Immunomodulatory Effects of a M2-Conditioned Medium (PRS^® CK STORM): Theory on the Possible Complex Mechanism of Action through Anti-Inflammatory Modulation of the TLR System and the Purinergic System

Abstract

Co-culture of primary or mesenchymal stem cells (MSC) with M2 macrophages produces a very special conditioned medium with a recognizable and stable cytokine pattern (PRS CK STORM), independent of the donor, with unique anti-inflammatory properties. This product can regulate certain pathways of inflammation in an anti-inflammatory manner, including TLR3, TLR4, the inflammasome, and the purinergic system. The anti-inflammatory action of PRS CK STORM is demonstrated both by its composition and by its action in in vitro and in vivo inflammatory models. The study of the mechanism of action showed changes in the pattern of toll-like receptors (TLR) and purinergic receptors, with an increase in the relative expression of mRNA encoding A2a and A3 receptors, together with a decrease in the relative expression of mRNA encoding P2X7 receptors. Second, it mitigated the adverse effects of a systemic inflammatory process in mice, especially in comparison with a known anti-inflammatory drug (Anakinra). Thus, due to its profile in terms of biosafety and efficacy, PRS CK STORM may be a strong candidate to treat inflammatory processes, such as cytokine storm associated with severe infectious processes, including COVID-19.

1. Introduction

Inflammation is the response of an organism’s immune system to damage caused to its cells and vascularized tissues by bacterial pathogens and by any other biological, chemical, physical, or mechanical aggressor. Such an inflammatory response must be self-limiting in time and intensity since, if this is not the case and if there is no perfect coordination between the innate and adaptive immune systems, a severe systemic inflammatory syndrome with positive feedback systems will occur, eventually causing a cytokine storm that can lead to multi-organ failure [1, 2]. In the establishment, maintenance and termination of this cytokine storm, at the molecular level, in cases of sepsis and severe viral infections such as that associated with COVID-19, the toll-like receptors (TLRs), NOD-like receptors (NLRs), and RIG-like helicase receptors (RLRs), cytokines, chemokines and growth factors, and the purinergic system will be fundamental in the establishment, maintenance, and termination of this cytokine storm.

In the late 1990s, the ability of infectious agents (bacteria, viruses, zoonoses, or parasitic and/or fungal infections) to trigger cytokine storm syndrome (CSS) was first described with the recognition of a case series of hemophagocytic lymphohistiocytosis (HLH) of viral origin [3]. Such a cytokine storm is basically characterized by an exaggerated production of proinflammatory and profibrotic soluble mediators (especially IL-1β, IL-6, and TNF-α), together with an aberrant immunopathological reaction, involving an uncoordination between the innate and adaptive immunity system, there being generally an overactivation of the innate immune system, the main cellular actors being macrophages, dendritic cells, monocytes, neutrophils, and T lymphocytes [4, 5, 6, 7]. As a consequence of this cytokine storm, a situation of multi-organ hyperinflammation will be provoked, which usually affects mainly the lung and pancreas, among other organs, and which usually results in acute respiratory distress syndrome (ARDS) and/or acute lung injury (ALI), which can lead to multi-organ failure.

Although the association of increased levels of proinflammatory and profibrotic cytokines and chemokines with increased levels of morbidity and mortality following an infectious process is well known, we still lack a suitable drug to treat the cytokine storm [8].

The innate immune system is able to recognize molecular structures specific to viruses, bacteria, fungi, and other pathogens; these structures are known as pathogen-associated molecular patterns (PAMPSs) [9, 10, 11]. PAMPSs are small-molecule sequences that are repeated in groups of pathogens recognized by the so-called pattern recognition receptors (PRRs). These include the toll-like receptors (TLRs) family of membrane receptors, NOD-like receptors (NLRs) and RIG-like helicase receptors (RLRs), among which the NLRP receptors stand out, oligomeric structures called inflammasomes, responsible for generating the mechanism of pyropoptosis by hyperproduction of hyperinflammatory cytokines, used as a trigger for the hyperproduction of IL-1β and IL-18 [12]. These molecular patterns are essential for the recognition of microorganisms by innate immunity cells, which respond differently depending on the microorganism identified [9, 10, 11].

Analyzing the detection capabilities of all these receptors, both DAMPS and PAMPS, we conclude that the main receptors involved in innate immunity against infections are TLR2, TLR3, TLR4, TLR7, TLR9, NOD1, NOD2, RIGI, and NLRP3. In Table 1, we summarize the PAMPS and DAMPS that are able to activate them [13, 14, 15, 16, 17].

Table 1.

TLR 2 receptors are activated mainly by bacteria and fungi, and although their activation has been described in SARS COV-2 infection, it is possible that this is due not directly to the SARS COV-2 virus but to the existence of a concomitant bacterial infection.

TLR 3 receptors are activated mainly by viruses, although SARS COV-2 is not activated to the same extent as other viruses, such as influenza virus, respiratory syncytial virus (RSV), or herpes virus. TLR 4 receptors are activated mainly by bacteria, but activation is also seen in viruses, although it is unclear whether this is due to a posteriori DAMPS. TLR 7 receptors are mainly activated by viruses. TLR 9 receptors are activated by both RNA and DNA so that viruses, bacteria, fungi, or any pathogen can activate them. Cytoplasmic receptors, since they detect most pathogen fragments, including DNA and RNA, can be activated by both viruses and bacteria, although NOD1 is activated more strongly by bacteria. It appears that RIG I is only activated by viral and not bacterial RNA [13, 14, 15, 16, 17]. (++ maximal activation, + intermediate or mild activation, no activation).

Considering the different receptors involved in cytokine storms associated with infectious processes, we can deduce that the activation of the transcription factors AP-1 (activator protein 1) and NF-kβ (nuclear factor kappa light chain enhancer of activated B cells), both common denominators in almost all pattern recognition pathways, will provoke the dreaded cytokine storm, resulting in a state of generalized hyperinflammation. The most affected organs are lung or pancreas, with the consequent associated fibrotic reaction, producing irreparable anatomopathological damage with loss of function in the most affected organs.

A fact especially associated with the cytokine storm associated with SARS COV-2 is that a decrease in the production of type I interferons is observed, which causes dysregulation in the coordination of the innate and adaptive immunity systems, facilitating the appearance of the dreaded severe pneumonia that on many occasions determines the patient’s admission to the ICU [18].

It is very difficult to explain the existence of a cytokine storm by the activation of a single receptor. If this were so, treatment of the cytokine storm by a single monoclonal antibody, for example, a monoclonal antibody against IL-1β or against IL-6, would always be effective, and this we know almost never occurs. Moreover, even if in the first instance only one of the receptors is activated, the simple initiation of its metabolic cascade will provoke the appearance of DAMPS that will stimulate other receptors. If we add to this the fact that in the majority of cytokine storms associated with infections we do not see a single causative pathogen, but rather a group of them, we will understand that there is almost always a joint activation of several of these receptors, producing between them phenomena of agonism and synergy, as well as antagonism [19]. Any of them can have agonistic relationships with others, if they are stimulated at the same time. However, if these same receptors are activated with a significant time lag between them of hours or even days, the most likely mutual relationship they will establish will be one of antagonism [19]. Thus, the types of cytokines and chemokines that will be released as a result of the activation of the different receptors will depend on the sets of receptors that are primarily activated by PAMPS and, once initiated, such release of pro-inflammatory and profibrotic mediators will be prolonged and augmented over time by positive feedback from the same receptors or even the addition of others, by the stimuli elicited by DAMPS, which could lead to reactive phenomena even autoimmunity.

In the cytokine storm, we must also consider the intervention of the purinergic system [20, 21, 22]. Extracellular adenosine triphosphate (eATP) or its enzymatic degradation products, such as ADP, AMP, and adenosine, can stimulate a number of membrane receptors [23]. More specifically, the P2X7 receptor is widely distributed on innate cells of the adaptive immune system, a system that constitutes the first line of defense against invading pathogens. These cells are lymphocytes, granulocytes, macrophages (including microglia), and dendritic cells in peripheral tissues [24, 25, 26]. Activation of the P2X7 receptor has been associated with the establishment and prolongation of inflammation and cytokine storm in septic infections, including SARS-COV-2 infection [27, 28, 29]. The stimulation of the P2X7 receptor by adenosine triphosphate (ATP) causes the activation of the NLRP3 inflammasome, and consequently of caspase 1, stimulating this the exaggerated secretion of IL-1β and IL-18 [30]. For all these reasons, the ideal immunomodulatory treatment of the cytokine storm associated with moderate and severe infections should include the P2X7 receptor (generating antagonism) or P1-like receptors (generating agonism) as a therapeutic target [29].

The treatments tested to date to control cytokine storms associated with infectious processes have been based on the use of monoclonal antibodies used alone or in combination. The hypothesis put forward by our group proposes as a treatment a biological therapy based on the use of allogeneic-conditioned medium derived from M2-type macrophages and enriched with mesenchymal stem cells (MSCs). Mesenchymal stem cells, placed in co-culture with macrophages, not only respond to macrophages and adjust their secretome accordingly but also induce macrophages to respond to them, creating a feedback loop that contributes to immune regulation [31]. In the complex composition of this conditioned medium are present all growth factors, cytokines, and chemokines that are naturally produced by M2-type macrophages and MSCs, associated with innate immunity respecting natural pleiotropic relationships. The immunomodulatory cytokine profile of this medium confers a potent anti-inflammatory and anti-fibrotic action, and even, thanks to the results obtained with the secretome of MSCs on macrophages stimulated with TLR7/8 ligand, possibly antiviral [32]. Different studies have shown that the secretome of these two cell types is modified and modulated when co-cultures of these cells are performed [33]. The immunomodulation mechanisms mediated by MSCs are due, among other factors, to the release of PGE-2 (prostaglandin E-2) and TSG-6 (TNF-stimulated gene 6 protein) [34]. M2 (anti-inflammatory) macrophages secrete high levels of IL-10 and low levels of IL-12p70 and IL-17, in a process that is directly mediated by other factors produced by MSCs (such as IL-6 and HGF) [35, 36]. It has been experimentally demonstrated that factors secreted by pro- or anti-inflammatory macrophages activate the immunomodulatory potential of MSCs. In this regard, IL-10 release by anti-inflammatory macrophages activates MSCs to release PGE-2 [37], which in turn modulates macrophages producing a cascade of additive molecular interactions in favor of immunotolerant and anti-inflammatory mechanisms. Basically, the polarization of macrophages to M2 type with immunoregulatory phenotype will be maintained [38], which, in turn, will express more IL-6, IL-10, and IGF-1 and inhibit their production of IL-12 and TNF-α. MSCs are also capable, through secretion of these same factors, of inhibiting the migration, maturation, and differentiation of dendritic cells [39, 40, 41]. Similarly, monocyte-derived M2 macrophages co-cultured with MSCs have been shown to increase mitochondrial function and ATP turnover, both in vitro and in vivo, resulting in an increase in the ADP/ATP ratio [42]. In addition, MSCs maintain ATPase and CD73 enzymatic activities on their surface, converting ATP to ADP and AMP to adenosine, respectively [43]. Adenosine, the last molecule in these reactions, has immunoregulatory functions through the activation of the P1 receptor [44]. Importantly, activation of monocyte P1 receptors, such as A2A and A2B, inhibits TNF-α production [44].

Several previous experiences demonstrate how secretomes from both cell types possess immunomodulatory properties. For example, direct injection of the supernatant of cultured mesenchymal stem cells (MSCs) containing a variety of growth factors, prostaglandins, and cytokines can be applied to the treatment of kidney injury [45]. Both co-culture with M2 macrophages and treatment with M2 macrophage supernatant have also been shown to increase endothelial cell viability in a bacterial lipopolysaccharide-generated lung sepsis model [46]. In addition, the efficacy and safety of multiple sclerosis treatment by intravenous infusion of conditioned medium from mesenchymal stem cell culture have also been demonstrated [47].

The advantage of using the complete conditioned medium versus one of its purified components lies in the synergistic mechanism between its different components [48], the result of subjecting the cell populations to a culture that, in vitro, attempts to emulate the anti-inflammatory, anti-fibrotic, and regenerative immunomodulatory microenvironment that occurs in vivo in diseased tissue.

2. Material and methods

Production and characterization of allogeneic-conditioned medium derived from M2-type macrophages and enriched with MSCs.

First, to obtain MSCs, a lipoaspirate sample was obtained from which the stromal vascular fraction (SVF) was extracted, following the protocol described by Lapuente et al. [49]. SVF was harvested by centrifugation under the same conditions as earlier-mentioned, seeded at a density of approximately 30,000 cells per cm² in 100-mm diameter culture plates (this and all culture plastic used was from Corning, Corning, NY, USA) and cultured at 37°C and 5% CO₂ in culture medium (DMEM + 10% fetal bovine serum (FBS) + 1% P/S). At 24 h, the culture was washed with phosphate-buffered saline (PBS) to remove nonadherent cells and the adherent cell population, called processed lipoaspirate (PLA), was cultured to subconfluence under the same conditions as earlier-mentioned, changing the culture medium three times a week and performing the necessary passages with trypsin 0.05% (Gibco), until a homogeneous population of mesenchymal-type stromal cells, also called mesenchymal stem cells (MSCs), was obtained. After culture, the cells were frozen at a freezing ramp of −0.5°C/min to −80°C in freezing medium composed of 10% dimethyl sulfoxide (DMSO, Sigma) in FBS or culture medium, then immersed in liquid N₂ and maintained until use.

Secondly, monocytes were isolated from one altruistic blood donation bag of 450 ml with 12% citrate–phosphate-dextrose (Grifols, Barcelona, Spain) from the blood bank of the Fuenlabrada Universitary Hospital. To isolate the leukocytes, each bag was divided into 50-ml tubes (Corning) and centrifuged at 1500 x g for 10 min at room temperature (RT). The intermediate band, leukocyte buffy coat, was collected and deposited on a clean tube. Immediately, 24 ml of this concentrate was carefully placed on 18 ml of Ficoll Histopaque 1077 (Sigma) and centrifuged at 400 × g for 30 min at room temperature (RT) and without brake. The mononuclear cell band was collected, and after adding PBS in a 1:1 ratio, centrifuged at 300 x g for 5 min at RT. The supernatant was discarded and the resulting pellet was resuspended in a fivefold volume of erythrocyte lysis buffer and incubated at RT for 10 min. Subsequently, a 10-fold volume of PBS was added and centrifuged under the same conditions as earlier-mentioned to obtain the cell pellet after discarding the supernatant. This last wash was repeated once more and, after this, the resulting peripheral blood mononuclear cell pellet (PBMC) was resuspended in CTS-AIM-V medium (Gibco) supplemented with 0.1% Dipeptiven 200 mg/ml (Frenesius Kabi Austria GmbH, Graz, Austria) and cultured in T-175 culture flasks (Corning, approximately 200 million PBMCs per flask) at 37°C and 5% CO₂ atmosphere for 90 min. The next step was to wash the flasks twice with plenty of PBS to remove unattached cells. The cells were immediately lifted with a cell scraper (Corning) to obtain a cell suspension in PBS, which was centrifuged for 5 min at 300 × g at room temperature. The resulting pellet was resuspended in AIMV + Dipeptiven + 10 ng/ml M-CSF (R&D Systems, McKinley, Minneapolis, USA) for co-culture. All cell counts and viabilities (trypan blue exclusion method) were performed with an automatic counter TC20 (BioRad, Hercules, CA, USA), strictly following the manufacturer’s instructions, marking a lower threshold of 8 μm to disregard possible erythrocytes, platelets, and other contaminating cellular debris.

Third, co-culture was established to produce the conditioned medium. For this purpose, the obtained monocytes were seeded at a density of 500,000 cells/cm² in inserts (Transwels, with a polyethersulfone membrane of 1 μm pore size, from Corning) of 6-well plates and cultured under standard conditions for 4 days with the described medium. When the culture medium was removed, the inserts were washed twice with PBS and added to the plates on which the MSCs had been seeded and cultured in pass 4 (24 h earlier, in Corning 6-well plates at a density of 10,000 cells/cm² under standard conditions), previously washed twice with copious PBS and using CTS-AIMV-V medium supplemented with 0.1% Dipeptiven to maintain the co-culture under standard conditions of temperature and CO₂ concentration. The co-culture was maintained for 4 weeks by collecting the conditioned medium and adding fresh medium twice a week. To preserve the different collections, they were immediately frozen by immersion in carbonic snow and kept at −80°C until analysis. To perform the analysis, the cultures were thawed at 4°C and analyzed immediately after filtering through a 0.45-μm pore nitrocellulose filter (Merck KGaA, Darmstadt, Germany). To obtain secretome controls for MSCs and different monocytes, the different cell populations were cultured separately under the conditions described for co-culture.

Subsequently, phenotypic characterization of MSCs and monocytes was performed and samples were taken from both populations at time 0, 7 days, 14 days, and 28 days. MSCs were lifted with trypsin and monocytes with scraper as described earlier and, after centrifugation at 300 × g 5 min at 4°C, resuspended each cell type in PBS, permeabilized the monocytes with Perm/Wash buffer (BD Biosciences, Franklin Lakes, NJ, USA), and incubated the cells for 15 min at RT and in the dark with the following fluorochrome-conjugated antibodies (and their related isotypes as negative controls) at 1:50 concentration: CD73-APC, CD90-APC, CD45-FITC, HLA-DR-FITC, CD31-PE, CD68-FITC, and CD163-PE (all from BD Biosciences). The fluorescence minus one technique was used to adjust the voltages and compensate for fluorescence, and propidium iodide (Sigma) was used to determine dead cells according to the manufacturer’s instructions. A Guava EasyCyte flow cytometer (Merck) was used to acquire the samples and InCyte software (Merck) was used to analyze the results.

To quantify the secretome of both cell types and the co-culture, 30 growth factors, cytokines, and chemokines were quantified using either ELISA or Multiplex assay (ProcartaPlex 23 PLEX, Invitrogen, Grand Island, NY, USA), strictly following the manufacturer’s instructions. A Luminex Labscan 100 plate reader (Luminex Corporation, Austin, TX, USA) was used for the determinations. The molecules quantified by Multiplex were the following: MIP1-α, IL-2, IL-6, TIMP-1, IL-8, IL-10, IL-12 P70, IL-1 RA, RANTES, GM-CSF, leptin, HGF, MMP-3, MCP1, BNGF, EGF, adiponectin, TNF-α, MMP-1, TRAIL, FGF-2, PDGF-BB, and VEGF-A. For quantification of IGF-1, BMP-6, IL-1β, IL-4, TGF-β1, TGF-β3, and VEGF-C, a double sandwich ELISA technique was used following the manufacturer’s instructions (DuoSet ELISA kit, R&D) and quantification was determined using an iMark plate reader (BioRad).

2.1 Generation of the in vitroinflammation model

THP-1 cell line culture and subsequent differentiation to macrophages were performed to generate in vitro models of biosafety and efficacy. THP-1 monocytic cells (CellLineService, cat. No.: 300356) were cultured and expanded using RPMI 1640 (Lonza, Basile, Switzerland) supplemented with 10% fetal bovine serum (FBS) (Corning, NY, USA), 1% penicillin/streptomycin (P/S) (Lonza), 1 mM sodium pyruvate (Lonza), and 1% MEM nonessential amino acids (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) starting now THP-1 medium. Cells were maintained at a density ranging from 2.5 × 10⁵ to 10⁶ cells/ml to ensure adequate growth and a stable phenotype. Forty-eight hours prior to lipopolysaccharide (LPS) stimuli, cells were differentiated into resting macrophages using phorbol 12-myristate 13-acetate (PMA) (Sigma–Aldrich, Saint Louis, MO, USA) at 5 ng/ml in THP-1 medium as described in the protocol used by Park et al. [50]. After this differentiation process, the cells were used for our experiments. All cell cultures were maintained at 37°C in an atmosphere of 5% CO₂ and 98% relative humidity. The in vitro inflammation model was generated by differentiating 400,000 THP-1/ml cells in exponential growth phase into resting macrophages in 12-well plates (Nunc, ThermoFisher) (final volume 1 ml) and after 48-h pretreatment with PMA, the cells were washed three times with 0.5 ml of tempered THP-1 medium without PMA and allowed to incubate for 30 min before LPS stimuli. Once at rest, rest, the cells were treated with 10 ng/ml LPS (Sigma–Aldrich) in RPMI 1640 medium and the investigational product, which had been previously quenched at room temperature or quenched THP-1 medium, using as control the same THP-1 culture treated with the same amount of LPS, but adding in this case 10 μg/ml hydrocortisone. The final volume of each well was 1 ml with 400,000 cells each. Stimulation was carried out for 5 h.

2.2 Evaluation of the possible mechanism of action of the proposed conditioned medium

After 5 h of stimulation, supernatants were removed from each well, divided into aliquots, and flash-frozen by immersion in dry ice for further analysis. Total RNA was extracted from the cells using an RNeasy Plus Mini kit (Qiagen, Hilden, Germany), and extraction was carried out strictly according to the manufacturer’s instructions. This kit included a genomic DNA removal step. The resulting RNA was eluted from the columns using nuclease-free water, divided into aliquots, and stored at −80°C to avoid degradation by environmental RNAases. From the 40 μl of RNA solution from each sample, an aliquot was extracted to assess RNA integrity and concentration. Total RNA integrity was assessed by agarose gel electrophoretic run of total RNA on a 2% agarose-TBE gel for 30 min at 120 V and 400 mAh. Quantification of total RNA was performed by a fluorimetric method using a highly sensitive fluorimetric kit (Qubit HS RNA quantification kit, Applied Biosystems, ThermoFisher). The cDNA was synthesized from total RNA for quantitative PCR of our genes of interest. A high-capacity cDNA reverse transcription kit (Applied Biosystems) was used for synthesis, and a total of 150-ng total RNA was used, for each synthesis reaction. Each sample had 5 cDNA synthesis reactions to achieve sufficient volume for downstream applications. The synthesis protocol was performed using an RNAase inhibitor following the manufacturers’ recommendations, and their protocol was strictly followed. Random hexamers were used to perform reverse transcription of all mRNAs into double-stranded cDNA. After synthesis, the cDNAs were divided into aliquots and stored at −20°C, for later use. An aliquot of these cDNAs was extracted for quantification using a Qubit dsDNA HS Assay kit (Applied Biosystems). Primer concentrations were optimized using a cDNA pool to determine the most appropriate concentrations of the primers in the qPCR protocol. For such determination, a standard PCR was performed using a 2× PCR MasterMix (DreamTaq HotStart PCR MasterMix) (Applied Biosystems). Cycling conditions were 98°C for 3 min, then 35 cycles at 95°C for 45 s, 60°C for 30 s, and 72°C for 30 s. After these 35 cycles, the temperature was set at 72°C for 5 min and then held at 4°C indefinitely. The optimal primer concentration was determined by selecting the sharpest specific bands on agarose electrophoresis, uncontaminated by the presence of primer dimers at the front of the gel or nonspecific products. Ideal primer concentrations were 250 nM for forward and reverse primers. Primer sequences and amplicon sizes are attached in Table 2.

Gene name	Forward 5′→3′	Reverse 3′→5′	Gene ID	Amplicon size (bp)
TLR-2	CCACCTGCCTGGAACTCAG	CAGTCACCTGAGAGAACGCC	7097	216
TLR-3	AACGACTGATGCTCCGAAGG	CAGGGTTTGCGTGTTTCCAG	7098	207
TLR-4	TAGCGAGCCACGCATTCACA	TTAGGACCACCTCCACGCAG	7099	165
TLR-7	CCCTGGCCACAGACAATCAT	TCCTGTGACAGACGTTGGTG	51,284	210
TRAF6	CCGCGCACTAGAACGAGCAA	GGCAGTTCCACCCACACTATC	7189	157
MyD88	ATGCCTTCATCTGCTATTGCCC	GGCCTTCTAGCCAACCTCTTT	4615	175
RIPK1	ACTAGGTGGCAGGGTACAG	TGATCATGAGTCCCTGGGTT	8737	202
IKKβ	TGGACGTGGTCACAGACGGA	CGAGGAACCACCATGTGAGA	3551	203
NF-kβ	TTAGGAGGGAGAGCCCAC	AGTCGGATCTGTGGTTGAAA	4790	241
CASP1	CAGTCACACAAGAAGGGAGG	CCCCTTTCGGAATAACGGAC	834	227
NLRP3	GCTGGCATCTGGATGAGGAA	AAAGTTCTCCTGTTGGCTCG	114,548	247
GAPDH	GCACCACCAACTGCTTAGC	GCATGGACCGTGGTCATGAG	2597	131

Table 2.

Primer sequences and amplicon sizes.

The following Thermo Fisher primers (coupled to FAM) were also used for A2a (Hs00169123_m1), A3 (Hs04194761_s1), and P2X7 (Hs00175721_m1) receptors.

Subsequently to perform qPCR, total RNA was extracted from 400,000 THP-1 cells using a Qiagen RNeasy plus mini kit (Qiagen, Hilden, Germany). THP-1 cells had been previously differentiated to resting macrophages using 5 ng/ml phorbol 12-mystirate 13-acetate (PMA) (Sigma–Aldrich, Saint Louis, MA, USA) in cell culture medium 48 h prior to the experimental model. A total of 0.75 μg of RNA was retrotranscribed to cDNA using a high-capacity cDNA synthesis kit (ThermoFisherScientific, Waltham, USA) employing random hexamers. For qPCR, 10 ng of cDNA per reaction was amplified using a Power SYBR-Green PCR master mix (Applied Biosystems, Thermo Fisher) on a StepOnePlus real-time PCR machine (Perkin Elmer, Waltham, MA, USA) and using the primers listed in Table 1 (primer table). Thermal cycling conditions included an initial denaturation step at 95°C for 5 min, followed by 40 cycles at 95°C for 30 s, 60°C for 30s, and 72°C for 30s. Melting curve analysis of each qPCR was performed on the final products. Messenger RNA fold changes were calculated using the ΔΔΔCt method with GAPDH as a calibrator gene.

Finally, in order to have an approximation of the mechanism of action of our PRS^® CK STORM conditioned medium, two studies were performed. Firstly, a quantification of the ATP/ADP ratio contained in the drug. The Sigma–Aldrich colorimetric ADP/ATP ratio assay kit (Ref: MAK135) was used for this purpose, and the Biorad iMark plate reader was used for its reading. Secondly, a quantification of extracellular ATP in THP-1 cells placed in culture, comparing the results when stimulated by LPS and/or treated with PRS^® CK STORM conditioned medium. For this purpose, the ELISA ATP Assay Kit Colorimetric (Ref: ab83355) from Abcam was used, and the Biorad iMark plate reader was used for reading.

2.3 Generation of the in vivoinflammation model

To perform the experimental model of acute lung injury, the experimental model described by Stephens et al. [51] was used. For this model, 8–10 weeks old male C57BL/6 mice were used and administered 5 mg/kg of bacterial lipopolysaccharide (LPS) in 50 μl of physiological solution retro-orbital under anesthesia. To decrease the possible suffering of the mice due to LPS, they were administered buprenorphine hydrochloride in water at the established dose of 0.056 mg/ml. A total of 25 animals were used, with a number of animals per group of 5. The mice were conditioned 1 week prior to the procedure and were housed in standard conditions with access to food and water ad libitum with 12-h light/dark cycles at a temperature of 25°C and humidity greater than 40% over the course of the project.

• Basal group – animals given no LPS or treatment/basal data (n = 5; mice number 1–5).

• Vehicle control – animals administered 50 μl of vehicle (n = 5; mice number 6–10).

• Positive control group – animals given LPS but no treatment, vehicle only (n = 5; mice number 11–15).

• Gold standard-treated animals – LPS-treated animals given gold standard – Anakinra treatment at a dose of 200 mg/kg/day orally (defined in Stephens et al. [51]). (n=5; mice number 16–20).

• PRS^® CK STORM-treated animals – LPS-treated animals given the drug intravenously (n = 5; mice number 21–25).

The vehicle used in the vehicle control group is PBS (phosphate-buffered saline). The gold standard treatment consists of the administration of Anakinra (IL-1Ra) before and after inoculation with LPS. The test group received the established dose of imatinib 24 h prior to LPS administration. The animals were administered the drug every 24 h starting from the retro-orbital administration of LPS. Treatment was administered intravenously (40 μl), upon generation of the model and then every 24 h thereafter. Blood samples were taken 24, 48, and 72 h after generation of the model from the submaxillary sinus from which 120 μl were collected for hemodynamic and general biochemical study, which was performed with the Comprehensive Diagnostic Profile protocol (#500–0038), of the VetScan V2 device (Abaxis). After 72 h from the generation of the model, the animals were sacrificed and exsanguinated and plasma samples were collected for final biochemical and cytokine analysis. Specifically, TNF-α, IL-1β, and IL-6 as proinflammatory cytokines and IL-10 as anti-inflammatory cytokine were quantified. They were analyzed by Luminex: MILLIPLEX MAP Mouse Cytokine/Chemokine Magnetic Bead Panel – Immunology Multiplex Assay (Cat: MCYTOMAG-70 K) (Merck). During the study, hyperthermia/hypothermia, respiratory distress, weight loss, food and water consumption, as well as the existence of other behavioral disorders were monitored. After euthanasia and subsequent necropsy, the major organs (heart, lung, liver, kidney, and spleen) were removed and fixed and preserved in 4% formalin for histological study. A small portion of each organ, prior to fixation, was preserved directly by immersion in dry ice to study inflammatory cytokine content in future assays. All animals underwent Irwin’s test every 24 h to obtain neuropharmacology data following the protocol of Mathiasen et al. [52].

2.4 Statistical analysis

The MTT and cytokine release assays, as well as the cytokine analysis of the culture supernatants, the biochemical values of the blood of the different groups of treated mice, and the values obtained in the qPCRs, were subjected to statistical analysis. For these, a two-tailed Student’s t-test was performed to obtain the p-value between the different experimental groups and to analyze the existence of significant differences (p < 0.05). Statistical analysis was performed with Excel (Microsoft, Albuquerque, NM, USA).

3. Results

3.1 Isolation of MSCs and monocytes

The yield achieved in the isolation of MSCs was approximately 1 × 10⁵ cells per ml of lipoaspirate, and it was necessary to incubate for 16 days under the conditions described in the previous section to bring the culture to pass 4 (Figure 1a). The yield provided by the monocyte donor can be seen in Table 3.

Age	Sex	Complete blood (millions/μl)			Buffy volumen (ml)	PBMCs (millions)	Monocytes (millions)
		Luekocytes	Red blood cells	Platelets
45	M	54.3	2.4	69	145	668.8	65.65

Table 3.

Data and yields obtained from monocyte donors.

Figure 1.

Percentage of positive cells for each antibody tested.

Optical microscopy showed that the cell morphology of the adherent cells in the cultures corresponded to that of monocytes/macrophages.

The results of flow cytometric characterization of MSCs and monocytes from three co-cultures at the times studied are shown in Figure 1. Phenotypic characterization of MSCs shows the classic phenotype of CD90 > 90%, CD73 > 90%, CD31 < 2%, CD45 < 2%, and HLA-DR < 2%. The marker CD68 is used as macrophage identifier, CD163 is mostly present on M2 type macrophages, and CD39 is expressed on macrophages/monocytes in co-culture with MSC.

3.2 Cytokine characterization of the conditioned medium

The results of quantification by both ELISA and Multiplex are detailed in Table 4.

CK		MIP1-α	IL-2	IL-6	TIMP-1	IL-8	IL-10
Monocytes	AVG	9.5643367	<8.2	<10.5	35777,0862	235.261369	3.06669833
	SD	3.57120227	NP	NP	48725,6036	106.055289	1.34471437
Control MSC	AVG	<2.3	<8.2	542.274142	>167500	9.30152294	0.7036029
Co-culture MSC/monocytes	AVG	6.91058459	<8.2	1153.79376	>167500	330.463609	2.58248776
	SD	4.45252385	NP	309.210936	NP	175.835377	0.53200545
CK		IL-12 P70	IL-1 RA	RANTES	GMCSF	Leptina	HGF
Monocytes	AVG	2.08256741	476452.06	1.23761123	<17.3	49.1307951	1748.29405
	SD	0.04597914	826813.74	0.27823114	NP	1.9033707	412.510477
Control MSC	AVG	0.70858087	<41.4	4.7416231	<17.3	<16.3	145.833433
Co-culture MSC/monocytes	AVG	1.68319311	223795.156	5.39009081	<17.3	64.0308586	1630.39368
	SD	0.30017417	208257.922	2.88381852	NP	49.0665959	456.036333
CK		MMP-3	MCP1	BNGF	EGF	Adiponectin	TNF-α
Monocytes	AVG	<16.5	2658.5485	<7.4	1.93098264	189.27	<5.9
	SD	NP	1672.2932	NP	NP	111.942	NP
Control MSC	AVG	137.490877	573.667894	<7.4	<1.2	<89.3	<5.9
Co-culture MSC/monocytes	AVG	450.455082	3275.6692	0.48007142	<2.2	173.5	<5.9
	SD	322.742794	1773.64942	0.19422895	NP	23.5476	NP
CK		MMP-1	TRAIL	FGF-2	PDGF-BB	VEGF-A	IGF-1
Monocytes	AVG	32.0031459	1.63442629	<3	71.944999	15.6273891	760.452793
	SD	41.5287708	1.1152979	NP	37.2898086	26.2176915	1099.67079
Control MSC	AVG	560.670891	<1.2	<3	83.1238663	1838.37263	717.691553
Co-culture MSC/monocytes	AVG	1364.87744	3.18084865	<3	127.82401	239.64569	1903.91696
	SD	335.663559	2.00494659	NP	80.6592455	399,735639	881.409224
CK		BMP-6	IL-1β	IL-4	TGF-β1	TGF-β3	VEGF-C
Monocytes	AVG	<156	<1	<10	<31.2	<31.2	<62.5
	SD	NP	NP	NP	NP	NP	NP
Control MSC	AVG	<156	<1	<10	<31.2	<31.2	<62.5
Co-culture MSC/monocytes	AVG	<156	<1	<10	<31.2	<31.2	<62.5
	SD	NP	NP	NP	NP	NP	NP

Table 4.

Mean values of the molecules studied.

Values are shown in picograms per milliliter. > and < indicate that the value is above or below the detection limits (respectively). SD: standard deviation; NP: not applicable.

To obtain the pattern (Figure 2), those molecules that could be quantified because they were within the detection limits of the method used in each case were studied, and statistically significant differences were sought with respect to the values taken as control (conditioned medium of M2-like monocytes/macrophages). Those values that were significantly different in all the samples studied (five samples in triplicate) were considered to form a specific and reproducible pattern of monocyte secretome modification by co-culture with MSCs. To test whether the pattern obtained was specific to the cell type co-cultured with monocytes, the expression of the same secretome molecules was obtained under the same conditions, but co-culturing monocytes with the following cell types: osteocytes, chondrocytes, tenocytes, synoviocytes, myocytes, lymphatic vascular cells, and Schwann cells, was compared. From this comparison, eight different and characteristic secretomes could be specifically differentiated, quantifying a minimum of seven molecules: IL-6, leptin, HGF, MMP-1, MMP-3, adiponectin, and VEGF-A (data not shown).

Figure 2.

Cytokine expression patterns. Shown are 16 mean values ± standard deviation of molecular characterization (MIP-1α, IL-6, IL-8, IL-10, IL-1Ra, RANTES, Leptine, HGF, MMP-3, MCP-1, Adiponectine, MMP-1, TRAIL, PDGF-BB, VEGF-A, IGF-1). (A) Secretome of monocytes; (B) secretome of co-culture. Stars mark values where there is a statistically significant difference (p < 0.05).

3.3 In vitro anti-inflammatory potency test

THP-1 cells after 48 h of stimulation with PMA gain adherence to the culture plastic and take on a macrophage-like morphology. After 24 h of culture, the number of cells in suspension decreases and the number of cells adherent to the plastic increases, a symptom of correct differentiation. After 48 h, about 90% of the cells are adherent to the plastic and are used in the experimental model.

The in vitro inflammation model used in this research is based on stimulating the proinflammatory action of macrophages when exposed to LPS. Figure 3 shows that the addition of our conditioned medium to the culture of THP-1 cells transformed into macrophages can reverse the effect of LPS on these macrophages, and a statistically significant difference can be observed, in addition to the difference observed with the use of soluble hydrocortisone. It can also be observed that the cells are sensitive to LPS stimuli at the concentration used.

Figure 3.

Graph a shows the results in terms of IL-1β release in the in vitro model of inflammation. Graph b shows the results in terms of TNF-α release in the in vitro model of inflammation. For these four studies, the results were analyzed in three independent experiments with two replicates of the analytical technique, for each type of experiment. Error bars indicate the standard deviation between samples. Asterisks (*) mark values where there is a statistically significant difference (p < 0.05).

4. Biosafety and in vivo efficacy tests

The results of the Irwin test are analyzed to evaluate at a general level the effect produced by the conditioned medium administered intravenously to mice, in which the cytokine storm model had been previously induced, by retrorbital injection of LPS. A slight decrease in temperature was observed in all LPS-injected groups. LPS administration was observed to induce changes in reflexes and behavior, such as hunching, piloerection, and tremors, which were increased throughout the 3-day trial. In contrast, exploratory activity, reaction to contact, and aggressiveness were slightly decreased. In all LPS-treated mice (with or without drug administration), mild diarrhea occurred, which in untreated or gold standard-treated mice resulted in more severe dehydration than in mice treated with the conditioned medium. In general, there is a trend that the symptoms caused by LPS administration are dissipated by treatment with our conditioned medium (PRS^® CK STORM), demonstrating a beneficial effect on the mice without counterproductive effects (Table 5).

Table 5.

Results of the Irwin test; it includes the parameters studied on the third day after generating the model. Mice 11, 18 and 22 died before completing the 3 days of the experiment [52].

The variations observed have as reference value the parameters 1 day before generating the model. Mice 11, 18, and 22 died before completing the 3 days of the experiment.

From the blood obtained after euthanasia and exsanguination of the animals, biochemical tests were performed to determine the biochemical profiles, which are shown in Table 6. The Mann–Whitney test was used for statistical analysis, considering statistical significance as*p < 0.05 in the case of statistically significant minor differences with respect to the baseline value and **p < 0.05 in the case of statistically significant major differences with respect to the baseline value.

			Albumin	Alanineaminotranferase	Amylase	Totalbilirrubin	Creatinine	Glucose	Total protein	Globulin
Control	0 h	Average	41	23	945	4	34	20.3	52	12
		SD	3.5	3.1	20.1	24	9.3	2.2	2.5	4.2
	24 h	Average	40	25.8**	946	7.0**	21.0*	22	55	15.0**
		SD	4.1	3.7	23.7	2.8	10.9	2.6	2.9	5
	48 h	Average	48.0**	30.0**	912	10.0**	22.0*	21.4	58.0**	10.0*
		SD	3	2.7	17.3	2.1	8	1.9	2.1	3.6
	72 h	Average	40	25	960	5.0**	27.0*	23.7**	52	13
		SD	3.1	2.8	17.8	2.1	8.2	1.9	2.2	3.7
PBS	0 h	Average	43	30	869	5	18	20.9	54	11
		SD	2.2	3.4	29.5	2.4	3.4	2.5	1.5	1.5
	24 h	Average	40	28	866	6.0**	24.0**	22.8	54	14.0**
		SD	2.5	4.3	34.9	2.8	4	3	1.8	1.8
	48 h	Average	38.0*	28	910	6.0**	25.0**	19.9	52	14.0**
		SD	1.9	2.9	25.5	2.1	2.9	2.2	1.3	1.3
	72 h	Average	39	29	924.5	5	25.0*	25.6**	51	12
		SD	1.9	3	162.3	2.1	3	2.2	1.3	1.3
LPS	0 h	Average	33	33	758	5	20	13.1	50	17
		SD	1.6	3	71.1	3	1.4	2	1.3	2.4
	24 h	Average	29.0*	26.0*	615.0*	5	15.9	10.5*	51	22.0**
		SD	1.9	3.5	83.9	3.6	1.7	2.3	1.5	2.8
	48 h	Average	31	18.0*	768	6.0**	14.5	9.0*	49	17
		SD	1.4	2.6	61.3	2.6	1.2	1.7	1.1	2.1
	72 h	Average	31	84.0**	740	7.0**	16	8.8*	48	18
		SD	1.4	2.6	63	2.7	1.3	1.8	1.1	2.1
LPS + gold standard	0 h	Average	32	21	727	5	16.1	11	50	18
		SD	1.8	8.9	199.5	2.2	3.1	1.1	1.5	0.5
	24 h	Average	30	26.4**	854.0**	4.0*	15.4	9.5*	48	18
		SD	2.2	10.5	135.6	2.6	3.7	1.2	1.8	0.6
	48 h	Average	33	15.0*	1118.0**	5	14.6	9.0*	51	18
		SD	1.6	7.7	171.8	1.9	2.7	0.9	1.3	0.4
	72 h	Average	29	36.0**	1278.0**	9.0**	16	8.6*	48	19
		SD	1.6	7.9	176.6	2	2.8	0.9	1.3	0.4
PRS^® CK STORM	0 h	Average	29.5	48.5	787.5	8	23.5	15	49	19
		SD	1	17.8	118	2.6	7.1	3.3	2	3.1
	24 h	Average	31.5	32.5*	681.5*	5.0*	29.0**	12.9*	49	17.0 *
		SD	1.2	2.1	139.2	3.1	8.3	3.9	2.4	3.6
	48 h	Average	30	36.0*	823	9.0**	21.5	15.3	50.5	19
		SD	0.9	1.5	101.6	2.3	6.1	2.8	1.7	2.6
	72 h	Average	38.5**	48	783.5	6.5*	21.5	14.5	52.5	18
		SD	0.9	1.6	104.5	2.3	6.3	2.9	1.8	2.7

Table 6.

Results of the biochemical analysis of the mice, including the parameters studied pretreatment and posttreatment at 24, 48, and 72 h.

The Mann–Whitney test was used for the statistics, considering the statistical significance as *p < 0.05 in the case of minor statistically significant differences with respect to the basal value and **p < 0.05 in the case of major statistically significant differences with respect to the basal value.

Figure 4 shows in percentage the relative variations observed at the posttreatment times (24, 48, and 72 h), with respect to the baseline values observed in each group.

Figure 4.

Concentration values of the different metabolites that make up the biochemical profile of the mice, expressed as a percentage with respect to the baseline value observed in each group.

The experimental model employed uses LPS as a causative agent of acute lung damage, causing a cytokine storm in the organism of mice like that produced by COVID-19 disease. We focused on the quantification of a small amount of these cytokines (TNF-α, IL-1β, IL-6, and IL-10) to evaluate the effect of the PRS^® CK STORM. Figure 5 shows the evolution of these cytokines detected in the sera of the mice on each of the days that the treatment lasted.

Figure 5.

Serum values of the different cytokines after 24, 48, and 72 h from the administration of the first treatment expressed in pg/ml as the mean of the values of the mice in each of the experimental groups. a: TNF-α. b: IL-1β. c: IL-6. d: IL-10. Values not shown are lower than the detection limit of the assay (2.8 pg/ml).

Histopathological analysis of the samples obtained from various organs of the mice obtained after the corresponding necropsies showed patchy lung thickening of the interstitium in a large part of the sample in the LPS treatment, while in the group treated with PRS^® CK STORM it was observed that there was no lung damage, as in the control groups. Slight affectations were observed in liver and spleen, which the drug was also able to reverse. As for the heart and kidney, no pathological findings were detected. Figure 6 shows examples of the lung sections studied in the different groups of the experiment.

Figure 6.

Pulmonary anatomopathological study in the different groups of mice in the experiment. It can be seen that the group treated with PRS^® CK STORM (our conditioned medium), the appearance of the lung is very similar to that shown by the untreated control group. All sections have been stained with hematoxylin eosin. ×1 images are taken in MO at 4× magnification and ×2 images are taken in MO at 40× magnification.

4.1 Mechanism of action study

In order to approach the mechanism of action of our conditioned medium PRS^® CK STORM, qPCR study is performed in order to analyze the effects of our complex biological drug under investigation on all common molecules involved in TLR2, TLR3, TLR4, TLR7, TLR8, TLR9, NOD1, NOD2, and NLRP3 pathways, TLR9, NOD1, NOD2, and NLRP3, and it is found that the conditioned medium downregulates the hyperactivity of these pathways, immunoregulating the key proteins involved in these pathways, being very remarkable the decrease of expression observed in TRAF6, caspase-1, RIPK1, IKKB, NF-kβ, MyD88, and NLRP3 proteins.

Following the method described in the corresponding section of this chapter, the total mRNA of the proteins named in the previous paragraph common to various pattern recognition pathways was extracted from THP-1m cells in culture used as control, from those stimulated with LPS, and from those stimulated with LPS and treated with PRS^® CK STORM, obtaining by qPCR the relative expression of the genes at the mRNA level normalized against GAPDH, the results of which are shown in Figure 7.

Figure 7.

qPCR study showing the results of relative gene expression of proteins common to various pathways related to pattern recognition in infections, at the mRNA level normalized against GAPDH, considering statistical significance as *p < 0.05.

Similarly and under the same experimental methodology, the total mRNA of the pattern recognition proteins linked to main infectious processes TLR-2, TLR-3, TLR-4, and TLR-7 was extracted from THP-1m cells in culture used as control, from those stimulated with LPS and from those stimulated with LPS and treated with PRS^® CK STORM, obtaining by qPCR the relative expression of the genes at mRNA level normalized against GAPDH, whose results are shown in Figure 8.

Figure 8.

qPCR study showing the results of relative gene expression of major pattern recognition receptor proteins in infections, at the mRNA level normalized against GAPDH, considering statistical significance as *p < 0.05.

In order to estimate the possible action of our PRS^® CK STORM conditioned medium through the purinergic system (Figure 9), the results of the ATP/ADP ratio obtained were analyzed comparatively among three groups with three replicates for each group, being the first group formed by THP-1m cells placed in culture alone, THP-1m cells stimulated with LPS at the doses described in Section 2, and the same THP-1m cells stimulated with LPS but in culture in the PRS^® CK STORM conditioned medium (Figure 9a). The extracellular ATP contained in the same three groups of cultures with three replicates for each group was quantified (Figure 9b). Finally, following the method described in the corresponding section of this chapter, the total mRNA of purinergic receptors A2a, A3, and P2X7 was extracted from THP-1m cells in culture used as control, from those stimulated with LPS, and from those stimulated with LPS and treated with PRS^® CK STORM, obtaining by qPCR the relative expression of the genes at the mRNA level normalized against GAPDH (Figure 9c).

Figure 9.

(a) ATP/ADP ratio analysis in three groups of cell cultures (THP-1m, THP-1m + LPS, and THP-1m + LPS + PRS^® CK STORM). Asterisks (*) mark values where there is a statistically significant difference (p < 0.05). (b) Quantitative analysis of extracellular ATP in three groups of cell cultures (THP-1m, THP-1m + LPS, and THP-1m + LPS + PRS^® CK STORM). Asterisks (*) mark values where there is a statistically significant difference (p < 0.05). (c) qPCR study showing the results of the relative expression of purinergic A2a, A3, and P2X7 receptor genes at the mRNA level normalized against GAPDH, considering statistical significance as *p < 0.05.

5. Discussion

In the complex composition of this conditioned medium (PRS^® CK STORM), all the growth factors, cytokines, and chemokines that are naturally produced by M2-type macrophages and MSCs, associated with innate immunity, are present, respecting the natural pleiotropic relationships, with an immunomodulatory cytokine profile from which a potent anti-inflammatory action is expected. In addition, according to the results obtained in test studies, the mechanism of action on TLR-7 receptors may possibly include some antiviral activity [32]. In fact, several experiments have shown that the secretomes of both cell types possess immunomodulatory properties. For example, direct injection of the supernatant of cultured mesenchymal stem cells (MSCs), which contains a variety of growth factors, prostaglandins and cytokines, can be applied to the treatment of kidney injury [45].

The theoretical advantage of using the complete conditioned medium versus some of its purified components lies in the synergistic mechanism between its different components [48], the result of subjecting the cell populations to a culture that, in vitro, attempts to emulate the immunomodulatory and regenerative microenvironment that occurs in vivo in diseased tissue. However, the main handicap of biological drugs with complex natural compositions will always be the variability between different batches and the practical impossibility of achieving a complete characterization, both qualitatively and quantitatively, as well as functionally [48]. Despite the earlier-mentioned, our group has been able to prove the existence of a stable cytokine pattern or fingerprint, which depends directly on the type of co-culture established and the conditions of the same and not on the donor of origin.

The anti-inflammatory capacity and potency shown by the PRS^® CK STORM conditioned medium has been remarkable, showing statistically significant reductions in in vitro tests on TNF-α and IL-1β levels, these differences being very similar to those obtained with hydrocortisone. This anti-inflammatory immunomodulatory capacity, we have also been able to verify in the in vivo model, generated in mice. The mice treated with our PRS^® CK STORM conditioned medium have shown normal behavior and the rest of the parameters analyzed in the Irwin test have been very similar to the control groups, where the cytokine storm had not been provoked. In fact, the comparative results in the in vivo test between the gold standard treatment used (Anakinra) and the PRS^® CK STORM were clearly more favorable to the latter. It should be noted that in the group treated with PRS^® CK STORM practically no pulmonary lesions were observed, while in the group treated with Anakinra inflammatory and fibrotic infiltrates were observed in a minimum of 30% of the surface of the sections. This coincides with the observation made at the time of sacrifice of the animals under study where both the control animals, in which the cytokine storm had not been provoked, and those treated with PRS^® CK STORM, took more than 2 min to die in the CO₂ chamber, while the animals treated with Anakinra died in 40 s and those not treated in about 30 s, and these times can be directly related to the anatomopathological state observed at the pulmonary level.

The results obtained throughout this experiment suggest that the drug PRS^® CK STORM is safe for intravenous administration, since no significant adverse effects have been observed in the different parameters analyzed, those found being mild. In addition, the drug significantly attenuates the detrimental effects caused by the cytokine storm associated with LPS administration. Most of the proteins and metabolites analyzed follow the same trend, regardless of the experimental group observed. With respect to albumin, the main protein present in blood, there is a decrease in albumin in all groups to which LPS was administered, which fits with that described by Ballmer et al. [53], with hypoalbuminemia occurring when the organism undergoes sepsis due to infection. Related to this, there is a decrease in total protein: a decrease in albumin will cause a decrease in total protein since the former is at very high concentrations. A decrease in blood glucose is also observed in all mice treated with LPS and LPS + gold standard. However, those treated with PRS^® CK STORM managed to normalize the levels of glycemia, albuminemia, and total proteinemia 3 days after receiving the first dose. This fact was directly related to the clinical improvement observed in these animals subjected to the experimental treatment, given that they felt better than the rest of the mice, ate better, physical activity was normalized, and diarrhea was corrected. On the other hand, a significant increase in alanine amino transferase (ALT), total bilirubin, and amylase was observed during the experiment, which could be related to reactive hepatitis [54]. This increase was observed in all groups stimulated with LPS and was maintained in those treated with gold standard. However, in those treated with PRS^® CK STORM, the figures normalized 72 h after the first treatment.

In all the groups where LPS was administered, there was a rapid increase in all the cytokines analyzed. From this, we can conclude that stimulation with bacterial lipopolysaccharide (LPS) is able to induce the expected inflammatory response, being more pronounced in more acute phase, the day after treatment administration and decreasing over time. In the control group and the vehicle, the presence of proinflammatory cytokines is not observed, which confirms that LPS is the cause of this response. However, in the group treated with LPS + gold standard, a lower increase of IL-1β is observed with respect to the rest of the groups where LPS was administered. This fact can be explained by the previous administration of the gold standard, recombinant IL-1Ra. However, despite the lower increase in this proinflammatory cytokine, the decrease in IL-1β finally achieved by the gold standard is even lower than that obtained with PRS^® CK STORM treatment. In general terms, our PRS^® CK STORM conditioned medium achieves greater control of all the cytokines analyzed in the experiment.

The results of the mechanism of action study show that PRS^® CK STORM is able to immunomodulate in an anti-inflammatory way the expression of TLR-like pattern recognition pathways especially associated with infectious processes, such as TLR-2, TLR-3, TLR-4, and TLR-7. From the same study, it can be deduced that this effect is not only localized to these receptors but also acts at the level of various proteins common to these and other pathways, such as TRAF6, RIPK1, and IKKB, with the decrease in expression observed in the proteins NF-kβ, MyD88, caspase-1, and NLRP3 being very significant.

Many published studies have demonstrated the importance of the purinergic system in the inflammation associated with the cytokine storm caused by moderate/severe infection, including COVID-19, and have shown that using various purinergic system receptors as a therapeutic target can limit the negative effects of the cytokine storm [21, 22, 55, 56, 57]. Extracellular ATP at high concentrations becomes a true alarmin [58], a potent proinflammatory signal capable of overexpressing and stimulating P2X-type purinergic receptors, especially P2X7R, located on various immune cells (neutrophils, eosinophils, monocytes, macrophages, mast cells, and lymphocytes) [59]. Extracellular adenosine triphosphate (eATP) is a well-characterized DAMP that modulates function and plasticity [60, 61]. This nucleotide can be released by stressed, injured, and dying cells or in response to TLR activation, reaching high concentrations in the extracellular medium [62].

In contrast, it has been shown that the balance between ATP and adenosine concentration is crucial in immune homeostasis. CD39 and CD73 are two ectonucleotidases that cooperate in the generation of extracellular adenosine by ATP hydrolysis, thus tipping the balance toward immunosuppressive microenvironments. Extracellular adenosine through A2A receptor stimulation has the ability to prevent activation and proliferation of both macrophages and T cells, thereby dramatically decreasing cytokine production [63].

From the results observed both in the cytometric characterization of M2 macrophages used in the co-culture to produce our conditioned medium, where a strong expression of CD39 and CD73 is observed, and from the study of the ATP/ADP ratio, where a clear increase of ADP is observed, we can deduce that one of the mechanisms of action of PRS^® CK STORM is probably linked to the process of dephosphorylation of extracellular ATP, which is degraded by ectonucleotidases to adenosine, and the latter interacts with adenosine receptors, type A2A and A3, producing an immunomodulatory anti-inflammatory effect on the cytokine storm. This theory is supported by two further pieces of evidence; firstly by the statistically significant decrease observed in LPS-stimulated THP-1 cells treated with our PRS^® CK STORM conditioned medium with respect to the levels observed in untreated LPS-stimulated THP-1 cells; and secondly by the combination of the observed reduction in the relative expression of P2X7 receptor mRNA with the observed increase in the relative expression of A3 and A2a receptor mRNA in LPS-stimulated THP-1 cells treated with PRS^® CK STORM relative to untreated THP-1 cells.

In all assays used in the in vitro and in vivo models, both employed LPS as an inducer of inflammation. Although the immunomodulatory mechanisms induced by bacterial antigens with respect to viral antigens in immune cells are different at the mechanistic level, the inflammatory response at the level of innate immunity ends up sharing numerous points in common [64, 65]. Therefore, the effect that PRS^® CK STORM has on these models of inflammation with LPS can be exportable to what can happen, at the level of immune response, during a viral infection. In fact, it has been shown that monocytes and macrophages stimulated with LPS and ATP increase the release of IL-1β [66, 67].

All these results observed in the study on the possible mechanism of action of our complete conditioned medium (PRS^® CK STORM) demonstrate that the immunomodulatory anti-inflammatory effect observed is a direct consequence of the action of various molecules contained in this medium, acting in a synergistic and pleiotropic manner on various therapeutic targets associated with different proinflammatory pathways, managing to downregulate the activation of the inflammasome, the inflammatory activation of the purinergic system, the activation of various pathways of pattern recognition associated with infections, etc., thus avoiding the possible feedback effects that therapeutic approaches based on the inhibition of a single receptor or a single inflammatory pathway may have.

6. Conclusions

The co-culture of M2 macrophages with MSCs allows the simple production of a complete conditioned medium (PRS^® CK STORM), which has a clear anti-inflammatory profile. In line with this characterization, it has been demonstrated that PRS^® CK STORM is able to stop macrophage overactivation in an in vitro inflammation model, through several mechanisms, including the expression pattern of TLR’s and the purinergic system. Likewise, the action capacity of this drug has also been demonstrated in vivo, by improving the symptomatology and tissue damage induced in mice in another model of inflammation. Therefore, PRS^® CK STORM is proposed as an effective and safe treatment to treat cytokine storms associated with moderate/severe infectious processes of any etiology, including that associated with COVID-19.