|
|
||||||||
B Regulates Ig
Light Chain Gene Rearrangement1
Department of Microbiology and Immunology, Vanderbilt University School of Medicine, Nashville, TN 37232
| Abstract |
|---|
|
|
|---|
locus is regulated by the inducible transcription factor NF-
B.
In contrast to the Ig
locus, known transcriptional control elements
in the Ig
locus lack functional NF-
B binding sites. Consistent
with this observation, the expression of assembled Ig
genes in
mature B cells has been shown to be NF-
B independent. Nonetheless,
we now show that specific repression of NF-
B inhibits germline
transcription and recombination of Ig
gene segments in precursor B
cells. Molecular analyses indicate that the block in NF-
B impairs
Ig
rearrangement at the level of recombinase accessibility. In
contrast, the activities of known Ig
promoter and enhancer elements
are unaffected in the same cellular background. These findings expand
the range of NF-
B action in precursor B cells beyond Ig
to
include the control of recombinational accessibility at both L chain
loci. Moreover, our results strongly suggest the existence of a novel
Ig
regulatory element that is either directly or indirectly
activated by NF-
B during the early stages of B cell
development. | Introduction |
|---|
|
|
|---|
Despite the action of a common recombinase on highly conserved RSSs, the assembly of individual Ag receptor genes occurs in a stage- and tissue-specific manner. Prior studies have demonstrated that substrate specificity is imposed on the recombination process by modulating the accessibility of gene segment clusters within Ig and TCR loci (4, 5). In the vast majority of cases, recombinational accessibility of a given gene segment correlates temporally with its transcriptional status in vivo (4, 5, 6). Consistent with a causal relationship between these two processes, transcriptional promoter and enhancer activities are required for the efficient rearrangement of chromosomal gene segments in cis (5, 6, 7). In precursor lymphocytes, fluctuations in the activity of transcription factors that target promoter/enhancer elements likely play a major role in the developmental control of Ig and TCR locus accessibility (5, 8). The precise molecular mechanisms by which cis-acting elements regulate accessibility remain elusive. However, recent studies indicate that directed alterations in the chromatin associated with gene segment clusters act to relieve nucleosome-mediated repression of RSS cleavage by the RAG proteins (9, 10).
In the B lymphocyte lineage, pro-B cells must first rearrange a
functional IgH gene before recombination of the Ig
locus at the
pre-B cell stage (11). In pre-B cell models, the onset of
J
germline transcription and V
J
rearrangement requires the
induction of transcription factor NF-
B (12), which
binds to a consensus site in the Ig
intronic enhancer (iE
).
Individual pre-B clones that fail to generate an in-frame V
J
coding join subsequently initiate rearrangement of the second L chain
locus, Ig
(11). Relative to Ig
, little is known
about the transcription factors and DNA elements that regulate
recombination of Ig
gene segments. Two nearly identical Ig
enhancers (E
3-1 and E
2-4), which are active in mature B and
plasma cells, require binding by the PU.1/IFN regulatory
factor-4 transcription factor complex (13).
Importantly, E
3-1 and E
2-4 function in an NF-
B-independent
manner in these late-stage cells (14).
During the course of our studies on Ig
regulation (12),
we discovered that cells defective for NF-
B signaling also were
affected at the Ig
locus. Consistent with this initial observation,
we now demonstrate that the NF-
B signaling pathway is required to
induce germline transcription and recombination of all murine V
and
J
gene segments. In contrast, the activities of known Ig
regulatory elements are unaffected in NF-
B-arrested cells. These
findings highlight a novel mechanism by which NF-
B acts as a global
regulator of Ig L chain (IgL) gene assembly in precursor B cells.
| Materials and Methods |
|---|
|
|
|---|
The 103/BCL-2/4 pre-B cell line is conditionally
transformed with a temperature-sensitive (ts) mutant of the
v-abl oncogene (15, 16). Stable transfectants
of ts-abl expressing a dominant inhibitor of NF-
B, termed I
B
N
(
N.1,
N.7), and a control line expressing the histidinol
resistance gene (WT.24) have been described previously
(12). Cells were routinely propagated at 34°C in RPMI
1640 supplemented with 10% FCS, 50 µM 2-ME, and 0.01%
penicillin-streptomycin. Inactivation of v-abl was achieved
by incubation at the nonpermissive temperature (38.5°C) for either 24
or 48 h.
DNA PCR
Relative levels of V
J
and V
J
coding joins and total
DNA content (C
) were assessed by semiquantitative PCR in 25-µl
reactions containing 200 ng genomic DNA, 10 mM Tris (pH 8.3), 1 µg/ml
BSA, 200 µM dNTPs, Taq polymerase (1 U), and 25 ng of each
primer. Amplifications were performed as follows: 94°C (1 min),
60°C (C
) or 52°C (V
J
, V
J
) for 1 min, and 72°C (1.5
min) for either 24 (C
) or 28 cycles (V
J
, V
J
). All PCR
products were separated on 2% agarose gels and transferred to
ZetaProbe membranes (Bio-Rad, Richmond, CA) for hybridization with the
appropriate radiolabeled probes.
For V
and J
SEs, 2.5 x 106 cells were
cultured at the permissive or nonpermissive temperature for 24
or 48 h. Linker-ligated DNA plugs were prepared as described
previously (17). SE assays were performed on 3 µl of
molten plugs (95°C, 10 min) in 25-µl reaction mixtures as described
for the coding joins assays, except that reactions were incubated at
72°C (5 min) before the addition of Taq polymerase (1 U).
The 12-cycle amplifications consisted of 1 min at 92°C, 1 min at
52°C, and 1.5 min at 72°C. An aliquot of the primary PCR (2 µl)
then was amplified for 28 cycles under the same conditions, but with a
nested primer.
RT-PCR analyses
Total cellular RNA (3 µg) was treated with RNase-free DNase I
(Promega, Madison, WI) and subjected to reverse transcription with
random hexamer primers. To discount genomic DNA contamination, all PCR
assays included controls lacking reverse transcriptase. Samples were
normalized for cDNA input using a separate PCR specific for
-actin
cDNA sequences. PCR amplifications were performed as follows: 94°C (1
min), 58°C (
-actin), 50°C (V
1, V
2, V
), or 48°C
(J
1, J
2/3) for 1 min, and 72°C (1.5 min) for either 24
(
-actin) or 28 cycles (IgL germline transcripts).
PCR primers and probes
Oligonucleotide primers for PCR and Southern blot probes were as
follows: V
J
2 coding joins: primers V
(S)
(GGCTGCAGSTTCAGTGGAAGTGGGTC) and 3'J
2 (GTGAACAAGAGTTGAGAAGAC),
probe J
2 cod (TTCGGAGGGGGGACCAAGCTGG); V
J
coding joins:
primers V
R1 (ATGAATTCACTGGTCTAATAGGTGGTACCA) and J
R1
(TAGAATTCACTYACCTAGGACAG), probes V
R (CTGTGCTCTATGGTACAGCACCC),
V
1CP (GGATGAGGCAATAT), V
2CP (GATGATGCAATGTAT), J
1CP
(TTGGTGTTCTGGTGG), J
2CP (TATGTTTTCGGCGGT), and J
3CP
(TTTATTTTCGGCAGT); C
control: primers 5'C
(CAGAATTCACCTTCCYCTGARGAG) and 3'C
(GAGTCGACARACTCTTCTCCA), probe
C
P (TACGAGAACGACAGTCCCAG); V
1 SEs: primers BW-1H
(CCGGGAGATCTGAATTCCAC), 3'V
1-1 (GGTTCTCTTCTCAATG), and 3'V
1-2
(TATGTTGTGCCAAGTTGG), probe 3'V
1P (AAGTGGTAGTTATGAGACTGT); V
2
SEs: primers BW-1H, 3'V
2-1 (GTTGATAAACAAAGCTTGTC), and 3'V
2-2
(ATCAAGGCATAATTATTATAC), probe 3'V
2P (AGAAGATGGTAGTGAGACTG); V
SE
control: primers V
R1 (ATGAATTCACTGGTCTAATAGGTGGTACCA) and
3'V
1-2, probe V
P (GTGTAGATGGGGAAGTAGA); J
2/3 SEs: primers
BW-1H, 5'J
2/3-1 (TACCACCCACTKCWWS), and 5'J
2/3-2
(AGGTCAYAGCTCCACC), probes J
2SE (ACCAGGTGCTGGCCCCATAGG) and
J
3SE (CCCAGGTGCTTGCCCCACAGG); J
2/3 SE control: primers J
R1 and
3'J
2/3-2 (AGGTCAYAGCTCCACC), probe J
2/3P (GGTTGGGTTTYAGTCA);
V
1 RT-PCR: primers V
B (CACTTATACTCTCTCTCCTGG) and V
R,
probes V
1CP or V
2CP; V
RT-PCR: primers V
B
(GACATTCAGCTGACCCAGTCTCCA) and V
RT
(GGCCCGGGTTTWTGTTMWGRBYGTAKCACAGTG), probe V
R
(GTYCCWGAYCCACTGCCACTGAASC); J
1 RT-PCR: primers 5'J
1-2
(GATCTTTCAGTGATGTA) and J
R1, probe J
1CP; J
2 RT-PCR:
primers 5'J
2/3-2 and J
R1, probe J
2CP;
-actin RT-PCR:
primers 5'
-actin (AGAGCTATGAGCTGCCTGACGGCC) and 3'
-actin
(AGTAATCTCCTTCTGCATCCTGTC), probe 450-bp cDNA amplification product
of the 5'
-actin and 3'
-actin primers.
Reporter gene assays
The ts-abl clones (5 x 105
cells/transfection) were resuspended in 0.8 ml serum-free medium and
transfected transiently with a 0.2-ml mixture containing 12.5 µl of
Lipofectin (Life Technologies, Rockville, MD), luciferase plasmid (2.5
µg), and a control Renilla plasmid (0.5 µg, pRL-TK;
Promega). Samples were cultured at 34°C for 5 h, washed in
serum-free medium, and resuspended in complete medium. Subsequently,
one-third of the transfected cells were cultured at the permissive
temperature, and the remaining two-thirds were incubated at the
nonpermissive temperature for 22 h. Firefly and Renilla
luciferase activities were measured in protein extracts (25 µg) with
a dual assay kit (Promega). The pV
and pV
E
constructs contain
the V
2 promoter alone (14) or together with E
2-4
(14) in the BamHI site of the pGL2 basic
plasmid (Promega). The 6x
B plasmid contains six
B binding sites
located upstream of a minimal thymidine kinase promoter in pGL2-basic
plasmid (18).
| Results |
|---|
|
|
|---|
B regulates Ig
gene rearrangement
To investigate the transcriptional mechanisms that control IgL
gene assembly, we have used a conditionally transformed pre-B cell
model that closely mimics the physiology of primary pre-B cells.
Inactivation of the temperature-sensitive v-abl oncogene in
these cells (ts-abl cells) induces levels of RAG gene expression,
NF-
B, and IgL rearrangement that are comparable to those observed in
vivo (12, 15, 16). We generated transfectants of the
ts-abl line that stably express a mutant form of the NF-
B inhibitory
protein I
B
, termed I
B
N. This transdominant inhibitor is
resistant to signal-dependent proteolysis and blocks nuclear
translocation of NF-
B complexes containing the transactivating
subunits c-Rel and RelA (12). In prior studies, we found
that inhibition of the NF-
B signaling pathway in ts-abl cells
blocked germline J
transcription and V
J
rearrangement,
presumably by interfering with iE
function (12).
Unlike the Ig
locus, which contains
100 V
gene segments that
randomly rearrange with four functional J
elements, the murine Ig
locus primarily uses two V gene segments, V
1 and V
2
(19). In cells that fail to express functional Ig
protein, V
segments rearrange preferentially to their most proximal
set of J
-C
clusters (refer to diagram of the Ig
locus in Fig. 1
A). Prior studies had shown
that Ig
enhancers lack
B sites and function in plasma cells that
are deficient for nuclear NF-
B (14). Therefore, we
expected that levels of Ig
rearrangements would be unaffected in
ts-abl cells expressing a dominant repressor of NF-
B signaling. To
quantify Ig
rearrangements in control ts-abl cells (103 and WT.24
transfectant lacking I
B
N) and two I
B
N expressing clones
(
N.1 and
N.7), we used a set of degenerate PCR primers that
amplify all V
J
coding joins with equal efficiency. Surprisingly,
inactivation of v-abl in control cells induced levels of
V
J
rearrangement at 48 h that were at least 5- to 10-fold
greater than those observed for NF-
B-defective clones (Fig. 1
B, top, lanes 3, 6, 9, and
12). Recombinase activity was unaffected in I
B
N clones
as judged by RAG-1/2 expression and rearrangement of extrachromosomal
substrates (12). Similar decreases in V
J
rearrangement were observed when cells cultured at the nonpermissive
temperature were sorted for viability before DNA extraction (data not
shown). Thus, I
B
N-expressing cells in the process of V
J
or
V
J
recombination are not subject to significant negative
selection during the 48 h of v-abl inactivation.
|
locus is organized into two rearrangement cassettes
that each contain V
and J
gene segments as well as C
coding
exons (Fig. 1
locus contains distinct domains, each of which may be
differentially regulated by NF-
B. To test whether all gene segments
are equally impaired by c-Rel/RelA corepression, we probed Ig
rearrangements in ts-abl DNAs with oligonucleotides specific for each
V
or J
coding region. As shown in Fig. 1
B
N
clones exhibited a 5- to 10-fold reduction in coding joins that use
either V
1 or V
2. Similar results were obtained with J
1-,
J
2-, and J
3-specific probes (data not shown). Together with
published data (12), these findings indicate that
I
B
N acts as a global repressor of both V
J
and V
J
rearrangement in ts-abl pre-B cells.
c-Rel and RelA are required for efficient generation of V
and
J
SEs
The repressive effects of I
B
N on V
J
rearrangement in
pre-B cells might occur at two distinct levelsaccessibility of Ig
gene segments or efficient resolution of V
and J
CEs by DNA
repair complexes (20). To examine whether Ig
accessibility is regulated by NF-
B, levels of RAG-mediated cleavage
were monitored with ligation-mediated PCR (LM-PCR) assays that
specifically detect V
or J
SEs (17). As shown in
Fig. 2
, the kinetics of SE appearance for
all Ig
gene segments was similar to that observed for coding joins.
High levels of
SEs accumulated only after 2448 h of culturing at
the nonpermissive temperature (lanes 712).
Importantly, I
B
N-expressing clones exhibited a 5- to 10-fold
reduction in all V
and J
SEs when compared with the control
clones (Fig. 2
, A and B). Our prior studies of
extrachromosomal substrates showed that recombination of accessible
RSSs is unaffected by I
B
N (12). Together with the
data presented in Fig. 2
, these findings strongly suggest that NF-
B
is required for RAG-mediated cleavage of Ig
gene segments, the
initial step of V
J
rearrangement.
|
germline transcription is repressed in Rel-deficient pre-B
cells
Current evidence suggests that Ag receptor loci are
activated for rearrangement via the binding of developmentally
regulated transcription factors to Ig and TCR enhancers
(21). In turn, functional enhancers direct germline
transcription of linked gene segments, which is likely required for
their accessibility. Consistent with this regulatory model, we have
demonstrated previously a requirement for c-Rel/RelA in the activation
of J
-C
germline transcription and rearrangement in ts-abl cells
(12). To investigate accessibility mechanisms at the Ig
locus, we designed a series of semiquantitative RT-PCR assays that are
specific for transcripts derived from unrearranged Ig
gene segments.
Initially, total cDNA was prepared from each ts-abl clone cultured at
the permissive and nonpermissive temperatures. Subsequent to PCR
amplification, germline transcripts were detected with oligonucleotide
probes specific for the J
1 or J
2 gene segments, which reside in
separate Ig
cassettes. As shown in Fig. 3
A, levels of germline J
transcripts were significantly reduced in I
B
N-expressing pre-B
cells cultured at the nonpermissive temperature.
|
-actin (Fig. 3
-actin transcripts present in each sample.
Results from a representative set of experiments for each V
and J
assay are shown in Fig. 3
and J
germline transcripts were induced after inactivation of
the v-abl oncogene. By comparison, there was a consistent
reduction in the steady-state levels of germline Ig
transcripts
measured in I
B
N pre-B clones. The observed reduction in Ig
transcripts was not a general phenomenon, because a separate assay for
V
germline mRNAs revealed no significant differences between the
control and I
B
N clones at either temperature (Fig. 3
J
rearrangement, indicating a locus-wide repression
of Ig
accessibility in NF-
B-arrested cells.
Ig
promoter and enhancer activities are unaffected by I
B
N
Prior DNA sequence analyses provide no evidence for the presence
of NF-
B binding sites in the known Ig
enhancers (E
3-1 and
E
2-4). Moreover, both Ig
enhancers are fully functional in a
plasma cell line that lacks nuclear NF-
B (14). However,
it remained possible that NF-
B regulates E
function via an
indirect mechanism in earlier stages of B cell development. In this
regard, we have shown that NF-
B controls the inducible expression of
Oct-2, a B lineage-specific transcription factor that binds to critical
sites in all Ig promoters (22). To test the possibility
that NF-
B regulates known Ig
elements in ts-abl cells, we
monitored the activity of E
2-4 and the V
2 promoter (PV
) with a
series of luciferase reporter genes. As expected, luciferase expression
from a control construct driven by a hexamer of NF-
B sites was
severely impaired in I
B
N clones cultured at the nonpermissive
temperature (Fig. 4
). In contrast, the
function of both PV
and E
was similar in the I
B
N and
control cells. These data are in keeping with the prior finding that
E
activity is NF-
B-independent in mature B cell lines
(14). Collectively, our results strongly suggest that
Ig
germline transcription and recombinational accessibility are
regulated independently of E
function in pre-B lymphocytes.
|
| Discussion |
|---|
|
|
|---|
B, RAG expression, and V
J
rearrangement that are comparable
to those observed in bone marrow pre-B cells (15, 16). By
using a dominant mutant of I
B, we have shown previously that the
NF-
B signaling pathway is critically required for activation of
V
J
rearrangement in ts-abl cells (12). Unlike
traditional v-abl pre-B cells, the ts-abl lines also execute
high levels of Ig
gene assembly (23). This unique
feature has allowed us to explore the transcription factor requirements
for efficient V
J
rearrangement in the appropriate developmental
context. Our results clearly show that NF-
B is required for full
activation of Ig
germline transcription and the recombinational
accessibility of all five V
and J
gene segments.
What is the mechanism by which NF-
B exerts global control over the
Ig
locus in pre-B cells? Our functional data (Fig. 4
) exclude the
possibility that the core V
promoter (PV
) or E
elements are
regulated either directly or indirectly by NF-
B in extrachromosomal
vectors. However, V
germline transcription is clearly inhibited in
the NF-
B-arrested cells, whereas V
germline transcription is
unaffected (Fig. 3
). Therefore, it seems likely that endogenous PV
function is potentiated in a chromosomal context by unidentified
element(s) that respond to the NF-
B signaling pathway in ts-abl
cells. Similarly, NF-
B may control the activities of J
germline
promoters. Emerging studies have identified the presence of J
C
germline transcripts in pre-B cells (24), but the location
and functional architecture of the putative J
promoters remain
unknown. Given the results from germline transcription assays (Fig. 3
),
it is tempting to speculate that these promoter elements may be either
directly or indirectly regulated by NF-
B. Resolution of these
important issues awaits identification of the precise regulatory
sequences within the Ig
locus that are NF-
B-responsive in pre-B
cells.
In summary, our findings support a model for Ig
activation that is
highly reminiscent of the mechanisms that regulate Ig
activation.
Both L chain loci require c-Rel and RelA for initiation of V-J
rearrangement and for J-C germline transcription. During Ig
activation, NF-
B is thought to function primarily via binding to its
cognate site in iE
(25). Mouse knockout studies
indicate that iE
is critical for mediating V
J
rearrangement in
pre-B cells (26). In contrast, the distal 3'E
is
NF-
B-independent and plays a dominant role in constitutive
expression of rearranged Ig
genes in mature B cells
(27). Likewise, the distal Ig
enhancers (E
2-4 and
E
3-1), which are NF-
B-independent, are likely to be the major
regulators of Ig
expression in mature B cell subsets
(13). By analogy to iE
, unidentified NF-
B-dependent
elements may control initial locus activation in pre-B cells. In both
the ts-abl system and a second pre-B cell model (24), the
kinetics of Ig
germline transcription and rearrangement (2448 h)
are delayed relative to Ig
(1224 h, data not shown). Because
activation of NF-
B in pre-B cells induces Ig
gene assembly by its
direct binding to iE
, these kinetic data suggest that efficient
V
J
rearrangement may require the prior induction of other
transcription factors by NF-
B (e.g., Oct-2). Thus, indirect
regulation of a novel Ig
element by NF-
B may provide an
attractive explanation for the temporal order of L chain gene assembly
in pre-B cells.
| Acknowledgments |
|---|
/E
fragments, Larry
Kerr for the 6x
B construct, and D. Ballard and W. Khan
(Vanderbilt University, Nashville, TN) for valuable comments. | Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Eugene M. Oltz, Department of Microbiology and Immunology, Vanderbilt University School of Medicine, A-4203 Medical Center North, Nashville, TN 37232-2363. E-mail address: oltzem{at}ctrvax.vanderbilt.edu ![]()
3 Abbreviations used in this paper: RAG, recombination-activating gene; RSS, recombination signal sequence; CE, coding end; SE, signal end; iE
, Ig
intronic enhancer; LM-PCR, ligation-mediated PCR; ts, temperature sensitive. ![]()
Received for publication February 9, 2001. Accepted for publication April 18, 2001.
| References |
|---|
|
|
|---|
1 gene segment at the TCR
locus by a germline promoter. Immunity 10:313.[Medline]
-/- small pre-BII cells have no L chain gene rearrangements: detection by high-efficiency single cell PCR. Immunity 11:309.[Medline]
gene transcription and rearrangement in precursor B lymphocytes. Immunity 5:563.[Medline]
locus is duplicated and functionally independent of NF-
B. Genes Dev. 4:978.
B/Rel factors and blocks
gene transcription in pre-B lymphocytes. Genes Dev. 8:678.
B sites. Proc. Natl. Acad. Sci. USA 88:3715.
chains. Proc. Natl. Acad. Sci. USA 81:2484.
locus: differential effects on DNA cleavage and joining. Genes Dev. 12:2305.
gene rearrangement by CBF/PEBP2. J. Exp. Med. 185:1193.
B regulates inducible Oct-2 gene expression in precursor B lymphocytes. J. Biol. Chem. 272:28826.
and
for rearrangement at consecutive developmental stages. Eur. J. Immunol. 29:2167.[Medline]
immunoglobulin enhancer-binding protein NF-
B by a posttranslational mechanism. Cell 47:921.[Medline]
light chain intronic enhancer/matrix attachment region impairs but does not abolish V
J
rearrangement. Immunity 4:377.[Medline]
enhancer influences the ratio of Ig
versus Ig
B lymphocytes. Immunity 5:241.[Medline]
This article has been cited by other articles:
![]() |
L. R. Thomas, H. Miyashita, R. M. Cobb, S. Pierce, M. Tachibana, E. Hobeika, M. Reth, Y. Shinkai, and E. M. Oltz Functional Analysis of Histone Methyltransferase G9a in B and T Lymphocytes J. Immunol., July 1, 2008; 181(1): 485 - 493. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. M. Souto-Carneiro, R. Fritsch, N. Sepulveda, M. J. Lagareiro, N. Morgado, N. S. Longo, and P. E. Lipsky The NF-{kappa}B Canonical Pathway Is Involved in the Control of the Exonucleolytic Processing of Coding Ends during V(D)J Recombination J. Immunol., January 15, 2008; 180(2): 1040 - 1049. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Martone, G. Euskirchen, P. Bertone, S. Hartman, T. E. Royce, N. M. Luscombe, J. L. Rinn, F. K. Nelson, P. Miller, M. Gerstein, et al. Distribution of NF-{kappa}B-binding sites across human chromosome 22 PNAS, October 14, 2003; 100(21): 12247 - 12252. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. G. Cowell, M. Davila, K. Yang, T. B. Kepler, and G. Kelsoe Prospective Estimation of Recombination Signal Efficiency and Identification of Functional Cryptic Signals in the Genome by Statistical Modeling J. Exp. Med., January 20, 2003; 197(2): 207 - 220. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. L. Sikes, A. Meade, R. Tripathi, M. S. Krangel, and E. M. Oltz Regulation of V(D)J recombination: A dominant role for promoter positioning in gene segment accessibility PNAS, September 17, 2002; 99(19): 12309 - 12314. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |