Introduction

During B-cell differentiation, Ig heavy chain (Ig_H) proteins exist in two forms, differing only at their carboxyl termini. The membrane form of Ig (mIg) is found on the surface of early or memory B cells, whereas the secreted form (sIg) is produced by plasma cells. The membrane heavy (H) chains have a longer and highly hydrophobic C-terminal segment, which is essential for anchoring the receptor molecules to the lipid bilayer, whereas the secretory form of H chains have a shorter and hydrophilic C-terminal region. The mIg and sIg mRNA result from differential processing of a single transcript.¹ For all isotypes the secreted form uses an intronic polyadenylation signal associated with the exon encoding the terminal constant region domain. However, if the primary transcript includes the distal membrane exon, the sequence encoding the secreted form may be removed during RNA splicing and the cell surface form of Ig is produced. During B-cell development, there is a change in both the level and ratio of the two forms of mRNA.² The ratio of secreted to membrane mRNA is indicative of a B cell's developmental stage.³ Early B (i.e. pre-B) and memory cells and their tumour analogues (i.e. lymphomas) make approximately equal amounts of secreted and membrane Ig mRNA with the exception of IgA. In contrast, terminally differentiated plasma cells and their tumour counterparts (i.e. myelomas) make 10- to 100-fold more secreted mRNA than membrane mRNA. While translational and post-translational control mechanisms play a small role in contributing to the secreted versus membrane protein ratios of Ig-producing cells, the major contribution to the phenotype comes from the differential production of secreted and membrane forms of mature Ig heavy chain mRNA.^{4, 5, 6, 7} The pattern of 3' pre-mRNA processing is different in B cells that have undergone isotype switching to IgA because the secreted form of H chain (s) mRNA tends to predominate over the corresponding membrane form (m) regardless of the stage of B-cell differentiation. Nevertheless, the development of IgA-secreting cells is associated with both an increase in total -specific mRNA and the ratio of s : m mRNA.^{6, 8, 9, 10}

Real-time quantitative reverse transcriptase–polymerase chain reaction (RT-PCR) provides a simple and highly reliable method for quantitative analysis of gene expression.¹¹ This technique uses fluorogenic hybridization probes^{11, 12, 13} or dsDNA-specific fluorescent dyes^{14, 15} to detect PCR product during amplification (real-time detection) without purification or separation by gel electrophoresis. The TaqMan system exploits the 5' to 3' exonucleolytic activity of Taq polymerase, which removes a fluorescent dye species at the 5' end of a probe only when it is specifically hybridized with an amplified target sequence. The detection of such events relies on the fluorescence quenching of another dye molecule attached near the 3' end of the probe.^{12, 16} The sensitivity of these probe systems allows the measurement of PCR product during the exponential phase of amplification before critical reactants become limiting. Analysis of the resulting PCR growth curves^{11, 17} from real-time quantitative RT-PCR provides highly sensitive and precise quantification of the target sequence with a linear dose–response over a wide range of target concentrations.

Multiplex PCR can be used in relative quantification where one primer pair amplifies the target and another primer pair amplifies the endogenous reference in the same tube. The ABI Prism^® 7700 Sequence Detector (Applied Biosystems, Foster City, CA, USA) provides a wide spectrum of emission wavelengths. By using target and control probes containing different reporter fluorescent dyes (FAM, JOE (2,7-dimethoxy-4,5-dichloro-6-carboxyfluorescein), or TET (tetrachloro-6-carboxyfluorescein)), it is possible to detect both target and control DNA simultaneously in a single reaction tube. However, accurate quantification depends on how well the various dyes are spectrally resolved. For the most accurate quantification with two probes in one tube, the reporter dyes that have the largest difference in emission maximum should be used. It has been reported that, in practice, this approach is limited to concentrations of target and internal control RNA or DNA that are within 1000-fold of each other.¹⁸

The present studies were designed to monitor changes associated with B-cell differentiation in vitro by measuring the relative levels of the secreted versus membrane forms of the H chain message using the real-time quantitative RT-PCR on the ABI Prism^® 7700 Sequence Detection System (Applied Biosystems). The changes in both forms of mRNA were measured in a homogenous cell line (Dakiki) and in heterogenous B-cell populations (human tonsillar B cells) upon treatment with cytokines such as IL-2, IL-10 and IFN-. Furthermore, an important independent marker for monitoring B-cell differentiation, human plasma cell antigen-1 (plasma cell membrane glycoprotein; HPC1¹⁹), was also examined during B-cell differentiation induced by appropriate cytokines. Development of the assay included the design of the primers and probes, optimization of primer and probe concentrations and optimization of PCR conditions. In order to handle a large number of samples, the possibility of using multiplex PCR was also examined, and the limiting primer concentrations were also defined.

Materials and Methods

Cell culture

Dakiki (ATCC, Manassas, VA, USA) and NCI-H929 (European Collection of Cell Cultures (ECACC), Wiltshire, UK) are surface IgA-positive human B lymphoma and IgA plasmacytoma cell lines, respectively. Cells from both lines were cultured at 1 10⁵/mL in RPMI 1640 (Life Technologies Inc., Gaithersburg, MD, USA) with 10% heat-inactivated FCS, 2 mmol/L L-glutamine (Trace Scientific, Castle Hill, NSW, Australia) and 50 mol/L -mercaptoethanol (Sigma, St Louis, MO, USA). Sodium pyruvate (1 mmol/L) and 1 non-essential amino acids (NEAA; Trace Scientific) were also provided for culturing NCI-H929 cells. Pokeweed mitogen-activated human tonsillar sIgM⁺ B cells depleted for sIgA⁺ cells by anti-IgA panning²⁰ were cultured at 1 10⁶ cells/mL in RPMI 1640 medium with the additional supplement of 50 IU/mL to 50 g/mL of penicillin–streptomycin (Trace Scientific). Interleukin 2 (Cetus Corp., Emeryville, CA, USA), IL-10 (R&D Systems Inc., Minneapolis, MN, USA) and IFN- (R&D Systems Inc.) were used at concentrations of 10 ng/mL for each case (whether separately or in combination). B-cell supernatants were measured for IgA secretion using ELISA.²⁰

FACS analysis

Cells were washed three times in PBS–5% FCS solution by centrifugation (400 g for 5 min) at 4°C. Cells (10⁶) were then stained with FITC-conjugated mAb to either CD38 or CD40 (Pharmingen, San Diego, CA, USA) by following the manufacturer's instruction. The stained cells were then analysed using FACScan (BD Immunocytometry, San Jose, CA, USA).

Total RNA isolation and reverse transcription

Total RNA from cells of interest were extracted using the RNeasy Mini Kit (Qiagen, Inc., Chatsworth, CA, USA). First strand cDNA was synthesized (1 h; 37°C) in a standard 50 L reaction mix in the recommended reaction buffer (MBI Fermentas, Vilnius, Lithuania) containing 2.5 g of extracted total RNA from each sample, 1 mmol/L dNTPs mix, 450 pmol oligo(dT)_12–18 (Pharmacia Biotech, Uppsala, Sweden), 80 units of ribonuclease inhibitor (RNasin; Promega, Madison, WI, USA), and 40 units of M-MuLV reverse transcriptase (MBI Fermentas).

Real-time quantitative RT-PCR

Real-time quantitative RT-PCR was performed using specific TaqMan^® primers and probes (Table 1), the ABI Prism^® 7700 and SDS analysis software (PE Applied Biosystems, Foster City, CA, USA). Primers were designed from appropriate database sequences using the PrimerExpress software (Applied Biosystems). Reactions (50 L) were assembled in 96-well reaction plates using 25 L of TaqMan^® Universal PCR Master Mix (2); 5 L of each primer (300 nmol/L final) (Pacific Oligos, Lismore, NSW, Australia); 5 L of TaqMan^® Probe (100 nmol/L for P2.-actin (TET probe; Table 1), 250 nmol/L for P1.C (FAM probe; Table 1) and 200 nmol/L for P1.HPC1 (FAM probe; Table 1)); 5 L of template (cDNA); and 5 L of water. Polymerase chain reaction was performed under the following conditions: 50°C for 2 min; 95°C for 10 min; 40 cycles at 95°C for 15 s and 60°C for 1 min. This was applied in all cases except for relative quantification of HPC1 expression, in which 60 cycles were used during the last PCR stage (95°C for 15 s and 60°C for 1 min). Controls without template (no cDNA) were also performed on each PCR plate.

Table 1 - TaqMan primers and probes.

Full table

Multiplex real-time quantitative RT-PCR

Multiplex real-time quantitative RT-PCR was developed following the optimization of the limiting primer concentrations for the target mRNA (s or m; using primer pairs C/gen.F + C/Sec.R and C/ gen.F + C/Mem.R, respectively) and endogenous reference mRNA (-actin; using primer pair -actin.F + -actin.R; Table 1). Reactions (50 L) were assembled in 96-well reaction plates using 25 L of TaqMan^® Universal PCR Master Mix (2 ); 2.5 L of each primers (-actin with either s or m, total four primers at 60 nmol/L final for each primer); 5 L of each TaqMan^® Probe (100 nmol/L for P2. -actin, 250 nmol/L for P1.C; Table 1); and 5 L of template (cDNA). Polymerase chain reaction was performed using the same program as described earlier.

Quantitative analysis and statistics

Quantitative analysis of gene expression was done using the standard curve and comparative C_T (C_T) methods,²¹ in which C_T is the threshold cycle number (the minimum number of cycles needed before the product can be detected). The arithmetic formula for the C_T method is described as the amount of target (normalized to an endogenous reference) relative to a calibrator (normalized to the same reference mRNA), and is given by 2^–CT.²¹ The following term is an example for comparing s and m, where both are normalized to -actin: C_T = C_T(s/-actin) – C_T(m/-actin). Calculation of 2^–CT then gives a relative value when comparing the target to the calibrator,²¹ which we designate in this context as N_s/m, the value for the relative mRNA levels between s (target) and m (calibrator) both normalized to -actin (endogenous reference).

Standard deviations for each data point were also calculated. In the case of HPC1 expression, it proved necessary to carry out quantitative analysis of gene expression by the standard curve method. Such curves were generated using a range of dilutions of total reverse-transcribed RNA prepared from a control cell line, with the C_T values of unknown targets allowing the corresponding 'target amount' to be read from the corresponding standard curve. The amount of HPC1 and -actin in each target sample was determined from these curves, in which the HPC1 quantity was divided by that for -actin to obtain a normalized target value (N_HPC1/-actin); that is:

Standard deviation for such (N_HPC1/-actin) quotients were calculated from the standard deviations of HPC1 and -actin values using the following formula:

in which Sa, Xa and Sb, Xb are the mean and standard deviations for HPC1 (a) and -actin (b) values, respectively. When the comparative method was used (e.g. C_T), the standard deviation of the difference was calculated from the standard deviations of the s (or m) and -actin values using the following formula:

in which S1 and S2 are the standard deviations for s or m and -actin values, respectively. Statistical analyses were also done using unpaired t-test (StatView™ SE+^Graphics (1988; Abacus Concepts Inc., Calabasas, CA, USA)).

Results

Technical validation of the TaqMan^® System for quantification of the desired target mRNA

Because the secreted and membrane forms of H chain mRNA have common 5' regions but diverge in their 3' segments,^{9, 10} we used common 5' primers and probes for both (C/Gen.F and P1.C; Table 1) with divergent 3' primers (C/Sec.R; C/Mem.R; Table 1), allowing discrimination between the two forms (Figure 1). Because two membrane exon spliceforms for the human 1 gene are known,^{22, 23} we ensured that the 3' membrane exon primer (C/Mem.R; Table 1) was downstream from the second alternative splice site to allow amplification of both forms (Figure 1). Also, primers did not discriminate between the 1 and 2 isoforms;^{22, 24} hence, in the present work we compare membrane and secreted patterns for total H chain expression. Initially, for these (and all other primer pairs used in the present study) we showed that the predicted product gel bands were obtained after RT-PCR with the standard TaqMan cycle using reverse-transcribed RNA from tonsillar B cells. The sequences of such product bands were shown to match the predicted sequences from GenBank (automated ABI 373A sequencing). In the case of the membrane products, we obtained evidence indicating that both spliceforms were expressed, which is expected for 1.²³

Figure 1.

Splicing of secreted and membrane forms of 1 and 2 heavy chain (not to scale), and relative positions of primers and probe (designations as in Table 1). Dotted lines indicate alternative splicings possible.

Full figure and legend (7K)

To determine the linear dynamic range of the template and standard curves for each gene amplification, human B-cell cDNA was amplified across a range of concentrations. The standard curves for each gene amplification were plotted as log starting RNA (ng) versus threshold cycle number (C_T). The linear dynamic range refers to the range of initial template concentrations over which accurate C_T values are obtained. As shown in Figure 2a, within the total RNA range of 0.78–100 ng, the amplification of each target (m, s and -actin) was linear with respect to the target concentration, suggesting that the use of any template concentration within this range would result in a linear progression of the amplification. A quantity of 25 ng of template was selected for TaqMan–PCR amplification of s, m and the internal control -actin. Because HPC1 expression was low using 25 ng of template, further testing was carried out with the linear concentration range of 25–500 ng (Figure 2b). As a result of this, 200 ng of template for each sample was chosen for all TaqMan–PCR amplification of HPC1 expression and its internal control -actin.

Figure 2.

Linear dynamic range of templates. The range of template RNA was from (a) 0.78 ng to 100 ng or (b) 25 ng to 500 ng at serial dilutions (1:2) for each data point, which represents the average of triplicate reactions. (a) s (; primers C/gen-F + C/Sec-R; probe P1.C), m (; primers C/gen-F + C/Mem-R; probe P1.C) and -actin (; primers -actin-F + -actin-R; probe P2.-actin) expression in human tonsillar B cells. R² values are 0.999, 0.998 and 0.997, respectively. (b) Human plasma cell antigen-1 (HPC1) (; primers HPC1-F + HPC1-R; probe P1.HPC1) and -actin (; primers as for a) expression in human plasmacytoma cell line NCI-H929. R² values are 0.993 and 0.997, respectively.

Full figure and legend (8K)

In order to use the comparative C_T method for quantification analysis, the efficiencies of target and reference should be approximately equal, and the absolute value of the slope of log input amount versus C_T should be <0.1.²¹ Efficiency plots were derived by plotting measured average C_T values between s, m, or HPC1 and -actin against corresponding amounts of input RNA on a log scale (data not shown). The resulting slopes for efficiency plots between [s/-actin] and [m/-actin] were within this guideline (0.066 and 0.09, respectively) and, therefore, the comparative C_T method was applicable.

However, the absolute value of the slope for efficiency plots between HPC1 and -actin was 0.309. Therefore, for HPC1 expression in this context the relative standard curve method (discussed earlier) was appropriate. Examples of curves for both HPC1 and -actin expression generated using a range of dilutions of total reverse-transcribed RNA prepared from a control cell line (human plasmacytoma cell line: NCI-H929) are shown in Figure 2b.

Characterization of two human B-cell lines

The present study used two human B-cell lines, Dakiki and NCI-H929, as control standards. Dakiki cells were described originally as surface IgA⁺ human lymphocytes and are known to express the IgA1 isotype,^{22, 25} but have been referred to as plasma cells.²⁶ To properly characterize Dakiki and NCI-H929 cells, both lines were subjected to FACS analysis for the expression of CD38 (which is expressed on terminally differentiated plasmacytoid B cells)^{27, 28} and CD40 (which is expressed on all stages of B cells except plasma cells).^{27, 28} The cells were also examined for their relative levels of secreted (s) and membrane (m) forms of IgA H chain mRNA and HPC1 mRNA expression. Dakiki cells expressed surface IgA and CD40, but not CD38, with converse staining seen for NCI-H929 cells (Figure 3). Also, Dakiki cells showed only a small preponderance of s unlike the elevated levels for NCI-H929, and only the latter had significant levels of HPC1 mRNA (Figure 3). These results showed that Dakiki cells had characteristics consistent with a memory B lymphocyte origin,²⁸ and were clearly distinguishable from the plasmacytoid NCI-H929 cells.

Figure 3.

Characterization of Dakiki cells and NCI-H929 cells. (a,b) FACS analyses of Dakiki and NCI-H929 cells, respectively, for the indicated markers. X-axes show log fluorescent intensities. (c) Relative expression of s : m (N_s/m) and human plasma cell antigen-1 (HPC1) (N_HPC1/-actin) mRNA in Dakiki and NCI-H929 cells. Primers are as for Figure 2; relative expression calculations are as described in Materials and Methods.

Full figure and legend (83K)

Differentiation-related changes induced by IL-2 and IL-10 in Dakiki cells

Because data of the present study showed Dakiki cells conformed to the definition of memory cells by standard criteria, we used this cell line as a model to study the differentiation of IgA producers. The changes of s, m and HPC1 mRNA in IgA differentiation were investigated in Dakiki cells treated with IL-2 IL-10 because the combination of IL-2 + IL-10 has been proven to promote the differentiation of memory B cells into plasma cells.^{29, 30} As shown in Figure 4a, the relative expression of s versus m remained unchanged in culture without any treatment (see control samples at day 0 and day 7). However, upon cytokine treatment with either IL-2, IL-10 or IL-2 + IL-10, N_s/m increased with time (from day 5). In IL-2 + IL-10-treated cells, the increases were dramatic, being towards that of the level of NCI-H929 cells (plasmacytoma cells). Such results suggest that sIgA⁺ Dakiki cells can undergo differentiation-related changes towards a plasmacytoid phenotype when treated with IL-2 + IL-10. These changes also correlated with HPC1 mRNA expression. Baseline HPC1 expression in Dakiki cells was very low, but increased dramatically when the cells were treated with IL-2 alone, IL-10 alone or IL-2 + IL-10 (Figure 4b). No baseline changes in control cells occurred between day 0 and day 7. After 7 days in culture, IgA levels produced by Dakiki cells increased progressively (Figure 5). Without cytokines, IgA cumulative levels were significantly less (P < 0.01) at 7 days compared with the cytokines; whereas with IL-2 + IL-10, IgA levels peaked at day 6. The increase in IgA secretion induced in Dakiki cells by these cytokines occurred in concert with the increases observed in the relative quantities of s versus m mRNA and HPC1 mRNA expression in these cells (Figure 4), although the increase in antibody production was consistently less than the relative increase of s mRNA. In any case, antibody measurement alone cannot determine the relative cellular expression of membrane and secreted forms, and is thus a poor yardstick of differentiation.

Figure 4.

Time course of (a) s mRNA relative expression and (b) human plasma cell antigen-1 (HPC1) mRNA expression in Dakiki cells treated with IL-2 and IL-10 (Materials and Methods). Primers are as for Figure 2. Cells and supernatants were harvested at indicated times (D0, day 0, etc.). RNA was extracted from the cells, reverse transcribed, amplified and quantified for s : m and HPC1 mRNA relative expression. All data are expressed as the mean SEM from triplicate polymerase chain reaction amplifications, and are representative results from three assays. C, control; NCI, control NCI-H929 cells.

Full figure and legend (31K)

Figure 5.

Time course of IgA secretion in Dakiki cells induced by IL-2 and/or IL-10 (assayed by ELISA). Post-treatment supernatants were obtained as for Figure 4. D1, D2, etc. indicates the day of post-treatment; D7cont., cells maintained without treatment in tandem with the treated cells for 7 days.

Full figure and legend (62K)

A measured change in the relative amounts of s versus m must be a result of a change in the production of the respective forms at the level of splicing. However, in principle this could occur by a significant decrease in m without any change in s (i.e. without a net increase in the secreted form). To test this we compared the levels of s and m in Dakiki cells before and after cytokine treatments, each separately normalized to -actin. We found (Figure 6) pronounced net increases in s after IL-2 and IL-10 treatment, but levels of Dakiki m mRNA did not change significantly. Because the membrane and secreted forms are produced from a common transcriptional unit, this indicates that the observed changes in the relative quantities of s versus m result from both increased levels of steady state total mRNA and altered processing in favour of the secreted form.

Figure 6.

Relative expression of s () compared to m () (each normalized by -actin; see Materials and Methods) in Dakiki cells treated with IL-2 and/or IL-10. C, control. D0, D1, etc. indicate day 0, day 1, etc. after commencement of experiment. Primers are as for Figure 2.

Full figure and legend (35K)

Relative levels of s, m and HPC1 mRNA in freshly isolated human tonsillar sIgM⁺(sIgA⁺-depleted) cells

Human tonsillar sIgM⁺ B cells (depleted by panning for sIgA⁺ cells) were also examined for changes in s versus m mRNA expression and HPC1 mRNA levels induced by combinations of known cytokines. After panning the remaining sIgA⁺, cells were usually equivalent to 1% of the fractionated population, as detected by FACS analysis (data not shown). This IgM⁺ (sIgA⁺-depleted) population could be shown to still contain cells with both genomic class switch rearrangement (assessed by the digestion–circularization PCR method²⁵) and mature Ig transcripts (assessed by sensitive RT-PCR with a J_H consensus primer and 1 and 2 constant region primers; data not shown). It proved useful to isolate the sIgM⁺ (sIgA⁺-depleted) tonsillar cell population for these studies to reduce the background level of IgA production (including any antibody deriving from IgA- committed plasmacytoid cells initially present), while retaining sufficient cells precommitted by class switching to heavy chain for subsequent experimentation.

Freshly purified human tonsillar sIgM⁺ (sIgA⁺-depleted) cells showed a relatively low s : m mRNA ratio (Figure 7a), although this varied from batch to batch (data not shown). However, after treatment with IL-2 + IL-10, the s : m mRNA ratio (N_s/m) and IgA secretion were increased strongly in these cells (Figure 7a).

Figure 7.

Cytokines in regulation of (a) s relative mRNA expression (), IgA secretion () and (b) human plasma cell antigen-1 (HPC1) mRNA expression in human tonsillar sIgM⁺ (sIgA⁺-depleted) B cells. PWM-activated tonsillar sIgM⁺ (sIgA⁺-depleted) B cells were cultured with recombinant IL-2 + IL-10 or IL-2 + IL-10 + IFN- (10 ng/mL for each cytokine) at 1 10⁶/mL and 2 mL/well in triplicate wells for 6 days. Control cells were treated with PWM only. Cells and supernatants were harvested on days 0 and 6. RNA were extracted from the cells, reverse transcribed, amplified and quantified for s, m and HPC1 mRNA expression. Data for N_s/m determinations are representative of results from six assays. Data for N_HPC1/-actin are representative results from two assays. All data are expressed as the mean SEM from triplicate polymerase chain reaction amplifications. Primers are as for Figure 2. The supernatants were assayed for IgA by ELISA.

Full figure and legend (44K)

Relative expression levels of the HPC1 marker in fresh purified tonsillar sIgM⁺ (sIgA⁺-depleted) cells (Figure 7b) also showed batch variation. The N_{(HPC1/-actin)} values for different preparations of these tonsillar cells ranged from non-detectable to levels comparable with NCI-H929. However, treatment of such tonsillar cells with IL-2 + IL-10 consistently resulted in increased relative HPC1 expression (Figure 7b), but the magnitude of such increases themselves varied with different tonsillar cell batches.

Because of the known roles of IFN- in various aspects of B-cell regulation,³¹ we also examined the regulation of B-cell differention by IFN- in human tonsillar sIgM⁺ (sIgA⁺-depleted) B cells as assessed by TaqMan. The addition of IFN- suppressed the increase in HPC1 expression mediated by IL-2 + IL-10 treatment at day 6 (Figure 7b) without affecting IgA secretion or s : m mRNA ratios. This indicated that H chain and HPC1 transcription are, at least in part, independently regulated. Overall analysis of s : m mRN&Agr; expression also showed that s mRNA expression predominates over m for all tonsillar B-cell populations tested (16 independent samples; data not shown) as for Dakiki cells (Figure 6).

Discussion

Increased IgA levels from a cell population that has already undergone isotype switching to IgA can result from: (i) expansion of IgA memory cells; or (ii) their differentiation to IgA plasma cells. The possibility that only (i) is occurring can, in principle, be excluded by measuring secreted versus membrane (s vs m) H chain mRNA levels. Memory cells can indeed proliferate.^{32, 33} However, during proliferation, if the cell remains fixed at this differentiation stage, the average amount of antibody produced per cell would not change and neither would the s versus m ratio. However, with a mixed population of cells, it might be difficult, in practice, to tell whether IgA⁺ cells are significantly proliferating. Memory cells can also secrete antibody,^{32, 33} indicating that although the pathways for selectively promoting the production of secretory form mRNA are not active in memory cells (as opposed to plasma cells), the pathways for the secretion of mature Ig protein are functional.

Studies on the relative expression of s and m in Dakiki cells and in ex vivo tonsillar B cells indicate that the cytokine-mediated increase in the ratio of s versus m is because of strongly enhanced levels of the secreted form with little change in the membrane form relative to their original baselines. Because s levels after cytokine stimulation were much higher than the combined baseline s + m levels (Figure 6), this indicates that a net increase in steady-state levels of the s form occurred (either through enhanced chain transcription or message stabilization), as well as altered mRNA processing in favour of the secreted form. The observed H chain mRNA changes occur both in a homogenous cell line and in heterogenous B-cell populations, indicating that differentiation-related changes in mRNA processing can be measured accurately by automated quantitative PCR. As terminally differentiated plasma cells cannot proliferate, this approach can distinguish between clonal expansion of memory IgA cells and cells differentiating towards a plasmacytoid phenotype.

We observed significantly increased IgA production in Dakiki cells treated with IL-2 + IL-10 (Figure 5), but this was not as pronounced as the relative increase in the s mRNA form (Figure 4). Protein levels will not necessarily show a one-to-one correspondence with changes in abundances of the mRNA that encode them because they are subject to potential factors such as post-transcriptional regulatory influences, translational efficiency, translational control mechanisms and protein stabilities. Also, in the present study we have assessed secreted protein only; clearly additional factors could modulate the efficiency of the secretory pathway. Any of these potential influences could also be abnormally strong in Dakiki cells as a consequence of their transformed state. However, using IgA measurement and an independent marker (HPC1), we have shown that certain changes characteristic of B-cell differentiation occur in Dakiki cells treated with appropriate cytokines, and that these effects occur in tandem with strong s : m changes that can be measured by quantitative PCR. More work is required to assess how closely the observed differentiation-related events in Dakiki cells correspond to true differentiation in normal B cells, or whether additional signals are required for an optimal differentiation response.

The results also indicate that freshly isolated sIgA⁺ B cells, sIgM⁺ (sIgA⁺-depleted) B cells and Dakiki cells have higher steady-state levels of the secreted than the membrane form of IgA mRNA. This result is in agreement with other studies, which find that s mRNA predominates regardless of the stage of B-cell differentiation.^{9, 10} This may be due to enhancement of the efficiency of s polyadenylation by a specific sequence within the s–m intron or, possibly, also due to pre-mRNA processing by differential splicing.^{8, 9} (The production of either secreted or membrane forms of mRNA results from competition between utilization of the secreted form polyadenylation signal and splicing between the donor and acceptor splice sites for the membrane form.)

Data also showed that IL-2 + IL-10 induces the sIgM⁺ (sIgA⁺-depleted) B cells to develop toward a phenotype with features suggestive of plasma cells with a high rate of IgA secretion, a high ratio of s versus m, and a high expression of HPC1 mRNA. As no stimuli capable of effecting class switching were applied, it is evident that the remaining sIgA⁺ cells in the fractionated tonsillar population underwent differentiation-related changes in IgA processing and production. The addition of IFN- increased both IgA secretion and the ratio of s verus m from the fractionated tonsillar cells, but dramatically reduced HPC1 mRNA relative levels (Figure 7b). Although IFN- can thus modulate B-cell differentiation induced by IL-2 + IL-10, it had opposing effects on H chain mRNA processing versus HPC1 expression Figure 7. As such, these two processes that are linked to B-cell differentiation are clearly not subject to identical regulatory influences.

Our general results indicate that quantification of secreted versus membrane forms of human IgA mRNA allows differentiation-related changes in B-cell populations to be monitored in vitro. As all Ig_H chains have membrane and secreted forms, this approach can be applied generally to all isotypes and, in conjunction with analysis of other B cell markers such as HPC1, has utility in defining cytokine- mediated effects on cell differentiation. In combination with quantitative assays for Ig class switching,²⁵ we have used the TaqMan–PCR approach to help distinguish cytokine- mediated B-cell differentiation from true de novo switching in tonsillar cell populations (SD Xiang, EM Benson, Ian S Dunn, unpubl. data, 2000).

References

1. Early P, Rogers J, Davis M et al. Two mRNAs can be produced from a single immunoglobulin gene by alternative RNA processing pathways. Cell 1980; 20: 313–19. | Article | PubMed | ISI | ChemPort |
2. Lamson G, Koshland ME. Changes in J chain and chain RNA expression as a function of B cell differentiation. J. Exp. Med. 1984; 160: 877–92. | Article | PubMed | ISI | ChemPort |
3. Guise JW, Galli G, Nevins JR, Tucker PW. Developmental regulation of secreted and membrane forms of immunoglobulin chain. In: Honjo FWA, Rabbitts TH, (eds). Immunoglobulin Genes. San Diego: Academic Press Inc, 1989; 275–302.
4. Galli G, Guise M, McDevitt A, Tucker PW, Nevins JR. Relative position and strengths of poly(A) sites as well as transcription termination are critical to membrane versus secreted -chain expression during B-cell development. Gene Dev. 1987; 1: 471–81. | PubMed | ChemPort |
5. Kobrin BJ, Milcarek C, Morrison SL. Sequences near the 39 secretion-specific polyadenylation site influence levels of secretion-specific and membrane-specific IgG2b mRNA in myeloma cells. Mol. Cell. Biol. 1986; 6: 1687–97. | PubMed | ChemPort |
6. Lassman CR, Matis S, Hall BL, Toppmeyer DL, Milcarek C. Plasma cell-regulated polyadenylation at the IgG2b secretion-specific poly(A) site. J. Immunol. 1992; 148: 1251–60. | PubMed | ChemPort |
7. Peterson ML, Perry RP. The regulated production of _m and _s mRNA is dependent on the relative efficiencies of _s poly(A) site usage and the C4 to M1 splice. Mol. Cell. Biol. 1989; 9: 726–36. | PubMed | ISI | ChemPort |
8. Coyle JH, Borinstein SC, Woodward EO, Lebman DA. Predominant usage of the proximal poly(A) site in a mRNAs is not intrinsic to the 3' termini. Int. Immunol. 1998; 10: 669–78. | Article | PubMed | ChemPort |
9. Seipelt RL, Peterson ML. Alternative processing of IgA pre-mRNA responds like IgM to alterations in the efficiency of the competing splice and cleavage–polyadenylation reactions. Mol. Immunol. 1995; 32: 277–85. | Article | PubMed | ChemPort |
10. Lebman DA, Park MJ, Fatica R, Zhang Z. Regulation of usage of membrane and secreted 3' termini of mRNA differs from mRNA. J. Immunol. 1992; 148: 3282–9. | PubMed | ChemPort |
11. Gibson UEM, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR. Genome Methods 1996; 6: 995–01. | ChemPort |
12. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Methods 1996; 6: 986–94. | ChemPort |
13. Tyagi S, Kramer FR. Molecular beacons: Probes that fluoresce upon hybridization. Nature Biotechnol. 1996; 14: 303–8. | Article |
14. Higuchi R, Dollinger G, Walsh PS, Griffith R. Simultaneous amplification and detection of specific DNA sequences. Biotechnology 1992; 10: 41317.
15. Wittwer CT, Hermann MG, Moss AA, Rasmussen RP. Continuous fluorescence monitoring of rapid cycle DNA amplification. Biotechniques 1997; 22: 134–8.
16. Livak KJ, Flood SJ, Marmaro J, Giusti W, Deetz K. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods Appl. 1995; 4: 357–62. | PubMed | ISI | ChemPort |
17. Higuchi R, Fockler C, Dollinger G, Watson R. Kinetic PCR analysis: Real-time monitoring of DNA amplification reactions. Biotechnology 1993; 11: 1026–30. | Article | PubMed | ChemPort |
18. Williams SJ, Schwer C, Krishnarao AS, Heid C, Karger BL, Williams PM. Quantitative competitive polymerase chain reaction: Analysis of amplified products of the HIV-1 gag gene by capillary electrophoresis with laser-induced fluorescence detection. Anal. Biochem. 1996; 236: 146–52. | Article | PubMed | ChemPort |
19. Goding JW, Shen FW. Structure of the murine plasma cell alloantigen PC-1: Comparison with the receptor for transferrin. J. Immunol. 1982; 129: 2636–40. | PubMed | ChemPort |
20. Bertolini JN, Bertolini J, Thean E, Benson EM. Human low molecular weight B cell growth factor induces surface IgM⁺/A^– B cells to express and secrete IgA. J. Immunol. 1992; 149: 1771–8. | PubMed | ChemPort |
21. PE-Applied-Biosystems. Relative quantitation of gene expression. In: User Bulletin Vol. 2. 1997; 3, 11–14.
22. Yu L, Peng C, Starnes SM, Liou RS, Chang TW. Two isoforms of human membrane-bound Ig resulting from alternative mRNA splicing in the membrane segment. J. Immunol. 1990; 145: 3932–6. | PubMed | ISI | ChemPort |
23. Leduc I, Drouet M, Bodinier MC, Helal A, Cogne M. Membrane isoforms of human immunoglobulins of the A1 and A2 isotypes: Structural and functional study. Immunology 1997; 90: 330–6. | Article | PubMed | ChemPort |
24. Flanagan JG, LeFranc M-P, Rabbits TH. Mechanisms of divergence and convergence of the human immunoglobulin 1 and 2 constant region gene sequences. Cell 1984; 36: 681–8. | Article | PubMed | ChemPort |
25. Weckert HA, Hughes JAI, Benson EM, Dunn IS. Quantifiable analysis of human immunoglobulin heavy chain class-switch recombination to all isotypes. J. Immunol. Methods 2000; 233: 141–58. | Article | PubMed | ChemPort |
26. O'Dorisio MS, Shannon BT, Fleshman DJ, Campolito LB. Identification of high affinity receptors for vasoactive intestinal peptide on human lymphocytes of B cell lineage. J. Immunol. 1989; 142: 3533–6. | PubMed | ChemPort |
27. Tsubata T, Nishikawa SI. Molecular and cellular aspects of early B cell development. Curr. Opin. Immunol. 1991; 3: 186–92. | Article | PubMed | ChemPort |
28. Janeway Jr, CA, Travers P. The development of B lymphocytes. In: Palmer, R. (ed.). Immuno Biology: The Immune System in Health and Disease. London: Current Biology Ltd., 1994; 1–33.
29. Rousset F, Garcia E, Defrance T et al. IL-10 is a potent growth and differentiation factor for activated human B lymphocytes. Proc. Natl Acad. Sci. USA 1992; 89: 1890–3. | Article | PubMed | ChemPort |
30. Rousset F, Peyrol S, Garcia E et al. Long-term cultured CD40-activated B lymphocytes differentiate into plasma cells in response to IL-10 but not IL-4. Int. Immunol. 1995; 7: 1243–53. | PubMed | ChemPort |
31. Jelinek DF, Splawski JB, Lipsky PE. The role of interleukin-2 and interferon-gamma in human B cell activation, growth and differentiation. Eur. J. Immunol. 1986; 16: 925–32. | PubMed | ChemPort |
32. McHeyzer-Williams MG, Nossal GJ, Lalor PA. Molecular characterization of single memory B cells. Nature 1991; 350: 502–5. | Article | PubMed | ChemPort |
33. McHeyzer-Williams MG, Ahmed R. B cell memory and the long-lived plasma cell. Curr. Opin. Immunol. 1999; 11: 172–9. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We are grateful to the Biomolecular Resource Facility (BRF) Unit in the John Curtin School of Medical Research at the Australian National University (Canberra, ACT, Australia) for providing the ABI PRISM^® 7700 Sequence Detection System. All real-time quantitative PCR were performed at the BRF.