Enforced CD19 Expression Leads to Growth Inhibition and Reduced Tumorigenicity

The tumorigenic KMS-519 and the nontumorigenic U-26620 human myeloma cell lines and the human erythroleukemia cell line K-562 were maintained in RPMI-1640 (Nissui, Tokyo, Japan) supplemented with 10% fetal calf serum (FCS; M.A. Bioproducts, Walkersville, MD). Growth curves of U-266 were performed in the presence or absence of exogenous recombinant human interleukin-6 (rhIL-6; 2 ng/mL). For cell culture experiments in serum-free medium, we used AlbuMAX (200 μg/mL), bovine insulin (2 μg/mL), and human transferrin (2 μg/mL; all purchased from GIBCO BRL, Grand Island, NY).

Full-length CD19 cDNA was cloned by reverse transcriptase-polymerase chain reaction (RT-PCR) from the human Burkitt’s lymphoma cell line (Raji) using primers CD19 F1 and CD19 R1 described below. The cloned PCR product was directly ligated into the TA vector pCR-TOPO (Invitrogen, San Diego, CA) to give pCR-CD19. A truncated form of CD19, which has only the extracellular and transmembrane domains, but lacks the cytoplasmic domain, was amplified by PCR from pCR-CD19 using primers CD19 F1 and CD19 R2 described below and directly cloned to the TA-vector to give pCR-ΔCD19 (Fig 1A). Both cDNA inserts in pCR-CD19 and pCR-ΔCD19 were sequenced by the dyedeoxy termination method using (ABI PRISM) dye terminator cycle sequencing ready reaction kit (Perkin Elmer, Applied Biosystems Division, Foster City, CA) according to the manufacturer’s instructions. The cloned cDNA fragments were then eluted and subcloned into theEcoRI site of the mammalian expression vector pCI-neo (Promega, Madison, WI) driven by cytomegalovirus (CMV) immediate early enhancer/promoter to give pCI-CD19 (full-length) and pCI-ΔCD19 (truncated form). The orientation of the inserts in the expression vector was verified by Xho I restriction for the full-length construct and Sma I for the truncated form (pCI-ΔCD19).

The human CD19⁻ KMS-5 myeloma cell line was transfected by pCI-neo vector (neo-control), pCI-CD19 (K-19⁺), or pCI-ΔCD19 (K-Δ19) by lipofection. Cells were plated at 3 × 10⁵ in 6-well plates overnight, and on the following day, cells were washed in serum-free medium and transfected by a mixture of 5 μg DNA and 20 μg lipofectin (GIBCO BRL) and incubated at 37°C with 5% CO₂. After 10 hours of incubation, complete medium with 10% FCS was added and incubation was continued for a total of 72 hours. Transfected cells were then diluted 1:10 in medium and selected in geniticin (GIBCO BRL) at 500 μg/mL. By 8 weeks, resistant clones started to appear, and subcloning was performed at least twice to obtain single-cell clones by limiting dilution in 96-well plates. The K-562 erythroleukemia cell line was transfected by lipofection in an identical manner. The cells were selected in 1 mg/mL geniticin. The U-266 human myeloma cell line was transfected by electroporation of 3 × 10⁷ cells with 300 μg of each recombinant plasmid in 600 μL of Hank’s buffered solution. Electroporation was performed by the Gene Pulser (Bio-Rad Laboratories, Tokyo, Japan) at 210 V and 1,440 μF capacitance, followed by selection in medium containing 1 mg/mL geniticin. Transfected U-266 and K-562 expressing the surface CD19 were sorted after 2 weeks of selection, maintained in selection medium, and subsequently used for experiments.

Cells were harvested and stained with either phycoerythrin (PE)-labeled anti-IgG isotypic control or PE-labeled anti-CD19 (Coulter, Hialeah, FL) as described previously21and were analyzed by the cell sorter (Epics Elite ESP; Coulter).

mRNA was extracted by oligo-dT latex beads (Oligotex-dT30; Takara, Kyoto, Japan) according to the manufacturer’s instructions. Isolated mRNA was reverse transcribed at 37°C for 60 minutes in a reaction mixture containing 100 pmol/L random hexamer (Pharmacia Biotech, Tokyo, Japan), 0.5 mmol/L each deoxynucleotide-5-triphosphate (dNTPs), RT buffer (50 mmol/L Tris-HCl, pH 8.8, 7.5 mmol/L KCl, 3 mmol/L MgCl₂), 10 mmol/L dithiothreitol (DTT), 20 U of human placental ribonuclease inhibitor (Takara), and 20 U of reverse transcriptase Superscript II (GIBCO BRL, Gaithersburg, MD).

Full-length CD19 cDNA was amplified by RT-PCR using primers designed according to the published sequence5 as follows: CD19 F1: 5′-GGA GAG TCT GAC CAC CAT GCC ACC T-3′ (nucleotides 1-25) and CD19 R1: 5′-AAG GGG ACT GGA AGT GTC ACT GGC AT-3′ (nucleotides 1878-1853). The truncated construct was amplified from pCI-CD19 using primers CD19 F1 and CD19 R2: 5′-GGC TCT TTG AAG ATG AAG AAT GCC CAC-3′ (nucleotides 964-938). For detection of integration of the pCI-neo into the genome of mock transfectants, the following pCI-neo vector specific primers (see Fig 1A) were used to amplify a 222-bp fragment from genomic DNA; pCIF: 5′-ATC CAC TTT GCC TTT CTC TCC ACA-3′ (nucleotides 998-1021), and pCIR: 5′-CTG CAT TCT AGT TGT GGT TTG TCC-3′ (nucleotides 1219-1146). PCR cloning of the full-length CD19 was performed in a 2-step cycle of 94°C for 30 seconds denaturation and 72°C annealing/extension for 4 minutes (7 cycles), followed by 32 cycles of denaturation at 94°C for 30 seconds and annealing/extension at 67°C for 4 minutes with rTth (Perkin Elmer) and Vent (New England Biolabs, Beverly, MA) mixture of DNA polymerases and a manual hot start by AmpliWax (Perkin Elmer). Thermal cycling for the pCI-neo sequences was performed by rTaq (Takara) at 94°C denaturation for 1 minute, annealing at 65°C for 1 minute, and extension at 72°C for 1 minute. The truncated fragment was amplified using the same conditions except for annealing at 57°C for 1 minute.

Soft agar (1.6% wt/vol) was prepared by autoclaving Bacto-agar (DIFCO, Detroit, MI) in distilled water just before use. The bottom agar layer (2.1 mL/well) contained 1.6% agar:2× RPMI/20% FCS:1× RPMI/10% FCS without cells in a volume ratio of 1:1:1, respectively, as a final agar concentration of 0.53%. The top agar layer (0.9 mL/well) contained 1.6% agar:2× RPMI/20% FCS:1× RPMI/10% FCS with cells in a ratio 1:1:2, respectively, with a final agar concentration of 0.4%. The number of cells plated for each clone was 5 × 10²/mL or 5 × 10³/mL. Plates were incubated at 37°C with 5% CO₂ for 2 weeks and the colonies were counted and photographed under a phase contrast microscope (Nikon, Tokyo, Japan).

Severe combined immunodeficiency (SCID)-hIL-6 transgenic mice22 provided by Chugai Pharmaceutical Co, Ltd (Shizuoka, Japan) were used for experiments between 4 to 6 weeks after birth. The tumorigenic KMS-5 transfectants (K-19⁺ and neo-control cells) were harvested (5 × 10⁶), suspended in 200 μL phosphate-buffered saline (PBS), and subcutaneously inoculated into the flanks or the back of the mice on the same day. Mice were observed for tumor formation every 4 days. After 5 weeks, mice were killed, tumor was resected, and its growth was evaluated by volume according to the formula: 1/2ab2 (in microliters), where “a” is the longest and “b” is the shortest axes of the tumor. Statistical analysis was performed by the Student’s t-test (n = 5 for each clone). This experiment was reviewed by the Committee of the Ethics on Animal Experiment in Yamaguchi University School of Medicine and performed under the control of the Guideline for Animal Experiment in Yamaguchi University School of Medicine and The Law (No. 105) and Notification (No. 6) of the Government.

To characterize the transfected cells and confirm the expression of the CD19 transgene, we used flow cytometry to screen clones obtained by limiting dilution. Twenty-seven single-cell clones expressing variable levels of surface CD19 were identified, and of those, 3 representative clones were further identified and examined. One clone that expressed low-level CD19 (K-19⁺ clone #8) and 2 clones expressing high levels (K-19⁺⁺ clones #3 and #11) were used for the following experiments. We also characterized another clone that expressed high levels of truncated CD19 (K-Δ19) on its surface (Fig 1B). The mRNA expression of the transgenes was examined by RT-PCR. Amplification of reverse transcribed mRNA using the CD19 F1 and CD19 R1 primer pair produced a single band of the expected length from K-19⁺ cells, but not from K-Δ19 or neo-control transfected cells (Fig 1C, top). On the other hand, CD19 F1 and CD19 R2 primers produced amplification products from all clones except neo-control transfected cells (Fig 1C, middle). Neo-control transfected cells were also cloned by limiting dilution to isolate single-cell clones. The integration of the expression vector into the genome of 3 independent neo-control transfectants (M no. 1, M no. 2, and M no. 3) was verified by amplifying genomic DNA using vector specific primers from 3 mock clones, but not from the KMS-5 parent cells (Fig 1D). These data show the expression of varying levels of CD19 in stably transfected clones.

To examine the biological effects of CD19 expression on myeloma cells, we first evaluated the growth rate of transfected clones cultured in RPMI 1640 supplemented with 10% FCS. Interestingly, K-19⁺clones showed slower growth rate compared with either neo-control or K-Δ19 cells (Fig 2). Moreover, the growth-inhibitory effect seemed to be related to the level of CD19 expression. High CD19 expression (clones #3 and #11) was associated with markedly slower growth, and low expression (clone #8) was associated with less marked growth inhibition. Three neo-control clones (M no. 1, M no. 2, and M no. 3) that we used in our experiments had an almost identical growth pattern. The growth rate of neo-control clones and K-Δ19 cells was almost identical. These data show that CD19 appears to exert a growth-inhibitory effect on myeloma cells in an expression level-dependent manner. These effects are likely related to intracellular signaling through CD19, because they were not observed in the truncated CD19 transfectants or in neo-control clones.

Another biological feature of malignant cells that we have investigated is anchorage-independent growth capability. As evaluated by soft agar colony assay, K-19⁺ cells showed markedly lower capability for colony formation. Plating 5 × 10² cells/mL in triplicate experiments, both neo-control and K-Δ19 cells formed at least 3 large colonies in each well after 2 weeks, whereas K-19⁺ cells formed lower numbers of smaller colonies (Fig 3 and Table 1). Low CD19 expression clone (K-19⁺ #8) formed 2 small colonies in only 1 well, and high CD19 expression clones (K-19⁺⁺ #3 and #11) did not form any colonies, even when the cultures were kept for a longer incubation period. Increasing the number of plated cells to 5 × 10³ cells/mL also showed a lower capability for colony formation by K-19⁺ clones compared with neo-control clones (Table 1). These data show that the expression of CD19 in myeloma cells resulted in reduced capacity for anchorage-independent growth. The level of this inhibitory effect appears to be well-correlated with the level of CD19 expression in myeloma cells.

We next tried to verify if CD19 exerts a similar growth inhibitory effect in vivo by subcutaneous inoculation of CD19 transfectants in SCID-hIL-6 transgenic mice. Mice were observed for tumor formation every 4 days. The earliest observed tumors appeared after 17 days in mice injected with neo-control clones. After 5 weeks, mice were killed, and the tumor was resected and evaluated according to the equation described in Materials and Methods. The time required for tumor formation and the tumor volume correlated well with the in vitro growth characteristics of the different K-19⁺ clones. The clones expressing CD19 formed tumors at least 1 week (K-19⁺ #8) or more (K-19⁺⁺ #3 and #11) slower than neo-control clones. Also, the tumor volume of K-19⁺ clones was significantly smaller than that of mock clones. The statistical data are drawn from 5 animals injected with each clone, and the P value is shown above the horizontal bars (Fig 4). When K-19⁺ clones (K-19⁺ #8, K-19⁺⁺ #3, and #11) were compared with each other, there was significant difference between them, specifically between clone #8, expressing low-level CD19, and the high expression clones #3 and #11. These in vivo data support and confirm the in vitro observations of the growth-inhibitory effect of CD19 in myeloma cells. It also shows that the level of expression of CD19 correlates with the growth inhibitory effect.

To verify whether the growth-inhibitory effect of CD19 is specific to myeloma cells, we have used 2 cell lines, U-266 nontumorigenic myeloma cell line and K-562 human erythroleukemia cell line, to express the intact and truncated forms of CD19. Because the U-266 cells proliferate in response to exogenous IL-6, we have tested these cells both in the presence and absence of exogenous rhIL-6 (2 ng/mL). Intact CD19 induced growth inhibition in U-266 cell line both in the presence or absence of exogenous IL-6 (Fig 5A). However, this growth-inhibitory effect was also observed in K-562 (Fig 5B). Taken together, these data further support the observation of the growth-inhibitory effect exerted by CD19 and suggest that this effect is not restricted to myeloma or B-lineage cells.

An important feature of human MM plasma cells is the specific loss of the B-cell surface protein, CD19, in contrast to normal plasma cells, which do express this molecule.2 The loss of CD19 in human myeloma cells was correlated with the absence of its positive regulatory factor, B-cell–specific activator protein (BSAP), encoded by Pax-5 gene.3 However, the biological significance and the consequences of loss of CD19 in human myeloma has been unclear. In this study, we have generated clones of myeloma cell lines that stably express CD19 and studied the effects of transgene expression on cell growth and tumorigenicity. We observed a striking growth inhibition in cells expressing intact CD19, with the degree of inhibition related to the level of expression of the transgene (Fig 2); high expression clones had slower growth rates. This in vitro growth inhibition was further confirmed by in vivo studies performed by subcutaneous inoculation in SCID-hIL-6 transgenic mice (Fig 4). We observed a good correlation between in vitro and in vivo growth characteristics of the CD19-transfected cells. The parental cell line, KMS-5, is IL-6 independent and rapidly proliferates with a doubling time of less than 2 days.23 This might explain the fact that cell growth was not completely arrested, thus allowing isolation of stable transfectants.

Introduction of intact and truncated CD19 transgenes into another myeloma cell line, U-266, and a myeloid cell line, K-562, confirmed that the growth-inhibitory effect is specific to CD19 (Fig 5). The in vivo effect of CD19 could not be confirmed in U-266 because this line is nontumorigenic. However, these experiments showed that this effect was not specifically restricted to myeloma cells because it could be observed in the erythroleukemia cell line.

Another biological feature clearly modified by the introduction of CD19 into myeloma cells was anchorage-independent growth in soft agar. The number and size of colonies formed by CD19 expressing clones was significantly lower than that of neo-control and truncated CD19 clones (Fig 3 and Table 1). The failure of high expression clones to form colonies in soft agar was not merely due to their slower growth rates, because they did not form colonies even when kept in culture for longer periods. A 10-fold increase in the number of plated cells did not result in marked increase of the number of neo-control colonies, but intact CD19 transfectants could form few small colonies (Table 1). Also, this effect seems to be related to intact CD19 cytoplasmic domain and hence to CD19 signaling, because the truncated mutant form failed to exert similar effects.

The mechanism(s) by which CD19 induces these effects remains to be elucidated. Systematic study of the major candidate signaling pathways (mainly PI-3 kinase and MAP kinase) downstream of CD19 is ongoing in our laboratory. Preliminary experiments on K-19⁺ cells stimulated with serum showed that phosphorylated ERK-1 and ERK-2 MAP kinases were downregulated in K-19⁺⁺ clones, but were constitutively phosphorylated in neo-control and K-Δ19 clones (data not shown). This suggests that CD19 might be working as a molecular switch, which turns the growth signals off in appropriate time, and its absence may leave constitutively activated proliferation-promoting signals on, at least regarding the MAPK pathway. Nevertheless, it is surprising that expression of a surface protein such as CD19, unlike other known tumor suppressor gene products, could inhibit the growth of these cell lines both in vitro and in vivo. Taken together, CD19 may act in human myeloma as a new type of tumor suppressor gene product that is a transmembrane-signal transducing molecule.

We also observed decreased sensitivity of CD19 transfected cells to apoptotic stimuli such as dexamethasone, insulin, and serum starvation in comparison to neo-control and truncated CD19 clones. The difference was less striking than that of growth inhibition and tumorigenicity. Because the parental cell line, KMS-5, is already resistant to apoptotic stimuli, the observed effect could be due to CD19 expression, CD19-mediated amplification of antiapoptotic signals elicited by another pathway, or alternatively, due to the proliferation-inhibitory effect of CD19 expression (data not shown).

Our data suggest that loss of CD19 expression in human MM could be an important factor leading to the expansion of the malignant plasma cells. The availability of these clones should help in the dissection of signal transduction pathways in plasma cells, which no longer express the B-cell antigen receptor, and hence their signaling pathways via CD19 could be different from those of mature B cells. Finally, although the growth-inhibitory effect of CD19 is not specifically exerted on myeloma cells, the probability of specific transfer of CD19 gene into myeloma cells could contribute to the development of new strategies for the treatment of human MM.