performed on the obtained cDNA to identify the presence and relative expression (by comparing the fluorescence signal strength) of selected characteristic genetic markers CDKN2A, CLDN1, DNMT3B, ESM1, HAND2, HEY1, MYO3B, NANOG, PDGFRA, POU5F1, TDGF1, TWIST1, ZBTB16 (PrimePCR 96W custom plate, BioRad) according to the protocol presented in Table 1 below.

Table 1. RT-PCR reaction protocol conditions.

Stage Temperature Time Number of cycles Activation 95°C 2 min 1 Denaturation 95°C 5 sec 40 Annealing/lengthening 60°C 30 sec Melting curve* 65-95°C (0.5°C increments) 5 sec/step 1

* Melting curve step is for SYBR® Green analysis only

WJ-MSC cells, depending on the donor, showed expression at different levels for selected genetic markers in the pluripotent panel (TDGF1, NANOG, POU5F1, DNMT3B), MSC-specific panel (CDKN2A, HAND2, PDGFRA, TWIST1, ZBTB16), mesodermal panel (ESM1, HEY1), ectodermal panel (MYO3B), endodermal panel (CLDN1). The results of the analysis are shown in Figure 2. The correlation between the selected genetic markers is shown in Figure 3.

In the graph, the size of the points indicates the absolute value of the correlation, the shape of a circle indicates a positive correlation, and the shape of a rhombus indicates a negative correlation. Additionally, the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***.

Example 3

Determination of WJ-MSC phenotype using flow cytometry

To define the extended cytometry panel, prepared cells were washed with DPBS (Gibco, Bleiswijk, The Netherlands) and the pellet was resuspended in Brilliant Stain Buffer Plus (Becton Dickinson, Franklin Lakes, USA) at a minimum density of 1 million cells/ml. The cell suspension was incubated with fluorochrome-conjugated antibodies (FITC, BV421, PE, AF647) against CD119, CD120b, CD274, CD273, CD200, CD72, CD39, H O (Becton Dickinson, Franklin Lakes, USA) for 30 min on ice in the dark. To exclude nonspecific binding, the appropriate IgG2b κ isotype antibodies and unstained cells were used as controls. Cell analysis was performed using a Celesta flow cytometer and FACSDiva software (Becton Dickinson, Franklin Lakes, USA). Results are expressed as percentage of cells positive for the respective markers relative to isotype control and unstained cells.

WJ-MSC cells, depending on the donor, have a different percentage distribution of selected genetic markers in the extended panel. The results are presented in Figure 4. The CD72 marker < 0.2% is present at a negligible level in all donors. The CD119 < 3% and CD120b < 5% markers are also present at low levels. In the case of CD200, we can distinguish 3 groups CD200 range 12-14% (ALS393 and ALS439), range 30-50% (ALS613, ALS414, ALS487) and the largest group range 59-88% (ALS366, ALS397, ALS410, ALS432, ALS558, ALS628, ALS633, ALS634, ALS502). The average percentage for the CD273 marker among all donors is 76% (+/- 14). The percentage of CD39 also divided them equally into two groups depending on the donor. CD39 < 10% (ALS397, ALS393, ALS9, ALS628, ALS613, ALS414, ALS487) and CD39 range 12-30% (ALS66, ALS410, ALS432, ALS558, ALS633, ALS634, ALS502). CD274 marker - its average percentage among all donors is 34% (+/- 12). Heme oxygenase-1 (HO) divided all donors into 3 groups. HO < 5% (ALS397, ALS393, ALS439, ALS628, ALS613, ALS414, ALS487), range 6-10% (ALS366, ALS432, ALS634) and range 11-17% (ALS410, ALS558, ALS633, ALS502).

The correlation between the extended panel markers is shown in Figure 5.

In the graph, the size of the points indicates the absolute value of the correlation, the shape of a circle indicates a positive correlation, and the shape of a rhombus indicates a negative correlation. Additionally, the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***.

Example 4

WJ-MSC secretome analysis

After seeding 0.5 million cells in 3 wells of a 6-well plate (167,000 cells per well), the culture was placed in an incubator for 48 hours under standard conditions, i.e. at 37°C and 5% CO2. The culture was carried out in the same complete medium that was used for the manufactured cell products. After 48 hours, the medium was removed from the cells and washed twice with DPBS. 4 ml of complete medium was added to each well and the culture was placed in an incubator for 4 days under standard conditions, i.e. at 37°C and 5% CO2. After 4 days of culture, the supernatant from each well was collected. After 4 days of culture, the supernatant from each well was collected and centrifuged at 600 x g, 10 min. The resulting supernatants were cryopreserved at -80°C until analysis. The secretion of the following analytes by 14 WJ-MSC cell lines was analyzed: BDNF, HGF, IL6, IL8, MCP1, RANTES, TGFb1, TGFb2 and was performed using Luminex Multiplex assays (Merck). All assays were performed according to the manufacturer's protocol and read on the Lminex instrument. The obtained cytokine levels were analyzed in the Belysa® Software program.

WJ-MSC cells, depending on the donor, have different levels of secretion of BDNF, HGF, IL6, IL8, MCP1, RANTES, TGFb1, TGFb2, which may indicate their biological variability (Figure 6). The correlation between the analytes is presented in Figure 7.

In the graph, the size of the points indicates the absolute value of the correlation, the shape of a circle indicates a positive correlation, and the shape of a rhombus indicates a negative correlation. Additionally, the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***.

A significant correlation was observed for HGF and MCP1. The more cells secreted HGF, the more MCP1 they secreted. Another significant correlation was RANTES ver. MCP1. The greater the secretion of RANTES, the greater the secretion of MCP1. The last significant correlation was TGFb2 ver. HGF. The more TGF, the less HGF secretion.

Example 5

Immunomodulation assay based on co-culture of WJ-MSC with mononuclear cells

One of the elements of the co-culture were thawed MNC mononuclear cells (isolated by centrifugation in a concentration gradient from peripheral blood mononuclear cells - PBMC from a healthy donor). Mononuclear cells were thawed (on the same day as WJ-MSC cells) in a water bath at 37°C, then the freezing medium was removed by centrifugation and the cells were suspended in a lymphocyte expansion medium - ImmunoCult™-XF T Cell Expansion Medium (StemCell Technologies) + 1% (v/v) Antibiotic/Antimycotic (Gibco) + 56 UL/ml premium human IL2 IS (Miltenyi Biotec). After determining the number and viability of cells using an automated cell counter, 8 million cells were collected in 4 ml (maintaining a cell density of 2 million/ml) and plated in a 12-well plate. The cells were activated by adding 25 μl/ml ImmunoCult™ Human CD3/CD28/CD2 T Cell Activator (StemCell Technologies) and the plate was placed in an incubator for 72 hours under standard conditions of 37°C and 5% CO 2 . After 72 hours of activation, the cell suspension was diluted eight-fold and the culture was placed in an incubator for 48 hours for expansion. After 48 hours, the cell suspension was diluted two-fold for further expansion and placed in an incubator for another 48 hours. After 48 h and 7 days of expansion, the activated mononuclear cells were ready for co-culture with 14 WJ-MSC cell lines. For the 48 h culture of WJ-MSC, the inventors had to determine their number in a well of a 6-well plate. This information is needed to seed the correct number of mononuclear cells, which should be 10 times higher in a well than the number of WJ-MSC. For this purpose, cells were detached from one of the four wells using TrypLE™ Express Enzyme (1X) (Gibco, Bleiswijk, The Netherlands). Cells were counted using an automated cell counter. WJ-MSC culture medium was removed from the remaining two wells with cells and washed twice with DPBS. Co-culture was performed in ImmunoCult lymphocyte medium (Stem Cell Technologies) to ensure the best possible conditions for their expansion. To test the immunomodulatory properties of WJ-MSC, mononuclear cells were placed on a 0.4 μm insert (Greiner Bio-One) into a well with WJ-MSC. Ten times more mononuclear cells were placed on the insert than WJ-MSC. The insert allows cocultured cells to interact indirectly by secreting growth factors and cytokines in 4 ml of lymphocyte medium. One of the wells with WJ-MSC acted as a control for lymphocyte expansion with 4 ml of ImmunoCult™-XF T Cell Expansion Medium (StemCell Technologies) + 1% (v/v) penicillin/streptomycin and amphotericin B solution (Gibco) + 56 µL/ml premium human IL2 IS (Miltenyi Biotec). The next control was a well with an identical number of mononuclear cells (in 4 ml of lymphocyte medium) to which the co-culture was placed. All controls and the co-culture in a 6-well plate were placed in an incubator for 4 days under standard conditions, i.e. at 37°C and 5% CO2. After 4 days, the immunomodulatory test was determined as the number of mononuclear cells and their phenotype. The number and viability of mononuclear cells from the control and co-culture were determined using an automatic cell counter. Mononuclear cells were washed once with DPBS (Gibco, Bleiswijk, The Netherlands) and labeled for selected lymphocyte subpopulations and markers characteristic of the level of activation. The pellet was suspended in Brilliant Stain Buffer Plus (Becton Dickinson, Franklin Lakes, USA) at a minimum density of 1 million cells/ml. The cell suspension was incubated with fluorochrome-conjugated antibodies (BV421, BV605, BV650, BV786, BB515, PE-CF594, BB700, APC, APC-R700, APC-H7, PE, PerCP, AF647) against CD25, HLA-DR, CD3, CD69, CD45RO, CD16, CD56, CD4, CD45RA, CD19, CD8, CD40, CD45, CD14, CD183, CD196, CD127, CD194 (Becton Dickinson, Franklin Lakes, USA) for 30 min on ice in the dark. In parallel, internal cytokines IFN-γ (for Th1), IL-4 (for Th2) and IL-17A (for Th17) were assayed using the Human Th1/Th2/Th17 Phenotyping Kit (Becton Dickinson, Franklin Lakes, USA). One million cells (D4 control and D4 co-culture) were suspended in 3 ml of ImmunoCult™ -XF T Cell Expansion Medium (StemCell Technologies) + 1% (v/v) penicillin/streptomycin and amphotericin B solution (Gibco) + 56 UI/ml premium human IL2 IS (Miltenyi Biotec) and seeded per well of a 6-well plate. Then GolgiStop Protein Transport Inhibitor - final concentration of 1 μl/ml (part of the Human Th1/Th2/Th17 Phenotyping Kit Becton Dickinson, Franklin Lakes, USA), PMA - final concentration of 50 ng/ml (Sigma-Aldrich Saint Louis, Missouri, USA) and ionomycin - final concentration of 1 μg/ml (Sigma-Aldrich Saint Louis, Missouri, USA) were added and incubated for 5 hours in an incubator under standard conditions, i.e. at 37°C and 5% CO2. After 5 hours of incubation, the cell suspensions were centrifuged for 5 min at 500 x g and washed once with DPBS (Gibco, Bleiswijk, the Netherlands). Cell pellets were suspended in 0.5 ml of cold BD Cytofix Fixation Buffer (part of the Human Th1/Th2/Th17 Phenotyping Kit, Becton Dickinson, Franklin Lakes, USA) and incubated for 15 min at room temperature. After incubation, the suspensions were centrifuged for 5 min at 500 x g, the pellet was suspended in Perm/Wash buffer (part of the Human Th1/Th2/Th17 Phenotyping Kit, Becton Dickinson, Franklin Lakes, USA), and then a cocktail of CD4 antibodies PerCP-Cy5.5, IL-17A PE, IFN-γ FITC, IL-4 APC (part of the Human Th1/Th2/Th17 Phenotyping Kit, Becton Dickinson, Franklin Lakes, USA) was added and incubated for 30 min at room temperature. Cell analysis was performed using a Celesta flow cytometer and FACSDiva software (Becton Dickinson, Franklin Lakes, USA). Results are presented as percentage of positive cells for the respective markers relative to the control and unstained cell isotype. Inhibition of MNC cell expansion in coculture

PL 247114 Β1 due to the immunomodulatory properties of WJ-MSC cells is shown in Figure 8A. The viability of MNC cells is shown in Figure 8B. The average viability of MNC cells on the day of co-culture establishment (DO) was 84% (+/-5).

At the end of co-culture (D4), the average viability of MNC cells in the control and co-culture was 68% (+/-8).

According to the statistical analysis, the result of co-culture is shown in Figure 9 as the value

D4 CO-BREEDING—D4 CONTROL

To where D4 co-culture, D4 control and DOO were expressed in %, while the co-culture result was expressed as a coefficient illustrating the trend of increase or decrease in relation to the baseline or control. All MNC between day 0 (DO) and day 4 (D4) showed cell proliferation. On D4, only one donor - ALS613 did not inhibit the expansion of MNC cells in co-culture. The percentage of individual subpopulations and activation markers are presented in Figures 10-14.

According to the statistical analysis, the result of co-culture is shown in Figure 15 as the value

D4 CO-BREEDING - D4 CONTROL

Down

T lymphocytes, as a result of co-culture with WJ-MSC from most donors, the % CD3+ on MNC increased. B lymphocytes, as a result of co-culture with WJ-MSC from most donors, the % CD19+ on MNC decreased. NK, as a result of co-culture with WJ-MSC from most donors, the % CD16+CD56+ on MNC decreased. NKT, as a result of co-culture with WJ-MSC from most donors, the % CD3+CD16+CD56+ on MNC decreased. Th lymphocytes, as a result of co-culture with WJ-MSC from most donors, the % CD4+ on MNC increased. Tc lymphocytes, as a result of co-culture with WJ-MSC from most donors, the % CD8+ on MNC decreased. Naive Th lymphocytes, as a result of co-culture with WJ-MSC from most donors the % CD4+CD45RA+ on MNC increased. Memory Th lymphocytes, as a result of co-culture with WJ-MSC from 7 donors the % CD4+CD45RO+ on MNC increased, with one WJ-MSC donor the percentage did not change, with WJ-MSC from 6 donors the % decreased. Naive Tc lymphocytes, as a result of co-culture with WJ-MSC from half of the donors the % CD8+CD45RA+ on MNC increased. Memory Tc lymphocytes, as a result of co-culture with WJ-MSC from half of the donors the % CD4+CD45RO+ on MNC decreased or remained unchanged. Early activation on Th lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD4+CD69+ decreased, only with WJ-MSC from donor ALS628 it increased slightly. Proper activation on Th lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD4+CD25+CD69+ increased. Late activation on Th lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD4+CD25+CD69+ increased. Non-activated Th lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of non-activated Th cells increased, silencing of activation by WJ-MSC. Early activation on Tc lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD8+CD69+ decreased. Proper activation on Th lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD8+CD25+CD69+ increased. Late activation on Tc lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD8+CD25+CD69+ increased. Non-activated Tc lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of non-activated Tc cells increased, silencing of activation by WJ-MSC. Regulatory T lymphocytes, as a result of co-culture with WJ-MSC from most donors the percentage of CD4+CD25+CD1271owFoxp3+ cells decreased. Th1 lymphocytes, as a result of co-culture with WJ-MSC the percentage of Th1 cells increased in half of the donors. Th2 lymphocytes, as a result of co-culture with WJ-MSC in all donors the percentage of Th2 cells increased quite significantly. Th17 lymphocytes, as a result of co-culture with WJ-MSC in most donors the percentage of Th17 cells increased. In the case of internal cytokines in Th cells (CD4+) participating in the Th1 response, a decrease in % IFN-γ was observed in co-culture with WJ-MSC donor ALS 613 and ALS628. In the remaining co-cultures the % IFN-γ increased slightly or remained unchanged. The percentage of IL-4 responsible for the Th2 response increased for co-culture with WJ-MSC for donors ALS393, ALS439, ALS613, ALS397, ALS366, ALS487, in the remaining co-cultures a decrease in % IL-4 was observed. The percentage of IL-17 involved in the Th17 response was increased by co-culture with WJ-MSC from 7 donors (ALS393, ALS439, ALS628,

PL 247114 Β1

ALS634, ALS366, ALS633ALS414), remained unchanged from ALS397 or decreased with donors ALS613, ALS558, ALS432, ALS410, ALS502, ALS487.

Changes in MNC expansion and phenotype in co-culture with WJ-MSC indicate immunomodulation of MSCs. Greater/lower immunomodulatory capacities are associated with biological variability of MSC donors and affiliation to different therapeutic clusters.

The correlation between MNC cell subpopulations and their activation level is shown in Figure 16.

In the graph, the size of the points indicates the absolute value of the correlation, the shape of a circle indicates a positive correlation, and the shape of a rhombus indicates a negative correlation. Additionally, the significance of the correlation at the level of 0.05 is marked with *, at the level of 0.01 is marked with **, and at the level of 0.001 is marked with ***.

Example 6

Statistical analysis

The analyzed problem requires a statistical procedure that includes two important stages. In the first stage, the division into clusters (subpopulations) was determined. Then, in the second stage, a pool of parameters was determined that best characterize the indicated division.

Once all the data has been loaded, some of it needs to be preprocessed. Preprocessing includes, for example, the secretome data. Duplicates are removed as part of the preprocessing. Other steps in this step include removing outliers, constant components, or other noise. Then, for each variable, the mean value is determined for each sample (MSC cell product).

For the MNC phenotype data and the MNC number and viability data, the variable (parameter) values were determined according to the formula:

D4 K0H0D0WLA-D4 CONTROL while in the case when the Do value was not available only the difference was determined. Do values were not available for variables Th1, Th2, Th17 and their internal cytokines.

Before starting the principal component analysis, missing data were filled in using the regularized iterative PC A algorithm [Josse, J & Husson, F. (2013). Handling missing values in exploratory multivariate data analysis methods. Journal de la SFdS. 153 (2), pp. 79-99.]

Stage 1 (division into clusters)

1) In order to determine therapeutic clusters for the division of MSC bank cells, hierarchical cluster analysis (‘complete’ method) was used, the result of which is illustrated in Fig. 17. The initial division of MSC cells into 3 clusters was determined. The parameters were grouped as follows: a) MSC phenotype, b) MNC phenotype, c) MNC viability, d) percentage of colony-forming cells - CFU, e) expression of 12 genes, f) secretory activity of MSC cells,

2) Principal component analysis was used to check the obtained division (Fig. 18).

The input data for this stage is processed into a matrix, where the columns are successive variables (parameters), while the rows are samples (MSC cell lines). A distance matrix is determined from this matrix. The result of this stage is the assignment of each sample to one of 3 clusters (division of samples into 3 subpopulations), which are used in the next stage.

Stage 2 (defining the parameter pool)

As in the first stage, the parameters are combined into a matrix of the same dimensions. Each sample is assigned to one of three clusters - pro-regenerative, hyperactive and anti-stress. The final effect of this stage is a set of several parameters that best divide the samples into specific clusters.

1) The Boruta algorithm was used to find the significant parameters. The Boruta algorithm [Miron B. Kursa, Witold R. Rudnicki (2010). Feature Selection with the Boruta Package. Journal of

Statistical Software, 36(11), p. 1-13.] is used with a random forest classifier to find the importance of each parameter. For this purpose, the parameter importance measure is determined by the algorithm based on the classification error coefficients of each parameter. Boruta's algorithm includes the following steps:

a) Creating a dataset based on the input data with spermuted labels for groups (so-called ‘shadow features’).

b) Using a random forest classifier on the extended dataset and determining a measure of the importance of a given variable (by default, this is Mean Decrease Accuracy). Then, the same thing is done for the data without spermuted labels. The differences between the two sets are then averaged across all trees and normalized by the standard deviation of the differences (this is the Z-score).

c) In each iteration, we check whether the Z-score for the actual data is higher than that for the augmented data after spermuting the label for a given variable (i.e. whether the predictor variable has a higher Z-score than the maximum Z-score of its shadow features), and continuously remove variables that are considered highly unimportant.

d) The algorithm stops when all variables are confirmed or rejected.

2) After selecting the important parameters, to verify the obtained results, the second algorithm for searching for important parameters was used - DALEX [Biecek P (2018). “DALEX: Explainers for Complex Predictive Models in R.” Journal of Machine Learning Research, 19(84), 1-5]. It is used in the example similarly as in the case of Boruta together with the random forest classifier. The selection of parameters is done here using the parameter importance measure based on the loss function of the so-called cross entropy.

3) Only parameters that are significant in both algorithms were taken into further analysis. Then, the significance of correlations between selected parameters was checked.

4) Then, those parameters that are strongly correlated were excluded from the set of significant parameters. This increased the efficiency of the classification procedure and reduced the complexity of the model.

5) Based only on the selected parameters, it was checked whether the use of hierarchical cluster analysis would give the same division into clusters that was obtained at the beginning of the procedure (Stage I) using all parameters.

6) The efficiency of the parameters was checked in the classification process, for which the ensembl classifier [Datta, Susmita, et al. “An adaptive optimal ensemble classifier via bagging and rank aggregation with applications to high dimensional data.” BMC Bioinformatics, vol. 11, 18 Aug. 2010, p. 427] was used, which is a combination of 3 classifiers: support vector method, random forests and recursive partitioning. For each cell line, only selected parameters were taken into account (5 most important: HO level, POU5F1 expression level (OCT3/4), change in % of Th cells under the influence of MSC; change in % of Th cells positive for CD69 under the influence of MSC and change in % of ThN (naive) cells) as the input set for the classifier in order to predict whether a given MSC line belongs to one of the three previously selected clusters. The final result of this stage is to provide the probabilities of belonging of the cell line to a given cluster.

Example 7

Using the procedure described in Example 6, 3 clusters were selected and characterized. In the cluster analysis, the cut-off level for clusters was assumed to be 11.85 (Fig. 17). Three selected clusters were additionally confirmed based on PCA analysis. The first cluster consists of MSC cell products for which the value of the y coordinate in the new coordinate system is negative, while the x coordinate is positive or negative (Fig. 18). The second cluster consists of MSC cell lines for which the value of the y coordinate in the new coordinate system is positive, while the x coordinate is negative. The third cluster consists of MSAC cell lines for which the values of the x and y coordinates in the new coordinate system are positive (Fig. 18).

Basic descriptive statistics for the selected 5 parameters are presented in Table 2:

PL 247114 Β1

Table 2. Summary of basic statistics for the 5 most significant parameters for each of the 3 considered clusters. Results are presented as percentage [%] or fold change with respect to the reference line [FC

Parameter Group (cluster) Median Mean Minimum Maximum HO [%] 0(1) 3.7 5.27 3,4 10.1 HO [%] 1(2) 2.9 2.93 2 3.9 HO [%] 2(3) 15.2 14.7 11.4 17 POU5F1[FC] 0(1) 1.12 1,1 0.84 1.4 POU5F1[FC] 1(2) 0.84 0.78 0.5 1.01 POU5F1[FC] 2(3) 0.68 0.65 0.55 0.7 Th.CD69+L%] 0(1) -5.2 -5.69 -10.6 0.59 Th.CD69+[%] 1(2) -17.5 -16,11 -17.83 -13 Th.CD69+[%] 2(3) -14.67 -14.56 -27.33 -1.59 Th.naive[%] 0(1) 0.19 0.21 0.08 0.31 Th.naive[%] 1 (2) -0.28 -0.26 -0.35 -0.15 Th. naive [%] 2(3) 0,1 0.03 -0.22 0.16 Th[%] 0(1) 0.06 0.05 -0.02 0.09 Th[%] 1(2) 0.06 0.11 0.05 0.23 Th[%] 2(3) 0,1 0,1 0.09 0.12

CLUSTER 1 (group 0)

PRO-REGENERATION CLUSTER (7 MSC lines)

General characteristics: high POU5F1/low Th[%]/high positive Th.naive[%]/lowest decrease in Th.CD69+[%]

Based on the determined average higher expression of the POU5F1 (OCT3/4) gene in cluster 1, it can be assumed that this subpopulation potentially demonstrates increased pro-regenerative potential. Higher expression of OCT3/4 as one of the main pluripotency factors may be correlated with greater plasticity, more open chromatin, and thus characterize MSC populations with a higher differentiation potential (including transdifferentiation) than in clusters 2 and 3. It should be noted that although OCT3/4 expression is determined as higher, it does not reach the levels determined in pluripotent lines, and thus is not associated with an increased risk of in vivo tumorigenesis. Additionally, it should be mentioned that the MSC population from the pro-regenerative cluster in contact with MNCs caused an increased percentage of naive T helper lymphocytes (CD3+CD4+CD45RA+) as described in the literature; such an increased percentage of naive Th is correlated with a better prognosis for ALS patients (13, 14, 15, 16).

CLUSTER 2 (group 1)

OVERACTIVE CLUSTER (3 MSC lines)

General characteristics: low HO/ negative Th.naive[%]

The MSC subpopulation selected on the basis of low HO expression level, with low POU5F1 (OCT3/4) expression and in contact with MNC causes a decrease in the percentage of naive Th lymphocytes. Based on preliminary assessments from the clinical trial, the MSC product derived from the line from cluster 2 did not induce the expected therapeutic effect (3 administrations of MSC from cluster 2 - two patients). As demonstrated in a mouse model of sepsis, the level of HO produced by MSC may translate into a therapeutic effect in vivo (17).

CLUSTER 3 (group 2)

ANTI-STRESS CLUSTER

PL 247114 Β1 (4 MSC lines)

General characteristics: high HO/low POU5F1/high Th[%]

The subpopulation showing the highest percentage of cells with membrane expression of the heme oxygenase 1 gene (ΗΜΟΧΙ). The ΗΜ0Χ1 protein is involved in intercellular communication, immunomodulation and response to oxidative stress. There are reports in the literature about the introduction of forced expression of this gene into MSC cells as an approach to enhance the therapeutic effect of mesenchymal cells. Therefore, the MSC population from cluster 3 will be characterized by a relatively high potential to reduce oxidative stress in the diseased site. It should be mentioned that oxidative stress is very often associated with an active inflammatory state. Populations from cluster 3 show an increased percentage of Th lymphocytes in interaction with activated MNCs, which has also been presented as a positive prognostic factor for patients with ALS (16, 18, 19).

Table 3. Prediction efficiency of classification – probability of belonging of a given cell product series to the appropriate therapeutic cluster no. 1, 2 or 3.

Cluster number real Cluster number designated Cluster membership probability Cluster 1 Cluster 2 Cluster 3 ALS366 1 1 0.98 0.00 0.02 ALS393 2 2 0.34 0.54 0.12 ALS397 1 1 1.00 0.00 0.00 ALS410 3 3 0.18 0.06 0.76 ALS414 1 1 1.00 0.00 0.00 ALS432 1 1 1.00 0.00 0.00 ALS439 2 2 0.18 0.60 0.22 ALS487 1 1 1.00 0.00 0.00 ALS502 3 3 0.25 0.00 0.75 ALS558 3 3 0.09 0.01 0.90 ALS613 2 1 0.84 0.13 0.03 ALS628 1 1 0.99 0.00 0.01 ALS633 3 3 0.37 0.00 0.63 ALS634 1 1 1.00 0.00 0.00

Bibliography

1. DeVeaux SA, Ogle ME, Vyshnya S, Chiappa NF, Leitmann B, Rudy R, et al. Characterizing human mesenchymal stromal cells' immune-modulatory potency using targeted lipidomic profiling of sphingolipids. Godoy JAP, Paiva RMA, Souza AM, Kondo AT, Kutner JM, Okamoto OK. Clinical Translation of Mesenchymal Stromal Cell Therapy for Graft Versus Host Disease. Dev Biol Cell Front. 2019;7(November): 1-14.

2. Stroncek DF, Jin P, McKenna DH, Takanashi M, Fontaine MJ, Pati S, et al. Human Mesenchymal Stromal Cell (MSC) Characteristics Vary Among Laboratories When Manufactured From the Same Source Material: A Report by the Cellular Therapy Team of the Biomedical Excellence for Safer Transfusion (BEST) Collaborative. Dev Biol Cell Front. 2020;8(June):1-13.

3. Chen Y-S. Mesenchymal Stern Cell: ConsiderationsforManufacturing and Clinical Trialson Cell Therapy Product. Int J Stern Cell Res Ther. 2016;3(1): 1-12.

4. Wilson A, Webster A, Genever P. Nomenclature and heterogeneity: Consequences for the use of mesenchymal stem cells in regenerative medicine. Regen Med. 2019;14(6):595-611.

5. Tomecka E, Lech W, Zychowicz M, Sarnowska A, Murzyn M, Oldak T, et al. Assessment of the neuroprotective and stemness properties of human Wharton's jelly-derived mesenchymal stem cells under variable (5% vs. 21%) aerobic conditions. Cells. 2021; 10(4).

6. Munoz-Perez E, Gonzalez-Pujana A, Igartua M, Santos-Vizcaino E, Hernandez RM. Mesenchymal stromal cell secretome for the treatment of immune-mediated inflammatory diseases: Latest trends in isolation, content optimization and delivery avenues. Pharmaceuticals. 2021; 13(11).

7. Viswanathan S, Shi Y, Galipeau J, Krampera M, Leblanc K, Martin I, et al. Mesenchymal stem versus stromal cells: International Society for Cell & Gene Therapy (ISCT®) Mesenchymal Stromal Cell committee position statement on nomenclature. Cytotherapy [Internet], 2019;21(10): 1019-24.

8. Garcia-Bernal D, Garcia-Arranz M, Yanez RM, Hervas-Salcedo R, Cortes A, FernandezGarcia M, et al. The Current Status of Mesenchymal Stromal Cells: Controversies, Unresolved Issues and Some Promising Solutions to Improve Their Therapeutic Efficacy. Front Cell Dev Biol. 2021;9(March):1-18.

9. Petrenko et al. (2020); A Comparative Analysis of Multipotent Mesenchymal Stromal Cells derived from Different Sources, with a Focus on Neuroregenerative Potential; Scientific Reports volume 10, Article number: 4290.

10. Li et al. (2014) Comprehensive characterization of four different populations of human mesenchymal stem cells as regards their immune properties, proliferation and differentiation; International Journal of Molecular Medicine.

11. Ouafy et al. (2020); Evaluation of the therapeutic potential of mesenchymal stromal cells during in vitro production: search for markers related to immunomodulatory properties; Cytotherapy, vol. 22, no. 5, supplement, 2020, pages 80-88.

12. Igor Matic et al., Expression of OCT-4 and SOX-2 in Bone Marrow-Derived Human Mesenchymal Stem Cells during Osteogenic Differentiation, Open Access Maced J Med. Sci. 2016 Mar 15; 4(1): 9-16.

13. Sei-Myoung Han et ah, Enhanced proliferation and differentiation of Oct4- and Sox2-overexpressing human adipose tissue mesenchymal stem cells, Exp Mol Med. 2014 Jun; 46(6): e101.

14. Aleksandra Musiał-Wysocka et al., Molecular and Functional Verification of Wharton’s Jelly Mesenchymal Stem Cells (WJ-MSCs) Pluripotency, Int J Mol Sci. 2019 April 12;20(8):1807.

15. Michael P. Gustafson et al., Comprehensive immune profiling reveals substantial immune system alterations in a subset of patients with amyotrophic lateral sclerosis, PLoS One. 2017 Jul 25;12(7):e0182002.

16. Yujun Xu et ah, SPION-MSCs enhance therapeutic efficacy in sepsis by regulating MSCexpressed TRAF1-dependent macrophage polarization. Stem Cell Res Ther. 2021 Oct 9;12(1):531.

17. Jiwon Oh et ah, Mesenchymal stem cells genetically engineered to express platelet-derived growth factor and heme oxygenase-1 ameliorate osteoarthritis in a canine model, J Orthop Surg Res. 2021 January 11; 16(1):43.

18. Dimitrios Mougiakakos et al., The effect of inflammatory licensing on heme oxygenase-1mediated induction of regulatory T cells by human mesenchymal stem cells, Blood. 2011 May 5; 117(18):4826-35.

19. Josse, J & Husson, F. (2013). Handling missing values in exploratory multivariate data analysis methods. Journal de la SFdS. 153 (2), pp. 79-99.

20. Miron B. Kursa, Witold R. Rudnicki (2010). Feature Selection with the Boruta Package. Journal of Statistical Software, 36(11), pp. 1-13.

21. Biecek P (2018). “DALEX: Explainers for Complex Predictive Models in R.” Journal of Machine Learning Research, 19(84), 1-5.

22. Datta, Susmita, et al. “An adaptive optimal ensemble classifier via bagging and rank aggregation with applications to high dimensional data.” BMC Bioinformatics, vol. 11, 18 Aug. 2010, p. 427.

Claims (10)

1. A method for classifying MSC cells with respect to therapeutic properties, characterized in that it comprises the following steps: a. measurement of the level of heme oxygenase presence, POU5F1 expression level, change in the share of Th lymphocytes under the influence of MSC cells; change in the share of CD69-positive Th lymphocytes under the influence of MSC cells and change in the share of naive Th lymphocytes under the influence of MSC cells, wherein the change in the share of Th lymphocytes under the influence of MSC cells, change in the share of CD69-positive Th lymphocytes under the influence of MSC cells and change in the share of naive Th lymphocytes under the influence of MSC cells is determined after co-culture of MSC cells with mononuclear cells in comparison to the control consisting of mononuclear cells that were cultured in identical conditions but without contact with MSC cells, b. classification of MSC cells as belonging to one of at least three therapeutic clusters: a pro-regenerative therapeutic cluster with regenerative potential and differentiation potential, or a hyperactive therapeutic cluster that shows no therapeutic effect, or an anti-stress therapeutic cluster with the potential to reduce oxidative stress based on the results of the measurements from step a. 2. The method according to claim 1, characterized in that in step b - if the level of MSC cells showing the presence of heme oxygenase is from 2% to 3.9% and the level of POU5F1 expression is from 0.5 CF to 1.01 CF and the change in the share of CD69-positive Th lymphocytes under the influence of MSC cells is from -17.83% to -13% and the change in the share of naive Th lymphocytes is from -0.35% to -0.15% and the change in the share of Th lymphocytes under the influence of MSC cells is from 0.05% to 0.23%, the MSC cells are classified into the overactive cluster, - if the level of MSC cells showing the presence of heme oxygenase is from 11.4% to 17% and the level of POU5F1 expression is from 0.55 CF to 0.7 CF and the change in the share of CD69-positive Th lymphocytes under the influence of MSC cells is from -27.33% to -1.59% and the change in the share of naive Th lymphocytes is from -0.22% to 0.16% and the change in the share of Th lymphocytes under the influence of MSC cells is from 0.09% to 0.12%, the MSC cells are classified into the anti-stress cluster, - if the level of MSC cells expressing heme oxygenase is from 3.4% to 10.1% and the level of POU5F1 expression is from 0.84 CF to 1.4 CF and the change in the proportion of CD69-positive Th lymphocytes under the influence of MSC cells is from -10.6% to 0.59% and the change in the proportion of naive Th lymphocytes is from 0.08% to 0.31% and the change in the proportion of Th lymphocytes under the influence of MSC cells is from -0.02% to 0.09%, the MSC cells are classified into the pro-regenerative cluster. 3. The method according to claim 1 or 2, characterized in that in step a the change in the share of Th lymphocytes under the influence of MSC cells, the change in the share of CD69-positive Th lymphocytes under the influence of MSC cells and the change in the share of naive Th lymphocytes under the influence of MSC cells is determined after co-culture of MSC cells with mononuclear cells in a quantitative ratio of 1:10, respectively. 4. The method according to claim 3, characterized in that the co-culture is carried out at a temperature of 37°C and in a 5% CO2 atmosphere. 5. The method according to claim 3, characterized in that the co-cultivation is carried out for 4 days. 6. The method according to claim 1, characterized in that the MSC cells are derived from Wharton's jelly or from bone marrow or from adipose tissue, preferably from Wharton's jelly. 7. The method according to claim 1, characterized in that step b. is computer-aided and comprises the following steps of building a classifier: a1. determining at least three therapeutic clusters: a pro-regenerative therapeutic cluster with regenerative potential and differentiation potential, or a hyperactive therapeutic cluster that does not demonstrate a therapeutic effect, or an anti-stress therapeutic cluster with the potential to reduce oxidative stress, wherein the determination comprises measuring MSC cell parameters comprising: i. CFU assay ii. determination of the teratogenic potential of MSC cells by measuring the expression of genetic markers CDKN2A, CLDN1, DNMT3B, ESM1, HAND2, HEY1, MYO3B, NANOG, PDGFRA, POU5F1, TDGF1, TWIST1, ZBTB16 iii. determination of the phenotype of MSC cells by measuring the genetic markers CD119, CD120b, CD274, CD273, CD200, CD72, CD39, HO iv. analysis of the secretome of MSC cells by measuring the secretion of BDNF, HGF, IL6, IL8, MCP1, RANTES, TGFb1, TGFb2 v. determining the immunomodulation of MSC cells by co-cultivating MSC cells with mononuclear cells, wherein the co-culture is preferably carried out in a quantitative ratio of 1:10, at a temperature of 37°C and in a 5% CO2 atmosphere for 4 days, wherein after the co-culture the number of mononuclear cells, their viability and the change in the % share of T lymphocytes, B lymphocytes, NK, NKT, Th, Tc, naive Th, memory Th, naive Tc, memory Tc, Th CD69+, Th CD25+CD69+, Th CD25+, non-activated Th, Tc CD69+, Tc ‘CD25+Cd69+, Tc CD25+, non-activated Tc, Th1, Th2, Th17, internal cytokines for: Th1, Th2 and Th17, and T regulatory mononuclear cells are determined vi. division of MSC cells into at least three therapeutic clusters: hyperactive, anti-stress and pro-regenerative based on the results from stages i.-v., b1. building a classifier, which includes: i. reducing the number of parameters to at least 5 parameters by identifying significant parameters based on classification error coefficients or on the basis of a measure of variable significance based on a loss function, ii. performing correlation of parameters from stages a1. i.-v. and eliminating parameters that are correlated with each other, iii. determining the probability of membership in a specific cluster on the basis of comparing the classification results using the parameters selected in stage b1. i. with membership in the designated cluster in stage a1. vi. and the steps of MSC classification, including: c1. providing MSC cells, d1. measurement of selected significant parameters determined in step b1. i. of the MSC cells from step c1., e 1. classification of the MSC cells using the classifier determined in step b1. as belonging to one of the therapeutic clusters on the basis of the measurement results from step d1./ wherein steps a1. and b1. are performed once for a given population of MSC cells, step a1. vi., b1., e1. are performed by means of calculations in the computer system to which the data from the measurements in step a1. i.-v. or d1 are transferred. 8. The method according to claim 7, characterized in that in step b1. i. only those parameters are selected that have been indicated as significant. 9. The method according to claim 7, characterized in that in step a1. vi. the division into clusters is performed using hierarchical cluster analysis. 10. The method according to claim 7, characterized in that in step b1. i. the identification is performed using the Boruta algorithm with a random forest classifier.