Here, a protocol for the isolation and characterization of CD4+ T-cell subsets from human peripheral blood is described. Purified CD4+ T-cells are analyzed by flow cytometry to determine proportions of T-follicular helper cell subsets.
Aberrant T-follicular helper (Tfh) cell activity is detectable in autoimmune conditions and their presence is associated with clinical outcomes when the lymph node microenvironment in B-cell non-Hodgkin's lymphoma is analyzed. Subsets of circulating T-follicular helper cells (cTfh), the circulating memory compartment of Tfh cells in the blood, are also perturbed in disease and therefore represent potential novel predictive biomarkers. Peripheral blood-based testing is advantageous because it is relatively non-invasive and allows simple serial monitoring.This article describes a method for isolating CD4+ T-cells from human blood, and further analysis by flow-cytometry to enumerate cTfh cells and the proportions of their various subsets (cTfhPD-1-/+/hi, cTfh1,2,17 and cTfh1/17). The level of these subsets was then compared between normal subjects and patients with lymphoma. We found that the method was robust enough to obtain reliable results from routinely collected patient material. The technique we describe for the analysis can be easily adapted to cell sorting and downstream applications such as RT-PCR.
T-follicular helper cells (Tfh) are a CD4+ T-cell subset that was initially characterized in lymphoid tissues1. These cells express PD-1 and CXCR5 surface receptors, secrete IL-21 and IL-4 and show nuclear expression of the transcription factor, BCL-62,3. As their name suggests, they are found in germinal centers and are essential for high affinity antibody production1.
Dysregulated Tfh responses have been implicated in disease pathogenesis, most notably autoimmune disease, where they promote the expansion of autoreactive B cells4. They also play a role in the tumor microenvironment of both solid5,6 and lymphoid cancers7. Conversely, genetic defects of surface proteins essential for Tfh function such as inducible T-cell costimulator (ICOS) result in human immunodeficiency syndromes8. CD4+ CXCR5+ cells in human peripheral blood are termed circulating T-follicular helper cells (cTfh) and are believed to be the memory compartment of Tfh cells in tissues9. The purpose of the method described here is the analysis of cTfh subsets following CD4+ cell isolation from peripheral blood samples.
Several cTfh subsets have been defined and the efficiency with which they provide B-cell help differs from one subset to another9,10,11,12. The relative proportions of these subsets are altered in a number of diseases, most prominently autoimmune disease in which there is almost always a relative increase in the more functional PD-1+/hi and/or cTfh2 or cTfh17 subsets in comparison to the less functional PD-1– and/or cTfh1 subsets12. The extent of these changes frequently associate with clinical parameters including disease activity and autoantibody titers, indicating a potential role of cTfh subset distribution as a prognostic biomarker in disease, which may reflect the activity of Tfh in lymphoid tissues9,12,13,14. Additionally, taking blood samples from participants is quick, safe and acceptable, and so allows serial monitoring for the analysis of disease progression or response to therapy.
The use of isolated CD4+ T-cells over traditional peripheral blood mononuclear cell (PBMNC) suspensions enables higher throughput flow cytometry experiments by reducing the time required to acquire a substantial number of cTfh cells for analysis. This is particularly helpful when sorting cells from rare cTfh subsets using flow activated cell sorting (FACS). To aid the efficiency, these suspensions can be cryopreserved to enable "batching" of samples to be used in the flow cytometry experiment. On testing, the CD4+ purity was not reduced by cryopreservation.
While different laboratories used different markers to categorise cTfh cells in the early stages of their discovery, the method presented here makes use of a unified scheme of two groups of cell-surface markers as proposed by Schmidt et al.12,15 to enable the simultaneous identification of cTfh and their nine recognized subsets in a single flow cytometry experiment.
As only cell surface markers are used, the cells do not require fixation or permeabilization, and thus can remain alive for downstream functional studies. This could be facilitated by cell sorting using FACS with the same antibody panel. This panel could be expanded to include other markers, allowing for the restrictions of the flow cytometer being used.
The analysis of multi-color flow cytometry experiments can be challenging due to the inherently subjective nature of gating on 2-dimensional dot plots, especially when cell populations do not have a clear bi-modal distribution in marker fluorescence, as is the case for cTfh cells and their subsets. For this reason, it is imperative to set up effective controls to reduce the artefacts to enable better resolution of populations and to set gating strategies confidently. As such, antibody panel design and the set-up of basic controls for a flow cytometry experiment, i.e., using compensation and FMO controls are outlined in Step 3.4.2 and 3.4.3,respectively.
All cTfh cells are defined as CD4+ CXCR5+ CD45RA–. The level of expression of the characteristic Tfh activation marker PD-1 can then be determined to identify the subsets of PD-1–, PD-1+ or PD-1hi cTfh cells. Then, using a combination of the chemokine receptors CXCR3 and CCR6, which are differentially expressed by traditional Th1,2 or 17 cells, cTfh can be characterized as cTfh1,2 or 17-like by a profile of CXCR3+ CCR6–, CXCR3– CCR6–, and CXCR3– CCR6+, respectively.
The antibody panel used by our laboratory is displayed in Table 1. The user may have to adapt their fluorophore selection to account for the laser and light filter configuration available on their local flow cytometer.
The following considerations influence the choice of fluorophores. Use bright fluorophores where possible. In particular, use the brightest available fluorophores on the dimmest (less highly expressed) markers. Dimmer markers include PD-1, CXCR3 and CCR6, and to a lesser extent, CXCR5. We specifically made use of the newer BB, BV and BUV fluorophores which provide excellent brightness and thus enable easier resolution of distinct populations of the cells.
Spread the fluorophore selection across the emission spectra as much as possible to minimize spectral overlap and thus the level of compensation required. A free, online tool that can be used to assist designing a flow cytometry panel can be found here: http://www.bdbioscienes.com/us/s/spectrumviewer. To save space on the emission spectrum, we employed a "dump channel" by using a viability dye with an emission wavelength that overlaps with that of APC-H7 (conjugated to CD45RA) to enable the detection (and exclusion) of both dead and/or CD45RA+ using a single detector.
Here, a protocol is presented for the isolation of peripheral blood CD4+ T-cells and their subsequent analysis by flow cytometry to determine the proportions of the different and recently described circulating subsets.
Blood samples were obtained from normal subjects (NS) (n = 12) as well as patients with marginal zone lymphoma (MZL) (n = 7) and other types of B-cell non-Hodgkin's lymphoma (BNHL) (6 FL patients, 2 lymphoplasmacytic lymphoma patients and 1 low-grade B-cell non-Hodgkin's lymphoma not otherwise specified patient). Patients were recruited from the hematology clinics at Leicester Royal Infirmary after having given informed, written consent, with ethical approval in place for all studies. Ethical approval was obtained by Leicestershire, Northamptonshire and Rutland Research Ethics Committee 1, reference 06/Q2501/122 for patient samples and the Health Research Authority (HRA) NRES Committee East Midlands-Derby, reference 14/EM/1176 for normal subjects.
1. Isolation of CD4+ T-cells from Whole Peripheral Blood
2. Flow Cytometry
3. Flow Cytometry Controls and Set-up
4. Data Analysis
High CD4+ purity was achieved using the CD4+ isolation protocol, which was reliable across all blood samples tested by us (mean: 96.6%, SD: 2.38, n=31) (Figure 3).
Identification of cTfh (CD4+ CXCR5+ cells) in a representative normal subject is presented (Figure 4A). The proportion of total cTfh cells within CD4+ cells had a median value of 29.4% (inter-quartile range (IQR) = 10.8) across 12 normal subjects. Multiple studies across several different diseases including hepatocellular carcinoma17,20, systemic lupus erythematous11,21, and rheumatoid arthritis11,14 have been conflicting in the ability to detect a difference in overall cTfh between heathy controls and patients. No significant differences were found in overall cTfh between normal subjects and MZL or BNHL patients (Figure 4B)21.
Identification of PD-1 expression within cTfh cells from a representative normal subject and BNHL patient for comparison is shown in Figure 5A. PD-1 expression was significantly higher in MZL and BNHL patients than normal subjects (Figure 5B)21. Similar increases in PD-1 expression have been demonstrated in multiple autoimmune disorders11,12,13,14.
Identification of cTfh1,2,17 and 1/17 within the population of cTfh cells using CXCR3 and CCR6 expression from a representative normal subject is shown in Figure 6A. The proportion of cTfh1 cells was significantly higher in MZL and BNHL patients than normal subjects (Figure 6B)21.
Figure 1: Illustration of overall gating strategy used to identify cTfh cells and their subsets. From top left, a population of lymphocytes are distinguished, and the doublets are excluded. Live CD45RA– cells are selected using a "dump channel". CD4+ cells are gated and CXCR5+ cells are identified as cTfh. cTfh are divided into cTfh1, 2 or 17-like using CXCR3 and CCR6 expression. PD-1 expression within these subsets is then distinguished. Figures taken from a representative normal subject blood sample. FSC-A, FSC Area; SSC-A SSC Area; FSC-W, FSC Width. Please click here to view a larger version of this figure.
Figure 2: Illustration of FMO controls. Each biaxial flow cytometry plot shows CD4+ cells stained with all fluorophores except the one of interest to demonstrate the minimum at which gates for fluorescence positivity can be set. Please click here to view a larger version of this figure.
Figure 3: Biaxial flow cytometry plot showing the identification of CD4+ cells after gating to live lymphocytes. Taken from a representative normal subject blood sample. Please click here to view a larger version of this figure.
Figure 4: Identification of cTfh cells. (A) Biaxial flow cytometry plot showing CXCR5+ expression on cells gated for CD4+ CD45RA–. Taken from a representative normal subject. (B) Relative percentages of cTfh within total CD4+ cells in normal subjects (n = 12), MZL (n = 7), BNHL (n = 9). Horizontal lines represent the median, and error bars represent inter-quartile range. No significant differences were found between groups using the Mann-Whitney U test. This figure has been modified from Byford et al.21. Please click here to view a larger version of this figure.
Figure 5: PD-1 expression and relationship to cTfh cells. (A) Flow cytometry histogram used to determine PD-1 expression within total cTfh cells. Taken from a representative normal subject and BNHL patient. (B) PD-1+ cells as a proportion of total cTfh cells. Horizontal lines represent the median, and error bars represent inter-quartile range. Medians are significantly (Mann-Whitney U-test) different between normal subjects (21.5%, IQR = 10.8, n = 12) and lymphoma patients (MZL 54.1%, IQR = 21.2, n = 7, p = 0.0008 and BNHL 45.2%, IQR =11.4, n = 9, p = 0.0003). This figure has been modified from Byford et al.21. Please click here to view a larger version of this figure.
Figure 6: CXCR3 and CCR6 expression and their relationship to cTfh1 numbers. (A) Biaxial flow cytometry plot showing the expression of CXCR3 and CCR6 on CD4+ CD45RA– CXCR5+ cells. Taken from a representative normal subject and BNHL patient. (B) cTfh1 cells as a percentage of total cTfh cells. Horizontal lines represent the median, and error bars represent inter-quartile range. Medians are significantly (Mann-Whitney U-test) different between normal subjects (20.8%, IQR = 6.7, n = 12) and lymphoma patients (MZL 32.1%, IQR = 6.8, n = 7, p = 0.013, and BNHL 35.4%, IQR = 7.6%, n = 9, p = 0.0056). This figure has been modified from Byford et al.21. Please click here to view a larger version of this figure.
Table 1: Flow cytometry antibody panel.
This protocol represents a simple and efficient way to analyze peripheral blood cTfh cells, enabling the detection of all relevant subsets identified in the literature thus far. Blood samples can be easily and efficiently obtained as part of standard out-patient clinics and serial samples can be collected in parallel with clinical data. In turn, this enables prospective studies evaluating cTfh subsets as biomarkers for disease progression or response to treatment. These studies would be particularly warranted in disease where Tfh dysregulation is implicated in pathogenesis such as autoimmunity and certain types of solid and haematological cancer. In addition, the changes in cTfh function in disease could be investigated by sorting cTfh cells using flow cytometry as only cell surface markers are used in this protocol. Sorting cTfh cells in MZL patients enabled us to find differences in gene expression profiles when compared to normal subjects21.
We found that efficient CD4+ T-cell isolation improved the data analysis. We describe the steps involved in detail because minor issues such as centrifuge braking and speed were critical to the isolation procedure. Another important step, as for all flow cytometry analysis, is to set the compensation and FMO controls.
PD-1 expression varies on cTfh cells and those cells with the highest PD-1 might be the most functional and therefore relevant, especially for the study of autoimmune disease11,12. One important issue was to decide the gate for the identification of PD-1hi cells, the activated cTfh subset, as there is no standard limit defined in the literature. This important but minor subset probably reflects active Tfh differentiation in lymphoid tissue11. To overcome this challenge, the threshold was set such that it was the same as that required to detect bona-fide Tfh cells from human tonsil with the same antibody panel. We recognize that obtaining tonsils might be problematic for some users. An alternative, would be to expand the antibody panel used in this protocol by the addition of anti-ICOS, as only PD-1hi cells are ICOS+ 12.
Here, we present an antibody panel to detect surface markers characteristic of circulating CD4+ T-cells. Circulating T-follicular regulatory cells (cTfr) are the blood memory compartment of T-follicular regulatory cells (Tfr) that are resident in lymphoid tissue and have important roles in regulating the germinal center reaction22,23. cTfr are blood CD4+ CXCR5+ cells that co-express T-regulatory cell markers including the transcription factor FoxP3. Measuring cTfr alongside cTfh may provide a more complete picture of activity in the germinal center, and could present a biomarker in disease in its own right24. Although cTfr numbers are inherently low (mean: 1.82%, SD: 1.40, n = 24 of total CD4+ cells in our own experiments), separating cTfr from cTfh would also increase the specificity of cTfh analysis. Adding a combination of specific T-regulatory cell-surface markers to our panel such as CD25 and CD12725 would enable the detection of cTfr in addition to all cTfh subsets in the same experiment using a very similar protocol. Alternatively, the more classical intracellular regulatory marker FoxP3 could be used, though this requires fixation and permeabilization preventing downstream functional assays. In concentrating on cTfh analysis, however, there are also advantages to defining a minimum panel that can be employed in conjunction with a standard laboratory flow cytometer to obtain clinically or experimentally useful results.
The authors have nothing to disclose.
The work was supported by a grant from Leukaemia UK to ETB and MJA.
RosetteSep Human CD4+ T Cell Enrichment Cocktail | STEMCELL TECHNOLOGIES | 15062 | |
Ficoll-Paque PLUS | GE Healthcare Life Sciences | 17144003 | |
BUV395 Mouse Anti-Human CD183 Clone 1C6/CXCR3 | BD Horizon | 565223 | |
BV711 Mouse Anti-Human CD196 (CCR6) Clone 11A9 | BD Horizon | 563923 | |
BV421 Rat Anti-Human CXCR5 (CD185) Clone RF8B2 | BD Horizon | 562747 | |
BB515 Mouse Anti-Human CD4 Clone RPA-T4 | BD Horizon | 564419 | |
PE Mouse Anti-Human CD279 Clone MIH4 | BD Pharmingen | 557946 | |
APC-H7 Mouse Anti-Human CD45RA Clone HI100 | BD Pharmingen | 560674 | |
LIVE/DEAD Fixable Far Red Dead Cell Stain Kit, for 633 or 635 nm excitation | ThermoFisher Scientific | L34973 | |
Brilliant Stain Buffer | BD Horizon | 563794 | |
Anti-Mouse Ig, κ/Negative Control Compensation Particles Set | BD CompBead | 552843 | |
FACS Aria II Flow Cytometer | BD Biosciences | 644832 | |
FACSDiva 6.1.3 | BD Biosciences | 643629 | Flow Cytometer Acquisition Software |
FlowJo 10.2 | Treestar Inc. | Flow Cytometry Data Analysis Software | |
Anti-CXCR3 antibody | BD Horizon | 565223 | |
Anti-CCR6 antibody | BD Horizon | 563923 | |
Anti-CXCR5 antibody | BD Horizon | 562747 | |
Anti-CD4 antibody | BD Horizon | 564419 | |
Anti-PD-1 antibody | BD Pharmingen | 557946 | |
Anti-CD45RA antibody | BD Pharmingen | 560674 | |
Viability Marker | ThermoFisher Scientific | L34973 |