Document wDG16Nr0DLb4KJpJkw421rjLQ

Pre-B Cells in Peripheral Blood of Multiple Myeloma Patients L" By Linda M. Pilarski, Michael J. Mant. a n d Bernard A. Ruether Although multiple myeloma is a disease of plasma cells, abnormalities have been detected in both B and T lymphocytes in peripheral blood. Although multiple myeloma patients are deficient in surface Ig (s1g)-positive B lymphocytes, analysis of lymphocytes present in blood indicates an abnormally large pool of circulating pre-B cells. These pre-B cells express BA-1, do not bear slg, and contain cytoplasmic p chains. High numbers of pre-B cells occur in 88% of individuals with frank myeloma and in 44% of individualswith monoclonalgammopathy of undetermined significance. Pre-B cells bearing BA-1 differ between patients in their expression of HLA-DR and receptors for peanut agglutinin (PNA). Those pre-B cells in myeloma patients a r e either BA-l+ PNA- HLA-DR+ (54% of patients) o r BA-l+ PNA' HLA-DR- (30% c f patients), w mixture of phenotypes (14% of patients). Pre-B c d h PNA- phenotype are almost alwal s HLA-DR', and p)u, 4tppa6rrtee0iv---B6feocdlcdeeelfllalisnpniodtaiportuenelsmaH.tpiTLoohAnre-vaDalnRrsui-ehm.sifbWtbesyirtobhofofintsphi-iterqhf-uaeBacnescaetipmltlashetimevpeanaytaio-ninwdaW.shIcr, tBdheAet-e1Bc'tleydPm.NpTAhh-oecstyoeteoBdbAifs-fe1er'rveanPttiNioanAtis+oninopdraizvtahictweeaayvnelerasabadninohgra"vt`oe%kbrrn cells in the periphery that vary in r umber and cell rurh~, phenotype, and that are unable to express slg. w 1985 by Grune & Stratton. Inc. I THE M A J O R I T Y of patients with multiple myeloma are deficient in their ability to mount a humoral immune response'-' and in secretion of serum immunoglobulin (Ig) other than the monoclonal protein! In addition, they show severely reduced numbers of B cells in peripheral blood as defined by their expression of surface immunoglobulin (sIg), HLA-DR, and an antigen uniquely expressed on sIg' B cells (41H.16).7 It seems likely that the inability to manifest humoral immunity'" and susceptibility to infections* are direct results of the deficiency in sIg+ B lymphocytes. The reasons for the decreased numbers of sIg+ B cells are unknown, but could result from a shift in differentiation toward plasma cells, a block in B cell development at the stem cell stage, or a maturation block a t some point after commitment to the B cell lineage.7*e" In this study we demonstrate that in multiple myeloma patients, differentiation of slg- pre-B cells to sIg+ B cells is blocked, and pre-B cells appear in the peripheral blood. Although the majority of the 47 patients studied have decreased numbers of sIg+ cells among peripheral blood lymphocytes (PBLs), they exhibit normal numbers of BA-I+ cells. BA-I is a marker that defines mainly slg+ B cells in normal donors, although a small subset of slg- BA-I + lymphocytes is detectable." Gathings et all3 have described sIg- cells with low amounts of cytoplasmic p chains as pre-B cells. W e present evidence here to show that the slg-BA-I+ subset of cells in patients' PBLs represents cp pre-B cells. MATERIALS AND METHODS Fifty-five patients with multiple myeloma (nine with light chain, two with nonsecretory, three with IgA, and 41 with IgG M- From the Departments of Immunology and Medicine. University of Alberta, Edmonton. and the Department of Medicine, University of Calgary, Alberta. Canada. Supported .by a grant from the National Cancer Institute of Canada (L.M.P.). Submitted Sept 10. 1984; accepted Feb 19. 1985. Address reprint requests to Dr L.M. Pilarski, Department of Immunology, University of Alberta. Edmonton. Alberta. Canada T6G 2H7. o I985 by Grune & Stratton, Inc. 0006-4971/85/6602~31$03.Oo/0 416 components), 20 patients with monoc1on;tlgammopathy of.-u a)mined significance (MGUS), and 23 no'mal donors (age 40 to were studied. Informed consent was obti ined prior to p a r t i d p h in the study. Of the multiple myeloma p Itients, 13 were u n t r a u . 36 were receiving intermittent chemotherapy, and six had no chemotherapy treatment for at least si ( months before the timed L sampling. Purification of PBLs. Heparinized I~loodsamples were fuged and the buffy coat was collected. This was diluted eightfd in commercially prepared RPMI growth medium (GIBCO, G a Island, NY) combined with 20% fetal cz If serum, and layend ova Isolymph (Gallard-Schlesinger, Carle PI ice, NY)followed by m. trifugation. The white cells on the RPM I/lsolymph interface W- collected, washed three times in the RI'MI growth medium, a d subjected to a viable count and a differential count of May- i 1 ! t Grunwald-Giemsa-stained smears to determine the percentage d lymphocytes in the sample. Recovery of 1 ymphocytes after pu-. tion was routinely 90%to 100%. In additi in, all immunofluo- (IF) assays included aliquots of cells inc ubated with anti-granub cyte or anti-monocyte monoclonal antibcdies to assess the purity PBLs actually screened by IF. Most prep irations had 4%granub. cytes and/or monocytes. Plasmacytoid cc 11s were only rarely seenin ! cytospin preparations. Immunojuorescence. For this study the assay method was 111 indirect IF assay. Every set of assays incliided a normal PBL s a m e as a control to ensure that B cells were d#:tectablewith the reagents used on a particular day. Purified PB1.s ( 5 x 10' to 1@) wcre aliquoted into v-bottomed microtiter v,ells, pelleted, and resw pended in 50 pL of the test murine anti- iuman antibody diluted in normal saline to an appropriate concentr:ition. and incubated for 60 minutes at 4 OC. Plates were centrifuged ind washed twice in saline, and the cell pellet was resuspended in 5 1 pL of a 1/20 dilution of F(ab), fragments of sheep anti-mouse Ig !abeled with fluorescein isothiocyanate (FITC) or tetramethyl rhodamine isothiocyanate (RITC) (Tago, Burlingame,Calif), incubated for 60minutes at 4C. and washed twice. When indicated, cells were resuspended in warm ' saline and incubated for ten minutes at 37 OC to allow cap formition. followed by fixation of cells in 1% formalin. If the capping step was not to be included, cells were kept in 0.02`6sodium azide througbout the procedure and fixed in 1% formalin alter the final wash. ' IF was examined using a Zeiss microscope with f l u o r c s ~ ~ ~ epi-illumination, and selectivefilters. Cell samples were counted fOr total cells in the field and for total numlar of cells with capped 01 I ring fluorescence using a hemacytometer A capped cell was definod as one having a polar aggregation of flu xescence and at least One third of the cell lacking fluorescence. Caps were generally d i s t i d covered approximately one third of the ell surface, and were VCry .- t B M ,VOI66, AO2 i PRE-B CELLS IN BLOOD OF MYELOMA PATIENTS 47 bright. The percentage of positive cells was calculated after screening a minimum of 1,OOO cells. In samples with less than 0.1% of caps or rings, a minimum of 5,000 cells were screened. Assays for detection of sIg involved enumeration of capped cells as has been described.' Assays for BA-1 involved ring fluorescence. For the cp chain determinations, a modification of described procedures was used.'' Cytospin preparations of PBLs were fixed in carbon tetrachloride followed by washing and then treatment with FITC-sheep anti-human IgM (Tago), or RITC rabbit anti-p (Jackson Immunochemicals, Avondale, Pa); comparable results were obtained with both reagents. Negative controls were always done on patient samples using FITC-sheep anti-mouse Ig, and every test included slides of Nalm-6, a c p + pre-B cell line, as a positive control on the reagents used; these controls were uniformly completely negative or completelypositive, respectively. We found that reproducible results were obtained only with the carbon tetrachloride fixation procedure; use of acid-ethanol, when it was succesful,gave comparable intensity and quantity of cp+ cells, but frequently no fluorescence was observed even for the Nalm-6 control slides. The number of BA-I+ cells per milliliter of blood was calculated as (fraction of BA-I+ cells in purified PBLs) x (fraction of -lymphocytes in WBCs) x (number of WBCs per milliliter of blood) number of BA-1+ cells per milliliter of blood. Double IF (dIF). Aliquots of PBLs were tested for expression of two different markers by either indirect IF or a combination of indirect and direct IF. Cells were incubated in sodium azide with BA-I antibody or with saline, washed, and treated with RITCF(ab), fragment of sheep anti-mouse Ig. After washing again, cells were treated with the indicated second antibody or saline, washed, and incubated with an FITC-F(ab), fragment of sheep anti-mouse lg. For the controls in which saline was used instead of a second antibody, no double-labeled cells were detected. To determine the number bearing both BA-I and PNA receptors, cells were treated Jequentially with BA-I, RITC-F(ab), sheep anti-mouse Ig, and f i d l y PNA-FITC. Cells exhibiting both red and green fluorescence were enumerated. Every test included a negative control in which salinewas substituted for the BA-I antibody. These did not show red fiuorescence. dIF to detect cp chains was done using suspension of PBLs. Cells were tested sequentially as follows: BA-I, RITC-F(ab), sheep anti-mouse Ig, fixation in ethanol, and finally FITC-sheep antihuman IgM (Tago)." Cells with red ring fluorescence and intracellular green fluorescence were considered to be BA-I+ cIg+. Every assay included a negative control with saline instead of the BA-I antibody. Antibodies. BA- 1 is a monoclonal antibody specificfor a marker on sIg+ B cells and some granulocytes" (Hybritech, San bego). Anti-rc and a n t i 4 were monoclonal reagents from BRL (Gaithersburg, Md). OKT3, OKT8, and OKTlO were from Ortho Pbrmaceutical Corp (Raritan, NJ). 80H-3 and 80H-5'' are monocronal antibodies specific for granulocytes, and 82H-316 is specific 'or monocytes; these were a gift from Dr Patrice Mannoni. University of Alberta. 7H-3 is a monoclonal antibody specific for mono- Whit determinants on HLA-DR," provided by Dr BM Longen- ecker, University of Alberta. PNA-FITC was from Sigma (St his). RESULTS sIg+ cells, the number of BA-I+ lymphocytes was within tlie normal range (Fig 1 and Table 1). In 78% of the patients, tlie ratio of BA-I+ to sIg+ cells was significantly higher th: n observed in normal individuals where Ba- 1 and sIg+ levc Is + were comparable (Fig 1 and Table 1). The discrepancy between the number of BA- 1+ cells ar d sIg+ cells did not correlate with treatment status (Table I ) , and was not restricted to frank myeloma, as 44% of patients with MGUS also exhibited an elevated BA-l+:sIg+ ratia. Abnormally high ratios of Ba-l+ relative to sIg+ cells we-e seen in 50% of patients with light chain disease, and in tv'o *patients with nonsecretory myeloma, (mean, 4.9% 1% BA-1+ cells), indicating that the high levels of BA-I+ cells are not due simply to the presence of large amounts If M-component in serum. Usually, in patients with MGU3, the number of BA-1 cells in PBLs, and the number of BA-I + cells per milliliter of blood were within the normal ranl:e (Table l), although values from myeloma patients varicd considerably more than those from normal donors or patients with MGUS (Fig 1). The presence or quantity of BA-I+sIg cells showed no detectable correlation with type of man)clonal protein expressed or with its location in serum v uriiie (data not shown). 1100.0 88 -10.0 vw8- ) u* a0X. I -5 0 0 00 I0 n m 10- - 0 0 ;le 0 0 8 0 0 0 0 0 8 08 0 I: 0 '0 1 mma 1111 1 1 aI nl nl '0 I !01 01-10 11-60 3111 8.31 3 Wln P4TlENTS MGUS NORMAL CATILNTS WMAL with IgG, IgA, or light chain disease had reduced numbers of described.' Mean normal ratio, 1.1 gO.36SD. 418 Table 1. EA-1 + Cells in Patients ClassifiedAccording t o Treatment Status Patmnts Untreated Intermittent chemotherapy Off treatment MGUS Normal % BA-1 Cells in PBLs (No. of Samples) *4.4 1.9(12) 8.6 f 1.2 (64) 5.9 i 2.5 (6) 5.5 i 1.3 (9) 9.7 f 1 (23) % slg' Calls in PBLs 0.3 f 0.13 *0.3 0.4 1.7 f 0.6 3.1 f 1 6.34 f 0.5 No. of Samples With Abnamal BA-l+:slg* 8/12 50164 516 419 0123 -BA- 1 -+ a l a x io-*/m of Bbod 11*7 -9.9 k 3.8 8.2 2 2.4 *6.8 4 Values are the means SE. An abnormalratioof BA-1' to slg was defined as one exceedingthe mean of the ratio in noT nal donors by (22.2). The number of BA-1' cells in blood was calculated from the WBC and differential counts on the blood to be pi rified, or on a diperipheral blood counts done on a separate tube of blood taken at the same time. ma Variation in BA-I+ levels. A series of patients was assessed for the number of BA- 1 cells in PBLs at regular + intervals for a nine-month period (Table 2). Considerable variation was observed in absolute numbers of BA-I + and in the ratio of BA-I+ to sIg+ cells. Data from case No. 086 suggests possible cyclical variation. Other phenotypic markers on BA-I + cells. dIF studies with BA-1 and a panel of antibodies directed to other B and pre-B cell markers were performed to identify the BA-I + subpopulation of PBLs (Table 3). Pre-B cells have been shown to be sIgM- and cp."." BA-I + cells from case No. 086 were 95% sIgM- as defined by a lack of ring fluorescence with the FITC-anti-IgM on RlTC/BA-1 -labeled cells. However, 88%of the BA-1' cells were weakly positive for cIgM in double IF determinations. The majority of BA-I+ cells did not bear markers expressed on granulocytes or monocytes and were negative for expression of OKTIO, or receptors for PNA. In a series of double IF tests on PBLs from different individuals, each of three patients had BA-I+ sIgM- cells. Ten of 12 patients had 3% to 20% BA-1+ cells but t0.6% OKTlO+ cells. In a series of cp determinations, 18 of 22 myeloma patients whose PBLs expressed low numbers of sIg+ cells had, on average, 22-fold higher numbers of cp+, and twofold more cp+ cells than BA-I+ cells (Table 4). Patients with MGUS and normal donors had numbers of cp+ cells that correlated with the number of sIg+ cells and, on average, twofold more BA-I + cells than cp+ cells. Thus, the pattern in patients with MGUS and normal donors was considerably different from that of myeloma patients (Table 4). Preliminary data sug- gest that the BA- 1- sIg- cp+ popul ition seen in some of I& myeloma patients represents a very early stage of Pre-B al adevelopment (L.M.P., unpublished Ibservations, 1984). majority of the cIgM+ cells had a large, sometima anve luted, nucleus and a thin rim of CI:M+ cytoplasm. IF with the anti-p or anti-IgM reagents was dim to m d m % although in two patients a very low number of cells with & bright I F and morphology characte ktic of plasma cells observed. In six of eleven patients ar alyzed, tl% of c ~ a+lh were positive for the expression of c rtoplasmic light chainre Heterogeneity in the expression c f receptorsfor PNA a& of HLA-DR on BA-I + cells from p ztients. Using dIF, & BA-I+ cells in a number of patients were screened for simultaneous expression of HLA-DI t and receptors for PNA (Table 5). Although the BA-I + cell i within PBLs of a @veri individual were in most cases homog eneous, heterogeneity in the expression of PNA receptor2 was seen among the patients. For multiple myeloma, WI: detected either BA-1' PNA+ or BA- 1 PNA- cells. Norm a1 donors and those wib + MGUS had mainly BA-1 PNA- cells. In all categories we + detected some patients with a mixti re of BA-1' PNA' a d BA-I+ PNA- cells; most of the mi ltiple myeloma patienu had BA-I + PNA+with a small subs%of BA-I+ PNA-cellr, while normal donors and MGUS pat lents with mixed populations had the reverse pattern. Heter )geneity among patients was also observed in the expressicn of HLA-DR. BA-I+ PNA- cells in myeloma patients. MGUS patients, and normal donors were HLA-DR+ (Ta ,le 5 ) . In all cases where BA-I+ HLA-DR- cells were seen, those correlated very closely with the number of BA-I + P NA+ cells in aliquots of t f Table 2. Temporal Variation in the Number of EA-1+ slg- Cells in PELSof Multiple Myeloma Patients Percentage of BA-1+/SIB+in lndvidualSampler of PBLs onNo. Julv AUO Sept Nov Dm FEb MRdl 086 510.3 013 1.612.4 075 3.0.10.32 109 - 082 810.4 093 087 097 - 101 2.810.2 080 710.9 1510.05 0.0311.6 - - - 5.910.33 - 1610.28 - 2.110.7 1710.08 03411.5 - - 1910.06 - 18.310.08 7.610.1 1 0.610.16 - 2810.02 14/13 1510.07 - - 2.0/0.08 2.910.15 13.510.11 - 5.810.02 - - 20.10.75 - 1210.03 - - 22l0.02 - - - - 7.210.1 - 4.310.9 2 410.07 - 2. 4/o.o9 4.310.7 - 6-10.02 26f0.06 1.910.3 - 110.09 - 2.310.7 0.1610.2 13.40.9 *PBLs taken from patients while still untreated. Patient No. 075 had light chain disease, and patient No. 080 had an I( A M-cornponent;the rest d had lgG M-components. r t I n E - B CELLS IN BLOOD OF MYELOMA PATIENTS 49 Table 3. Phenotypic Markers on EA-1 Cells + M&Ol Dcubk-Ldmbd CellslNo. of BA-1' Cdls 1%) phsnotypeof Ma* EA-l* Sllbrst - Double immunofluoreecence slgM 101195 (5.11 slgM- CP PNA receptor 80H.3 (granulocyte) 80H.5 (granulocyte/myeloid)) 82H.3 monocyte) 43/49 (88) 2/132(1.51 79/299 (26) 16/101 (161 115/38 1 (30) CP+ PNA80H.380H.582H.3- Single immunofluorescence OKT10:BA-1 0:250 ( ~ 0 . 0 4 ) OKTlO- ~ PBLs were from patient No. 086 (1gG-r: on treatment) and were 70% lymphccytesand 30% polymorphsQ monocytes by Giemsa staining. Of to al PBLs, 22% wwe BA-1'. For all markersexcept OKT10, double immunofluorescence,performedat 4 O C in the presence of sodium azide, was used :o define the 8A-1' population. For OKT10, separate aliquots of cells were tested which yielded 22% 8A-1' cells and <0.05% OKTlO' cells, makii 1g double fluorescence studies unnecesm. the same PBL preparations. In the case of normal donors, MGUS patients, and four myeloma patients, BA-I+ cells that appeared to be sIg+ B cells were also PNA-. Temporal variation in the PNA/HLA-DR phenotype of BA-l+ cellsfrom individual patients. In addition to quantitative temporal variations in BA-1+ cells (Table 2), the slg- BA-I population in individual patients exhibited quali- + tative shifts in the PNA/HLA-DR phenotype over time (Table 6). During a seven-month period, patient No. 086 exhibited a shift from a PNA- to a PNA+ BA-I + population (U defined by dIF studies. Both populations were sIgClgM'. In PBLs from patient No. 075,BA-I + cells showed a transition from a PNA+ to a PNA- phenotype. Patient No. 097 maintained a stable BA-1+ phenotype. The level of sIg- BA-l+ cells in PBLs does not correlate Mth T cell subset distribution. Since patients have a *kewed ratio of OKT4- to OKT8-bearing T cellsigvmthat involvts reduced numbers of OKT4+ cells coupled with normal numbers of OKT8' cells, we searched for a relation'hip between the fluctuation in BA-l+ cells and the T cell distribution. The majority of patients studied here had M u a d numbers of OKT4 and normal numbers of OKT8 ells: the total number of T cells (OKT3+) varied consideraMY:l We observed no correlation between the absolute "umber of BA- 1 cells or the ratio of BA- 1+:sIg-to absolute + a u m b of OKT4+or OKT8' cells (data not shown), and no correlation between BA- 1+:sIg- and the proportion of total r cells that are OKT8+ (OKT8:OKT3, Fig 2). The level of slg- BA-1+ cells does not correlate with tl e concentration of polyclonal Ig in serum. Since B cell differentiation is arrested in most myeloma patients, t seemed possible that the level of sIg- BA-I+ cells migl t correlate with the concentration of normal serum Ig. Tl-e amount of serum IgM, as the uninvolved class for IgG, IgP ., and light chain disease, was quantitated and compared to the percentage of BA- 1+ cells.There was no correlation between number or PNA+/- phenotype and reductions in the concer - tration of serum IgM (Table 7). DISCUSSION Lymphocytes from multiple myeloma patients appeared t 3 be frozen in their differentiation from sIg- pre-B cells to sIg ' B cells, and pre-B cells were abnormally located in th: peripheral blood. We detected high numbers of BA- 1+ sIg cp lymphocytes in the PBLs of myeloma patients and in 445 of patients with MGUS. We also detected temporal fluctuation in BA- I + levels of up to 460-fold in some patients. Thes : observations suggest that the B lymphocyte lineage c f patients bears superficial similarity to the developmenta I Table S. Heterogenietyin Expression of PNA and HLA-OR Markers on EA-1 Cells From Dflerent Individuals: Double + Immunofluorescence Studies Table 4. cp+ Cells in Multiple Myeloma. MGUS. No. of IndividualsWith and Normal Donors BA-1' PNA' BA-1+ PNA' Mixed. Pacentagoof PBLsIRanga) Multiple myeloma 14/52 29/52 9/52 No. of MGUS Normal -5120 12/20 15/16 312 0 1/16 EA-1' HLA-OR++ BA-1' HLA-OR-$ Mixed. *a 18 15.6 t 4.3 0.7 c 0.4 7.0 c 1.4 Multiple myeloma 10126 8/26 8/26 -W J S h \ Thr (2-801 4 C l 0.19 c 0.1 0.67 c 0.3 9 3.3 2 0.7 3.1 f 1 6.2 2 1.2 (<1-71 3 6.3 c 1.2 4.4 0.5 11.9 f 0.8 (4- 10) Values reported above accurately r e k t the pattern of MGUS Nwnal 518 118 218 516 - 116 A population of SA-1 cells was consideredto be mixed if the relevant + subsets exceeded 20% of the total BA- 1+ cells. *Patients with a mixture of PNA+'- phenotypes had a mixture c f HLA-DR+/-. tAll patients with 8A- 1 HLA-OR+ phenotype also expressed th,9 + \ *%bv In indiwdual samples. Marker expression was deterIF on three aliquots of each mdtndual PBL sample. BA-l+ PNA- phenotype. $All patients with BA-1' HLA-DR- were also B A - l + PNA'. 420 Table 6. Temporal Stability of Surface Phenotype in Some but not all Patients Patient No. 097 086 075 Date Dec 1983 Feb 1984 March 1984 Oct 1983 Nov 1983 Feb 1984 March 1984 April 1984 Oct 1983 Feb 1984 March 1984 PNA EA- 1+ +Cella1%) - __ 26 74 14 86 34 66 18 8 2 10 90 75 25 81 19 75 25 86 14 69 31 5 95 Predominant Phenotype PNAPNAPNAPNAPNAPNA' PNA' PNA+ PNA' PNA' PNA- All determinations of PNA bindingwere done in double IF with 8A-l+ cells. Patient No. 0 9 7 had 3% to 25% 8A-l+ cells, patient No. 086 had 2% to 28% 8A-l+, and patient No. 095 had 3% to 26% BA-l+ cells over the course of these assays (see Table 2). stage of an early fetus when pre-B cells predominate and sIg+ B cells are low or ~ndetectable.""~ BA- 1 is a monoclonal antibody generated in response to a pre-B ALL which expresses the slg- cp+ phenotype.12 The BA-I antigen is expressed on sIg+ B cells and is lost on terminal differentiation to a plasma cell. PBLs from normal donors included 6% to 10% BA-I+ sIg+ cells and 0% to 1% BA-I+ sIg- cells." Our results confirmed that BA-I+ cells from normal donors were sIg+ B cells. Finally, normal and chronic lymphocytic leukemia progenitor cells able to produce B lymphocyte colonies in vitro bore BA-1." To further identify the BA-I+ cells in patients, we used dIF. Since patients are deficient in sIg+cells: BA-1+ cells do not have a high density of slg. However, it was possible that they represented immature B cells with a low density of sIg, or B cells able to rapidly shed sIg.mv2W3 e performed d I F in sodium azide to prevent shedding of sIg and found very few BA-I' cells, if any, that were sIgM+. They did, however, exhibit moderately positive intracytoplasmic staining with FITC-sheep anti-IgM (Table 3). Also consistent with a 5.07 - 12.0 0. 0. 0 f0 02 0 1 IO loo loo0 RATIO OF 80-It: dgt Fig 2. Numbers of Ba-1' cells do not correlate with the number of OKT8+ cells. Closed symbols represent valuer in individual patients; opep symbols represent values in normal patients. Table 7. Lack of Correlation Betweet Percentageof SA-1 + vand Serum IgM .eve15 Mean % of .(lr Serum IgM (mg/mL) <0.18 EA-1+ Cells (No.of Samples) 6.1 r 0.9 (44) Ra lge (%I 0.1 3-26 *, -phenotype (No.of PNA+ (7) 0.19-0.39 6.6 t 0.8(17) 0.5-18.8 PNA` (13) PNA+ (4) PNA- (11) 0.4-2.6 6.4 t 1.0(54) 0.3 1-28 PNA+ (14) -PNA- (17) '*Serum IgM levels were quantitatedby nephelometry (Beahring ,nostics of Canadian Hoechst Ltd, Montrei I). BA-1 + cells and murn were always quantitated at the same tin le point. The normal ran- serum IgM is 0.4 to 2.6 (mean ? 2 SD). Values for BA-l+ cells we means ? SE. definition of the BA-I + cells as prt -B cells in many individ- ual patients was their expression E f a low density of HLA. DR (non-capping s~bset)."'~ Although OKTlO is expressed on sIg+ B cells in bone marrow,24we were unable to deteci expression of OKTlo on the BA-I + PBLs observed here, inc licating that they are not equivalent to immature bone mar'ow B cells. In contrast, BA- 1 PBLs from some patients e: pressed receptors for the + lectin PNA, which in normal patien ts occur only on B cells of bone marrow origin.I8*"In a series >fstudies using a murine model, Osmond and his have shown that the B cell maturation pathway includes large PNA+ pre-B cells which give rise to small PNA+ Ire-B cells, followed by differentiation to PNA- B cells. Th :subpopulations of pre-B cells in the PBLs of multiple myelc )ma patients appeared to parallel the murine pattern. The PI IIA- pre-B cells observed may represent an intermediate st age between the PNA+ pre-B cell and a PNA- B cell. The multiple myeloma patients exhibited heterogeneity i I that some had BA-I+ PNA- cells and others had BA-1' PNA+ cells. MGUS and normal donors had all or mainly PF A- BA- 1+ cells which in normal and most MGUS patients v ere sIg+ B cells. HLA-DR expression on the B1L-1+ population differed between individual patients (Tablc: 5 ) . BA-I+ PNA- cells were always HLA-DR+, while EA-I+ PNA+ cells were HLA-DR-. Both phenotypic sets were cp+sIg-. The pheno- typic patterns seen in patients and normal donors are sum- marized in Table 8. In two patients sequentially typed over a six- to seven-month period, we observed a shift in the BA-I+ phenotype from PNA- to PNA+ in patient No. 086, and from PNA+to PNA- in patient Nc I 095. It seems likely that such temporal shifts in phenotypc account for the mixed PNA+/- HLA-DR+/- population I we observed in some patients. The functional significan :e of the various BA-1' populations and the acquisition or IC ss of markers is currently being investigated. They may reflec :stages of developmental blockade, and suggest that the arr :st in differentiation is a dynamic process. Results analogous to or consistent with our own have been obtained in other systems. A devl :lopmental arrest in the expression of sIg by pre-B cells has been shown in congenital a g a ~ m a g l ~ b u l i n e m i aw, ~it~h ~p'n r-B cells located in the periphery." sIg- PBLs expressing a B cell antigen were PRE-B CELLS IN BLOOD OF MYELOMA PATIENTS 4: 1 c-gw Myeloma MGUS Normal k, mixed population. Table 8. Phenotype of BA-l+ Cells in Myeloma, MGUS, and Normal Donors BA-1 PNA ++ +++ + - or ? +- +++ +- HLA-DR - + + -or* + + + + slgM - - - + + - +or- + No. of Individuals (Type ot BA-1+ Cells1 14 (Pre-8) 2 5 (Pre-B) 7 (Pre-6) 6 (B cells) 5 !B cells) 12 (Pre-8) 1 (Pre-B)/2 (B cells) 16 (B cells) - - described in one myeloma patient by Turesson et aI3* as reflecting a lower degree of B cell maturity. Mayumi et ai" have described an unusual population of bone marrow pre-B cells in leukemia patients. The apparent cycling in PBLs of the BA-1+ cells seen in patient No. 086, which appears to . correlate with shifts in the PNA-binding phenotype, is similar to that reported for bone marrow pre-B cells in a patient with cyclic neutropenia." Reports that B cells from myeloma patients showed enhanced sensitivity to T cell suppression3' prompted us to search for correlations between the fluctuation pattern of BA-I+ cells and OKT4' or OKT8' cells. No such correlations were detected. Neither could we detect a correlation between number or phenotype of pre-B cells and the degree of reduction of serum IgM. It is not clear whether the Ba-I+ pre-B cells have entered the peripheral circulation as a result of disordered differentiation or immunoregulatory influences, or if it is a consequence of extensive and diffuse marrow infiltration by the Plasma cell neoplasm. The existence of pre-B cells in the periphery of some patients with MGUS, and patients in very early stages of the disease, both of which have very felv plasma cells in the marrow, suggests that tumor cell infiltra - tion of the intramedullary space does not play a role in th: exodus of pre-B cells from the bone marrow. In addition, th: number or phenotype of pre-B cells in circulation does nc t correlate in any obvious way with the stage of disease. Experiments are in progress to explore whether these abnor . mally located cells are pre-malignant by determining thi : monoclonal or polyclonal nature of the pre-B cell populatioi I and to determine the reasons for the existence of pre-B cell ; in the periphery. ACKNOWLEDGMENT We thank Dr Edna Rapp for providing access to her patient!. Excellent technical assistance was provided by Donna Jean Gibne), Margret Krezolak, Parviz Somji, and Gaila Hinz. Invaluable help i I patient accrual came from Maggi Staworski and Pauline Ashle). We are particularly grateful to Dr Ted Zipf for providing th: resources of his lab to facilitate transport of samples. REFERENCES 1. Zinneman HH, Hall WH: Recurrent pneumonia in multiple myeloma and some observations on immunologic response. Ann Intern Med41:1152, 1954 2. Cone L, Uhr JW: Immunological deficiency disorders asso- ciated with chronic lymphocytic leukemia and multiple myeloma. J Clin Invest 43:224i, 1964 3. Fahey JL, Scoggins R, Utz JP, Szwed C F Infection, antibody 'esPonSe and gamma globulin components in multiple myeloma and ~croglobulinemiaA. m J Med 35:698, 1963 4. Heath RB, Hamilton Fairley G, Malpas JS: Production of antibodies against viruses in leukaemia and related diseases. Br J Hatmato1 1 0 3 6 5 , 1 9 6 4 5 . Harris J, Alexanian R,Hersh E, Migliore P: Immune function in multiple myeloma: Impaired responsiveness to keyhole limpet # kmwYanin. Can Med Assoc J 104:389,1971 6. Pruzanski W, Gidon MS, Roy A: Suppression of polyclonal immunoglobulinsin multiple myeloma: Relationship to the staging and other manifestations at diagnosis. Clin Immunol Immunopathol '7:280, 1980 7. Pilanki LM, Mant MJ, Ruether BA, Belch A: Severe defiof a mature B cell subset in peripheral blood from multiple myeloma patients. J Clin Invest 74:1301. 1984 8, Cohnen HJ, Rundles RW: Managing the complications of Plasma Ccll myeloma. Arch Intern Med 135:177, 1975 9. Ullrich S, Zolla-Pazner S: lmmunoregulatory circuits in Clin Haematol 11:87, 1975 10. Lindstrom FD, Hardy WR, Eberle BJ, Williams RC: Multi. ple myeloma and benign monoclonal gammopathy: Differentiatioi I by immunofluorescence of lymphocytes. Ann Intern Med 78337 1973 1 1 . Chen Y, Bhoopalam N, Yakulis V, Heller P Changes ii lymphocyte surface immunoglobulins in myeloma and the effect o ' an RNA-containing plasma factor. Ann Intern Med 83:625, 1975 12. Abramson CS, Kersey JH, LeBein TW: A monoclonal anti body (BA-I) reactive with cells of human B lymphocyte lineage.. lmmunol 126:83, 1981 13. Gathings WE, Lawton AR, Cooper MD: Immunofluorescen studies of the development of pre-B cells, B lymphocytes anc immunoglobulin isotype diversity in humans. Eur J Immunol7:804 1977 14. Zipf T: (submitted for publication) 15. Mannoni P, Janowska A, Turner AR, McGann L, Turc JM Monoclonal antibodies to human myeloid antigens. Hum Immunol 5:309, 1982 16. Mannoni P, Janowska A, Fromont P, Weiblen B, Turner AR. Turc JM: Myeloid differentiation studied by monoclonal antibodies, in Dausset J, Milstein C, Schlossman S (eds): Proceedings of the First International Symposium on Leukocyte Differentiation Antigens. Paris, Karger, 1983, p 410 17. MacLean GD, Mosmann RT, Akabutu JJ, Longenecker BM: Preference of the early murine immune response for polymorphic determinants on human lymphoid-leukemia cells and the potential -.. 422 PIURSKI. MANT. AND RUE^ use of monoclonal antibodies to these determinants in leukemia- 28. Osmond DG, Melchers F. Pai3e CJ: Pre-B cells in moore typing panel. Oncodev Bioi Med 3:223, 1982 bone marrow: In vitro maturation of peanut agglutinin binding 18. Cooper MD: Pre-B cells: Normal and abnormal development. lymphocyte precursors separated fron bone marrow by flu- J Clin Immunol 1:81, 1981 cence-activated cell sorting. J Immuno I 133:86, 1984 19. Mills KH, Cawley JC: Abnormal monoclonal antibody 29. Osmond DG:Pre B cells in bon :marrow: Peanut aggluti,,i,, defined helper/suppressor T-cell subpopulations in multiple myelo- binding and separation of cytoplasmic p chain-bearing cell ppuh. ma: Relationship to treatment and clinical state. Br J Haematol tions in normal, post-irradiation and P Aycythemic mice using flu* 53:271, 1983 rescence-activated cell sorting. Eur J 11nmunol 14495,1984 20. Mellstedt H, Holm G, Pettersson D. Bjorkholm M, Johansson B, Lindemalm C, Peest D, Ahre A: T cells in monoclonal gammopathies. Scand J Haematol29:57, 1982 21. Pilarski LM, Mant MJ, Ruether BA, Carayanniotis G, Otto D. Krowka J F (submitted for publication) 22. Perri RT, Royston I, LeBein TW, Kay NE: Chronic lymphocytic leukemia progenitor cells carry the antigens T65, BA-I, and la. Blood 61:871, 1983 30. Pezrl ER, Vogler LB, Okos A I, Crist WM, Lawton AR Cooper MD: B lymphocyte precursors in human bone marrow analysis of normal individuals and path :nts with antibody deficitncy states. J lmmunol1201169,1978 31. Dosch HM, Percy ME, Gelfand EW: Functional differentia. tion of B lymphocytes in congenital ag immaglobulinemia:Genm. tion of hemolytic plaque-forming cells. I Immunol 119:1959, 1977 23. Melchers F, von Boehmer H, Phillips RA: B-lymphocyte 32. Turesson I, Berntorp E, Zettevall 0 The exprmion of subpopulations in the mouse. Transplant Rev 25:26, 1975 24. Janossy G. Tidman N, Papageorgious ES,Kung PC,Gold- stein G: Distribution of T lymphocyte subsets in the human bone human B-lymphocyte antigen and surfa :e membrane immunoglobulin by lymphoid cells from patients with lymphocytic lymphom, multiple myeloma and benign monoclo la1 gammapathy. Acta M d marrow and thymus: An analysis with monoclonal antibodies. J Scand 206:37, 1979 Immunol 126:1608, 1981 33. Mayumi H, Kubagawa H, Omu a GA, Gathings WE, Kcar. 25. Gobbi M, Caligaris-Cappio F, Janossy G: Normal equivalent ney JF, Cooper MD: Studies on the cl mal origin of human B all cells of B cell malignancies: Analysis with monoclonalantibodies. Br leukemia using monoclonal anti-idiot:'pe antibodies. J Immunol J Haematol54:393, 1983 129:904, I982 26. Osmond DG: Production and differentiation of B lymph* cytes in the bone marrow; in Battisto JR, Knight KL ( 4 s ) : 34. Engelhard D, Landreth KS. K,ipoor N, Kincade PW, De Immunoglobulin Genes and B Cell Differentiation. New York, Bault LE, Theodore A, Good RA: Cyc ing of peripheral blood and I Elsevier/North Holland, 1980. p 135 marrow lymphocytesin cyclic neutroper ia. Proc Natl Acad SciUSA 27. Osmond DG, Owen J J T Pre-B cells in bone marrow: Size 80:5734, 1983 distribution profile, proliferation capacity and peanut agglutinin 35. Perri RT, Oken MM, Kay N E I Cnhanced T cell suppression binding of cytoplasmic p-chain bearing cell populations in normal is directed towards sensitive circulating 3 cells in multiple myeloma. I and regenerating bone marrow. Immunology 51:333, 1984 J Lab CIin Med 99512, 1982 I