Molecular Brain Research 72 Ž1999. 65–79
www.elsevier.comrlocaterbres
Research report
Identification of NFI-binding sites and cloning of NFI-cDNAs suggest a
regulatory role for NFI transcription factors in olfactory neuron gene
expression 1
Hans Baumeister a , Richard M. Gronostajski
b,c
, Gary E. Lyons d , Frank L. Margolis
e,)
a
Deutsches Institut fur Ernahrungsforschung,
Bergholz-Rehbrucke,
14558, Germany
¨
¨
Department of Cancer Biology, Research Institute, CleÕeland Clinic Foundation, CleÕeland, OH 44195, USA
c
Department of Biochemistry, Case Western ReserÕe UniÕersity, CleÕeland, OH 44106, USA
d
Department of Anatomy, UniÕersity of Wisconsin Medical School, Madison, WI 53706, USA
Department of Anatomy and Neurobiology, UniÕersity of Maryland School of Medicine, HSF 280, 685 West Baltimore Street, Baltimore, MD 21201,
USA
b
e
Accepted 22 June 1999
Abstract
Olfactory receptor neurons are responsible for the detection and signal transduction of odor ligands. Several genes associated with this
activity are preferentially or exclusively expressed in these neurons. Among these genes are those coding for olfactory receptors, adenylyl
cyclase type III, the cyclic nucleotide gated olfactory channel 1 ŽOcNC-1., Ga olf and the olfactory marker protein ŽOMP.. Promoter
analyses of these genes identified a binding site for the new transcription factor family OrE whose initial member, Olf-1, is abundantly
expressed in olfactory neurons. We report here that the proximal promoters of three of these genes, that are selectively expressed in
olfactory neurons, each contains a functional NFI binding site and that the sites have different affinities for NFI proteins indicating a
regulatory role for NFI proteins in olfactory gene expression. We further demonstrate, by cloning, that all four NFI genes are expressed in
the olfactory nasal mucosa. Analysis by in situ hybridization illustrates that at least three of these gene products are expressed in the
neuroepithelium in which the olfactory neurons reside. NFI proteins are capable of functioning as positive or negative regulators of
transcription depending on the tissue, cell-type, age, and gene in question. These multivalent functions of NFI could be achieved by
temporally and spatially regulated expression of distinct subsets of NFI isoforms. It now remains to characterize the tissue and cell
specific patterns of expression of distinct NFI transcription factors during ontogeny and their roles in regulating gene expression. q 1999
Elsevier Science B.V. All rights reserved.
Keywords: Upstream binding element; Olfactory gene promoter; Nuclear Factor I; Olfactory receptor neuron; Gene expression; Mobility shift assay
1. Introduction
AbbreÕiations: UBE, Upstream binding element; Cyclase III, Adenylyl cyclase type III; UCY, Upstream binding element in the cyclase III
gene; FIB, Consensus NFI binding site; mFIB, Mutated FIB; ORN,
Olfactory receptor neuron; oCNC-1, Olfactory cyclic nucleotide-gated
channel subunit 1; UC, Upstream binding element in the oCNC-1 gene;
OR, Olfactory receptor gene; OMP, Olfactory marker protein; NFI,
Nuclear factor I; Olf-1, Olf-1 transcription factor; EBF, Early B-cell
factor; NEM, N-ethyl maleimide; DTT, Dithiothreitol; EMSA, Electrophoretic mobility shift assay; Mash-1, Mouse homologue of the
achaete-scute gene
)
Corresponding author. Fax: q 1-410-706-2512; E-mail:
[email protected]
1
The sequences described here have been assigned Accession numbers
AF112455–AF112459.
The olfactory mucosa in the nasal vault contains the
olfactory neuroepithelium, a tissue that is highly specialized for odorant detection. Olfactory neurons in this stratified neuroepithelium express several genes, specifically or
predominantly, enabling them to respond to odorants and
to transduce this information into action potentials directed
to their terminals in the olfactory bulb. In mammals,
several of these genes encode the components of a cyclic
AMP-mediated pathway including 7TM-membrane receptors w10,67x, Ga olf , a stimulatory G-protein w35x, type III
0169-328Xr99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 9 - 3 2 8 X Ž 9 9 . 0 0 2 1 0 - 7
66
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
adenylyl cyclase w6x and the olfactory cyclic nucleotidegated ion channel 1 ŽOcNC-1. w19,37x. Together with
functional studies, these observations provide evidence for
a model w9,12,13,31,75x in which odorant binding to putative membrane receptors stimulates the synthesis of cAMP
that in turn activates the ion channel resulting in depolarization of the sensory neuron. The diversity of odorants is
matched by a family of up to a thousand receptor genes in
mammals w11x. However, a single olfactory neuron expresses just one or a small subset of receptor genes
w61,62,68x assigning an important role in olfactory coding
to the control of gene expression in olfactory neurons
w17,22,44,68,76,80x.
An additional novel characteristic of mature olfactory
neurons, that is distinct from virtually all other differentiated neurons, is their ability to be replaced in a process of
continuous turnover throughout adult life w8,26x. The key
to this process is a population of globose basal cells
located at the basal region of the neuroepithelium that act
as progenitor cells. These globose basal cells undergo
mitosis and subsequent migration in the apical direction
and progressive differentiation to mature olfactory neurons
replacing those which degenerate and die. This process is
associated with the induction of olfactory neuron-specific
gene expression w49,55,56,74,77x. Mature olfactory neurons are identified by the presence of the olfactory marker
protein ŽOMP. w50–52,77x, a 19-kDa cytoplasmic protein
thought to be a modulator of signal transduction w15x
although a recent report suggests it may participate in
regulating mitosis in olfactory neuronal precursors w20x.
Genes whose products are responsible for the differentiated properties of mature neurons, i.e., the olfactory receptors, Ga olf , type III cyclase, and OcNC-1, must be induced
during the course of this differentiation in a coordinated
fashion.
The molecular mechanisms underlying olfactory neuron-specific differentiation remains unknown. An important regulator may be the mouse achaete-scute homolog-1
ŽMash-1., a transcription factor that was localized to early
stages of the developing mouse olfactory neuroepithelium
w28x. Targeted deletion of the Mash-1 gene in mice resulted in a loss of olfactory neuron precursors w29x. But
forced expression of the Mash-1 gene in P19 embryonal
carcinoma cells is insufficient to drive their neuronal differentiation, although these cells express endogenous
Mash-1 when induced to differentiate to neurons by retinoic
acid w34x. Therefore, Mash-1 may be a necessary but not
sufficient factor for olfactory neuronal differentiation w34x.
A candidate to interact with Mash-1 is Olf-1, the first
member of the OrE family w80x of transcription factors of
the helix–loop–helix ŽHLH. family like Mash-1, that is
localized to the nuclei of immature and mature olfactory
neurons w79x. Olf-1 binding sites were identified in several
genes associated with the olfactory phenotype: OMP Žtwo
sites., Ga olf , type III cyclase, OcNC-1, 50.06, and 50.11
w43,78x. The gene encoding Olf-1 is additionally expressed
in early B-cells giving rise to the nearly identical transcription factor EBF Žearly B-cell factor. w30x. Mice with a
partial deletion of Olf-1rEBF are B-cell-deficient but do
not suffer from deficits in olfactory gene expression w48x.
This apparent discrepancy may reflect the presence of
multiple members of the OrE family in olfactory neurons
but not in the lymphoid system w80x or it may indicate that
additional factors are important for olfactory gene expression. One such factor may be Roaz, a zinc finger protein
that interacts with Olf-1rEBF to regulate gene expression
w76x.
In the present study, we demonstrate that members of
the NFI family of transcription factors are present in
olfactory neurons and that they bind specific sites in
olfactory specific gene promoters. The upstream binding
element ŽUBE. within the OMP gene promoter and two
sites ŽU-sites. within the OcNC-1 and the type III cyclase
gene promoters were described previously w43,78x. All
three sites were identified by footprint analyses but the
transcription factors that bind these sites remained unknown. The significance of the UBE site is emphasized by
the fact that it is conserved in the rat, mouse and human
OMP genes at nearly the same position w14x.
2. Materials and methods
2.1. Electrophoretic mobility shift assay (EMSA)
Nuclear extracts were prepared from the olfactory neuroepithelium of 3-week-old CD rats ŽCharles River Laboratories. as described previously w43x. Two microgram
protein of the nuclear extract Ž1 mgrml protein concentration. were incubated at room temperature for 15 min in a
binding buffer containing 25 mM HEPES pH 7.5, 150 mM
NaCl, 5 mM MgCl 2 , 4 mM DTT, 125 mgrml polyŽdI-dC.
ŽPharmacia, Upsala, Sweden., 0.5 mgrml BSA ŽBoehringer-Mannheim, Germany. and 0.5% Nonidet P-40. After the first incubation, samples were placed on ice, 1 ml of
the radiolabeled DNA Žabout 0.5 ng of DNA containing
100,000 c.p.m.rml. was added to a final volume of 20 ml
and the samples were incubated for further 15 min at room
temperature. After the second incubation, samples were
placed on ice and 3 ml of dye Ž0.25% bromophenol blue,
0.25% xylene cyanol FF, 30% glycerol. were added. The
samples were electrophoresed in a 5% nondenaturing
acrylamide gel containing 0.5 = TBE Ž44.5 mM Tris-Base,
44.5 mM Boric acid, 1 mM EDTA, pH 8.3. as gel and
running buffer. For autoradiography, the dried gels were
exposed at y808C for 2 h and overnight.
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
The DNA-probes were generated by annealing pairs of
synthetic oligonucleotides as below:
67
mobility pattern compared to an unirradiated protein standard.
2.3. RT-PCR
FIB and mFIB were previously described w24x. Oligonucleotides were purified by electrophoresis in a 20% acrylamide gel before annealing. DNA Ž40 ng. was labeled with
w a32 PxdCTP with the Klenow polymerase resulting in
100,000 c.p.m.r0.8 ng DNA ŽUBE., 100,000 c.p.m.r0.5
ng DNA ŽUCY. and 100,000 c.p.m.r0.4 ng DNA ŽFIB
and UC., respectively.
Supershift EMSAs were performed with 1 ml of the
rabbit antiserum or preimmune serum Ža8199. provided
by N. Tanese w72x. Antiserum or preimmune serum were
added to the samples containing the DNA probes and were
incubated for an additional 7 min at room temperature.
To analyse the binding activity of NFI-C220 by EMSA,
we used the condition described by Novak et al. w63x,
applying 5 ml of the enriched protein and about 1 ng DNA
labeled with 200,000 c.p.m., respectively. Fresh DTT and
BSA were added separately.
To analyse the NEM-sensitivity of the UBE-binding
activity, the nuclear extract was incubated with 10 mM
NEM for 10 min at 48C. Excess NEM was then inactivated
by addition of 50 mM DTT and incubation for 10 min
more at 48C. Samples treated with NEMrDTT were analysed by EMSA as described above except that 50,000
c.p.m.rreaction were used. The sensitivity to NEM was
also analysed by altering the order of addition of the
various components.
2.2. Cross-linking of DNA-protein complexes
Nuclear extract Ž10 mg. was incubated with 100,000
c.p.m. of the radiolabeled DNA probe as described for
EMSA. Competitors were added in 50-fold excess. For
cross-linking, the samples were exposed to short wave UV
light Ž256 nm. for 5 min at room temperature. All samples
were analysed on a 10% SDS polyacrylamide gel. 14 Clabeled rainbow colored proteins ŽAmersham, Arlington
Heights, USA. were used as molecular weight standards.
Irradiating the protein standard with UV light under the
same conditions as used for cross-linking did not alter their
To extract total RNA from olfactory neuroepithelium,
we used RNAzol according to the manufacturer’s description ŽAGS, Heidelberg, Germany.. Synthesis of first strand
cDNA was performed in the presence of 5 mg total RNA,
100 pmol dŽ N .6-primer ŽPharmacia., 1 = buffer Žsupplied
with the enzyme., 10 mM DTT, 0.5 mM dNTP Žeach
nucleotide, respectively. and 400 U of recombinant
Moloney Murine Leukemia Virus ŽMoMuLV. reverse transcriptase ŽSuperscript, BRL, Bethesda, USA. in a total
volume of 20 ml. After 1 h of incubation at 378C, the
enzyme was inactivated by incubation at 528C for 30 min
followed by ethanol precipitation. One-tenth Ž5 ml. of the
redissolved precipitate was used for PCR which was performed in the presence of 1 = buffer ŽPerkin-Elmer, USA.,
200 mM dNTP Žeach nucleotide, respectively., 2.5 U Taq
Polymerase ŽPerkin-Elmer. and 150 pmol of each primer
deg1 and deg2 Ždeg1: TTCCGGATGAŽGrA.TTŽCrT.CA ŽCrT .CCITT ŽCrT .AT ŽCrT .GA ŽGrA .GC, deg2:
AATCGATŽGrA.TGŽArG.TGŽCrTrG.GGCTGIAŽCrT.
GrA.CAIAG. in a final volume of 50 ml. These primers
were previously used to amplify fragments of NFI-cDNAs
from mouse w16x. The PCR amplification program was 1
min at 948C, 2 min at 508C and 2.5 min at 728C for 30
cycles. The same conditions were used to amplify NFI
cDNA fragments from 1 ug of plasmid DNA from a
cDNA library of rat olfactory neuroepithelium. The PCRproducts of 490 bp was isolated and cloned into pBluescript ŽStratagene. at the SmaI site. Cloned PCR-products
were analysed by automated nucleotide sequencing using
the DNA sequencer A373 ŽABI..
2.4. Cloning of NFI-A2
A portion of a rat olfactory epithelium cDNA library
Ž50,000 clones at 5000 clones per plate. subcloned into
pSport ŽBRL. was screened for full size cDNA-clones
encoding NFI of the A-type. As probe we used the NFI-A
cDNA fragment which was amplified by PCR from the
library. Replica filters were prehybridized in 50% formamide, 20 mM HEPES pH 7.0, 5 = SSC, 5 = Denhardt’s
solution, 0.1% SDS and 50 ugrml salmon sperm DNA.
For hybridization 50 ng of the NFI-A DNA fragment were
radiolabeled using the Ready-to-Go kit ŽPharmacia. which
resulted in 1.4 = 10 8 c.p.m.rmg. The radiolabeled DNA
fragment was added to 100 ml of prehybridization solution
Ž7 = 10 4 c.p.m.rml. and hybridization was carried out
over night at 428C. The replica filters were washed twice
for 5 min in 2 = SSC, 0.1% SDS at room temperature and
twice for 10 min in 0.2 = SSC, 0.1% SDS at 688C. After
autoradiography overnight at y808C, three positive signals
68
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
were detected, two of which were verified on second round
screening with the same probe. Nucleotide sequencing
confirmed that both clones contained the same cDNA
encoding NFI-A2.
2.5. In situ hybridization
In situ hybridization was performed exactly as previously described w16x. 35 S-labeled antisense probes were
selected from the 3X coding and non-coding regions of NFI
clones that differ between the four NFI transcripts.
3. Results
3.1. UBE, UC and UCY are binding sites for NFI transcription factors
On comparison of the nucleotide sequences of the three
UBE sites present in the human, mouse and rat OMP
promoter, respectively, with the nucleotide sequence of the
UC-site, described within the promoter of the rat olfactory
cyclic nucleotide-gated channel ŽOcNC. we found a common palindromic motif of 5X-CTGGŽ N .7-8CCAG-3X resembling the palindromic motif present in the NFI binding
site FIB ŽTable 1.. The UCY-site, described within the
promoter of the rat type III adenylyl cyclase, shares a 9-bp
identity with the UC-site and contains a palindromic motif
of 5X-GGCAŽ N .3TGCC-3X that is almost identical to the
consensus sequence for NFI binding sites ŽTable 1.. This
degree of homology prompted us to analyse whether NFI
proteins bind to UBE andror the U-sites.
The EMSA was used to analyse the binding activity of
nuclear proteins of rat olfactory neuroepithelium toward
synthetic double-stranded DNA probes containing the high
affinity NFI binding site FIB, the rat UBE, the UC and the
UCY sites, respectively ŽFigs. 1 and 2.. Nuclear proteins
of the olfactory neuroepithelium are able to bind FIB ŽFig.
1A. illustrating that NFI transcription factors are present
within this tissue. DNA-protein complexes formed with
FIB, UBE, UC and UCY, respectively, all migrate with the
same electrophoretic mobility. To confirm that NFI transcription factors bind to UBE, UC and UCY, we
performed EMSAs in the presence of a specific NFI
antiserum ŽFig. 1B., in the presence and absence of competitors ŽFig. 1C., and in the presence of NEM, an alkylating agent ŽFig. 2A.. The addition of unlabelled FIB totally
abolished binding to all the sites. By contrast, mFIB, the
mutated NFI binding site with a single base alteration
ŽTable 1 and Section 2. that is devoid of any NFI-binding
activity, was ineffective in altering binding to FIB, UBE,
UC or UCY. A polyclonal antiserum that recognizes the
Table 1
Sequence homologies between UBE, UC and UCY and the NFI-consensus sequence
a
UBE, UC and UCY were described elsewere.
Bold characters mark matching nucleotides between UBE, UC, UCY, FIB and the NFI-consensus sequence. The point mutation in mFIB is marked by an
arrow head.
c
The position refers to the transcription start site ŽUC and UCY. and the translation start site ŽUBE., respectively. The UCY-site is located within the
X
5 UTR on the noncoding DNA strand.
d
Footprint and EMSA analyses were published elsewhere w14,44,78x. FIB and mFIB do not represent genomic sites for protein binding w63x. The
NFI-consensus sequence was published w63x.
b
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
amino-terminal portion of NFI-CrCTF1, a human NFI
transcription factor, was analysed for its ability to form
69
supershifts with the UBE-, UC- or UCY-binding proteins
in rat olfactory epithelial extracts. The amino-terminal
region contains the DNA binding domain known to be
conserved among all NFI transcription factors so far identified w41x. The results of this experiment ŽFig. 1B. demonstrate that the NFI-antiserum specifically recognizes proteins bound to UBE, UC and UCY, respectively, resulting
in ternary complexes with reduced mobility Žsupershifts.
of the respective DNA, the bound proteinŽs. and the
antibody. Preimmune serum does not form any supershift,
nor is the antiserum able to bind on its own to any DNA
used in these experiments. Thus, the immunochemical
results and binding characteristics indicate that UBE-, UCand UCY-binding proteins belong to the family of NFI
transcription factors. Finally, the demonstration that the
Fig. 1. ŽA. Binding of rat olfactory epithelial nuclear proteins with
putative NFI-binding sites from genes expressed abundantly within the
olfactory epithelium. EMSAs were performed as described with 2 ug of
nuclear extract of rat olfactory epithelium and radiolabeled synthetic
DNA as probe Ž100,000 c.p.m. per binding reaction, for nucleotide
sequence, see Section 2.. Protein-bound DNA is visible in lanes 1–6. In
all cases, the band representing the rapidly migrating unbound DNA has
been trimmed from the Figures. Lane 1: binding to FIB containing the
NFI-binding site described previously w24x. Lanes 2–4: binding to putative NFI-binding sites, UC Žlane 2. contains the U-site from the OcNC
gene w78x, UBE Žlane 3. contains the UBE-site of the OMP gene w43x, and
UCY Žlane 4. contains the U-site from the adenylate cyclase type III
gene, respectively. To demonstrate binding to UBE and UCY, the same
gel was exposed for 2 h Žlanes 1–4. and overnight Žlanes 5 and 6.. As
negative controls, the binding reactions were performed in the absence of
nuclear proteins. Lane 7 represents the control experiment with UCY.
The same results were obtained with all four probes. ŽB. Recognition of
UBE-, UC-, and the UCY-binding proteins by an antiserum directed
against NFI protein. The binding reaction with the radiolabeled DNAprobe, the nuclear extract of rat olfactory epithelium and the NFI-antiserum or preimmune serum were performed as described in Section 2.
The location of protein-bound DNA Žbound DNA. and protein-bound
DNA including antibodies Žsupershift. are indicated. Lane 1: w32 PxFIBDNA incubated with nuclear extract of rat olfactory epithelium in the
absence of any serum; lane 2: same as lane 1 but in the presence of
NFI-antiserum; lane 3: same as lane 2 but with w32 PxUC-DNA as probe;
lane 4: same as lane 2 but with w32 PxUCY-DNA as probe; lane 5: same as
lane 2 but with w32 PxUBE-DNA as probe; lane 6: w32 PxFIB-DNA incubated with NFI-antiserum in the absence of nuclear extract; lane 7: same
as lane 1 but in the presence of preimmune serum. ŽC. Efficacy of FIB
and mFIB as competitors. EMSAs were performed as described for this
figure. FIB Žlanes 1–3., UC Žlanes 4–6., UBE Žlanes 7–9. and UCY
Žlanes 10–12., respectively. The competitors FIB or mFIB Ž2.5 ng or
about five-fold excess over the amount of probe present in each reaction.
were each added to the binding reactions presented in lanes 2, 5, 8 and 11
ŽFIB. and in lanes 3, 6, 9 and 12 ŽmFIB.. The DNA mFIB contains the
NFI-binding site with one point mutation Žsee Section 2. that abolishes
the ability of NFI to bind to this site. This property of mFIB is shown in
lanes 1–3: lane 1 shows binding of the nuclear proteins to w32 PxFIB
without competitor. mFIB has no effect on this binding activity Žlane 3.
but FIB as competitor reduces greatly the amount of protein-DNA
complexes with w32 PxFIB Žlane 2.. Only the bands containing the protein
bound DNA are shown. Lane 13 gives the result of one binding reaction
without nuclear extract as a negative control.
70
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
UBE-binding activity is sensitive to the thiol-specific alkylating reagent NEM ŽFig. 2A. offers additional strong
support that these observations reflect the presence of NFI,
as it is known that the site-specific DNA binding activity
of NFI is abolished by NEM w63x.
To obtain independent confirmation of these findings,
we evaluated the ability of a recombinant protein Žexpressed in Escherichia coli . representing the amino-terminal DNA-binding domain of NFI-CrCTF1 to bind UBE,
UC and UCY. Analysis by EMSA shows ŽFig. 2B. that
this protein forms complexes of identical electrophoretic
mobilities with FIB and with all three DNAs. The differing
intensities of the shifted complexes correspond to those
observed when nuclear proteins of the olfactory epithelium
were used.
The difference in the intensities of the shifted DNA-protein complexes shown in Fig. 1A indicated that NFI proteins bind to the three sites with different affinities. To
analyse this in more detail, we evaluated the abilities of
UBE, UC and UCY to compete in EMSA with FIB as
radiolabeled probe. The results ŽFig. 2C. demonstrate that
UC is most effective in competing with FIB for protein
binding and is as efficient as FIB competing with itself
Žresult not shown.. By contrast, UBE and UCY are at least
10-fold less effective as competitors compared to UC but
are quite similar to each other. This indicates that NFI
proteins bind UC with high affinity comparably to the
affinity of NFI proteins to its canonical binding site FIB.
UBE and UCY are bound by NFI proteins with lower but
very similar affinity.
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
3.2. NFI transcription factors are present within the olfactory neuroepithelium
The presence of NFI transcription factors within the
olfactory neuroepithelium is strongly indicated by the
FIB-binding activity of the nuclear extract shown in Fig.
1A. However, NFI comprises a family of transcription
factors consisting of four gene products NFI-A, -B, -C, and
-X and multiple alternative splice variants Žfor more detail,
see Section 4.. To determine whether one or several
isoforms of NFI proteins are present within the olfactory
neuroepithelium, we analysed the nuclear extract after UV
cross-linking with radiolabeled FIB-DNA as probe by
SDS-polyacrylamide gel electrophoresis ŽFig. 2D.. At least
four different DNA-protein complexes could be detected
with an apparent molecular weight of 36–40 kDa. Crosslinking with radiolabeled UBE-DNA resulted in detection
of two DNA-protein complexes of 40 and 43 kDa. The
specificity of these cross-linked products is demonstrated
by the ability of non-radioactively labeled FIB-DNA but
not mFIB-DNA, to compete and by the absence of any
cross-linked products in the absence of nuclear proteins.
Taken together, these results indicate, but do not prove,
that several NFI isoforms are present in this tissue that are
capable of binding to UBE.
Compelling evidence for the presence of NFI isoforms
in the olfactory neuroepithelium would derive from
demonstration of appropriate PCR products and isolation
and sequencing of NFI cDNA clones from this tissue.
Degenerate primers Žsee Section 2. that anneal within the
conserved, amino-terminal region of all NFI-cDNAs, were
used to identify and amplify transcripts of all four genes
known to encode NFI proteins. Analysis of RT-PCR products by electrophoresis in 1% agarose gels showed a single
band, slightly smaller than 500 bp, corresponding to the
expected size of 486 bp ŽFig. 3.. DNA products of the
same size could be amplified directly by PCR of a rat
71
olfactory neuroepithelium cDNA library. The RT-PCR
products could only be detected after reverse transcription
of RNA confirming the absence of contamination by genomic DNA.
Both the RT-PCR products Ž12 clones., and the PCRproducts directly amplified from the cDNA library Ž21
clones., were cloned into pBluescript and sequenced. These
33 clones revealed four different nucleotide sequences
ŽFig. 4. which were identified as NFI-like sequences,
because of their high homology with nucleotide sequences
of the corresponding NFI-cDNA fragments described in
chicken w40,71x and mouse ŽRef. w16x, Fig. 4 and Table 2..
One nucleotide sequence is identical with the corresponding sequence of the rat NFI encoding cDNA ŽrNFI-L.
cloned from liver, identifying it as an NFI-A type sequence. The identification of the remaining three NFI like
nucleotide sequences as products of the NFI-B, -C, or -X
gene, respectively, is based on sequence comparisons with
corresponding sequences of mouse ŽFig. 4. and chicken
NFI-cDNAs ŽTable 2.. The mouse and rat nucleotide
sequences are highly homologous Ž97% identical nucleotides for NFI-A, 96% for NFI-B, 95% for NFI-C, and
99% for NFI-X, Fig. 6.. In contrast to the high inter-species
Žmouse vs. rat. sequence identity for each NFI isoform the
nucleotide sequence homology between the four rat cDNA
fragments is significantly lower Ž79%–80%.. Comparison
with the chicken sequences ŽTable 2. revealed the highest
homology among the NFI-B sequences Ž91% identity..
Whereas the homology between the NFI-C, and -X sequences range between 83%–86% identity. In total, 33
cDNA clones were sequenced. Sequences of the NFI-A, B,
-C and -X type were present in 13, six, seven and seven
clones, respectively. In summary, the olfactory neuroepithelium of 3-week-old rats express all four NFI genes A,
B, C and X Žas they are classified in chicken and mouse..
Using the rNFI-A probe to screen a cDNA library from
the olfactory neuroepithelium of adult rats, we isolated two
Fig. 2. ŽA. Inactivation of the UBE binding activity by NEM. EMSA was performed as described except that 50,000 c.p.m. of radiolabeled DNA were
used per reaction. The order of addition of components for each reaction is indicated by numbers above each lane. Free sulfhydryl residues were
inactivated by incubation of the nuclear extract of rat olfactory epithelium with 10 mM NEM for 10 min at 48C. Excess NEM was then inactivated by
incubation with 50 mM DTT for 10 min at 48C before the UBE probe was added for further incubation Žlane e.. To ensure that the activity of NEM is
essential for this effect, we inactivated 10 mM NEM by pre-incubation with 50 mM DTT for 10 min at 48C before addition of the nuclear extract and the
UBE probe Žlane c.. Further, the inactivation by NEM occurred after the proteinŽs. were bound to DNA Žlanes d and f.. Lane a: w32 PxUBE probe alone;
lane b: nuclear extract with w32 PxUBE; lane c: NEM and DTT added prior to nuclear extract; lane d: NEM and DTT added after incubation of nuclear
extract with w32 PxUBE; lane f is lane d exposed longer; lane e: NEM and DTT added to the nuclear extract prior to w32 PxUBE. ŽB. Binding of recombinant
NF-I C220 to UC, UCY and UBE. EMSA was performed as described except that 5 ml of the NF-I C220 extract were incubated with 200,000 c.p.m. of the
radiolabeled DNA-probe at 48C for 30 min. Lanes 1–4: EMSA with NF-I C220 and w32 PxFIB, w32 PxUC, w32 PxUCY and w32 PxUBE, respectively. Lane 5:
w32 PxFIB without protein. The exposure time was 2 h Žlanes 1–3. or overnight Žlanes 4 and 5.. ŽC. Binding affinity of UC-, UCY- and UBE-binding
proteins. The binding affinity of NF-I to UC, UCY and UBE, respectively, was analysed by their ability to compete with w32 PxFIB as probe. EMSA was
performed as described with 2 ng of competitor DNA Žfive-fold excess, lanes 2, 4 and 6. or 40 ng of competitor DNA Ž100-fold excess, lanes 3, 5 and 7..
Lane 1: w32 PxFIB with the nuclear extract; lanes 2 and 3: with UC-DNA as competitor; lanes 4 and 5: with UCY-DNA as competitor; lanes 6 and 7: with
UBE-DNA as competitor; lane 8: w32 PxFIB alone in the absence of extract protein. ŽD. Cross-linking of UBE- and FIB-binding proteins. Nuclear extract
proteins Ž10 ug. were cross-linked with w32 PxFIB and w32 PxUBE Ž100,000 c.p.m.rreaction., respectively, Žsee Section 2.. The cross-linked products were
analysed on a 10% SDS-polyacrylamide gel. Competitors were present at 50-fold excess. The relative molecular weights of protein standards Žlanes M1
and M2, where M1 was treated as were a–h for cross-linking. are indicated. Lanes a–d, cross-linking with w32 PxUBE without nuclear extract Ža.; with
nuclear extract Žd.; with additional competitors FIB Žc. and mFIB Žb.. Lanes e–h, cross-linking with w32 PxFIB without nuclear extract Že.; with nuclear
extract Žh.; with additional competitors FIB Žg. and mFIB Žf.. Specific cross-linking products are marked with arrow heads.
72
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
clones representing the same rNFI-A cDNA of 3389 bp.
We designate the new cDNA rNFI-A2 because of its high
homology to mNFI-B2 ŽFig. 5., which is an NFI cDNA
isolated from mouse brain and classified as a product of
alternative splicing of the NFI gene A transcript w32x. The
open reading frame ŽORF. chosen corresponds to the ORF
described in NFI-B2 and predicts a protein of 532 amino
acids with a calculated molecular weight of 58.6 kDa. The
amino acid sequence derived from NFI-A2 contains within
the amino-terminal region the highly negatively charged
motif as well as the four conserved cysteines found in all
NFI proteins. The carboxy-terminal 100 amino acids are
very proline-rich Ž24%. including one stretch of seven
prolines. The proline-rich C-terminus of CTF-1, a human
homolog of NFI-C, was reported to activate transcription
w38,57,82x. The nucleotide sequence of rNFI-L is identical
to rNFI-A2 ŽrNFI-A2 is about 1.7 kb longer than rNFI-L.
except for a 16-bp oligonucleotide at the 5X end of rNFI-L
where there is only about 20% nucleotide identity. The
break point matches exactly with the location of an exon–
intron boundary described in the rat and porcine genes
w3,54x and with alternative splicing products described in
chicken and mouse w16,41x. Thus, rNFI-A2 seems to be
another splicing variant of rNFI-L derived from the NFI-A
gene.
3.3. NFI transcripts are present in olfactory receptor
neurons
Our ability to isolate transcripts of all four of the NFI
genes demonstrates their presence in olfactory mucosa but
does not prove their presence in ORNs. To confirm the
presence of NFI transcripts in ORNs, in situ hybridization
was performed on prenatal nasal tissue at embryonic day
16.5 ŽFig. 6.. Transcripts of all four NFI genes, A, B, X
and C are all present in the nasal mucosa ŽFig. 6.. The
most abundant expression is observed in nasal glandular
tissue, a site of active protein synthesis and secretion. At
this age, transcripts of NFI-A, -B and -X are clearly
associated with the olfactory neuroepithelium where the
ORNs reside. Transcripts of all four NFI genes are expressed in olfactory neurons, although with differing developmental profiles ŽBehrens et al., ms. submitted.. These
observations provide further support for the role of NFI in
olfactory neuron gene transcription.
4. Discussion
Fig. 3. RT-PCR with primers for NFI encoding cDNAs. Total RNA Ž5
ug. isolated from olfactory epithelium of 3-week-old rats was reverse
transcribed with random hexamers as primers. The first strand cDNA
ŽssDNA. as well as double-stranded cDNA ŽdsDNA. from an olfactory
epithelial cDNA-library were used as templates for PCR with degenerate
primers. These primers were designed from a conserved 486 bp domain
common to all four NFI encoding Žsee Section 2.. The PCR-products
were analysed on a 2% agarose gel containing ethidium bromide. Lanes 1
and 2: PCR-products derived from the dsDNA Ž1 ug. of the library as
template w1x, from the ssDNA Ž10% of the synthesized ssDNA. as
template ŽRT-PCR, 2.. Lane 3: to ensure that no genomic DNA was
amplified, the RNA that was used for RT-PCR was also used for PCR
directly in the absence of RT Ž0.5 ug.. Lane 4: no nucleic acids added.
DNA-standard ŽM.: 1 kb ladder ŽBoehringer-Mannheim. with fragments
of 506 and 396 bp marked.
We report here Ž1. the identification of three functional
binding sites of NFI transcription factors in three distinct
promoters that drive gene expression specifically or preferentially in olfactory neurons; Ž2. expression of four NFI
genes within the rat olfactory epithelium; and Ž3. cloning
of a new full length rat NFI cDNA from a rat olfactory
epithelium cDNA library. Together with our preliminary in
situ localization data, these results suggest that NFI transcription factors participate in olfactory neuron gene expression.
Genes that are preferentially, or even specifically, expressed within the olfactory neuroepithelium include the
putative olfactory receptors w10,67x, the GTP binding protein Ga olf w35x, the adenylyl cyclase type III w6x, the
OcNC-1 w19,37x, the OMP w50–52,77x, and additional genes
of unknown function Ž50.06, 50.11. w78x. Promoter analyses of six of the rat genes by footprint assays and EMSA
revealed four protein binding sites that might function as
cis-regulatory elements in olfactory gene expression. These
sites are: Ž1. Olf-1rEBF Žpresent in the Ga olf , type III
cyclase, OcNC, OMP, 50.06, and 50.11 genes., the bind-
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
73
Fig. 4. Expression of NFI genes in rat olfactory epithelium. RT-PCR products and PCR products derived from the cDNA-library were subcloned into
pBluescript and sequenced. Of 12 clones derived from RT-PCR, six contained the sequence designated as rNFI-X Žsee text., while two each contained the
sequences designated as rNFI-A, -B and -C. Of 21 clones derived from the cDNA-library 11, 5, 4 and 1 clones contained the sequences designated as
rNFI-A, -C, -B and -X, respectively. The sequence of each rat isoform was aligned with the corresponding mouse NFI sequence w16x. For each isoform,
nucleotides that are not conserved between mouse and rat are in bold.
ing site of the transcription factor Olf-1rEBF, that has
been confirmed as a transcriptional activator w30,43,78–80x;
Ž2. the UBE localized in the OMP gene w14,43x, Ž3. the
U-site localized in the OcN-channel gene ŽUC. w78x; and
Ž4. the U-site localized in the adenylyl cyclase gene ŽUCY.
w78x. The identity of the proteins that interact at the U and
UBE sites, and the function of these sites was unknown.
We have now demonstrated that these sites bind to members of the ubiquitously expressed NFI family of transcriptional regulators. Further, we illustrate that these binding
proteins are members of a complex set since distinct
DNA-protein complexes are formed with nuclear extracts
from different tissues w43,78x.
The UBE site is conserved in nucleotide sequence and
position in the promoters of the cloned human, rat and
mouse OMP genes ŽTable 1.. Furthermore, sequence comparison of these and the rat U-sites identified a homologous, palindromic motif that corresponds to a consensus
sequence for binding sites of NFI transcription factors
ŽTable 1.. Five lines of evidence from this study converge
to confirm that the UBE, UC, and UCY-binding proteins
of the rat olfactory neuroepithelium are indeed members of
the NFI family of transcription factors. They are: Ž1. the
DNA-protein complexes formed with FIB-DNA, containing a previously described NFI-binding site ŽTable 1., and
with UBE, UC and UCY-DNA, migrate identically in a
native EMSA gel ŽFig. 1A.; Ž2. an antiserum directed
Table 2
Comparison of nucleotide sequences of NFI-cDNA fragments derived
from chicken Žc-. and rat Žr-.
% Identity a
Rat
a
rNFI-A
rNFI-B
rNFI-C
rNFI-X
Chicken
cNFI-A
cNFI-B
cNFI-C
cNFI-X
85
76
74
79
78
91
77
79
78
79
85
86
77
78
83
85
The nucleotide sequences shown in Fig. 4 were compared using the
Pearson and Lipman algorithm which also provides the percentage of
identical nucleotides.
74
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
against NFI specifically recognizes the DNA-protein complex formed between olfactory tissue extracts and FIB,
UBE, UC, and UCY-DNA ŽFig. 1B.; Ž3. FIB-DNA, but
not the mutated mFIB DNA, is a very efficient competitor
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
75
Fig. 6. Localization of NFI transcripts in E16.5 rat nasal mucosa by in situ hybridization. Micrographs of parasagittal sections of nasal mucosa hybridized
with 35 S-antisense probes to NFI-A, NFI-B, NFI-X and NFI-C. Strong signal is seen in the nasal glands Žg. and in the olfactory mucosa Žom., the site of
the olfactory neuroepithelium where olfactory receptor neurons reside. Signal is also evident in the olfactory bulb Žb. that is separated from the nasal side
by the portion of the skull identified as the cribriform plate Žcp.. The scale bar is 300 mm. At this age, NFI-B expression seems most robust in the
neuroepithelium.
for UBE, UC, and UCY-binding proteins ŽFig. 1C.; Ž4. the
UBE-binding activity is NEM sensitive ŽFig. 2A., a characteristic of NFI proteins w63x; and Ž5. the recombinant
DNA-binding domain of the human NFIrCTF expressed
in E. coli interacts with FIB, UBE, UC, and UCY-DNA in
EMSA to give shifts of indistinguishable electrophoretic
mobility ŽFig. 2B..
These results demonstrate that all three sites, UBE, UC,
and UCY, are recognized by NFI proteins. The role this
transcription factor plays in regulating the expression of
the corresponding genes ŽOMP, OcNC-1, and type III
cyclase. remains to be elucidated. Recently, Glusman et al.
w22x reported the presence of an NFI binding motif ŽTable
1. in the putative promoter region of an olfactory receptor
gene. The fact that four promoters driving olfactory neuron
specific gene expression contain an NFI binding site suggests that NFI is an important regulator of olfactory neurons gene expression. All three sites, UBE, UCY, and UC
contain a CCAG motif, while in many NFI-binding sites, a
CCAA motif is conserved w36x. This latter motif is the
basis for the alternative identification of NFI as CAAT
binding transcription factor ŽCTF. w72x. Recently, a slow
cAMP response element was localized to an NFI-binding
site ŽTGGGCGCCTTGCCAG. w83x that also contains the
CCAG motif and that is almost identical Ž12 of 15 nucleotides. with the UCY-site. This is intriguing because
cAMP plays a major role as a second messenger in olfactory signal transduction w9,31x, suggesting that it could be
involved in regulating gene expression of some signal
transduction components Žtype III cyclase and OcNC-1.
distinct from the well-documented transcriptional regulation by the cAMP response element CRE.
NFI proteins bind the three U-sites with varying affinities ŽFig. 1A,C.. Evaluation of UBE, UC, and UCY-DNA,
as competitors for NFI binding to FIB-DNA ŽFig. 2C.
demonstrated that UC has the highest affinity comparable
Fig. 5. Nucleotide sequence of the full length rat cDNA NFI-A2. The nucleotide sequence of rNFI-A2 was generated by independently sequencing two
identical clones obtained by screening a rat olfactory neuroepithelial cDNA-library Žsee Section 2.. The homology between the nucleotide sequences of
rNFI-A2, rNFI-L, and mNFI-B2 is illustrated and identical nucleotides are marked by a dot. Three amino acid exchanges between mNFI-B2 and rNFI-A2
are indicated above the translated rNFI-A2 ORF. The positions of the degenerate primers used for PCR are underlined. Note that although long polyA runs
X
X
X
are present at the 3 end no appropriate polyadenylation signal sequence was found indicating that this 3 end may not represent the 3 end of the
corresponding mRNA.
76
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
to FIB in its binding strength Ždata not shown.. In contrast,
UBE and UCY are of lower affinity by about a factor of
10. The binding affinity of NFI proteins to different sites
are reported to range from 10 8 to 10 11 My1 w53x. A
well-known site of relative low affinity w54,64x is the half
site of the canonical NFI binding site, TGGCA, that has
been shown to be an active cis-regulating element in
several promoters w65,69,71x. Nevertheless, confirmation
of the activity of the three U-sites as functional cis-regulating elements in olfactory gene expression remains to be
elucidated. Interestingly, the high affinity site UC and the
low affinity site UCY differ by just 1 bp from the NFI
consensus sequence. In the case of UCY, a contact point
for NFI binding w18x, the first T of the consensus sequence,
is replaced by a G. In UC, the fourth bp is exchanged as is
common in NFI binding sites w18x. All three UBE sites
contain a stretched linker sequence Ž4 instead of 3 bp.
compared to the other NFI binding sites presented here
ŽTable 1.. Gronostajski w27x reported a slightly different
consensus sequence ŽTGGGCrANCrGNNGrTGCCAA.
for NFI binding sites Žin HeLa cell extracts. with a stretched
linker that matches very well with all three UBE sequences. In agreement with our results, these sites were
found to have a lower binding affinity to NFI proteins.
NFI is a family of structurally related proteins that are
involved in regulation of viral replication w23,58,59x and
transcription of certain viral w4,46x and cellular genes
w21,25,32,47,69,70,83x. The diversity of the NFI family
was first discovered in chicken when cloning of several
NFI encoding cDNAs indicated the presence of four closely
related but distinct genes cNFI-A, cNFI-B, cNFI-C, and
cNFI-X w40,71x and alternative splicing products of each
gene w5,41x. Additionally, post-translational modification
by glycosylation w33x and oxidation w7,39x as well as
heterodimer formation between individual members of the
family provides mechanisms for further modulating the
activities of these DNA-binding proteins w42x. The NFI
encoding cDNAs cloned from frog w66x, porcine w54x,
mouse w16,32x, hamster w21x, rat w65x and human w5,72,81x
libraries give rise to corresponding genes and splice variants between vertebrate species w41x. Sequence comparison
of NFI encoding cDNAs revealed a highly conserved
region of about 600 bp that is located 5X to very divergent
sequences that are characteristic for each gene and splicing
variant w71x. The conserved region codes for the aminoterminal domains within NFI proteins that are necessary
for DNA-binding, regulation of the viral replication and
protein dimerization w23x, while some but not all NFI
proteins contain a carboxy-terminal proline-rich region that
is considered to contain the transcriptional activator domain w2,38,57,82x. Although NFI expression is considered
to be ubiquitous, RNA blot, RNase protection assays, in
situ hybridization and biochemical analyses in mouse, rat
and human suggest a tissue, cell-type and developmentally
regulated distribution of distinct NFI transcripts and proteins w16,32,45,65x. Indeed such a distribution seems essen-
tial to explain the role NFI proteins play in regulating the
tissue and cell-type specific expression of several target
genes w1,5,25,32,45,47,73x.
Electrophoretic analysis of UV-cross-linked products
formed between the radioactive FIB-DNA Žcontaining one
NFI binding site. and nuclear proteins of the rat olfactory
neuroepithelium revealed the presence of multiple NFI
proteins within this tissue ŽFig. 2D.. This was confirmed
when we analysed the number and identity of NFI genes
that are expressed in the rat olfactory neuroepithelium.
Using both the RT-PCR and conventional PCR techniques
with degenerate primers, we amplified cDNA-fragments
from common to the conserved region of all NFI cDNAs.
Cloning and sequencing of the PCR products revealed four
distinct nucleotide sequences of 486 bp ŽFig. 4.. Identities
ranging from 74%–91% and 95%–99% compared to the
corresponding regions of the cNFI genes ŽTable 2. and
mNFI cDNAs, respectively, prove that all four genes,
NFI-A, -B, -C and -X, are expressed within the rat olfactory epithelium.
In our search for rNFI encoding cDNA fragments, we
mostly detected sequences of the NFI-A type suggesting
that among the four NFI genes, the NFI-A gene is predominantly expressed in the rat olfactory neuroepithelium.
Interestingly, Osada et al. w64x reported a DNA-binding
site for expressed mouse NFI-A that resembles the UBE
and U-sites. Therefore, we screened a cDNA library derived from rat olfactory epithelium for a full length cDNA
using the PCR product identified as a fragment of a
rNFI-A cDNA. Two identical cDNA clones were obtained
of 3389 bp ŽFig. 5. containing an ORF of 1596 bp, and 5X
and 3X untranslated regions ŽUTR. of 195 and 1595 bp.
This represents the longest NFI encoding cDNA so far
identified. The ORF encodes for a protein of 532 amino
acids. Both, the nucleotide sequence and the amino acid
sequence are highly homologous with NFI encoding sequences ŽFig. 5.: The sequence of the rNFI-L cDNA Ž1712
bp. is identical over 1689 bp and the sequence of mNFI-B2
w32x, one of six cDNA clones from mouse brain, shows
100, 97.7, and 92.4% identity to the 5X UTR, ORF, and
3X UTR, respectively. This includes AT rich sequences
within the two 3X UTRs that have been suggested as signals
for controlled degradation of NFI transcripts w60x. The
amino acid sequences derived from the two cDNAs are
identical except for three residues ŽFig. 5.. The high
sequence identity with mNFI-B2 indicates that we have
cloned the rat homologue of the mNFI-B2 splicing product
of the NFI-A gene, which we designate rNFI-A2. rNFI-L,
that lacks the 5X UTR and the 5X end of the ORF w65x,
differs from rNFI-A2 by 23 nucleotides at the 5X end of
rNFI-L. The first amino acid Žaspartate. that is encoded by
both sequences matches exactly with the first amino acid
encoded by exon 2 of the rNFI-L gene w3x, indicating that
rNFI-A2 transcripts contain an alternative exon 1 compared to rNFI-L. Thus, the NFI-A gene is expressed in the
olfactory epithelium as well as in liver but distinct tran-
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
scripts are generated in each tissue by the mechanism of
alternative splicing and possibly by the use of alternate
promoters. Differential use of two promoters could explain
a tissue specific distribution of NFI transcripts. This results
in a protein that is 24 residues longer at the amino-terminus
than rNFI-L. The functional difference, if any, between the
two proteins derived from rNFI-A2 and rNFI-L is unknown.
In summary, we have shown that the proximal promoters of three genes that are selectively expressed in olfactory neurons each contain an NFI binding site and that NFI
proteins are present in olfactory tissue extracts. We have
demonstrated, by cloning and in situ hybridization, that all
four NFI genes are expressed in olfactory tissue and that at
least three of these are expressed in the neuroepithelium in
which the olfactory neurons reside. The functional relationships in olfactory neuron gene expression among these
elements remains to be elucidated. NFI proteins are capable of functioning as positive or negative regulators of
transcription depending on the tissue, cell-type, age, and
gene investigated w1,16,32,84x. These multivalent functions
of NFI could be achieved by temporally and spatially
regulated expression of distinct subsets of NFI isoforms as
recently demonstrated w16x. All four NFI genes are expressed in the olfactory neuroepithelium. It now remains to
characterize the tissue and cell specific patterns of expression of distinct NFI transcription factors during ontogeny
and their roles in regulating gene expression. In the olfactory epithelium NFI binding to the U-sites could modulate
the function of Olf-1 as a transcriptional activator or could
be involved in induced regulation of transcription by cAMP
or the growth factor TGFb w69,78x.
w6x
w7x
w8x
w9x
w10x
w11x
w12x
w13x
w14x
w15x
w16x
w17x
w18x
Acknowledgements
We thank N. Tanese for the generous gift of anti-NFI
antiserum, R. Wurzburger for DNA sequencing and Maik
Behrens for valuable discussions. Supported in part by
HD34908 to RMG and DC03112 to FLM.
References
w1x A.D. Adams, D.M. Choate, M.A. Thompson, NF1-L is the DNA-binding component of the protein complex at the peripherin negative
regulatory element, J. Biol. Chem. 270 Ž1995. 6975–6983.
w2x H. Altmann, W. Wendler, E.-L. Winnacker, Transcriptional activation by CTF proteins is mediated by a bipartite low-proline domain,
Proc. Natl. Acad. Sci. U.S.A. 91 Ž1994. 3901–3905.
w3x R. Ammendola, F. Gounari, G. Piaggio, V. De Simone, R. Cortese,
Transcription of the promoter of the rat NF-1 gene depends on the
integrity of an Sp1 recognition site, Mol. Cell. Biol. 10 Ž1990.
387–390.
w4x D. Apt, T. Chong, Y. Liu, H.-U. Bernard, Nuclear factor 1 and
epithelial cell-specific transcription of human papillomavirus type
16, J. Virol. 67 Ž1993. 4455–4463.
w5x D. Apt, Y. Liu, H.-U. Bernard, Cloning and functional analysis of
w19x
w20x
w21x
w22x
w23x
w24x
w25x
77
spliced isoforms of human nuclear factor I-X: interference with
transcriptional activation by NFIrCTF in a cell-type specific manner, Nucleic Acids Res. 22 Ž1994. 3825–3833.
H.A. Bakalyar, R.R. Reed, Identification of a specialized adenylyl
cyclase that may mediate odorant detection, Science 250 Ž1990.
1403–1406.
S. Bandyopadhyay, R.M. Gronostajski, Identification of a conserved
oxidation-sensitive cysteine residue in the NFI family of DNA-binding proteins, J. Biol. Chem. 269 Ž1994. 29949–29955.
W. Breipohl ŽEd.., Ontogeny of Olfaction, Springer, Berlin, 1986.
L.J. Brunet, G.H. Gold, J. Ngai, General anosmia caused by a
targeted disruption of the mouse olfactory cyclic nucleotide-gated
cation channel, Neuron 17 Ž1996. 681–693.
L. Buck, R. Axel, A novel multigene family may encode odorant
receptors: a molecular basis for odorant recognition, Cell 66 Ž1991.
175–187.
L. Buck, Identification and analysis of a multigene family encoding
odorant receptors: implications for mechanisms underlying olfactory
information processing, Chem. Sens. 18 Ž1993. 203–208.
L.B. Buck, S. Firestein, R.F. Margolskee, Olfaction and taste in
vertebrate: molecular and organizational strategies underlying
chemosensory perception, in: G.J. Siegel, B.W. Agranoff, W. Albers, P.B. Molinoff ŽEds.., Basic Neurochemistry, Raven Press, New
York, 1994, pp. 157–177.
L.B. Buck, Information coding in the vertebrate olfactory system,
Annu. Rev. Neurosci. 19 Ž1996. 517–544.
O.I. Buiakova, N.S.R. Krishna, T.V. Getchell, F.L. Margolis, Human and rodent OMP genes: conservation of structural and regulatory motifs and cellular localization, Genomics 20 Ž1994. 452–462.
O.I. Buiakova, H. Baker, J.W. Scott, A. Farbman, R. Kream, M.
Grillo, L. Franzen, M. Richman, L.M. Davis, S. Abbondanzo, C.L.
Stewart, F.L. Margolis, Olfactory marker protein ŽOMP. gene deletion alters physiological activity of olfactory neurons, Proc. Natl.
Acad. Sci. U.S.A. 93 Ž1996. 9858–9863.
A.Z. Chaudhry, G.E. Lyons, R.M. Gronostajski, Expression patterns
of the four nuclear factor I genes during mouse embryogenesis
indicate a potential role in development, Dev. Dyn. 208 Ž1997.
313–325.
A. Chess, I. Simon, H. Cedar, R. Axel, Allelic inactivation regulates
olfactory receptor gene expression, Cell 78 Ž1994. 823–834.
E. De Vries, W. van Driel, S.J.L. van den Heuvel, P.C. van der
Vliet, Contactpoint analysis of the HeLa nuclear factor I recognition
site reveals symmetrical binding at one side of the DNA helix,
EMBO J. 6 Ž1987. 161–168.
R.S. Dhallan, K.-W. Yau, K.A. Schrader, R.R. Reed, Primary
structure and functional expression of a cyclic nucleotide-activated
channel from olfactory neurons, Nature 347 Ž1990. 184–187.
A.I. Farbman, J.A. Buchholz, E. Walters, F.L. Margolis, Does
olfactory marker protein participate in olfactory neurogenesis?, Ann.
N.Y. Acad. Sci. 855 Ž1998. 248–251.
G. Gil, J.R. Smith, J.L. Goldstein, C.A. Slaughter, K. Orth, M.S.
Brown, T.F. Osborne, Multiple genes encode nuclear factor 1-like
proteins that bind to the promoter for 3-hydroxy-3-methylglutarylcoenzyme A reductase, Proc. Natl. Acad. Sci. U.S.A. 85 Ž1988.
8963–8967.
G. Glusman, S. Clifton, B. Roe, D. Lancet, Sequence analysis in the
olfactory receptor gene cluster on human chromosome 17: recombinatorial events affecting receptor diversity, Genomics 37 Ž1996.
147–160.
F. Gounari, R. De Francesco, J. Schmitt, P.C. van der Vliet, R.
Cortese, H. Stunnenberg, Amino-terminal domain of NFI binds to
DNA as a dimer and activates adenovirus DNA replication, EMBO
J. 9 Ž1990. 559–566.
N. Goyal, J. Knox, R.M. Gronostajski, Analysis of multiple forms of
nuclear factor I in human and murine cell lines, Mol. Cell. Biol. 10
Ž1990. 1041–1048.
R.A. Graves, P. Tontonoz, S.R. Ross, B.M. Spiegelman, Identifica-
78
w26x
w27x
w28x
w29x
w30x
w31x
w32x
w33x
w34x
w35x
w36x
w37x
w38x
w39x
w40x
w41x
w42x
w43x
w44x
w45x
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
tion of a potent adipocyte-specific enhancer: involvement of an
NF-1-like factor, Genes Dev. 5 Ž1991. 428–437.
P.P.C. Graziadei, G.A. Monti-Graziadei, Continuous nerve cell renewal in the olfactory system, in: M. Jacobson ŽEd.., Handbook of
Sensory Physiology: IX. Development of Sensory Systems, Springer,
Berlin, 1978, pp. 55–83.
R.M. Gronostajski, Site-specific DNA binding of nuclear factor I:
effect of the spacer region, Nucleic Acids Res. 15 Ž1987. 5545–5559.
F. Guillemot, A.L. Joyner, Dynamic expression of the murine
achaete-scute homologue Mash-1 in the developing nervous system,
Mech. Dev. 42 Ž1993. 171–185.
F. Guillemot, L.C. Lo, J.E. Johnson, A. Auerbach, A.L. Joyner,
Mammalian achaete-scute homolog 1 is required for the early
development of olfactory and autonomic neurons, Cell 75 Ž1993.
463–476.
J. Hagman, C. Belanger, A. Travis, W. Turck, R. Grosschedl,
Cloning and functional characterization of early B-cell factor, a
regulatory of lymphocyte-specific gene expression, Genes Dev. 7
Ž1993. 760–773.
J.G. Hildebrand, G.M. Shepherd, Mechanisms of olfactory discrimination: converging evidence for common principles across phyla,
Annu. Rev. Neurosci. 20 Ž1997. 595–631.
T. Inoue, T. Tamura, T. Furuichi, K. Mikoshiba, Isolation of complementary DNAs encoding a cerebellum-enriched nuclear factor I
family that activates transcription from mouse Myelin basic protein
promoter, J. Biol. Chem. 265 Ž1990. 19065–19070.
S.P. Jackson, R. Tjian, O-glycosation of eukariotic transcription
factors: implications of mechanisms of transcriptional regulation,
Cell 55 Ž1988. 125–133.
J.E. Johnson, K. Zimmerman, T. Saito, D.J. Anderson, Induction
and repression of mammalian achaete-scute homolog ŽMASH. gene
expression during neuronal differentiation of P19 embryonal carcinoma cells, Development 114 Ž1992. 75–87.
D.T. Jones, R.R. Reed, Golf : an olfactory neuron-specific G protein
involved in odorant signal transduction, Science 244 Ž1989. 790–795.
K.A. Jones, J.T. Kadonaga, P.J. Rosenfeld, T.J. Kelly, R. Tjian, A
cellular DNA-binding protein that activates eukariotic transcription
and DNA replication, Cell 48 Ž1987. 79–89.
U.B. Kaupp, The cyclic nucleotide-gated channels of vertebrate
photoreception and olfactory epithelium, Trends Neurosci. 14 Ž1991.
150–157.
T.K. Kim, R.G. Roeder, Proline-rich activator CTF1 targets the
TFIIB assembly step during transcriptional activation, Proc. Natl.
Acad. Sci. U.S.A. 91 Ž1994. 4170–4174.
L. Knoepfel, C. Steinkuhler,
M.-T. Carrı,
¨
` G. Rotilio, Role of zinccoordination and of the glutathione redox couple in the redox
susceptibility of human transcription factor Sp1, Biochem. Biophys.
Res. Commun. 201 Ž1994. 871–877.
U. Kruse, F. Qian, A.E. Sippel, Identification of a forth nuclear
factor I gene in chicken by cDNA cloning: NFI-X, Nucleic Acids
Res. 19 Ž1991. 6641.
U. Kruse, A.E. Sippel, The genes for transcription factor nuclear
factor I give rise to corresponding splice variants between vertebrate
species, J. Mol. Biol. 238 Ž1994. 860–865.
U. Kruse, A.E. Sippel, Transcription factor nuclear factor I proteins
form stable homo- and heterodimers, FEBS Lett. 348 Ž1994. 46–50.
K. Kudrycki, C. Stein-Izsak, C. Behn, M. Grillo, R. Akeson, F.L.
Margolis, Olf-1-binding site: characterization of an olfactory
neuron-specific promoter motif, Mol. Cell. Biol. 13 Ž1993. 3002–
3014.
K. Kudrycki, O.I. Buiakova, G. Tarazzo, E. Walters, F.L. Margolis,
Effects of mutation of the Olf-1 motif on transgene expression in
olfactory receptor neurons, J. Neurosci. Res. 52 Ž1998. 159–172.
S. Kulkarni, R.M. Gronostajski, Altered expression of the developmentally regulated NFI gene family during phorbol ester-induced
differentiation of human leukemic cells, Cell Growth Differ. 7
Ž1996. 501–510.
w46x K.U. Kumar, A. Pater, M.M. Pater, Human JC virus perfect palindromic nuclear factor 1-binding sequences important for glial cellspecific expression in differentiating embryonal carcinoma cells, J.
Virol. 67 Ž1993. 572–576.
w47x S. Li, J.M. Rosen, Nuclear factor I and mammary gland factor
ŽSTAT5. play a critical role in regulating rat whey acidic protein
gene expression in transgenic mice, Mol. Cell. Biol. 15 Ž1995.
2063–2070.
w48x H. Lin, R. Grosschedl, Failure of B-cell differentiation in mice
lacking the transcription factor EBF, Nature 376 Ž1995. 263–267.
w49x T. Margalit, D. Lancet, Expression of olfactory receptor and transduction genes during rat development, Dev. Brain Res. 73 Ž1993.
7–16.
w50x F.L. Margolis, A brain protein unique to the olfactory bulb, Proc.
Natl. Acad. Sci. U.S.A. 69 Ž1972. 1221–1224.
w51x F.L. Margolis, A marker protein for the olfactory chemo-receptor
neuron, in: R.A. Bradshaw, D.M. Schneider ŽEds.., Proteins of the
Nervous System, Raven Press, New York, 1980, pp. 59–84.
w52x F.L. Margolis, Molecular cloning of olfactory specific gene products, in: F.L. Margolis, T.V. Getchell ŽEds.., Molecular Neurobiology of the Olfactory System, Plenum, New York, 1988, pp. 237–265.
w53x M. Meisterernst, I. Gander, L. Rogge, E.-L. Winnacker, A quantitative analysis of nuclear factor IrDNA interactions, Nucleic Acids
Res. 16 Ž1988. 4419–4435.
w54x M. Meisterernst, L. Rogge, R. Foeckler, M. Karaghiosoff, E.-L.
Winnacker, Structural and functional organization of a porcine gene
coding for nuclear factor I, Biochemistry 28 Ž1989. 8191–8200.
w55x B.P. Menco, F.D. Tekula, A.L. Farbman, W. Danho, Developmental
expression of G-proteins and adenylyl cyclase in peripheral olfactory
systems. Light microscopic and freeze-substitution electron microscopic immunocytochemistry, J. Neurocytol. 23 Ž1994. 708–727.
w56x B.P.M. Menco, Ultrastructural aspects of olfactory signaling, Chem.
Sens. 22 Ž1997. 295–311.
w57x N. Mermod, E.A. O’Neill, T.J. Kelly, R. Tjian, The proline-rich
transcriptional activator of CTFrNF-I is distinct from the replication
and DNA binding domain, Cell 58 Ž1989. 741–753.
w58x A. Monaghan, A. Webster, R.T. Hay, Adenovirus DNA binding
protein: helix destabilising properties, Nucleic Acids Res. 22 Ž1994.
742–748.
w59x K. Nagata, R.A. Guggenheim, T. Enomoto, J.H. Lichy, J. Hurwitz,
Adenovirus DNA replication in vitro: identification of a host factor
that stimulates synthesis of the preterminal protein-dCMP complex,
Proc. Natl. Acad. Sci. U.S.A. 79 Ž1982. 6438–6442.
w60x G. Nebl, N. Mermod, A.C.B. Cato, Post-transcriptional down-regulation of expression of transcription factor NF1 by Ha-ras oncogene, J. Biol. Chem. 269 Ž1994. 7371–7378.
w61x P. Nef, I. Hermans-Borgmeyer, H. Artieres-Pin, L. Beasley, V.E.
Dionne, S.F. Heinemann, Spatial pattern of receptor expression in
the olfactory epithelium, Proc. Natl. Acad. Sci. U.S.A. 89 Ž1992.
8948–8952.
w62x J. Ngai, A. Chess, M.M. Dowling, N. Necles, E.R. Macagno, R.
Axel, Coding of olfactory information: topography of odorant reception expression in the catfish olfactory epithelium, Cell 72 Ž1993.
667–680.
w63x A. Novak, N. Goyal, R.M. Gronostajski, Four conserved Cysteine
residues are required for the DNA binding activity of nuclear factor
I, J. Biol. Chem. 267 Ž1992. 12986–12990.
w64x S. Osada, S. Daimon, T. Nishihara, M. Imagawa, Identification of
DNA binding-site preferences for nuclear factor I-A, FEBS Lett. 390
Ž1996. 44–46.
w65x G. Paonessa, F. Gounari, R. Frank, R. Cortese, Purification of a
NF1-like DNA-binding protein from rat liver and cloning of the
corresponding cDNA, EMBO J. 7 Ž1988. 3115–3123.
w66x M. Puzianowska-Kuznicka, Y.-B. Shi, Nuclear factor I as a potential
regulator during postembryonic organ development, J. Biol. Chem.
271 Ž1996. 6273–6282.
w67x K. Raming, J. Krieger, J. Strotmann, I. Boekhoff, S. Kubick, C.
H. Baumeister et al.r Molecular Brain Research 72 (1999) 65–79
w68x
w69x
w70x
w71x
w72x
w73x
w74x
w75x
w76x
Baumstark, H. Breer, Cloning and expression of odorant receptors,
Nature 361 Ž1993. 353–356.
K.J. Ressler, S.L. Sullivan, L.B. Buck, Information coding in the
olfactory system: evidence for a stereotyped and highly organized
epitope map in the olfactory bulb, Cell 79 Ž1994. 1245–1255.
P. Rossi, G. Karsenty, A. Roberts, N.S. Roche, M.B. Sporn, B. de
Crombrugghe, A nuclear factor 1 binding site mediates the transcriptional activation of a type I collagen promoter by transformation
growth factor-b, Cell 52 Ž1988. 405–414.
R.J. Roy, S.L. Guerin,
The 30-kDa rat liver transcription factor
´
nuclear factor 1 binds the rat-growth hormone proximal silencer,
Eur. J. Biochem. 219 Ž1994. 799–806.
R.A.W. Rupp, U. Kruse, G. Multhaup, U. Gobel,
K. Beyreuther,
¨
A.E. Sippel, Chicken NFIrTGGCA proteins are encoded by at least
three independent genes: NFI-A, NFI-B and NFI-C with homologues
in mammalian genomes, Nucleic Acids Res. 18 Ž1990. 2607–2616.
C. Santoro, N. Mermod, P.C. Andrews, R. Tjian, A family of human
CCAAT-box-binding proteins active in transcription and DNA replication: cloning and expression of multiple cDNAs, Nature 334
Ž1988. 218–224.
G. Schweizer-Groyer, A. Groyer, F. Cadepond, T. Grange, E.-E.
Baulieu, R. Pictet, Expression from the tyrosine aminotransferase
promoter Žnt y350 to q1. is liver-specific and dependent on the
binding of both liver-enriched and ubiquitous trans-acting factors,
Nucleic Acids Res. 22 Ž1994. 1529–1583.
J. Schwob, E. Mieleszko, K.E. Szumowski, A. A Stasky, Olfactory
sensory neurons are trophically dependent on the olfactory bulb for
their prolonged survival, J. Neurosci. 12 Ž1992. 3896–3919.
G.M. Shepherd, Current issues in the molecular biology of olfaction,
Chem. Sens. 18 Ž1993. 191–198.
R.Y.L. Tsai, R.R. Reed, Cloning and functional characterization of
Roaz, a zinc finger protein that interacts with Olf-1rEBF to regulate
w77x
w78x
w79x
w80x
w81x
w82x
w83x
w84x
79
gene expression: implications for olfactory neuronal development, J.
Neurosci. 17 Ž1997. 4159–4169.
J. Verhagen, A.B. Oestreicher, M. Grillo, Y.-S. Khew-Goodall,
W.H. Gispen, F.L. Margolis, Neuroplasticity in the olfactory system:
differential effects of central and peripheral lesions of the primary
olfactory pathway on the expression of B50rGAP43 and the olfactory marker protein, J. Neurosci. Res. 26 Ž1990. 31–44.
M.M. Wang, R.Y.L. Tsai, K.A. Schrader, R.R. Reed, Genes encoding components of the olfactory signal transduction cascade contain
a DNA binding site that may direct neuronal expression, Mol. Cell.
Biol. 13 Ž1993. 5805–5813.
M.M. Wang, R.R. Reed, Molecular cloning of the olfactory neuronal
transcription factor Olf-1 by genetic selection in yeast, Nature 364
Ž1993. 121–126.
S.S. Wang, R.Y.L. Tsai, R.R. Reed, The characterization of the
Olf-1rEBF-like HLH transcription factor family: implications in
olfactory gene regulation and neuronal development, J. Neurosci. 17
Ž1997. 4149–4158.
S. Wenzelides, H. Altmann, W. Wendler, E.-L. Winnacker, CTF5
— a new transcriptional activator of the NFIrCTF family, Nucleic
Acids Res. 24 Ž1996. 2416–2421.
H. Xiao, J.T. Lis, H. Xiao, J. Greenblatt, J.D. Friesen, The upstream
activator CTFrNF1 and RNA polymerase II share a common element involved in transcriptional activation, Nucleic Acids Res. 22
Ž1994. 1966–1973.
X. Zhang, R. Miskimins, Binding at an NFI site is modulated by
cyclic AMP-dependent activation of myelin basic protein gene expression, J. Neurochem. 60 Ž1993. 2010–2017.
L. Zou, R.K. Menon, A member of the CTFrNF-1 transcription
factor family regulates murine growth hormone receptor gene promoter activity, Endocrinology 136 Ž1995. 5236–5239.