CD71 surface analysis of T cells: a simple alternative for extracorporeal photopheresis quality control

Extracorporeal photopheresis (ECP) is a leukapheresis‐based cellular therapy that is used with increasing frequency worldwide to treat various T‐cell‐mediated diseases. Currently, the inhibition of T‐cell proliferation after photopheresis is analysed frequently using time‐consuming assays including radioactive thymidine assays or carboxyfluorescein succinimidyl ester (CFSE) staining. We investigated whether simple surface T‐cell staining using surrogate markers of T‐cell proliferation can replace time‐consuming measurement of T‐cell proliferation in ECP quality control.


Introduction
Extracorporeal photopheresis (ECP) is an immunomodulatory cellular therapy that has been used successfully for the first-line treatment of erythrodermic cutaneous T-cell lymphoma (CTCL) [1-3] for more than 30 years. Several studies demonstrated that ECP promotes different immunomodulatory effects including induction of regulatory T cells as well as modulation of dendritic cell activation [4,5]. Since UVA-associated apoptosis induction is a central component of the various clinically used photopheresis methods, reliable and simple quality control is essential [6]. Over time, the range of indications has been continuously expanded, so that ECP is now performed in other T-cell-mediated diseases, including chronic graftversus-host disease (GvHD) [3, 7,8] and allograft rejection after organ transplantation [9][10][11][12][13][14]. The ECP procedure is a combination of leukapheresis followed by cell exposure to 8-methoxypsoralen , followed by UVA. It involves the ex vivo collection of leucocytes by apheresis, exposure to the photosensitizing agent 8-MOP plus UVA light and finally, reinfusion into the patient [15]. Although ECP is currently being performed in over 200 centres worldwide [16], there is no consensus on standardized cell-based quality control [17]. Since 2003, various methods for ECP quality control have been described, such as analysis of proliferation inhibition by using 3 H-thymidine [6,18,19] and CFSE (5,6-carboxy fluorescein diacetate succinimidyl ester) [6,18,20] or measurement of ECP-induced cell apoptosis [20,21]. Due to its use of radioactive material, more recent studies avoided the 3 H-thymidine assay [20,21]. CFSE-based proliferation assays require labelling of leucocytes and a minimum culture period of 3 days after T-cell stimulation and are therefore relatively complex and time-consuming methods. The principal disadvantage of T-cell proliferation assays is the possible impaired proliferation capacity of T cells in patients receiving immunosuppressants or other antiproliferative drugs [18,21]. These aspects demonstrate that no ideal test procedures are currently available to analyse T-cell proliferation inhibition after ECP. However, apoptosis quantification could represent a possible alternative, as indicated in recent studies [20,21]. Here, we have investigated a novel approach: analysis of the suppression of surrogate markers for T-cell proliferation after ECP. Furthermore, we compared the results with classical ECP quality control assays including the CFSE T-cell proliferation test and apoptosis quantification. Finally, to address the possible confounding impact of immunosuppressive drugs or other patient-related factors, we performed our analyses in peripheral blood mononuclear cells (PBMCs) from both healthy donors and clinical patient samples.

Experimental photopheresis
Peripheral blood mononuclear cells were incubated for 30 min (37°C) with 8-MOP (300 ng/ml, prepared by the hospital pharmacy) and exposed to UVA light (2 J/cm 2 ) in an irradiation chamber (BIO-LINK â BLX-365, Vilber Lourmat) in a 12-well plate in 1 ml. UVA dose was controlled independently by using a calibrated UVA dosimeter (RM-21, Dr. Gr€ obel UV-Elektronik). The interval between irradiation and cell stimulation was 2 h unless otherwise stated. Results were validated on PBMCs of 11 healthy blood donors. In all samples, haematocrit was <0Á2%; mean lymphocyte content was 71 % (range: 36Á5%-97Á4%).

Carboxyfluorescein succinimidyl ester staining and T-cell analysis
Peripheral blood mononuclear cells were centrifuged to remove any culture medium residues. The cell pellet was resuspended in 1 ml PBS, mixed with CFSE (0Á5 lM final concentration) and incubated in a water bath for 10 min at 37°C with gentle mixing. For patient samples, a higher CFSE concentration (1 lM) was chosen due to lower dye uptake. Labelling was blocked by addition of warm culture medium (37°C) supplemented with 10% FBS. The samples were centrifuged (at 400 g for 5 min), and warm culture medium was added for 15 min in a water bath at 37°C. Staining was monitored 1 day later in unstimulated CFSE-stained cells by flow cytometry. The determination of the percentage of division (PD) of all CFSE-stained cells was quantitated with the following formula reported by Faivre et al. [6] (Fig. 1a):

FACS analysis
Cells were analysed with a FACS Canto II Flow Cytometer (BD Bioscience). Samples (125 ll) were transferred into FACS tubes (Sarstedt) and incubated with 10% vol/vol immunoglobulin G for 10 min at room temperature (RT) to block nonspecific binding. Monoclonal antibody staining was carried out with CD4, CD8 and CD71 (BioLegend) unless stated otherwise, for 15 min at RT in the dark. Cells were washed and either analysed or prepared for apoptosis staining. To quantify the effect of photopheresis on lymphocyte activation, the difference between preand post-photopheresis samples was analysed by calculating the change in expression (D Expr) between pre-and post-photopheresis values.

Apoptosis quantification
Apoptotic cells were quantified by staining with Annexin-V (BioLegend) and SYTOX Dead Cell Stain (Thermo Fisher Scientific) (Fig. 1b) in 100 ll of Annexin binding buffer containing calcium and 1 µl of Annexin-V (stock 4 µg/ml) for 10 min at RT. SYTOX was added 5 min prior to FACS analysis, and Annexin-V + cells were referred to as 'early apoptotic cells'. Annexin-V + /SYTOX + or Annexin-V -/SYTOX + were referred to as 'late apoptotic cells'. Total apoptotic cells were determined by summing the early and late apoptotic cells.

Statistics
Analyses were performed using GraphPad Prism software  errors (SEM). Significances were tested with a two-sided t-test, one-way ANOVA or two-way ANOVA. Subsequent post-tests are indicated in the respective legends. A probability of P < 0Á05 was considered significant.

Ki-67 suppression by photopheresis
Ki-67 is a well-known cell proliferation marker that is used for different cell populations, including tumour cells [23]. Regarding Ki-67 upregulation, we analysed different T-cell stimulants, including phytohaemagglutinin (PHA), CD3/28 and concanavalin A (Con A; Fig. 2a). CD3/28 and Con A stimulation induced the strongest Ki-67 upregulation (P < 0Á001 vs. unstimulated control) (Fig. 2a). Next, we investigated the suppression of Ki-67 upregulation by photopheresis. Photopheresis significantly impaired Ki-67 expression 22 h after stimulation in CD4 + and CD8 + T cells. (Fig. 2b/c). Furthermore, no notable Ki-67 upregulation was observed in unstimulated controls. These data suggest that Ki-67 expression analysis might represent, first and foremost, a suitable and rapid quality control marker for ECP. However, one drawback of Ki-67 expression analysis for ECP quality control is the relatively low and slow upregulation in stimulated samples. After combined Con A 2Á5/ CD3 CD28 stimulation, less than 10% of the T cells were positive for Ki-67 within 22 h.

CD71, CD25 and CD69
Since Ki-67 expression analysis was not ideal for ECP quality control, we investigated other mid-to early-expressed markers of T-cell activation, such as CD25, CD69 and CD71 [24,25]. For this purpose, PBMCs from blood donors underwent experimental photophoretic treatment with subsequent stimulation, and the kinetics of CD71, CD25 and CD69 upregulation were determined. Our data show a strong and rapid upregulation of these markers both in CD4 + and CD8 + T cells within 22 h (P < 0Á001), indicating their suitability as quality control markers for photopheresis ( Fig. 3a-d).
As described in detail [24][25][26], CD69 is rapidly upregulated after T-cell stimulation. After only 4 h of stimulation, pronounced CD69 expression of CD4 + and CD8 + T cells was detectable (Fig. 3a). In addition, markedly reduced CD69 expression in UVA/8-MOP-treated CD4 + and CD8 + T cells was already recognizable 4 h after stimulation (P < 0Á001; Fig. 3a). However, further analyses revealed the inferior robustness of CD69 expression with respect to environmental effects (variation in the serum added to the medium) when compared with CD71. Our experiments revealed a marked dependency and variation of CD69 expression with respect to medium composition and UVA exposure in unstimulated T cells (Fig. 3e). A change of the serum manufacturer (serum 1 versus serum 2) led to a marked increase in CD69 expression in UVA-treated but unstimulated CD4 + T cells (P < 0Á01). Also, in serum-free medium (vs. medium containing serum 1), a marked increase in CD69 expression after irradiation was observed (P < 0Á05). In contrast, CD71 expression was neither affected by differences in serum manufacturers nor by the omission of serum from the culture medium (Fig. 3f). These data suggest that CD69 expression is affected by cell culture additives and is therefore not optimal for quality control purposes.
In a direct comparison with CD71 and CD69, CD25 exhibited slower upregulation in stimulated CD4 + cells ( Fig. 3b). An obvious effect of photopheresis was visible 22 h after stimulation (P < 0Á001). In contrast, irradiated CD8 + T cells expressed higher CD25 baseline levels than controls up to 8 h after stimulation.
The best results were obtained with the surface marker CD71 due to its consistency and reliability. Over the entire culture period, CD71 expression was very high in stimulated controls, in contrast with the low or nearly absent CD71 expression observed in UVA/8-MOP-treated CD4 + /CD8 + T cells (Fig. 3c). As early as 6-8 h after stimulation, stimulated control CD4 + cells showed a marked increase in CD71 expression with a dominant and marked inhibitory effect of photopheresis (P < 0Á01; Fig. 3c). Expression of CD71 by CD8 + T cells was slightly delayed after stimulation, whereas a pronounced difference from the irradiated group was detectable after 22 h of stimulation (P < 0Á001). To determine the earliest possible time-point and a late time-point at which CD71 could be suitable as a quality control marker for photopheresis, we analysed CD71 expression 7 h, 16 h and 4 days after stimulation. These experiments revealed strong suppression of CD71 as early as 16 h after stimulation in photopheresis-treated CD4 + and CD8 + T cells (P < 0Á001; Fig. 3d), remaining stable for up to 4 days (P < 0Á001) (Fig. 3d). CD71 expression was also not affected by culture medium components (Fig. 3f). In contrast, CD69 expression exhibited marked variability under different culture medium conditions (Fig. 3e).
Validation of the CD71 assay and comparison with the CFSE proliferation assay CD71-based measurement as a marker of proliferation inhibition after experimental photopheresis was validated using samples from 11 blood donors. For each column (pre-and post-exP), proliferation levels after CD3/CD28/IL-2 stimulation are displayed as relative values (Delta PD) (Table 1). Absolute values (percentage of division; PD) were also collected but are not displayed. Our results demonstrated marked suppression of CD71 expression in all tested samples of stimulated PBMCs after experimental photopheresis. CD4 + T cells showed a mean DExpr of CD71 of 45Á3% (range: 36Á4%-54Á9%) after a stimulation period of 16 h (Table 1/ Fig. 4a), which increased to 70Á8% (range: 40Á6%-88Á3%) 4 days after stimulation (Table 1/ Fig. 4b). As in preliminary experiments, CD8 + T cells showed a slightly slower DExpr after experimental photopheresis. The mean DExpr was 34Á0% (range: 16Á1%-70Á8%), increasing to a mean of 61Á7% (range: 36Á1%-81Á0%) 4 days after stimulation. Inhibition of T-cell proliferation after photopheresis was confirmed by labelling the same samples with CFSE fluorescent dye and FACS analysis (Table 1/Fig. 4c). CD4 + T cells showed a DExpr of 60Á3% with a range of 25Á9%-79Á4% 4 days after stimulation. Analogously to the CD71 analysis, CD8 + T cells presented a slightly reduced DExpr of 51Á2% (range: 26Á6%-68Á7%) in the CFSE assay. Direct comparison of results obtained 4 days after stimulation in    the CD71 and CFSE assays showed a strong correlation of DExpr in CD4 + (r = 0Á8717; P = 0Á0005) and CD8 + T cells (r = 0Á9114; P < 0Á0001), indicating that the simple CD71 assay may be suitable to replace the CFSE proliferation assay for quality control in photopheresis (Fig. 4d).

Comparison of CD71 and apoptosis analysis
Finally, the utility of CD71 expression analysis in PBMC was directly compared to that of apoptosis quantification, since several laboratories have used apoptosis detection as a quality control method in photopheresis. The photopheresis-induced increase in total apoptosis in CD4 + and CD8 + T cells was significant after 16 h of culture (P < 0Á001 for CD4 + and P < 0Á01 for CD8 + ; data not shown). Comparison of overall induced DCD71 and DApoptosis in PBMCs after photopheresis indicated the superiority of CD71 analysis to detect the effect of photopheresis in non-patient samples, especially for CD4 + T cells (P < 0Á001) (Fig. 4e).

Prospective CD71 analysis in patient photopheresis samples and comparison with the standard CFSE fluorescence proliferation test
Photopheresis patients frequently take different immunosuppressive drugs targeting T-cell proliferation or activation. In these patients, stimulation-dependent quality control tests sometimes cannot indicate photopheresis-dependent proliferation inhibition [18]. Therefore, applicability of the new CD71 assay to patients was prospectively validated on clinical ECP procedures and directly compared to the 'gold' standard CFSE proliferation assay. All patients received combinations of different immunosuppressive drugs. The resting period between ECP and T-cell stimulation was extended to 26 h due to the observation of a delayed photopheresis effect in preliminary results with patient samples. CD71 expression was analysed 16 h and 3 and 4 days after stimulation (Table 2). In patient samples, a stimulation time of 16 h was not optimal to obtain a clear DExpr of CD71 in CD4+ (mean: 8Á4%, SD: 4Á8%) and CD8 + T cells (mean: 6Á0%,   SD: 4Á4%). Three days after stimulation, a marked photopheresis effect was observed in 8 out of 10 ECP procedures, with a DExpr of CD71 of 20Á0%-62Á9% (CD4+) and 17Á4%-56Á5% (CD8 + ). Two ECP treatments (1a, 2b) showed no clear CD71 suppression after ECP throughout the whole experiment. In a direct comparison, the CFSE proliferation assay revealed 4 out of 10 ECP procedures with a DExpr of only <10% in CD4 + and CD8 + T cells (1a, 2b, 5d, 7e). In 6 out of 10 ECP treatments, the DExpr was 18Á3%-50Á3% (CD4 + ) and 12Á9%-40Á3% (CD8 + ). These results confirmed that the CD71 assay represents a potential simple alternative to the CFSE assay, if analysis of Tcell proliferation inhibition is required. Based on the data, we suggest a Δ PD (%) expression threshold (Pre-ECP delta PD% -Post-ECP delta PD%) of 10% for patients under immunosuppressive therapy and 30% for non-patient PBMCs for the CD71 assay.

Apoptosis detection represents an alternative to proliferation-dependent quality control tests
Finally, we investigated the suitability of apoptosis quantification in patient samples. The kinetics of the induction of total apoptosis in patients' PBMCs were assessed in pre-ECP samples and compared to apoptosis levels after ECP (Fig. 5a-c). The percentage of total apoptotic CD4 + and CD8 + T cells was determined 24 h, 42 h, 4 days and 5 days after ECP. The change in apoptosis (Dapoptosis) from pre-to post-ECP was determined (Table 3). CD4 + and CD8 + T cells showed a small and non-significant increase in total apoptosis 24 h after ECP (P > 0Á05). From 42 h after ECP, CD8 + T cells showed a marked and significant increase in total apoptosis (P < 0Á01) with a mean Dapoptosis of 16Á0% (SD: 10Á9%). Four days after ECP, a more pronounced elevation in apoptosis was observed in CD4 + and CD8 + T cells, which increased even further 5 days after ECP to a Dapoptosis of 60Á2% (SD: 17Á6%) in CD4 + T cells and 64Á6% (SD: 15Á0%) in CD8 + T cells (P < 0Á001). Measurements of late apoptosis only showed no marked increase in apoptosis after 24 and 42 h (P > 0Á05, n.s.) but provided significant results 4 and 5 days after ECP (Fig. 5b).

Discussion
Here, we present data on a novel, simple alternative quality control assay for ECP using CD71 expression as a surrogate marker of T-cell proliferation. In addition to CD71, we have investigated Ki-67, CD25 and CD69 as possible quality control markers for extracorporeal photopheresis. The expression levels were systematically investigated after UVA/8-MOP treatment in PBMCs isolated from blood donors and in patient samples. To validate our results, CFSE and apoptosis assays, representing accepted quality control methods [6,18,21], were also performed and directly compared. CD71 was found to be a highly reliable, rapid and simple test marker to display UVA/8-MOP-related effects. As early as 16 h after stimulation, sufficient T-cell activation and suppression after experimental photopheresis was detectable. The ability of CD71 to display T-cell proliferation has been documented in several previous studies [24,25,27,28]. In a direct comparison between Ki-67 and CD71, Motamedi et al. demonstrated that both markers display a comparable expression pattern after stimulation, and thus, the simpler CD71 measurement can replace the intracellular Ki-67 measurement beyond a certain stimulation time [24]. The slower expression of Ki-67 displayed in our results might be due to the de novo synthesis of the protein in the first cell cycle after stimulation [29]. The novel CD71 assay was subsequently validated in samples from blood donors. Overall, strong and uniform CD71 expression could be triggered in each sample after CD3/28/IL-2 stimulation, which was dramatically reduced in the UVA/8-MOP-treated samples. The CFSE assay was found to be a reliable test system, in agreement with previously reported results [20]. We observed a strong inhibitory effect in all tested samples of healthy donors similar to the results of Szczepiorkowski et al. [30]. In the context of ECP quality control, CD71 was examined for the first time in this study; therefore, no comparative values are available in the literature. Possible applications of the fast and simple CD71 assay are conceivable for the routine quality control of ECP procedures, as well as with respect to the effective validation of novel ECP equipment, such as tubing systems and apheresis machines. A major advantage of the CD71 assay is the rapid availability of distinctive test results within 24 h. In patient specimens, T-cell activation was <30% in 3 (CD4 + ) and 4 (CD8 + ) out of 6 patients 16 h after stimulation, probably because these patients had received immunosuppressive therapy, which inhibited T-cell activation and proliferation. Accordingly, the proportion of proliferating cells determined via CFSE assay was also low, at 29Á1% (CD4 + ) and 23Á7% (CD8 + ). The partially reduced proliferation behaviour of patient PBMCs in photopheresis is in agreement with the studies of Faivre et al. [6] and Evrard et al. [18]. All patient ECP procedures investigated in our study were under concurrent potent dual or triple immunosuppressive regimens including tacrolimus (a calcineurin inhibitor), mycophenolate mofetil, prednisolone and everolimus (a mammalian target of rapamycin (mTOR) inhibitor. Despite the lower proliferation levels, the CD71and CFSE-assay display an inhibitory ECP effect in almost all tested samples. Our results are comparable to the study of Faivre et al., who recommends a minimum inhibition rate of the CFSE assay of approximately 70% [6]. In patient samples, however, the CD71 assay is not faster than the CFSE assay due to incubation times, but it is easier to perform. Direct apoptosis detection might represent an alternative quality control method, since we found a marked photopheresis effect after 4 days in all patient samples. Furthermore, since PBMCs can remain unstimulated and only Annexin-V and SYTOX staining is performed, this assay is very simple to perform. However, on the other hand, apoptosis detection in cultured blood samples definitively requires strict negative controls, since Kinetics of total apoptosis in patient samples after ECP (a-c); n = 10 experiments with six patients; two-way ANOVA with Bonferroni post-test; * P < 0Á05; ** P < 0Á01; *** P < 0Á001; d, days; h, hours; n.s., not significant. significant apoptosis will occur in all samples over time.
In summary, our study revealed that CD71 detection represents a simple and reliable quality control assay for extracorporal photopheresis in normal PBMC samples devoid of T-cell suppressive drugs.