The protocol described here represents an easy and reproducible method that employs reverse phase high-performance liquid chromatography (RP-HPLC) to measure purine metabolism on chronic lymphocytic leukemia (CLL) cells cultured under different conditions.
This method describes a sensitive, specific, reliable and reproducible reverse phase high-performance liquid chromatography (RP-HPLC) assay developed and validated for the quantification of extracellular purine nucleotides and nucleosides produced by purified chronic lymphocytic leukemia (CLL) cells under different culture conditions. The chromatographic separation of adenosine 5'-monophosphate (AMP), adenosine (ADO) and inosine (INO) is performed at RT on a silica-based, reversed-phase column that is used for polar compound retention. The method includes a binary mobile phase, which consists of 7 mM ammonium acetate and acetonitrile with a flow rate of 1.00 ml/min. The eluates are monitored using a Photodiode Array UV detector set at 260 nm. A standard calibration curve is generated to calculate the equation for the analytical quantification of each purine compound. System control, data acquisition and analysis are then performed. Applying this protocol, AMP, INO and ADO elute at 7, 11 and 11.9 min, respectively, and the total run time for each sample is 20 min. This protocol may be applied to different cell types and cell lines (both suspension and adherent), using culture media as matrix. The advantages are easy and fast sample preparation and the requirement of a small amount of supernatant for analysis. Furthermore, the use of a serum-free medium allows skipping the protein precipitation step with acetonitrile that impacts the final concentration of purine compounds. One of the limitations of the method is the requirement of the equilibration column run before each single sample run, making the total run time of the experiment longer and preventing high throughput screening applications.
Adenosine (ADO) is a purine nucleoside with an adenine molecule attached to a ribose sugar molecule moiety through a glycosidic bond. When present in the extracellular environment, it protects cells from excessive damage by the action of the immune system. This role has been highlighted using different disease models, such as colitis1, diabetes2, asthma3, sepsis4, and ischemic injury5. One of the main ADO functions is the inhibition of immune responses in the tumor microenvironment, contributing to tumor immune evasion6. For this reason, the mechanisms involved in ADO formation and signaling are of considerable therapeutic interest7.
ADO levels in the tissue microenvironment are relatively low under normal physiologic conditions and certainly below the sensitivity threshold of immune cells. However, during hypoxia, ischemia, inflammation, infection, metabolic stress and tumor transformation they rapidly increase8. The elevated extracellular ADO levels in response to tissue-perturbing signals have a dual function: to report tissue injury in an autocrine and paracrine way and to generate tissue responses that can be generally viewed as cytoprotective.
Extracellular ADO can be formed through a variety of mechanisms, which include release from intracellular compartments mediated by nucleoside transporters9 or accumulation because of impaired degradation operated by adenosine deaminase. The main pathway leading to increased extracellular ADO levels involves the action of a cascade of ectonucleotidases, which are membrane associated ectoenzymes generating ADO by phosphohydrolysis of nucleotides released from dead or dying cells. This pathway proceeds through the sequential action of CD39 (ectonucleoside triphosphate diphosphohydrolase-1) that converts extracellular adenosine 5'-triphosphate (ATP) or adenosine 5'-diphosphate (ADP) to adenosine 5'-monophosphate (AMP) and of CD73 (5'-nucleotidase), which converts AMP to ADO10.
Extracellular ADO elicits its physiological responses by binding to four transmembrane ADO receptors, namely A1, A2A, A2B and A3. Each receptor has different affinities for ADO and specific tissue distribution. All the receptors have seven transmembrane domains and are G-protein coupled to intracellular GTP-binding proteins (G proteins), that can induce (Gs protein) or inhibit (Gi protein) adenylate cyclase activity and, subsequently, the production of intracellular cAMP. Therefore, changes in cytoplasmic cAMP levels impact on intracellular protein kinase activity during physiological responses11. Under physiological conditions extracellular ADO is below 1 µM, which can activate indiscriminately A1, A2A and A3 receptors. However, the activation of A2B subtype requires considerably higher concentrations of the nucleoside, such as those generated under pathophysiological conditions. Alternatively, extracellular ADO can be degraded to inosine (INO) by adenosine deaminase (ADA) and CD26, an ADA complexing protein localizing ADA on the cell surface. Another possibility is that ADO is internalized by the cell through the equilibrative nucleoside transporters (ENT) and phosphorylated to AMP by ADO kinase protein12,13.
The aim of this protocol is to describe an analytical method of reverse phase high-performance liquid chromatography (RP-HPLC) to quantify in a single run the substrate AMP and the products ADO and INO, as generated by human lymphocytes. Our experience was initially obtained using cells from chronic lymphocytic leukemia (CLL) patients, which are characterized by the expansion of a mature population of CD19+/CD5+ B lymphocytes constitutively expressing CD3914,15. We showed approximately 30% of CLL patients express the CD73 ectoenzyme and that this phenotype correlates with a poor prognosis16. This subpopulation of leukemic cells co-expressing CD39 and CD73 can actively produce extracellular ADO from ADP and/or AMP. Preincubation of CD73+ CLL cells with α,β-methylene-ADP (APCP), a known inhibitor of CD73 enzymatic activity, completely blocks extracellular ADO synthesis confirming that CD73 represents the bottle-neck enzyme of that cascade16.
CLL cells also express ADA and the ADA complexing protein CD26, which are responsible for the conversion of ADO into INO. By using specific ADA inhibitors, such as erythro-9-(2-Hydroxy-3-nonyl)I wiadenine (EHNA) hydrochloride and deoxycoformycin (dCF), it is possible to block extracellular ADO degradation into INO. Furthermore, pretreatment with an ADA inhibitor in combination with dipyridamole, that blocks nucleoside transporters, enhances ADO accumulation in cell supernatants.
We have then extended this protocol to cells derived from other lineages, including T lymphocytes and myeloid cells, confirming CD73-dependent ADO production. These findings suggest that this HPLC protocol is highly versatile and that it can be applied to different cell lineages and to different culture conditions (Figure 1).
Figure 1. Schematic representation of the enzymatic machinery responsible for extracellular ADO production. Adenosine 5'-triphosphate (ATP) and/or adenosine 5'-diphosphate (ADP) can be degraded by CD39 to adenosine 5'-monophosphate (AMP), which in turn is converted by CD73 into the nucleoside adenosine (ADO). Once ADO is produced in the extracellular space, it may reenter the cell through the nucleoside transporters (ENT), be converted into inosine (INO) or bind to different types of P1 ADO receptors. Please click here to view a larger version of this figure.
CLL blood samples are obtained in accordance with Institutional Guidelines and Declaration of Helsinki.
1. Isolation of Leukemic Lymphocytes from Blood Samples of CLL Patients
2. Purification of Leukemic B Cells by Negative Isolation
3. Preparation of Standard and Inhibitors Stock Solutions
4. Program the HPLC Method
Time | Flow rate (ml/min) | %A | %B |
1.00 | 100 | 0 | |
1.24 | 1.00 | 100 | 0 |
6.22 | 1.00 | 2 | 98 |
18.65 | 1.00 | 2 | 98 |
Table 1: Equilibration column method. Schematic representation of solvent changes for the equilibration of the column. Buffer A: 7 mM ammonium acetate, pH 3.0. Buffer B: acetonitrile.
Time | Flow rate (ml/min) | %A | %B |
1.00 | 0 | 100 | |
3.74 | 1.00 | 0 | 100 |
13.71 | 1.00 | 15 | 85 |
17.00 | 1.00 | 100 | 0 |
20.00 | 1.00 | 100 | 0 |
Table 2: Run sample method. Schematic representation of solvent changes for HPLC measurement of purine compounds. Buffer A: 7 mM ammonium acetate, pH 3.0. Buffer B: acetonitrile.
5. Generation of a Standard Calibration Curve for Each Compound
Retention Time | λmax | |
AMP | 8.00 min | 260 |
INO | 11.00 min | 260 |
ADO | 11.90 min | 260 |
Table 3: Retention times of purine compounds. Typical retention times observed for AMP, ADO and INO. The UV detector is programmed to read at 260 nm.
Figure 2. Generation of an internal standard curve. Representative calibration standard curve for ADO and the relative equation obtained. Please click here to view a larger version of this figure.
6. Pretreatment with the Inhibitors and Incubation with the Substrate (AMP)
7. Samples Preparation for HPLC
8. HPLC Measurements of Purines
To evaluate the percentage (%) of leukemic cells in freshly purified PBMCs from a representative CLL patient, cells are marked with anti-CD19 and anti-CD5 antibodies. The left panel of Figure 3 represents a cytofluorimetric dot plot with a selective gate on live cells. Figure 3 shows an example of PBMC from a CLL patient before (middle panel) and after (right panel) B cell purification.
An example of HPLC quantification of purine compounds in CD73+ CLL cells is represented in Figure 4A. The different retention times of AMP, INO and ADO at 260 nm allow the simultaneous quantification of the compounds. When the assay is run correctly, all the nucleotides and nucleosides have an adequate peak separation. Starting from a concentration of 200 µM AMP as substrate, the peaks obtained for ADO and INO are clearly visible and the concentrations are in the range of the standard calibration curve. However, using also lower concentrations of the substrate (e.g., 100 µM) will give good results. After peak integration, the areas of each peak are converted into µM concentration and finally in nmoles of compound produced by 106 CLL lymphocytes through the use of standard curves.
The HPLC chromatogram of ADO produced by CD73+ cells after 60 min of pretreatment with 10 µM APCP indicates a complete blockade of extracellular ADO synthesis (Figure 4A). Because of the rapid enzymatic conversion of ADO into INO operated by the ADA/CD26 complex, pretreatment of CLL cells with 10 µM EHNA hydrochloride or 10 µM dCF is required to inhibit ADA degradation. Under these conditions extracellular ADO concentrations are significantly increased. Representative chromatograms of Figure 4 highlights how pretreatment with EHNA hydrochloride and/or dCF (Figure 4B–C) completely inhibits ADO conversion into INO.
Figure 3. Cytofluorimetric analysis of the % of CD19+/CD5+ leukemic cells in freshly purified PBMCs of a representative CLL patient. After gating on live cells (left panel), the % of CD19+/CD5+ CLL cells is evaluated. Before the negative isolation of B cells (middle panel), the % of leukemic cells corresponds to the 75%. After purification (right panel) the % of CLL B lymphocytes reaches 95%. Please click here to view a larger version of this figure.
Figure 4. Quantification of extracellular purine nucleotides and nucleosides consumed and produced by leukemic cells after exposure to AMP and different inhibitors. (A) Representative chromatograms of ADO and INO generated by CD73+ leukemic lymphocytes with or without pretreatment with α,β-methylene-ADP (APCP, 10 µM, 60 min), a specific inhibitor of CD73 enzymatic activity. (B) Representative HPLC chromatogram obtained from supernatants of CD73+ CLL cells pretreated or not with the adenosine deaminase (ADA) inhibitor erythro-9-(2-Hydroxy-3-nonyl)adenine (EHNA) hydrochloride (10 µM, 30 min) before incubation with AMP as substrate (200 µM, 30 min). (C) Representative chromatogram obtained from supernatants of CD73+ CLL cells after pretreatment or not with the ADA inhibitor deoxycoformycin (dCF, 10 µM, 30 min) before incubation with AMP (200 µM, 30 min). Data shown in panel (B) and (C) are obtained from a patient with a 98% of CD73 expression. Please click here to view a larger version of this figure.
The protocol described here permits to evaluate the activity of the CD39/CD73 adenosinergic machinery in cell culture media from purified human leukemic cells. Through this HPLC method we can follow and quantitatively measure the enzymatic generation of ADO (CD73-dependent) and its subsequent degradation to INO (CD26/ADA dependent). The use of enzyme inhibitors allows to control the protocol and to have internal controls. The advantages and novelties of this protocol are that i) it may be applied to cells that are growing in culture, that ii) it requires a small amount of culture media and that iii) the sample preparation is extremely easy.
Our system consists of a separation module equipped with a quaternary low-pressure mixing pump and inline vacuum degassing along with an autosampler. The chromatographic separation of AMP, ADO and INO is achieved on a silica-based, reversed-phase column that is used for polar compound retention. The binary mobile phase system consists of a Buffer A (7 mM ammonium acetate in double deionised water, pH 3.0 adjusted with hydrochloric acid) and Buffer B (acetonitrile).
This HPLC method is sensitive, reliable and reproducible and can be applied to different cell culture models, both adherent and suspension cells. It could be also applied to compare different culture conditions, such as normoxia (21% of O2) and hypoxia (1-3% O2) or to compare different activation states of cells. Here we perform treatment of cells with purine compounds in cell culture plates, but can be also used microcentrifuge tubes for short time treatments.
The method here described allows the quantification of extracellular purine nucleotides and nucleosides consumed and produced in the reaction avoiding the use of radiolabeled compounds such as those employed in the thin-layer chromatography (TLC) assays10. Moreover, the HPLC system is also more sensitive and selective as compared to the TLC method. However, a RP-HPLC system with tandem mass spectrometric (MS/MS) detection is now considered the best choice for quantitative measurements of nucleotides and nucleosides. A limitation of this HPLC method is the requirement of an equilibration column run before each single sample run, making the total run time of the experiment longer and significantly limiting the possibilities for high throughput screenings. Faster alternative methods are represented by the Malachite Green Phosphate assay, a colorimetric method that allows the indirect measurement of AMP conversion into ADO, and bythe luciferase-based method20. However, both these methods do not offer a direct measurement of adenosine.
The sample processing for this assay is minimal and protein precipitation with acetonitrile is not required, because the protocol is based on the use of a serum-free medium. An alternative to the serum-free medium could be to resuspend cells in PBS 5 mM glucose before the incubation with the substrate.
The main critical step within the protocol is to obtain a pure population of B cells. Indeed, the measurements of AMP, ADO and INO can be false due to the presence of other cell populations expressing the ectoenzymes involved in this cascade (e.g., T lymphocytes). A second critical point is to inject filtered culture media into the HPLC system without any cell contamination that can affect the integrity and the life of the column.
In conclusion, this assay is relatively fast and highly reliable, allowing for real-time monitoring of nucleotide consumption and nucleoside generation in different cell populations.
The authors have nothing to disclose.
This work is supported by Associazione Italiana Ricerca Cancro (IG #12754).
Human blood | |||
Milli-Q water | Millipore | double deionised water | |
Ficoll-Paque Plus | GE-Healthcare | 17-1440-03 | |
purified anti-CD3, -CD14, -CD16 | made in-house | mouse monoclonal | |
PE-labeled anti-CD19 | Miltenyi Biotec | 120-014-229 | |
FITC-labeled anti-CD5 | Miltenyi Biotec | 130-096-574 | |
Dynabeads sheep anti-mouse IgG | Invitrogen | 11031 | |
Phosphate-buffered saline (PBS) | Amresco | E404-200TABS | tablets |
bovine serum albumin (BSA) | ID bio | 1000-70 | standard grade |
isolation buffer | PBS 0.1 % BSA 2 mM EDTA, pH 7.4 | ||
AIM V serum free medium | GIBCO | 12055-091 | liquid (research grade) |
adenosine 5’-diphosphate (ADP) | Sigma-Aldrich | A2754 | |
adenosine 5’-monosphate (AMP) | Sigma-Aldrich | A1752 | |
adenosine (ADO) | Sigma-Aldrich | A9251 | |
inosine (INO) | Sigma-Aldrich | I4125 | |
α,β-methylene-ADP (APCP) | Sigma-Aldrich | M8386 | CD73 inhibitor |
EHNA hydrochloride | Sigma-Aldrich | E114 | adenosine deaminase inhibitor |
Deoxycoformycin (dCF) | Tocris | 2033 | adenosine deaminase inhibitor |
Dimethyl sulfoxide (DMSO) | Sigma-Aldrich | D2650 | |
Dipyridamole | Sigma-Aldrich | D9766 | nucleoside transporter inhibitor |
acetonitrile (CHROMASOLV Plus) | Sigma-Aldrich | 34998 | HPLC-grade |
ammonium acetate | Sigma-Aldrich | 9688 | 7 mM, pH 3.0 |
hydrochloric acid | Sigma-Aldrich | 30721-1L | min. 37 % |
Name | Company | Catalog Number | Comments |
Equipment | |||
Bürker cell counter | VWR | 631-0920 | hemocytometer |
DynaMag-15 Magnet | Invitrogen | 12301D | Dynal magnetic bead separator |
microcentrifuge safe-lock tubes | Eppendorf | 030-120-0086 | 1.5 ml |
PET centrifuge tubes | Corning | 430053/430304 | 15 – 50 ml |
Minisart RC4 syringe filters | Sartorius Stedim Biotech | 17821 | membrane 0.2 µm |
short thread vials | VWR | 548-0029 | 1.5 ml/glass |
micro-inserts | VWR | 548-0006 | 0.1 ml/glass |
screw caps | VWR | 548-0085 | 9 mm/PP blue |
Atlantis dC18 Column | Waters | 186001344 | 5 µm, 4.6 x 150 mm |
Atlantis dC18 Guard Column | Waters | 186001323 | 5 µm, 4.6 x 20 mm |
Waters Alliance 2965 Separations Module | Waters | HPLC separation module | |
Waters 2998 Photodiode Array (PDA) Detector | Waters | UV detector | |
Waters Empower2 software | Waters |