| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
) and mb-1 (Ig
) Promoters1
Department of Microbiology and Immunology and Molecular Biology Institute, University of California School of Medicine, Los Angeles, CA 90095
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) and mb-1 (Ig
) gene
products are B cell-specific essential components of the B cell
receptor that are coexpressed at all stages of B cell differentiation,
with the exception of plasma cells, which lack mb-1
expression. Transcription of both genes is governed by a similar
cassette of interactive transcription factor-binding elements,
including octamer motifs, in TATA-less promoters. In this study, we
show the B cell-specific B29 gene promoter is
transactivated in B and non-B cells by cotransfection with the B
cell-specific octamer cofactor gene, Bob1 (OCA-B/OBF-1). The expression
of Bob1 is also sufficient to override the silencing effects of the
B29 silencer. This indicates that Bob1 plays a critical
role in B cell-specific B29 promoter expression. In
contrast, coexpression of Bob1 had no effect on mb-1
promoter activity. Bob1 transactivation only occurs with select octamer
sequences that have an adenosine at position 5 (ATGCAAAT). The
B29 promoter conforms to this consensus octamer motif,
while the mb-1 promoter octamer motif does not. Octamer
motif swapping between B29 and mb-1
promoters renders B29 unresponsive to Bob1
transactivation and makes mb-1 competent for Bob1
transactivation, thereby indicating that the B29 octamer
motif is solely responsible for Bob1 interaction. Additionally, the
mb-1 construct containing the B29 octamer
motif is expressed in a plasmacytoma cell line, while the wild-type
mb-1 promoter is not. Bob1 transactivation of
B29 and the lack of this transactivation of
mb-1 account for the differential expression of
B29 and mb-1 in terminally differentiated
plasma cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Bob1 is essential for normal patterns of Ig expression such that Bob1-/- mice are crippled in their Ag-dependent responses, but show normal Ag-independent responses. Hence, Bob1 is dispensable for normal, Ag-independent B cell differentiation and B cell receptor gene expression, but is essential for Ag-dependent maturation of B cells. Specifically, the proliferative response to surface IgM cross-linking is severely impaired, as is the production of secondary Ig isotypes due to the reduced levels of transcription from normally switched Ig H chain loci (9, 10, 11).
The B29 (Ig) gene is strictly B cell specific and
expressed at all stages of B cell differentiation (12).
The mb-1 gene is also B cell specific and is expressed at
all stages of B cell development, except the plasma cell stage
(13). The products of the B29 and
mb-1 genes are essential components of the B cell receptor
and play critical roles in B cell development (reviewed in Refs.
14, 15, 16). Both B29 and mb-1 gene
transcription is controlled by TATA-less promoters containing almost
identical cassettes of interactive transcription factor-binding
elements that collectively impart B cell-specific expression, including
essential octamer motifs (17, 18, 19, 20, 21, 22, 23). Additionally, the
B29 promoter activity is modulated by three upstream
silencer elements; FROG, TOAD, and the A + T-rich octamer-binding motif
that coordinately act to govern B29 gene expression
(24, 25). In this study, we show that the B29
promoter is transactivated in B and non-B cells by concomitant
expression of the octamer cofactor gene, Bob1. We also show that the
mb-1 promoter is not transactivated by Bob1 under identical
conditions. This differential responsiveness was shown to be controlled
by the octamer motifs in the two promoters. Together, these data
suggest a role for Bob1 in determining the differential expression of
B29 and mb-1 in plasma cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All cell lines were propagated in RPMI 1640 supplemented with sodium pyruvate (Life Technologies, Gaithersburg, MD), nonessential amino acids (Life Technologies), 50 µM 2-ME (Sigma-Aldrich, St. Louis, MO), 5 mM glutamine (Sigma-Aldrich), and 5% FCS (Gemini Scientific, Tarzana, CA).
Plasmid construction and mutagenesis
B29 promoter constructs were B29 promoters introduced into pCAT basic (Promega, Madison, WI), as described (17, 24, 25, 26) or B29 promoters from -164 to +32 in relation to the start of transcription introduced into SacI-HindIII of pGL3 basic (Promega). mb-1 promoter constructs were mb-1 promoters introduced into pCAT basic (Promega), as described (26), or mb-1 promoters from -252 to +48 in relation to the start of transcription introduced into HindIII of pGL3 basic (Promega). Bob1 construct (27) was kindly provided by M. Peterlin (San Francisco, CA). Mutagenesis of the B29 and mb-1 octamer motifs was performed with the Quik Change kit (Stratagene, La Jolla, CA), as described (17, 24, 25, 26). Octamer motif swapping between B29 and mb-1 was performed by mutagenesis using the Quik Change kit (Stratagene) and the following oligonucleotides and their complements: 5'-GGGTCTCAATTTGCCATGGCAGGAAG-3'-B29 and 5'-GCCCACACATATGCAAATAAAGGGCC-3'-mb-1.
Transfections, CAT assays, and luciferase assays
M12 B cell line was transfected by the DEAE-dextran method, as described (28), using 5 µg chloramphenicol acetyltransferase (CAT)3 plasmid and 5 µg pRSV-luciferase plasmid, or 10 µg pGL3 plasmid and 5 µg pRL SV40 plasmid. BW5147 T cell and J558L plasmacytoma cell lines were transfected by electroporation, as described (29), using 5 µg CAT plasmid and 5 µg pRSV-luciferase plasmid, or 10 µg pGL3 plasmid and 5 µg pRL SV40 plasmid. NIH 3T3 fibroblast cell line was transfected by the SuperFect (Stratagene) method using 5 µg CAT plasmid, and 5 µg pRSV-luciferase. Bob1 construct (27) was used in transfections as 5, 10, or 20 µg, as described in the figures. Total amounts of DNA transfected were equalized by the addition of pBluescript in samples that do not include Bob1 expression construct DNA. All transfections were harvested 40–48 h posttransfection. For pCAT constructs, extracts were prepared and CAT assays were performed, as described previously (17), with the exception of quantification by PhosphorImager (Molecular Dynamics, Sunnyvale, CA) analysis. Results were normalized to pRSV-luciferase values. For pGL3 constructs, dual luciferase assays were performed as described in the Dual Luciferase Reporter Assay System (Promega). All transfection values are ± SD of at least three transfections using at least two preparations of DNA. Values of p were calculated by the Student two-sided t test.
EMSA
Oct-1 and Oct-2 in vitro translate was prepared using the
TNT quick coupled reticulocyte lysate system (Promega). EMSA was
preformed as described (17) using 2 µl in vitro
translate. EMSA probes were double-stranded oligonucleotides 5' end
labeled with [
-32P]ATP and purified by
Microspin G25 columns (Amersham Pharmacia Biotech, Piscataway, NJ).
EMSA complementary double-stranded oligonucleotide probes were as
follows: 5'-CCTGCCATGCAAATTGAGAC-3' (B29OCT);
5'-CCTGCCATGCAAcgTGAGAC-3' (B29mOCT);
5'-CCTGCCATGgCAAATTGAGAC-3' (B29(mb-1OCT));
5'-CCACACATATGGCAAATAAAGGGCC-3' (mb-1OCT);
5'-CCACACATATGGCAAgcAAAGGGCC-3' (mb-1 mOCT); and
5'-CCACACATATGCAAATAAAGGGCC-3' (mb-1(B29OCT)). Underlines
represent octamer consensus.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B29 promoter activity is dependent on a functional octamer motif, as there is significantly reduced transcriptional activity from the B29 minimal promoter when this site is destroyed by mutagenesis (17). This critical B29 octamer motif matches the consensus for Oct-1/2-Bob1-DNA ternary complex formation because it contains the essential "A" in position 5 of the octamer consensus (6, 7).
Transient transfections of M12 B cells with the wild-type
B29 promoter (-164) and the B29 -164 promoter
with a mutated octamer motif (-164 mOCT) were compared with transient
transfections of the identical constructs along with increasing
concentrations of a Bob1 expression construct. Fig. 1
A shows that the addition of
Bob1 significantly increased (up to 10-fold) the expression of the
wild-type B29 (-164) in B cells (p
< 0.05) but did not affect the expression of the variant
B29 (-164 mOCT) construct that cannot interact with octamer
factors (17). These data suggest that the Bob1
transactivation effect is exerted specifically through the octamer
motif in the B29 -164 promoter.
|
We next asked whether the concordant expression of Bob1 could
transactivate the B29 promoter in non-B cells and break the
B cell specificity of the promoter. Fig. 1
, B and
C, shows that the inclusion of the Bob1 expression construct
significantly (up to 5-fold) increased expression of the wild-type
B29 (-164) promoter over its negligible expression in
transfections of BW5147 T cells and NIH 3T3 fibroblasts. Again, this
transactivation was octamer motif dependent, as the B29
octamer mutant (-164 mOCT) promoter showed no increase in activity
with the addition of Bob1 in T cells (Fig. 1B) or
fibroblasts (Fig. 1C).
Bob1 transactivation overrides the silencing effects of the TOAD, FROG, and A + T-rich motifs in the B29 promoter
Expression from the B29 promoter was previously
shown to be governed by three independent, but cooperative upstream
silencer elements that function in B cells as well as in non-B cells
(24, 25). We used our cotransfection system to determine
whether Bob1 would transactivate B29 promoter constructs
containing the TOAD, FROG, and A + T-rich silencer elements. Fig. 2 shows the effects of Bob1 expression on
the B29 (-354) construct containing the TOAD silencer
element and the B29 (-565) construct containing the TOAD,
FROG, and A + T-rich silencer elements in comparison with these
constructs alone in M12 B cells (dark stippled bars). These results are
also directly compared with the effects of Bob1 on the wild-type
B29 promoter (-164) and on the B29 octamer
mutant (-164 mOCT) (Fig. 2). Cotransfection of Bob1 with both the
B29 (-354) and (-565) silencer constructs resulted in a
significant increase in activity (Fig. 2, compare -354 with -354/Bob;
and compare -565 with -565/Bob). These Bob1-transactivated promoter
activities (-354/Bob and -565/Bob) are not significantly different
from the B29 (-164) promoter construct (Fig. 2).
|
shows that when the B29 (-164)
promoter octamer motif was mutated within the B29 (-565)
promoter construct (-565 mOCT), the addition of Bob1 expression
construct did not transactivate this construct. These data suggest that
the A + T-rich octamer motif alone was not sufficient for Bob1
transactivation of this construct. In contrast, the complementary
construct, wild-type B29 (-164) promoter octamer motif and
mutated A + T-rich octamer motif in the context of the B29
(-565) promoter (-565 mA + T), was transactivated by Bob1 to an
extent equivalent to the wild-type B29 (-565) construct
(Fig. 2). These data show that Bob1 transactivation is not mediated
through the A + T-rich octamer motif. The B29 (-565)
promoter construct with both octamer sites mutated showed no
significant difference from the B29 (-164 mOCT) or the
B29 (-565 mOCT) promoter constructs when Bob1 expression
constructs were cotransfected (data not shown).
The identical constructs were transfected into BW5147 T cells to
determine whether the coexpression of Bob1 was sufficient to override
the B29 silencer element effects in T cells (Fig. 2, light
hatched bars). Fig. 2 shows that Bob1 transactivated the B29
(-354) and B29 (-565) promoter constructs in T cells,
breaking the B cell specificity of the B29 promoter. These
levels of Bob1 transactivation are significantly less than the
transactivation seen with the B29 (-164) promoter (Fig. 2).
Again, mutation of the B29 (-164) octamer motif in the
B29 (-565) promoter (-565 mOCT) did not allow
transactivation of this construct by Bob1, and mutation of the A +
T-rich octamer motif had no effect on the transactivation potential of
the B29 (-565) promoter (-565 mA + T) in T cells.
Bob1 does not transactivate mb-1 promoter activity in either B or non-B cells
The mb-1 minimal promoter is strikingly similar to the
B29 -164 promoter (see Fig. 7), and is also strongly
dependent on its octamer motif for maximal activity (26).
Unlike the B29 octamer motif, the mb-1 octamer
motif does not have the essential sequence required for Bob1
interaction. Specifically, the mb-1 octamer motif lacks an
adenine at position 5 of the octamer consensus (see Fig. 7). We tested
the mb-1 promoter in our Bob1 transactivation system to
determine whether the mb-1 octamer motif would interact with
Bob1 and result in transactivation of the promoter. Fig. 3 shows that Bob1 did not transactivate
the mb-1 promoter in either B cells (M12 B cell line; Fig. 3, dark hatched bars) or T cells (BW5147 T cell line; Fig. 3, light
hatched bars) when increasing amounts of the Bob1 expression construct
were cotransfected with the wild-type mb-1 promoter
(mb-1) and mutant octamer mb-1 promoter
(mb-1 mOCT) constructs. In B cells, the wild-type
mb-1 promoter showed significant activity over pCAT basic,
while the addition of Bob1 did not show a significant increase in
expression of this construct (Fig. 3). The mutant octamer
mb-1 promoter showed no significant activity above pCAT
basic, and the addition of Bob1 did not change the expression level
(Fig. 3). In T cells, neither the wild-type mb-1 promoter
nor the mutant octamer mb-1 promoter showed activity above
pCAT basic (Fig. 3). The addition of Bob1 to either of these constructs
did not result in any significant transactivation of these promoters
(Fig. 3). These data are consistent with the rules for Bob1 interaction
and function, and suggest that in our system, Bob1 functions according
to its determined specificity.
|
|
Based on the inability of Bob1 to transactivate the
B29-mutated octamer site promoter (-164 mOCT) and the
mb-1 promoter (mb-1), we have demonstrated that
the Bob1 transactivation activity functions through the octamer
consensus sequence (Fig. 1). In this study, we directly show that the
mb-1 9-bp octamer consensus sequence (ATGGCAAAT), when put
into the context of the B29 octamer site (-164(mb-1OCT)),
no longer allows Bob1 transactivation of B29 in either B or
T cells (Fig. 4). In the same way, the
B29 8-bp octamer consensus sequence (ATGCAAAT), when put
into the context of the mb-1 octamer site (mb-1(B29OCT)),
confers Bob1 sensitivity and transactivation to mb-1 in both
B and T cells (Fig. 4). These data also show that even though the
mb-1 octamer site is a functional octamer site based on
mutagenesis results (26), the B29 octamer
consensus when swapped with the mb-1 consensus
(mb-1(B29OCT)) confers more overall activity to the mb-1
promoter (Fig. 4A). The opposite is true for mb-1
octamer site placed in the B29 promoter (-164(mb-1OCT))
(Fig. 4A). In the context of the B29 sequences
(-164(mb-1OCT)), the mb-1 consensus does not appear to
function as an octamer motif, as the activity level is similar to that
seen for the B29-mutated octamer motif promoter (-164 mOCT)
(Fig. 4A).
|
A), possibly reflecting
intrinsic differences in the B29 and mb-1 octamer
sites. In contrast, the mb-1 with the B29 octamer
consensus motif (mb-1(B29OCT)) showed greater binding to IVT Oct-1 than
the wild-type mb-1 octamer motif (Fig. 5B). This
result may explain why the mb-1(B29OCT) construct has higher activity
in transient transfections than the wild-type mb-1 promoter
construct (Fig. 4).
|
Previous studies have shown that the mb-1 promoter has
no activity in transient transfection of plasma cell lines (19, 23), while B29 has high activity (18). We tested
our octamer site swap constructs (-164(mb-1OCT) and mb-1(B29OCT)) in
J558L plasmacytoma cell line to determine whether the change in the
octamer sites alone would change the expression patterns of the
B29 and mb-1 promoters. Fig. 6 shows that the Bob1-responsive
B29 octamer site conferred activity onto the
mb-1 promoter (mb-1(B29OCT)) in plasmacytoma cells.
Additionally, the Bob1-unresponsive mb-1 octamer site shut
down expression of the B29 promoter (-164(mb-1OCT)) in
plasmacytoma cells. These data suggest that the octamer site alone
controls differential expression of the B29 and
mb-1 promoters in terminally differentiated plasma cells.
Interestingly, ectopic Bob1 expression had no significant effect on the
activity of any of the B29 or mb-1 promoter
constructs in plasma cells (Fig. 6). The increased level of endogenous
Bob1 expression in J558L (4-fold over M12 Bob1 expression) could
preclude any further effect by ectopically expressed Bob1 (data not
shown) (30).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Early reports of Bob1 specificity
suggested a role for the TATA box in controlling which promoters were
targets for Bob1 (1). Our data showing transactivation of
the TATA-less B29 promoter by Bob1 suggest that other
criteria must be responsible for selective Bob1 transactivation in the
context of TATA-less promoters. Our study describes a new mechanism for the differential expression patterns of B29 and mb-1 in B cell development. The B29 gene is expressed throughout B cell development (31), while the mb-1 gene is only expressed through B cell development up to the plasma cell stage (19, 20). An early hypothesis purported that the transcription factor, early B cell factor (EBF), controlled the differential expression of B29 and mb-1 (19, 20). This inference was based on the identical expression profiles of mb-1 and EBF (19, 20, 32, 33), and the apparent lack of interaction of EBF and the B29 promoter (17). Recently, it was shown that EBF did interact with the B29 promoter (18), and that B29 is not expressed in EBF-/- knockout mice (34). In light of these discoveries, EBF cannot account for the differential expression of B29 and mb-1. B cell lineage-specific activator protein (BSAP; Pax5) was shown to interact with the mb-1 promoter (21, 22). Like EBF, BSAP is not expressed in terminally differentiated plasma cells (35). Unlike EBF, BSAP does not appear to affect B29 promoter activity because normal levels of B29 expression were seen in BSAP-/- mice (22), and the B29 promoter sequence does not contain a BSAP binding site (17). The lack of BSAP in plasma cells may contribute to the loss of mb-1 expression in addition to the unresponsiveness of the mb-1 promoter to Bob1 transactivation.
The concordant expression of Bob1 and B29 naturally points to a role for Bob1 in B29 gene expression. Our evidence for the differential interaction of Bob1 with the B29 and mb-1 promoters poses a plausible explanation for the continued expression of B29 and for the extinction of mb-1 in the last stage of B cell development. Our study showed that Bob1 transactivated the B29 promoter, while Bob1 was unable to affect expression from the mb-1 promoter. These data support our proposal that while a lack of BSAP may contribute to the extinction of mb-1, the presence of Bob1 is responsible for the continued expression of B29 in plasma cells. Data showing that Bob1 promoter expression and Bob1 protein levels are highest in plasma cells (30) support our proposal that Bob1 transactivation supports ongoing B29, but not mb-1 gene expression in plasma cells.
Our data specifically showed that Bob1 was a potent transactivator of B29 in the lymphoid and nonlymphoid cell types tested, and that this effect was mediated specifically through the octamer motif. Equal amounts of Bob1 expression construct showed no effect on the activity of the mb-1 promoter, even though the mb-1 octamer site is functional and has been shown to be required for maximal mb-1 promoter activity (26). We directly showed that the sequences responsible for Bob1 transactivation lie within the octamer consensus sequence. The B29 and mb-1 octamer consensus site swap constructs showed that Bob1 transactivation is mediated by the B29 consensus octamer site (ATGCAAAT), regardless of whether the sequences outside this site were derived from B29 or mb-1. Additionally, the octamer swap constructs demonstrated that the B29 consensus octamer site controlled the plasma cell-specific activity of the promoters; the mb-1 promoter alone had little activity in plasma cells, while the mb-1 with the B29 octamer consensus site (mb-1(B29OCT)) had significantly greater activity than the wild-type mb-1 promoter. In fact, the mb-1(B29OCT) promoter had activity that was not significantly different from the wild-type B29 promoter in plasma cells. Furthermore, the B29 with the Bob1-nonresponsive mb-1 octamer consensus site (B29(mb-1OCT)) had little activity in plasma cells. These data support our proposal that Bob1 expression in plasma cells controls the differential expression of the B29 and mb-1 promoters.
Our data showing that Bob1 overrides the effects of the B29 silencer elements in B cells and non-B cells suggest that Bob1 also plays a critical role in regulating B cell expression of the B29 gene. The B29 silencer elements have been shown to govern B29 expression, but not control cell type specificity because they are equally active in both B and non-B cells (24). The expression of Bob1 may act to negate the effects of the B29 silencers in B cells in which Bob1 and B29 are expressed. In this mechanism, Bob1 would act as an antisilencer countering the 5' B29 silencers, thereby controlling cell type specificity of the B29 gene and restricting B29 expression to B cells only. This combination of cis-acting silencers and a trans-acting transcriptional coactivator functioning as an antisilencer represents a novel mechanism for controlling B cell gene specificity.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Randolph Wall, University of California School of Medicine, 539 Boyer Hall, 611 Charles E. Young Drive East, Los Angeles, CA 90095. E-mail address: rwall{at}mbi.ucla.edu
3 Abbreviations used in this paper: CAT, chloramphenicol acetyltransferase; BSAP, B cell lineage-specific activator protein; EBF, early B cell factor; IVT, in vitro translated.
Received for publication August 30, 2001. Accepted for publication January 31, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Science 272:411.[Abstract]
-chain) gene is a genetic target for early B-cell factor. Mol. Cell. Biol. 19:392.) promoter are neither promoter- nor cell-type-specific. Proc. Natl. Acad. Sci. USA 94:12314.) gene expression. J. Immunol. 164:2550.) promoter. Mol. Immunol. 37:321.[Medline]
-lineage cells. Proc. Natl. Acad. Sci. USA 85:6890.This article has been cited by other articles:
|
|
 |
|
  R. R. Shen, D. O. Ferguson, M. Renard, K. K. Hoyer, U. Kim, X. Hao, F. W. Alt, R. G. Roeder, H. C. Morse III, and M. A. Teitell Dysregulated TCL1 requires the germinal center and genome instability for mature B-cell transformation Blood, September 15, 2006; 108(6): 1991 - 1998. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Bartholdy, C. Du Roure, A. Bordon, D. Emslie, L. M. Corcoran, and P. Matthias The Ets factor Spi-B is a direct critical target of the coactivator OBF-1 PNAS, August 1, 2006; 103(31): 11665 - 11670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ushmorov, F. Leithauser, O. Sakk, A. Weinhausel, S. W. Popov, P. Moller, and T. Wirth Epigenetic processes play a major role in B-cell-specific gene silencing in classical Hodgkin lymphoma Blood, March 15, 2006; 107(6): 2493 - 2500. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Brunner, H. Laumen, P. J. Nielsen, N. Kraut, and T. Wirth Expression of the Aldehyde Dehydrogenase 2-like Gene Is Controlled by BOB.1/OBF.1 in B Lymphocytes J. Biol. Chem., November 14, 2003; 278(46): 45231 - 45239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Gordon, C. M. Kanegai, J. R. Doerr, and R. Wall Somatic hypermutation of the B cell receptor genes B29 (Igbeta , CD79b) and mb1 (Igalpha , CD79a) PNAS, April 1, 2003; 100(7): 4126 - 4131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Sigvardsson, D. R. Clark, D. Fitzsimmons, M. Doyle, P. Akerblad, T. Breslin, S. Bilke, R. Li, C. Yeamans, G. Zhang, et al. Early B-Cell Factor, E2A, and Pax-5 Cooperate To Activate the Early B Cell-Specific mb-1 Promoter Mol. Cell. Biol., December 15, 2002; 22(24): 8539 - 8551. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
