Octamer independent activation of transcription from the kappa immunoglobulin germline promoter.

Abstract

Previous analyses of immunoglobulin V region promoters has led to the discovery of a common octamer motif which is functionally important in the tissue-specific and developmentally regulated transcriptional activation of immunoglobulin genes. The germline promoters (Ko) located upstream of the J region gene segments of the kappa locus also contain an octamer motif (containing a single base pair mutation and referred to as the variant octamer) which has been shown previously to bind Oct-1 and Oct-2 transcription factors in vitro. To further elucidate the role of this variant octamer motif in the regulation of germline transcription from the unrearranged kappa locus, we have quantitated the relative binding affinity of Oct-1 and Oct-2 for the variant octamer motif and determined the functional role of this octamer motif in transcriptional activation. We find that, although the variant octamer motif binds Oct-1 and Oct-2 in vitro with 5-fold lower affinity than the consensus octamer motif, mutation of the variant octamer motif to either a consensus octamer or non-octamer motif has no effect on transcriptional activation from the germline promoter. We also find significant differences in activation of germline and V region promoters by kappa enhancers. Our results suggest that the germline promoters and V region promoters differ in their dependence on octamer for activation and respond differently to enhancer activation. These findings have important implications in regulation of germline transcription as well as concomitant activation of the V-J recombination of the kappa light chain locus.

Full text

Page 1

 1996 Oxford University Press
4805–4811
Nucleic Acids Research, 1996, Vol. 24, No. 23
Octamer independent activation of transcription from
the kappa immunoglobulin germline promoter
Anila Prabhu1, Darin P. O’Brien2, Georgia L. Weisner+, Regan Fulton3 and
Brian Van Ness1,2,*
1Institute of Human Genetics, 2Department of Biochemistry and 3Department of Lab Medicine and Pathology,
University of Minnesota, 515 Delaware Street SE, Minneapolis, MN 55455, USA
Received July 10, 1996; Revised and Accepted October 8, 1996
ABSTRACT
Previous analyses of immunoglobulin V region pro-
moters has led to the discovery of a common octamer
motif which is functionally important in the tissue-
specific and developmentally regulated transcriptional
activation of immunoglobulin genes. The germline
promoters (Ko) located upstream of the J region gene
segments of the kappa locus also contain an octamer
motif (containing a single base pair mutation and
referred to as the variant octamer) which has been
shown previously to bind Oct-1 and Oct-2 transcription
factors in vitro. To further elucidate the role of this variant
octamer motif in the regulation of germline transcription
from the unrearranged kappa locus, we have quantitated
the relative binding affinity of Oct-1 and Oct-2 for the
variant octamer motif and determined the functional role
of this octamer motif in transcriptional activation. We find
that, although the variant octamer motif binds Oct-1 and
Oct-2 in vitro with 5-fold lower affinity than the con-
sensus octamer motif, mutation of the variant octamer
motif to either a consensus octamer or non-octamer
motif has no effect on transcriptional activation from the
germline promoter. We also find significant differences in
activation of germline and V region promoters by kappa
enhancers. Our results suggest that the germline pro-
moters and V region promoters differ in their
dependence on octamer for activation and respond
differently to enhancer activation. These findings have
important implications in regulation of germline tran-
scription as well as concomitant activation of the V-J
recombination of the kappa light chain locus.
INTRODUCTION
The study of immunoglobulin (Ig) genes has provided important
insights into the molecular mechanisms involved in the activation
of transcription by tissue specific and developmentally regulated
transcriptional regulatory regions. In addition, key develop-
mental stages of B cell development have been elucidated and
defined by the study of Ig genes. One such stage of development
is the pre-B cell stage, defined by the presence of rearrangements
at the heavy chain locus and the absence of rearrangements at the
light chain locus (1,2). Furthermore, transcriptional activation of
the unrearranged (germline) kappa light chain locus occurs at this
stage of development (3). Interestingly, induction of germline
transcription of the kappa locus occurs concomitantly with the
activation of V-J recombination, leading to the suggestion that
transcriptional activation of kappa is important in targeting the
locus for recombination (4). Therefore, this hypothesis suggests
that the early activation of germline transcription of the kappa
locus plays an important role in B cell development.
The transcriptional activation of the kappa locus involves a
number of transcriptional regulatory elements, including pro-
moters and enhancers (Fig. 1A). There is a promoter located 5′ of
each V region gene segment as well as two germline promoters
(Ko) 5′ of the J gene segments in the unrearranged locus. Although
less is known about the regulation of kappa germline promoters,
comparison of germline and V region promoters indicates that
there are a number of important differences as well as some
similarity between these promoters. One of the common features
of V region promoters is the presence of a highly conserved
octamer sequence (ATTTGCAT). The importance of the octamer
motif in transcription from V region promoters has been docum-
ented previously (5–8). Comparison of V region and germline
promoters suggests little sequence conservation other than the
presence of this octamer motif. However, the octamer motif in the
germline promoters contains a single transition, has an inverted
orientation (ATGTAAAT), and is referred to here as the variant
octamer. Furthermore, the V region promoters contain a TATA box
while the germline promoters do not. In addition to these sequence
differences, it has been reported that there are functional differ-
ences between the germline and V region promoters since the rate
of transcription from the V region promoters is approximately
three to six times greater than the germline promoters (3).
Sequence analysis of the kappa germline promoter of the murine
and human loci led to the identification of a highly conserved 80 bp
region flanking the variant octamer which has a 78% homology
between the two species (9). In addition to the sequence similarities
observed in this region, the germline transcripts in these two
different species are processed in a similar manner (9).
Based on the functional importance of the octamer in the V
region promoters and the presence of the variant octamer in the
*To whom correspondence should be addressed at: UMHC, Box 206, University of Minnesota, Minneapolis, MN 55455, USA. Tel: +1 612 624 9944;
Fax: +1 612 626 7031; Email: vanness@lenti.med.umn.edu
+Present address: Center for Human Genetics, 11001 Cedar Avenue, Suite 510, Case Western Reserve University, Cleveland, OH 44106, USA

Page 2

4825
Nucleic Acids Research, 1994, Vol. 22, No. 1
Nucleic Acids Research, 1996, Vol. 24, No. 23
4825
Figure 1. Organization of the endogenous kappa light chain locus and reporter
vectors. (A) Organization of the endogenous kappa locus. Coding regions,
including variable gene segments (V), joining gene segments (J), and the
constant region gene segment (C) are indicated. Transcriptional regulatory
regions including the intron enhancer (IM), the 3′ kappa enhancer (3′KE), the
germline promoters (5′Gp and 3′Gp) and a representative Vκ region promoter
(Kp) are shown (not drawn to scale). (B) Schematic representation of the
reporter gene vectors used in transient transfection assays. Luciferase
expression vectors driven by either the germline (Gp) or Vk21E (Kp) promoter
with either the intron enhancer, the 3′ enhancer or both enhancers are shown.
Reporter vector nomenclature is indicated to the right of each vector.
germline promoters, it was our hypothesis that the variant octamer
motif within the germline kappa promoter was responsible for the
3–6-fold difference observed between the transcriptional activity
of the germline and V region promoters. Moreover, it was possible
that this difference in the transcriptional activity from the germline
and V region promoters could be attributed to the differences in the
affinity of the germline and V region promoters for octamer
binding factors. To test this hypothesis, we quantitated the
difference in binding affinity between germline and V region
octamer motifs and determined the functional significance of the
octamer motif in transcriptional activation of germline promoters.
Our results suggest that unlike the octamer dependent tran-
scription from the V region promoters, transcriptional activation
of kappa germline promoters appears to be octamer independent.
This clearly distinguishes the germline promoters from the
octamer dependent transcription of the V region promoters.
Moreover, we note that the germline and V region promoters
show differences in response to the kappa enhancers.
MATERIALS AND METHODS
Cell culture and nuclear extracts
The pre-B cell lines 3-1, 38B9, 1-8 and the myeloma cell line
S194 have been characterized previously (10,11). The pre-B cell
line 38B9 was kindly provided by Eugene Oltz (Vanderbilt
University). Cells were maintained in RPMI 1640 (Gibco BRL,
Gaithersburg, MD), 10% FCS, 25 U/ml penicillin, 25 µg/ml
streptomycin, 2 mM L-glutamine and 50 µM β-mercaptoethanol.
Nuclear extracts were prepared using a previously described
modification (12) of the method outlined by Dignam et al. (13).
Vector construction and transient transfections
The construction of luciferase expression vectors (Kp.luc,
IM.Kp.luc, 3′KE.IM.Kp.luc, Gp.luc, IM.Gp.luc, 3′KE.IM.Gp.luc,
and IM.Gp.(OCT)luc used for transient transfections were de-
scribed previously (14). IM.Kp.oct–.luc and IM.Gp.oct–.luc were
constructed by overlap PCR mutagenesis with oligos containing a
3 bp mutation in the octamer site (5′-ACA GCC AAT CTG AAC
ATT TCA GAG GCT TAG-3′ for IM.Gp.oct–.luc and 5′-TCT
GCT GAG CTG ATG TTC AGA TGG AGT CAT-3′ for
IM.Kp.oct–.luc) and confirmed by sequencing.
Mouse B-cell lines were transfected by a modification of the
DEAE–dextran method (15,16). Briefly, 10 × 106 cells per
transfection were washed once in sterile TS (137 mM NaCl, 5 mM
KCl, 0.6 mM Na2HPO4.7H2O, 25 mM Tris, 0.5 mM MgCl2,
0.7 mM CaCl2, pH 7.4) and resuspended in 50 µl TS at 37?C.
Ten µg test plasmid and 10 µg of the control plasmid, pCH110, a
β-galactosidase expression vector (Pharmacia, Piscataway, NJ),
purified by two CsCl gradients, were resuspended in 1.5 ml TS
containing 10 µl of 10 mg/ml DEAE–dextran (67 µg/µl final
concentration) and warmed to 37?C. The cells were added to the
plasmid/TS/DEAE–dextran solution and mixed by pipetting. The
mixture was allowed to sit at 37?C for 20 min. Cells were washed
once in sterile TS and placed in a 10 ml volume at 37?C and 7%
CO2. Ten µg/ml of LPS (Difco, Detroit, MI) was used for induction
of pre B cell lines. Transfection of each plasmid construct was
performed in triplicate. Cells were harvested 18 h after transfection
and luciferase and β-galactosidase assays were performed as
previously described (17). Briefly, we assayed luciferase and
β-galactosidase activities from each of the three transfections and
luciferase activity was normalized to β-galactosidase activity for
each sample (luciferase RLU/β-galactosidase RLU). The mean of
the normalized luciferase activity and the standard error of the
mean are indicated for each plasmid construct.
Probes
The probes used for mobility shift assays were end-labeled with
T4 polynucleotide kinase (NEB) to a specific activity of 0.5 × 108
to 1 × 109 c.p.m./µg. A 260 bp fragment containing the variant
octamer of the human germline promoter was amplified with
primers 5′-ACA AAG CTT TGA GGG GCT CGT-3′ and
5′-CAC GTC GAC ATG CGT GGG ACA CAA AT-3′, and
digested with PstI at 37?C for 2 h prior to gel purification. The
130 bp probe containing the variant octamer from the mouse
germline promoter was amplified with primers 5′-ACA CGT
CTA GAG TAA GCT GGA-3′ and 5′-CCC GGA TCC CTG
CAG CCA TTT CAC TTC AGA TGA A-3′.
Mobility shift assay
Approximately 10 000 c.p.m. (1–5 fmole of DNA) of end-labeled
DNA probe were added to a 15 µl binding reaction containing
10 mM Tris–HCl (pH 7.4), 5% (v/v) glycerol, 50 mM NaCl, 1
mM EDTA, 1 mM DTT, 0.3 mM ATP, 2.5–5 µg of sheared
salmon sperm DNA (ssDNA, Sigma) and 5–10 µg nuclear extract
protein (18). After incubation at 25?C for 30 min, the binding
reaction was loaded on a 4.5% polyacrylamide gel containing
40 mM Tris, 0.384 M glycine and 2 mM EDTA. The samples
were subjected to electrophoresis at 8 V/cm, and the gel was dried
and exposed to X-ray film for 12–24 h.
Competitive binding assays were performed by adding 100 or
250-fold amounts of unlabeled 24 bp octamer containing fragments
to the binding buffer and incubating at room temperature for 5 min
prior to addition of the γ-32P-labeled consensus octamer fragment.
The reaction was incubated for an additional 30 min and

Page 3

Nucleic Acids Research, 1996, Vol. 24, No. 23
4826
Figure 2. The germline promoter is a weaker promoter than a V region promoter in pre-B and mature B cell lines. Luciferase reporter vectors used for the transient
transfection experiments are shown in Figure 1B. Results of transfection of the 3-1 pre-B cell line with constructs containing either the V region promoter (A) or the
germline promoter (B) are shown. Solid bars represent uninduced activity and hatched bars represent activity induced with 10 µg/ml of LPS. Additionally, transfection
of the mature B cell line S194 with vectors containing either the V region promoter (C) or the germline promoter (D) is also indicated. Transfection efficiency was
normalized by co-transfection with the β-galactosidase expression vector, pCH110, and dividing luciferase activity by β-galactosidase activity. The ratio is labeled
luminescence and error bars represent the standard error of the mean.
subjected to electrophoresis as described above. Alternatively, the
competition reaction was performed by proportionately increasing
the binding reaction volume 10-fold, adding the γ-32P-labeled
consensus octamer fragment as probe and incubating for 30 min.
Cold competitor (250-fold) was then added to the binding
reaction and 15 µl aliquots were loaded on a polyacrylamide gel
in 5–15 min intervals over 90 min. After autoradiography,
densitometric analysis was performed to determine the relative
density of the bands corresponding to the protein–DNA complexes.
DNase I protection
Binding reactions used for mobility shift assays were increased
5–10-fold and contained 25–50 µg of induced or uninduced nuclear
extract from 3-1 pre B cells and either a 160 bp human or 130 bp
mouse germline labeled fragment as previously described (19).
After incubating for 30 min at room temperature, 2.5 U of DNase
I in a buffer containing 10 mM Tris–HCl (pH 7.6), 5 mM CaCl2 and
10 mM MgCl2 were added to the solution and incubated for 3–5 min
on ice and terminated by the addition of 2 µl of 0.5 M EDTA. The
complex was then either gel purified or immediately phenol
extracted and ethanol precipitated with 20 µg of carrier tRNA. If the
complexes were gel purified, the solution was applied to a 5%
polyacrylamide gel and subjected to electrophoresis. Bands were
identified by autoradiography, excised from the wet gel, and the
DNA was recovered by diffusion over 6–10 h into a buffer
containing 0.5 M sodium acetate, 1 mM EDTA, 10 mM magnesium
acetate and 0.1% SDS. After centrifugation through siliconized glass
wool, the products were phenol extracted and ethanol precipitated
with 20 µg of tRNA. The DNA was resuspended in 5 µl of loading
buffer and the fragments were separated on a 6%
polyacrylamide–8 M urea denaturing gel next to lanes containing
free, or unbound probe treated with DNase I, and an A+G sequence
of each region (12), dried and exposed to X-ray film.
Statistical analysis
Statistical analysis of the data was performed by Student’s t-test
and Statswork software. Probability (P) values ?0.05 were
considered significant.
RESULTS
Germline promoters are weaker than V region
promoters in pre-B and mature B cell lines
To determine the role of the octamer motif in activation of
germline transcription, we utilized reporter test constructs with
either the Vk21E promoter (designated Kp) or the kappa 5′
germline promoter (designated Gp) to drive expression of a
luciferase reporter gene. Derivative vectors were made by
inserting either the kappa intron enhancer (IM) or the 3′ kappa
enhancer (3′KE) or both as shown in Figure 1B.
In the absence of LPS, a strong inducer of kappa transcription,
the vectors containing either the V region promoter (Kp) or the
germline promoter (Gp) alone (no enhancers present) show
relatively little activity in pre-B cell lines (Fig. 2). Even in the
presence of LPS, the germline and V region promoters by
themselves or in the presence of the 3′ kappa enhancer are not
significantly inducible (Fig. 2A and B). On the other hand, we
observe that in pre-B cells, both the germline and the V region
promoters were activated by LPS when paired with the intron
enhancer (Fig. 2A and B). Comparison of LPS induced transcrip-
tional activity of IM.Gp.luc and IM.Kp.luc indicates a 2–4-fold
greater increase in the latter case (P < 0.026) in the pre-B cell line
3-1 (Fig. 2A and B). Identical results were seen in another pre-B
cell line, 1-8 (data not shown).
In the pre-B cell line 3-1, the presence of both the enhancers with
the V region promoter (3′KE.IM.Kp.luc) resulted in transcriptional
activity equivalent to the activity observed with the intron enhancer
alone (IM.Kp.luc) (Fig. 2A). However, constructs with the intron
enhancer and 3′ kappa enhancer paired with the germline promoter
(3′KE.IM.Gp.luc) resulted in a statistically significant decrease

Page 4

4827
Nucleic Acids Research, 1994, Vol. 22, No. 1
Nucleic Acids Research, 1996, Vol. 24, No. 23
4827
(P < 0.047) in transcriptional activity in 3-1 cells compared with the
presence of the germline promoter with the intron enhancer alone
(IM.Gp.luc) (Fig. 2B). Figure 2B is representative of multiple
experiments that consistently show suppression of the germline
promoter with the addition of the 3′KE sequence elements. This
decrease is due to suppression of both uninduced and induced
activity. This suppressive effect of 3′KE on the germline promoter
was also observed in the pre-B cell line 1-8 (data not shown).
Transfections of the myeloma cell line S194 demonstrate that both
the intron and the 3′ kappa enhancer are active at this stage of B cell
development (Fig. 2C and D). Furthermore, we observed a 2–5-fold
difference in the reporter gene expression from the germline and V
region promoters which is consistent with differences from
endogenous V region and germline promoters noted previously (3).
The difference between activation of the germline and V region
promoters may lie in the ability of the 3′ enhancer to preferentially
activate the V region promoter. Consistent with this explanation, we
found that the transcriptional activity of the germline promoter with
both enhancers is the additive sum of either enhancer alone
(Fig. 2D). However, with the V region promoter we observed that
the presence of both enhancers results in transcriptional activity that
is more than the additive sum of the individual transcriptional
activities observed with either the intron or the 3′ kappa enhancer
alone (Fig. 2C). Based on these results we conclude that, (i) in pre-B
cells, both the germline and the V region promoters are dependent
on the intron enhancer, (ii) in both pre-B and mature B cells, the
germline promoter is weaker than the V region promoter, (iii) in
pre-B cells, the 3′ enhancer has no effect on the LPS inducibility of
the V region promoter but appears to suppress the germline
promoter and (iv) enhancer synergy is observed only in mature B
cells with the V region promoter and not with the germline promoter.
The Oct-1 and Oct-2 transcription factors bind to the
variant octamer of the germline promoter
It has been shown previously that nuclear factors Oct-1 and Oct-2
bind to the variant octamer present in the germline promoter of
the mouse kappa locus (20). Although there is a remarkable
amount of sequence conservation between human and mouse
germline promoters in the region containing the variant octamer,
there are some differences in the region flanking the variant
octamer motif. To determine the effect of these differences on the
ability of the variant octamer in the human germline promoter to
bind Oct-1 and Oct-2, we performed gel shift analysis with a
probe corresponding to the human germline promoter. The results
of this analysis suggest that a 24 bp probe containing the human
variant octamer (Fig. 3A, II) forms complexes indistinguishable
from those seen with a consensus octamer probe (Fig. 3A, I).
Furthermore, we find that similar analysis of Oct-1 and Oct-2
complexes with a human 160 bp probe or a murine 130 bp probe
containing the germline promoters resulted in complexes very
similar to complexes observed with the 24 bp probe (Fig. 3A,
H-160 versus M-130). We have also observed that the signal
intensities of complexes with the variant octamer are consistently
less than those observed with the consensus octamer probe. These
results suggest that the human germline variant octamer binds
Oct-1 and Oct-2 in vitro and that sequences flanking the variant
octamer do not contribute significantly to the binding of Oct-1
and Oct-2. The failure to detect complexes other than Oct-1 and
Oct-2 with the 160 bp and 130 bp probes suggests that only Oct-1
and Oct-2 are capable of binding to this region in vitro.
To further evaluate the potential contribution of sequences
flanking the variant octamer to the binding of Oct-1 and Oct-2 or
other proteins not detected by gel shift analysis, we used DNase I
footprinting to analyze protection of specific residues in protein-
bound complexes. If differences in transcriptional activity between
germline and V region promoters are due to differences in nuclear
factors bound by the octamer motif which are not detected by gel
shift analysis, such differences may be detected by DNase I footprint
analysis. Footprint analysis of uninduced and induced complexes
bound to the variant germline promoter reveals an equivalent 18 bp
region of protection (Fig. 3B). The pattern of protection observed
with the variant octamer motif is similar to that observed with the
consensus octamer motif (21).
Oct-1 and Oct-2 have a decreased relative affinity for
the human and the mouse variant octamer motif
We have consistently found that octamer complexes with the
mouse and human germline promoter containing the variant
octamer produce bands of weaker intensity in comparison with
complexes with the consensus octamer motif. The correlation
between apparently weaker octamer binding to the variant octamer
with decreased transcriptional activity of germline promoters
suggested to us that the difference in affinity for octamer binding
proteins may have important functional consequences. In order to
quantitate the differences in octamer protein binding to the
consensus and the variant octamer sites, we performed the gel shift
competition experiments shown in Figure 4A. The 24 bp human
V region probe containing the consensus octamer motif was used
as a probe for detection of octamer binding activity in nuclear
extracts from LPS induced pre-B cells. Competition for binding to
this probe was performed with increasing concentrations of either
unlabeled V region (consensus octamer) or germline promoter
(variant octamer) sequences. The results of this analysis suggest
that the consensus octamer competes better than the variant
octamer (Fig. 4A). Comparison of the slopes of the lines derived
from the relative density of Oct-1 for the consensus and the variant
octamer provides an indication of the relative affinity of the Oct-1
protein for the consensus and the variant octamer (Fig. 4B). The
results of this analysis suggest that the relative affinity of the
octamer binding proteins for the variant octamer is 5-fold lower
than that of the consensus octamer. Our results indicate that the fold
difference observed in the binding of octamer proteins correlated
with the fold differences observed in the transcriptional activity
from V region and germline promoters.
Functional analysis of the variant octamer in germline
transcription in pre-B and mature B cells
To further test whether the relatively low level of transcription
from the germline promoter compared with the V region
promoter was due to the differences in the binding affinity of
Oct-1 and Oct-2, we first mutated the variant octamer in the
germline promoter to a consensus octamer site, [designated
Gp.(oct).luc]. The functional effect of this mutation was then
tested by transient transfection in pre-B and mature B cell lines.
In addition, we also created a null mutation of the consensus
octamer in the V region (designated Kp.oct–.luc) to produce a
non-octamer site that is unable to bind octamer binding proteins
(data not shown).

Page 5

Nucleic Acids Research, 1996, Vol. 24, No. 23
4828
Figure 3. Protein binding to the octamer sequence in the germline promoter. (A) Mobility shift assays of the human and the mouse germline promoter. Oligonucleotide
probes (24 bp) containing the human consensus (I) or variant (II) octamer motif were incubated with nuclear extracts from uninduced (–) or LPS induced (+) 3-1 mouse
pre-B cells, or in vitro translated Oct-2a protein (Oct-2). The observed difference in mobility between the in vitro translated Oct-2 protein and the Oct-2 from the extracts
is due to the difference in size between human and mouse Oct-2 protein as described previously (34). Alternatively, a 160 bp human fragment (H-160) or a 130 bp
mouse fragment (M-130) containing the variant octamer was used as probe (extracts and proteins as indicated). Specific octamer containing complexes (Oct-1 and
Oct-2) and free probe (F) are indicated by arrows. (B) DNase I footprint analysis of the mouse germline promoter fragments. The 130 bp fragment containing the mouse
variant octamer motif was end labeled on the antisense strand and incubated with nuclear extracts from uninduced and LPS induced 3-1 pre-B cells. DNA from bound
complexes from binding reactions containing extracts from either uninduced (BU) or induced (BI) cells was gel purified and analyzed on a denaturing acrylamide gel
(see Materials and Methods). Free probe was purified as a control and ran next to an A/G sequencing reaction for identification of sequences protected. The bracket
indicates the location of the variant octamer motif within the protected area.
As expected, the null mutation of the octamer motif in the V
region promoter resulted in a dramatic decrease in transcriptional
activity in both pre-B and mature B cell lines (Fig. 5),
demonstrating that the octamer motif plays an important role in
the activation from these promoters. Interestingly, the mutations
of the variant octamer to a consensus octamer in the germline
promoter had no significant effect on transcriptional activity from
the germline promoter (Fig. 5). This result was inconsistent with
our expectation that conversion of the variant octamer to a
consensus octamer site would increase transcriptional activity
from germline promoters. Therefore, to further elucidate the role,
if any, of the variant octamer motif in transcription from the
germline promoter, we mutated the variant octamer to the null
mutation described above in the V region promoter. Interestingly,
the functional analysis of this mutant construct (Gp.oct–.luc)
shows that there was no significant difference in transcriptional
activity from the germline promoter containing the variant
octamer or the one containing the octamer with the null mutation
(Fig. 5). Therefore, our results indicate that in pre-B and mature
B cell lines, it was of little consequence whether a variant,
consensus or a mutated octamer (null mutation) motif was present
in the germline promoter since the transcriptional activity from
the germline promoter was comparable in all three cases. These
results demonstrate that although the variant octamer has some
capacity to bind Oct-1 and Oct-2 in vitro, it does not have any
demonstrable functional role in the transcriptional activation of
germline promoters.
DISCUSSION
In this study we have utilized vectors containing the germline
promoter (Gp) or V region promoter (Kp) with either the kappa
intron enhancer (IM), 3′ enhancer (3′KE), or both to address the
observed difference in promoter strength of the germline and V
region promoters. Our results indicate that the binding affinity of
Oct factors to the variant octamer within the germline promoter
is 5-fold less than that of the consensus octamer present in the V
region promoter. Therefore, it was our hypothesis that the
difference between the transcriptional activity of the germline and
V region promoters could be explained by the presence of the
variant octamer in the germline promoter.
We found that the intron enhancer mediated activation of both the
V region and germline promoters in pre B cells. However, we detect
little or no activation of either promoter by the 3′ enhancer in pre-B
cells. Furthermore, reporter constructs containing both enhancers
with the V region promoter (3′KE.IM.Kp.luc) had activity similar
to that of the intron enhancer alone with the V region promoter
(IM.Kp.luc), however with the germline promoter we observed a
significant decrease in transcriptional activity upon addition of the
3′ enhancer. These results suggest the possibility that the 3′ enhancer
is able to suppress activation of germline promoters in pre-B cells.
This result indicates that there is a promoter preference regarding the
negative regulation by the 3′ kappa enhancer.
In the mature B cell line, S194, we observed that the promoters
by themselves had a low basal level of transcriptional activity.

Page 6

4829
Nucleic Acids Research, 1994, Vol. 22, No. 1
Nucleic Acids Research, 1996, Vol. 24, No. 23
4829
Figure 4. Comparison of Oct-1 and Oct-2 binding affinity of the germline
variant octamer and the Vκ consensus octamer. (A) Competition of consensus
octamer binding activity with consensus and variant octamer motifs. The Vκ probe
containing the consensus octamer was incubated with nuclear extracts from
uninduced or LPS induced cells (–LPS, +LPS). Competition of octamer binding
activity in LPS induced extracts with increasing molar excess of cold consensus,
variant or unrelated Kb competitor is also shown. (B) Densitometric analysis of
consensus and variant octamer competition of Oct-1 binding activity. A semi-log
plot of the relative density of the Oct-1 bands versus fold excess of unlabeled
consensus octamer (open square) or variant octamer (triangles) was used to
quantitate binding diffrences between the consensus and variant octamer.
Presence of the intron or the 3′ kappa enhancer increased the
transcriptional activity of both the V region and the germline
promoter. However, significantly higher transcriptional activity
was observed in the presence of both the enhancers with the V
region promoter (3′KE.IM.Kp.luc) than both enhancers with the
germline promoter (3′KE.IM.Gp.luc). In both the pre-B and
mature B cell lines we observed that the germline promoter was
consistently weaker than the V region promoter. These results are
consistent with a previous analysis of kappa transcription from
germline and rearranged alleles in plasmacytomas in which it was
found that transcription from the rearranged alleles was 3–6-fold
higher than the germline allele (3).
It was our hypothesis that the difference in transcriptional
activity between the germline and V region promoters was due to
the presence of the variant octamer in the germline promoter. The
octamer motif is an important regulatory sequence present 5′ of
virtually all Ig V genes of the heavy and light chain locus and is
also present in the enhancer elements in Ig genes (19,22–24). The
octamer motif is highly conserved and has been implicated in the
induction of transcription from the heavy and light chain V region
promoters by deletion analysis (5,7,23). We and others have also
recognized that the mouse and human germline promoters
contained a variant octamer motif (9,18). In this study we
demonstrated that the single base pair transition in the variant
octamer resulted in a decreased binding affinity but did not
abolish the recognition of this variant octamer motif by Oct-1 and
Oct-2 nuclear factors. Our data suggest that the relative affinity
of Oct-1 and Oct-2 for the variant octamer motif was 5-fold lower
than the consensus octamer and this difference closely correlates
with the differences in the promoter activity of the germline
promoter and V region promoter in transient transfection assays.
Non-consensus octamer motifs have been identified in other
genes that demonstrate decreased binding affinity to the Oct-1
and Oct-2 nuclear factors and also a decrease in transcriptional
activity (25). Atchinson et al. reported the presence of a variant
octamer in the V region promoter of Vk19. This variant octamer
binds to nuclear factors Oct-1 and Oct-2 weakly compared with
the variant octamer present in germline promoters of the kappa
light chain. It was observed that to compensate for an ineffectual
octamer present in the Vk19 promoter, an upstream K-Y element
was important for the normal transcriptional activity (20).
Similarly, Eaton and Calame have demonstrated the role of
flanking heptamer and pyrimidine rich sequences in octamer
dependent transcription from the V region promoters of the Ig
heavy chain (26). An additional DNA element, called the BAT
box with AT rich sequences, has been implicated in transcription
of the CD20/B1 promoter which contains a variant octamer motif
(27). We have found no homology to the K-Y element, heptamer
or pyrimidine rich sequences in the kappa germline promoters
upon computer sequence analysis. In addition, footprint analysis
of the human germline promoter as well as previous analysis of
the mouse octamer motif (20) reveals an 18 bp region which is
protected and this footprint is not affected by additional flanking
sequence. Surprisingly, we observed that the deletion of the
variant octamer present in the germline promoter had no effect on
transcription from this promoter. These results are consistent with
the results from Oct-2 knockout mice which have no detectable
Oct-2 protein binding activity (28). These mice have apparently
normal numbers of pre-B cells but have a greatly diminished
capacity to secrete immunoglobulin at later stages of B cell
development. Our finding that the variant octamer site in the
germline promoter is not required for germline promoter
activation but is required for V region promoter activation is
consistent with a role for Oct-2 in later stages of B cell
development but not early B cell development.
The presence of the TATA box and a functionally important
octamer in the V region promoters is an important distinction
between the V region and germline promoters. Although the
distance between the germline and the V region promoters from
the enhancers is different, the difference in transcriptional activity
between the V region and the germline promoter is not position
dependent because equal promoter strengths have been observed
over considerable distance from the enhancer (29). If the presence
of the variant octamer in the germline promoter is not functionally
relevant, the question arises as to what constitutes the germline
promoter. Our data thus far indicate that the germline promoter
serves as a minimal promoter, completely dependent on activation
by the intron enhancer. If accessibility and V-J recombination of the

Page 7

Nucleic Acids Research, 1996, Vol. 24, No. 23
4830
Figure 5. Functional analysis of the octamer motif in the germline and V region promoters. The pre-B cell lines 3-1 ( A), 1-8 (B), and 38B9 (C) and plasmacytoma cell line
S194 (D) were transiently transfected with plasmid constructs with the intron enhancer and either the germline or V region promoter (IM.Gp.luc/IM.Kp.luc). Reporter
constructs containing mutations (to either a consensus octamer site or a non-octamer site which lacks octamer binding activity) in the germline and V region promoters were
also used for transfection. These vectors include, IM.Gp.(oct).luc which contains a mutation in the variant octamer motif to give a consensus octamer motif and
IM.Gp.oct-.luc/IM.Kp.oct-.luc which contain a null mutation of the octamer motif. Transfection efficiency was normalized by co-transfection of the β-galactosidase expression
vector, pCH110, and dividing luciferase activity by β-galactosidase activity. Mean of triplicate transfections and standard error of the mean are shown.
kappa locus is dependent on germline transcription, then our data
suggest that the developmentally regulated accessibility of the kappa
locus is solely dependent on the activation of the intron enhancer.
Our functional data from transient transfection experiments
indicate that in pre-B cells, transcription from the germline promoter
proceeds at a significantly higher rate in the presence of the intron
enhancer compared with parallel experiments in which 3′ enhancer
was paired with the germline promoter. We consistently observed a
negative regulation by the 3′ enhancer when present with the
germline promoter. The presence of the 3′ enhancer with the V
region promoter did not have any significant effect on transcriptional
activity from this promoter in pre-B cells. This raised the possibility
that the negative regulation that we observed in the case of the 3′
enhancer was promoter specific. LPS inducibility of the 3′ enhancer
has been reported in some earlier studies (30,31). However, in these
studies the functional role of the enhancers in development was
studied by pairing them with heterologous promoters. Our studies
are more relevant since we have used either a germline or the V
region promoter to address this issue. Consistent with our observa-
tion are reports that have implicated the 3′ enhancer in preventing
precocious rearrangement of the kappa locus (32,33).
ACKNOWLEDGEMENTS
We thank Dr Eugene Oltz (Vanderbilt University) for pre B cell
line 38B9. This work was supported by NIH Grant GM37687.
REFERENCES
1 Staudt, L.M. and Lenardo, M.J. (1991) Annu. Rev. Immunol., 9, 373–398.
2 Alt, F.W., Blackwell, T.K., DePinho, R.A., Reth, M.G. and Yancopoulos,
G.D. (1986) Immunol. Rev., 89, 5–30.
3 Van Ness, B.G., Weigert, M., Coleclough, C., Mather, E.L., Kelley, D.E.
and Perry, R.P. (1981) Cell, 27, 593–602.
4 Blackwell, T.K., Moore, M.W., Yancopoulos, G.D., Suh, H., Lutzker, S.,
Selsing, E. and Alt, F.W. (1986) Nature, 324, 585–589.
5 Bergman, Y., Rice, D., Grosschedl, R. and Baltimore, D. (1984) Proc.
Natl. Acad. Sci. USA, 81, 7041–7045.
6 Falkner, F.G., Neumann, E. , and Zachau, H.G. (1984) Hoppe Seylers Z.
Physiol. Chem., 365, 1331–1343.
7 Mason, J.O., Williams, G.T. and Neuberger, M.S. (1985) Cell, 41, 479–487.
8 Wirth, T., Staudt, L. and Baltimore, D. (1987) Nature, 329, 174–178.
9 Thompson, A., Timmers, E., Kenter, M.J., Kraakman, M.E., Hendriks,
R.W. and Schuurman, R.K. (1992) Eur. J. Immunol., 22, 3167–3171.
10 Kim, K.J., Kanellopoulos Langevin, C., Merwin, R.M., Sachs, D.H. and
Asofsky, R. (1979) J. Immunol., 122, 549–554.
11 Martin, D.J. and Van Ness, B.G. (1989) Mol. Cell. Biol., 9, 4560–4562.
12 Nelms, K., Hromas, R. and Van Ness, B. G. (1990) Nucleic Acids Res., 18,
1037–1043.
13 Dignam, J.D., Lebovitz, R.M. and Roeder, R.G. (1983) Nucleic Acids Res.,
11, 1475–1489.
14 Fulton, R. and Van Ness, B.G. (1993) Nucleic Acids Res., 21, 4941–4947.
15 Fulton, R. and Van Ness, B.G. (1994) Nucleic Acids Res., 22, 4216–4223.
16 Schwartz, O., Virelizier, J.L., Montagnier, L. and Hazan, U. (1990) Gene,
88, 197–205.
17 Fulton, R. and Van Ness, B.G. (1993) Biotechniques, 14, 762–763.
18 Martin, D.J. and Van Ness, B.G.(1990) Mol. Cell. Biol., 10, 1950–1958.
19 Nelms, K. and Van Ness, B.G. (1990) Mol. Cell. Biol., 10, 3843–3846.
20 Atchison, M.L., Delmas, V., and Perry, R.P. (1990) EMBO J., 9, 3109–3117.
21 Pierani, A., Heguy, A., Fujii, H. and Roeder, R.G. (1990) Mol. Cell. Biol.,
10, 6204–6215.
22 Parslow, T.G., Blair, D.L., Murphy, W.J. and Granner, D.K. (1984) Proc.
Natl. Acad. Sci. USA, 81, 2650–2654.
23 Falkner, F.G. and Zachau, G.H. (1984) Nature, 310, 71–74.
24 Currie, R.A. and Roeder, R.G. (1989) Mol. Cell. Biol., 9, 4239–4247.
25 Baumruker, T., Sturm, R. and Herr, W. (1988) Genes Dev., 2, 1400–1413.
26 Eaton, S. and Calame, K. (1987) Proc. Natl. Acad. Sci. USA, 84, 7634–7638.
27 Thevenin, C., Lucas, B.P., Kozlow, E.J. and Kehrl, J.H. (1993) J. Biol.
Chem. 268, 5949–5956.
28 Corcoran, L.M., Karvelas, M., Nossal, G.J., Ye, Z. S., Jacks, T. and
Baltimore, D. (1993) Genes Dev., 7, 570–582.
29 Atchison, M.L. and Perry, R.P. (1986) Cell, 46, 253–262.
30 Meyer, K.B., Sharpe, M.J., Surani, M.A. and Neuberger, M.S. (1990)
Nucleic Acids Res., 18, 5609–5615.
31 Pongubala, J.M. and Atchison, M.L. (1991) Mol. Cell. Biol., 11, 1040–1047.
32 Hiramatsu, R., Akagi, K., Matsuoka, M., Sakumi, K., Nakamura, H.,
Kingsbury, L., David, C., Hardy, R.R., Yamamura, K. and Sakano, H.
(1995) Cell, 83, 1113–1123.
33 Goodhardt, M., Cavelier, P., Akimenko, M.A., Lutfalla, G., Babinet, C. and
Rougeon, F.(1987) Proc. Natl. Acad. Sci. USA 84, 4229–4233.
34 Radomska, H.S., Shen, C-P., Kadesch, T. and Eckhardt, L.A. (1994)
Immunity, 1, 623–634.