Detection and characterization of plasma cells in peripheral blood: correlation of IgE⁺ plasma cell frequency with IgE serum titre

Abstract

In atopic patients and patients with hyper-IgE syndrome (HIE) highly elevated IgE serum levels can be detected. Due to their very low frequency little is known about IgE-producing plasma cells (PC) in peripheral blood. We used CD138 MACS microbeads to enrich plasma cells from peripheral blood of normal donors, atopic patients and one HIE patient. CD138⁺ cells were mainly CD45⁺, CD44⁺⁺, CD19^dim, CD38⁺⁺, CD27⁺⁺, CD86⁺, HLA-DR^+/++, CD71^dim, VLA-4⁺, VLA-5⁻, CD28⁻, CD25⁻, CD69⁻, CLA⁻, CD20⁻, CD21⁻ and CD22⁻. They show weak expression of surface Ig but high levels of intracellular Ig and they secrete Ig in culture. Thus CD138⁺ cells from peripheral blood show characteristics of early plasma cells. IgE⁺ CD138⁺ plasma cells could be detected in 19 of 24 normal donors with an average frequency of 0·06% IgE⁺ cells among CD138⁺ cells. Higher frequencies were detected in atopic patients, atopic patients with markedly elevated serum IgE levels and the hyper-IgE patient with an average of 0·32%, 7·21% and 6·54%, respectively. Additionally, using the recently developed cellular affinity matrix technology, we were able to detect IgE secreting plasma cells and thereby could demonstrate that most of the IgE secreting cells express CD138. The frequency of IgE⁺ CD138⁺ cells among PBMC correlated highly significantly with serum IgE titres (r = 0·8532***), indicating that IgE secreting CD138⁺ cells in peripheral blood are directly related to the plasma cell pool contributing to the IgE titre.

Introduction

Humoral immune responses result in the activation of antigen-specific B cells in secondary lymphoid organs and their differentiation into plasma cells (PC), which secrete immunoglobulin (Ig). Following activation B cells enter the bloodstream migrating to tissues such as bone marrow, skin, spleen and mucosa, which are believed to be the major sites of Ig production by PC. Recently it was shown in a mouse model system that long-living PC in the bone marrow can maintain IgG serum titres for more than 1 year [1–5]. However, little is known about circulating PC in peripheral blood because the direct analysis of normal PC in peripheral blood, e.g. by flow cytometry, is hampered by their low frequency, therefore many studies have focused on diseases with increased PC numbers in peripheral blood. First, of all malignant PC in multiple myeloma [6–8], and more recently PC in reactive plasmacytoma, a disorder with high frequencies of non-malignant plasmablasts and plasma cells, were analysed [9,10]. It is, however, not clear whether the phenotype of PC in these diseases reflects the phenotype of PC in healthy donors. In a few studies antibody secreting cells (ASC) in peripheral blood have been analysed using ELISPOT assays [11–14]. A drawback of using ELISPOT assays for the detection of ASC is that it neither allows direct phenotypic characterization nor the isolation of the cells.

Therefore, we used CD138 monoclonal antibody conjugated to superparamagnetic microbeads to magnetically enrich PC directly from peripheral blood for phenotypic and functional characterization. CD138 (syndecan-1) was reported to be expressed on PC in bone marrow and peripheral blood but not on pregerminal centre B-cells [15,16]. CD138 is a heparan sulphate proteoglycan, which mediates cellular adhesion to collagen type 1 [17–20] and might play a role in adhesion to bone marrow stromal cells [21,22]. Enriched CD138⁺ cells were stained for various surface as well as intracellular markers using an improved method for phenotypic characterization of enriched cells while they are immobilized on the magnetic enrichment column. After elution cells were directly analysed by flow cytometry.

In allergic diseases induction of IgE synthesis is a critical step and in some atopic patients very high serum IgE concentrations exceeding 10–100 times the normal level can be found [23–25]. Similarly, in hyper-IgE syndrome extremely elevated IgE levels can be observed [26–28]. Several investigators were able to detect spontaneously IgE-secreting cells in the blood of patients with IgE-related diseases [23–25,29–32]. Using ELISPOT assays or ELISAs IgE-secreting B cells were found in patients with atopic dermatitis and HIE, but only very rarely in normal donors. Furthermore, the elevated IgE serum titres could be correlated to the number of detected IgE-secreting B cells.

We enriched PC magnetically from peripheral blood of normal donors, atopic patients and patients with HIE using CD138 microbeads. CD138⁺ cells were stained for intracellular IgE and the frequency of IgE⁺ CD138⁺ cells was correlated with the IgE serum titres.

Additionally we used the recently developed cellular affinity matrix technology [33] (Arce et al., manuscript in preparation) to detect and analyse IgE-secreting PC directly from peripheral blood. There a catching reagent specific for human Ig light chain was attached to the surface of all cells. Cells were allowed to secrete Ig, which was caught by the affinity matrix. Then the caught Ig was labelled with an IgE-specific detection antibody for flow cytometric analysis. The results were compared with those obtained with the CD138 enrichment procedure.

Materials and methods

Patients and cell preparation

We studied 22 patients with atopic disease, one hyper-IgE patient and 24 normal donors. Donors were classified as atopic according to an IgE serum titre>150 kU/l. The hyper-IgE patient had recurrent skin and pulmonary abscesses and extremely elevated levels of IgE (19440 kU/l) and a typical facial phenotype as recently described [28]. Forty ml of venous blood was obtained from all patients. The blood samples were drawn using heparin as an anticoagulant, and processed within 6 h after collection. Buffy coats (about 50–70 ml) from normal healthy donors were obtained from the Institute for Transfusion medicine (Community Hospital Merheim, Cologne, Germany) and processed within 6 h. All patients were informed about the objectives of the study, and gave their consent. The study was approved by the ethical committee of the University Hospital, Cologne.

PBMC were prepared using the LeucoSep system (Greiner, Frickenhausen, Germany). Fifteen ml Ficoll-Paque (Pharmacia, Uppsala, Sweden) were pipetted onto the filtering disk of a 50-ml tube. After 30 s of centrifugation at 300 g the Ficoll-Paque was beneath the disk and 20 ml of blood was poured on top of the disc. After centrifugation for 10 min at 1000 g, PBMC and serum were collected separately above the disk. PBMC were washed twice with PBS. Serum was stored at −20°C until analysis.

Monoclonal antibodies

The following MoAbs were used for flow cytometry: CD3 (OKT3) from Ortho Biotech Products (South Raritan, NJ, USA); CD14 (MΦP9), CD20 (Leu-16), CD27 (L128), CD69 (L78), CD71 (L01·1) and anti-IgD (TA4·1) from BD Biosciences (San Jose, CA, USA); CD14 (TUK4), CD19 (SJ25-C1), CD138 (B-B4), CD45 (5B1), anti-HLA-DR (910/D7), anti-NP and anti-lambda (HP6054) from Miltenyi Biotec GmbH (Bergisch Gladbach, Germany); CD123 (AC145) from Amcell (Sunnyvale, CA, USA); CD21 (B-ly4), CD28 (CD28·2), CD40 (5C3), CD44 (G44-26), anti-VLA-5/CD49d (9F10), anti-β7 intergrin (FIB504), anti-IgA (G18-1) and anti-IgE (G7-26) from BD Pharmingen (San Diego, CA, USA); anti-VLA-5/CD49e (SAM-1), anti-IgG (JDC-10) and anti-IgM (SA-DA4) from SBA (Birmingham, AL, USA); CD22 (6B11) and CD86 (HA5·2B7) from Coulter-Immunotech (Marseilles, France); CD80 (DAL-1) from CLB (Amsterdam, the Netherlands); anti-IgG (TM15) from Serotec (Oxford, UK); anti-CLA (HECA-452) kindly provided by E. Butcher (Department of Pathology, Stanford University, Stanford, CA, USA); CD25 (4E3) kindly provided by W. Knapp (Institute of Immunology, University of Vienna, Vienna, Austria).

Immunomagnetic isolation and staining of cells

PC from peripheral blood were isolated using CD138 microbeads (Miltenyi Biotec) according to the manufacter's instructions. Briefly, PBMC were incubated at a concentration of 1 × 10⁸ cells/ml with CD138 microbeads at a titre of 1:5. After 30 min of incubation at 4°C cells were washed once with PBS buffer containing 5 mm EDTA and 0·5% BSA (PBS/EDTA/BSA). After resuspending cells in 2 ml PBS/EDTA/BSA cells were separated using two sequential MS-columns (Miltenyi Biotec). To characterize the enriched cells immunophenotypically we used either MoAb coupled to FITC or PE or an indirect approach using MoAb conjugated with biotin or digoxigenin. In the latter cases we used streptavidin PE or anti-digoxigenin PE to stain the cells. Surface and intracellular staining of the enriched CD138⁺ cells was performed on the second MS-column prior to elution. In case of surface staining cells eluted from the first column were applied directly to the second column and washed once with 500 µl buffer. Cells were then incubated on the column for 10 min at RT with 100 µl staining solution and then washed again with 500 µl buffer. For intracellular staining cells eluted from the first column were fixed for 20 min with Inside Fix (Miltenyi Biotec) at RT and then put directly onto the second column. After washing the cells with 500 µl PBS/EDTA/BSA buffer they were stained first on the surface using 100 µl staining solution for 10 min at RT, washed again with PBS/EDTA/BSA buffer and then with 500 µl Inside Perm (Miltenyi Biotec) to permeabilize the cells. Cells were stained intracellularly using 100 µl staining antibodies diluted in Inside Perm for 10 min at RT, then washed once with Inside Perm and once with PBS/EDTA/BSA buffer before elution of the cells with 500 µl of PBS/EDTA/BSA. Eluted cells were then directly applied to flow cytometric analysis.

Flow cytometry

Flowcytometric analysis was performed on a FACScan or FACScalibur (BD Biosciences, San Jose, CA, USA). Forward scatter (FSC) and side scatter (SSC) signals were recorded in linear mode, fluorescence signals in logarithmic mode. Dead cells and debris were gated out using scatter properties of the cells and additionally, unless intracellular staining of the cells was performed, using propidiumiodide (PI) at a concentration of 1 µg/ml. Data were analysed using CellQuest software (BD Biosciences).

Determination of frequencies of IgE⁺ CD138⁺ cells among PBMC

Two approaches were used for the determination of the frequency of IgE⁺ CD138⁺ cells among PBMC:

1. The total number of IgE⁺ CD138⁺ cells in the positive fraction after magnetic enrichment of CD138⁺ cells was divided by the total number of PBMC before separation.

2. The frequency of IgE⁺ cells among CD138⁺ cells was determined by flow cytometry in the positive fraction after magnetic enrichment. The frequency of CD138⁺ cells among PBMC was determined by flow cytometry before separation. Therefore PBMC were stained with CD38 FITC, CD138 PE, CD14 PerCP and CD19 APC and 500 000 events were aquired for flow cytometric analysis.

Cultivation of the isolated CD138⁺ cells

To measure Ig secretion of the isolated CD138⁺ cells, enriched cells, unseparated PBMC and CD138 depleted PBMC from one atopic and four normal donors were cultured for 24 h in RPMI-1640 (Life Technologies, Paisley, UK) supplemented with 10% FCS (Life Technologies), 100 U/ml penicillin/streptomycin (Life Technologies), 1 µm l-alanyl-glutamine (Life Technologies), 0·05 mmβ-mercaptoethanol (Life Technologies) and 1 mm pyruvat (Life Technologies) (RPMI-medium). Isolated CD138⁺ cells were cultured at concentrations between 1 × 10⁴ and 1 × 10⁵ cells/ml. The supernatants were harvested and stored at −70°C.

ELISA

The concentration of serum IgE in patients and controls as well as the concentration of secreted IgE in the culture supernatants was measured by a commercially available ELISA (Pharmacia, Freiburg, Germany) according to the manufacturer's instructions. The lower detection limit of the ELISA is 2 kU/l.

Cellular affinity matrix technology

PBMC were depleted of T cells, monocytes and basophils in order to deplete background giving cells. Briefly, cells were incubated with CD123 biotin and washed once with PBS/EDTA/BSA. Cells were then incubated with CD3, CD14 and streptavidin-coupled microbeads, washed and separated on a LD depletion column (Miltenyi Biotec). The depleted cells were then biotinylated on their surface using 2 mg of sulpho-NHS-LC-biotin in 2 ml of PBS for 10 min at 37°C. After adding 2 ml of RPMI with 10%, FCS cells were incubated for additional 10 min at 37°C and then washed three times with PBS/EDTA/BSA to remove free biotin. Subsequently cells were incubated 10 min on ice with 200 µg/ml unlabelled anti-IgE detection antibody to block surface-bound IgE and after washing with cold buffer 100 µg/ml antihuman lambda antibody was added. After 5 min on ice cells were incubated with 70 µg/ml avidin coupled with antihuman lambda antibody for a further 10 min on ice. Cells were then put into RPMI with 10% FCS for 30 min of secretion at 37°C with a cell concentration of 1 × 10⁶/ml. After culture cells were put on ice for 10 min to stop further secretion and they were incubated with 1 µg/ml NipCap coupled anti-IgE antibody for 10 min and washed with cold buffer. Finally secreted IgE was detected using 2 µg/ml anti-NP APC and further characterization of the cells was performed using fluorochrome-labelled anti-CD38 and anti-CD138 antibody.

Statistical analysis

We used the Spearman's rank correlation test to determine the correlation between serum IgE titres and the frequency of IgE⁺ CD138⁺ cells among PBMC. The Mann–Whitney U-test was used to calculate whether or not the observed differences in the frequencies of IgE⁺ CD138⁺ cells between normal donors and atopic patients were significant. P < 0·05 was regarded as significant. All statistical analyses were performed using GraphPad Prism version 3·0a for Macintosh, GraphPad Software, San Diego California USA, http://www.graphpad.com.

Results

Enrichment and characterization of CD138⁺ cells from peripheral blood

The frequency of CD138⁺ cells among PBMC in normal donors is estimated at 1 × 10⁻⁴ making it difficult to analyse them by direct immunofluorescence. Therefore CD138-expressing cells were isolated from PBMC by magnetic labelling with CD138 microbeads and enrichment of labelled cells using magnetic cell sorting (MACS) for phenotypic and functional analysis. The average purity of CD138⁺ cells from 49 experiments was 81·8 ± 10·4%. The frequencies of CD138⁺ among PBMC, estimated from the total numbers of isolated CD138⁺ cells, ranged from 7·8 × 10⁻⁴ to 1·5 × 10⁻⁵. According to scatter properties the isolated CD138⁺ cells are larger than normal resting B lymphocytes (Fig. 1). The immunophenotype of the CD138⁺ cells is summarized in Table 1. The B cell marker CD19 is expressed on CD138⁺ cells, although lower than on mature B cells, whereas CD20, CD21 and CD22 are down-regulated on most cells (>90%). They highly express CD38, an antigen found on activated B cells and plasma cells but not on resting, naive or memory B cells. All cells homogeneously express CD44 and CD45. The CD138⁺ cells express CD40, CD86, but no CD80 and to a varying degree HLA-DR. While all cells brightly express CD27 and VLA-4, they do not express CD28 and VLA-5 (CD49e). They also do not show CD25 and CD69 expression, but weakly express CD71. Remarkably they do not express the cutaneous lymphocyte antigen (CLA), reported previously to be expressed on about 50% of antibody-secreting cells [34]. Approximately 35% of the CD138⁺ cells express integrin β7, a molecule associated with mucosa homing. According to DNA staining, only few cells are in S or G2/M phases of the cell cycle, indicating that these cells do not proliferate strongly (data not shown).

Fig. 1.

Magnetic isolation of CD138⁺ cells from peripheral blood. The flow cytometry dot plots are a representative example of 49 samples. Left panels (a,c) show CD138 versus CD19 staining. Scatter properties of the cells are shown in panels (b) and (d). Cells in panels (a) and (c) are gated to exclude dead cells and debris (gates not shown). Before CD138 enrichment (a,b) only few CD138-expressing cells can be detected. 100 000 cells were analysed. CD138-enriched cells express CD19 (c) and show distinct scatter properties (d).

Table 1.

Expression of surface molecules on CD138⁺ plasma cells isolated from peripheral blood of normal donors. At least three different donors for each marker were analysed

Antigen	Clone	Expression
CD19	SJ25-C1	+
CD20	Leu-16	− (<5% +)
CD21	B-ly4	−
CD22	6B11	−
CD25	4E3	−
CD27	L128	+ +
CD28	CD28·2	−
CD38	T16	+ +
CD40	5C3	+
CD44	G44-26	+ +
CD45	5B1	+ +
CD56	2G11·3	−
CD69	L78	−
CD71	L01·1	dim
CD80	DAL-1	−
CD86	HA5·2B7	+ +
CLA	HECA-452	−
HLA-DR	910/D7	+/++
surface IgA	G18-1	54·4% dim
surface IgD	TA4·1	−
surface IgG	TM15	27·4% dim
surface IgG	JDC-10	27·4% dim
surface IgM	G18-1	9·7% dim
VLA-4	9F10	+
VLA-5	SAM-1	−
β-7	FIB504	35% +

Surface immunoglobulin (sIg) expression on CD138⁺ cells is weak, but they are brightly stained for intracellular Ig (iIg) (Fig. 2). Normal donors had median frequencies of 27·4 ± 7·4%, 54·4 ± 8·9% and 9·7 ± 5·3% iIgG⁺, iIgA⁺ and iIgM⁺ cells, respectively (n = 14).

Fig. 2.

Expression of immunoglobulin classes by CD138⁺ plasma cells enriched from peripheral blood mononuclear cells of a normal donor using MACS. Upon enrichment cells were fixed and stained for CD138 and surface immunoglobulins, then permeabilized and further stained for intracellular Ig. Flow cytometric analysis was restricted to CD138⁺ lymphocytes by gating according to scatter properties and CD138 staining. Panels (a), (b) and (c) show staining of surface Ig (x-axis) versus intracellular Ig (y-axis) for IgA, IgG and IgM of gated CD138⁺ cells, respectively.

The frequencies of iIgG⁺, iIgA⁺ and iIgM⁺ cells found in the hyper-IgE patient were 22·7%, 43·9% and 26·5%, respectively.

To analyse whether the CD138⁺ cells secrete Ig, we cultured the CD138-enriched fraction, the CD138-depleted PBMC and unseparated PBMC for 24 h in vitro and then analysed the supernatant for secreted IgM, IgG and IgA by ELISA. No secreted Ig could be detected in supernatants of unseparated and CD138-depleted PBMC (detection limit 4 ng/ml). On the other hand, CD138-enriched PC produced up to 500 ng/ml Ig (data not shown). Thus the CD138⁺ cells in normal blood show characteristics of early plasma cells.

Detection of IgE⁺ CD138⁺ cells in peripheral blood of atopic patients and normal donors

In atopic patients and hyper-IgE patients serum levels of IgE were highly elevated. Therefore we investigated whether IgE-positive cells could be detected among CD138⁺ cells in peripheral blood of these patients. Surface IgE was undetectable on CD138⁺ cells of normal donors as well as atopic patients and the hyper-IgE patient.

However in 19 of 24 normal donors we were able to detect at least five iIgE positive cells among the CD138⁺ cells (Fig. 3). The specificity of staining was confirmed by preincubating the FITC conjugated monoclonal anti-IgE antibody for 30 min with a 1000-fold excess of IgE immunoglobulin. In all experiments an almost complete block of IgE staining could be observed (Fig. 3). In normal donors (n = 24; IgE < 150 kU/l) IgE⁺ cells were detected at a frequency of 0·06% (range 0·01–0·26%) among CD138⁺ cells. In marked contrast to the normal donors, dramatically increased frequencies of IgE⁺ CD138⁺ cells were found in the hyper-IgE patient (serum IgE = 19440 kU/l), with a mean frequency of 6·54% (range 5·66–8·6%) IgE⁺ cells among CD138⁺ cells in three consecutive experiments. In the group of atopic patients (n = 13; serum IgE = 150–1000 kU/l) the mean frequency was 0·32% (range 0·04–1·27%) whereas atopic patients with markedly elevated serum IgE levels (n = 9; IgE > 1000 kU/l) had a mean frequency of 7·21% (range 0·76–30·88%) among CD138⁺ cells.

Fig. 3.

Increased frequency of IgE⁺ cells among CD138⁺ cells in a hyper-IgE patient. CD138⁺ cells were enriched from PBMC of a hyper-IgE patient (IgE serum titre 19940 kU/l) (a,b) and a normal donor (IgE serum titre 36 kU/l) (c,d) using MACS. Enriched cells were fixed, stained for CD138, permeabilized and then stained for intracellular IgE. All dot plots show staining of intracellular IgE (x-axis) versus CD138 (y-axis) of gated lymphocytes. The right dot plots (b,d) show the (blocking) control stainings, i.e. preincubation of anti-IgE antibody with an excess of IgE. Only few IgE-expressing cells can be detected among CD138⁺ plasma cells in the normal donor (c) whereas a high frequency of IgE⁺CD138⁺ cells can be observed in the hyper-IgE patient (a).

We found a highly significant difference between the frequencies of IgE⁺ CD138⁺ cells among PBMC in normal donors compared to patients with markedly elevated serum IgE (>1000 kU/l) (P < 0·0001). The frequencies of IgE⁺ CD138⁺ cells among PBMC correlated highly significantly with serum IgE titres (r = 0·8532; p < 0·0001; n = 24) (Fig. 4). Additionally the frequencies of IgE⁺ cells among CD138⁺ cells correlated highly significantly with the IgE serum titres (r = 0·7901 with P < 0·0001).

Fig. 4.

IgE serum titres correlate with the frequencies of IgE⁺CD138⁺ plasma cells among PBMC. Frequencies of IgE⁺CD138⁺ plasma cells were calculated as described in Material and methods (approach 2). Normal donors are represented by open rectangles, atopic patients by solid rhombi and atopic patients with markedly elevated IgE titres by solid rectangles. Spearman's rank correlation test was used to determine the correlation (r = 0·8532; P < 0·0001).

Fig. 5.

CD138-enriched plasma cells from atopic patients spontaneously secrete IgE upon culture. CD138-enriched plasma cells, CD138-depleted PBMC and unseparated PBMC from four atopic patients (solid rectangles; mean of 9·23% IgE⁺ cells among CD138⁺ plasma cells) and four normal donors (open rectangles; mean of 0·15% IgE⁺ cells among CD138⁺ plasma cells) were put into culture for 24 h without additional stimuli. IgE level in the supernatant was analysed by ELISA. All values were calculated on the basis of 1 × 10⁵ cells/ml.

To test whether IgE⁺CD138⁺ cells secrete IgE, the enriched CD138⁺ cells, unseparated PBMC and CD138-depleted cells were cultured for 24 h from four atopic patients (mean serum IgE titre of 11250 kU/l and 9·23% IgE⁺ cells among CD138⁺) and four normal donors (mean serum IgE titre of 63·5 kU/l and a mean of 0·15% IgE⁺ cells) (Fig. 4). IgE production could be detected clearly by ELISA in the CD138-enriched fractions of the atopic patients (mean of 105 kU/l IgE). In contrast, only very low levels of IgE could be detected in the CD138-enriched fractions of the normal donors (mean of 8·8 kU/l IgE), whereas IgE was undetectable in all other samples (detection limit 2 kU/l). For one atopic patient known to be highly allergic against house dust mite, further analysis of the supernatants with allergen-specific IgE ELISAs revealed that the enriched CD138⁺ cells produced IgE specific for Dermatophagoides pteronyssinus and D. farinae, two major antigens of house dust mite (data not shown).

Most IgE secreting cells in peripheral blood of atopic patients express CD138

There are several indications that not all ASC express CD138 [9,34]. Therefore we used cellular affinity matrix technology [33] to detect all IgE-secreting cells in atopic patients and analysed them for expression of CD138 (n = 3). Figure 6 shows the typical picture for the analysis of IgE-secreting PC. While in the control sample without catching matrix (Fig. 6a), only a few background-giving cells are detected in region R1, a distinct population of IgE-secreting CD38⁺ PC is detected in R1 when using the anti-lambda catching matrix (Fig. 6b). Approximately 70% (73·6%, 71·4%, 67·8%) of the IgE-secreting CD38⁺ PC express CD138 (Fig. 6c) and the cells have the same scatter properties as CD138-enriched PC (Fig. 6d). The lower background staining in panel (b) compared to panel (a) results probably from partial blocking of surface Ig-staining through the anti-lambda catch matrix. Side-by-side comparison of the cellular affinity matrix technology and the intracellular staining on the same donors with respect to the obtained frequencies of IgE⁺ cells among CD138⁺ PC revealed a very good correlation, with mean frequencies of 4·9% and 4·8%, respectively (n = 3). Thus this new technology can be used for the direct analysis of individual, viable Ig-secreting cells.

Fig. 6.

Detection of IgE secreting plasma cells in an atopic patient (IgE serum titre 20000 kU/l) using the cellular affinity matrix technology. PBMC were first depleted for T cells, monocytes and basophils. A catching reagent with specificity for human lambda Ig was then attached to the surface of the cells. Cells were put into culture for 30 min to let them secrete Ig. The caught IgE was then detected using an anti-human IgE MoAb (see Material and methods). Panel(a) shows the negative control without catching reagent. Cells being stained here for IgE are cells exhibiting IgE on their surface. Stainings of CD38 versus secreted IgE are shown for both samples (a,b). A distinct population of CD38⁺⁺ IgE-secreting plasma cells could be detected in the sample with catching reagent (R1 in (b)). Expression of CD138 and scatter properties are shown for these CD38⁺⁺ IgE-secreting plasma cells in (c) and (d),respectively (gated by R1 in (b)). A large fraction of these cells express CD138 (c) and their scatter properties are like those of CD138 enriched plasma cells (d).

Discussion

Apart from major sites such as bone marrow, PC or ASC have been detected in peripheral blood, e.g. by ELISPOT assays [23,24,29–32]. The cells are thought to be found transiently in the circulation and are considered to be on their way to their final destination, e.g. bone marrow, gut or skin. The homing potential of PC seems to depend on the site of activation, i.e. route of immunization [11,12,35]. However, the detection and analysis of PC in peripheral blood of normal donors has long been hampered by their low frequency of 1 × 10⁻⁴ among PBMC in normal donors. As shown here, the enrichment of PC from peripheral blood circumvents this problem and allows a direct and detailed analysis of the enriched plasma cells. Very few CD138⁺ cells could be stained for various surfaces as well as intracellular markers using an improved method for phenotypic characterization of enriched cells while they are immobilized on the magnetic enrichment column.

CD38, and more recently CD27, have been shown to be highly up-regulated on PC [36–39] (our unpublished data). However, both antigens are expressed to a varying degree on other cells, such as activated T cells and basophils for CD38 [40–42] or T cells and memory B cells for CD27 [43–46], hampering sensitive detection and simple enrichment of PC. CD138, however, is expressed only on plasma cells but not on circulating B cells or other cells in peripheral blood, thus allowing a simple enrichment and detection of CD138 expressing PC from peripheral blood. CD138-enriched cells from several normal donors had a similar phenotype, mainly being CD38⁺⁺, CD19^lo, CD20⁻, CD21⁻, CD22⁻, CD27⁺⁺, CD28⁻, CD40⁺, CD44⁺, CD45⁺ and CD49e⁻. This is consistent with the reported phenotype of PC in reactive plasmacytoma [9], but they clearly differ from malignant PC in peripheral blood (or bone marrow), which do not express CD19 [47] (and our own unpublished results) but often express CD28 [16]. However, they also differ from normal mature PC in bone marrow tonsils [48] which express VLA-5 [38]. Also CD45 has been reported to be absent from plasma cells in bone marrow [37] or myeloma cells [47]. Additionally they still have weak levels of immunoglobulin on their surface. Furthermore, we showed that CD138⁺ PC do not express CLA, a receptor believed to be responsible for lymphocyte homing to skin [49,50]. This is in contrast to the results of Kantele et al., indicating that about half of the ASC from peripheral blood express CLA [34]. Notably also in their experiments, only about half of the ASC seemed to express CD138. This correlates to data from Jego et al. [9] and to our results obtained with the Ig secretion assay, which show that between 50% and 80% of Ig-secreting, CD38⁺⁺ cells express CD138. Taken together, these results imply that there are CLA⁺ CD138⁻ ASCs in peripheral blood, which presumably migrate to skin, beside CLA⁻ CD138⁺ ASCs, which presumably home to other tissues than skin, e.g. bone marrow. While the molecule CD138 has been implicated to be involved in adhesion to bone marrow stromal cells [21,22], the fact that a very high frequency of CD138⁺ cells express IgA and β7 integrin (35%; IgA⁺ as well as IgA⁻ cells) (data not shown) indicate that a large part of the CD138⁺ cells might also home to mucosal tissue [51]. Analysis of intracellular and surface Ig expression shows that approximately 60% of the isolated PC express IgA, 30% express IgG and 10% express IgM isotype. Interestingly, the frequency of IgA⁺ cells among CD138⁺ PC (60%) is much higher than the frequency of IgA⁺ cells among memory B cells (approximately 6%) [52], further supporting the idea that PC and memory B cells are independently regulated [53]. When put into culture for 24 h the isolated CD138⁺ PC spontaneously secrete high amounts of Ig.

In conclusion, these results show that CD138 can serve as a marker for the rapid detection, analysis and isolation of functional PC from peripheral blood. The phenotype of these CD138⁺ PC is in accordance with the assumption that these are early PC, presumably on their way to their final destination, e.g. bone marrow, spleen or mucosa, and are therefore only transiently in the circulation. This is supported further by recent findings in our laboratory upon monitoring of PLA₂-specific CD138⁺ PC during bee venom immunotherapy of allergic patients. A few days after antigen injection a 10-fold increase in frequencies of CD138⁺ cells in peripheral blood was detected. While no PLA₂-specific PC were detectable among CD138⁺ cells before immunotherapy, up to 20% PLA₂-specific PC were detected among CD138⁺ cells on day 10, which corresponds to a> 2000-fold increase of the total number of PLA₂-specific CD138⁺ PC in peripheral blood. These PC completely vanished 14 days after the last antigen challenge (Horst et al. unpublished results).

From peripheral blood of normal donors, atopic patients and patients with HIE CD138⁺ PC were enriched and analysed for expression of IgE by intracellular staining. We were able to detect at least five IgE⁺CD138⁺ PC in peripheral blood of 1/1 HIE patient, 22/22 atopic patients and in 19/24 normal donors, showing the high sensitivity of the method. Previous attempts to detect IgE⁺ B cells using intracellular immunofluorescence of sorted and EBV-transformed B cells were limited due to the low frequency of the cells and to complicated protocols [54,55]. Using ELISPOT assays IgE-secreting B cells could be detected in allergic diseases with high IgE serum titres such as HIE, atopic dermatitis and others, but very rarely in normal donors [23,24,29–32].

Although we frequently found IgE producing CD138⁺ PC in normal donors, there are enormous differences between the frequencies of IgE⁺CD138⁺ in the normal donors compared to the different patient groups. Especially in atopic patients with markedly elevated IgE serum titres more than 100-fold higher frequencies than in normal donors could be detected. Our data show that the frequency of IgE producing PC correlates highly significantly with the detected IgE serum titres (r = 0·8532; P < 0·0001), which is in accordance with data obtained by other groups using ELISPOT assays [23,24,29–32].

Why does one find such a correlation?

PC in peripheral blood are likely to be not the only major source of IgE in the serum, because IgE-secreting PC have been found in many different tissues, such as lamina propria of nasal mucosa, tonsils and mesenteric lymph nodes (depending on the method of immunization) [51,56,57]. For many years it has been believed that PC are short-lived and continuous generation of new PC would be required to maintain Ig serum titre. In this case a correlation between the frequency of – presumably recently generated – PC found in peripheral blood and the serum titre would have been expected. A constant generation of fresh IgE⁺ PC is certainly conceivable for some atopic patients due to a repeated encounter of antigens, e.g. for house dust mite and for HIE patients due to repeated/ongoing immune responses with a strong bias towards IgE production. However, recent experiments by Manz et al. and Slifka et al. have shown that IgG-producing PC in mice can be long-lived [2–4]. In this case a correlation between the frequency of PC found in peripheral blood and the serum titre would occur if a certain part of long-living PC is in a constant recirculation between different tissues via the blood or simply ‘leaching’ out of tissues into the blood. Although not formally excluded, the phenotype of peripheral blood CD138⁺ PC indicates that most of them are early PC. It has been shown recently for memory T cells by Selin et al. [58] that potentially long-living memory T cells can be replaced or reduced by newly generated memory T cells. If a similar mechanism of homeostasis existed for PC, the lifespan of a PC, although theoretically being long-lived, could also depend on how many newly generated PC enter their compartment. Due to limited space and/or availability of survival factors newly arriving PC may replace older PC.

It becomes evident that for a conclusive answer to these questions it is important to be able to distinguish between long-lived and short-lived PC and to understand which signals determine the lifetime and the homeostasis of PC. The CD138-based enrichment and staining method and the cellular affinity matrix technology provide powerful tools for further investigation of PC. While the cellular affinity matrix technology is a sophisticated approach for the comprehensive analysis of Ig-secreting cells, the CD138-based procedure allows the direct and rapid isolation and analysis of rare PC even from peripheral blood.