Brief Communication. R. BECK and P. R. L. LAM-PO-TANG. Immunology and Cell Biology (1994) 72,

Transcription

1 Immunology and Cell Biology (1994) 72, Brief Communication Comparison of cord blood and adult blood lymphocyte normal ranges: A possible explanation for decreased severity of graft versus host disease after cord blood transplantation R. BECK and P. R. L. LAM-PO-TANG Bone Marrow Transplant Laboratory. Children's Leukaemia and Cancer Research Centre. Prince of H ales Children's Hospital, Randwick, New South Wales. Australia Summary Lymphocyte subpopulations in cord blood (CB; collected at birth from full-term babies) were compared with that of adult blood (AB) and found to contain significantly different numbers and percentages of lymphocyte subpopulations. The absolute lymphocyte count was greater in CB (4.8 ±1.1 X 10^/L)than in AB (1.69 ±0.38 x 10^/L), with CB having significantly higher absolute numbers of lymphocyte subsets even though CB percentages were significantly lower. Significant differences in percentages were found between cord and adult T cells (CB 58% v.v AB 74%), NK cells (CB 19% v.v AB 7%) and their subsets. CD38, a marker of activation and immaturity, was present on virtually all cord T cells and approximately half the adult T cells. CD45RA, a marker considered to define unprimed or naive cells, was expressed on 82% of cord lymphocytes as compared with 48% in AB. CD45RO was expressed on 16% of CB lymphocytes and 49% of AB lymphocytes. Cord blood contains a higher percentage and total number of immature and immunologically naive lymphocytes than AB. Key words: cord blood, lymphocyte subsets. Introduction Bone marrow transplantation is a potentially curative treatment for a number of haematopoietic disorders. An HLA compatible sibling donor is the donor of choice when considering a transplant for haematological malignancy or marrow failure, however only 30% of patients have a compatible donor and graft versus host disease (GVHD) can be a major problem.' Alternative sources of normal stem cells for haematoiogical rescue include partially matched related bone marrow donors, or matched unrelated bone marrow donors. The limiting factor in the use of these alternative donors is the increased likelihood of GVHD and the increased morbidity and mortality associated with the transplantation procedure. Correspondence: Dr P. R. L. Lam-Po-Tang, Bone Marrow Transplant Laboratory, Children's Leukaemia and Cancer Research Centre, Prince of Wales Children's Hospital. Randwick, NSW 2031, Australia. Received 1 7 January 1994: accepted 3 March The use of umbilical cord blood (CB) as a source of stem cells for transplantation has been described in at least six published cases. Graft versus host disease grade 2 (in a two antigen sibling mismatch) was the highest grade of GVHD observed thus far.--^ This study was undertaken to evaluate why GVHD may be less severe after CB transplantation. To do this we analysed the immunophenotype of CB lymphocytes, with particular reference to cell surface markers of lymphoid immaturity and naivety. These results were compared with those obtained from adult blood (AB). Method and materials Specimens Cord blood specimens from 26 healthy babies were collected into anti-coagulant citrate dtwtrose solution formula A (ACD-A; approximately 1 ml ACD:10mL blood) after clamping and cutting of the cord and by venepuncture or direct drainage from the umbilical vein following normal

2 Cord and adult lymphocyte subsets 441 delivery. The specimens were transported to the laborator\' at room temperature and processed within 8 h of collection. Where ACD-A was used, a correction of the absolute count was made for the dilution effect. Normal peripheral blood was collected into EDTA from 26 volunteer adult blood donors (16 female, 10 male) at a Red Cross blood collection centre: the mean age of donors was 37.5 years (range years). White cell counts and differentials were performed using either a Sysmex NE8000 or KIOOO blood counter. All specimens were stored at room temperature and stained within 6 h of collection. Monoclonal antibodies All conjugated mab were purchased from Becton Dickinson (BD, Sydney, NSW, Australia). The panel of mab included: (i) FITC: CD3, 8, RA, ap TCR and IgGl y control: (ii) phycoerythrm (PE): CD4, 14, 16, 16/56, 19, 33, 38, 45RO. y6-l TCR and IgG2a y control; and (iii) peridinin chlorophyll protein (Per CP): CD3, HLA DR and IgG2a y control. Flow cytometer The flow cytometer used was a BD FACScan using LYSIS II software (BD Immunocytometry Systems, San Jose, CA. USA). Staining procedure A white cell count and differential were performed prior to staining of the whole blood and each sample was diluted with PBS so that after dilution, 100 il contained approximately 1 X 10^ cells for phenotyping. This was done to ensure saturation binding at a constant reaction volume for all specimens. The appropriate mab were incubated with 1 x 10^ cells/ tube for 10 min at room temperature. White cells were then fixed by the addition of 100 il isotonic formalin (7.5%, ph 7.25) to each tube for 60 s at room temperature, then RBC were lysed by adding I ml ol distilled water/tube and incubating the tubes at 37 C for 3 min. The cells were then washed twice in PBS and resuspended in 300 ml of 1% isotonic paraformaldehyde. Samples were stored at 4 C in the dark until flow cytometric analysis (within 24 h). A minimum of 3000 events falling within the lymphocyte gate were counted for each tube. Accuracy of staining with the mab panel was evaluated by: (i) calculating a lymphosum (T cells% + NK cells% + B cells%) and monitoring its proximity to 100%; and (ii) by monitoring non-lymphoid cells (CD14 and CD33 positive) falling within the lymphocyte gate. Statistical analysis The unpaired Student's Hest was used to evaluate statistical significance. A /'value of less than 0.05 was considered to be significant. Results were expressed as the mean ± s.d. Results The percentages and absolute counts for CB and AB lymphocyte subpopulations are summarized in Table 1. Contamination of the lymphocyte gate by nonlymphoid leucocytes was minimal. The mean expression of CD33 (a myeloid marker) within the lymphocyte gate was 3.2% for CB and 3.5% for AB, whilst CD 14 expression (a monocyte marker) was virtually absent in both CB and AB specimens. The normal range of the adult phenotype that we established is similar to that used at this hospital and acts as a control for our findings. The mean absolute lymphocyte count was higher in CB than in AB. The mean absolute number of cord T cells, including both CD4* and CD8* T subsets, was significantly greater than adult levels (/"). In spite of a lower percentage of T cells in CB, the mean absolute number of virtually all T cell subsets was significantly greater in CB than in AB. The exception was CD3^CD56' T cells, whose function is related to MHC unrestricted cytotoxicity. The CB mean percentage and absolute number of CD3*CD56* T cells was significantly lower than that observed in AB {P ). The mean percentage and absolute number of NK cell markers was significantly higher in CB (?). Interestingly, the total mean CD8 percentage was similar for both CB and AB (/*= 0.688). Percentage T cell expression of CD8 was significantly lower in CB, whilst the percentage of CB CD8 * NK cells (as defined by CDl 6) was significantly higher than AB. however, proportionally there was little difference. There were no significant differences between CD4 CDS double positive populations in AB or CB. CD38 was co-expressed on 95% and 48% of cord and adult T cells, respectively. Similarly 96% of cord CD8 * lymphocytes coexpressed CD38 whereas only 50% of adult lymphocytes co-expressed this marker. Whilst B cell percentages were not significantly different between the two cell populations (P= 0.263), CB had significantly higher absolute numbers of B cells (P). The percentage of cells expressing HLA DR was significantly higher in AB (P ). Since B ceils always express class II MHC antigens, the significantly higher percentage of HLA DR in AB can only be explained by the presence of activated T or NK cells. Eighty-two per cent of cord lymphocytes as compared with 48% of AB lymphocytes were CD45RA* CD45RO. Adult blood had a significantly higher percentage of CD45RA-CD45RO* lymphocytes as compared with CB, 40.1 and 7.1%, respectively (-P). The percentage of cells co-expressing CD45RA and CD45RO was not significantly ditlerent

3 442 R. Beck and P. R. L. Lam-Po-Tang Table I. Comparison of CB and AB lymphocyte subset normal ranges. CB AB P CB AbsLymx 10^/L AB AbsLymx 10^/L P Marker CD 3 CD4 CD3*8^ CDS total CD4*8" CDl 6^56*3- CD16 CDSM6* CD3*56^ CD38 CD3-38^ CD8-^38* CD19 HLADR aptcr y6tcr CD45RA*RO- CD45RA RO* CD45RA'RO* CD33 Lymphosum 57.7± ± ± S± ± ± ± ±4 0.15±0.l4 93.4±3.S 54.6±8.S 23.8± ± ± ± ± ± ± 5 9.0± ,8± ± ± ± ± ± ±4.l 6.6±3.8 2.S± ± ± ± ± ± ± ± ± ± ± ± ±1.S NR4.8± ± OO±O ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±0.05 NR 1.69± ± ±0.1 O.39±O.]3 0.43± ± ± ± ± ± ± ± ±0.n O.23±O.O7 0.29± ± ± ± ±0.I ± ±0.02 <0.00I Results are from theflowcytometric phenotyping of 26 cord blood and 26 adult blood samples, using a whole blood technique. Results are e.xpressed as the mean ± s.d. P < 0.05 was considered significant. The lymphosum - T ce!l% + B cells % + NK cells% (theoretical values should equal 100%). NR = normal range. Abs Lym = absolute lymphocyte count. in the two cell populations (P = 0.B,09). However, we found significantly higher absolute values in the CB as compared with AB (/^). Discussion Immunophenotypic analysis of the lymphoid subsets in CB and AB demonstrates that CB has a significantly more immature immunophenotype than AB. These findings may help identify why GVHD is less severe after CB transplantation. The mean iymphosum for AB was significantly greater than that found with CB (95.7±1.8% vs 92.8 ±4.0%, P= 0.003). The increased variability of the CB lymphosum is thought to be associated with the presence of immature non-lymphoid cells such as erythroblasts. which weakly express CD45. Even with stringent gating, a specimen that has a leucogate of greater than 98% CD45 bright cells could have a lymphosum of 90%. Overcoming contamination of the lymphocyte gate in CB whilst phenotyping is a major problem and requires the use of special lysing techniques. Red blood cells were lysed using a method developed by L. Peters (Flow Cytometry Unit, Southpath, St George Hospital, Sydney, NSW, Australia). This method was demonstrated to give a 'cleaner' CB sample than other methods with no more adverse effects on white cells (L. Peters, J. Wotherspoon pers. comm. 1993). Differences found between CB and AB included expression of CD38 on T cells. CD45RA, CD45RO and NK markers. CD3^CD56^ lymphocytes (cytotoxic MHC unrestricted T cells) were virtually absent from CB and in low numbers in AB. This is consistent with the findings reported by Rabian-Herzog et al} In contrast, the same group found less than 0.5% y5-l T cells in CB, whereas we found a definite population, albeit in a reduced percentage as compared with AB. Absolute numbers of y6-l T cells were not significantly different between CB and AB {P ). Keever et al. found a significant difference between B cell percentages in contrast to our findings, in their cord and adult NR.9 Both groups documented T cell percentages that were higher than ours.**** Some groups have published CB lymphocyte normal ranges using density gradient separated mononuclear cells.^-'*' It is known that selective cell loss occurs through the use of density gradients and this probably accounts for some of the differences.

4 Cord and adult lymphocyte subsets 443 Table 2. Comparison of published CB lymphocyte subset ranges. Marker CD3 CD4 CD8 aptcr ystcr CDI6 CD16*56*3 CD19 CD38 Beck et ai* 57.7± ± ± ±8.9 I.2± ± ± ± ±3.8 Rabian-Herzog** 44.8 ±13.3 3].0± ± < ±II ± ± 10.2 Keei/er^+ 70 ± ± 9 21± 3 Racadot'**^ 7O± ± ± ± ± ± ± ±7 Beck et al.* (Abs Lym x lovl) 2.71 ± ± ± ± ± ± ± ± ±0.98 Rabian-Herzog^* (AbsLym±IO^/L) 1.7 ±0.60 I.19± ±a ± ± ± ±0.98 Only markers common to most groups were chosen. It is assumed that the CD8 values quoted by other groups are CD8 total. *Whole blood technique; ^mononuclear cell preparation. Abs Lym = absolute lymphocyte count. All studies are in agreement with respect to expression of CD45RA on the majority of cord lymphocytes and CD38 on the majority of cord T cells. Table 2 compares our findings with those of other published CB lymphocyte subset ranges. It should be noted that only one other group (Rabian-Herzog et al.) includes absolute numbers. Our mean CB absolute lymphocyte count is 4.8± 1.1 x lo^cells/l, theirs is 4.0± 1.3 x 10^ cells/l. The values we generated for both CD3 and CD4 also differ. Little information has been provided by other groups regarding the clear definition of NK cells. We used both CD 16 and CD 16/56 and found that the CD 16 population comprises approximately 92% of total NK cells in CB and 70% in AB. No other group has clearly differentiated CD8 expression and we presume that where they present CD8 values otherwise undefined, they mean total CDS. We differentiate between total CD8. T cell CD8 and NK cell CD8 expression. The need for such discrimination is explained in part by the fact that measurement of CD8 alone ignores the existence of at least two distinct lymphocyte populations. Research by Fuchshuber and Lotzova has shown that oncolytic activity is not equally distributed amongst NK cell subsets (including CDl6^CD8 + lymphocytes) and may be tumour dependent." This is interesting in light of the possible role of NK cells in GVHD, since NK cells have been implicated as possible effectors in CrVHD.'- The decreased severity seen in CB transplantation may be due in part to NK cell immaturity, a possibility that we are investigating. The function of CD38 remains unclear, its presence is thought lo indicate immaturity or T cell activation or suppression.'-'"'-'^ Similarly the functions of CD45RA and CD45RO have yet to be fully understood. Current theory attributes T cell priming and the ability to respond to recall antigens with CD45RO expression and T cell naivety with CD45RA expression."' The co-expression of CD45RA and CD45RO may even be linked to cell cycle.'^ The controls incorporated into the study lead us to conclude that our results are accurate and the methodology sound. It would seem this study is the first to incorporate the concept of a 'lymphosum' into flow cytometric findings for lymphocytes. This type of quality control which was easily incorporated into our flow cytometric studies, shows that we were counting what we said we were counting. In conclusion our results show significant differences in lymphocyte subpopulations between CB and AB. Cord blood contains more cells with markers associated with immaturity. If phenotypic immaturity translates into decreased alloreactivity then CB. an already proven source of stem cells, may become the transplant material of choice where a familial donor is unavailable. Acknowledgements The authors thank the staff of The Royal Hospital for Women labour ward, particularly Srs B. Michie, E. Lloyd and C. Pankhurst for the collection of cord blood, and the Sydney Red Cross Blood Transfusion Centre for allowing us to use their facilities for the collection of adult blood. This work was supported by a grant from the Richard Arcus Fund. References 1. Vowels, M. R., Honcyman, M. and Ziegler. J / Paediatr. Child Health. 28: Gluckman. E.. Broxmeyer. H. E., Auerbach, A. E. et al A'. Engi J. Med. 321:

5 444 R. Beck and P R. L. Lam-Po-Tang 3. Gluckman, E., Devergie, A., Bourdeau-Esperou. H. et al Nouv. Rev. Er. Haematoi. 32: Vilmer, E., Broyart. A., Lescoeur, B., Rahimy, C, Denamur. E. and Blot, P Transplantation 53: Wagner, J. E., Broxmeyer. H. E.. Byrd. R. L. et ai Blood 19: Tiedeman. K. and Tucker, D. P Proceedings of Ihe Isl International Symposium on Cytokines in Bone Marrow Transplantation. 7. Vowels. M. R., Lam-Po-Tang. P. R. L., Berdoukas, V. el ai N. EnglJ. Med. 329: Rabian-Herzog, C, Lesage, S., Gluckman. E. and Charron, D / Hematotherapy 2: , 9. Keever, C. A., / Hemaiotherapy 2: Racadot, E., Schaal, J. P., Van Lemmens, P., Billot. M. and Herve. P / Hematotherapy 2: , 11. Fuchshuber. P. R. and Lotzova. E Lymphokine CyiokineRes. II: Johnson. B. D. and Triutt. R. L Transplantation 54: Gerli, R.. Bertotto, A.. Spinozzi, E., Cernetti, C. Grigani. E. and Rambotti. P Clin. Immunoi Immunopathol. 40: Eunaro. A., Spagnoli. G. C. Ausiello. C M. et al / Immunol. 145: Clement.. L. T.. Vink, P. E. and Bradley. G. E / Immunol. 145: Hassan, J. and Reen. D. J Res. Immunoi 144: LaSalle. J. M. and Hatler, D. A Celi Immunoi 138:

6