In this study, the long-term storage effects of Cryopreservation and Transfix on peripheral and brain-derived immune cell percentages and proportions in prospectively collected CSF samples were assessed by flow cytometric analysis and compared to freshly processed CSF (current gold standard). Unlike previous studies which only cryopreserved lumbar CSF from adults16, 17, our study was the first to attempt long-term cryopreservation of ventricular CSF obtained from children with CNS infections. Ventricular CSF is pauci-cellular compared to lumbar CSF, but more demonstrative of ongoing disease processes at the site of disease2. In keeping with this observation, this is the first study to detect brain-derived cells (microglia and astrocytes) in cryopreserved CSF. Given that the role resident brain cells play in pathology is largely poorly understood, these findings are promising for future studies focused on elucidating site-of-disease pathophysiological processes.
Significant differences in viability, CD11b++, CD45+CD11b+, and CD3+ cell percentages between Fresh and Cryopreservation methods were only observed in experimental samples and not patient samples. Although significant, these differences were small, and, in some cases, median values were higher in Cryopreserved CSF than in Fresh CSF. Furthermore, these cell populations could still be accurately and clearly defined in cryopreserved CSF. Overall, the Cryopreservation method demonstrated highly comparable results to the Fresh method in both patient and experimental samples with small variation in the results and little bias in the variability (respective mean differences of 3.19 and 0.13). In combination these results suggest that cryopreservation may serve as an acceptable alternative method to fresh processing of samples.
Transfix yielded significantly lower percentages of many cell populations in both patient and experimental samples, certain sub-populations were no longer identifiable, and the Bland Altman analysis suggested large variation in the results with clear biases in variation for specific cell types from the Transfix method. This did appear to be related to the duration of storage; the maximum storage periods used in previous studies were 18–72 hours11, 19, and our 24 hour storage of CSF in Transfix yielded significantly higher percentages of CD45+, CD11b++, CD161+, and MAIT cells relative to Fresh CSF. This is similar to the findings of De Jongste et al who reported significantly higher absolute counts of lymphocytes in Transfix-treated CSF after 18 hours of storage compared to CSF collected in serum-containing media11. Nonetheless, the Fresh method still yielded significantly higher percentages compared to the 24 hour storage time for CD45+CD11b+ (non-microglia), and B cells, and clear separation of these cells were not possible after 1 week of storage in Transfix. These results suggest that Transfix, even following 24 hour storage, may not be a suitable cell stabilizing agent for markers of major cell populations such as B cells, NK cells, and monocytes, and minor cell populations including MAIT cells or activation markers. Additionally, due to the fixative in the Transfix, no additional functional assays can be performed on CSF samples.
Significantly lower MFI values were observed in cryopreserved and Transfix-treated CSF, however, the difference in mean ranks between the two separate comparisons was greater in Fresh vs Transfix than in Fresh vs Cryopreservation methods. The lower MFI could be due to the length of storage time of cells in cryo-solution (1 month) and Transfix (2 weeks), the freeze-thaw cycle in the Cryopreservation method, and the additional wash steps included in both methods which may have caused a loss of epitopes on the cell. Fixation with formaldehyde, although advantageous, can mask epitopes through crosslinking and make it more difficult for antibodies to bind to their target sites. All three methods included a fixation step, but as shown in a previous study20, the timing of this step appears to be of significance. The fixation step in both Fresh and Cryopreservation methods was included after antibody staining, the difference in mean ranks was smaller between these methods; whereas in Transfix, fixation occurs immediately once CSF is added to the tube, and a greater difference in mean ranks was observed. Therefore, fixation prior to antibody staining may reduce available epitopes on cell surfaces. The loss of fluorescent signal in cryopreserved CSF did not, however, appear to negatively impact on determining cell proportions and cell populations could be easily distinguished, unlike in Transfix-treated CSF.
Low percentages of viable cells were observed in freshly processed and cryopreserved CSF in both patient and experimental samples. This was an unexpected finding, a possible explanation for this in patient samples could be the presence of exotoxins within CSF. Most exotoxins are polypeptides produced by pathogenic Gram-positive and Gram-negative bacteria21 that are responsible for inducing apoptotic and- or necrotic cell death of host immune cells during infection22. Thus, the low cell viability observed in patient samples may be a result of cells having undergone cell death prior to CSF sampling and may therefore, not be a reflection of systematic error introduced by the processing methods. The addition of streptomycin/penicillin to media and cryo-solution may improve cell viability, these antibiotics were included in the serum-containing media used by De Graaf et al who reported improved viability in CSF cells10. Low cell viability in the experimental samples may have been a result of the multiple freeze thaw cycles; the first occurred when the PBMCs were isolated and cryopreserved for subsequent spiking, and the second was during the Cryopreservation method.
Low percentages of GFAP+ astrocytes were quantified in experimental CSF samples, which was unexpected given that these samples were spiked with peripheral leukocytes only. Although the CSF samples were obtained from patients with non-infectious CNS conditions, 85% of this cohort had hydrocephalus. Hydrocephalus is a condition which develops from the excessive accumulation of CSF in the ventricles which increases the intracranial pressure23. Significantly elevated levels of GFAP have been previously found in the CSF of patients with hydrocephalus, which may be indicative of reactive astrogliosis in response to raised pressure, especially affecting the ependyma (tissue surrounding the ventricles)24. The presence of GFAP+ astrocytes in our experimental CSF samples could, therefore, be a result of the brain’s response to hydrocephalus.
Results in experimental samples demonstrated greater differences between CSF processing methods relative to patient samples, possibly because of smaller patient sample numbers, which may have masked significant differences in cell percentages, and because PBMCs underwent two freeze-thaw cycles which may have contributed to cell death.
While this study is the first to report the effects of cryopreservation on CSF cells, there are certain limiting factors. The small sample size for patient samples may have resulted in significant differences in cell percentages being missed. Experimental samples represent an artificial environment and the PBMCs used to spike the CSF were drawn from healthy adult volunteers, these cells may differ from paediatric patients that have an ongoing infectious processes. Absolute numbers of cell populations were not assessed using true count beads in this study; however, the aim of this study was to assess whether cell proportions and phenotypes could be accurately identified following cryopreservation or Transfix storage. CD11b was selected as a marker for microglia but it is not specific to brain-derived microglia, and is also expressed by lymphocytes, monocytes, and neutrophils, likely explaining the presence of CD11b + cells in experimental samples. A more specific marker, such as Transmembrane Protein (TMEM) 119, could be considered for future studies. Furthermore, a loss in neutrophils was observed in cryopreserved CSF, therefore subsequent gating was not performed.
In conclusion, this study shows that cryopreservation is an acceptable alternative where fresh CSF processing is not feasible. Downstream functional assays should be feasible in cryopreserved CSF, unlike Transfix, seeing as cells are isolated from the supernatant first before the addition of cryo-solution. This capacity increases translational research potential, particularly in countries faced with high disease burden of CNS infections and limited laboratory resources and may be extended to a broad spectrum of paediatric CNS conditions where the application of flow cytometry is currently limited by resource constraints and low cell counts. This method also allows CSF samples to be easily transported between sites in multi-centre studies and shared between research units, promoting collaborations between institutions.