pre-b cells in bone marrow: size distribution profile, proliferative

Immunology 1984 51 333

Pre-B cells in bone marrow: size distribution profile, proliferative capacityand peanut agglutinin binding of cytoplasmic y chain-bearing cell populations

in normal and regenerating bone marrow

D. G. OSMOND & J. J. T. OWEN, Departments of Anatomy, University of Birmingham, Birmingham,U.K. and McGill University, Montreal, Quebec, Canada

Acceptedfor publication 22 August 1983

Summary. Pre-B cell populations in mouse bonemarrow, identified by double immunofluorescencelabelling of cytoplasmic and surface p chains (cy, sy),have been characterized by cell size, proliferativecapacity and the binding of peanut agglutinin (PNA).In the normal steady state of lymphocyte productionthe size distribution profile of cytocentrifuged cp+sju-cells was bimodal. A population of large cells in rapidcell cycle was revealed by arresting cells in mitosis withvincristine. Many cp+sp- cells, however, formed anondividing population of small lymphocytes, resem-bling sp+ cells in size distribution. During regene-ration from sublethal whole body X-irradiation (150rads) a marked enrichment of large cp+su- cellspreceded small cp+sp- and sp+ cells; progressivechanges in cell size distribution reflected a wave of Blymphocyte genesis. The cp+sp- cells in foetal liverresembled those in regenerating marrow. Surfacebinding ofPNA characterised all cp+sy- cell popula-tions in normal and regenerating bone marrow and infoetal liver, whereas only a minority of sy+ cells and,u-negative marrow cells bound PNA strongly. Thepresent size distribution analyses allow a correlation

Abbreviations: cp, cytoplasmic p chains; FCS, foetal calfserum; FITC, fluorescein isothiocyanate; LPS, lipopolysac-charide; PBS, phosphate-buffered saline; PNA, peanut agg-lutinin; sp, surface p chains.

Correspondence: Dr D. G. Osmond, Department ofAnatomy, McGill University, 3640 University Street, Mon-treal, Quebec, Canada H3A 2B2.

with other cytological and functional studies of mar-row lymphocyte precursors in defining the place ofpre-B cells in B lymphocyte genesis.

INTRODUCTION

The production of B lymphocytes in murine bonemarrow maintains a pool ofshort-lived primary B cellsin the peripheral lymphoid tissues (Osmond, 1980).Primary humoral immune responsiveness thusdepends upon the continuous proliferation and differ-entiation of B lymphocyte precursors, pre-B cells, inthe marrow. These precursor cell populations remainill-defined, however, despite the use of various criteriato assay them.The first evidence for the existence of precursors

producing large numbers of lymphocytes in the mar-row was provided by radioautographic DNA labellingof individual cells in vivo and in cultures (Osmond &Everett, 1964; Brahim & Osmond, 1970; Yoshida &Osmond, 1971). In cultured marrow cell fractions,lymphocyte precursors were identified within a popu-lation of large lymphoid cells visualized in radioauto-graphic stained smears (Yoshida & Osmond, 1971).Analyses of the size distribution ofDNA synthesizingcells revealed populations of dividing large lymphoidcells and non-dividing small lymphocytes, demon-strating a precursor-product relationship betweenthem. The development of surface IgM molecules on B

333

D. G. Osmond & J. J. T. Owen

cells did not coincide with the transition from dividingprecursors to non-dividing small lymphocytes, how-ever, but developed after a distinct post-mitotic period(Osmond & Nossal, 1974) together with other Blymphocyte surface markers (Yang, Miller &Osmond, 1975). Further analysis of lymphoid precur-sors based on morphological criteria alone was limitedby their heterogeneity in proliferative and hemopoieticpotential (Miller & Osmond, 1973; Rosse, 1976), andby the lack of specific markers for B lineage precur-sors.Double immunofluorescence labelling of cells from

bone marrow and sites of foetal B cell genesis hasidentified individual cells containing cytoplasmic pchains (cy) without detectable surface membrane pchains (sp), as presumptive late B lineage cells follow-ing the rearrangement and expression of immunoglo-bulin pu heavy chain genes (Raff et al., 1976; Owen etal., 1977b; Cooper & Lawton, 1979). Such cp +sy-pre-B cells are heterogeneous in size and in prolifera-tive capacity, as shown by [3H]-thymidine uptake(Owen et al., 1977b; Burrows et al., 1978; Pearl et al.,1978; Landreth, Rosse & Clagett, 1981), but theconstituent populations have not been fully character-ized. A proposed large to small cell sequence ofdevelopment of cp+sp,- cells and their role in thegenesis of sp+ small B lymphocytes remain based oncompelling but circumstantial evidence.

Pre-B cells have also been assayed indirectly by theircapacity to give rise to functionally responsive Blymphocytes after an appropriate maturation periodeither in vivo or in vitro (Lafleur, Miller, & Phillips,1972; Melchers, 1977; Paige et al., 1981). These assayshave analysed pre-B cell populations by their sedimen-tation velocity at unit gravity (Lau et al., 1979) and byhydroxyurea deletion to reveal dividing cells (Rus-thoven & Phillips, 1980; Freitas et al., 1982). Whilethey define some properties of pre-B cell populations,notably their relative cell size distribution as reflectedby sedimentation velocity profiles, these assays do notdistinguish individual pre-B cells from other cell typesin the same sedimentation fractions nor the prolifera-tive sequence of cell generations in B lymphocyteproduction.

Cytoplasmic p chains provide a direct cytologicalmarker for B lineage cells, yet such cell populationsremain less well characterised than, and difficult toequate with, those defined by the other precursor cellcriteria. The present work analyses the size distribu-tion profiles of cp+sp- and sp+ cells in mouse bonemarrow, as detected by double immunofluorescence

labelling and measurement of individual cells understandardised cytocentrifuge conditions, as well as theproliferative subset of cp+sp- cells identified aftermetaphase arrest by vincristine. Cu+sp- cell popula-tions in the steady state of B cell genesis in normalmouse marrow are compared with those propagatingwaves of activity in regenerating marrow and in foetalliver. In each case, the binding of the lectin, peanutagglutinin (PNA), to individual cells is correlated withp chain expression, as a potential means ofdistinguish-ing immature lymphoid cell populations (Reisner,Linker-Israeli & Sharon, 1976; Rose et al., 1980;Newman & Boss, 1980). The results characterizefurther the cp,+sp- cell populations in bone marrow,indicating the place of these cells in a scheme of Blymphocyte differentiation and providing a basis forcomparison with other pre-B cell assays.

MATERIALS AND METHODS

AnimalsMale and female CBA mice were used at 7-9 weeksafter birth and during foetal life at 14-19 days ofgestation.

CellsBone marrow cells were flushed from the femoralshafts of pairs of mice, suspended in HEPES-bufferedRPMI- 1640 with 20% foetal calf serum (RPMI-FCS),pooled, washed by centrifugation through FCS at 4°and resuspended in RPMI-FCS. Foetal liver cells weretreated similarly after teasing the organ inRPMI-FCS.

p-Chain labellingTo label cell surface p chains, cell aliquots (2 5 x 106)were centrifuged to a pellet, resuspended in 20p1 anti-pantibody conjugated with fluorescein isothiocyanate(FITC) (Meloy Laboratories, Springfield, VA; 1: 15dilution), incubated on ice for 30 min and centrifugedtwice through FCS. To label intracytoplasmic p chainsin addition to sp, as described (Raff et al., 1976),aliquots of cells (1-5 x 105) were next cytocentrifugedonto microscope slides, fixed in pre-cooled 5% aceticacid in ethanol (40, 10 min) and washed by threeimmersions in phosphate-buffered saline (PBS; roomtemperature, 15 min). After removing excess PBS,each cytospot was covered with 10 p1 anti-p antibodyconjugated with rhodamine isothiocyanate (affinitycolumn purified anti-p antibody supplied and rhoda-

334

Pre-B cells in bone marrow

mine conjugated by Dr M. D. Cooper, Birmingham,AL) incubated at room temperature for 30 min in ahumidified chamber and washed in three changes ofPBS. Cytospot preparations were mounted in 30%glycerol in PBS, sealed and examined by epifluores-cence with lOOX oil immersion objective for thepresence of individual cells with rhodamine fluores-cence alone (cp only) and with double labelling byfluorescein plus rhodamine (sp).

Mitotic arrestTo examine the entry of p-bearing cells into mitosis,mice were given vincristine sulphate ('Oncovin', EliLilly, Basingstoke; 01 mg/ml sodium chloride injec-tion) intraperitoneally in a dose of 1 mg per kg bodyweight at 08.30 to 10.00 hours. After precisely timedintervals of 2 hr and 4 hr, femoral marrow cells frompairs ofmice were sampled and analysed for cp and sp,as above.

IrradiationMice were given a single whole body X-irradiation of150 rads (50 + 4 rads/min, 300 kV, 5 mA; filters,copper 1-5 mm, aluminium, 0-5 mm) and maintainedunder conventional conditions with addition of Neo-mycin to the drinking water. Femoral marrow cellsfrom pairs of mice were pooled at 3, 4, 5, 6, 7 and 10days after irradiation and labelled for sy and cp.

Peanut agglutinin bindingTo examine the binding of PNA to cp+sp- cells andsu+ cells, aliquots of 25 x 106 washed cells fromnormal marrow, regenerating marrow and foetal liverwere exposed to 20 pl PNA-FITC (L'Industrie Biolo-gique Frangaise, Paris, Lot H536; 2 mg/ml in PBS)(1:10) for 30 min on ice, washed twice through FCSand then either exposed to anti-p rhodamine in cellsuspensions (30 min on ice) to label sy, or cytocentri-fuged and exposed to anti-p rhodamine (30 min, roomtemperature) after fixation with cold 5% acetic acid inethanol to label all p chain-bearing cells (cp +sp), asabove. Cells were examined by epifluorescence forfluorescein labelling, indicating PNA binding, with orwithout simultaneous rhodamine labelling for eithersp or cp + sp, respectively.

Cell analysisIn every cell preparation, at least 1000 nucleated cellswere examined in each ofthree ways: by phase contrastmicroscopy and epifluorescence excitation for fluores-cein and rhodamine labelling, respectively. Metaphase

figures were scored and the mean diameter of each cellwas measured with an ocular micrometer scale. Thesize distribution profile and incidence of total p+ cells,cp+sy- cells, sy+ cells and cp +sp - cells in metaphasewere calculated.

RESULTS

Cells containing cytoplasmic p chains in the absence ofdetectable surface membrane u chains were readilyidentified and distinguished from sp+ cells by doubleimmunofluorescence labelling. In the marrow of 7- to9-week-old CBA mice the incidence of cu+sp- pre-Bcells, detected under the prevailing technical condi-tions, was 5-5 + 0.5% and that of sp+ B lymphocytes,8.2 + 0.6% (Table 1).

Table 1. Incidence of cells bearing cytoplasmic and surface juchains in postnatal bone marrow and foetal liver

Incidence of p+ cells (%)

Tissue Age Cytoplasmic p Surface p Total p

Bone marrow 7 weeks 4-3 9 0 13 38 weeks 57 71 1289 weeks 64 84 148

Foetal liver 19 days 10-2 4 5 14 7

Size distribution profile of pre-B cells in normal marrowDirect measurements of individual pre-B cells and Blymphocytes in cytocentrifuged cell preparationsrevealed characteristic population size profiles (Fig.1). Cells bearing sp were all small lymphocytes,ranging from 5 to 10 pm in diameter, with a singlesharp distribution curve, peaking at 7 5 pm. Incontrast, cp +sp- cells showed a broad size distribu-tion, ranging widely in size from 7 to 15 pm indiameter, and tending to give a biphasic curve.Apart from the presence ofp chains, indicating their

B lymphocyte lineage, the largest cu+sp- cellsappeared only as large undifferentiated blasts with anucleus of circular, indented or irregular outline andcopious cytoplasm. Intermediate size cp+sp- cellsshowed the morphology of large lymphoid or transit-ional cells, having a relatively small cytoplasmicvolume and a circular or slightly indented nucleus withfine chromatin pattern. The small cp+sp- cells (< 10pm), however, were typical small lymphocytes with acoarse pachychromatic nucleus and high ratio of

335


151

Cr

*12

,5 10 IS

Cell diameter (i.m)Figure 1. Size distribution of cells bearing (0) cytoplasmicand (0) surface

pchains in the bone marrow of normal mice.

nucleus to cytoplasm. On morphological groundsalone, these cells would be indistinguishable from sp+cells.

Size distribution profile of proliferating pre-B cellsIn normal marrow large cpu+sp - cells were occa-

sionally observed by nuclear morphology to be inmitosis. To examine more fully the size characteristicsof this cycling population, vincristine sulphate was

administered to block cells in metaphase (Wright &Appleton, 1980). Four hours later, arrested meta-phases in cu+sp- cells were readily distinguishableunder phase contrast by a characteristic dense andirregular chromosomal mass. Virtually all the mitoseswere among the large cp+sp- cells, forming a well-defined and broad cell size distribution profile rangingfrom 10 to 15 pm in diameter in cytospots (Fig. 2).Within the large cp+sp- cell population ( > 10 pm) theproportion of total cp+sp- cells of each size categoryarrested in metaphase increased progressivelythroughout the size range, almost all the largest cells(> 13 pm) having entered mitosis within 4 hr aftervincristine administration. At the same time, the sizedistribution profile of the total cp+sp- cell populationshowed a shift to the right, with a relative decrease in

105 1O 15-

Cell diameter (mm)Figure 2. Size distribution of cells bearing (0) cytoplasmicand (0) surface p chains and of(v) cytoplasmic p chain-bear-ing cells in metaphase in mouse bone marrow, 4 hr afteradministration of vincristine sulphate.

small cp/+sp- cells and increase in large cp+sp- cellsrelative to normal mice not given vincristine (compareFigs 1 and 2).The rate of entry of cp+sp- cells into mitosis was

remarkably high. In one pair of mice, 2 hr aftervincristine, 12-9% of all cp+sp- cells were in meta-phase. Two hours later in two further pairs ofmice thevalues had risen to 27.2% and 39.7%, respectively,indicating a rapid turnover ofthe cycling cp+sp- cells.

Thus, the cp+sp- pre-B cells comprise two subsets,dividing and resting, distinguishable on the basis oftheir size distribution profile. The large cp+sp- pre-Bcells ( > 10) are entirely proliferating. The largest cellsof all are late in the cell cycle, within 4 hr of enteringmitoses. While other large cp+sp- cells do not allreach metaphase within 4 hr, their proportions relativeto those arrested in metaphase (Fig. 2), coupled withthe evidence for a short total cell cycle time, areconsistent with all these cells being at earlier stages ofthe cell cycle. In contrast, small cp+sp- cells (< 10pm), coinciding in size distribution and morphologywith sp+ small lymphocytes, rarely if ever entermitosis. The possiblity is not excluded that a few smallcp +sp - cells, somewhat less than 10 um in diameter,may be GI cells which might subsequently enlargeduring cell cycle to exceed 10 pm in diameter by the

336


time they enter mitosis. At least the large majority ofthe numerous small cy+sp- cells encompassed by thesize distribution curve below 10 pm in diameter,however, appear to be non-dividing cells.

Pre-B cell subsets in regenerating bone marrowThe size distribution profile ofcu+su- cells in normalmice (Fig. 1) represents the balance achieved betweenthe cycling and non-dividing pre-B cell subsets and Blymphocytes in the steady state in vivo, cell productionbalancing cell loss. To examine the relationshipsbetween these populations more fully their size distri-bution profiles were examined sequentially in a waveof B lymphocyte genesis during post-irradiationregeneration.

Following their initial depeletion after 150 radswhole body X-irradiation, cp+sps- cells were detectedin small numbers at 3-4 days, increased rapidly by 6days and thereafter continued markedly to exceednormal incidences at 7 and 10 days (Table 2). At 6 dayssy+ cells were infrequent, but thereafter their inci-dence increased progressively to high normal values by10 days.

Table 2. Incidence of p-bearing cells in bonemarrow after 150 rads whole body X-irradiation

Incidence of p + cells (%)Days afterirradiation Cytoplasmic p Surface p Total p

4 10 27 375 3.4 7.4 1086 140 1-4 1547 123 36 159

10 117 7.7 19.4

The first cp +sp- cells to reappear at 3 days (data notshown) and 4 days (Fig. 3) were exclusively large,putatively cycling cells, ranging in diameter from 10 to15 pm. Some small cp+sp- cells appeared at 5 days.Subsequently, as the number of cp+sp- cells in-creased, their size distribution profile shifted progres-

sively to the left, so that by 7-10 days a major portionof the curve corresponded to the subset of smallnon-dividing cp+sp - cells (Fig. 3). In contrast, the sizedistribution profile of sp+ cells remained essentiallyconstant in shape.

Size distribution profile of cp+sp- and su+ cells infoetal liverIn 19 day foetal liver, the cy+sp- and sp+ cells showedthe same size ranges as in adult bone marrow (Fig. 4).The shape of the size distribution profile of sp+ Blymphocytes was essentially identical to that in themarrow. Cu+sp- cells overlapped considerably thesize distribution of sp+ small lymphocytes and tendedto show a bimodal distribution including many largeblasts, generally resembling the marrow (Figs 1, 4). Inincidence, relative to sp+ cells (Table 1), and sizedistribution profile, however, the cp+sp- cells in 19day foetal liver most closely resembled those ofregenerating marrow, 6-7 days after 150 rads X-irra-diation.

PNA binding by B lineage cellsIn normal bone marrow certain cells bound PNA-FITC strongly, sometimes distributed uniformlyaround the cell circumference but often concentratedeither in a diffuse cap or on a localized cell protrusionor uropod.By double immunofluorescence labelling, combin-

ing PNA-FITC and anti-p rhodamine binding in cellsuspensions, the sp+ cells were predominately (90%)PNA negative (Table 3). In contrast, when totalp-bearing cells (cp+ sp) were labelled by exposingfixed cells to anti-p rhodamine after PNA-FITCbinding in cell suspension the p-bearing cells includedmany PNA-binding cells (Table 3). By subtracting theincidence of sp+ cells it may thus be concluded that alarge majority of cp+sp- cells showed PNA bindingunder the experimental conditions.

Regenerating cp+sp- cells also bound PNA (Table3). At 7 days after 150 rads X-irradiation the totalp-bearing cells (cp+ sp) showed a considerable enrich-ment of PNA binding cells, suggesting that everycp+sp- cell bound detectable PNA, the small numberofp+ cells not bindingPNA representing the incidenceof sp+ cells.

In one experiment, p-bearing cells also bound PNAin 19 day foetal liver (Table 3). The labelling tended tobe weaker than in the bone marrow, however. Some,but not necessarily all, cp+sp- cells appeared to binddetectable PNA under the experimental conditions.

In contrast to the cytoplasmic p-bearing cells, only aminority (I 1-15%) of the completely p-negative cellsin normal marrow, regenerating marrow or foetal livershowed PNA binding (Table 3), and this was often ofrelatively low intensity.

337


o 'J o1 o\g1 .5 days 6 days

0

2 0~~~~~

5 10 15 5 10 15

Cell diameter (jim)Figure 3. Size distribution ofbone marrow cells bearing (-) cytoplasmic and (0) surface Muchains at various times after sublethalwhole body irradiation (I150 rads).

DISCUSSION

The present results provide a series of size distributionprofiles for cp+sy- pre-B cells in normal and develop-ing bone marrow; they define cycling and non-cyclingsubsets and permit a correlation of immunofluores-cence data with other marrow lymphocyte studies andpre-B cell assays in formulating a model of B lympho-cyte genesis.Many of the cp+sp pre-B cells in bone marrow are

actually small cells coinciding in size distribution andmorphology with sp+ B small lymphocytes (Owen etal., 1977a; Landreth et al., 1981). Of the larger cp+s/-cells, some are morphologically large lymphoid(transitional) cells, as previously described (Yoshida &

Osmond, 1971; Rosse, 1971; Owen et al., 1977a), whilethe largest are blasts that would not otherwise berecognized as of lymphoid lineage (Pearl et al., 1978).The overall range in size of c+sjy- cells is consistentwith other reports (Owen et al., 1977a; Landreth et al.,1981). The bimodal size profile of cp+sp- cells innormal marrow as well as the size distributions of cellsaccumulating in metaphase arrest and in regeneratingmarrow all define a subpopulation of cycling pre-Bcells. The size distribution profiles of non-cycling andcycling subsets of cp+sp - cells correspond with thoseestablished for small lymphocytes and large lymphoidcells, respectively, based on radioautographic analysesof DNA-synthesizing cells in conventionally stainedcytocentrifuged marrow preparations (Yang et al.,

338


10

CFu i

CTI

0

Cell diameter (1m)

Figure 4. Size distribution of cells bearing(0 cytoplasmic

and (0) surface p chains in 19 day foetal mouse liver.

1978). Small cp-sp- cells correspond with the sy-small lymphocytes previously detected by doubleradioautographic labelling in stained preparations,mainly the most newly formed, post-mitotic smalllymphocytes (Osmond & Nossal, 1974).Combining present and previous studies, the place

ofthe cp+ sp- cells in marrow B cell genesis appears tobe as shown in Fig. 5. A subset of large cp+sp- cells,themselves formed from unknown precursors under-going p heavy chain gene rearrangement and ex-

pression, divide to give rise to a population of smalllymphocytes which at first continue to exhibit cp aloneand subsequently express sp, thus becoming identifi-

able as B lymphocytes with no further division or

change of cell size. Consistent with this model is theobserved shift in size distribution after vincristine; therelative number of small cu+sp- cells declines due totheir continuing post-mitotic maturation into sp+small lymphocytes while not being replenished fromthe large dividing cu+sp- cells, arrested in metaphase.The progressive increase in metaphase index with cellsize among the large cp+sp - cells resembles a compar-

able increase in incidence of DNA-synthesizing cellswith size of morphologically defined large lymphoidcells in stained radioautographs (Miller & Osmond,1973). Thus, the largest cells in these size distributionprofiles are predominantly late in cell cycle. Thepresent findings and proposed cell sequence reconcilepreviously disparate reports of the extent to whichpre-B cells as a whole are cycling and the place of celldivision in the transition of pre-B cells to B lympho-cytes (Osmond & Nossal, 1974; Owen et al., 1977a;Pearl et al., 1978; Lau et al., 1979; Rusthoven &Phillips, 1980; Landreth et al., 1981; Paige et al., 1981).The size distribution profiles of cp+sp- and sp+

cells are compatible with sedimentation velocityanalyses both ofcp+sp- cells and of functional assays

of pre-B cells, respectively. From the data of Owen et

al. (1977a) sedimentation velocity profiles for sp + andcp+sp- murine marrow cells can be derived (Fig. 6).The profile of su+ small B lymphocytes and therelative position and shape of the profile of cp+sp-pre-B cells resemble closely the present curves derivedfrom direct measurements of cytocentrifuged cells. Inturn, the profiles of size distribution and sedimen-tation velocity of cp+sp - cells resemble the sedimen-tation velocity profiles of pre-B cells defined by

Table 3. Binding ofpeanut agglutinin by p-bearing and p-negative cells in normalbone marrow, regenerating bone marrow and foetal liver

Incidence of cells (0)

sit + (cp + sp) Others

Tissue PNA+ PNA - PNA + PNA - PNA + PNA -

Bone marrowNormal 0-6 5-6 15-8 78-0

6-4 10-8 9-6 73-2Regenerating* 6 days 7-3 6-7 8-7 77-3

7days 15-3 1-5 113 71-9Foetal liver 19 days 5 16 8 71

* After 150 rads whole body X-irradiation.

339


Figure 5. Scheme of cytoplasmic and surface u-bearing cells in B lymphocyte genesis in bone marrow.

being cycling pre-B cells which develop into LPS-res-ponsive B lymphocytes during a maturation period invitro of 3-4 days. The most rapidly sedimenting cellfractions have been selected experimentally as sourcesof pre-B cells (Lau et al., 1979; Rusthoven & Phillips,1980). It now appears that these would represent onlya portion of the pre-B cell pool, which also may bepartially synchronized late in cell cycle. The extent towhich functionally defined pre-B cells are identicalwith cp +sy- cells is unknown. Using cell size distribu-tion as a common denominator, however, the presentstudies show a consistency between cytological andfunctional assays, and suggest that general compari-sons between microscopic and sedimentation analyses

o _ of marrow lymphoid cells are justified.The size distribution profiles of cp+spu cells in the

normal steady state ofcell renewal represent a summa-tion of all cells with this p chain status, from which it isnot possible to distinguish individual generations or

~-~* ..,..*f~...- ~ + stages in development. To test some of the cell2 3. 4 5 6 7 interrelationships implicit in the model (Fig. 5) and

analyse the sequence of B lineage cells in relation toSedimentation velocity (mm/hr) time the steady state was perturbed by sublethal

imentation velocity distribution of cells bearing X-irradiation and the marrow was examined duringnic and (0) surface p chains in the bone marrow regeneration. Whole body X-irradiation (150 rads)ice (derived from data of Owen et al., 1977a). produces severe depletion of the marrow lymphocyte

population, followed by a rapid recovery and pro-nounced overshoot before returning to normal cell

assays (Lafleur et al., 1972; Lau et al., numbers (Osmond et al., 1966; Osmond & Evoy,lure, immediately reactive, B lymphocyte 1977). Small lymphocytes during this process show as, giving a peak response in vitro to LPS in sequential appearance of sy-, weak sp+ and strongresent a single sharp profile of small cells. so+ forms (Osmond & Evoy, 1977). The progressiveing peak responses at longer intervals (7-8 shift in shape and area of the profiles of cp+sp- and-ver, show an additional broader profile of sp+ cells with time now reveals an orderly wave offly sedementing, larger cells, deleted by pre-B and B cell genesis, providing additional evidencea (Lau et al., 1979; Rusthoven & Phillips, for the developmental sequence; large cycling cp+spas et al., 1982). These are interpreted as small cp+sp-, small sp+. This accords with observa-

201

0

Uc

o~r

LL.

l1-

Figure 6. Sedi(0) cytoplasrof normal mi

functional1979). Matpopulations4-5 days, pCells reachidays), howemore rapidhydroxyure1980; Freit;

340

30[

Pre-B cells in bone marrow 341

tions based on post-cyclophosphamide (Burrows etal., 1978) and ontogenic (Owen et al., 1977a,b; Cooper& Lawton, 1979) sequences.The number of times each cp+sp- pre-B cell divides

will determine the degree of expansion of B cell clonesfollowing the determination of their ju chain speci-ficity. This may have implications both for the size ofindividual clones and the ultimate diversity of the Bcell repertoire. The number of pre-B cell generations isunknown. The present metaphase arrest data indicatethat the large cu+sp- cells are cycling rapidly. Thepost-irradiation regenerative sequence suggests a lagof several (3-4) days between the first cycling cp+sy-cells and B lymphocytes, time for several (4-8) divi-sions and thus a considerable expansion ofeach clone.Although their size distributions have not been fullyanalysed, foetal and post-cyclophosphamide studiesshow a similar time lag between pre-B and B celldevelopment, as do functional assays in vitro(Melchers, 1977; Owen et al., 1977b; Burrows et al.,1978; Cooper & Lawton, 1979; Lau et al., 1979),consistent with several pre-B cell generations. Incontrast, from radioautographic data it has beensuggested that cp/++su- cells represent only the ter-minal mitosis ofthe B lineage (Landreth et al., 1981). Itcannot be excluded that the number of generationsoccurring in a progressive wave of B cell genesis maydiffer from those occurring in steady state equilibrium.Further studies are needed to test this point directly.Of practical interest is the finding of a considerably

increased incidence of large cp+ssu- cells with only afew sp+ cells at a certain well defined stage (6-7 days)of post-irradiation (150 rads) regeneration. Suchbiologically enriched marrow provides a convenientsource ofcycling pre-B cells for experimental purposes(Osmond, submitted).

Foetal liver generally resembles adult bone marrowin the size distribution of sp+ and cp+sp- cells. Thedetailed size profile of cu+su- cells in 19 day foetalliver most closely resembles that of a particular phaseof regenerating marrow, however, rather than normalmarrow. This serves to emphasise that B lymphocytegenesis in mammalian foetal liver represents a limitednumber of generations producing a transient waverather than a steady state of B cell production(Melchers, 1977). The composition of various Blineage cells may thus change markedly with foetalage, in contrast with the relatively stable populationsmaintained in dynamic equilibrium in post natalmarrow.PNA is a lectin binding specifically to glycoproteins

with terminal galactose residues. Such binding sitesare exposed on certain surface molecules of immaturethymocytes (Reisner et al., 1976), germinal centre cells(Rose et al., 1980) and immature haemopoietic cells(Newman & Boss, 1980; Nicola et al., 1980), but notperipheral lymphocytes on which the sites presumablyare masked by sialic acid. Immunofluorescence doublelabelling now demonstrates that PNA binding nor-mally characterises a small minority of sp+ smalllymphocytes and most of the cp+sp- pre-B cells notonly in normal marrow (Newman & Boss, 1980) butalso in regenerating marrow and foetal liver. Thus,high intensity PNA binding consistently characterizescp+su- cells under a variety ofconditions, although itdoes not apparently discriminate between cycling andnon cycling subsets or other developmental stages ofpre-B cells. Because PNA binding to non lymphoidcells in the marrow is generally weak and of lowincidence, however, this property is potentially usefulas a surface marker to separate viable cp+sp- cellsfrom the bone marrow (Osmond, submitted).

ACKNOWLEDGMENTS

The generous gift of rhodamine conjugated, affinitycolumn purified anti-p chain antibody from Dr M. D.Cooper is gratefully acknowledged. The work wassupported by the Medical Research Councils of theUnited Kingdom and Canada.

REFERENCES

BRAHIM F. & OSMOND D.G. (1970) Migration of bonemarrow lymphocytes demonstrated by selective bonemarrow labeling with thymidine-H3. Anat. Rec. 168, 139.

BURROWS P.D., KEARNEY J.F., LAWTON A.R. & COOPERM.D. (1978) Pre-B cells: bone marrow persistence inanti-p-suppressed mice, conversion to B lymphocytes,and recovery after destruction by cyclophosphamide. J.Immunol. 120, 1526.

COOPER M.D. & LAWTON A.R. (1979) Pre-B cells: Normalmorphologic and biologic characteristics and abnormaldevelopment in certain immunodeficiencies and malig-nancies. In: Cells of Immunoglobulin Synthesis (eds B.Pernis and H. J. Vogel), p. 411. Academic Press, NewYork.

FREITAS A.A., ROCHA B., FORNI L. & CouTINHo A. (1982)Population dynamics of B lymphocytes and their precur-sors: demonstration of high turnover in the central andperipheral lymphoid organs. J. Immunol. 128, 54.

LAFLEUR L., MILLER R.G. & PHILLIPS R.A. (1972) Aquantitative assay for the progenitors of bone marrow-associated lymphocytes. J. exp. Med. 135, 1363.

342 D. G. Osmond & J. J. T. Owen

LANDRETH K.S., ROSSE C. & CLAGETT J. (1981) Myelogenousproduction and maturation of B lymphocytes in themouse. J. Immunol. 127, 2027.

LAU C.Y., MELCHERS F., MILLER R.G. & PHILLIPS R.A.(1979) In vitro differentiation of B lymphocytes frompre-B cells. J. Immunol. 122, 1273.

MELCHERS F. (1977) B lymphocyte development in fetal liver.II. Frequencies of precursor B cells during gestation. Eur.J. Immunol. 7, 482.

MILLER S.C. & OSMOND D.G. (1973) The proliferation oflymphoid cells in guinea pig bone marrow. Cell TissueKinet. 6, 259.

NEWMAN R.A. & Boss M.A. (1980) Expression of bindingsites for peanut agglutinin during murine B lymphocytedifferentiation. Immunology, 40, 193.

NICOLA N.A., BURGESS A.W., STABER F.G., JOHNSON G.R.,METCALF D. & BATTYE F.L. (1980) Differential ex-pression of lectin receptors during hemopoietic differen-tiation: enrichment for granulocyte-progenitor cells. J.Cell. Physiol. 103, 217.

OSMOND D.G. (1980) The contribution ofbone marrow to theeconomy of the lymphoid system. In: Essays of theAnatomy and Physiology of Lymphoid Tissue (eds Z.Trnka and R. N. P. Cahill; 'Monographs in Allergy', Vol.16), p. 157. Karger, Basel.

OSMOND D.G. & EVERETT N.B. (1964) Radioautographicstudies of bone marrow lymphocytes in vivo and indiffusion chamber cultures. Blood, 23, 1.

OSMOND D.G. & Evoy K. (1977) Differentiation of bonemarrow lymphocytes: regeneration of immunogloblin-bearing and other lymphocytes in mouse bone marrowand spleen after sublethal gamma irradiation. Anat. Rec.187, 672.

OSMOND D.G. & NOSSAL G.J.V. (1974) Differentiation oflymphocytes in mouse bone marrow. 2. Kinetics ofmaturation and renewal of antiglobulin-binding cellsstudied by double labeling. Cell. Immunol. 13, 132.

OSMOND D.G., ROYLANCE P.J., LEE W.R., RAMSELL T.G. &YOFFEY J.M. (1966) The effect of unilateral limb shieldingon the haemopoietic response of the guinea pig to gammairradiation (150r). Br. J. Haemat. 12, 365.

OWEN J.J.T., JORDAN R.K., ROBINSON J.H., SINGH U. &WILLCOX H.N.A. (1977a) In vitro studies on the gene-

ration of lymphocyte diversity. Cold Spring HarborSymposia on Quantitative Biology, 41, 129.

OWEN J.J.T., WRIGHT D.E., HABU S., RAGG M.C. & COOPERM.D. (1977b) Studies on the generation ofB lymphocytesin fetal liver and bone marrow. J. Immunol. 118, 2067.

PAIGE C.J., KINCADE P.W., SHINEFELD L.A. & SATO V.L.(1981) Precursors ofmurine B lymphocytes. Physical andfunctional characterization, and distinctions from mye-loid stem cells. J. exp. Med. 153, 154.

PEARL E.R., VOGLER L.B., OKos A.J., CRIST W.M., LAWTONA.H. & COOPER M.D. (1978) B lymphocyte precursors inhuman bone marrow: An analysis of normal individualsand patients with antibody-deficiency states. J. Immunol.130, 1169.

RAFF M.C., MEGSON M., OWEN J.J.T. & COOPER M.D. (1976)Early production of intracellular IgM by B-lymphocyteprecursors in mouse. Nature (Lond.), 259, 224.

REISNER Y., LINKER-ISRAELI M. & SHARON N. (1976) Sepa-ration of mouse thymocytes into two subpopulations bythe use of peanut agglutinin. Cell. Immunol. 25, 129.

ROSE M.L., BIRBECK M.S.C., WALLIS V.J., FORRESTER J.A. &DAvIES A.J.S. (1980) Peanut lectin binding properties ofgerminal centres of mouse lymphoid tissue. Nature(Lond.), 284, 364.

ROSSE C. (1976) Small lymphocyte and transitional cellpopulations in the bone marrow: their role in themediation of immune and hemopoietic progenitor cellfunctions. Int. Rev. Cytol. 45, 155.

RUSTHOVEN J.J. & PHILLIPS R.A. (1980) Hydroxyurea kills Bcell precursors and markedly reduces functional B cellactivity in mouse bone marrow. J. Immunol. 124, 781.

WRIGHT N.A. & APPLETON D.R. (1980) The metaphasearrest technique. A critical review. Cell Tissue Kinet. 13,643.

YANG W.C., MILLER S.C. & OSMOND D.G. (1978) Matu-ration of bone marrow lymphocytes. II. Development ofFc and complement receptors and surface immunoglobu-lin studied by rosetting and radioautography. J. exp.Med. 148, 1251.

YOSHIDA Y. & OSMOND D.G. (1971) Identity and prolifer-ation of small lymphocyte precursors in cultures oflymphocyte-rich fractions of guinea pig bone marrow.Blood, 37, 73.

pre-b cells in bone marrow: size distribution profile, proliferative

Documents