measuring lymphocyte kinetics in tropical field settings

Transactions of the Royal Society of Tropical Medicine and Hygiene (2005) 99, 675—685

Measuring lymphocyte kinetics in tropicalfield settings

Hala Ghattasa, Bakary M. Darboeb, Diana L. Wallacec,George E. Griffina, Andrew M. Prenticeb,d, Derek C. Macallana,∗

a Department of Cellular and Molecular Medicine, St George’s Hospital Medical School,Cranmer Terrace, London SW17 0RE, UKb Medical Research Council Laboratories, Keneba, The Gambiac Edward Jenner Institute for Vaccine Research, Compton, Newbury, Berkshire RG20 7NN, UKd Medical Research Council International Nutrition Group, London School of Hygiene and TropicalM

R

0d

edicine, Keppel Street, London WC1E 7HT, UK

eceived 3 December 2004; received in revised form 2 February 2005; accepted 7 February 2005

KEYWORDST-lymphocyte;Kinetics;Stable isotope labelling;Cell sorting;MACS;The Gambia

Summary Studies involving in-vivo labelling of lymphocyte DNA with 6,6-2H2-glucose to track T-cell turnover have contributed to understanding lymphocytehomeostasis in health and disease. Applying such studies in tropical settings (wherediseases that affect T-cells are prevalent) requires protocol modifications includingnon-intravenous label administration, applicability in outpatient facilities, and T-cell sorting methods independent of a fluorescence activated cell sorter (FACS).Such protocols were validated in UK pilot studies and applied in The Gambia.Healthy adult subjects (n = 12) were recruited from three Gambian villages. 6,6-2H2-glucose was administered orally in an outpatient clinic and T-cell subpopulationsisolated from peripheral blood using plastic adherence, and MultisortTM magneticcell sorting (MACSTM) to obtain CD8 +CD45R0+, CD8−CD45R0+, CD8 +CD45R0− andCD8−CD45R0− subsets. To achieve high cell purity and yield, CD45R0— cells werereincubated with CD45R0 beads. T-cell proliferation and disappearance were quan-tified using gas chromatography mass spectrometry. Results were consistent withthose of other studies showing higher turnover in memory (CD45R0+) than in naıve(CD45R0−) T-cell subsets, and an association between recent cell proliferation andsusceptibility to cell death. Cell kinetics research is applicable in tropical settings,and can contribute to further understanding the regulation of adaptive immunity inresponse to infections and other insults.© 2005 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd.All rights reserved.

* Corresponding author. Tel.: +44 20 8725 0283; fax: +44 20 8725 3487.E-mail address: [email protected] (D.C. Macallan).

035-9203/$ — see front matter © 2005 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.oi:10.1016/j.trstmh.2005.02.005

676 H. Ghattas et al.

1. Introduction

The measurement and description of lymphocytekinetics is an important component for the under-standing of lymphocyte homeostasis in normal anddisease states. Several pathological and physiolog-ical conditions affect lymphocyte numbers and dis-tribution. How this occurs has been the subjectof studies investigating the underlying dynamics oflymphocyte pools (Macallan et al., 2003a, 2004;McCune et al., 2000; Wallace et al., 2004). Method-ologies have been devised and studies conductedin developed countries, but the need for suchapproaches may be more imminent in developingcountries, where up to 59% of mortality is relatedto infectious causes (WHO, 2004) and where fac-tors that influence immunity, such as HIV, malaria,helminthic infections, TB and malnutrition, are farmore prevalent. There is therefore an imperativeto formulate novel investigative tools applicable toresearch in resource-limited settings.

There is evidence for altered leukocyte composi-tion and lymphocyte phenotype in African popula-tions (Tsegaye et al., 1999, 2003; Tugume et al.,1995; Worku et al., 1997), where people were

In order to achieve this, two aspects have beenaddressed: label administration and cell process-ing. In the first phase of lymphocyte kinetic studies,plasma glucose label enrichment is maintained at ameasurable constant level. Previously, this has beenachieved using a 24-hour primed intravenous infu-sion of 6,6-2H2-glucose (Macallan et al., 2003a).Such intravenous administration of label is lim-ited by the need for sterile preparation, in-patientovernight facilities and clinical supervision, as wellas by the cultural acceptability of venesection andcanulation. We therefore sought to develop a pro-tocol for oral administration of deuterated glucose,which would keep blood glucose enrichment lev-els as constant as possible, and allow estimation ofplateau enrichment levels from minimal blood sam-pling.

Studies have shown that T-cell phenotype defineskinetically heterogeneous cell populations. In bothyoung healthy controls and in the elderly, CD8+and CD4+ CD45R0+ memory lymphocyte popula-tions have faster proliferation rates than theirrespective CD45R0− (naıve) counterparts (Macallanet al., 2003a, 2004). Fluorescence activated cellsorting (FACS), utilized in several previous stud-i2mbetlauftf

avorlict

2

PaitwL

found to have low white blood cell counts, reducedCD4/CD8 ratios and increased activated CD8+ cellscompared with populations in developed countries.Such an effect could reflect immune adaptation inresponse to chronic exposure to certain infections(Worku et al., 1997), less common in developedcountries, or may arise from genetic differences.

T-cell pool size and phenotypic distribution arethe result of the underlying kinetics of T-cells,and immunocompetence depends on how T-cellpools are up- or downregulated in response toan antigenic stimulus. Specifically, whether thisoccurs within discrete pools of cells and how theimmune system integrates new antigen-specificmemory cells without continually increasing overallpool size, and retaining diversity and functionality,remain unanswered questions.

Several different approaches have been takenfor measuring in-vivo cell kinetics in humans(Hellerstein, 1999; Hellerstein et al., 1999;Macallan et al., 1998; Michie et al., 1992; Neeseet al., 2001). We have recently used a stable isotopelabelling approach using deuterium-labelled glu-cose, which is incorporated into the DNA of dividingcells (Macallan et al., 1998, 2003a, 2003b; Wallaceet al., 2004). Although highly equipped dedicatedclinical research facilities do exist in many develop-ing countries, which would enable the applicationof published methods, a wider-scale applicationwould require the development of simpler proto-cols.

es (Macallan et al., 2003a, 2004; Neese et al.,001, 2002), is a high-cost method inaccessible inany clinical/field settings. Such approaches haveeen used because of the requirement for high lev-ls of cell purity, as contamination with rapidlyurning-over cells may give falsely high levels ofabelling in slow turnover populations. Alternativepproaches, not requiring a cell sorter, include these of antibody-coated magnetic beads. We there-ore set out to establish a cell separation protocolhat would achieve high purity and yield of theseour different T-cell subsets.Achieving such a methodological approach would

llow the investigation of lymphocyte kinetics in di-erse disease states and settings. The measurementf lymphocyte kinetics (proliferation and deathates) is important in the understanding of normalymphocyte physiology and the changes that occurn T-cell pools in response to different immunologi-al stresses (e.g. pathogen exposure, chronic infec-ion, immunosuppression, malnutrition).

. Materials and methods

ilot studies were conducted in the UK to validaten oral dosing regime for 6,6-2H2-glucose admin-stration and a magnetic cell sorting protocol forhe separation of T-lymphocytes. These approachesere then applied at the Medical Research Councilaboratories in Keneba, The Gambia.

Measuring lymphocyte kinetics in tropical field settings 677

2.1. Pilot studies

Four healthy young control subjects (all male, agerange 18—24 years) were recruited into pilot studieswith informed consent according to local researchethics committee approval. A 10ml baseline bloodsample was withdrawn for measurement of base-line glucose and DNA enrichment. Subjects thenreceived 0.63 g/kg bodyweight 6,6-2H2-glucose asan oral solution (20% w/v in distilled water) admin-istered in half-hourly doses over a period of 10 hfollowing a prime equivalent to 2 h dosage. Dur-ing the protocol, subjects received a diet consist-ing of small, low-energy meals at 2.5 h intervalsstandardized for calorie and carbohydrate content(200 kcal permeal, including 15 g carbohydrate and8 g fat). In order to assess constancy of labelling,2ml venous blood were taken at 2-hourly intervalsduring the period of glucose administration.

Plasma glucose enrichment was measured fol-lowing derivatization to the aldononitrile acetatederivative with hydroxylamine/pyridine and aceticanhydride, monitoring ions m/z 328 and 330 by gaschromatography mass spectrometry (GCMS; Agilent5973/6890, HP-225 column).

3c(ustAtC

pttyae

2

Acsivar2nte

within the separation columns. A small macrophageor monocyte contamination within the lymphocytepopulation would lead to a large error in estimationof lymphocyte proliferation rates. The adherencestep was therefore necessary to ensure that thecell suspension to which magnetic beads would beadded was free of monocytes and macrophages.

Two cell-sorting approaches were assessed withthe aim of separating PBMC into CD8 +CD45R0+,CD8−CD45R0+, CD8 +CD45R0− and CD8−CD45R0−T-cell subsets. Initial separations using dual param-eter magnetic cell sorting technique (Partingtonet al., 1999), with two different sized magneticbeads and magnetic separators achieved purities of62% for CD4 +CD45R0− and 65% for CD8 +CD45R0−cell populations. Similarly, the use of the MACStechnique (Miltenyi Biotec) with single incubationsand single column passages failed to achieve ade-quate purities (>91% for CD4 +CD45R0− and >95%for CD8 +CD45R0−). A modified protocol was there-fore developed.

Non-adherent cells, resuspended in PBS/BSA/EDTA, were incubated with CD8 multisort beads andsorted on a mini-MACS magnet (Miltenyi Biotec).The depleted population, consisting of CD8− cells,watrwtttsCpCptwAs

t1psbCtCpCP

t

Follow-up blood samples were taken at daysand 10 post-labelling. Ficoll-Paque (Pharma-

ia) isolated peripheral blood mononuclear cellsPBMC) from 30ml heparinized blood were sortedsing a MoFlo cytometer (Cytomation). Cells weretained with CD3-RPE (Serotec), then anti-CD8-bio-in (Serotec) + Streptavidin-Allophycocyanin (Sav-PC) (PharMingen) and CD45R0-RPE-CY5 (Serotec)o yield CD8 +CD45R0+, CD8−CD45R0+, CD8 +D45R0− and CD8−CD45R0− subsets.T-cell subset DNA enrichment was measured as

reviously described by DNA extraction, digestiono nucleosides, SPE purification, derivatization tohe aldononitrile acetate derivative with hydrox-lamine/pyridine and acetic anhydride, and GCMSnalysis monitoring ions m/z 198 and 200 (Macallant al., 1998, 2003a).

.2. Magnetic cell sorting

fter PBMC isolation by Ficoll-Paque gradient,ells were resuspended in RPMI (5% fetal calferum) and incubated at 37 ◦C, 5% C02 for 75minn order to remove plastic adherent cells. Pre-ious experiments have shown that monocytesnd macrophages have extremely fast proliferationates compared with lymphocytes (Neese et al.,001). Due to the small size of the MultisortTM mag-etic cell sorting (MACSTM) beads (Miltenyi Biotec),here is a possibility that phagocytic cells wouldngulf the beads and become positively selected

as further incubated with CD4 multisort beadsnd sorted on a mini-MACS magnet, yielding a posi-ively selected population of CD4+ T-cells. Multisortelease reagent and Stop Release (Miltenyi Biotec)ere then added consecutively and in parallel tohe CD4+ and CD8+ populations in order to removehe bound magnetic beads from cells. Cells werehen washed, incubated with CD45R0 beads andorted on a mini-MACS magnet into CD8+CD45R0+,D8 +CD45R0−, CD4 +CD45R0+ and CD4 +CD45R0−opulations. CD45R0− cells were reincubated withD45R0 beads and further separated. Cells wereassed through two columns at each separation stepo ensure all cells bound to beads were retainedithin the column (protocol detailed in Figure 1).liquots of the four subsets were retained for puritytaining.Sorted cells were counted on a haemocytome-

er using Trypan Blue dye exclusion. Aliquots of.5× 104 cells were taken from each cell subset forurity evaluation by flow cytometry (Becton Dickin-on FACScan) after staining with the following com-inations of antibodies (all from Becton Dickinson):D45RA-FITC, CD3-PE, and CD8-Cychrome, usedo assess purity of the cell subset; CD45RA-FITC,D45R0-PE and CD8-Cychrome, to ensure no doubleositive (CD45RA +CD45R0+) had been counted asD45R0− in the purity analysis; CD14-FITC and CD3-E, in order to detect any monocyte contamination.Target purities were estimated by modelling

he effects of contamination between CD45


R0− and R0+ cells using data from previ-ous studies. Minimum required purity values forthe four subsets CD4 +CD45R0−, CD8 +CD45R0−,CD4 +CD45R0+ and CD8 +CD45R0+ were calculatedto be 91, 95, 72 and 76%, respectively, in orderto achieve a less than 20% margin of error inlymphocyte proliferation rates. Greater purity wasrequired in the slow turnover (CD45R0−) subsetsthan in the high turnover subsets, because of thedisproportionate impact of a small contaminantof high turnover cells in a low turnover popula-tion, as described above for monocytes. Cell yieldtargets were dependent on the minimum detec-tion threshold for the DNA derivative by GCMS(5× 105 cells).

When cell purities and yields were assessedusing the approach described, the efficacy ofmacrophage and monocyte removal by adher-ence was found to be >99% in all samples. Mod-ification of the sorting protocol using multipleincubations and column separations yielded puri-ties in excess of minimum modelled require-ments. A typical phenotypic profile is shown inFigure 2.

Ft(aCaa(cnrfntfwptuatcrdcCf1fCt

igure 1 Flow chart for sorting protocol. (1) CD8 mul-isort beads are added to peripheral blood lymphocytesPBL) at a rate of 20�l/106 cells and incubated for 20mint 10 ◦C. (2) CD8+ cells bind to magnetic beads. (3)ells are washed in PBS/BSA/EDTA and passed throughminiMACS magnet: CD8− cells go through the columnnd CD8+ cells are retained magnetically in the column.4) The column is removed from the magnet and CD8+ells are flushed out (steps 3—4 are repeated using aew column to increase purity). (5) Multisort releaseeagent is added to CD8+ cells and incubated at 10 ◦Cor 15min. (6) The release reagent cleaves the mag-etic beads from the cells. (7) CD8+ cells are passedhrough a column to ensure removal of magnetic beadsrom the cell suspension. (8) CD8+ cells are incubatedith CD45R0 beads for 15min at 10 ◦C. (9) Cells areassed through a column: CD8 +CD45R0+ cells remain inhe column; CD8 +CD45R0− cells pass through. The col-mn is removed from the magnet and CD8 +CD45R0+ cellsre flushed out (step 9 is repeated for CD45R0+ cellso increase purity). (10) CD8 +CD45R0− cells are rein-ubated with CD45R0 beads for 20min at 10 ◦C to ensureemoval of any remaining CD8 +CD45R0+ cells and RA/R0ouble-positive cells. (11) Cells are passed through aolumn: any CD8 +CD45R0+ cells remain in the column;
D8 +CD45R0− cells pass through (step 11 is repeatedor CD8 +CD45R0− cells to increase purity). (1—11) Steps—11 are followed in parallel to separate CD4+ cellsrom the remaining cells and to separate CD45R0+ fromD45R0− cells. This is conducted by adding CD4 multisorto CD8− cells at step 1.


Figure 2 Sample phenotype of magnetically separated lymphocytes from a single timepoint in a single individual. Allsubsets were stained with CD3-PE, CD8-Cychrome and CD45RA-FITC. Dot plots are of purified ungated cell populations.

2.3. Field studies

Field studies were conducted within the Medi-cal Research Council field station in Keneba, TheGambia. Twelve healthy young (18—24 years old)men were recruited from Keneba and neighbour-ing villages by the study co-ordinator (H.G.) anda locally trained Mandinka-speaking fieldworker.Written informed consent (or thumbprint if not lit-erate) was obtained before screening and all pro-cedures were approved by the Joint Gambia Gov-ernment/MRC Gambia Ethics Committee and theScientific Coordinating Committee (SCC) of the MRCGambia.

Screening included a clinical medical historyand examination, fingerprick blood sampling forhaemoglobin and malaria parasites, and urinaly-sis for glucose. Subjects were excluded if therewere any current medical conditions, anaemia, par-asitaemia or glycosuria.

A baseline venous blood sample of 8ml wastaken before label administration for backgroundenrichment in plasma glucose and cell DNA. Sub-jects then received glucose label and a standard-ized local meal (chicken durango or fish stew) fol-lowing protocols described above for pilot studies.A further 2ml of blood were taken at time = 9 hto estimate the enrichment of labelled plasmaglucose. This was frozen, before shipping to theUK for GCMS analysis, as for the aldononitrileacetate derivative described above. Two follow-up blood samples of 28ml each were withdrawnon days 3 and 10 post-labelling. Fresh cellswere separated on site by antibody coated mag-netic beads, resuspended in RNAlater (Ambion),frozen and shipped to the UK for DNA extrac-tion, derivitization and GCMS analysis as above.Aliquots of cell subsets were stained for purityanalysis, fixed and stored at 4 ◦C, and anal-ysed on a FACSCalibur flow cytometer (Becton


Dickinson) at the MRC Laboratories in Fajara within7 days.

Published modelling approaches (Asquith et al.,2002; Macallan et al., 2003a) were adapted todescribe the appearance and disappearance oflabelled cells. In summary, this involves the esti-mation of two parameters: proliferation (p) anddisappearance (d) rate constants. Where F is themeasured fractional enrichment, � is the length oflabelling (in days), t is the time of sampling (indays), the following equations are derived:

During label administration:

F(t) = p

d(1− e−dt)

After label administration:

F(t) = p

d(1− e−d�)(e−d(t−�))

(where p≥ 0 and d≥ 0).Using non-linear least squares regression

(Levenberg-Marquardt method), these equations

3. Results

3.1. Pilot studies

Pilot studies were designed to assess labelling ofglucose in terms of sufficiency, constancy, appro-priate priming and predictability.

In terms of sufficiency, levels of plasma glu-cose enrichment achieved by a repeated oral dos-ing schedule are illustrated in Figure 3. The oraldosing protocol was able to achieve levels rangingbetween 20 and 37% enrichment (mean 27%). Suchlevels are comparable to those achieved in previousintravenous administration experiments (Macallanet al., 2003a).

In terms of constancy, the feeding protocol wasfound to have a relatively small impact on glu-cose enrichment. This was investigated in detail inone subject, in whom enrichment was measuredevery 15min following a meal, as shown by thefour timepoints between 5 and 6 h in subject 1(Figure 3). Although there was a fall in glucoseenrichment, consistent with increased postprandialglucose clearance, this was relatively small (2%).

The prime dose was found to increase enrich-mw

were used to fit the model to experimental dataand estimate p and d.

Figure 3 Glucose enrichment profiles in plasma during admihalf-hourly intervals following a priming dose in four young h

ent beyond plateau levels in subjects 2 and 4, andas therefore reduced from twice the hourly dose

nistration of small doses of 6,6-2H2-glucose repeated atealthy subjects.


Figure 4 Individual fractional enrichment curves for CD4 +CD45R0−, CD8 +CD45R0−, CD4 +CD45R0+ andCD8 +CD45R0+ T-cell subsets from pilot subjects 2—4 (a—c), and mean value from six control subjects dosed intra-venously for 24 h (d). Lymphocyte enrichment levels are comparable in range to those achieved through intravenousstudies.

to 1.8 times the hourly dose in subsequent labellingstudies.

Anticipating restrictions on blood sampling infuture field studies, the predictive value of a singletimepoint as a proxy measure for average plasmaenrichment achieved during 10-hour labelling wasevaluated. Excluding the priming period, time = 0 hto time = 1 h, a single measure at time = 9 h wasfound to predict mean glucose enrichment with abias of −1.3% and an error of 3%.

Ultimately, it is not glucose but cellular DNAenrichment that needs to be measurable. Peakenrichments were in the range of 0.01—0.7 atompercent enrichment (APE) for deoxyadenosine(where APE is a measure of the proportion ofcellular DNA that has incorporated deuterium).Correcting for glucose enrichment to derivefractional cellular labelling (F) yielded curvesfor CD4 +CD45R0−, CD8 +CD45R0−, CD4 +CD45R0+and CD8 +CD45R0+ T-cell subsets in pilot sub-

Table 1 Cell subset yields and purities in field studies (n = 12)

Cell subset CD4 +CD45R0− CD4 +CD45R0+ CD8 +CD45R0− CD8 +CD45R0+

Mean yield (×106 cells) 1.93 3.85 2.12 0.89SD 1.24 1.74 1.06 0.46Mean purity (%) 91 91 94 88SD 4 5 6 5


jects 2—4, which were comparable to those fromsix control subjects dosed intravenously for 24 h(Figure 4).

3.2. Field studies

Twelve young healthy men were studied from ascreened population of 13, one being excludedbecause of an intercurrent upper respiratory tractinfection.

Following the oral dosing regimen, averageplasma glucose enrichment measured at time = 9 hwas 23.2% (SD = 3.7%). This implies that over theperiod of 10 h of label administration, an averageof about 23% of the total circulating glucose poolconsisted of deuterium-labelled glucose. These lev-els are comparable to averages achieved in the UKand are high enough to label a sufficient propor-tion of circulating glucose for enrichment to bedetected in cellular DNA on follow-up days 3 and10.

Mean cell purities and yields achieved bythe magnetic cell sorting method are listed inTable 1. Purity targets were achieved in virtuallyall CD45R0+ samples, and in most of the CD45R0−Ta

ble

2Proliferation(p)anddisappearance(d)ratesof

lymphocytesubsetsin

UKstudiesandGam

bian

field

studiesusingoral

orintravenouslabelling

Cellsubset

CD4+CD

45R0—

CD8+CD

45R0—

CD4+CD

45R0+

CD8+CD

45R0+

Setting

nFollo

w-uppoints

Labelling

p(%/day)

d(%/day)

p(%/day)

d(%/day)

p(%/day)

d(%/day)

p(%/day)

d(%/day)

Gam

bia

122

Oral

0.57

10.02

0.96

9.14

1.97

6.63

3.08

6.80

UK—pilotstudy

32

Oral

0.29

0.00

1.01

6.62

1.63

12.15

9.91

19.71

UKa

84

IV0.59

7.25

0.45

11.80

2.65

7.37

5.09

8.83

UK

62b

IV0.58

8.00

0.53

4.66

2.43

8.26

3.88

7.47

aPreviouslypublisheddata

(Macallanet

al.,2003a).

bSamedata

setas

group3withday4andday21

timepointsexcluded

from

analysis.

samples using strict criteria from isotype controlsettings as cut-offs for quadrants on FACS analysis.All subsets were >98% CD3+.

Analysis of deuterium enrichments in deoxy-adenosine in these studies were in the range0.001—0.6% APE, similar to those achieved in thepilot studies. Modelling of enrichment data bytime yielded proliferation and disappearance ratesshown in Table 2 for the four lymphocyte subsetscollected. Such values were similar to previous val-ues published from intravenous studies in the UKusing four timepoints (Table 2). The latter data arealso shown re-analysed using two timepoints only(day 3 and day 10, n = 6). This did not substan-tially alter the proliferation rate estimates but didoverestimate the disappearance rates of one cellsubset (CD8 +CD45R0−). Oral UK studies were notperformed with the intention of a quantitative com-parison by p and d with other studies, as the groupsize was too small, but data are similar to expectedranges (Table 2).

4. Discussion

This study was the first to measure in-vivo lym-phocyte kinetics in a developing country setting.In-vivo T-cell dynamics in young Gambian menwere assessed using oral administration of 6,6-2H2-glucose and an on-site magnetic cell sorting


method to sort fresh cells into four different T-cellsubsets. We have validated and demonstrated themodifications required to published protocols toallow such studies to be performed.

Strict adherence to an oral dosing and dietaryregime was able to maintain glucose enrichmentat relatively constant levels. This enabled theuse of one blood sample to provide an accept-able estimate of plasma glucose enrichment dur-ing the labelling period. However, as plasma glu-cose enrichment is used to estimate cellular DNAenrichment, the error estimated for using a sin-gle measurement (1± 3%), would be carried overto the estimates of proliferation and disappear-ance rates and lead to increased uncertainty inthese estimates. An alternative approach would beto take smaller, more frequent, blood samples. Wehave recently validated the use of fingerprick bloodsamples, which have been found to be sufficientfor GCMS analysis of percent precursor enrichment.Such a technique allows for repeated sampling andwould reduce the error of the estimation of glucoseenrichment. Such a sampling method is advanta-geous for use in the field, where fingerprick sam-pling appears to be more acceptable than venipunc-tm

l3oe(Trdpp

tsddewogoDlsadamq

the shorter labelling period of the oral glucoseapproach used in this study, infrequently divid-ing cell populations may not incorporate suffi-cient label to be reliably measured. This meansthat where specific cells with long lifespans arethe subject of interest, a short labelling proto-col such as this will not reliably distinguish pro-liferation rates near to zero. In such situations,alternative longer labelling protocols should beconsidered.

The current study used magnetic cell sorting,hence removing dependency on a flow sorter. How-ever, cell purity assessments that require the useof a flow cytometer do need to be done in orderto ensure that antibody-coated magnetic beadsare effective, and that separated cell populationsare not contaminated with other subsets that mayaffect the accuracy of kinetic estimates. Mag-netic cell sorting experiments are also labour-intensive, and are dependent on antibodies with alimited shelf life that require constant cold stor-age, which may not be available in certain fieldsettings.

One study used a combination of Rosette-Sep andMACS methods to isolate CD4 cells for measure-mHeww(asCpipce1

ati(asacac

icco

ure, due to its routine use in the diagnosis ofalaria parasitaemia.Deuterium incorporation into the DNA of pro-

iferating cells was measurable on follow-up daysand 10. Results were consistent with those ofther studies (Macallan et al., 2003a; Wallacet al., 2004), showing higher turnover in memoryCD45R0+) T-cell subsets than in naıve (CD45R0−)-cells, and confirming the association betweenecent cell proliferation and susceptibility to celleath; d >p is interpreted as reflecting increasedropensity of recently divided cells to die or disap-ear.Oral dosing of deuterated glucose overcomes

he need for in-patient facilities and is less inva-ive than intravenous dosing. Subsequent to theevelopment of these protocols, an approach usingeuterated water (2H2O) has been described (Neeset al., 2002). The key advantages of deuteratedater include: (1) the significantly lower costf deuterated water compared with deuteratedlucose; and (2) the ease of oral administrationf deuterated water within out-patient settings.euterated water studies, however, require pro-onged commitment to the labelling part of thetudy (daily dosing for several weeks). Addition-lly, high turnover subsets may become saturateduring long-term 2H2O labelling. The oral glucosepproach only requires a short subject time com-itment for labelling and avoids saturation of fre-uently dividing cell populations. However, with

ent of deuterium incorporation and cell kinetics inIV-infected patients (Busch et al., 2004); althoughxcellent for studies of CD4 in HIV, there are muchider questions that need to be addressed thatould include effects on CD8 kinetics. Our studiesMacallan et al., 2003a) have shown large discrep-ncies in kinetics between memory and naıve T-cellubsets and demonstrate a need for separation ofD45R0− cells from CD45R0+ cell subsets. This isarticularly relevant for the conduct of such stud-es in an African setting due to the suggestion thathenotype distribution may be different in Africansompared with Caucasian populations (Tsegayet al., 1999; Tugume et al., 1995; Worku et al.,997).Although the use of an oral dosing regime

nd magnetic cell sorting techniques decreaseshe dependence on clinical and high-tech facil-ties, kinetics studies remain highly expensivedeuterated glucose and antibody costs), requirelaboratory, a flow cytometer and trained per-

onnel, and are time-consuming and hence onlyppropriate for small mechanistic studies withlearly defined groups for comparison. Sampleslso need to be shipped to facilities with GCMSapacity.Blood sampling remains a limitation in kinet-

cs studies, especially in settings where ethi-al issues may limit the volume of blood thatan be withdrawn from individuals. We couldbtain only two samples (days 3 and 10) for the


modelling of cell proliferation and disappearancerates. Although this appears to be adequate forcalculation of p, it is sub-optimal for estimationof d; more points would be required for moreaccurate estimates. Moreover, the use of a shortlabelling period of 10 h decreases the likelihoodof labelling slow turnover cell subsets, the kinet-ics of which could have been overlooked by thisstudy.

However, this study shows that in spite of thelimited number of follow-up points for the mea-surement of T-cell turnover, measured rates arewithin similar ranges to those for 24-h intravenousstudies. T-cell homeostasis in this Gambian pop-ulation appears to be extremely robust, althoughdifferences in T-cell repertoire may exist. Also,we found that for the Gambian population, as foryoung healthy individuals in the UK, there is a con-siderable physiological inter-individual variation inproliferation and disappearance rates of differentT-cell subsets, which does not appear to be disease-related.

Homeostasis of lymphocyte subpopulationsunderlies normal immune function. In the trop-ical setting, altered lymphocyte kinetics are

Busch, R., Cesar, D., Higuera-Alhino, D., Gee, T., Hellerstein,M.K., McCune, J.M., 2004. Isolation of peripheral blood CD4+T-cells using RosetteSep and MACS for studies of DNA turnoverby deuterium labelling. J. Immunol. Methods 286, 97—109.

Hellerstein, M.K., 1999. Measurement of T-cell kinetics:recent methodologic advances. Immunol. Today 20, 438—441.

Hellerstein, M., Hanley, M.B., Cesar, D., Siler, S., Papageor-gopoulos, C., Wieder, E., Schmidt, D., Hoh, R., Neese, R.,Macallan, D., Deeks, S., McCune, J.M., 1999. Directly mea-sured kinetics of circulating T lymphocytes in normal andHIV-1-infected humans. Nat. Med. 5, 83—89.

Macallan, D.C., Fullerton, C.A., Neese, R.A., Haddock, K., Park,S.S., Hellerstein, M.K., 1998. Measurement of cell prolifera-tion by labeling of DNA with stable isotope-labeled glucose:studies in vitro, in animals, and in humans. Proc. Natl. Acad.Sci. USA 95, 708—713.

Macallan, D.C., Asquith, B., Irvine, A.J., Wallace, D.L., Worth,A., Ghattas, H., Zhang, Y., Griffin, G.E., Tough, D.F., Bever-ley, P.C., 2003a. Measurement and modeling of human T cellkinetics. Eur. J. Immunol. 33, 2316—2326.

Macallan, D.C., Wallace, D.L., Irvine, A.J., Asquith, B., Worth,A., Ghattas, H., Zhang, Y., Griffin, G.E., Tough, D.F., Bev-erley, P.C., 2003b. Rapid turnover of T cells in acuteinfectious mononucleosis. Eur. J. Immunol. 33, 2655—2665.

Macallan, D.C., Wallace, D., Zhang, Y., De Lara, C., Worth, A.T.,Ghattas, H., Griffin, G.E., Beverley, P.C., Tough, D.F., 2004.Rapid turnover of effector-memory CD4(+) T cells in healthy

M

M

N

N

P

T

T

T

likely to mediate the effects of malnutrition andinfectious agents on T-cell immunity. The avail-ability of methodological approaches, such as thatdescribed in this study, creates new opportunitiesfor investigating human lymphocyte kinetics inpopulations where malnutrition and infection arecommon.

Conflicts of interest statementThe authors have no conflicts of interest concerningthe work reported in this paper.

Acknowledgements

We are grateful to those who volunteered to takepart in this study, both in Keneba and in London; toDr Yan Zhang for GCMS support; to Andrew Worthand Cathy DeLara for FACS sorts; to Becca Asquithfor mathematical advice; to Molipha Jammeh andKebba Bajo for subject recruitment and transla-tion. This work was supported by a grant from theMRC. D.M. was supported by an MRC-Glaxo Well-come Clinician Scientist Fellowship.

References

Asquith, B., Debacq, C., Macallan, D.C., Willems, L., Bang-ham, C.R., 2002. Lymphocyte kinetics: the interpretation oflabelling data. Trends Immunol. 23, 596—601.

humans. J. Exp. Med. 200, 255—260.cCune, J.M., Hanley, M.B., Cesar, D., Halvorsen, R., Hoh, R.,Schmidt, D., Wieder, E., Deeks, S., Siler, S., Neese, R.,Hellerstein, M., 2000. Factors influencing T-cell turnoverin HIV-1-seropositive patients. J. Clin. Invest. 105, R1—R8.

ichie, C.A., McLean, A., Alcock, C., Beverley, P.C., 1992. Lifes-pan of human lymphocyte subsets defined by CD45 isoforms.Nature 360, 264—265.

eese, R.A., Siler, S.Q., Cesar, D., Antelo, F., Lee, D., Misell,L., Patel, K., Tehrani, S., Shah, P., Hellerstein, M.K., 2001.Advances in the stable isotope-mass spectrometric measure-ment of DNA synthesis and cell proliferation. Anal. Biochem.298, 189—195.

eese, R.A., Misell, L.M., Turner, S., Chu, A., Kim, J., Cesar,D., Hoh, R., Antelo, F., Strawford, A., McCune, J.M., Chris-tiansen, M., Hellerstein, M.K., 2002. Measurement in vivo ofproliferation rates of slow turnover cells by 2H2O labeling ofthe deoxyribose moiety of DNA. Proc. Natl. Acad. Sci. USA99, 15345—15350.

artington, K.M., Jenkinson, E.J., Anderson, G., 1999. A novelmethod of cell separation based on dual parameter immuno-magnetic cell selection. J. Immunol. Methods 223, 195—205.

segaye, A., Messele, T., Tilahun, T., Hailu, E., Sahlu, T., Doorly,R., Fontanet, A.L., Rinke de Wit, T.F., 1999. Immunohema-tological reference ranges for adult Ethiopians. Clin. Diagn.Lab. Immunol. 6, 410—414.

segaye, A., Wolday, D., Otto, S., Petros, B., Assefa, T., Ale-bachew, T., Hailu, E., Adugna, F., Measho, W., Dorigo,W., Fontanet, A.L., van Baarle, D., Miedema, F., 2003.Immunophenotyping of blood lymphocytes at birth, duringchildhood, and during adulthood in HIV-1-uninfected Ethiopi-ans. Clin. Immunol. 109, 338—346.

ugume, S.B., Piwowar, E.M., Lutalo, T., Mugyenyi, P.N.,Grant, R.M., Mangeni, F.W., Pattishall, K., Katongole-Mbidde, E., 1995. Hematological reference ranges among


healthy Ugandans. Clin. Diagn. Lab. Immunol. 2, 233—235.

Wallace, D.L., Zhang, Y., Ghattas, H., Worth, A., Irvine, A.,Bennett, A.R., Griffin, G.E., Beverley, P.C., Tough, D.F.,Macallan, D.C., 2004. Direct measurement of T cell subsetkinetics in vivo in elderly men and women. J. Immunol. 173,1787—1794.

WHO, 2004. The World Health Report 2004 — Changing History.World Health Organization, Geneva.

Worku, S., Christensson, B., Bjorkman, A., Islam, D., 1997.Higher proportion of CD8+ T cells in the blood in healthyadults from Ethiopia and Bangladesh compared with Sweden.Trans. R. Soc. Trop. Med. Hyg. 91, 618—622.

measuring lymphocyte kinetics in tropical field settings

Documents