Controlled-rate freezer cryopreservation of highly concentrated peripheral blood mononuclear cells results in higher cell yields and superior autologous T-cell stimulation for dendritic cell-based immunotherapy
- First online:
Availability of large quantities of functionally effective dendritic cells (DC) represents one of the major challenges for immunotherapeutic trials against infectious or malignant diseases. Low numbers or insufficient T-cell activation of DC may result in premature termination of treatment and unsatisfying immune responses in clinical trials. Based on the notion that cryopreservation of monocytes is superior to cryopreservation of immature or mature DC in terms of resulting DC quantity and immuno-stimulatory capacity, we aimed to establish an optimized protocol for the cryopreservation of highly concentrated peripheral blood mononuclear cells (PBMC) for DC-based immunotherapy. Cryopreserved cell preparations were analyzed regarding quantitative recovery, viability, phenotype, and functional properties. In contrast to standard isopropyl alcohol (IPA) freezing, PBMC cryopreservation in an automated controlled-rate freezer (CRF) with subsequent thawing and differentiation resulted in significantly higher cell yields of immature and mature DC. Immature DC yields and total protein content after using CRF were comparable with results obtained with freshly prepared PBMC and exceeded results of standard IPA freezing by approximately 50 %. While differentiation markers, allogeneic T-cell stimulation, viability, and cytokine profiles were similar to DC from standard freezing procedures, DC generated from CRF-cryopreserved PBMC induced a significantly higher antigen-specific IFN-γ release from autologous effector T cells. In summary, automated controlled-rate freezing of highly concentrated PBMC represents an improved method for increasing DC yields and autologous T-cell stimulation.
KeywordsDC Dendritic cells Cryopreservation Cellular immunotherapy Cell yields
Based on their unique abilities to initiate, tolerate, or abrogate immune responses, dendritic cells (DC) have become major candidates for immunotherapy. Especially, the capacity of DC to take up and present antigenic epitopes on major histocompatibility complex (MHC)-I and MHC-II to CD4+ and/or CD8+ T cells account for their promising potential in clinical trials [1, 2]. Most research has focused on DC-based cancer immunotherapy [3, 4], which is spearheaded by the FDA approval of sipuleucel-T (APC8015), a cellular vaccine for patients with metastatic prostate cancer based on enriched autologous antigen-presenting cells briefly cultured with a fusion protein of prostatic acid phosphatase and GM-CSF . Other areas of intense research for DC-based or DC-targeted immunotherapy include infectious diseases, allergy, autoimmunity, and transplantation medicine [6–8].
Ex vivo manipulation of DC for reinfusion into patients as well as monitoring of cellular function during immunotherapeutic clinical trials usually require cryopreservation of cells. Over the last several years, cryopreservation protocols for fully differentiated DC after a broad variety of maturation agents, antigen loading, and costimulatory strategies were validated [9–11]. Recently, an in-depth analysis of cryopreservation of different DC and progenitor cells revealed superiority of cryopreserved monocytes for DC generation over all other types of cryopreserved immature, semi-mature, and mature DC regarding their cell viability, surface markers, and functional properties . DC cryopreservation is usually performed by standard cryopreservation containers where cell tubes are surrounded by isopropyl alcohol (IPA) and simply placed inside a −80 °C freezer.
Generally, cell damage in freeze–thaw processes is either caused by extensive cell dehydration (“solution effect”), by intracellular ice crystallization (“mechanical damage”), or a combination of both. Cell dehydration commonly occurs during low-rate freezing, while rapid freezing usually results in ice crystallization [13, 14]. Cryopreservation by controlled-rate freezers (CRF) is thought to minimize both effects through continuous adjustment of the temperature decline according to the actual temperature of the cryo specimen thus compensating for fusion heat and minimizing supercooling effects [15–17]. Although CRF are routinely used for cryopreservation of autologous peripheral blood progenitor cells (PBPC) for transplantation , this approach has not been applied for clinical studies with DC immunotherapy. To the best of our knowledge, a head-to-head comparison of phenotypic and functional properties of PBMC for DC immunotherapy after uncontrolled- and controlled-rate freezing procedures has not been reported.
Materials and methods
PBMC preparation and cryopreservation
PBMC from IPA and CRF cryopreservation are termed “IPA-PBMC” and “CRF-PBMC” throughout this manuscript. DC generated from non-cryopreserved (NC) PBMC is referred to as “NC-DC”, and DC generated from IPA- and CRF-cryopreserved PBMC as “IPA-DC” and “CRF-DC”, respectively.
DC generation and maturation
Immature DC (iDC) were generated from a plastic-adherent cell fraction of PBMC. Adherent cells were grown in RPMI 1640 supplemented with 10 % FCS, 100 IU/ml penicillin, 100 μg/ml streptomycin (PAA, Pasching, Austria), 800 IU/ml GM-CSF, and 1,000 IU/ml IL-4 (Immunotools, Frisoythe, Germany) and cultured for 6 days as described earlier . For maturation of iDC, day 6 cultures were supplemented with 18 μg/ml of poly(I/C) (Sigma-Aldrich, St. Louis, Missouri, USA) for 48 h. As control, iDC were cultured for another 2 days in RPMI 1640 containing 10 % FCS, 800 IU/ml GM-CSF, and 1,000 IU/ml IL-4. On day 8, all supernatants were collected and stored at −20 °C for cytokine profile analysis.
Antigen expression was determined using antibodies against the following antigens: CD3, CD4, CD8, CD14, CD19, CD83, CD86, HLA-DR, or the respective isotype-matched controls (all Invitrogen, Carlsbad, CS, USA except CD83-Ab from Beckman Coulter, Krefeld, Germany). Integrity of the plasma membrane and thereby cell viability was assessed by propidium iodide staining (Sigma-Aldrich, St. Louis, MO, USA). Cells were washed and immediately analyzed by flow cytometry (FACSCanto II with DIVA software, BD Biosciences, San Jose, CA, USA). Dead cells and debris were gated out on the basis of their light scatter properties.
Total protein quantification
During DC generation, the adherent cells were harvested on day 1, day 4, and day 6 by repeated scraping to collect all cell material for protein quantification (since whole cell harvesting was not possible due to adhesion). Pellets were stored at −20 °C until total protein quantification by Bradford protein assay (Bio-Rad Laboratories, Munich, Germany).
Cytokine profiling in cell culture supernatants
Cells were plated on day 6 at a density of 2 × 106 iDC per well in 6-well plates in RPMI 1640 containing 10 % FCS, 800 IU/ml GM-CSF, and 1,000 IU/ml IL-4. After maturation by poly(I/C) for 48 h, cell culture supernatants were collected and stored at −20 °C for cytokine profile analysis. Nitrocellulose membranes spotted with capture antibodies against 36 human cytokines (R&D, Minneapolis, MN, USA) were used for cytokine profiling of cell culture supernatants. The assay was performed according to the manufacturer’s instructions. After adding a chemiluminescent agent (ECL Plus Detection Reagents, GE Healthcare, Buckinghamshire, UK), blots were recorded by image reader LAS-4000 (Fujifilm Europe, Düsseldorf, Germany) and submitted to an automated digital analysis using Fujifilm Multi Gauge software (Fujifilm Europe, Düsseldorf, Germany).
For verification of results concerning the IL-12p70 (bioactive heterodimer) concentration in supernatants, quantitative analysis was performed by ELISA assay (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions.
Proliferative response of allogeneic CFDA-SE-labeled T cells
In order to measure proliferative responses of allogeneic T cells to cryopreserved or freshly prepared DC, respectively, T cells were separated from PBMC by magnetic cell sorting (MACS) using CD3-MicroBeads (Mitenyi Biotec, Bergisch Gladbach, Germany). FACS analysis revealed a purity of >95 %. For carboxyfluorescein diacetate N-succinimidyl ester (CFDA-SE) proliferation assays, CD3+ T cells were stained with CFDA-SE (Invitrogen, Oregon, USA) according to the manufacturer’s instructions. Day 8 DC were cultured with allogeneic CFDA-SE-labeled CD3+ T cells in triplicates in flat-bottom 96-well plates in a final volume of 200 μl/well RPMI 1640 with 10 % FCS. 2 × 105 T cells were stimulated with 2 × 104 DC (ratio 10:1) . After 5 days, cocultures were washed and assessed for CFSE staining of T cells by flow cytometry.
Enzyme-linked-immunospot (ELISPOT) assay of DC with autologous PBMC
During DC maturation between day 6 and 8, DC cell cultures were supplemented with 50 μl/ml tetanustoxoid (TTX, a kind gift of Novartis Behring, Marburg, Germany) as recall antigen. On day 8, TTX-loaded DC resulting from the different cryopreservation protocols were washed separately and cultured in a 24-well plate with freshly thawed autologous PBMC (all cryopreserved by IPA protocol) for 7 days for expansion of antigen-specific memory T cells [2 × 106 PBMC cocultured with 2 × 105 DC (ratio 10:1), or 6.6 × 104 DC (ratio 30:1)]. For improving further expansion and survival of T cells, 12 U/ml IL-2 (R&D Systems, Minneapolis, MN, USA) were added on day 10. After 7 days of coculture, cells were washed and analyzed by ELISPOT assays in order to quantitate TTX-driven IFN-γ release of effector T cells.
For IFN-γ ELISPOT, nitrocellulose-bottomed 96-well plates were pre-coated with capturing anti-IFN-γ mAb (Mabtech, Nacka Strand, Sweden). Briefly, 5 × 104 non-adherent cells of re-stimulated cultures were incubated in triplicates with 5 × 104 TTX-loaded autologous PBMC in 200 μl ELISPOT medium (RPMI 1640, 10 % FCS, 100 U/ml penicillin, 100 μg/ml streptomycin) for 48 h at 37 °C and 5 % CO2. Spots were stained using a Vectastain® Elite kit (Vector Laboratories, Burlingame, CA, USA) and counted by a computer-assisted video imaging analysis system (AID EliSpot Reader System, Autoimmun Diagnostika, Strassberg, Germany). After configuration of software, all ELISPOT plates were analyzed in one single pass. No spots were added or removed manually.
All figures show mean ± standard error of observed values or mean and observed values. The F tests of the ANOVA model were performed using the option ANOVAF, which is similar to the method described by Brunner et al. . The considered outcome measures were regarded as logarithmized to achieve a better model fit. Normality was checked visually (via histogram and q–q-plot of the residuals) and by performance of a Shapiro–Wilk test. The level of significance was set to α = 0.05. If the closed test principle could not be applied, we corrected for multiple testing by Tukey–Kramer. Additionally, we performed a nonparametric Kruskal–Wallis test to ensure homogeneity in individual (directly observed) divergence of method-related specific INF-γ spots between the ratios 1:10 and 1:30 of the ELISPOT experiments. All analyses were performed using SAS 9.2 software (SAS Institute Inc., Cary, NC, USA).
Cell yield of immature DC depends on method of PBMC cryopreservation
To reassess the results of day 6 cell counts, additional whole protein analysis of culture plate contents after repeated thorough scraping and microscopically ensured clearance of all cells was performed on days 1, 4, and 6. Whole protein quantification showed significantly higher protein amounts per culture plate for CRF-PBMC in comparison with IPA-PBMC and thus confirmed results of cell counts (Fig. 2c, d). Of note, an increased protein amount of CRF-PBMC was already observable after 1 day of culture. Flow cytometry analysis of NC-iDC, IPA-iDC, and CRF-iDC on day 6 revealed no differences in cell size or granularity (data not shown), underlining that the additional protein amount of CRF-iDC resulted from higher cell numbers.
CRF-iDC and non-cryopreserved iDC show comparable phenotypic properties and maturation markers
Cytokine profiles of CRF-mDC, IPA-mDC and freshly prepared mDC
Expression of GM-CSF and IL-4 was not evaluated due to external supplementation with these cytokines during cell culture. For the majority of the 36 cytokines an inter-donor variability was noted.
CRF-mDC demonstrate identical allogeneic, but superior autologous T-cell stimulatory capacity compared with IPA-mDC
We present data in favor of using highly concentrated PBMC, cryopreserved by a computer-assisted controlled-rate freezer (CRF), to reach a maximum yield of viable iDC after 6 days of culture. In our experiments, uncontrolled freezing by using standard isopropyl alcohol (IPA) resulted in significant cell loss during differentiation into iDC. Increased cell recovery after cryopreservation is a crucial aspect during clinical application of DC immunotherapy, especially with respect to human and financial resources and to prevent additional leukapheresis for the patient. It is thought that only a small fraction of injected DC finally reaches the lymph nodes depending on the injection route (e.g., ≤4 % after intradermal injection ). Moreover, a recent meta-analysis revealed a positive influence of higher DC doses on the clinical response rate in DC-based immunotherapy against cancer . Generally, one to several million DC are used per single injection, and this procedure is repeated in defined intervals for several times . Good manufacturing practice (GMP) guidelines demand regular quality controls of DC before administration, resulting in higher DC cell demand during the manufacturing process. In several clinical trials, even cautiously calculated requirements for DC cell numbers were not achieved [25, 26], or all available DC per patient were administered resulting in a high variability of injected cell numbers [27–29]. Therefore, cryopreservation of PBMC by CRF aims to ensure an improved DC yield as one central prerequisite for successful DC immunotherapy.
Plastic adherence of PBMC obtained from leukapheresis for generation of monocyte-derived DC is the most widely used technique because of easy handling and cost-effectiveness and has been extensively validated and optimized for closed cell culture systems according to GMP guidelines for patient application [30, 31]. However, the variability in DC purity after this approach was addressed by recent publications [32–34], and other technologies such as positive and negative immunomagnetic selection and monocyte elutriation, on the basis of counter-flow centrifugation, have been evaluated as alternatives for cell preparation [33, 35]. Nevertheless, since methods differ significantly in need of financial and human resources, and concerns exist regarding activation signals by extended pretreatment or labeling with xenogeneic antibodies [36, 37], the optimal isolation method for DC generation it is a matter of ongoing controversy . During the years 1997–2008, more than 600 melanoma patients have been treated with DC immunotherapy, the vast majority of patients received monocyte-derived DC after plastic-adherence of PBMC . For evaluation of more recent trials, our literature research retrieved 32 published immunotherapy trials using dendritic cells (MEDLINE, search terms: dendritic cells, immunotherapy, limits: clinical trials, English language, date range: 01/01/2008-01/01/2010). Three clinical trials out of these were not accessible for the detailed study protocol. The method of plastic adherence for monocyte isolation was used in 23/29 trials, magnetic beads for cell isolation in 5/29 studies, and elutriation in one trial. Remarkably, all researchers working with cryopreserved cells did not use controlled-rate freezing procedures.
We developed further the results of one earlier study that showed the superiority of cryopreserved monocytes over pre-differentiated iDC or mDC for DC immunotherapy . In our experiments, we continued to increase DC yields by introducing a computer-assisted controlled-rate freezer (CRF). Our finding that total protein per plate of IPA-PBMC is already significantly reduced 24 h after culture of adherent cells led to the hypothesis that fewer IPA-PBMC recover compared with CRF-PBMC. It was shown that viability of cells after standard freezing procedures was significantly reduced . Another contributing factor for reduced IPA-iDC numbers might be cryopreservation-dependent downregulation of adherence molecules on monocytes and therefore an increased loss of CD14+ cells during pre-plating. Earlier studies suggested the involvement of CD62L (L-selectin), as it mediates monocyte adhesion to cytokine-stimulated endothelial monolayers , and its downregulation was regularly observed during cryopreservation of hematopoietic stem cells and PBMC [40, 41].
The applied method of CRF for cryopreservation of PBMC was analyzed in one earlier study. The authors concluded that CRF-PBMC can be stored for at least 12 years with no general tendency toward cell loss and no statistically significant changes in the percentage of viable cells in correlation with the time of storage . Interestingly, the authors also observed that IPA-PBMC had an estimated loss of 2.23 % of viability per year, compared with a significant lower 0.01 % reduction of viability per year for CRF-PBMC. Therefore, the selected method of cryopreservation does not only affect the cells at the time of the procedure but shows a prolonged influence on quantity and quality of cryopreserved cells depending on the time of storage.
Our comparative phenotypical analysis of NC-iDC, IPA-iDC, and CRF-iDC and corresponding mDC did not reveal significant differences in the expression of surface molecules including CD83 and CD14. As expected, the induction of maturation by poly(I/C) resulted in statistically significant differences in CD83 expression between immature and mature DC . Therefore, our results confirm an earlier study that also reported no phenotypical differences between NC-DC and DC generated from CRF-PBMC .
In addition, after the different methods of cryopreservation, we found no differences in cytokine secretion during DC maturation by semi-quantitative cytokine array screening for 36 cytokines with two exceptions. First, IL12p70 secretion into the supernatant was lower for NC-DC and higher for CRF-DC. However, exact quantification by ELISA technique did not show significance of this observation. To the best of our knowledge, no other studies comparing IL12p70 secretion of differently cryopreserved DC after maturation by poly(I/C) have been published. Second, CXCL1 (Groα) secretion was twofold higher by NC-DC compared with cryopreserved DC on array analysis. This finding was not further evaluated in our experiments because of the rather limited role of CXCL1 in DC chemotaxis .
Phenotypical and functional analysis of DC after the different cryopreservation protocols was completed by assays of allogeneic and autologous T-cell stimulation. Proliferation assays of IPA-DC and CRF-DC with allogeneic CFDA-SE labeled T cells revealed no differences in stimulatory capacity. Remarkably, after 1 week of re-stimulation of autologous PBMC with antigen-loaded mDC for expansion of specific T cells, determination of specific IFNγ-spots resulted in significantly more spots for CRF-mDC than for IPA-mDC. Moreover, three of six IPA-mDC batches (in two of three donors) and none of six CRF-mDC batches failed to induce an autologous T-cell response with antigen-specific IFNγ-spots. Given that the induction of specific autologous T-cell responses is crucial in DC-based immunotherapy , this observation might be important for the design of future clinical studies. Although several studies emphasized that cryopreservation does not alter phenotypic or functional properties of DC [9, 45, 46], these studies evaluated already differentiated iDC or mDC for IPA-cryopreservation and may not reflect all aspects of cryopreserving cells for DC immunotherapy.
In conclusion, this study demonstrates that automated controlled-rate freezing of PBMC for DC generation results in higher cell yields of immature DC as compared with standard uncontrolled freezing of cells. Moreover, CRF-DC do not differ from DC after standard uncontrolled freezing or freshly prepared DC regarding the expression of relevant surface markers, the potential of full DC maturation, cytokine production, and allogeneic T-cell stimulatory capacity. Importantly, DC generated from CRF-cryopreserved PBMC induced a significantly higher antigen-specific IFN-γ release from autologous effector T cells in ELISPOT assays as compared with DC after standard freezing procedures.
We thank A. Knüppel for excellent technical assistance and the Department of Transfusion Medicine of the University Medical Center Göttingen for supply with the leukapheresis. Furthermore, we highly appreciate the contribution of all leukapheresis donors. The work was supported by a faculty grant of the University Medical Center Göttingen for young researchers to TB.
Conflict of interest
The authors have no conflicting financial interests.
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.