Mesenchymal stem cells are usually obtained from bone marrow and require expansion culture. When samples clot before processing, a protocol using the (enzymatic) thrombolytic drug urokinase can be applied to degrade the clot. Thus, cells are released and available for expansion culture. This protocol provides a rapid and inexpensive alternative to resampling.
Mesenchymal stem cells (MSCs) – usually obtained from bone marrow – often require expansion culture. Our protocol uses clinical grade urokinase to degrade clots in the bone marrow and release MSCs for further use. This protocol provides a rapid and inexpensive alternative to bone marrow resampling. Bone marrow is a major source of MSCs, which are interesting for tissue engineering and autologous stem cell therapies. Upon withdrawal bone marrow may clot, as it comprises all of the hematopoietic system. The resulting clots contain also MSCs that are lost for expansion culture or direct stem cell therapy. We experienced that 74% of canine bone marrow samples contained clots and yielded less than half of the stem cell number expected from unclotted samples. Thus, we developed a protocol for enzymatic digestion of those clots to avoid labor-intense and costly bone marrow resampling. Urokinase – a clinically approved and readily available thrombolytic drug – clears away the bone marrow clots almost completely. As a consequence, treated bone marrow aspirates yield similar numbers of MSCs as unclotted samples. Also, after urokinase treatment the cells kept their metabolic activity and the ability to differentiate into chondrogenic, osteogenic and adipogenic lineages. Our protocol salvages clotted blood and bone marrow samples without affecting the quality of the cells. This obsoletes resampling, considerably reduces sampling costs and enables the use of clotted samples for research or therapy.
Mesenchymal stem cells (MSCs) play a major role in regenerative medicine and tissue engineering. They can migrate, differentiate into various cell types 1 and engraft, which renders them the ideal candidates for autologous therapies 2,3. Lately, clinical trials using MSCs for bone and cartilage repair, graft versus host disease or heart disease were launched 4. These MSCs can be harvested from the umbilical cord or adipose tissue but most promising results were obtained from bone marrow derived stem cells 5.
The iliac crest allows to collect a considerable amount of bone marrow and therefore serves as main site of aspiration 6. However, the quality of the aspirate decreases with increasing volume of bone marrow withdrawn. While the first 5 ml of bone marrow aspirate contain MSCs of high quality, withdrawal of larger volumes leads to dilution of the aspirate with peripheral blood from the highly vascularized bone 7. Because of the present megakaryocytes and platelets, bone marrow aspirates are prone to clotting, unless anticoagulants are used. But even with anticoagulants, clots may occur.
In bone marrow, MSCs represent only a small proportion of the total cell pool 8 and have to be expanded in culture for most tissue engineering or therapeutic applications 4. The quality of such a culture largely depends on the initial cell pool, i.e., diversity and a high starting number 9. Low numbers of MSCs from withdrawals may be partly explained by donor variability. On the other hand, MSCs from low quality samples require longer time in culture and extended passaging to reach the desired number of cells. In either case, extended passaging is a source of cell senescence and can lead to the loss of differentiation potential 10. Therefore, optimized protocols that can maximize cell yield and prevent from detrimental effects have to be developed 11,12.
When we began to work with canine MSCs, we were astonished to see that about three in four canine bone marrow samples contained clots, while fortunately clotted human samples (one in ten) were less frequent. On the other hand it was no surprise, that we observed much lower yields of MSCs from clotted samples. To solve the recurring issue of clotted samples, we developed the protocol using the thrombolytic drug urokinase instead of resampling.
Thrombolytic therapies can counteract life threatening situations such as occlusion of blood vessels causing heart attack, stroke or embolisms because of unwanted clotting. They work by degradation of the clots through enzymatic cleavage of fibrin by plasmin and enzymatic plasminogen activators. Despite the wide use for treatment of patients, only very few publications exist that utilized thrombolytic activities for laboratory applications to rescue clotted samples, mostly focusing on lymphocytes. In 1987, Niku et al. described the use of streptokinase for dissolving blood clots resulting in functional lymphocytes 13 and four years later, De Vis et al. extended the use of streptokinase to isolate leukemia cells from blood and bone marrow for flow cytometric applications 14. A more recent publication suggests the use of Alteplase for cancer diagnostics 15. While using the same enzymatic approach, our protocol focuses on the isolation of multipotent MSCs form bone marrow to provide a tool for researchers in the stem cell field.
NOTE: Human bone marrow aspirates from the iliac crest were collected from consenting donors with the approval of the ethics committee of the canton of Lucerne. Canine bone marrow aspirates from the iliac crest were collected with dog owner’s consent.Human (approx. 20 ml) and canine (approx. 10 ml) bone marrow aspirates were anti-coagulated by addition of 15 ml of 3.8% sodium citrate immediately after withdrawal in the operation theatre. The samples were transferred to the laboratory environment for processing the same day as withdrawn.
1. Preparation of Urokinase (Prior to 1st Use)
2. Preparative Steps (Prior to Every Bone Marrow Treatment)
3. Enzymatic Digest of the Clot
4. Seeding for Expansion Cell Culture and CFU Plates
5. Giemsa Stain for CFU Assay
6. MSC Differentiation
The facts that 74% of canine bone marrow samples (n=54) contained clots when they arrived in our laboratory (Figure 1A) along with decreased MSC yields from these samples, made us believe that a considerable number of MSCs was trapped within the clots. Indeed, a simple DAPI-stain of sectioned clot material confirms the presence of nucleated cells in high density (Figure 1B). This ultimately leads to low numbers of MSCs available for expansion culture, which triggered us to develop the protocol using urokinase. Typically clots disappear almost completely when applying our protocol. However, during method development we addressed the question of clot digest systematically by weighing before and after digest. These experiments revealed that the undigested remainders were 15% (dog) or 9% (human) of the initial clot weight (Figure 2A).
A typical feature of bone marrow derived MSCs is the ability to adhere to cell culture dishes. Researchers make use of this characteristic to select for MSCs. As a consequence, a colony forming assay allows to evaluate the quality of the cell pool in a simple, quantitative and reliable way. In our laboratory, the colony forming assay described herein has been applied routinely to all bone marrow samples processed (also non-clotted ones). This allowed us to use the assay as main criterion for determining the efficacy of the urokinase digest. To allow direct comparison, cells from clotted sample filtrates (i.e. as if clotted samples were just filtered but not enzymatically treated) and digested clots were seeded on separate 10 cm dishes followed by incubation for two weeks. When visualizing colony forming units (CFU) with Giemsa stain (as shown in Figure 2B), we observed 3.8 times more CFU from the urokinase reactions of canine samples than from the corresponding initial filtrates (Figure 2C). Although less prominent, human samples followed a similar trend with 1.6 fold more CFU from treated samples, confirming the suitability of the protocol. Indeed, the plates of digested clot illustrated in the right row in Figure 2B correspond to the typical result seen for a pooled sample. However, donor variation may easily result in CFU numbers from half to double of what is depicted.
Taking the colony-size as indicator of cell pool quality, more medium (2-5 mm) and large (>5 mm) colonies appeared on the plates seeded from clots compared to the CFU from filtrates (Figure 2C). This generally indicates that MSCs grow normally after urokinase treatment. Also in the aggregate, the comparison of digested canine specimens (n = 21) to samples without clot (n = 7) yielded comparable numbers of total MSCs for treated samples, although a big inter-sample variation was observed due to the heterogeneous donor population (Figure 2D).
As a last functional test of suitability, we induced differentiation of MSCs derived from digested bone marrow samples. Applications in autologous stem cell therapies are based on MSC differentiation into cartilage, bone or adipose tissue or MSC paracrine signaling 18. Cells can be guided towards the desired path of differentiation by choosing the appropriate culture conditions for adipogenic differentiation 19, osteogenic differentiation 20 or chondrogenic lineage 17. Hence, we compared MSCs from bone marrow clot to MSCs derived from unclotted bone marrow sample. After four weeks in culture, it was possible to differentiate MSCs into all three above mentioned lineages. This was histologically tested with Von Kossa (for osteogenic lineage), Oil Red O (adipogenic) and Alcian Blue (chondrogenic) stainings, showing no differences in the grade of differentiation between the groups (Figure 3).
Figure 1. Bone marrow clots. Percentage of canine and human bone marrow samples in partially clotted state upon arrival (A). Furthermore, we found that MSC yields from clotted samples were strongly reduced due to a high number of cells trapped within the clots. A high number of nucleated cells are present within the clots as demonstrated by DAPI-staining of a cryosection from a canine bone marrow clot (B). Scale bar represents 200 µm.
Figure 2. Isolation of MSCs from bone marrow clots digested with urokinase. (A) Typically, bone marrow clots disappear almost completely upon urokinase digest. Upon method development, we assessed clot weights for canine (n = 5, mean ± SD, **p ≤0.01) and human samples (n = 3, mean ± SD, n.s. = not significant p = 0.10) that were strongly reduced upon urokinase digest. (B) A simple CFU assay can serve as an informative tool for assessing the quality of an MSC preparation. To assess the efficacy of the digest, the CFU assay was performed for filtrates and digested clots from bone marrow aspirates separately. After two weeks in culture, 10 cm control plates were stained with GIEMSA and CFU were counted. The assay confirmed that high number of CFU can be released from the clot by urokinase digest and remain functional. This is also confirmed by the high frequency of large colonies (class-division: >5 mm in black, 2-5 mm in dark grey and <2 mm in light grey. Error bars represent SD of total CFU (n = 5 for dog, n = 3 for human, **p ≤0.01, n.s. = not significant, p = 0.17). (C) Pictures from Giemsa-stained plates confirm the results shown. The plates shown for digested clot correspond to a typical result for a digested bone marrow sample – however, numbers can vary largely due to donor variability. (D) In the sum, applying the urokinase digest protocol (n = 21) results in comparable total MSC yields after expansion culture to naturally clot free samples (n = 7, mean ± SD). Statistical analysis for the entire Figure 2 was performed using Student’s t-test.
Figure 3. Comparison of differentiation potential of digested versus natively unclotted samples. Canine MSCs isolated from clotted bone marrow treated with Urokinase (top row) and unclotted bone marrow (bottom row) were differentiated in osteogenic phenotype and stained by Von Kossa (bar = 200 µm), adipogenic phenotype was stained by Oil Red O (bar =100 µm; in the insets larger magnification of fat vacuoles), while chondrogenesis was revealed by Alcian blue staining (bar =100 µm; counterstaining nuclear fast red). Please click here to view a larger version of this figure.
Routinely we sample bone marrow while the patient is undergoing surgery (in our case mainly spine surgery), with the advantage that only little additional work has to be carried out by the personnel in the operation theatre. Even though the samples are mixed with sodium citrate immediately after withdrawal, many samples were partially clotted when they arrived in the laboratory for processing. At this stage, resampling to replace clotted specimens would be a separate additional intervention necessitating again local or general anesthesia 6. This requires the willingness of both the clinical staff and the donor to contribute and consumes a lot of resources 21.
Here we provide a protocol that uses recombinant human urokinase on clotted bone marrow samples to isolate multipotent MSCs. We could demonstrate that the urokinase treatment does not harm the cells and the differentiation potential is retained. The protocol is short since it prolongs sample processing only by approximately 1 ½ hr as compared to unclotted samples while yielding comparable cell pools. To date, this protocol has been successfully applied on bone marrow clots from different individuals and dogs using urokinase from several lots and has shown robust and reproducible success. Apart from this, urokinase acts indirectly via activation of sample-intrinsic plasminogen. This implies that the urokinase reaction requires the plasma-environment. It is therefore unadvisable to introduce steps of clot rinsing or alike with e.g., PBS. This would dilute the serum environment and lead to a significant slow-down of the enzymatic reaction.
The decision to use urokinase for our method and not another thrombolytic drug was trivial: we developed the protocol based on the thrombolytic drug readily available in our hospital pharmacy. For other research groups it may therefore be easier to use a tissue-type plasminogen activator (i.e., a different drug product like alteplase, reteplase or tenecteplase) if urokinase is not available. However, potentially important differences between the drugs exist: unlike tissue-type plasminogen activators, urokinase can trigger cell activation and proliferation when bound to the urokinase-type plasminogen activator receptor (uPAR) 22. Upon binding, intracellular signaling cascades are activated that lead to cell migration and proliferation. In a physiological situation this is further supported by the secretion of matrix metalloproteinases by activated cells. However, in our system, urokinase is diluted and gradually removed upon expansion culture without showing detrimental effects. Still, with respect to the intended use of isolated MSCs, the suitability of any of the thrombolytic drugs needs to be determined individually.
Generally, high dosages for rapid and complete clot digest are recommended to minimize unwanted selection during clot processing. Noteworthy, many previous studies used bacterial streptokinase for digesting blood and bone marrow clots 13,14. However, this drug may provoke unwanted activation and response of the immune system, which may be a major drawback especially when application of cells to patients is intended.
We therefore believe that our protocol does not only help researchers and medical professionals to save time and reduce costs. Using urokinase – an approved enzymatic drug without known antigenicity – may even provide a valuable tool for many translational research laboratories aiming at therapeutic use of MSC preparations.
The authors have nothing to disclose.
This work was supported by the Swiss National Foundation Grant CR3I3_140717/1 and the Swiss Paraplegic Foundation.
Basal Medium Components | ||
PenStrep 100X | Gibco | 15140122 |
Human FGF-basic | Peprotech | 100-18B |
MEM Alpha w/ Nucleoside, w/ stable Glutamine | Amimed | 1-23S50-I |
FBS Heat Inactivated | Amimed | 2-01F36-I |
Amphotericin B | Applichem | A1907 |
Adipogenic Medium Components | ||
DMEM-HAM F12 + GlutaMAX | Amimed | 1-26F09-I |
Insulin | Sigma | I5500 |
Rabbit serum | Gibco | 16120099 |
Dexamethasone | Applichem | D4902 |
3-Isobutyl-1-methylxanthine | Sigma | I5879 |
Biotin | Sigma | B4639 |
Rosiglitazone | Sigma | R2408 |
Pantothenate | Sigma | P5155 |
Oil Red-O | Sigma | O0625 |
Osteogenic Medium Components | ||
L-ascorbic acid 2-phosphate | Sigma | A8960 |
ß-glycerophosphate | Sigma | G9422 |
Silver nitrate (AgNO3) | Sigma | S6506 |
Chondrogenic Medium Components | ||
Biopad – sponge shaped medical device | Euroresearch | |
L-proline | Sigma | P5607 |
Insulin-Transferrin-Selenium X | Gibco | 51500056 |
Human transforming growth factor-β1 | Peprotech | 100-21 |
Alcian Blue 8GX | Sigma | A3157 |
Nuclear fast red | Sigma | N8002 |
Generic | ||
Tri-Sodium citrate dihydrate | Applichem | A3901 |
PBS | Applichem | 964.9100 |
Urokinase | Medac | 1976826 |
0.5% Trypsin-EDTA | Gibco | 15400054 |
Giemsa stain | Applichem | A0885 |
Formaldehyde | Applichem | A0877 |
Sulfuric acid (H2SO4) | Applichem | A0655 |
Dimethyl sulfoxide (DMSO) | Applichem | A1584 |
Magnesium chloride (MgCl2) | Applichem | A3618 |
Guanidine hydrochloride | Applichem | A1499 |
Consumables | ||
50 mL reaction tube | Axygen | SCT-50ML-25-S |
10 mL syringe | Braun | 4606108V |
Sterican needle (22G) | Braun | 4657624 |
1.7 mL Microtubes | Brunschwig | MCT-175-C |
100 μm cell strainer | Falcon | 6.05935 |
sterile forceps | Bastos Viegas, SA | 489-001 |
sterile scalpel | Braun | 5518059 |
Primaria cell cuture dish | Falcon | 353803 |
C-Chip Neubauer Improved | Bioswisstech | 505050 |
cell culture flask – Flask T300 | TPP | 90301 |
Equipment | ||
Microbiological biosafety cabinet class II | Skan | 82011500 |
water bath | Memmert | 1305.0377 |
Stripettes Serological Pipette 5ml | Corning | 4487-200ea |
microscope | Olympus | CKX41 |
humidified incubator Heracells 240 | Thermo scientific | 51026331 |
Heraeus Multifuge 1S-R | Thermo scientific | 75004331 |