The authors have declared that no competing interests exist.
Conceived and designed the experiments: CEL. Performed the experiments: CEL CG ASG FJE VM AAS JMGV. Analyzed the data: CEL CB CG ASG FJE VM GF IM ML AAS JMGV. Contributed reagents/materials/analysis tools: CEL ASG FJE VM IM JMGV. Wrote the paper: CEL.
Human adipose stem cells (hASCs) play a crucial role in the fields of regenerative medicine and tissue engineering for different reasons: the abundance of adipose tissue, their easy harvesting, the ability to multipotent differentiation and the fact that they do not trigger allogeneic blood response or secrete cytokines that act as immunosuppressants. The vast majority of protocols use animal origin reagents, with the underlying risk of transmitting infections by non-human pathogens. We have designed a protocol to isolate and maintain the properties of hASCs avoiding xenogeneic reagents. These changes not only preserve hASCs morphology, but also increase cell proliferation and maintain their stem cell marker profile. On the other hand, human serum albumin (HSA), Tryple® and human Serum (HS), do not affect hASCs multipotent differentiation ability. The amendments introduced do not trigger modifications in the transcriptional profile of hASCs, alterations in key biochemical pathways or malignization. Thus, we have proven that it is possible to isolate and maintain hASCs avoiding animal reagents and, at the same time, preserving crucial culture parameters during long term culture. Thereby we have revealed a novel and effective tool for the improvement of clinical, cell-based therapies.
Mesenchymal stromal cells (MSCs) have been the most widely used in preclinical and clinical assays so far
A number of clinical trials involving the treatment of a variety of diseases hASC is already ongoing
One of the key points for their use in clinics regards on the cell isolation and expansion procedures. The vast majority of protocols use reagents derived from animal sources, with the underlying risk of transmitting infections by non-human pathogens
Use of bovine derivatives (serum as supplement for cell growth medium or serum albumin for cell washing saline solution) can result in bacterial, viral and prion infections
Human adipose tissue samples were obtained at private plastic surgery clinic (Clinica Dra. Isabel Moreno) from lipoaspiration procedures from 8 healthy patients under surgery by aesthetic or beauty reasons (two of them were man and the rest were woman), aged between 18 and 35, following written informed consent and ethical research project approval by both Clinica Dra Isabel Moreno and Principe Felipe Research Center (CIPF) ethical boards. All the patients were previously screened for human immunodeficiency virus (HIV), hepatitis C and other infectious diseases. Cells were obtained following the protocol established from Planat-Benard
Human serum (HS) group A was obtained from voluntary healthy blood donors at the Valencia Regional Transfusion Centre. Donors were screened for disease risk factors using a health history questionnaire and laboratory tests, fulfilling law requirements for blood donations. Human sera (non lipemic and negative for irregular antibody screening test) were pooled (80–100 donors), centrifuged (2900 g for 10 min) and stored at −30°C, until use.
We established four points of analysis per patient: isolation point, that is, immediately after the isolation procedure; passage 1 (P1), or human adipose adherent cells after their first passage (3–5 days in culture); passage 3 (P3), after 10–12 days in culture); and passage 5 (P5), following the fifth expansion around 20 days after the isolation. This organization scheme was used throughout the entire project for all experiments performed. The total number of patients used in this study was n = 8, from which three were used for the growth curve, proliferation assays and differentiation experiments. Samples from the 8 patients were used for the rest of studies.
To assess the proliferative capacity of hASCs, parallel cultures from three donors were analyzed through eleven serial passages using two different culture conditions: 6% HS and 10% fetal bovine serum (FBS). Cells were counted and subcultured every four days. At each passage, the population doubling (PD) rate was determined using the formula
A growth curve was carried out in parallel using hASCs from n = 3 donors, starting at passage 3. Two hundred cells per square centimeter were plated in P24 plates (Beckton Dickinson). Every day the cells from two wells were harvested and counted.
hASCs from the 8 patients were analyzed for a list of genes summarized on additional
For fine ultrastructural analysis, cells were cultured in chamber slides and then serially washed in a 0.1 M phosphate buffer (PB; pH 7.4) solution, prior to their fixation for Transmission Electron Microscopy (TEM). Fixation was performed in 3% glutaraldehyde solution in PB for 30 minutes at 37°C and postfixed in 2% OsO4 in PB. Dehydration was achieved by a graded series of ethanol solutions and a final rinse with propylene oxide (Lab Baker, Deventry, Holland). Finally, plates were embedded in araldite (Durkupan, Fluka) overnight. Following polymerization, embedded samples were detached from the chamber slide and glued to Araldite blocks. Serial semi-thin (1.5 µm) sections were cut with an Ultracut UC-6 (Leica, Heidelberg, Germany), mounted onto slides and finally stained with 1% toluidine blue. Ultrathin (0.07 µm) sections were prepared with the Ultracut and stained with lead citrate. Photomicrographs were obtained under a transmission electron microscope (FEI Tecnai Spirit G2), using a digital camera (Morada, Soft Imaging System,Olympus).
Differentiation assays into adipogenic and chondrogenic lineages were performed for hASCs from all culture conditions at passage 1, from a total of n = 3 donors. The assay was then repeated at passage 7 to determine whether cells differentiation capacities were retained throughout long-term culture, following protocols previously established
Hybridization, washing, staining, and scanning of the arrays was performed according to the manufacturer’s instructions (Agilent Technologies “one –color microarray-based gene expression Analysis, 5.5 Version). The array contained 41000 probes.
We used total RNA obtained from lipoaspirates of eight separate donors (aged 18–35 years). The samples obtained from cell cultures consisted in total RNA isolated from 1.6×106 cells. RNA was obtained using the RNeasy kit (Qiagen), according to manufacturer’s instructions, and treated with DNAse. RNA concentration was measured by spectroscopy using Nanodrop with an A260/A280 ratio of 1.8. Prior to performing the hybridization procedure Agilent p/n 5188–5977 “one-color microarray-based gene expression analysis”, we did a quality and quantity control assay of total RNA, using electrophoresis in an Agilent Bioanalizer. The RNA Integrity Number (RIN) was between 8.0 and 9.9 in all the samples used for hybridization. Total RNA was standardized among patients by pooling together the different samples for each of the established stages. We established four time points for the analysis (Isolation and passages 1, 3 and 5), using four primary culture RNA preparations from 4 different patients per time point. We repeated the hybridization twice per group.Microarrays from hASC xenogeneic free cells are deposited under GEO number GSE46314.
Statistical analysis was carried out using the R-project software
Statistically significant differences between groups were identified using the empirical Bayes method, also implemented as
Number ofgenes BH<0.05 | KEGG Pathway | Genes in KEGG pathway | Fisher exactp-value | |
Passage 1 versus Isolation | 63 | ECM-receptor interaction | COL1A2, FN1,SDC1 | 2.5·10−3 |
Biosynthesis of unsaturatedfatty acids | ELOVL6, FADS1 | 2.3·10−3 | ||
Passage 3 versus passage 1 | 111 | Piruvate metabolism | ACAT2, ALDH1B1, PDHB | 1.3·10−3 |
Butanoate metabolism | E2F5,SMAD2,ANAPC1,CDC16,CDNK2A, ORC4L,TFDP2 | 1.0·10−4 | ||
Cell cycle | ACAT2, ALDH1B1, GLO1, PDHB | 1.1·10−4 | ||
Passage 5 versus passage 3 | 209 | Cell division pathway | BBC3, CHEK1, FAS, THBS1 | 7.2·10−3 |
Ribosome | MRPL13, RPL22L1, RPS10, RPLP0 | 1.7·10−2 | ||
Pyruvate metabolism | ACAT2, ALDH7A1, ME1 | 1.0·10−2 | ||
Regulation of actin cytoskeleton | WASF1, CFL2, FGF2, PFN2, RRAS2 | 3.5·10−2 |
Number of genes BH<0.05 | KEGG Pathway | Genes in KEGG pathway | Fisher exact p-value |
97 | Jak-STAT Signaling pathway | EZR, JAM PIK3R3 | 1.0·10−2 |
Leucocyte transendothelial migration | ARGDIB,NTRK2 PIK3R3 | 1.2·10−2 | |
Neurotrophin signalling pathway | IFNGR1,LIFR, PIK3R3 SPRY1 | 2.7·10−3 |
Number of genes BH<0.05 | KEGG Pathway | Genes in KEGG pathway | Fisher exact p-value |
136 | Neurotrophin signaling pathway | PLCG2, NTRK2, NFKB1, PIK3R3, ARHGDIB | 2.1·10−3 |
Epithelial cell signaling in Helicobacterpylori infection | PLCG2, HBEGF, NFKB1, JAM2 | 1.5·10−3 | |
Viral myocarditis | HLA-DRB1, FYN, HLA-DMA, HLA-DQA1 | 1.8·10−3 | |
Cell adhesion molecules (CAMs) | HLA-DRB1, CD34, JAM2, HLA-DMA, HLA-DQA1 | 2.8·10−3 | |
Asthma | HLA-DRB1, HLA-DMA, HLA-DQA1 | 1.2·10−3 | |
Antigen processing and presentation | HLA-DRB1, HLA-DMA, HLA-DQA1, B2M | 3.2·10−3 | |
Hematopoietic cell lineage | IL1R2, HLA-DRB1, CD34, CD14 | 3.6·10−3 | |
ErbB signaling pathway | PLCG2, HBEGF, PIK3R3, NRG2 | 3.8·10−3 | |
Allograft rejection | HLA-DRB1, HLA-DMA, HLA-DQA1 | 2.3·10−3 | |
Graft-versus-host disease | HLA-DRB1, HLA-DMA, HLA-DQA1ç | 2.9·10−3 | |
Type I diabetes mellitus | HLA-DRB1, HLA-DMA, HLA-DQA1 | 3.6·10−3 | |
Intestinal immune network for IgA production | HLA-DRB1, HLA-DMA, HLA-DQA1 | 5.6·10−3 | |
Autoimmune thyroid disease | HLA-DRB1, HLA-DMA, HLA-DQA1 | 6.3·10−3 | |
Leukocyte transendothelial migration | EZR, PLCG2, PIK3R3, JAM2 | 1.1·10−2 | |
Pathogenic Escherichia coli infection | EZR, FYN, CD14 | 8.5·10−3 | |
Adipocytokine signaling pathway | NFKB1, POMC, CAMKK2 | 1.3·10−2 | |
Complement and coagulation Cascades | VWF, F13A1, CFH | 1.4·10−2 |
Number of genes BH<0.05 | KEGG Pathway | Genes in KEGG pathway | Fisher exact p-value |
237 | Cell adhesion molecules (CAMs) | ITGA9, PTPRM, HLA-DRB1, CD34, CLDN5, HLA-DRB5, HLA-DPB1, JAM2, HLA-DMA, HLA-DQA1, HLA-F | 5.5·10−7 |
Allograft rejection | HLA-DRB1, HLA-DRB5, GZMB, HLA-DPB1, HLA-DMA, HLA-DQA1, HLA-F | 2.2·10−7 | |
Graft-versus-host disease | HLA-DRB1, HLA-DRB5, GZMB, HLA-DPB1, HLA-DMA, HLA-DQA1, HLA-F | 4.0·10−7 | |
Type I diabetes mellitus | HLA-DRB1, HLA-DRB5, GZMB, HLA-DPB1, HLA-DMA, HLA-DQA1, HLA-F | 6.8·10−7 | |
Asthma | HLA-DRB1, IL13, HLA-DRB5, HLA-DPB1, HLA-DMA, HLA-DQA1 | 1.2·10−6 | |
Autoimmune thyroid disease | HLA-DRB1, HLA-DRB5, GZMB, HLA-DPB1, HLA-DMA, HLA-DQA1, HLA-F | 2.7·10−6 | |
Antigen processing and presentation | HLA-DRB1, HLA-DRB5, HLA-DPB1, HLA-DMA, HLA-DQA1, B2M, HLA-F | 7.0·10−5 | |
Hematopoietic cell lineage | IL1R2, HLA-DRB1, CD34, CD33, HLA-DRB5, EPOR, CD14 | 8.8·10−5 | |
Viral myocarditis | HLA-DRB1, HLA-DRB5, HLA-DPB1, HLA-DMA, HLA-DQA1, HLA-F | 2.4·10−4 | |
Intestinal immune network for IgA production | HLA-DRB1, HLA-DRB5, HLA-DPB1, HLA-DMA, HLA-DQA1 | 3.4·10−4 | |
Systemic lupus erythematosus | HLA-DRB1, HLA-DRB5, HLA-DPB1, HLA-DMA, HLA-DQA1 | 7.8·10−3 | |
Jak-STAT signaling pathway | SPRY1, IL10RA, STAT5A, IL13, EPOR, PIK3R3 | 1.3·10−2 | |
Aldosterone-regulated sodium reabsorption | HSD11B2, IGF1, PIK3R3 | 1.4·10−2 | |
ErbB signaling pathway | STAT5A, HBEGF, PIK3R3, NRG2 | 2.3·10−2 |
Immediately after their obtention in the surgery room, lipoaspirates (≅3 l per patient), were transported to the laboratory and splitted in 2 sterile containers inside a culture hood. For digestion, one was treated with human derivatives (HSA for washing and HS during seeding) and the other with bovine reagents (BSA and FBS). After discard blood derivatives, mature adipocytes and chirurgical infiltration solution, the isolated hASC fractions were seeded under their respective conditions (for summary of the protocol see
(
Based on previous studies
Concerning proliferation kinetics, comparison of the expansion rates for HS-hASCs and FBS-hASCs showed significant differences (
hASCs were analyzed by PCR for transcriptional evidence of genes associated with adipose mesenchymal stem cells, early development, hematopoietic and adipose tissue markers [see
We wanted to confirm that cells were expressing markers characteristic of their germ layer and undifferentiated state. We followed the evolution of these markers throughout the culture, analyzing the cells at passages 3 and 5. We first analyzed (
(
To fully characterize the procedure, a morphological analysis of the cells was also required. Optical microscopy did not reveal major differences between cells cultured with HS versus those cultured with FBS. A more detailed study of the morphology at a semithin section level (
To determine whether the lack of animal reagents could diminish this potential, we performed differentiation assays using protocols previously established, to trigger differentiation towards the adipose and chondrogenic lineages
Phase contrast images of hASCs subjected to a chemically-defined method for differentiation to adipose and cartilage tissues.Results corresponds to differentiation protocols triggered in cells at 7 passage. (
Chondrocyte differentiation was induced with specific media
To further characterize possible changes or abnormalities induced in the cells by the increase of proliferation kinetics, we next studied their transcriptional profile evolution during culture with HS, using microarrays to compare possible variations of their profile in the different passages [Analysis and data available are provided at
The top network functions were Regulation of actin cytoskeleton, ECM-receptor interaction, cell division signaling pathway and ribosome. ABI1 = abl-interactor 1; ABL1 = c-abl oncogene 1, receptor tyrosine kinase; AKAP13 = A-kinase anchor protein 13; CASP8AP2 = caspase 8 associated protein 2; caspase = apoptosis -related cystein peptidase; CEBPA = CCAAT/enhancer-binding protein alpha; CHEK1 = CHK1 checkpoint homolog (S. pombe); CLEC11A = C-type lectin domain family 11; CRHR1 = corticotropin-releasing factor receptor 1; DYNLT3 = dynein light chain Tctex-type 3; ERK1/2 = mitogen activated protein kinase; FAS = Fas (TNF receptor superfamily, member 6); FGF2 = fibroblast growth factor 2 (basic); FN1 = fibronectin 1; GLO1 = glyoxalase I; IL36A = interleukin-36 alpha; IL36B = interleukin-36 beta; IL36G = interleukin-36 gamma; LITAF = lipopolysaccharide-induced TNF factor; MICA = MHC class I polypeptide-related sequence A; NFkB = NF-kappa-beta; P2RY6 = pyrimidinergic receptor P2Y, G-protein coupled, 6; PFN2 = profilin-2; Pkc(s) = protein kinase C; PPP2RA = protein phosphatase 2; RIN1 = ras and Rab interactor 1; RPLP0 = ribosomal protein, large, P0; RRAS2 = related RAS viral (r-ras) oncogene homolog 2; SATB1 = DNA-binding protein SATB1; SDC1 = syndecan 1; SMPD2 = sphingomyelin phosphodiesterase 2, neutral membrane (neutral sphingomyelinase); Sphk = sphingosine kinase; WASF1 = WAS protein family, member 1; WASF2 = WAS protein family, member 2; WASF3 = WAS protein family, member 3.The grey nodes are the genes classified as significant. The asterisk (*) indicates the degree of up-regulation.
The top network functions were Hematopoietic cell lineage, Cell adhesion molecules, Leucocyte transendothelial migration and Complement and coagulation cascades. B2M = beta-2-microglobulin; BCR = breakpoint cluster region; CD14 = monocyte differentiation antigen CD14; CD3 = T-cell surface glycoprotein CD3 epsilon chain; EPOR = erythropoietin receptor; ERK1/2 = mitogen activated protein kinase; EZR = ezrin; F13A1 = coagulation factor XIII, A1 polypeptide; FCRLA = Fc receptor-like A; FXN = frataxin, nuclear gene encoding mitochondrial protein; FYN = tyrosine-protein kinase Fyn; GRM1 = glutamate receptor, metabotropic 1; GZMB = granzyme B (granzyme 2, cytotoxic T-lymphocyte-associated serine esterase 1); HBEGF = heparin-binding EGF-like growth factor; HLA-DRB1 = MHC class II antigen HLA-DRB1 beta 1; HLA-F = HLA class I histocompatibility antigen, alpha chain F; IFNGR1 = Interferon gamma receptor 1; IGF1 = insulin-like growth factor 1 (somatomedin C); IL10RA = interleukin 10 receptor, alpha; IL13 = interleukin 13; IL1R2 = interleukin 1 receptor, type I; Immunoglobulin = Immunoglobulin; Karyopherin beta = nucleo cytoplasmic transporter; lgG = inmunoglobulin G; lgm = immunoglobulin M; Mapk = Mitogen-activated protein kinase; NFKB = NF-kappa-beta; NFKB1 = nuclear factor NF-kappa-B p105 subunit; P38MAPK = map kinase p38; PI3K = phosphatidylinositol 4-phosphate 3-kinase; PIK3R3 = phosphatidylinositol 3-kinase regulatory subunit gamma; PLCG2 = 1-phosphatidylinositol 4,5-bisphosphate phosphodiesterase gamma-2; STAT5A = signal transducer and activator of transcription 5B; TCR = T cell antigen receptor; VEGF = vascular endothelial growth factor. The grey nodes are the genes classified as significant. The asterisk (*) indicates the degree of down-regulation.
Number of genes BH<0.05 | KEGG Pathway | Genes in KEGG pathway | Fisher exact p-value |
62 | Neurotrophin signalling pathway | ARHGDIB, NTRK2,PIK3R3 | 4.2·10−3 |
Given the wide clinical potential that hASCs had demonstrated, our objective was to design a new protocol for the isolation and maintenance of clinical grade safe cells avoiding xenogeneic reagents without affecting their characteristics and multipotency. Use of animal origin reagents is a controversial issue to be taken into account, because of zoonosis disease transmission. The risk of prion infection can be avoided by using New Zealand FBS. However, use of an average of 20% FBS in cell culture, regardless of the type results in hMSCs carrying between 7 to 30 mg of bovine serum proteins
In addition, the expression of some murine derived syalic derivatives was observed in human cells that had been cultured in contact with mouse fibroblasts, with unknown implications
On the other hand, differences in gene expression levels between MSCs cultured with FBS versus HS have been reported
The amendments introduced in our protocol during the isolation and culture did not compromise the viability, growth or characteristics of the hASCs obtained. Instead, they did improve some of these aspects. Use of HS induced high proliferative growth that increased steadily until passage 18 (33.9±4.1 HS versus 13.9±3.3 in FBS). Moreover, the introduction of HS did not interfere with the hASCs morphology, which retains a characteristic fibroblast–like morphology and cellular organization
The introduction of wash and spin cycles with HSA during the isolation, also proved to be beneficial, eliminating the majority of mature adipocytes.
In addition, we did not observe any increase in the number of mature adipocytes during the maintenance and expansion of the cell cultures due to uncontrolled differentiation, indicating that both the media and the conditions under which the crops are kept helped preserve cell multipotency. Regarding expression patterns, detectable levels of mesenchymal stem cell markers were still observed after expanding the cell cultures 5 times. This result confirmed that the conditions used for cell maintenance were adequate to preserve the undifferentiation of hASCs. The expression of hematopoietic stem cell markers such as CD45 or CD14, was also detected in the initial stages of isolation and culture in certain samples but decreased progressively throughout the passages. This observation was confirmed with the microarray analysis. At the same time, expression of adipocyte progenitor cell markers was detected in the culture during isolation and expansion. These observations indicate that the vast majority, if not all, of the stem cells present in the culture after the first passage, had a clear adipose origin and were not derived from hematopoietic stem cells.
Concerning differentiation studies, our results confirmed that hASCs cells obtained using our isolation protocol maintain their multipotent capacity. Detailed ultrastructure and gene analysis of differentiated cells using TEM revealed that their morphology was compatible with mature adipocytes and cartilage. In addition, we wanted to be sure that the culture conditions did not alter hASCs properties or induce malignization, so we performed microarray analysis. We observed that in case of up-regulated genes hASCs cell behavior followed the classical steps for in vitro culture adaptation. This observation correlates with the proliferation curve data, which indicates that during passage 5 hASCs are in exponential growth phase. In summary, expression patterns indicate that during the initial isolation phase hASCs first adhere to the culture surface, progressively adapt to the in vitro environment conditions and finally turn on the proliferation pathways. We did not detect up-regulation of pathways related with cell death or senescence in any of the passages. Down-regulation of gene expression was also analyzed. Immediately between isolation and the first passage a down regulation of genes related with migration and proliferation was observed. This result is coherent with the previous observation that adhesion pathways were instead up-regulated. Comparison between passage 3 versus passage 1, and passage 5 versus passage 3, allowed us to detect a reduction in the expression of biochemical pathways related with immune reaction. This observation can be explained both by the progressive elimination of blood cells through the sequential washing and passaging, and due to the fact that hASCs act in vitro and in vivo as immunomodulatory cells
This is the first protocol for hMSCs performed without using any animal origin reagent. With this work we have proven that these amendments support the isolation and maintenance of hASCs, preserving or improving crucial culture parameters such as viability, cell morphology and identity during long term culture, as well as proliferation and differentiation ability. To our knowledge, this is the nearest good manufacture practice condition protocol developed for hASCs, and it is a low-cost method that can be routinely and easily used for clinic and research.
(PDF)
(PDF)
(PDF)
(XLS)
(PDF)
We are grateful to Dr. Felipe Prosper, Dr. M Jose Escribá, Jorge Oliver de la Cruz, Dr. Deborah Burks and Dr. Ruben Moreno for their assistance.