Stability testing is performed to ensure that products remain potent at the time of administration. Since the life of most cells is relatively short (sometimes only hours), stability testing must document cells that are still viable and potent at the end of the several hours required to release the product from the cell therapy laboratory's inventory, transport the product from the laboratory to the patient's bedside, and administer the product to the recipient.

Using the same protocol, there may still be considerable lot-to-lot variability in clinical cellular therapies produced. Consistency testing, therefore, measures variability among products. This variability may be due to differences in starting material or differences in the manufacturing process. Differences in the starting material may be due to variability in the collection process. There may also be differences in the starting material as a result of variability among donors with genetic differences. Genetic polymorphisms in cytokines, growth factors, and their receptors affect the cellular immune response [12-15]. It is likely that these polymorphisms affect the response of cells to cytokine and growth factor stimulation in vitro and the behavior of cells during culture. Epigenetic changes may also be important. The same type of cells obtained from different donors, time points, and physiological conditions could vary significantly due to genetic heterogeneities, epigenetic differences, or transcription regulation diversities. When material is collected from patients for autologous use, there may also be differences due to the patients' disease stage or type, or due to previous therapies.

Variability in the product could also be introduced in the manufacturing process due to differences in equipment, supplies, reagents, and the techniques used by the laboratory staff. Variability due to such production factors can be assessed by measuring the consistency of multiple products manufactured from the same donor or patient. Differences due to donor factors can be assessed by comparing the differences among multiple products manufactured from each donor. Environmentally caused differences can be measured by comparing multiple products manufactured from the same donor or patient at different time points or in different batches.

It is often necessary to compare products manufactured using different methods. During the scale up of the initial production of a cellular therapy, it is necessary to demonstrate that cells produced on a small scale are the same as those produced with a modified method used for manufacturing cells on a much larger scale. The scaling up of methods often requires the use of different containers, reagents, instruments, and media. These changes could affect the potency of the final product.

Once a specific cellular therapy is being manufactured for patient use, it may be necessary to change the manufacturing process or manufacturing facility due to a change or lack of availability of a reagent, device, or instrument used in the manufacturing process. Cells manufactured using a revised method should be compared with those produced using the original method to ensure that the 2 methods yield comparable cells.

Gene expression microarrays can be utilized to predict the quality of cells used to manufacture biological products. Human embryonic kidney (HEK) 293 cells are often used to manufacture products such as adenoviral gene therapy vectors and vaccines [28]. These cells can be grown in bioreactors, tissue culture flasks, and roller bottles. However, when HEK 293 cells have grown to form a confluent monolayer, their phenotype changes as does the quality of the vector or vaccine produced by these cells. Cell confluence can be readily assessed by visual inspection of cells grown in flasks and roller bottles, but for cells grown in bioreactors, the assessment of confluence by visual inspection is not always possible or clear. Gene expression profiling can be used to identify genes whose expression predicts cell confluence [28]. HEK 293 cells that have been grown to 90% confluency have a unique gene expression signature compared to those grown to 40% confluency and therefore a set of 37 of these signature genes is able to predict that quality and confluence of HEK 293 cells.

Human embryonic stem cells (hESCs) have the potential to be useful for a number of clinical applications. Since cultured hESCs may undergo spontaneous differentiation, it is important to determine if cultured hESCs have maintained their stem cell qualities or if they have begun to acquire properties of more differentiated cells. Gene expression profiling may thus be useful for assessing cultured hESCs since it has been used to identify genes that are uniquely expressed by hESCs [29]. Player et al have found that 1715 genes were differentially expressed between hESCs and differentiated embryonic cells [29]. This finding is likely to be useful in determining if cells in culture have maintained their embryonic stem cell characteristics.

Embryonic stem cells must be differentiated before they can be used clinically. One of the first steps in the differentiation of hESCs into mature cells and tissues for clinical use is the production of embryoid bodies (hEBs), which involves the aggregation of embryonic cells but the prevention of the differentiation of cells into germ lines by plating them onto a non-permissive substrate. After these hEBs are isolated, they can be induced to generate several different types of cells including hematopoietic cells, neuronal, myogenic, and cardiac muscle cells. A comparison of gene expression profiles of hESCs and hEBs has found 194 genes whose expression was more than 3-fold greater in hEBs than in hESCs [30]. This unique set of genes should also be useful in assessing hESCs differentiation.

hESCs and hEBs also have unique miR expression profiles. Unique miRNA expression patterns have been identified in undifferentiated hESCs. A comparison of miRNA expression profiles among samples from hESCs and hEBs and five different adult cell types revealed that 104 miRNAs differentially expressed miRNA [31]. Seven miRs within miR-302 cluster on chromosome 4 and 21 miRs within the miR-520 cluster on chromosome 19 were highly expressed in undifferentiated hESCs compared to hEBs and the adult cells. The members of these 2 clusters share a consensus 7-mer seed sequence and their targeted genes had overlapping functions. Therefore, miR in the miR-302 and miR-520 clusters are likely to be biomarkers for hESC differentiation. Interest in reprograming adult cells to function back to pluripotent stem cells, induced pluripotent stem (IPS) cells, for regenerative medicine applications is growing and gene and miR expression analysis will also likely be effective in assessing these cells.

The production of clinical cellular therapies often involves multiple centers. Cells such as PBMCs are collected from a donor or patient at one center and is then shipped to a specialized cellular therapy laboratory where they are processed and become DCs. When production of the cellular therapy at the centralized center is complete, the product is shipped to the site where it is administered to a specific patient. But the centralized cell processing laboratory may be hundreds or thousands of miles from the collection center and it may even take 48 hours or more after the collection is complete for the starting cellular material to reach the cell processing laboratory.

To evaluate whether immature DCs (iDCs) produced from fresh PBMCs are similar to those produced from PBMCs that have been stored for 48 hours at 4℃ and treated with GM-CSF and IL-4, we evaluated gene expression profiles of both types of iDCs [32]. Global gene expression analysis revealed that 48 hours of storage at 4℃ had some effects on the PBMCs, but had no effect on the iDCs. The striking similarity in gene expression among iDCs derived from fresh and stored PBMCs suggests that there is no difference in the function of these cells and that the changes in PBMCs during storage did not effect the ability of these cells to be used to produce iDCs.

Some adoptive immune therapy studies have used iDCs, however, mature DCs (mDCs) are increasingly being used for clinical immune therapy because they express increased levels of co-stimulatory molecules, produce greater quantities of cytokines, and are superior in the stimulation of cytotoxic T-cell response compared to iDCs [34]. mDCs with effective anti-tumor responses have high levels of expression of co-stimulatory molecules CD80, CD83, and CD86, have high migratory activity in response to the lymphoid organ chemokines CCL19 and CCL21, [35, 36] and have high capacity to produce interleukin-12p70 (IL-12p70) which is the major factor driving Th1 reactions [37, 38].

To produce mDCs, imDCs are treated with various stimuli known to induce DC maturation [37, 39, 40]. DC maturation can be induced using inflammatory signals such as tumor necrosis factor (TNF), interleukin-1beta (IL-1β), or bacterial derivatives such as lipopolysaccharide (LPS). Double stranded RNA, interferons and postaglandins can also induce DC maturation [41, 42]. The "gold standard" DC maturation cocktail has been IL-1β, IL-6, TNF-α, and prostaglandin E₂ (PGE₂) [43]. This cocktail generates mDCs with high co-stimulatory and migratory functions, however they produce relatively low levels of IL-12 [44]. As a result, a number of ex vivo maturation cocktails have been developed for clinical use that generate mDC that produce higher levels of IL-12p70 than the IL-1β, IL-6, TNF-α and PGE₂ cocktail [37].

LPS is a toll-like receptor (TLR) ligand and a microbial compound originating from the outer membrane of the gram-negative bacterial cell wall. It is being used as a DC maturation stimulant since it is a good inducer of IL-12p70 production [38, 40]. LPS stimulates Toll-like receptor 4 (TLR4), mediates the activation of nuclear factor-kappa B (NF-kB) and mitogen activated protein kinases (MAPKs), and finally initiates DC maturation [45, 46]. Another important DC maturation stimulant is IFN-γ which enhances the TLR signaling pathway and up-regulates DC inflammatory cytokines. It has been also reported that IFN-γ can significantly augment IL-12p70 production [47].

In order to determine if the maturation of DCs by the combination of LPS plus IFN-γ can be improved by the addition of IL-1β or IL-1β plus TNF-α, we compared mDCs produced from PBMCs with these agents. Monocyte-derived DCs were prepared by the elutriation of monocytes from PBMCs followed by incubation in GM-CSF and IL-4 to produce iDCs. To produce mDCs, the iDCs were stimulated with LPS plus IFN-γ; LPS, IFN-γ, plus IL-1b; and LPS, IFN-γ, IL-1β plus TNF-α. All three DC maturation cocktails produced mDCs that expressed greater levels of commonly used markers of DC maturation: CD80, CD83, CD86, and CCR7. In addition, there were no differences in the production of IL-12p70 and IL-10 by the mDCs produced with the 3 different cocktails.

Although traditional assays used to assess DCs, flow cytometry and ELISA, did not reveal any differences between the three groups of mDCs, the function of DCs is complex and still not completely understood. It is likely that DC function is influenced by many other factors in addition to the expression of CD80, CD83, CD86, and CCR7 and the production of IL-12p70 and IL-10. To more thoroughly compare the mDCs produced with the three maturation cocktails, we analyzed the mDCs using gene expression microarrays with more than 36,000 probes. While we found that 9,576 genes were differentially expressed between iDCs and mDCs, the expression of only 13 genes differed between the three groups of mDCs. Multivariate analysis also suggested that there were more differences between DCs from different donors than between maturation cocktails [33]. These results thus show that there is no benefit of adding IL-1β and TNF-α to the LPS plus IFN-γ DC maturation cocktail.

Potency assays used for HSC products should measure the ability of the product to reconstitute bone marrow hematopoietic cells and peripheral blood cells in the transplant recipient. The potency assay should also reflect the period of time that neutrophil, platelet, and red blood cell counts return to and remain above specified levels independent of transfusion therapy.

Liquid culture of long-term culture initiating cells (LTC-IC) and the repopulation of marrow in nonobese diabetic (NOD)/severe combined immunodeficiency (SCID) mice assays are considered to be the best measure of the quantity and quality of hematopoietic stem cells. However, these assays require highly specialized reagents, a highly trained staff, and several weeks to complete. As a result, these assays have seldom, if ever, been used as potency assays. The measurement of myeloid, erythroid, and mixed colony formation in methylcellulose culture systems has been used for assessing bone marrow and PBSC HSC products, however, these assays take approximately 14 days to complete and do not produce reliable quantitative data. Traditionally, total nucleated cell counts were used to assess the potency of bone marrow and are still used as a measure of potency of umbilical cord blood components prepared for transplantation. The measurement of CD34+ cells by flow cytometry has, however, become the universal assay for measuring the potency of HSC products. One limitation of the use of CD34+ antigen expression as a potency marker is that HSCs expressing CD34 antigen do not represent a homogenous population. Several distinct subpopulations or phenotypes of CD34+ cells have been described [48]. Some subpopulations are more primitive while others are more likely to differentiate into myeloid cells, erythroid cells or megakaryocytes.

Recently, the agent plerixafor has been approved by the US Food and Drug Administration for use in combination with G-CSF to mobilize stem cells for autologous transplantation. The mechanisms by which plerixafor and G-CSF alter HSC trafficking and mobilization are different, suggesting that HSCs with different intrinsic properties may be mobilized by these agents. Plerixafor, as a CXCR4 antagonist, mobilizes HSCs within 6 hours [49-51] by disrupting the engagement of stem cell surface CXCR4 [49, 52, 53] with its ligand SDF-1 (CXCL12) which is expressed on marrow osteoblast [54, 55]. In contrast, G-CSF mobilizes stem cells indirectly by down-regulating the expression of CXCL12 on marrow osteoblasts and by releasing neutrophil and monocyte proteolytic enzymes including neutrophil elastase, cathepsin G, and matrix metalloproteinase-9 which in turn degrade important HSC trafficking and adhesion molecules c-kit, VCAM-1 and CXCR4 [56].

To explore the possibility of using gene and miRNA expression profiles assess HSC potency, we used a rhesus macaque HSC-mobilization model to compare CD34+ cells mobilized by G-CSF, by plerixafor, and by the combination of G-CSF and plerixafor. The CD34+ cells were compared using global gene and miR expression analysis. While many of the same genes were up-regulated in CD34+ cells mobilized by either plerixafor or G-CSF, there were also a large number of genes expressed that differed [57]. G-CSF-mobilized CD34+ cells were more likely to express neutrophil and mononuclear phagocyte genes than those mobilized by plerixafor. In addition, G-CSF-mobilized CD34+ cells also expressed high levels of miR-155, which is expressed in maturing DCs [26, 58]. In contrast, plerixafor-mobilized CD34+ cells were more likely to express B, T, and mast cell genes. These results suggest that G-CSF mobilizes a greater proportion of HSCs that are committed toward myeloid differentiation while those mobilized by plerixafor are more likely to be committed toward B, T, and mast cells.

Surprisingly, the combination of plerixafor plus G-CSF mobilized a population of CD34+ cells that were distinct from CD34+ cells mobilized by either agent alone. CD34+ cells mobilized by plerixafor plus G-CSF were more likely to express B cell and T cell genes than G-CSF-mobilized and plerixafor-mobilized CD34+ cells. In addition, plerixafor plus G-CSF-mobilized CD34+ cells were also remarkable for their increased expression of miR142-3p and miR-142-5p. miR-142-3p and miR-142-5p are highly expressed in T cells [59] and miR-142-3p has been found to be increased in childhood B-cell precursor acute lymphoblastic leukemia [60]. These results thus suggest plerixafor plus G-CSF mobilizes a greater proportion of B and T cell precursors than either G-CSF or plerixafor alone.

Other studies have also found that plerixafor, G-CSF, and plerixafor plus G-CSF mobilize different types of CD34+ cells. In a mouse model, Pitchford and colleagues have found that G-CSF mobilized greater quantities of hematopoietic progenitor cells than plerixafor while plerixafor mobilized greater quantities of endothelial progenitor cells than G-CSF [61]. They also found that the combination of G-CSF plus plerixafor mobilized greater quantities of hematopoietic progenitors than either agent alone, but not endothelial progenitors. Additionally, Jin et al have found that HSCs in healthy subjects mobilized with plerixafor or plerixafor plus G-CSF differ from those mobilized by G-CSF alone [62]. Furthermore, flow cytometry analysis by Rettig and colleagues of CD34+ cells mobilized by Plerixafor alone identified a unique CD34^dimCD45RA+ subset that was not mobilized by G-CSF alone [63]. The results of these studies demonstrate a role for gene and miR expression analysis for HSC potency testing.