|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, 15 June 2006, Vol. 107, No. 12, pp. 4907-4916. Prepublished online as a Blood First Edition Paper on February 28, 2006; DOI 10.1182/blood-2005-08-3531.
NEOPLASIA Proteasome inhibitors induce a terminal unfolded protein response in multiple myeloma cellsFrom the Department of Microbiology and Immunology, University of Miami Miller School of Medicine, Miami, FL; and the Sylvester Comprehensive Cancer Center, University of Miami Miller School of Medicine, Miami, FL.
Multiple myeloma (MM) is an incurable plasma cell malignancy. The 26S proteasome inhibitor, bortezomib, selectively induces apoptosis in MM cells; however, the nature of its selectivity remains unknown. Here we demonstrate that 5 different MM cell lines display similar patterns of sensitivity to 3 proteasome inhibitors (PIs) but respond differently to specific NF-  B inhibition. We further show that PIs initiate the unfolded protein response (UPR), a signaling pathway activated by the accumulation of misfolded proteins within the endoplasmic reticulum (ER). Consistent with reports that prosurvival/physiologic UPR components are required for B-cell differentiation into antibody-secreting cells, we found that MM cells inherently expressed the ER chaperones GRP78/Bip and GRP94/gp96. However, bortezomib rapidly induced components of the proapoptotic/terminal UPR, including PERK, the ER stress–specific eIF-2 kinase; ATF4, an ER stress–induced transcription factor; and its proapoptotic target, CHOP/GADD153. Consistent with our hypothesis that PIs induce the accumulation of misfolded ER-processed proteins, we found that the amount of immunoglobulin subunits retained within MM cells correlated with their sensitivity to PIs. These findings suggest that MM cells have a lower threshold for PI-induced UPR induction and ER stress–induced apoptosis because they constitutively express ER stress survival factors to function as secretory cells.
Multiple myeloma (MM), the second most commonly diagnosed hematologic malignancy in the United States, is an essentially incurable malignancy of terminally differentiated B cells or plasma cells.1,2 Bortezomib (Velcade, PS-341) is a novel therapeutic agent that has been shown to selectively induce apoptosis in malignant cells.3,4 Bortezomib is particularly toxic to MM cells,5,6 but it has a favorable toxicity profile and was approved by the US Food and Drug Administration in 2003 for the treatment of relapsed refractory disease.7
Bortezomib is a potent and selective inhibitor of the 26S proteasome,8,9 a multisubunit protein complex present in the nucleus and the cytoplasm of all eukaryotic cells10 that is responsible for the degradation of ubiquitinated proteins.11 In addition to damaged or obsolete proteins, the proteasome is responsible for the degradation of proteins involved in cell-cycle regulation, oncogenesis, and apoptosis.12-20 Previous reports have demonstrated that proteasome inhibition by bortezomib abrogates degradation of I One of the defining features of plasma cells is an expansive and highly developed rough endoplasmic reticulum (ER) that is specialized for the production and secretion of thousands of antibody molecules per second.27 In fact the detection of large amounts of monoclonal immunoglobulin or light chain in the serum or urine is one of the diagnostic features of MM.28 Conditions that disrupt protein folding in the ER, such as a chemical insult or nutrient deprivation, activate a stress signaling pathway known as the unfolded protein response (UPR).29,30 UPR induction results in both an initial decrease in general protein synthesis, to reduce the influx of nascent proteins into the ER, and increased transcription of ER resident chaperones, folding enzymes, and components of the protein degradative machinery to prevent the aggregation of the accumulating misfolded proteins. These misfolded proteins are recognized by ER quality control systems and retained in the ER, preventing them from proceeding further through the protein maturation process.31-33 If these proteins cannot be properly refolded, they are targeted for ER-associated protein degradation (ERAD), which involves the retrograde translocation or dislocation of the misfolded proteins out of the ER and subsequent degradation by cytosolic 26S proteasomes.34,35 The UPR enables the cell to survive reversible environmental stresses. However, if the stress is severe or prolonged, UPR activation eventually leads to cell-cycle arrest36,37 and the induction of apoptosis.38-41 The retrograde translocation of misfolded proteins from the ER has been shown to be dependent on functioning cytosolic proteasomes.42-46 Thus, treatment of cells with proteasome inhibitors (PIs) results in the accumulation of misfolded proteins within the ER. We therefore hypothesized that treatment of MM cells with PIs initiates the UPR by inhibiting the retrograde translocation of misfolded proteins from the ER and that MM cells are highly sensitive to these agents because they produce large amounts of protein, namely immunoglobulin, that must be processed within the ER. Interestingly, we found that MM cells constitutively express high levels of UPR survival components, but that PI treatment leads to the rapid induction of proapoptotic UPR genes. We further demonstrate that the amount of immunoglobulin subunits retained in PI-treated MM cells correlates with their level of sensitivity to bortezomib. These data suggest that the secretory function of MM cells makes them more sensitive than other cell types to PI-induced UPR activation and ER stress-induced apoptosis.
Multiple myeloma–derived cell lines The 8226/S and U266 human MM cell lines were purchased from the American Type Culture Collection (Manassas, VA). The MM.1S cell line was obtained from Dr Steven Rosen (Northwestern University, Chicago, IL), and the KMS-11 and KMS-18 cell lines were provided by Dr P. Leif Bergsagel (Mayo Clinic, Scottsdale AZ). All cell lines were cultured in RMPI 1640 supplemented with 10% fetal bovine serum, 1% L-glutamine, and 1% penicillin/streptomycin (Mediatech, Herndon, VA). Reagents Bortezomib (PS-341, Velcade) was kindly provided by Millennium Pharmaceuticals (Cambridge, MA). Tunicamycin and melphalan were purchased from Sigma-Aldrich (St Louis, MO). Epoxomycin, PSI, brefeldin A, BAY 11-7082, and staurosporine were obtained from EMD Biosciences (La Jolla, CA). Detection of cell death For all cell death assays 2.5 x 105 cells/mL were treated with the indicated concentration of drug for up to 24 hours. Cells were washed twice in phosphate-buffered saline and then stained with Annexin V–FITC (Biovision, Mountainview, CA) and propidium iodide (Sigma-Aldrich) according to the manufacturer's instructions. Samples were acquired on a fluorescence-activated cell scanner (FACScan) flow cytometer (Becton Dickinson, Franklin Lakes, NJ) to assess viability. Electrophoretic mobility shift assay
Electrophoretic mobility shift assays for NF- Immunoblotting
Whole-cell lysates (5.0 x 106 cells) were prepared in modified RIPA buffer containing protease inhibitors (170 µg/mL PMSF, 2 µg/mL aprotinin, and 1 µg/mL leupeptin; Sigma-Aldrich). Fifty micrograms of protein were resolved on 8% to 15% SDS–polyacrylamide gels and transferred to nitrocellulose membranes (Schleicher and Schuell, Keene, NH). The following primary antibodies were used: anti-GRP78/BiP (BD Biosciences, Pharmingen, San Diego, CA), anti-GRP94/gp96 (Stressgen Bioreagents, Victoria, BC), anti–total eIF-2 Enzyme-linked immunosorbent assay
The Human Lambda (bound and free) ELISA (enzyme-linked immunosorbent assay) Quantitation Kit (Bethyl Laboratories, Montgomery, TX) was used to detect secreted and intracellular
Different multiple myeloma cell lines have different levels of sensitivity to proteasome inhibitors
Previous studies have shown that bortezomib induces a caspase-dependent apoptotic cell death in different MM cell lines.5 We further expanded on these studies by examining apoptosis induction in MM cell lines treated with 3 different proteasome inhibitors. As nonsecretory MM occurs in less than 1% of patients, each of the MM cell lines examined secreted immunoglobulin components. The 8226/S, MM.1S, and KMS-18 myeloma lines are
Proteasome inhibition does not produce the same response as NF-
Constitutive NF-
Similarly, the response of the 5 MM cell lines to the irreversible inhibitor of I B phosphorylation, BAY 11-7082, is not the same as their response to PIs. After 12 hours of treatment, NF-B DNA binding is inhibited in all 5 MM cell lines treated with either 5 µM BAY 11-7082 or 10 to 100 nM bortezomib (Figure 2A). However, after 24 hours of treatment, significant reductions in cell viability are seen in bortezomib-treated cells (Figure 1), but at 5 µMofBAY 11-7082 only the MM.1S and KMS-11 cell lines are sensitive to the drug (Figure 2B). There is a slight decrease in the viability of the 8226/S, U266, and KMS-18 cell lines at this concentration, and these 3 cell lines progressively undergo apoptosis as the concentration of BAY 11-7082 is increased up to 10 µM. However, in contrast to PI treatment, their responses are indistinguishable. Taken together, these studies indicate that, although the inhibition of NF-B has effects on MM cell viability, it cannot completely explain the selectivity of PIs for MM cells. Proteasome inhibitors marginally increase the expression of physiologic UPR components in MM cell lines MM cells are malignant plasma cells that, similar to their normal counterparts, produce and secrete large amounts of immunoglobulin.1,2,27,48 The ER is the site where integral membrane proteins and secretory proteins are folded into their proper tertiary structures, and multimeric proteins, such as immunoglobulins, are assembled.31-33 The inhibition of cytosolic proteasomes has been shown to cause the accumulation of misfolded proteins within the ER lumen.42-46 We therefore hypothesized that proteasome inhibition may lead to UPR induction in secretory MM cell lines, presumably by inhibiting the retrograde translocation of misfolded proteins from the ER.
One of the ways the UPR enables any cell to respond to ER stress is by up-regulating the expression of ER molecular chaperones and folding enzymes51,52 to prevent the aggregation of misfolded proteins and to help them refold. ER chaperone induction was examined in each of the MM cell lines treated for 12 and 24 hours with the 3 different proteasome inhibitors. Because these cell lines have different levels of sensitivity to proteasome inhibition, the highest concentrations of the PIs were used to treat the 8226/S and U266 MM cell lines, intermediate concentrations of PIs were used to treat the KMS-11 and KMS-18 cell lines, and the lowest concentrations of PIs were used to treat the MM.1S cell line. Additionally, we compared the induction of the ER molecular chaperones GRP78/BiP and GRP94/gp96 in PI-treated cells with cells treated with known ER stress agents or other apoptosis-inducing agents. Tunicamycin, an inhibitor of N-linked glycosylation, and brefeldin A, an inhibitor of ER to Golgi transport, are 2 chemical agents known to induce ER stress and the UPR.29,30 The alkylating agent melphalan53,54 and the nonspecific kinase inhibitor staurosporine55 were used as negative controls. Interestingly, we found that all 5 MM cell lines constitutively expressed high levels of both ER molecular chaperones, and that the levels of these proteins remained unchanged or were marginally increased in response to treatment with PIs or known ER stress agents, even as these cells began to undergo apoptosis (Figure 3). Thus, GRP94 was induced 1.0- to 3.8-fold and GRP78 was induced 0.9- to 1.8-fold in MM cells treated with classical ER stress agents compared with a 0.9- to 3.7-fold GRP94 induction and a 0.8- to 2.3-fold GRP78 induction in PI-treated cells. Cytotoxic agents that do not induce ER stress also had little effect on ER chaperone expression, causing only a 0.7- to 2.1-fold induction in GRP94 and a 0.8- to 1.2-fold induction in GRP78.
Our finding that MM cells constitutively express ER chaperones is consistent with previous reports that certain components of the UPR are induced during plasma cell development and are required to be constitutively expressed for these cells to function properly.56-58 It has been demonstrated that the expression of GRP78 and GRP94 is induced and maintained in mature B cells as they differentiate into antibody secreting plasma cells, whereas UPR components associated with decreased protein synthesis and apoptosis were not induced under these conditions.56-59 The specific induction of UPR genes that enable cells to differentiate into professional secretory cells capable of tolerating the constitutive production of high amounts of ER-processed proteins has been defined as a "physiologic" UPR. This UPR is distinct from the "ER stress" or "terminal" UPR, which is induced by nutrient deprivation or chemical agents that cause severe or prolonged ER stress.56,57 Taken together, these data suggest that prosurvival/physiologic UPR components are already highly expressed in MM cells and that their expression cannot be significantly altered by agents that cause additional ER stress within the cell.
Upstream terminal UPR component and eIF-2
The UPR also enables cells to survive stressful conditions through a transient decrease in de novo protein synthesis to reduce the client protein load in the ER. This is accomplished through phosphorylation of eukaryotic translation initiation factor-2
In addition to ER stress, the phosphorylation of eIF-2
Proteasome inhibitors induce the expression of transcription factors associated with ER stress–induced apoptosis
Although the phosphorylation of eIF-2 Although ATF4 is responsible for up-regulating the transcription of various prosurvival stress response genes involved in amino acid transport and the oxidative stress response, ATF4 also increases the transcription of the proapoptotic transcription factor GADD153/CHOP in response to ER stress and UPR induction.70 GADD153 induction has been shown to be specifically involved in ER stress–induced apoptosis,38,71 and mouse embryonic fibroblasts (MEFs) from GADD153-deficient mice have been shown to be resistant to PI- and ER stress–induced apoptosis.38,72 Consistent with these reports, GADD153 was rapidly induced after as little as 4 hours in MM cell lines treated with PIs or known ER stress agents (Figure 6A). Moreover, GADD153 induction is specific to ER stress agent and PI treatment because it is not induced after up to 24 hours of treatment with other cytotoxic agents (Figure 6B). Taken together, these results demonstrate that PIs cause the rapid and sequential induction of UPR components that are associated with severe or prolonged ER stress. These data further suggest that PIs, similar to classical ER stress agents, specifically induce proteins associated with a terminal UPR and ER stress–-induced apoptosis in MM cell lines.
Accumulation of intracellular  light chain correlates with sensitivity to proteasome inhibitors
Although PIs can induce the terminal UPR in each of the cell lines examined, the amount of the drug required for terminal UPR induction varies between different MM cell lines. Thus, a 10-fold higher concentration of bortezomib (100 nM) is required to induce apoptosis and the terminal UPR in the cell line that is most resistant to PIs (8226/S) compared with the cell line that is least resistant to PIs (MM.1S). Both of these cell lines are As we have hypothesized that MM cells are more sensitive to PIs than other malignant cells because they produce large amounts of ER-processed proteins, one explanation for the differences in sensitivity observed between MM cell lines is that cells that are more sensitive to PIs produce more immunoglobulin. However, when immunoglobulin secretion was quantitated by ELISA, we found that MM.1S cells secrete half as much light chain as 8226/S cells (Table 1). This assay measures the amount of immunoglobulin produced by an equal number of cells over a defined period of time without accounting for cell size, whereas visual inspection and forward scatter analysis have shown that 8226/S cells are much larger than MM.1S cells (data not shown).
In addition to cell size, ELISAs performed on cell supernatants also do not address the efficiency of light chain folding within MM cells. This is important because inhibition of the 26S proteasome leads to the accumulation of misfolded proteins in the ER lumen, whereas properly folded proteins can proceed along the secretory pathway. This is in contrast to the inhibition of ER-to-Golgi transport by brefeldin A, which inhibits all protein secretion. To distinguish between retained and presumably misfolded light chains and the total amount of light chain that is processed in the ER, whole-cell lysates were isolated from untreated, PI-treated, or brefeldin A–treated MM cells, and the amount of light chain present within the lysates was quantitated by ELISA. Additionally, the ratio of treated cells to untreated cells was used to normalize the data to make comparisons between cell lines. When compared with untreated controls, 10-fold less bortezomib caused the retention of 3- to 4-fold more light chain in MM.1S cells compared with 8226/S cells (Table 2), suggesting that higher amounts of misfolded light chains, which require the proteasome for degradation, are produced by the MM.1S cell line. The continued secretion of properly folded proteins combined with the activation of the PERK pathway (which decreases protein synthesis) is probably why the ratio is less than 1 in PI-treated cells. This is in contrast to brefeldin A–treated cells, in which the ratio of treated to untreated cells is greater than 1 because this drug inhibits protein secretion. Consistent with the secretion ELISA analysis, more protein accumulates within the brefeldin A–treated 8226/S cells compared with treated MM.1S cells. Taken together these data suggest that, although all MM cells are extremely sensitive to proteasome inhibition, the efficiency of protein folding within the ER can affect the level of sensitivity of individual MM cell lines to PIs.
NF- B activity has been shown to enhance MM cell survival and resistance to chemotherapeutic agents,26 and functional proteasomes are required NF-B activation.5,21-25 However, we found that despite that 5 human MM cell lines responded similarly to low concentrations of 3 different proteasome inhibitors (Figure 1), the effect of the irreversible inhibitor of IB phosphorylation, BAY 11-7082, on these MM cell lines differed from that of proteasome inhibition (Figure 2). These findings are similar to reports that bortezomib and a specific IB kinase inhibitor, PS-1145, do not have the same effects on cell proliferation in MM cell lines22 and suggest that the inhibition of NF-B cannot completely explain the selective anti-MM activity of proteasome inhibitors. Because MM cells, similar to their normal counterparts, produce significant amounts of ER-processed proteins,27 we reasoned that these cells may be sensitive to perturbations in protein degradation, which would result in the activation of the UPR. One of the hallmarks of UPR induction is the increased transcription and translation of ER molecular chaperones.29,30,51,52 These genes are induced by the UPR transcription factors XBP1 and ATF6.52,73 Although XBP1 splicing and its resulting activation have been shown to be inhibited in PI-treated MM cells,74 our finding that the high constitutive expression of 2 XBP1 target genes products, GRP78 and GRP94, is not reduced by PI treatment (Figure 3) and the observation by Lee et al74 that the XBP1-dependent UPR target gene ERdj4 was normally induced by PIs suggest that the UPR remains functional in PI-treated MM cells. Because both XBP1 and ATF6 can bind to ER stress response elements in the promoters of UPR target genes,52,73 it is possible that ATF6 may compensate for decreased XBP1 activity in PI-treated MM cells.75 Consistent with this, it has been shown that the induction of GRP78 and GRP94 is only slightly impaired in XBP1–/– B cells59 and that the expression of GRP94 requires either, but not both, ATF6 or XBP1.75 The high constitutive expression of the ER resident chaperones GRP78 and GRP94 in MM cell lines is consistent with reports that physiologic UPR gene expression is required for professional secretory cell function.56,57,59,76 Elevated levels of ER chaperones are characteristic of plasma cells,77,78 and their expression is essential for proper antibody assembly and secretion. GRP78 has been shown to stably bind to immunoglobulin heavy chains that have not yet associated with immunoglobulin light chains and to assist in immunoglobulin assembly.79-81 Furthermore, both GRP78 and GRP94 are important for immunoglobulin light chain folding and targeting unassembled subunits for degradation.82-86 The fact that the expression of GRP78 and GRP94 is only slightly increased in MM cells treated with PIs and classical ER stress agents suggests they already express near-maximal levels of cytoprotective UPR proteins to function as secretory cells. Thus, these cells may have a lower threshold (compared with nonsecretory cells) for induction of the terminal UPR following any additional stress to the ER.
Although PI treatment of MM cells does not alter the expression levels of ER chaperones, it does lead to the sequential induction of terminal UPR components. Preliminary data suggest that both bortezomib and tunicamycin rapidly activated PERK, the ER stress-–specific eIF-2 The fact that MM cell lines rapidly induce terminal UPR components without further increasing the expression of prosurvival UPR components when treated with PIs suggests that the secretory function of these cells makes them extremely sensitive to ER stress induced by this class of drug. This may explain the finding that murine myeloma cells can be sensitized to lower concentrations of PIs if they are also treated with tunicamycin,74 because the inhibition of N-linked glycosylation by tunicamycin would interfere with immunoglobulin/protein folding within the ER, creating an even larger amount of ERAD substrates. The selectivity of PIs for secretory cells is also supported by observations that PIs activate PERK much later in MEFs.72,88 Jiang and Wek72 were unable to detect PERK activation in MEFs after up to 6 hours of treatment with the proteasome inhibitor MG132, and Fribley et al88 only identified detectable levels of PERK activation in MEFs after 16 hours of bortezomib treatment.
Although the high levels of antibody synthesis in MM cells may indicate that myeloma cells, in general, have a greater need for the retrograde translocation of misfolded ERAD substrates, the differences in sensitivity observed between MM cell lines also appears to involve the efficiency of immunoglobulin folding. As each malignant myeloma clone arises from a mature B cell with a unique rearrangement of its immunoglobulin genes, one would expect that certain immunoglobulin protein subunits may fold or assemble more efficiently than others. Mutations that cause the inefficient or improper folding of ER-processed proteins have been associated with the pathology of diseases such as cystic fibrosis, and cytosolic proteasomes have been shown to be required for the removal and the degradation of these misfolded proteins from the ER.44,45 Consistent with the idea that immunoglobulin folding efficiency can affect the sensitivity of MM cells to PIs, higher amounts of
These studies further suggest that malignant cells can be sensitized to PIs by combining them with agents that affect protein folding within the ER or the cytoplasm. The accumulation of misfolded proteins in the ER lumen of PI-treated cells is thought to occur as a consequence of feedback inhibition, where the accumulation of ubiquitinated proteins that cannot be degraded in the cytosol prevents the retrograde translocation of additional unfolded proteins from the ER.89,90 The presence of misfolded proteins in the cytosol may partially explain why proteasome inhibitors also induce the expression of heat shock–factor proteins including HSP27, HSP70, and HSP90.6,91-95 HSP90 regulates the stability and function of many protein kinases, including IRE1 Taken together, these data suggest that MM cells are inherently sensitive to PIs because of their large volume of immunoglobulin production, which requires the constitutive expression of physiologic UPR genes. This appears to lower their threshold for the induction of a terminal UPR in response to PI-induced ER stress. In addition to the amount, PI sensitivity also appears to involve the efficiency of immunoglobulin folding within MM cells. These parameters may lead to the more specific identification of MM tumors that will respond to bortezomib. Furthermore, more resistant myeloma clones as well as other nonsecretory malignancies may be sensitized to bortezomib by combining it with agents that interfere with the UPR or agents that destabilize cytosolic proteins.
We thank Bill Dalton, Leif Bergsagel, and Steve Rosen for kindly providing us with cell lines. We also thank Robert Levy, David McConkey, Linda Hendershot, and Alan Diehl for helpful advice.
Submitted September 1, 2005; accepted February 13, 2006.
Prepublished online as Blood First Edition Paper, February 28, 2006; DOI 10.1182/blood-2005-08-3531.
Supported in part by National Institutes of Health (grants R0-1 GM68513 and R0-1 CA97243; L.H.B. and K.P.L.), a Senior Research Award from the Multiple Myeloma Research Foundation (L.H.B.), and a Predoctoral Fellowship from the Howard Hughes Medical Institute (E.A.O.).
The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked "advertisement" in accordance with 18 U.S.C. section 1734.
Reprints: Lawrence H. Boise, Department of Microbiology and Immunology, The University of Miami Miller School of Medicine, PO Box 016960 (R-138), Miami, FL 33101; e-mail: lboise{at}med.miami.edu.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
32P-labeled double-stranded NF-
